Greg Sargent is right to note that 43% of Americans think it likely that House Speaker Nancy Pelosi was misled by the CIA during those now much-publicized briefings–not incidental, considering that much of the establishment has decided a story that should be about how torture was used by the highest levels of the US government to elicit false information to serve a political agenda. But, unfortunately, Sargent ignores just how absurd the rest of this four-question Rasmussen poll is.

The absurdity begins with question one: “How closely have you followed recent new stories about House Speaker Nancy Pelosi and CIA Briefings?” Rather then frame this as a story about torture, it’s a story about briefings; rather than present it as a story about the previous administration, it is a story about the current speaker.

Question two: “Do you have a very favorable, somewhat favorable, somewhat unfavorable or very unfavorable opinion of the CIA?” Hey, 63% of Americans find the CIA at least somewhat favorable! You know, people need to read a little more about the CIA—but, more on topic, this is important now because. . . ? Is this about who is more popular equating with who is more right? If you ran the world, would you like to see your foreign intelligence agency work to win a popularity contest?

Question three is the one Sargent highlights: “How likely is it that the CIA misled Pelosi about the use of waterboarding when interrogating prisoners?” Sargent actually isn’t doing America’s long-mathematically challenged establishment media a service by headlining his post with the declaration that “more” say it is likely that the CIA misled Pelosi—it is statistically a draw—but it is important to note that a goodly number of Americans do understand that the CIA maintains a large staff of professional liars.

But, it is the fourth and final question that is perhaps the most ridiculous: “How likely is it that waterboarding and other harsh techniques helped secure valuable intelligence information?” Is this a question that is to be decided by a poll? We actually know, from years of clinical study and unfortunate example, how effective it is torture someone with the goal of getting “valuable intelligence.” The answer is: not very.

While I suppose it might be “useful” to know where the debate is heading on torture (that there is a debate at all, to my mind, is a grim testament), this question is so overbroad and loaded as to make it meaningless. It doesn’t qualify what makes for valuable information, it doesn’t quantify how much valuable information, and it doesn’t contrast this with what could have been learned through less reprehensible means.

Can you say that torture never yields truth? No, but how can you be sure? How can you separate the one bit of real intelligence from the piles of inaccurate information that history shows is coughed up during torture? The answer to that question is not opinion pollable.

But, of course, the epic fail of this poll is the meta one. Rasmussen operates as a legitimate polling organization, but has, in essence conducted and published a push poll. It is perhaps not as obviously as bad as asking “If I told you that Nancy Pelosi cheats at Yahtzee, would you still say that the CIA misled Congress?”–but framing this as a dispute about the truthfulness of briefings and the justifiability of Pelosi’s “harsh” accusations is a naked attempt to shift the focus away from Cheney, Bush, and their use of torture in the service of their own ambitions. It doesn’t even cover the disturbing possibility that a quantifiable part of the population no longer embraces “America does not torture” as a core value.

The other half of the failure is on the part of all the news organizations that pick up Rasmussen and uncritically report its results. In reporting what is a perhaps a small positive from this poll, Sargent (someone that I have a lot of respect for, by the way), perhaps unwittingly, services a larger negative.

Gregg Levine

Gregg Levine