Re: Article in the Chronicle of Higher Education

["Dr. Bruce R. McFarling" <Bruce.McFarling@xxxxxxxxxxxxxxxx>]

On Thu, 30 Jan 2003 13:29:58 -0500,
"Gunnar Tomasson" <gunnar.tomasson@xxxxxxxxxxx>
wrote:

>I conclude - and so advised Samuelson in the late 1970s -
>that the methodological pre-suppositions of his
>'Foundations of Economic Analysis', as reflected in
>the modus operandi of contemporary mainstream,
>monetarist, and PKT economists, are invalid.

Let me stress that I am not taking issue with
Gunnar's argument that:

>1. The last paragraph lays out the WHOLE case for
>the admissibility in principle of partial derivative
>analysis in general equilibrium reasoning and, more
>generally, the use of Comparative Statics; the
>capitalized words signal Samuelson's recognition that
>there exists NO CASE IN LOGIC for the exception which
>he claims to exist with "respect to limiting rates of
>change (partial derivatives)."

and therefore cannot be taking issue with the
following argument that:

>5.  As for the Laplacian construction itself, it
>was mirrored precisely in Leon Walras' statement
>late in life (reproduced in appendix to Don Patinkin's
>book on 'Money, Prices, etc.') to the effect that,
>insofar as real-world economies may be viewed as
>General Equilibrium Systems, they must be judged to
>be in general equilibrium AT ALL POINTS IN TIME.

To the extent that there is a contradiction
between general equilibrium analysis on the
one hand, and comparative statics and partial
equilibrium analysis on the other, that tells
us that one might have one, the other, neither,
but not both.  In itself the argument only
serves to undermine approaches such as
Samuelsonian economics or monetarism that
try to have it both ways at once.

However, since Post Keynesian economics *rejects*
General Equilibrium fantasies regarding the
economy, a contradiction between the requirements
of GE fantasies and some other forms of analysis
that some PK economists may use is of no import.
Certainly I have seen few clearer descriptions of
why GE modelling is beside the point than:

>2.  For, as underscored by the 19th century
>Laplacian general-equilibrium construction of
>Newtonian Universal Gravitational Mechanics,
>there is NO place for "external data or parameters"
>(let alone "changes" therein) in systems where
>"the relationships between our variables are
>strictly those of mutual interdependence."

If, as this argues, the GE system is casually
closed, that is enough reason in and of itself
to skip past it to the next approach on offer.