Is Colorado Republican Bob Beauprez really up 10 points? Probably not.
Well, you can have one of two reactions to Quinnipiac's new polls of
Colorado and
Iowa that the university released on Wednesday. One is to panic and assume the sky is falling on Democrats. The other response is to say, "Hmm. These polls don't look like any others we've seen lately. That means they're probably off-base."
The latter option is, of course, the correct one, so let's see what we've got here, starting with Colorado. There, Quinnipiac finds Republican Bob Beauprez implausibly surging out to a 50-40 lead on Democratic Gov. John Hickenlooper, who's generally held small leads. In fact, Beauprez's only ever led in two polls, and never by more than a single point. One of those was from Rasmussen and the other from ... Quinnipiac, back in July, when they had Beauprez ahead 44-43. Since then, they've switched to a likely voter model, and it seems that their screen must simply be bonkers.
And as we've noted before, the school has had trouble polling Colorado in the past, often finding much more positive numbers for Republicans than other outfits have. In Aug. 2012, they found Mitt Romney up 5 points, the most optimistic poll he ever saw there, and their final poll still had him on top by 1, even though he lost by more than 5. Likewise, in the summer of 2008, they also had John McCain leading at a time when pretty much no one else did.
And if Beauprez had internal numbers that were nearly this good, don't you think he'd have released them by now? He hasn't—but Democrats were ready with their own polling to counter Quinnipiac. Project New America released data from Myers Research showing Hickenlooper with a 51-44 advantage. That's actually the biggest lead Hick's seen in a while, but if you think those numbers are too gaudy, then that's a good reason to reject Quinnipiac's as well.
Iowa also requires the same level of skepticism. There, Quinnipiac has Republican Joni Ernst up 50-44 on Democrat Bruce Braley, the largest lead she's seen since just after she won the GOP primary. But again, Quinnipiac's switch from registered voters to likely voters has caused a dramatic shift, as Braley was up 4 in their last poll in June. Most recent polling has put Braley up a touch, and perhaps more importantly, Republicans are privately confessing that Ernst has slipped behind. (Politico's headline: "GOP frets over Harkin seat.")
And to be crystal clear, we're urging caution when interpreting these results not because they're bad for Democrats and we want people to plug up their ears and clap louder, but because they don't make sense in light of all the other available evidence. We'd be insisting on the same thing if Quinnipiac had suddenly found Pat Quinn or Mary Landrieu up 10. The point, as always, is that you can never rely on a single data point to draw conclusions, and Quinnipiac just gave us two good lessons as to why.