The website FireDogLake has commissioned a new survey to ask the constituents of Democrat Vic Snyder in the Arkansas 2nd district, one of the most conservative to have voted for health care reform, about about their opinion of the individual mandate. The survey fails to provide context about the individual mandate, and arguably biases the respondent against it through its choice of question wording and question order. Although the survey finds Snyder in a very tough position, it shows little evidence for a further decrease if an individual mandate is adopted as part of health care reform. Let's go through the survey in detail.And then he does, in typical excruciating Nate Silver "detail-oriented Nate Silver is detail-oriented" fashion, and then comes up with this at the end:
So, for all that work, the poll shows a whopping 4-point decline in Snyder's poll numbers, and a 2-point increase in Griffin's -- not even outside of the margin of error. We don't know how much of that has to do with opposition to the mandate versus the balance of the bill since the poll doesn't unpack them -- they could have asked an additional question or two to tease this out, but they didn't.You don't tug on Superman's cape, you don't spit into the wind, you don't pull the mask off the ol' Lone Ranger and you don't mess around with Nate Silver on the inherent accuracy of polling methodology and question bias.
And that 4-point decline -- which may or may not be statistically significant and which may or may not have anything to with the individual mandate -- comes only after they'd asked five or six questions in a row that framed the mandate in a negative light, and also reminded people for no particular reason about just happy they are with their coverage in the status quo -- all while using robopolling technology that was never really designed to ask complex sets of policy questions like these.
Great work, guys!
FireDogLake, what the hell has happened over there?