At 01:14 AM 2/1/00 -0500, you wrote:
>For those who are not in the know, the Coase article referred ask a simple
>question. If the market is the most efficient mechanism for allocation,
why do
>firms not use the market to allocate resources internally? The answer is that
>using the market costs. These costs are called transactions costs. Allocation
>will then be done by the method that is cheapest. I.e., by the market or by
>command. Transactions cost then determine what is allocated by the market and
>what is internal to the firm.
>The problem is that the reasoning is circular. It is a tautology. The problem
>has been given a name, but no demonstration, either logical or empirical, has
>been given. It sounds nice, "Oh yeah, transactions costs, that makes
sense." But
>it means nothing.
I don't think this is a sufficient criticism. There's nothing wrong with
tautology if one can use it to help with the understanding of something,
going beyond tautology. The physics equation Force = mass x acceleration is
tautological in that each two of the terms define the third. Somehow
physicists use it to get a lot of mileage. The neoclassical concept of
"economic rationality" is tautologically true, but it seems productive for
them. In my interpretation, Marx's "law of value" (a.k.a., "labor theory of
value") is tautologically true. But it helps us understand capitalism
better. Even Coasian transactions costs can help us understand the world
better, as shown by some of the later work of Douglass North.
The key thing is to avoid excessive reliance on tautology. That, however,
is what the Chicago-school does best, usually hiding it in a welter of math
and/or econometrics. The biggest logical circle is that of the MF. His
"positive economics" says that it doesn't matter how unrealistic one's
assumptions are as long as they predict well. So he assumes perfect
competition in markets (unless the gov't meddles). But then his empirical
work is poor, not really testing the assumption against alternative ones.
He then uses the "success" of his econometrics to validate his unrealistic
assumption and argues that the government shouldn't "meddle."
BTW, the transactions costs emphasized by Coase and the Coasoids are
nothing new. Classical economists were aware of them, as were Robinson and
Chamberlain when they developed theories of monopolistic competition. As
Doug points out, they are only important if one's theoretical base-line is
the silly Walrasian general equilibrium model (or hallucinogenic visions of
the Invisible Hand). Maybe the role of transactions costs have moved some
economists away from such silliness. But usually it doesn't do so unless
other "imperfections" (i.e., deviations of the real world from the ideal
forms) are brought in. Douglass North produces more interesting and
revealing results because he assumes "bounded rationality" and the like
along with transactions costs. He also is an economic historian, so that
the concern with the actual history of the US economy slowly pushed him
away from his Chicago-school economics.
Of course, weakening two or more of the hegemonic assumptions and
inductively bringing in reference to the real world gets one beyond the
journal article form.
Jim Devine [EMAIL PROTECTED] & http://liberalarts.lmu.edu/~jdevine
