Political Animal

Blog

May 05, 2014 5:08 PM Polling in the Shadows

By Ed Kilgore

Many close observers (yours truly included) of the red-hot GA GOP SEN contest have raised eyebrows at a couple of recent polls from Insider Advantage that showed underfunded former Secretary of State Karen Handel surging into contention and then into a putative runoff spot. Sure, Handel had benefited from earned media generated by a leaked video of opponent (and consensus front-runner in the primary) David Perdue mocking her lack of a college education, and from endorsements by Sarah Palin and Erick Erickson (or more properly re-endorsements, since both backed her 2010 gubernatorial run). But was that enough to explain breakthrough gains against candidates with big financial advantages? The curiosity about the polls was reinforced by IA’s opacity about its methodology, and the fact that IA proprietor Matt Towery is himself a long-time GOP political mover-and-shaker.

Well, it seems the folks at Huffpollster were curious, too, so Mark Blumenthal and Ariel Edwards-Levy looked into it with some help from Matt Towery, Jr., who helped conduct the survey. Turns out IA supplemented its robocalling with selective Internet interviews, which is not an unusual practice these days, but is also a bit of a shot in the dark. Asked to comment on IA’s methodology, highly regarded Marquette pollster Charles Franklin had this to say:

When polls abandon probability sampling they lose the theory (and theorems) that prove samples can be generalized to populations. There is not yet an accepted theory for how to generalize from non-probability samples, including internet samples, though there are a number of interesting approaches being tested. Some of these rely on weighting by a variety of demographic information. Others rely on estimating relationships in the sample and then applying that model to a known population (usually from census data or voter lists.) And some have remarkably ad hoc approaches. Those based on explicit models can be replicated and tested in a variety of settings but how well they work is an empirical question. The more ad hoc the approach the more impossible it becomes to assess. In effect we have polls with no theoretical basis to claim legitimacy. Maybe they work. Maybe they don’t. We don’t know.

Maybe we’ll have a better idea on May 20, when we find out if IA’s polling in Georgia turned out to be accurate. In the meantime, Karen Handel gets to enjoy some buzz, and perhaps some late transfusions of money and hope as well.

Ed Kilgore is a contributing writer to the Washington Monthly. He is managing editor for The Democratic Strategist and a senior fellow at the Progressive Policy Institute. Find him on Twitter: @ed_kilgore.

Comments

(You may use HTML tags for style)

comments powered by Disqus