Dead even, for all practical purposes. The Real Clear Politics average gives Obama a lead of 2.3 points, nationally. That figure, which is of dubious value except that it is frequently reported in the national news, is based on six polls, four of which show a small Obama advantage, one a small McCain advantage, and one a tie. But the polls that favor the Democrat are mostly polls of registered voters, and one has a very small sample size. The polls more favorable to McCain show likely voters, which is a better indicator of outcomes. The CBS poll, giving Obama a five point advantage, has a reasonable sample size of about 1100 voters, but it includes 404 Democrats v. 305 Republicans. That is probably based on recent voter ID stats, but the Republican brand has recovered substantially since the convention, and that kind of weighting may not be justified anymore. All this suggests that the race is pretty much tied nationally.
Of course the President is not elected by national vote, but by the several states through their Electoral College votes. Polls in the battle ground states are the only way to get a direct read on this. The RCP gives Obama/Biden 219 votes, McCain/Palin 189, with 130 toss up states. It is pretty clear that, if Obama wins all the states where he has any lead at all, he wins more than the 270 he needs to be elected. How do we game this?
There are really only two ways. The national polls, despite their legal irrelevance, have usually been the best indicator of the outcome. A candidate who is ahead by more than five percentage points in most polls will probably win all the states where the local polls are close. But here, obviously, the national polls are too close to give an indication. The other way is to look at how the battleground states have gone in recent elections. According to that kind of analysis, you'd expect Ohio and Florida to go Republican, Pennsylvania and New Mexico to go the other way. But the map is rather confused on that score. Florida, Ohio, Indiana, and Virginia are in the toss-up category. That bodes ill for McCain. But if the election remains close nationally, I expect that McCain will win all of these. He had better, if he wants to win. On the other hand, Minnesota and Wisconsin are also toss-up states, and Obama should be well ahead in both places.
So right now it is hard to get a good read by any of the traditional methods. If it's another squeaker, I won't be getting much sleep on election night.
But there is one more consideration that is probably disturbing the sleep of a lot of Obama's court soothsayers right now. It's the Bradley Effect. In 1982, Tom Bradley, an African American candidate for Governor of California, was well ahead of George Deukmejian in the polls. But Deukmejian won the election. Pollsters offered this explanation for their failure: a lot of voters said they were voting for Bradley but instead voted for Deukmejian. Why would an unusual number of respondents lie to pollsters?
It is frequently said that the motive was racism. Respondents didn't want to admit that they were voting against Bradley just because he was Black, but they knew in their hearts that they were going to do just that. But that confuses two questions: why they lied when polled, if that is what they did, and why they voted the way they did. It may well be that a lot of California voters felt that they should say they were voting for Bradley. That would reflect a consciousness of the racial issue that is problematic, but nothing to be ashamed of. But the Bradley Effect tells us nothing about what their motives were for voting as they did, once they were alone in the booth. No doubt race was a factor, but there is no way to tell how much of a factor it was.
So what if the Bradley Effect is working now. A lot of voters nationally are telling the voice on the phone that they are voting Obama, but they really won't? Some analysts believe that Obama can't win unless he is a good five points or more ahead in the polls. That would give him a cushion to make up for the Bradley Effect voters. If that is right, then Obama needs to gain some ground fast.
If Obama loses, Democrats will cry far and wide that it was racism that defeated him. That will make good political sense, as it will undercut the legitimacy of McCain's election. But it will be nonsense. Racism will no doubt be a factor in the election. But so will doubts about McCain's age, contempt for Governor Palin's Christianity, women who like Palin because she is one of them, etc.
I have no idea whose ahead at this moment, and that is a novel experience for me at this stage in an election. Isn't this exciting?
Recent Comments