[OPE-L:6533] Re: Learning beyond "static methods"

From: Steve Keen (s.keen@uws.edu.au)
Date: Thu Feb 07 2002 - 14:45:11 EST


Hi Alejandro,

Yes, Sraffians should also drop statics. This was the point of my 1998 ROPE 
paper "Answers (and questions) for Sraffians (and Kaleckians)", which 
basically showed that Steedman's 1992 critique of the Kaleckian theory of 
markup pricing was invalid in a dynamic setting.

As for why most economists of all persuasions are reluctant to abandon 
statics, I'd come down to about 3 reasons: inertia/familiarity, apparent 
definitiveness, and ideology.

Economics began using static tools--Ricardo's comparative advantage is 
entirely static, neoclassical theory ignores time, etc.--and this habit is 
maintained by our teaching systems (and the absence of any experimental 
feedback [such as applies in sciences] that could show that model outcomes 
aren't static ones).

Static answers are also definitive: "monopolies reduce welfare...", "the 
equilibrium price vector is..." whereas dynamics is far less prescriptive. 
The reliance on statics has kept the profession in Laplace's grand conceit 
that the future of the world can be predicted, whereas sciences have long 
since moved past that view to a far more humble statement of science's 
possibilities.

And of course, for neoclassicals in particular, statics--by presuming that 
everything happens in equilibrium--neatly hides the theory's ideological 
content in its mathematics. If you assume equilibrium in a spot market 
system to begin with, and endow that equilibrium with all sorts of 
normative conditions--utility maximisation, cost minimisation, etc.--then 
you are promulgating the view that the market system is perfection, without 
even being aware of ideology.

My own escape came courtesy of being introduced to Rostow's model of 
economic growth as a school student (and loving the nascent dynamics of 
it), doing maths outside economics and thus learning of differential 
equations independent of the appalling tuition economists receive in 
"quantitative methods", and being introduced to the holes in neoclassical 
economics as an undergraduate (specifically the theory of the second best). 
With that melange I was aware that dynamics was vital to doing economics 
properly.

Hicks stuffed up dynamics big-time!

His so-called dynamic model of cycles in output:

Y[t]= (1-s+c) Y[t-1] -c Y[t-2]

was supposedly derived by discretising Harrod's model:

G=s/c

But is actually based on an economic error. Harrod was attempting to 
provide a dynamic form of Keynes's model, which began with the convention 
that actual savings equals actual investment. Hicks's equation was derived 
by equating actual savings to *intended* investment!

Since both savings and intended investment were described as functions of 
output, his model answers the question "what level of output guarantees 
that actual savings and intended investment will be equal, if both are 
functions of output?" The answer is "zero output", but the question is 
nonsense.

So economists spent 20 years "doing" dynamics using a model which was 
nonsense as the basis of their instruction. Hicks, well-meaning and highly 
intelligent though he was, did more damage to economic logic than anyone 
other than Samuelson and Friedman.

A great reference on dynamic methods in general is Martin Braun's *Ordinary 
Differential Equations and their applications*, Springer-Verlag (I have the 
1993 edition, but there's bound to be later ones now). It's a superbly 
written maths text--reads like a novel often, rather than a text--with a 
competely self-contained introduction to differential equations.

To go beyond that, Ott's book on chaotic dynamics is excellent.

As for dynamics within economics, still the best book is Blatt's (reference 
in earlier post). It's out of print, but should be available in large city 
university libraries.

I'm not too fond of much else in the economics literature--I hate 
Gandalfo's text, for example--but things may be changing. I've just 
received Medio and Lines' "Nonlinear dynamics: a primer"(CUP 2001) and on a 
quick flick through, it looks pretty good. I'd also suggest Puu's books 
(1997 & 2000 Springer-verlag), not because I've read them but because of 
Puu's reputation.

Finally, I must confess that I haven't read the work you cite by Duncan! 
Maybe when I can find the time...

Cheers,
Steve
At 05:00 AM 8/02/2002 Friday, you wrote:
>Re Steve K 6530:
>
>Thanks for your interesting post and the references. Some set of questions:
>
>[...]
>
>You write:
>
>"Neoclassicals would also have to drop static methods and learn about
>differential equations, etc--something none of them show any real
>inclination to do."
>
>1. Pressumably, this would be the case of the Sraffians too, wouldn't it?
>How do you explain this situation? Why are most of the people so reluctant
>to abandon "static methods"? How did you personally escape from this?
>But, wasn't Hicks (a "neoclassical") interested in the kind of dynamics
>you're describing?
>
>2. Can you provide us with a list of references, both from Math and
>Economics books and articles, so that one can learn --from the most basic
>to the complex stuff-- about non "static methods"? What would be a "ideal
>program" for a course(s) to have a "non equilibrium centered" approach to
>the capitalist reality? Is there anything analogous to "Kurz & Salvadori"
>in the field you are describing?
>
>3. What do you think about Duncan's Money, Accumulation adn Crises (1986)
>and his "convolutions" approach to the capital circuit?
>
>A.



This archive was generated by hypermail 2b30 : Sat Mar 02 2002 - 00:00:04 EST