Tuesday, June 05, 2007

Arrow-Debreu Worth Generalizing To Model Power?

Power can be formally modeled in a framework of incomplete contracts, asymmetric information, strategic interaction, and principals and agents. Samuel Bowles & Herbert Gintis and Joseph Stiglitz, for example, have developed models of asymmetric information.

Is it worth developing models of power, as a generalization of the Arrow-Debreu model? I don't know that Stiglitz has ever addressed this question. His papers, as I understand them, do tend to be generalizations of the Arrow-Debreu model. This could be a rhetorical strategy. Bowles and Gintis (2000), however, assert that the Arrow-Debreu model is a detour. One can model power more directly.

Perhaps this contrast is an illustration of the different perspectives of economists that, until recently, would have been considered orthodox and heterodox. If so, it also illustrates that the boundary between orthodoxy and heterodoxy is murky.
  • Samuel Bowles and Herbert Gintis (2000). "Walrasian Economics in Retrospect", Quarterly Journal of Economics (November)
  • Samuel Bowles and Herbert Gintis (2007). "Power", University of Sienna, Working Paper number 495

3 comments:

YouNotSneaky! said...

(Currently I'm in need of an occasional break from grading papers)

Power can be formally modeled in a framework of incomplete contracts, asymmetric information, strategic interaction, and principals and agents
...
Is it worth developing models of power, as a generalization of the Arrow-Debreu model? I don't know that Stiglitz has ever addressed this question. His papers, as I understand them, do tend to be generalizations of the Arrow-Debreu model.


I've never thought of Stiglitz as a true-blood GE guy. To the extent that he's incorporated a lot of the issues mentioned above it's generally in a partial-equilibrium or toy-GE (i.e. some kind of representative agent/firm) or search/matching models.

Incomplete markets is a bit of a different ball game, which is/can be thoroughly Walrasian but I don't think Stiglitz is as much of a contributor to this area as some others (Oliver Hart, Magill and Quinzii). I don't think it's immiediately obvious how this can be extended to incorporate power but the idea isn't crazy. One aspect of this is that if you start out with incomplete contracts then add some more possible contracts, but don't go all the way to completeness basically anything can happen (welfare decreases, everyone worse off, as bad as you want it - i.e. financial innovation not always good). I think this is more general when considering something like power - the problem of the second best.

More generally the general feeling has been that market power is very hard to incorporate into GE, some fairly limited instances like monopolistic competition and limit pricing aside. I believe Makowski and Ostroy, among others, tried.

Robert Vienneau said...

I did not have "market power" in mind. As an example of power, consider an employer's ability to order an employee to do something, with the expectation that he will be obeyed.

YouNotSneaky! said...

I understand that. But market power is an obvious type of power. And also it really just boils down to "who controls the variables" - in the case of monopoly, the variable being price, and in the example you give the set of tasks that an employer can assign to a worker.