Monday, November 05, 2012


One thing I can predict with confidence is that Republicans are going to respond to this year's election coverage by creating lots and lots of polling firms in time for 2014 and 2016.

Republicans know that Nate Silver and other poll nerds repeatedly threw cold water on their pro-Romney narratives by citing poll averages; if this is the future of campaign coverage, Republicans will feel the need to mess with those poll averages by generating far more raw polling data than they do now. These GOP firms aren't going to generate unrealistically pro-GOP polls -- their numbers, like Rasmussen's, will look legit, if a tad skewed. Unlike Scott Rasmussen, these new pollsters won't write Republican-friendly op-eds, or have other blatant ties to the GOP; some of the pollsters may even be (nominal) Democrats. But their job will be to create a pro-GOP reality in Nerdville.

Right-wing news sources and pundits will point to poll aggregators who (unlike Nate Silver) average all poll numbers without assessing "house effects" (i.e., pro-Democratic or pro-Republican skews). Those aggregators will be said to be the honest, legitimate, unbiased ones, unlike Silver and his biased liberal soul mates. (Mainstream journalists may well fall for this.)

I think Silver and his fellow quants will be vindicated tomorrow, but this may be their last hurrah. The numbers they work with are going to be hopelessly politicized in the future.


Victor said...

"Mainstream journalists may well fall for this."

Fish, meet lure.

And Mr. Silver will be smart enough NOT to include these (white) johnny-come-latelies when aggragating his polls next time around.
He, and Mr. Wang, can take a snapshot of the existing legitimate polling outfits, and disregard the new wanna-be's, until they prove whether they're legit or not.

The guy who comes up in the last two weeks of the season, and hits
.638, shouldn't be eligible to win the batting title.

Anonymous said...

I just don't understand why a person would choose to be wrong. Sticking your head in the sand, and telling yourself everything is going to be okay -- sure, I get that impulse. And it does feel good when the polls tell you what you want to hear.

But skewing (excuse me, "unskewing") the data so it says what you want, even though it's wrong? Announcing to the world predictions that are obviously, empirically going to be spectacularly disproved? How does that help?

Nothstine said...

Sounds perfectly plausible to me. Of course it will cynically further cripple an already-crippled system, but that hasn't been an issue since Karl Rove met Junior. Or since Newt Gingrich first published "Language: A Key Mechanism of Control."

Jack said...

Those aggregators will be said to be the honest, legitimate, unbiased ones, unlike Silver and his biased liberal soul mates. (Mainstream journalists may well fall for this.)

Yes, they will. Especially because the GOP will continue to expand and refine its voter suppression efforts, with the result that their GOP-skewed polls will actually reflect the outcomes more accurately than real polls.

The one limitation of Silver's model is it doesn't take into account the tens of thousands, or hundreds of thousands, of votes that will be lost because of voter suppression or voter intimidation efforts. Can the GOP use these methods to shave 2-3% off Democrats' numbers? We'll find out tomorrow, I guess.

I was just thinking a couple of days ago that Americans are going to have to be taught that actual majorities don't matter and what polls show about those majorities must be ignored, because increasingly, in the years ahead, there will be a growing disparity between the two. But you have hit on an even "better" solution for the GOP: They'll just replace real polling with fake polling in the same way they replaced real news with fake news -- and then declared the real news to be the actual fake news, and their fake news to be the "real" news.

The media fell for it when it came to the media; they'll fall for it with polling.

Unknown said...

Steve, I completely agree with you about the first part, but find strong counsel with Victor as to the second part. Nate Silver is an unapologetic intellectual, which is the province of the "Left" in America, and his voice will be heard through effective channels. Trust me, he has a methodology for weighing the "data" from polling services who are not sufficiently forthcoming with respect to their sampling methods.

But the key point of your post--the imminent proliferation of right-wing polling corporations--is worth mentioning. Manufacturing a set of expectations is an critical part of the advertising/PR world of Republican politics. Get the notion "out there" that the Republican is "supposed" to win, and you have some advantage when it comes to the ballot counting shenanigans that Charles Pierce frequently talks about.

ZachPruckowski said...

Are there already accusations that some of the newer polling firms (Gravis, Purple Strategies) are already an implementation of this strategy?

Ambignostic - Many Republican strategists believe that appearing to be slightly ahead (or claiming to be slightly ahead) is a self-fulfilling prophecy, because it'll focus your restless base, help with fundraising, alter media coverage, and help persuade voters, all while demoralizing your opponents' volunteers/donors. Some of them even blame Nate Silver for Obama's win - after all, but for Silver's constant (and probably accurate) claim that Obama was still a slight favorite, Romney's "momentum" would have demoralized Democrats and powered Romney to a 10-point win.