Tuesday, February 17, 2004
Fun with exit polls
Edwards gets 36% of both men's and women's votes. Kerry gets 36% from men, but jumps to 41% among women. The pattern seems to be determined mostly by race, since Edwards polls two points higher among white women than among white men.
Kerry's support gets slightly stronger as the age of voters increases. Dean is strongest among the youngest voters, Edwards is weakest there.
Dean's support is pretty constant across racial lines. Kerry loses to Edwards by a point among whites, but beats Edwards by 39 points among African Americans and by 22 points among Latinos. (demographics: whites: 89%, african american: 6%, latino: 3%)
Kerry beats Edwards by 10 points among voters who earn less than $50,000, Edwards wins by two points among those who earn more.
Support for both Edwards and Dean increases with education level - though it's much more pronounced for Dean. Kerry does better among those with less education.
Kerry beats Edwards by six points in union households and by four points in veteran households.
Kerry beats Edwards by fifteen points among Democrats. Edwards carries the Republican and independent vote.
Kerry and Edwards both poll at 40% among moderates. Kerry wins the liberals, Edwards the conservatives.
Once again, Kerry wins the atheist vote, but this time by only three points. Catholics and Protestants vote for Kerry and Edwards in about equal numbers, but Kerry wins 'other Christians' by a whopping 15 points.
Edwards beats Kerry by 16 points among those who made up their minds in the last three days. Kerry wins by 29 points among those whose minds have been made up for awhile.
Kerry again wins big on 'can beat Bush.' Edwards beats Kerry 2 to 1 among those who are satisfied with the Bush administration, and beats Kerry 3 to 1 among those who are enthusiastic about the Bush administration.
Kerry and Edwards poll evenly among those who oppose gay marriage, whether or not this opposition is cashed out in terms of support for Civil Unions. Kerry beats Edwards by six points among those who favor gay marriage. Kucinich enjoys a six point bounce among those who favor gay marriage.
Edwards beats Kerry by 10 points among the 35% who support the war in Iraq. Kerry beats Edwards by 10 points among the 65% who oppose the war.
1% of respondents characterize the economy as excellent. Among those who think the economy is good, Edwards beats Kerry by 14 points. Kerry wins by three points among those who think the economy is not good and by 15 points among those who think the economy is poor.
Edwards beats Kerry by 2 to 1 among those who think the Bush tax cuts should not be repealed. They run about even among those who want to repeal tax cuts for the rich. Kerry wins among those who want all Bush tax cuts repealed.
Some people are trying to spin this as a Kerry loss since Dean+Edwards beats Kerry. Although Kerry does seem to be losing momentum, this strikes me as wishful thinking.
The idea that there's an anti-Kerry majority depends on the notion that Dean supporters will flock to Edwards if Dean drops out. I just don't see it.
To begin with, Wisconsin's open primary format makes the numbers look closer than they are. Edwards does well among constituencies that aren't likely to vote Democratic in November--war supporters, Republicans, and so on. Kerry beats Edwards soundly among women and minorities.
Other things being equal, conservatives prefer Edwards and liberals prefer Kerry. Dean's support is pretty evenly spread among conservatives and liberals, so you'd expect the conservatives to break for Edwards and the liberals to break for Kerry.
Things aren't equal, of course, because a sizeable number of Dean's supporters hate Kerry. So Dean's supporters will break for Edwards.
But they won't all go to Edwards, and they won't go 5 to 1 or even 3 to 1. Those are the kinds of numbers you need for an Edwards majority.