Monday, October 13, 2008

Why I Forecast that Krugman Would Not Get the Economics Nobel Prize

So, I accurately put international trade first as a field to get the Nobel this year and also forecast accurately that the prediction markets would be wrong. However, I said the trade prize would be to Bhagwati and Dixit with a third, but not Krugman. I was clearly very wrong on that. Why was I wrong about Jim Devine's former roommate?

Well, they have said that they gave it to him for applying his ideas to both international trade and location theory, and I must give him credit on that. The last person to get one for international trade did so also, the Swede, Bertil Ohlin, whose most famous book was _International and Interregional Trade_. I should have seen that one coming. He applied the Dixit-Stiglitz model to both fields to develop the respective "new trade theory" and "new economic geography."

So, why did I forecast he would not get it? Because he was not the first to apply that theory to those areas. For that matter, he did not invent the idea. Stiglitz and Dixit did. Now Stiglitz got his prize for asymmetric information, but Dixit has not gotten one, and it was for this paper that I had him on my list for a trade prize. This reminds me of Robert Lucas getting the prize for applying rational expectations to macroeconomics, while the inventor of the idea, John Muth, has to this day not gotten one.

As for applying, there were others, such as Elhanann Helpmann, who applied the D-S model to trade (Krugman always cited his and others work on this, although he took it much further than they did). As for applying it to location theory, that was first done by Masahisa Fujita of Japan, whom Krugman also cited. But it was Krugman who did it in both places and pushed the application far. That is why he got it.

Now, I have been on record harshly criticizing him regarding some of this, in particular in my review of his book, _Development, Geography, and Economic Theory_, which appeared in JEBO in 1996. In particular I was annoyed that he put himself forward as the first (except for Fujita, with whom he would later coauthor) to provide a mathematically rigorous explanation of agglomeration economies. Now D-S does this but on the basis of consumers liking a variety of goods, not on production efficiencies, which is what most observers think is what is really going on. However, there had been a long literature during the 1980s by non-economists in non-economics journals in fact developing mathematical models of regional dynamics with agglomeration economies (examples, the physicist Peter Allen, a student of Ilya Prigogine at the Free University of Brussels, and Wolfgang Weidlich of the Stuttgarth Institute of Theoretical Physics). Krugman never cited any of this work, and, frankly, this pissed me off.

Now, I must admit something important: I do not know for sure that Paul Krugman ever read any of that literature. It is possible he did not. Even if he did, he could defend himself on the grounds that although these models are mathematical, they were not drawn on neoclassical economic theory in a standard way. Dixit-Stiglitz is.

Now, I know that most readers here are probably cheering the award because of Krugman's long opposition to Bush's policies from his perch at the NY Times. I must give him credit for pointing out problems in Bush's policies at times when he was popular and few others were in the media. I also think he is a very brilliant and innovative guy, and I have long said that he probably eventually deserved a Nobel for his work on foreign exchange rate models.

While I think his excellent current commentaries on the current financial crisis may have played a role in his getting the prize, I do not think his criticisms of Bush did much, although commentators over on Marginal Revolution are frothing at the mouth over the award, saying it is "political." I would note that one of last year's winners, Hurwicz, was a student of Hayek, and the previous year's winners, Kydland and Prescott, were open supporters of Bush and his policies, especially Prescott who followed his award by writing a very stupid column in the Wall Street Journal praising Bush's fiscal policies and calling for more supply side tax cuts.

So, anyway, I am not against Krugman getting the award. He is indeed brilliant, and his applying the D-S model to both trade and location theory was very innovative. Congratulations to him.

3 comments:

Robert D Feinman said...

It is interesting that most awards in economics are given for the creation of models, while in physics theoretical physicists hardly ever win.

The Nobel academy has been burned a few times in the past when it gave an award to a theory which later proved to be false, so they prefer easy to verify activities like the discovery of some new effect or the creation of a device (like the transistor or laser).

Of course, if we waited for economic theories to be verified by actual data before giving out an award, as is the case in physics, then they would have to give up the prize altogether...

Personally, I consider it akin to the Oscars - an insider industry award without any input from truly independent judges.

However, it does allow economics to pretend to be a hard science even if the award is only a "Nobel".

YouNotSneaky! said...

"It is interesting that most awards in economics are given for the creation of models"

Not really. Well, maybe if by 'most' you mean a bit more than 50%.
Off the top of my head, you got your first Nobel prize going to empirical people like Frisch and Tinbergen. Leontief did both models and empirics and got the prize for input/output analysis which is also about how to measure empirical stuff. The economic history Nobel for Fogel and North was also based on empirical research. Kahneman did psychological experiments as well as develop models. V Smith did experiments. Engle and Granger devised new, better ways of measuring stuff (though I guess you could call some of this 'models'). Same with Heckman and McFadden. Out of the rest, most did probably both models and empirics, like Phelps or Sen, or the finance folks.

For a list of "pure theorists" you'd have Aumann and Schelling, Arrow, Samuelson, Debreau, Allais, Nash, Harsanyi, Selten, Lucas (along with Krugman, sort of theorists), Simon, and a few others.

Unknown said...

I propose that those who want to trash Krugman and claim that the decision was political should explain themselves and show some knowledge of Krugman's academic work.

------------------

ages

Internet Marketing