TODAY’S LESSON: THERE IS NO LESSON….I finally got around to looking at the overall 2006 exit poll data in detail, and it’s pretty interesting. Anyone planning to make any grand pronouncements about the “lesson” of the election really ought to spend a few minutes comparing the 2006 exit poll data to the 2004 exit poll data first. It turns out that the big lesson is that there’s no big lesson.

Here’s the baseline: the overall Democratic share of the congressional vote was about 5 percentage points higher than in 2004. And what you find from the exit polls is that Dems gained 2-7 points in practically every demographic group surveyed. It was an across-the-board sweep, not a victory that depended on any single big electoral shift.

So were there any big changes? Compared to the overall 5-point gain, did Dems get a bigger share of the white evangelical vote? No. Women? No. Young people? No. Low-income voters? No. Self-described conservatives? No. Suburban voters? No. The South? No. The Northeast? No. Any region? No. Dems gained a steady 2-7 points in all these groups.

In fact, I was only able to find a grand total of seven groups that broke for the Dems by substantially more than the overall gain of 5 points. Here they are:



No high school


Those rating the economy “good”






No religion


Income $200K+




Now, there are some stories here. Democrats obviously appealed to the middle a little better than in 2004. Republican pandering to the Christian right seems to have energized Jews and seculars to vote for Democrats. Latinos were pretty obviously turned off by the Republican hard line on immigration.

More interesting (though less important in raw numbers) is the fact that high-income voters broke for Democrats in large numbers. I’m scratching my chin over that one. And not only did the economy not help Bush, but apparently it actively hurt him. Those who rated the economy “good” voted much more strongly for Democrats than they did in 2004. (Those who rated it either excellent, not good, or poor voted about the same as last time.)

In the long run, I suppose the higher totals among Latinos and independents are the big news. Beyond that, there’s not much. Keep this in mind when you start reading anecdotal analyses of “what happened.” Most of it doesn’t hold water. Based on the exit poll data, it was just a broad-based wave of disgust against Republican rule.

UPDATE: A bit more discussion and some further explanation of what I meant is here.