Opinions, Context & Ideas from the TPM Editors TPM Editor's Blog
Back on Thursday I expressed my own uncertainty about just what to make of all this. On the one hand, the story seems so universal and there seems so little reason at this point to keep up the pretense, unless there's some reason I'm not aware of to keep pumping up retroactive erstaz Romney momentum. So the unanimity of the story gives it increasing credibility. On the other hand the scale of the goof just seems too much to believe and the Romney camp's own actions -- the desperate move into Pennsylvania, for instance -- don't really square with this purported confidence.
Now though we have yet another article in Politico reporting that not only did Romney's campaign have polls that were substantially off but that this was the case pretty much across all Republican campaigns.
I think at this point, I give up. I give in. I can't maintain my skepticism. I don't know quite how shell shocked Mitt and his crew were. But I don't see how not to believe that Republicans as a group were not working on the basis of internal polls that were just totally wrong. Cui Bono? Why lie this much? I simply don't see any purpose being served. I've heard many suggest that they need to keep up this pretense because these are the bogus numbers they were serving up to Adelson and the other multi-billionaires to keep the money coming in. But this doesn't really make sense. Are these guys going to react worse because they were lied to than they are if they think the guys they gave their money to are a bunch of incompetents and morons?
That just doesn't figure to me.
Many of the articles I've seen over the last few days has presented this as a divergence between the polls Republicans and Democrats were using. They each had very different takes on the composition of the electorate and the Democrats won big.
But that understates what happened here. This wasn't Democratic and Republican pollsters. It was everyone but Republican pollsters. As the folks behind PollTracker, Kyle Leighton, Tom Kludt and I lived in the polls for months. And the truth is that with the exception of Rasmussen and Gallup toward the end of the cycle, basically every public pollster had results broadly in line with the final result. That's why we at PollTracker and other poll aggregators like 538, Pollster, RCP etc were able to predict the election almost exactly right. (PollTracker got 49 out of 50 states right. Florida we missed by about 1% point.)
So there's no access to special information about voter turnout strategies that Democrats were using. Everybody -- and I mean EVERYBODY -- basically had this right except for Republican pollsters. That makes it more than just a bad call. Because what are the odds that every other public pollster, presumably indifferent to the partisan advantage of the result, did not make this same mistake?
So what happened? I'm still kind of dumbfounded by it. Wishful thinking must play a massive role. But one thing I have started thinking about more and more is the way you can read back into the election cycle and see certain talking points that popped up in the mainstream media that almost certainly came from spinning Republican pollsters and the campaigns that employed them.
There's one that I actually joked about a number of times as it became more prominent in the Fall -- the so-called "extremely likely voter."
Not surprisingly this novel sub-group of people, who were presumably about to go postal they were so eager to vote, juiced up Republican numbers. But frankly, while I've heard of likely and even very likely voters I don't think before this cycle, actually late in this cycle I'd ever heard of "extremely likely voters" as a voting segment anyone was following. As I think Democrats were pointing all along, 'extreme likelihood' to vote is perhaps a bit excessive. Emotional instability or obsessiveness doesn't make your vote count any more. So probably being 'likely' is enough.
It also seems very clear that the Unskewed Poll Movement originated among GOP pollsters since they seem to have been the ur-unskewers.
All I can wonder at this point is whether there wasn't some on-going self-spinning in GOP polling circles that took place over the last three or four months as Obama's narrow but persistent lead just appeared impossible to shake. Some growing effort to define away Obama's lead.
For now, that's really the only speculation I have since nothing else makes sense. But I do find it genuinely disturbing that this cocooning was apparently so profound that none of these folks seemed to be troubled by the fact that no other public polling organization but Rasmussen was generating numbers anything like theirs.