President Obama is leading by a bit in many of the national and state polls out over the past week. But I think if you see Obama up by a point or two in a poll, you can probably assume Gov. Mitt Romney will actually come out on top on Election Day. And if Obama is up 3 or maybe four points, it’s still likely just about a dead heat.
- I think some of the polls are understating the number of Republicans in the electorate, sometimes rather dramatically so. This is not going to be the 2008 race, and some polls are assuming similar turnout by Party identification. This is where the bias of individual pollsters can creep in.
- Pollsters are not successfully measuring the emotion out there among Republicans for electing Romney and deposing Obama. Republicans will be going to the polls in droves.
- Countering this somewhat is the Obama ground game, which is better organized than Romney’s. But Romney has a lot of people out in the field too, and I don’t think knocking on doors is going to overcome a surge of GOP passion.
- I think that some people don’t tell pollsters the truth because they don’t want them to think poorly of them for opposing Obama. There are several reasons for this, including that they don’t want to be perceived as racist and they don’t want to admit they support the less fashionable candidate.
- In 2008, pollsters overestimated Obama’s strength on average by about 1.4 percentage points. And this was during a race in which Obama had the vastly superior ground game and the passion was all on his side. Rasmussen, who today has the race tied both nationally and in Ohio and Wisconsin, did the most accurate final poll.
I don’t think there is a conspiracy among pollsters to support Obama, though I do think bias creeps in. And I don’t think it’s a good idea to ignore polling as completely unreliable, as some suggest.
But I do think Obama’s support is being overestimated, and that if Romney is very close, he wins.