You'd never know it given the timbre of political news coverage these days, but Democratic incumbents in the Senate are actually doing pretty well in poll averages, even in tough states like Arkansas.
That won't necessarily hold, and polling this far out probably doesn't tell us anything terribly useful about state-wide elections anyway, particularly in states with ongoing primaries. Other caveats apply as well. But if it does hold, conservatives are prepared to respond with
equanimity and a recalibrated political strategy another months-long poll unskewing temper tantrum.
I assume this based on the right's reaction to a New York Times/Kaiser Family Foundation poll that shows Democrats holding their own in swing states, and enjoying a commanding lead in Arkansas specifically.
Let me start by noting that absent a scandal or some other exogenous event, I'd be really surprised if Mark Pryor ends up defeating Tom Cotton by 10 points. But conservatives rushed straight to the internals to discredit the poll in almost exactly the same way they unskewed the 2012 polls to show Romney doing much better than he actually was.
— Logan Dobson (@LoganDobson) April 23, 2014
The obvious error here is an apples-oranges comparison between Romney's recorded share of the vote total with this after-the-fact, reported share of the voting-age population. In 2012, just over 30 percent of registered voters in Arkansas and over half of the voting age population didn't vote in Arkansas. Since the question was asked of all adults, it appears many people who didn't vote are now actually claiming to have voted for one of the candidates. And many adults, whether they voted or not, are claiming to have voted third party when they actually didn't. Eight percent of those surveyed say they voted for someone other than Obama or Romney. In reality third party candidates mustered a combined 2.5 percent of the vote (and a much smaller percentage of the voting age population) in Arkansas that year.
And as the Times' Nate Cohn notes in a strong defense of the poll, "there’s a well-known bias toward the victor in post-election surveys. Respondents who voted for the loser often say that they don’t remember whom they supported, or say they supported someone else."
If you're looking for cautious assessment of the poll, you might say there's some weirdness here, which cuts both directions, and polling averages don't show Pryor with such a significant lead. But conservatives went straight for the sample bias critique.
There's no denying that Democrats have a bunch of tough races to contest, but if in the coming weeks they perform better than expected by purveyors of conventional wisdom, it'll be fun to watch conservatives reckon with it.