Date: Wed, 12 Nov 1997 11:46:51 -0500 (EST)
From: JulioHuato@aol.com
To: andy@mira.net
Subject: Re: Great job

<< I really can't remember the issues in dispute now, that was in about 1983.
I have the articles in my file though.  >>

Where, specifically, Andy?  I browsed through your site, but no luck.

I'm sending you the Complexity book.  You may expect to receive it within a
couple of weeks.  In my opinion, the book provides a lite -- but fair --
presentation of the ideas about "complexity" developed by people now
connected to the Santa Fe Institute.  It's a short and very readable
story-telling book.

With regards to your attachment, unfortunately I don't know how to open it
up.  I assume it's a set of files bundled together as a MIME file.  I know
there's some shareware application out there to help me unbundle them and
all, but I don't know how to use it, even if I could find it.  Sorry, my
acquaintance with computers is post-Windows.  I suggest that you attach one
file per note.

<< I think the fractal mathematics and complexity stuff is probably much more
relevant to econometrics than Fourier analysis when it comes to cycles longer
than the short-term business cycle. However, the most important thing is to
distinguish the specific relations which determine the long-term trends from
the short-term dynamics. This cannot be solved by exptrapolation.>>

Besides the reference in the book "Complexity" and in a few other places, I'm
not familiar with fractal math at all.  I'll try to get acquainted.  Fourier
analysis is indeed widely used in econometrics (renamed as "spectral
analysis") and I took note of your comment about the invariance assumption.
 At work, I haven't done much time-series analysis.  Strictly speaking what I
do is closer to market research than it is to economics.  I deal mostly with
cross-sectional data.  

As to the critique of modern conventional economics, I don't think I'll be
unfair if I say that it has a relation to econometrics similar to that
between theoretical and experimental physics.  It is mostly an
axiomatic-deductive construction exposed in mathematical terms.  They use
econometric methods as when they make their models stochastic (eg, replacing
variables with probability distributions), but -- in general -- as an
approach to find economic regularities (induction) they despise econometrics.
 They believe the "axioms of choice" are sound enough to support the whole
edifice of their economics.

In the near future, I hope to expose to you my understanding of the new
classical economics school.  I gather that you have read Marx's Capital and
are familiar with the "value" question.  In an unpublished paper, I have
shown by reductio ad absurdum that the distinctive line of reasoning followed
by neoclassical economics leads necessarily to ... a labor theory of value.
 I believe that the best labor theory of value available is Marx's (you'll
say, "of course").  Unfortunately, my paper takes up the simplest
partial-equilibrium model.  I believe that my results (not formalized,
because I don't know how to formalize them and because I wrote it for a final
in a course of History of Economic Thought with Robert Heilbroner, who is not
into math) are generalizable to the general equilibrium case in its modern
version (as in Debreu).  

Now a days, conventional economics has assumed that the value question is
satisfactorily settled and moved on to refinements.  One of the reasons is,
probably, the undeniable eloquence and elegance of Debreu's axiomatization of
the (Walrasian) general equilibrium model.  You'll read about Gerard Debreu's
formalization of general equilibrium in "Complexity."

So, there is a vast amount of extensions derived from this tradition.  One of
the strands, which has taken up the "macroeconomic" questions posed by Keynes
and the monetarist critics, has become clearly (in my opinion) the dominant
school in modern economics in the US.  That's the "new classical economics"
school.  They are adamant about examining any macroeconomic problem in
question in a manner consistent with the "micro-foundations" laid out by
Debreu, et al.  Moreover, they refer to Debreu's static model only as a
modern evolutionary biologist would refer to Darwin's work, as a raw
precursor.  Their models are dynamic in the sense of having "agents" choosing
not one-time actions, but entire sequences of actions from a space of
functions that determine sequences of prices and quantities.

I'll leave it here.

J.