The Pollsters Got It Right

The mudslinging between the candidates and campaigns in this election was matched by the vitriol spewed over the polls. This one is oversampling Democrats, this one leans Right and this one is biased left.

We have heard it all over the last year. There was meant to be a landslide for Obama or a landslide for Romney in the popular vote. Romney was either going to get 350 electoral or under 200 – it had to be one or the other surely?

But in the sea of these outlier predictions by pundits from all across the political map, the pollsters managed to get it right. The overwhelmingly majority said that we were going to see a narrow popular vote win for the President and that he was going to win the crucial swing states giving him a healthy Electoral College win. As we all know now, it was, in fact, more than a healthy Electoral College win for Obama with Florida yet to be declared.

To be sure, there was debate over which way states like Ohio and Florida were going to go in the end. With polls showing a +1 or +2 edge for either of the candidates this is hardly surprising and, indeed, a necessary discussion.

In the end, though, the pollsters were able to get the right sample for who showed up on the polls on Tuesday. Young voters turned out at the same level, if not higher, than in 2008. And all those critics that said that were allegedly oversampling Democrats with a +7 or +8 spread were vindicated, at least in part. Exit polls show that Democrats were +6 on Tuesday.

Those responsible for the narrow analyses of the electorate like those used by Dean Chambers at his “Unskewed Polling” site were forced to admit fault in assuming that the electorate would lean Republican. Chambers said in a phone interview, “Nate Silver was right and I was wrong.” And Chambers is not the only one who has had to make that declaration over the past two days.

The key to great polling is to have the best voter turnout model. The Obama campaign had the best model out there, using a combination of the 2010 census and a 1.5 year long ground game in battleground states.

Obama’s team certainly benefited by not having to face a long, grueling primary season, but that doesn’t take away from the strength of their model. The model that the Republicans used paled in comparison, plain and simple.

Doubtless, pollsters everywhere are feeling the same thing that YouGov’s Joe Twyman expressed to me earlier today, “Whatever your feelings about the results, this is great news for pollsters.”