Wednesday, January 02, 2008

Who Changes The Prices?

An assumption in neoclassical models of perfect competition is that all agents are price-takers. That is, they take prices on the market as parametric data. If everybody takes prices as given, nobody is left to change the prices. So, for example, in general equilibrium theory, one has to assume the existence of an auctioneer to oversee the tâtonnement process. No trading, consumption, and production can take place outside of equilibrium.

This is not an original criticism. Bernard Guerrien calls this problem "the logical flaw in micro theory" and "the principal bug in microeconomics". (I do not agree. Neoclassical microeconomics has so many bugs it seems hard to justify calling one, even though important, the principal bug.) Apparently Kenneth Arrow brought this problem up in 1959. ("Toward a Theory of Price Adjustment", in The Allocation of Economic Resources (edited by M. Abramovitz), Stanford University Press).


Gabriel M said...

Arrow has been bringing this up since. There's a recent paper making the case that failing to incorporate explicit price setting into GE is a failure to live up to methodological individualism.

But... but how is that a criticism, as such? Different models have different limitations. Pick any model, any-dox, and we can find something that's unsatisfactory.

In any case, much macro is done with the markup-over-marginal-cost imperfect competition framework.

Plus "neoclassical economics" is what gave us the monopoly pricing model and the Cournot and Bertrand models of oligopoly, among other things.

Anonymous said...

"But... but how is that a criticism, as such? Different models have different limitations. Pick any model, any-dox, and we can find something that's unsatisfactory."

Having a "model" which cannot explain how prices change (i.e., are set) is hardly a "limitation" when that model seeks to explain how price is determined. It is not "unsatisfactory," it is nonsense.

But, then, neo-classical economics is not a science...

YouNotSneaky! said...

If all you want is perfect competition then all you really need is a Bertrand set up with an undifferentiated good and only two price setting firms. The trouble starts when you want to do general equilibrium with many interacting markets. Here immediately obvious problems arise, not the least of which is the fact that different markets can have different price-setting structures. One way out of this - at a cost of course since there's always a cost - is setting it up as a monopolistic competition model that Gabriel mentions.

How are prices set in the Sraffian framework? It seems to me like the requirement that there is a uniform rate of profit, combined with CRS and a single non produced factor pretty much forces ('determines') prices to be what they are. There doesn't seem to be much price setting going on either. In fact, my impression from the way you tell it is that it's not that different from 'neoclassical' way - firms minimize costs by choosing amounts of factors given their prices.

The reason why this doesn't quite work in the 'neoclassical' GE setting is because those conditions above (CRS, single non produced factor, linear technology perhaps) aren't made. But you still need something to 'replace' them - you need excess demand to be zero in all markets simultaneously and that's where the auctioneer comes in.

Robert Vienneau said...

Gabriel tells me that Bertrand provided a model of oligopoly. YNS tells me that the Bertrand setting is a model of perfect competition.

I've already provided an example of a Sraffian model with more than one non-produced input. In the first paragraph of his book, Sraffa says, "no such assumption" of constant returns is made. I often assume Constant Returns to Scale because I generally emphasize an internal critique of so-called neoclassical theory, not Sraffa's reconstruction of the "submerged and forgotten" logic of classical economics. I don't know why YNS implies that Dorfman, Samuelson, and Solow (1958) did not present an example of neoclassical General Equilibrium theory.

YouNotSneaky! said...

"Gabriel tells me that Bertrand provided a model of oligopoly. YNS tells me that the Bertrand setting is a model of perfect competition."

The "Bertrand Model" can come in several different varieties. You should know what the difference is between them and I think the context here makes it clear which one I am talking about and which one Gabriel is talking about. And if you do know then you shouldn't act like there's some kind of contradiction here when there isn't.

And, perhaps mistakenly calling it 'Sraffian', I was referring to the no-substitution theorem, which, I believe does require CRS. I may very well be completely off on this one though. In fact I probably am. Which is why I asked the question - 'how are prices set in other, non-neoclassical (and I'm including standard models of monopoly and oligopoly in the neoclassical cannon) models? Who sets the prices when it's all about technology coefficients?

Gabriel M said...

Prices "change" as to always equate supply and demand. The end. Have a great evening, everyone! :-)

What, you want a "because" in there? I'll need to see a secret handshake first.

In any case, correct me if I'm wrong, but I'm not aware of any non-"neoclassical" models of price setting that are consistent with the basic notion of rationality (SARP? GARP?)