Medicine Men

It often seems to me that the pollster more-or-less plays the role of witchdoctor in American politics. Ezra, for example, reflects on yesterday's business with Celinda Lake:

This poll wanted a result. It got it. It also could have gotten the opposite result. This happens all the time. It just depends on who's paying, and what they want to show. It's certainly true that good polling can be and often is, conducted, but far too much of it is of this type, and nether the polling industry nor the media polices these practices.

Right. For some reason, every advocacy group in town now-and-again stages an event where it commissions a poll with a reputable firm, the firm asks some questions designed to generate the result that the group's agenda is popular, and then it gets written up as a press release. All we learn from an exercise like this is that with proper framing you can get a poll to say just about anything. And everyone knows that. And given that the whole thing is fundamentally bogus, there's really no reason one should need to bother with the expense of hiring a reputable polling firm. You could just give me fifty bucks to make something up instead. A good pollster would be worth hiring if you really wanted an accurate read on public opinion, but that's not really the point in these situations. It's just a kind of ritual laying-on of hands.

Comments

Wateva wateva I'll do what I want.

Posted by Okgasm | September 24, 2007 8:42 AM

Hey everyone check out the lively discussion under Ganji's letter. Don't tell me you give more of a crap about Joe Biden than you do about the future of Iran!

Posted by Wurd | September 24, 2007 9:05 AM

It doesn't matter what the public wants.

Posted by steve duncan | September 24, 2007 9:13 AM

The point here is that in this poll as so many others, there was no effort to find out whether something was true or not, simply an effort to generate propaganda. Like one of those industry-funded studies that shows there's no such thing as global warming.

I feel like we're seeing the use of polls as propaganda more and more. And more and more, I hear people describe something as "out of the mainstream" or "extremist" simply because polls show -- or seem to show -- that something is a minority position.

Posted by Exile on Ericsson St. | September 24, 2007 9:17 AM

Well, the whole point of hiring a reputable polling firm is to pretend that you didn't just slant the questions to get your desired but actually conducted a "very serious poll" that just happened to produce results entirely consistent with your position.

Having a bogus poll conducted by a reputable polling firm is designed to mask the fact that the poll is fundamentally bogus.

Just giving you $50 to make something up defeats the entire purpose, since it negates the benefit one gets from pretending the poll is legit because it was done by a reputable firm.

Posted by KM | September 24, 2007 9:22 AM

"And given that the whole thing is fundamentally bogus, there's really no reason one should need to bother with the expense of hiring a reputable polling firm. You could just give me fifty bucks to make something up instead."

If they did that, they would get the results published in the Washington Post. Which is of course the whole point. It doesn't convince anyone of anything, but it stops the establishment papers from only (or at all) running stories about how the war is deeply unpopular (and going badly), which in turn reduces the pressure on wavering Republicans. Why the media plays along with this charade is another question.

Posted by Ginger Yellow | September 24, 2007 9:50 AM

Well, the whole point of hiring a reputable polling firm is to pretend that you didn't just slant the questions to get your desired but actually conducted a "very serious poll" that just happened to produce results entirely consistent with your position.

It's sort of like how you hire a guy from the Brookings Institute to write about how the surge is working because he's known as a "very serious" "war critic." You could get someone from AEI or Heritage to do it for less, but it wouldn't have the same impact.

Posted by Exile on Ericsson St. | September 24, 2007 9:54 AM

Something similar is true across a lot of fields where "empirical" evaluation is used. I was involved for a while in evaluating some federal education programs and it was quite clear that we were supposed to create some sophisticated-sounding incantations in support of the conclusion that the programs were quite effective, thank you very much. I get the impression from friends in the advertising biz that a lot (though not all) of the focus grouping and such that they do is similarly the functional equivalent of reading chicken entrails.

Posted by Bave Dee | September 24, 2007 10:40 AM

If unaware, this season of Frisky Dingo on Adult Swim is dealing with running for President and I think its take on pollsters is by far the correct one.

Posted by Rob | September 24, 2007 11:57 AM

All we learn from an exercise like this is that with proper framing you can get a poll to say just about anything. And everyone knows that.

No, actually everyone doesn't know that. It's precisely because back in the day people were shocked to learn things from polls, and that thereafter polls developed an almost religious aura of revealed truth around them (because they're 'scientific'!), that people can still be fooled by slanted polling. The polls run in the paper and the 'in-the-loop' journalists don't appear in the paper saying the whole thing is a fraud, so people who just read the paper take it as true.

I could point out that people of all ideological stripes do this, as it has gotten to the point that I wonder if any poll isn't slanted, but I'll just point out instead that newspapers still run op-eds written by person X working for lobbying organization Y under the byline of Z, who is invariably a supposed neutral party. But nobody mentions that in the paper either.

max
['Lies, lies and more lies.']

Posted by max | September 24, 2007 3:36 PM