Saturday, October 16, 2010

A "Nobel" Prize For Epicycles In Labor Economics?

Peter Diamond, Dale Mortensen, and Christopher Pissarides won the "Nobel" prize in economics this year. I do not have Bill Mitchell’s expertise on search theory and labor economics. Nevertheless, I thought I’d record my reactions.

As I understand it, Diamond, Mortensen, and Pissarides think labor would be described by by the interactions of well-behaved supply and demand functions for labor if it were not for the heterogeneity of workers and jobs and the time to form matches between them. The orthodox theory would be wrong even if workers and jobs were homogeneous. So I find puzzling why I should approve of this year’s award. I think my opinion is consistent with some of Alessandro Roncaglia’s observations of trends in mainstream economics.

David Ruccio and Richard McIntyre & Michael Hillard (Hat tip to Nick Kraft) have been critical of this year’s award. Paul Krugman has praised it. Doubtless, one could find more praise in the blogosphere.


BruceMcF said...

Yes, a prize for elaborating on a false theory to make it better simulate reality.

But simulation is not cause and effect explanation, and so is not science unless it is part of a research program aimed at arriving at cause and effect explanations.

Alex said...


I agree with Bruce completely.

This year's award only reiterates the fact that mainstream (marginal) economics is just too far away from reality (apart from their other problems).

Anonymous said...

Ooooh, come on, all you guys know that reswitching is a very, very rare phenomenon, don´t you? Please, don´t be intellectually dishonest.

Anonymous said...

I'm surprised that you have not mentioned the wonderful co-incidence that so often marks economics. Namely, that just when governments around the world are seeking to impose austerity the (non-)Nobel Prize folks give the award to people whose ideas just happen to back these attempts up.

As reported in the Guardian (who are just paraphrasing the award press release): “One of their conclusions is that more generous unemployment benefits give rise to higher unemployment and longer search times.”

As the Guardian usefully notes: “The research is particularly relevant now when looming drastic government spending cuts in Britain will lead to thousands of job losses in the public sector.”

Of course, that is just a co-incidence. Just as, although I'm sure the winners said other things, the award committee decided to stress the "cut unemployment benefit" aspect of their ideas...

An Anarchist FAQ

BruceMcF said...

Anon#1, first, regarding reswitching, how precisely would you know how common it is? In an econometric study that assumes it away at the outset, the observation cannot be made. The portion of an econometric model that relies on analysis from original assumptions must consist of both valid analysis and assumptions anchored in reality to be of use in actual empirical work.

If, on the other hand, we start by assuming a long since invalidated theory of human decision making, then there is no reason to believe that the modeling is more likely to reveal rather than obscure what is going on compared to an empirical framework that is naive regarding the underlying theory of choice.

Suppose the presumed model departs by construction from the actual underlying relation "only" 5% of the time. What does that imply for the true Type I error of an econometric parameter that is estimated to have a Type I error of 5% if the false model of the relation used in the analysis was, in fact, a true model?

Robert Vienneau said...

Thanks for the comments. I hadn't looked at your blog comments on this year's prize, Iain, before writing mine.

(Speaking of other blogs, I see Alex has a post on Krishna Bharadwaj, one of my favorite Indian economists.)