Sunday, October 9, 2011

Nobel predictions

It is Nobel time again.

My predictions for possible prizes in the near future, including this year:

Structural labor: John Rust and Ken Wolpin
Structural IO: some subset (possibly the whole set) of Barry, Levinsohn and Pakes
Larsometrics (i.e. GMM etc.): Lars Hansen
Institutional design: Al Roth
Set identification ("bounds") and peer effects and statistical treatment rules: Charles Manski

Farther in the future:

Field experiments in development: Esther Duflo

Various other predictions from the interwebs here and here (via Mankiw).

Nobel prize guessing humor from NPR (do they have a sense of humor?) (also via Mankiw).

I am puzzled by the predictions regarding Gordon Tullock and Jerry Hausman. Giving it to Tullock should have been done as a shared prize with Buchanan. Doing it now would be an admission of error. I think not. My sense is that Hausman has mostly consulted in recent years. There is nothing wrong with consulting but I don't think the Swedes will reward it.

For non-economists: GMM = generalized method of moments and IO = industrial organization.

2 comments:

Anonymous said...

It is interesting to have the comments of an econometrician on the Economics prize and your choices are all very eminent and any one would each be worthy Nobel laureates.

However, I am surprised by you omitting Hal White from your longlist. Quite a few recent predictions seem to have included him in their lists.
Some people argue that heteroskedasticity robust variance estimators are in Huber, but then so is GMM in both Burguete, Gallant and Souza (1982) as well as in White (1982) (and its implied form in Godambe (1960)). It is the elegance of the formalization in Hansen's paper and the generalization to dependent observations that sets his paper apart. Similarly for the heteroskedasticity robust variance estimator paper of Hal White.

And much like Debreu's Theory of Value, Hal White's 1984 book is one of the first to rigorously lay out the foundations of modern asymptotic econometrics. Again the large literature on misspecification testing (for which Hausman's paper is seminal) also finds it theoretical grounding in the book and papers of Hal White.

I wonder what your opinion on the matter is. :)

econjeff said...

I am a big Hal White fan but I left him off my list both because I wanted to mainly focus on people I think have a shot who are not widely discussed on other lists and because I have not had time to sit down myself and sort out the contribution relative to Huber. A similar issue arises with Hausman and the test that bears his name. It also had anticipators in the statistics literature. Again, I have not had time to sit down and study the matter carefully, though it would be interesting to do so.