“Political Polling Has Reached Its End Point”

That’s according to Time magazine’s Michael Scherer, who cites a new survey from Public Policy Polling showing that one of Mitt Romney’s improvised campaign appeals is making big inroads into Barack Obama’s base in electoral-vote-rich Michigan.

The PPP robo-poll of 500 Michiganders asked, ”In Michigan, do you think the trees are the right height, or not?”

“That’s right,” Scherer writes, ”2008 Obama voters are 17 points more likely to agree with Romney on the height of Michigan trees. It was a crossover vote play all along!” (Trees also polled well among women and young people.)

It would be fascinating to follow over the course of the campaign whether Michiganders bring their vote intentions into line with their, uh, spatial preferences or—as is more often the case—simply adopt the views of their favored candidate if and when they learn what those are. Alas, I don’t think PPP does robo-panel surveys; and in any case, political polling has reached its end point.

(Thanks to Chris Achen.)


My guess is that the question isn't there so much for it's answers, but as a way to "prime" people in order to affect the answers to later questions in a way that Romney can play up in the media.

It's like that Harvard experiment where they asked Asian females to take a math test. Before hand, they answered a questionaire with seemingly innocuous questions like, "Do you live in an all-girls dorm," and "What language does your family speak at home?"

It turns out that even such subtle questions could change outcomes, with those women reminded of their gender fulfilling the stereotypes of "girls are bad at math" and doing worse, whereas those primed to think about being Asian fulfilled another stereotype and did better (both in relation to a control group).

Here, I'd say it's just a tactic to "prime" respondents into remembering that Mitt grew up in Michigan and is "one of them." They likely believe that will make respondents more sympathetic to him.

It's easy to poke fun at this question, but the real issue is whether or not that question primed people in a way that may affect their answers later in the poll.

Re-reading the post, it's not actually who inserted that question (it may not be the Romney campaign), so I may have the intended effect backwards. For all I know, it may be there to remind respondents that Romney says silly things. Either way, the point still stands. This question may seem dumb on it's own, but it may actually have a surprising effect on how people answered later questions. The idea of priming seems a bit ludicrous, but the strength of it's effects are pretty well documented and it's not something one should quickly dismiss.

You need to be logged in to comment.
(If there's one thing we know about comment trolls, it's that they're lazy)