Political rant (I haven't done one of these for a long time)
"Where mistakes have been made, the responsibility rests with me."
This is the first time I've seen the guy as broken, as outwardly humiliated, as cautious as this. And I've never trusted him more. For once (and not without exceptions within the speech) he's stating his foriegn policy in realistic and practical terms as opposed to the mythologies of the past few years. If only he had started on this sooner.
Politics, in the broad sense, came up with Piotr and Hal on the car ride up to Maine. They are both roughly libertarian. I'm not sure where Hal comes from exactly, but he says that he "values liberty per se," so I think he may have some deontological convictions. Piotr is an apostate Randroid who now espouses a consequentialist pro-market view no doubt informed by Jeffrey Friedman. I picked a fight over paternalistic market intervention. It's an old fight--we've argued about it on this journal before, when they were regular readers and responders. My opinion is that if you are judging policy based on its effects, then to the extent we have any new information about how we can make policy work better, we ought to. If that means deviation from free marketeering, so be it--who died and made Adam Smith pope?
Anyway, the only reason why I bring that up is because it highlights, I guess, a newish kick about the role of information in politics. This is, for me, a logical meliorist step forward from JF's work on the role of ignorance in politics. It's not really a new view; economics of information is old. JF is drawing from Hayek who I've never read but who I understand gets his critique of socialism out of what could be understood broadly as an information processing view, arguing that market forces successfully use information a top-down planner could not. So be it. But the argument shouldn't stop here, as I get the impression the naturally libertarian leaning folks want it to. Rather than just cross our fingers with a market system that we have acknowledged to be imperfect, what we need is an in-depth analysis of where the packets of information lie across the terrain of politics and economy, and figuring out how to tap into them. Emissions cap and trade programs are an example of a brilliant merger between these ideas: a top-down environmental concern about emissions gets worked out in a policy which responds flexibly to the needs and intelligence that are the special province of those economic actors that are on the ground.
I think that this sort of innovation is real progress. It is effective political problem solving which I think has potential for radical advances in quality of life within the capitalist economy and in public policy in general. But it needs to be applied to all those funky aspects of the economy that the market handles so poorly: public goods, intellectual property, land and other infinitely inelastic goods, advertising, and ultimately, I think, human irrationality itself. And it needs to be applied to politics--immigration, foriegn policy, social policy, etc.
That's where Tetlock (and Bush) come in. Tetlock provides a measure of the success of political judgment (incidentally, one based largely on Bayesian belief updating), and finds out that foxes do better than hedgehogs. "What the hell does that mean?" What he's highlighting is a difference in cognitive style. Foxes are those that draw arguments and reasoning from a broad range of experiences and theories, hedgehogs are those that interpret everything according to a single defining idea. Another way to say this is that foxes cast a wider net for information in the world. Another is to say that they juggle contradictory view better, better following the principle of Epicurus by keeping undefeated theories alive and in consideration. Another is to say that they are better Bayesians due to pluralistic acceptance of a broad range of ideas.
The point is that foxes make more accurate political judgments, and the reason why is that, like prediction markets, they critically assess a wide range of ideas and have a built in sensitivity to the base rate of events that prevents extremism and overcomes the sort of narrative-based, base-rate-ignoring prediction that sucks so much. (Interesting note on prediction markets: apparently this economist has already coined the term for a government which evaluates policy via prediction markets, letting the electorate "vote on values, and bet on beliefs." The term is Futarchy. I think it's a dumb name, but whatever. It seems like the right idea except for its dependence on the dubious fact/value dichotomy....)
So...Bush. He was once the ultimate hedgehog, and presented himself forcefully as such. His image wasn't about careful consideration of competing ideas. It was about an unwavering, single-pointed crusade. And despite a certain amount of agnosticism about his policies, I (in an example of what, if I let myself fall prone to some kind of hindsight bias and more, would describe as a non-linguistic anticipation of what I now have the words to express) thought that his hedgehogness made him a stupid politician. In his most recent address, he seems weakened. He talks at length about how he considered other points of view. He expresses his reasons for rejecting them. He states the grim realities of the task ahead. He has abandoned the language of certainty for a language of contingency--"even if X, then Y..."
And I've never liked him better.