Wednesday, March 30, 2011

Therefore judge not their grasp of reality

When a person's pronouncements have a serious impact on how others behave, there may be good reason to be strategic about those pronouncements. We shouldn't expect them to be entirely honest.

When the Fed chairman comes out with a prediction about where the economy is headed, it has a serious impact on the he's not necessarily going to speak his mind. If the "prediction" later turns out to be wrong, we don't really get to turn around and call him a fool for believing what he believed, because we don't know what he really believed. We only know what he claimed to believe, and maybe what he said really was the best thing for the economy.

On the other hand, when there's a large number of people coming to a consensus -- so that no one person has a significant effect on what is believed in the aggregate -- then they may freely speak their minds. Absent other motives, we can be more sure that this consensus accurately reflects true beliefs.

So consider the difference between, say, a Democratic president on the one hand, and the Democratic party on the other. Or, the difference between a Democrat after he becomes president, versus before when he was just a member of the Democratic party (perhaps a congressman). Is there something predictable about the evolution of his stated beliefs? Some component that we can argue is due specifically to the increased impact his stated beliefs have on the world?

Political pronouncements are of course muddled by other sorts of motives, but that doesn't mean we can't say something interesting. It would also be hard to disentangle the fact that the president learns a lot of secret information that he did not have before, and which he cannot necessarily share with the public. What if when you become president, they sit you down in a room and expose you to a secret but compelling argument for why it's really sensible to start some war you never would have considered before? Perhaps we can never know what the president really believes, or whether we'd do the same thing if we were in his position, with the same information.


  1. Exogenous CombustionApril 2, 2011 at 12:44 PM

    You're right we might not know why, for example, a presidential candidate might claim that he'll close Guantanamo Bay and end military tribunals then change is mind on both. Not in our lifetimes at least.

    But the good news is that often the truth outs in the long run, either due to leaks of government records, Freedom of Information acts, revolution within country, revolution in allied governments, or FOIA acts in other countries (such as was recently the case in Britain). While this isn't useful for answer a question now, we can understand the question now given a distribution of similar issues in the past, where we do have information, do have the "secrets" told to presidents.

    For example, whether or not Kennedy sending military advisors to Vietnam may have been due to some great secret back in the day was unanswerable, and releasing the records could have been a mistake internationally. But now we have detailed records of the cabinet discussions on escalation and aid. We know what was said.

    These former revelations provide a distribution for actual convincing-ness of then-secret arguments. I might not know whether or not Barack Obama's decision to go to war in Libya was a mistake. But I can say that the distribution of "secrets" told to presidents in the past makes it highly unlikely that he was indeed told such a secret.

    In this case, we may be able to have a good idea whether or not we'd do the same thing in his position, because we have comparable full-information counterfactuals from the past.

  2. yes.

    but in a trivial sense, even if we never found much out later, we would always have *some* prior over possible justifications. You're saying that this prior can be even sharper than I acknowledged: true. But qualitatively it doesn't change the game at all. (that said, thank you for calling me out: I was not considering the info we learn later.)

    It isn't right to say "do not judge" as if we don't have *any* information with which to compute the likelihood of justification. I should really explicitly say "do not judge as if you have all information." The point is only that it's extremely tempting to take people at their word when we disagree with that word, in order to criticize them. If Bernanke says something you don't agree with, isn't it just enormously tempting to write a blog post registering your counterargument about why Bernanke is wrong? The difference between a post I do and don't like will be whether the words attack Bernanke himself, versus the statement he made.

    "He said it, therefore (he believes it and) I may now attack his grasp of reality for saying it." This is standard. You see it everywhere. Nobody questions it. boo.

  3. Exogeneous CombustionApril 4, 2011 at 3:16 PM

    I agree.

    I might add that there's also a (still noisy) real-time update on information. Prognosticators who, before an election or announcement, were accurately able to predict a change in opinions, should be weighted more heavily (I posit) due to their superior forecasting models (or worldview). That is, if a predictor says that Obama will have a change of heart and try individuals militarily due to some or another reason, and others say that he won't, then I suggest that the individuals who correctly predicted Obama's change of heart are more likely to be privy to the "secret" he has been told.

    That is, not all "secrets" are secrets. Just because Obama or Bernanke didn't know then going in, doesn't mean that outside individuals who are able to make predictions about their behavior aren't privy to the information that changed their mind.

    This example is said in a farcical manner, in light of my comment two days ago and today's announcement of Khalid Sheikh Mohammed's now-impending military tribunal.