Keynes deductive and inductive GT

["RICHARD P.F. HOLT" <holtri@xxxxxxxxxxxxx>]

>From Paul Davidson:


Dear Ric:

In his GT, Keynes used both the deductive and inductive methods in
developing his general theory.
Keynes certainly believed in the deductive method. As the Lynn T.
quote from the german edition to the GT indicates Keynes called his
theory a general theory because it required LESS axioms to  reach
its deductive conclusions. The classical theory can then be shown
to be a special deductive case that imposes additional restrictive
axioms on the General theory.

Once Keynes could show that the classical case was not a general
theory, then he could turn to induction. As he writes on p. 3 of
the GT "the CHARACTERISTICS of the special case assumed by the
classical theory happen NOT TO BE THOSE of the economic society IN
WHICH WE LIVE  , WITH THE RESULT THAT ITS TEACHING IS MISLEADING
AND DISASTROUS IF WE ATTEMPT TO APPLY IT TO THE FACTS OF
EXPERIENCE" (emphasis added}.

On pp.23-26 (middle of page) via equations (2.1) to {2.8) of my
book POST KEYNESIAN MACROECONOMIC THEORY, I demonstrate how Keynes
can logically show that his principle of effective demand is more
general (requiring less assumptions) than the classical case. In
the middle of p. 26 I then note that "the next logical task for
Keynes was to demonstrate that" the special classical case required
special characteristics that was in conflict with what Keynes
called "the facts of experience" no matter what the degree of
competition (i.e., even if real world competition was perfect). On
pp. 26-27 I briefly explain how Keynes could show that these
special characteristics would never match the facts of experience
even in perfect competition.

Keynes was therefore able to LOGICALLY demonstrate that as a
GENERAL THEORY unemployment equilibrium could exist in both the
short and long run.  Unemployment was not merely due to adding a
special assumption onto a general theory as some old classical
theorists and many mainstream theorists claim, i.e., it could not
be attributed to assuming the restrictive claim of non-flexible
prices, or nonlinear dynamic systems on a more general theory.
These latter faults could explain unemployment in a classical
system -- but Keynes insisted that such a system did not capture
the ESSENTIAL PROPERTIES of money and interest which were the
"facts of experience" WHICH WOULD CAUSE THE POSSIBILITY OF
UNEMPLOYMENT EQUILIBRIUM TO EXIST WHETHER THE SYSTEM WAS LINEAR OR
NOT AND WHETHER PRICES WERE FLEXIBLE OR NOT. Unemployment was not
due to some grand market failure as Mankiw claims.

Thus Keynes used BOTH the deductive and inductive aspects of
logical -- but at least he was logically consistent in his use of
these in developing a model that was general and could a case that
matched our inductive view of the real world.

I have since developed my pp. 23-26 argument further and will be
presenting the latest in a paper entitle "Setting The Record
Straight" in an Eastern Economic Association Session on Sunday
March 19, at 11AM (organized by Barkley Rosser with papers by
Barkley and David Colander as well).  In this further development
I can show that the mainstream fable that Keynes's analysis applies
only to a demand-constrained regime and the classical case applies
to a supply-constrained regime is logically incompatible with
Keynes's GT. That is why Keynes used his general theory to write
HOW TO PAY FOR THE WAR rather than apply the classical theory to
the supply-constrained war time economy of England. Moreover this
paper explains why Tobin's brand of Keynesianism (as Tobin states
it in recent articles in the JEP (1993) and EEJ (1992)) are not
compatible with Keynes's logical system AT ALL! But if you want to
hear more about that you will have to attend the session at the
EEA.

(Some may remember that over a year ago I had a big discussion with
Barkley, Jim Devine and others on this deductive analysis of less
vs more axioms. Barkley and  others were not interested in counting
axioms they want to immediately to "cut to the chase" of induction.
Who cares what is the axiomatic foundation of theory as long as one
can present a model that, from hindsight, seems to closely track
patterns we seem to observe in the real world.

Thus, we heard that the world is too complex for simple linear
equations, we can find complex chaotic systems that seem to
replicate patterns that we observe in the real world so why do we
care as to what axioms they "realistic" pattern followers are based
on, etc.-- Of course, classical theorists could also show
unemployment could occur if prices were not flexible -- and we all
know that prices are not flexible in the real world. (In fact Tobin
has taken this argument to the exteme to argue that it takes some
"real time" to change prices, ergo in the real world there can not
be perfectly flexible prices -- just like there can not be a
perfect vacumn where a feather and a stone will fall at the same
speed.
Nevertheless, the non flexibility of prices would be an additional
assumption added on to what classical theorists claim is the
general theory.  Classical theorists (old and new classical as well
as Keynesian) take the axioms of the Walrasian, arrow-Debreu world
as the general theory and then try to ASSUME  additional ad hoc
constraints (market failures) to explain real world unemployment.)