David Brooks' piece in The New York Times today on "Bush Democrats" just isn't terribly convincing about the alleged political salience of this phenomenom.

Consider the kind of evidence Brooks brings to bear--chiefly about splits in the Democratic ranks on Iraq-related poll questions and unity among Republicans. But that kind of result depends on which Iraq questions you look at.

For example, CBS News recently found that 74 percent of Democrats thought the result of the war with Iraq was not worth the loss of American life and other costs of attacking Iraq, compared to just 14 percent who thought it was. Pretty unified. And 68 percent of Republicans thought the Iraq war's result was worth cost, compared to 21 percent who didn't. Also pretty unified.

This kind of polarization is actually more common than not on polling questions these days--as many analysts have commented--and Brooks appears to have wilfully overlooked it in order to make his point.

Heck, you can even find examples that are the complete reverse of what Brooks points to. In the last ABC News poll, Democrats overwhelmingly believe (79 percent to 19 percent) that, given the goals and costs of the Iraq war, the level of military casualties has been unacceptable. But the Republicans, they're split! While 54 percent think the level of casualties has been acceptable, a healthy 40 percent think it has not.

So should we start talking about Dean (or Clark) Republicans?

DR doesn't think so, but it does suggest the problems with Brooks' logic and evidence.

Brooks also mentions that 20 percent of Democrats say they'll vote for Bush in a hypothetical Bush-Dean matchup, while Republicans are much more unified around their prospective nominee. But that's hardly suprising, given that Republicans know exactly who their nominee will be, while Democrats do not--in fact, many of them have not yet really focused on the upcoming presidential contest. In these circumstances, 20 percent support from Democrats is not all that impressive--especially given the unrealistically large lead Bush has in this particular horse race question (20 points). The race will wind up being a lot tighter and Bush's Democratic support will fall commensurately.

Finally, in the same poll cited by Brooks (the CBS News poll linked to above), just 11 percent of Democrats say they'd vote for Bush against "the Democratic candidate". That's probably a more reasonable estimate of Bush's current Democratic support. And how much support did Bush actually get from Democrats in 2000? You guessed it: 11 percent.

The more things change.....