Man, if the GOP nominates "Generic Republican,"
Barack Obama is in some serious trouble.
For most of the 2012 cycle to date, we have heard little from the undisputed champion of prolific right-wing meme setting via polling: the House of Rasmussen. About the only news that they had made this year was throwing a lot of their data behind a paywall.
But last week, after a tiptoe into the water in October to cover the Wisconsin Senate race, Rasmussen launched into their work for the 2012 cycle. But in doing so, they made, as the often do, a rather bizarre decision.
In four separate states (Ohio, Pennsylvania, Florida, and California), they trumpeted results which claimed Obama was trailing in the three swing states east of the Mississippi, and only up four points (45-41) in California.
The problem? The horse race data from these four states was based on a matchup between Obama and ... wait for it ... a generic Republican opponent. The "generic" candidate construction (or, as our pal David Nir puts it, "Republican Jesus") is a device that actually has merit in very certain situations. When a party does not have a name-brand candidate at the ready, or the field a party has to offer is relatively unknown, it makes more sense to test a known quantity against a generic foe from the other political party.
But that's not what we are talking about here. Given the national profile of a presidential race, to say nothing of the 238 or so nationally televised debates the nation has already been subjected to, it is safe to say that voters in the four states mentioned above have heard of people like Herman Cain, Mitt Romney, and Newt Gingrich.
Rasmussen knows that too, of course. Which is why they did conduct a named ballot test last week, as well. They polled the GOP leaders versus Obama in Missouri, finding Mitt Romney ahead of the president by a relatively modest three percent margin, while Obama led all other GOP comers.
The curiosity, then, is why the House of Ras saw fit to do normal polling in reddish-purple Missouri, but not in Florida or Ohio?
The most plausible answer lies in the polling evidence we have seen elsewhere--"Generic Republican" performs far better against Barack Obama then Actual Republican, be it in the form of Mitt Romney or one of the also-rans. The GWU/Politico/Battleground poll found the president deadlocked with "Generic Republican" at 43 percent. But the same sample preferred Obama to Romney by six points (49-43) and Obama to Cain by nine points (49-40).
Thus, relying on generic tests, except in the reddest state they tested last week, looks like an attempt to game the data to put the president in the most pessimistic possible light. And, assuming they haven't changed their ways from last cycle, their polling already has a pretty marked GOP lean to begin with.
As they are so often wont to do in their issue polling, an episode like this makes it hard not to conclude that Rasmussen is using their campaign polling as a conservative meme-setting operation now. Which, given the tendency of the traditional press to parrot their results without skepticism, is somewhat alarming.