With the McCain-Palin ticket retaining its convention bounce so far, the modest GOP leads in some of the surveys have, in my view, led some Republican partisans and conservative commentators to get carried away with irrational exuberance as well as some (if not many) Democratic partisans and liberal commentators to unwisely dismiss the new polls on the basis of the new-voter hypothesis. That’s the notion that pollsters’ likely voter screens do not adequately reflect the impact of Democratic voter-registration drives, particularly among young voters. Some pollsters use past-voting histories, either self-reported or from election records, which wouldn’t capture them. Therefore, there must be a systemic bias against Obama that masks his real lead among the electorate who will turn out on Election Day.
There’s enough plausibility to this critique not to dismiss it out-of-hand, but it’s also true that the major polling firms are very well aware of the issue, have taken steps to try to minimize or remedy any bias (through weighting by demographics and measures to capture cell phones), and have strong competitive reasons to keep working on it.
The point worth emphasizing here is that both sides are letting wishful thinking supersede the data. They argue that historical patterns should be ignored, because of social and political changes that make this election cycle unique. Some Republicans aren’t taking seriously the sheer unlikeliness of a party in power retaining power given the negative trends in economic performance and public sentiment. Some Democrats aren’t taking seriously the notion that pollsters are smart, constantly learning, and have usually pegged things more-or-less correctly in the past — and that when many separate surveys capture a similar trend (even if their absolute numbers aren’t comparable), the trend is probably real.
Just to be clear: I have no idea what’s going to happen in November. No one does. But as some conservatives found to their dismay two years ago: if you find yourself denying the evidence of your senses, in the belief that polling and other political information just can’t be reflecting reality, you are probably in denial of reality.






Comments