It's interesting to note that there's no dramatic difference in terms of support for presidential hopefuls between the Party supporters and Republicans generally. There is stronger support for candidates, but that stronger support is fairly evenly distributed--indicating that this is a group with a lot more early decision-making going on, as is often the case among those who are more dissatisfied:
It's hard to see Romney maintaining his lead position as other Republicans starting attacking him for being the inspiration for RomneyObamaCare, but his current high standing does attest to the power of his Gipper-like charisma hair. Meanwhile, what's up with Jindhal? He's the closest thing we have here to a true outlier, almost breaking into double-digits among the Teabaggers. Things to keep an eye on, one can only suppose.
But now to the more meta-discussion, which for once actually does have a legitimate claim on our attention.
At the National Journal, Mystery Pollster Mark Blumenthalrelayed criticism from ABC pollster Gary Langer, to wit that: - Introducing the statements as "things people have said about Obama" gives them added credibility.
- True-false questions take an "unbalanced" approach that academic research has shown to "overstate agreement with whatever's been posited."
- The statements were all "unrelentingly negative," rather than a mix of positive and negative.
- Respondents may have been using the questions to "to express their general antipathy toward Obama" rather than endorsing the truthfulness of each statement.
Now, there's something to be said for all of the above. These are well-known polling principles that professional pollsters ignore everyday. Usually that's a bad thing. But this time, I'm not so sure. It's not that I think there shouldn't be a poll that adopts all those corrective measures--I do. And, in fact, Mark relates that Kos has already offered to pay for a well-designed poll if Mark comes up with one. That would be a very good thing, and I really hope that it's done.
But in the meantime I have two responses in defense of the Harris poll. First, by far the bigger problem here is that virtually no one has asked most of these questions before--aside from Kos who has focused primarily on the Birther question--even as the circulation of these lies and others related to them has increasingly become a central part of the political discourse opposing the Democratic President. The failure to take note of this phenomena, draw attention to it and criticize is it a passive way of enabling it, and is a fundamental dereliction of duty of the press. As Media Matters reminded the Washington Post yesterday ("Attention Washington Post: Barack Obama still is not a Muslim"):
For some reason, it seems difficult for reporters at the Washington Post to understand that debunking a lie occasionally isn't good enough; you have to make clear that it is a lie every time you mention it. Unless, of course, you want people to believe the lie.
All the major pollsters except for Harris have been similarly derelict up until now. This failure on their part far overshadows any of the methodological quibbles they've raised.
Second, I'm not at all sure that this methodology is bad--or more properly inappropriate--for the phenomena it is measuring. Indeed, the very sorts of cognitive bias that pollsters generally try to screen out are intimately related to the spread of these beliefs in the first place. By allowing these biases to play out, as Harris did, one gets a raw, unfiltered picture of the crazy talk atmosphere that we all know is out there. Of course it's a good idea to also ask questions in a highly disciplined way intended to produce balanced, reflective responses. But if balance and reflection were the norm in America today, Harris would never have done this poll in the first place, because there wouldn't be such a rich pool of political lies to draw from.
In short, the ideal pollster practices to screen out distortions and cognitive noise may themselves produce a misleading picture of a cultural phenomena that is based on distortions and cognitive noise.
The pollster myopia here can be fairly deep. Over at Swampland, for example, Michael Scherer also weighed in somewhat tendentiously, making a big deal over a 9-point difference between Harris and DKos/R2000 over how many Republicans think Obama is foreign-born (45% for Harris, 36% for Dkos). Whichever figure is right, this should be a big deal that pollsters ought to be concerned about. But Sherer is more interested in the disparity, which he seems to think is huge ("Either one of the polls is wrong, or about one in ten Republicans found a reason in the last six weeks to doubt the legitimacy of Obama's birth certificate.") But given the MOEs for such sub-samples, and the continuing flow of media misinformation on all things Obama, it's really not such a huge dispariy. But Sherer's "explanation" is even more revealing of the problem with pollsters' myopia:
The latter seems unlikely. Another option: People answer polls not to say what they actually believe but to register their anger. The Harris poll was done during the height of the health care debate frenzy.
This may well be true. But if it is, it's all the more remarkable, and worthy of attention. If almost one in ten Republicans would respond to anger over the passage of REPUBLICAN-style health care reform by saying they think the President holds office illegitimately, then that should be a big deal in and of itself. And the failure to see this speaks volumes. Sherer goes on to note:
The Harris responses about Obama's religion are also odd. Pew polls in 2008 and 2009 found that the portion of Republicans who thought Obama was a Muslim unchanged: 17 percent. What would explain that number more than tripling in the last year to 57 percent? I have no idea, though it makes me wary of taking the Harris poll too seriously.
My response is the exact opposite. It's not that I think Harris is right and Pew is wrong. It's that any plausible explanation I can think of is troubling, and I'm further troubled that Sherer seems so incredibly blase about it all. One explanation is that a lot of Republicans don't really believe that Obama is Muslim, but they believe even less in the value of being truthful, honest and reflective. That's supposed to comfort us?
I'm reminded of the question of the size of an atom, which I learned about back in the Pleistocene Era There was a time when this was a very big deal, a topic of intense dispute. Then it gradually dawned on the physics community that there was no answer to the question, as they groped toward an understanding of what was going on. As it turned out, the size of atoms was a function of their electron shells, which could only be measured by shooting other particles at them, and measuring how they bounced off. But this was as much a function of the particles being shot at the atoms (how heavy/light they were, how much energy they had) as it was a function of the atoms' electron shells. If you wanted to know how "big" an atom was, you had to arbitrarily decide a standard for the particles used to measure it. This would produce a standardized measurement method, that everyone could use consistently. But it would not eliminate the fact that it wasn't measuring the size of the atoms it was supposed to be measuring. It was measuring a size, according to a measurement protocol.
This is the same position that pollsters find themselves in. Except for one thing. Normally, there are some very good reasons guiding pollsters' choice of measurement protocols, making their choice fundamentally less arbitrary than the choice faced by physicists. And yet, "normally" is not now. Now is anything but normal.
So, by all means, bring on the rigorously-constructed polls. But don't for a moment think that they will bring you the true picture. They will bring you an important picture. But it will take a lot more variations in methodology to even begin to get a picture that is true--if, indeed, that could ever be possible, given how much in flux our politics of fantasy is at the present point in time. |
STATE BLOGS
