Saturday, March 10, 2012

Comment on Paging Hugo Sonnenschein

Brad quotes Kevin Quinn noting that DSGE models almost always include a representative consumer while Debreu (not mentioned) and Sonnenschein proved that, with many different rational consumers, all one can conclude is that excess demand satisfies Walras's law.

I comment.

Can you say Geanokoplos Polymarchakis ? Neither can I (and the spelling is pretty much a guess -- I'm tired of looking their names up). OK but you can say Cass.

DSGE models also assume that markets are complete which is roughly off by the same factor as assuming that there is one consumer. One might ask if the basic results of general equilibrium theory with complete markets hold, or generically hold, or probably hold, when markets are incomplete.

The answer is no. G & P show that, for an open and dense set of economies with incomplete markets, it is possible to make everyone better off by restricting free trade by rational agents. Cass and many others show that market outcomes can easily be (generically ? I don't know) indeterminate if markets are incomplete. Sunspots can matter the assumption hidden in "*the* outcome ground out by the system of markets" is a sick joke.

The assumptions at the base of DSGE models are not at all those generated by GE theory post 1960. They are the assumption that things which can happen in General Equilibrium models can't happen, because if you don't assume them away, then you must conclude that with strong assumptions (acceptable to GE theorists) you get no conclusions. The people who want to base macro on GE theory do not bow down to GE theorists. They assume that when GE theorists say "we don't know" those theorists are full of it.

No comments:

Post a Comment