Everyone in the political world saw all the polls over the last several days, and they all pointed to a big Barack Obama win in New Hampshire. Given Hillary Clinton’s narrow victory, this has prompted many to suggest polls shouldn’t be taken seriously anymore. I think that’s a little rash under the circumstances.
Josh Marshall reminded us overnight that, by and large, polls are usually right.
[B]y and large they have a very good record. It’s silly to think that we — whether ‘we’ is reporters or political junkies or ordinary voters — are going to ignore the information that’s right in front of us. And why should we?
It’s true I guess that in an abstracted reality we could simply listen to the candidates, ignore all probabilistic data available, go to the polls with no idea of the result and learn the outcome the following morning. But that’s not the world we live in nor do I think it’s one I’d want to live in.
Agreed. Pollsters put surveys in the field, and tell us the results. The numbers offer us hints of what’s to come. When all the polls agree on a likely outcome, far more often than not, that’s what’s going to happen. Yesterday was obviously the exception, but that doesn’t necessarily mean it’s time to trash the rule.
On a related note, Matt Yglesias offers an interesting observation (and accompanying chart):
Commenter Brian makes an observation “No one is talking about how the polls actually nailed Obama’s number. Obama didn’t lose this election. He stayed steady and Hillary surged ahead.” That seems to be true.
It does, indeed. As of late Monday, averaging the various polls out of New Hampshire, we saw these averages: Obama 36%, Clinton 30%, Edwards 18%
And with more than 98% of precincts reporting: Clinton 39%, Obama 36%, Edwards 17%.
The polls got two of the three just right. It’s that winner they missed.
Is there a reasonable explanation? This might offer some clues:
Among voters who made up their minds yesterday, Hillary beat Obama by 39%-36%, suggesting that the last news cycle of the contest might have played a decisive role in shifting votes Hillary’s way. Yes, I admit it — I’m talking mainly about the wall-to-wall coverage of The Tears, which were effectively the last close look at Hillary these voters had before entering the voting booths.
Meanwhile, among voters who made up their minds in the last three days, Hillary also gained ground, coming in only one point behind Obama, 36%-37% — numbers that are way out of sync with the big lead Obama had in polls over the weekend and yesterday.
Since Hillary had a relatively solid lead of 41%-37% over Obama among voters who had already made up their minds before last weekend, the late-deciding voters enabled her to regain enough ground to win.
Bottom line: polls are not, all of a sudden, useless. Yesterday was a rare occurrence, but data still has value.