Some days I wake up grumpy . . . and some days, I let him sleep. No, no. Seriously, some days I wake up grumpy, and it usually has something to do with what's on the radio news as I am coming to in the wee hours. This morning, I am beyond grumpy. I am positively curmudgeonly.
NPR's "Weekend Edition Sunday" aired an interview this morning with Fank Newport of Gallup Polls, discussing public reaction to the passage of the health care reform bill. The general tenor of the discussion was that once the bill passed, a plurality of the public indicated support for the bill according to Gallup's own poll from last Monday. Mr. Newport added that even other polls which showed more people didn't like the bill than did like it constituted not "a vast majority of the American people," contrary to GOP talking points, but only a bare plurality of about 3 points [which, I note, statistically is within the margin of error--Ed.].
I have no quarrel with that. I do have a quarrel with the way all the polls discussed seem to have been formatted. If I understand the recap correctly, the polling was to learn the level of support for the bill. But it seems that what was asked was along the lines of "do you support the legislation, yes or no." That is no way to determine support for the bill. The only thing it indicates with any reliability is whether someone supports what s/he thinks the bill entails, and due to the months and months and months of negative spin spewed out by right-wing talking heads, what people think the bill entails and what the bill actually does are two very, very different things.
The only legitimate way to measure public support for the bill is to ask a series of questions about each key provision of the bill, to wit: do you support or oppose ending the ability of health insurers to deny coverage for "pre-existing conditions"? Do you support or oppose ending the ability of health insurers to drop your coverage if you get sick? Do you support or oppose allowing families to carry their adult children on their parents' policies until those children reach the age of 26? Do you support or oppose eliminating the so-called "doughnut hole" in Medicare's Part D prescription drug coverage? Do you support or oppose reducing the federal budget deficit by billions of dollars over the course of the next 10 years, as the non-partisan Congressional Budget Office has determined this bill will do?
I could go on, but I'll spare you. I will say that I am aware that the way I've phrased some of the questions I listed may not be "value-neutral." But I do not claim to be a pollster. I do think a pollster worth his or her salt could frame the questions in a valid value-neutral way. My point is not about the framing of specific questions in any event. My point is about the way to get the most accurate snapshot possible of the public's true opinions about the legislation, by asking questions about each of its provisions as opposed to asking a meaningless question about the bill as a whole.
Heck, even the overly broad question about the bill as a whole would be instructive IF there had been follow-up questions to determine what people actually think is in the bill in the first place. Without those follow-ups, the poll as it stands is meaningless. Anyone can take its results and twist them into whatever shape suits one's purposes, just as the CPA in the classic joke said "What do you want it to be?" when he was asked "What's two plus two?"
Mark Twain was right: there are three kinds of lies. Lies, damned lies, and statistics.
3 comments:
Well, polls are mostly for the benefit of politicians. So, if most people don't like the Health Care bill, that's all they need to know.
Whether it's because they think it costs too much money (a defensible critique) or they think it mandates "death panels" is of no importance to most politicians. The question is: "will supporting this make them like me?"
Your comments are absolutely correct regarding politicians, but I think there are other important uses for polls, which is why their results tend to make the neews. I discount polls, however, even ones from otherwise reputable pollsters, when their methodology isn't rigorous enough to support the conclusions being drawn from the "raw data."
"neews"? I meant "news." Sorry. I should never proofread my own stuff. I know what it's supposed to say, so I never see my own typos. Doesn't mean I think I don't make them. I KNOW I am not perfect. Just means that no one is 100% objective about everything all the time. T'ain't possible, Magee.
Post a Comment