Monday, October 15, 2007

PREDICTION MARKETS AND THE ECONOMICS NOBEL: OOOOPS!

One of the more innovative of the George Masonites is Robin Hanson, who runs the philosophy of science-oriented blog, overcoming bias. He is a big fan of the accuracy of prediction markets. But the various betting markets on the Sveriges Riksbank Prize for Economics in Memory of Alfred Nobel, such as intrade.com, were just way off. The top bets as of yesterday were in order, Fama, Barro, Tullokck, Helpmann, Grossman, Dixit, and Tirole. None of the actual winners: Hurwicz, Maskin, or Myerson, was even in the betting pool. Oooops!

I think what we see here is the hand of the Committee Chair, Juergen Weibull, a game theorist. I had heard scuttlebutt that Myerson and Maskin would be the winners of "the next Nobel in game theory," but figured that was a ways off because of that being it two years ago. So, Weibull figured out to give it to game theory for an application, in this case, mechanism design. A question arises if this is his last gasp before stepping off the committee, or if he is a new Assar Lindbeck, a dominating figure who will be around for years. In any case, for those who are not aware of it, Hurwicz's approach to mechanism design looks an awful lot like a an effort to figure out how to do central planning right.

7 comments:

Anonymous said...

maybe Juregen Weibull was "gaming" the BS-PESIMOAN (i.e., “Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel”)?

Jim Devine

Bruce Webb said...

Apropos of nothing. (pointy headed way of saying OT)

Back in the day Peking (not yet Beijing) had huge billboards showing left to right Marx, Engels, Lenin & Stalin. Since to the typical Chinese person all westerners look the same, it was dubbed 'hIstory of Shaving'

Just saying.

Anonymous said...

Hurwicz studied with Hayek and Mises, but clearly doesn't "get it" when it comes to the force of their arguments. See, for example, Hurwicz's comments on a paper by Israel Kirzner found here:


http://www.cato.org/pubs/journal/cj4n2/cj4n2-4.pdf

rosserjb@jmu.edu said...

Hurwicz's views are pretty complicated and not all that easy to characterize. My remark in the posting is overly simplistic, and indeed just the opposite can be argued, as Alex Tabarrock has done over at marginal revolution. The basis of that argument is that in his effort to discover incentive-compatible mechanisms to get people providing the inputs to the planners, he showed how difficult it was to engage in efficient central planning, very much a Hayekian point, which is indeed argued in the opening pages of the Nobel Committee's backup scientific paper.

Hurwicz did intersect with Hayek at the LSE in the 30s, where he got his Ph.D., although von Mises was not exactly hanging out there. To say he "studied with Mises" strikes me as not accurate at all.

BTW, a curious detail on Maskin and Myerson is that they both got Ph.D.'s in applied math in 1976 from Harvard. Have not figured out if they had the same major prof or not though. Suspect that is a first for the Sveriges Riksbank Prize in Economic Sciences, if not any of the other Nobels.

Barkley

rosserjb@jmu.edu said...

I have checked. Hurwicz definitely at least sat in on at least one course of Hayek's at the LSE in 1938-39. He also claimes to have attended the Mises seminar in 1940, although that would apparently have been only briefly. He was furthermore influenced by the market socialist Oskar Lange, who was at Chicago in 1940-42. All this comes in some comments he made on a paper by Israel Kirzner on Hayek. He seems to have wished to integrate the ideas of the competitors in the Socialist Controversy into a synthesis in his effort understand mechanism design.

rosserjb@jmu.edu said...

BTW, on further thinking about it, I think what happened here is that indeed it was Weibull who pushed this through, but what was the stick he used to beat it through was Hurwicz's age, at 90 the oldest Nobel recipient ever. The betting market had two other oldies up there, Tullock and William Baumol, and it is widely believed they are on the A-list. But Hurwicz is just plain older than either of them.

Anonymous said...

See letter to Nobel Prize Committee on
"Originality in Mechanism Design" Clarke weblog