ELECTION ANALYSIS 101. There is an aphorism, which I think comes from Alexander Herzen but I've seen it attributed to Mark Twain, which is something like, "There's nothing harder than trying to get someone to understand something when he's being paid to not understand it." And I'm beginning to think that aphorism applies to the group Third Way. I'm going to let Tom Schaller own the rebuttal of Third Way's latest report "proving" that affluent, white voters won the election for Democrats in 2006 and thus Democrats have to pay attention to them and throw benefits at them, but I want to highlight a slightly different aspect of it.

Tom points out one basic and fatal error, which is that Third Way compared the 2004 (presidential-year) electorate with the 2006 (congressional) electorate. The presidential year electorate is not only much larger (which Third Way corrects for), but because it consists of infrequent voters, it is by its very nature blacker and poorer. Thus the electorate in an off-year election is always going to be whiter and richer than in the presidential election that preceded it, and any Democratic gains will inevitably appear to be gains of white and affluent voters. To avoid that mistake, the appropriate baseline would be 2002.

But there is another point to make: 2006, while in part a nationalized election, was not a national election. Some areas had contested congressional elections and some did not. You might have heard that many of the most closely contested congressional races were in suburban districts -- districts that happen to have a lot of white and affluent voters. A hot congressional race increases turnout. So -- surprise! -- you have a lot of hot races in white, affluent areas, you're going to get a lot more white, affluent voters.

So even if Third Way had done the proper comparison with 2002 (as they do, briefly, in the conclusion), it would be misleading because the contested races were in different places. Just as in the future, the contests will be in other places, and the results of congressional races heavily weighted toward the suburban don't show how to win presidential or statewide elections.

It would take someone with more letters at the end of their name and more time than I have to do a comprehensive analysis of comparative turnout, but after the jump, I'll show some examples of how hot races drove up suburban voter participation: