The Panetta-Burns statistic

Chuckles all round about PPP´s poll asking for opinions on the nonexistent ¨Panetta-Burns¨ deficit reduction plan (8% for and 17% against) and whether defunct ACORN stole the 2012 election (49% of Republicans think so).

Perhaps this is more than a nice once-off gag. Double-blind clinical trials of new medicines calibrate them against placebos and white coats, which usually have some effect. Lie detection sessions are calibrated on questions known to be true and known to be false.

Including phony questions systematically in polls would give an indication of the attention the public is paying to the issue, or at any rate to the pollster´s question. Informed pollees may just be taking the mickey. But the mickey-takers are there anyway, creating error. A ringer question flushes some of them out.

Imagine a Presidential election poll.

If the election were held tomorrow, who would you vote for?
Mitt Romney, Republican Party
Barack Obama, Democratic Party
Gary Johnson, Libertarian Party
Jill Stein, Green Party
James Wimberley, Eurocrat Party

Since the election was very important and media coverage was beyond saturation, the ringer (me) would get a near-zero response. Not so, as PPP has shown, for a highly technical and artificial Beltway flap like the fiscal cliff. You could call the percentage response the ¨Panetta-Burns statistic¨.


  1. says


    Years ago, back in the pre-internet days, one of the Sunday morning news shows called around to various Congressional offices asking for a statement on the situation in Freedonia. More than a few issued fairly strong statements alluding to the importance of Freedonia, supporting or offering criticism of the President’s stand on the issue.

    Freedonia is, of course, the fictional nation in the Marx Brothers movie, Duck Soup.

  2. Ken Rhodes says


    James, I think your last paragraph contains a hidden bias. You would have received near zero votes because (a) nobody heard of you (except your loyal followers here, who would have voted for you), AND nobody would vote for the Eurocrat candidate.

    If, on the other hand, you offered a choice including James Wimberly, Freedom First party, I betcha you’d have gotten a significant vote.

  3. Ken Doran says


    Opinion polling actually gets an unjustifiably good name from election polling that really shouldn’t carry over to most other things. For better or worse, I am the world’s leading expert on who (whom?) I am going to vote for. I may change my mind, or forget to vote, but I am extremely unlikely to answer “wrong”. On almost any other subject, it is difficult to prevent a low information respondee from taking a wild flier; witness Panetta-Burns. My favorite evergreen example is that poll majorities always favor cutting foreign aid, but when asked what the level should be they pick an amount far higher than the actual rate.

    • Sean says


      1) I am the leading expert on who is the best candidate because he is the best candidate.

      2) I am the leading expert on whom I am going to vote for because I am going to vote for him.

      I know the joke’s about the Whom Do You Love? post, but just to clarify for anyone else.

  4. rachelrachel says


    Here’s the questioning of the PPP poll:

    Do you think that Barack Obama legitimately won the Presidential election this year, or do you think that ACORN stole it for him?
    Obama legitimately won……………………………. 66%
    ACORN stole it for him……………………………… 24%
    Not sure …………………………………………………. 10%

    Somebody who thinks that somebody stole the election, regardless of whether they thought ACORN did it, might very well take the second option, because it’s the best fit. They might think:

    (a) ACORN really did steal the election;
    (b) ACORN is no longer together, but the people who used to be in ACORN, along with others like them, stole the election
    (c) somebody stole the election, I don’t know who ACORN is, they must be the ones that did it
    (d) the election was stolen, I know damn well that ACORN doesn’t exist anymore, but this answer is the best alternative I have; stupid poll question!

    This sort of poll isn’t designed to get good information. It’s for creating headlines that say, “Ha ha ha! Look how stupid these people are!” A better constructed poll question would leave out the mention of ACORN.

    • J says


      Well, I agree with you that it’s badly worded. I also agree that people who said “ACORN stole it” could be saying that based on any of your reasons a, b, c, d. I also agree that people who are hyping this poll result as “haha, look at those dumb Republicans who don’t know that ACORN was disbanded” are being intentionally or unintentionally disingenuous.

      HOWEVER … all four of your options still involve the silly belief that the election was stolen. So that 24% (plus the not-sure 10%) are still deserving of ridicule, IMHO. Just not necessarily because of the word “ACORN”.

      • says


        So that 24% (plus the not-sure 10%) are still deserving of ridicule, IMHO.

        That 20% is what we used to call, when the web was young and nasty: Cheney’s Dead Enders.

        • Gray Woodland says


          Do we not have it on the authority of Kin Hubbard that twenty percent of the people are against everything all of the time? This is no mere neo-anything novelty, but a venerable tradition!

  5. JMG says


    “Lie detection sessions” are NOT calibrated because there is nothing to calibrate, any more than a dowsing rod can be calibrated. Lie detectors detect only the presence of a person whom the scammer giving the test hopes is a gullible dupe whose blind belief in the theatre of “lie detection” causes him to self incriminate. Sometimes there is no scammer present, just two dupes, one asking and one answering questions.

    • James Wimberley says


      You are right and I should have hinted that lie detection is a fraud. But it is pseudo-scientific, and don´t they try the knownm true and false questrions to start? That part of the procedure would be sensible, if the bssic theory held up.

    • James Wimberley says


      You are right and I should have hinted that lie detection is a fraud. But it is pseudo-scientific, and don´t they try the known true and false questrions to start? That part of the procedure would be sensible, if the bssic theory held up.

      • Brett Bellmore says


        It’s not so much that lie detection is a fraud or so I understand it, as that it’s a not terribly reliable art rather than science.

        Kind of like drug sniffing dogs, really; It’s not that dogs can’t smell pot, they can. But they’re so good at figuring out that their handler wants them to find something, that they can be practically useless outside of controlled conditions.

        False positives are just horribly easy to generate, even if you’re not meaning to.

        And if they were trying to be scientific, they wouldn’t start with known true and false questions, they’d intersperse them.