Despite all the polls favorable to Obama recently, I’m determined to prepare myself for the worst: I don’t really trust the polls this year, and there are still three weeks until election day, and you never know what kind of voter suppression or voting machine manipulation Republicans might accomplish. With those credentials as a determined pessimist established, I have to say that this post at fiverthirtyeight.com heartens me more than any poll numbers I’ve seen:
SurveyUSA has a lot of good habits as a pollster, and one of them is breaking out the results of early and absentee voting in states where such things are allowed. So far, SurveyUSA has conducted polling in five states where some form of early voting was underway. In each one, Barack Obama is doing profoundly better among early voters than among the state’s electorate as a whole:... Poll % Voted Non-Early State Date Early Early Voters Likely Voters ==================================================== NM 10/13 10% Obama +23% Obama +6% OH 10/13 12% Obama +18% Obama +4% GA 10/12 18% Obama +6% McCain +11% IA 10/9 14% Obama +34% Obama +10% NC 10/6 5% Obama +34% McCain +5%
We should caveat that these are not hard-and-fast numbers. Estimates of early voting results are subject to the same statistical vagaries as any other sort of subgroup analysis, such as response bias and small sample sizes.
Nevertheless, Obama is leading by an average of 23 points among early voters in these five states, states which went to George W. Bush by an average of 6.5 points in 2004.
Is this a typical pattern for a Democrat? Actually, it’s not. According to a study by Kate Kenski at the University of Arizona, early voters leaned Republican in both 2000 and 2004; with Bush earning 62.2 percent of their votes against Al Gore, and 60.4 percent against John Kerry. In the past, early voters have also tended to be older than the voting population as a whole and more male than the population as a whole, factors which would seem to cut against Obama or most other Democrats.
There’s a bit more analysis at the post. In addition to the sheer size of those numbers, I’m also heartened because I suspect that people are more likely to mislead pollsters about things they plan to do in the future (it doesn’t feel like a lie if you’re speaking hypothetically about a future action) than about things they have done in the past, so fears of the so-called “Bradley effect” (the supposed overestimation of support for black candidates in pre-election polls) might be unwarranted.
I’m still not going to count any chickens until the election is behind us, but there seems to be a very real possibility that the turnout models the pollsters have been using will be demolished by this election, and that enthusiasm for Obama among liberals, young people, blacks, and so on, combined with what has the potential to be a huge get-out-the-vote effort on Obama’s behalf, will create an Obama landslide.
I know that early voters aren’t representative of voters as a whole, and I know that we’re dealing with very small sample sizes here, but still, it’s getting harder and harder to find bad news for Obama these days. A 700-point drop in the Dow, while bad for my net worth, is presumably also good for Obama’s chances on the day of the final debate — at the very least, it gives him an easy comeback if McCain tries any character-based attacks. (“This is the kind of politics that the American people are tired of, John — on a day when they saw the value of their 401ks drop by over 5%, you keep trying to change the subject to a guy who did some despicable things when I was 8 years old…”)
Will someone please throw some cold water on these numbers, so that I can return to my moderately pessimistic equilibrium?