A very tired (despite the coffee) response

[pdavidso@xxxxxxxxxxxxxxxxxx]

The dialogue between myself and others must surely be reaching a high
degree of diminishing returns -- so I will end my contribution with this
last note -- and leave others to have the last word.
To Gonzalo: For something to be an axiom merely requires the model
builder to consider it as a "universal truth" not for you. For example,
the existence of God is a universal truth for a religious person and no
amount of "scientific evidence" will shake that person's belief in God.
For an atheist-scientist on the other hand the existence of God is not
"fundamentally self-evident".  In other words, self-evident universal
truths are in the eyes of the beholder -- not subject to testing by
Arrow, Debreu, McKenzie, Sonnenschein or even Fonseca.  Assumptions on
the other hand can be challenged.

Blanchard states "all the [New Keynesian] models we have seen impose
long-run neutrality of money as a maintained assumption. This is very much a
matter of faith, based on theoretical considerations rather than on
empirical evidence" (see O. Blanchard, "Why Does Money Affect Output?"
in HANDBOOK OF MONETARY ECONOMICS, VOL. 2, Edited by B. M. Friedman and
F. H. Hahn, p. 828). When a "maintained assumption" is "a matter of
faith" that cannot be tested by empirical evidence, that is an axiom in
my view.

 I have argued that a more general theory of EMPLOYMENT, interest, AND
MONEY (note the latter Ajit) would be one that does not require
neutral money in either the short-run or the long run. Nor does it
require any assumption about a specific degree of competition (see GT, p.
245).

Further Gonzalo, I did quote to you the basic axioms regarding production
and employment decisions in my last memo (from GT, p.47). You should have
noted it did not require specifying the number of commodities or markets
as you did in your previous communique. I really am not interested in all
the ancillary axioms of your neo-Walrasian economists" for one they
assume neutral money (e.g., only relative prices matter in real
decisions) or stochastic ergodicity or the ordering axiom of expected
utility theory, they are discussing very special cases -- and if they
offer additional "axioms" over neutral money and ordering axioms they are
developing even more specialized cases of the special case classical
analysis.
 Now do you really deny that the axiom that only relative prices matter in
real decisions and the ordering axiom are "universal truths" underlying
your  heroes "Arrow, Debreu, Mackenzie, Sonnenschein et AL". what
evidence would they require before they would drop the ordering axiom? or
that only relative prices matter?

I am arguing that one can develop a theory of employment, interest and
money without requiring the ordering axiom -- or that only relative
prices matter in real decisions. Is that your objection? what has
complexity to do with this? Or do you think you can hide ordering, etc
under complexity?

To Ajit- I am not interested in Ricardo neutral money system; I am
interested in a general theory of employment, interest and money.
There are an infinite number of ad hoc constraints that you can introduce
in a nonmonetary economy of the classical type (Ricardo was a classicist
Gonzalo?) to achieve an unemployment solution. All you need to do is to
presume fixed proportions production function between K (which Claudio
correctly asks what does K mean?) and L -- so that in the short-run no
change in the relative prices (all that are required in a neutral money
scheme) can put workers into employment in the short-run. In the
long-run, of course, the result will be to reduce the population of
workers until full employment is achieved.

To Barkley: Policy in a nonergodic environment involves developing (and
when necessary modifying and even dismantling) institutions so that they
act as a balancing wheel -- to offset socially undesirable movements
WHEN THEY OCCUR IN A NONERGODIC REAL WORLD. As Keynes noted Economists
should be humble persons -- like dentists (and perhaps unlike doctors?).
Dentists should not try to predict in advance what tooth shall decay --
rather they shall catch decay in early stage and restore full tooth
employment; Dentists should also advise us as to preventative dentistry
-- brush your teeth, floss,  two checkups a year, etc. but Dentists are
never surprised when the patient develops a cavity despite the
preventative schedule. As Shackle noted in EPISTEMICS AND ECONOMICS
economists should not try to predict what WILL happen, they should only
indicate that they can state what CAN happen under certain conditions --
a much more humble statement. (For example, Economists should not state
that if the price of commodity x is reduced by y percent total revenue
will increase by $z. Economists can say that if the price is reduced and
if the price elasticity is less than unity and if all other things --
including income, employment, etc are equal -- then total revenue will fall.

	To be concrete: what policy would you advocate for exchange rates
to close the US trade deficit?  I was asked this while testifying before
the Senator Bentsen's committee in the mid-1980s. Economists like Fred
Bergsten were predicting (on the basis of complex econometric models) that
if the dollar-yen ratio was allowed to slowly decline to approximately 95
yen to the dollar we
would achieve a soft-landing equilibrium trade balance with Japan -- and
probably with the rest of the world. Of course what Bergsten should have
been saying was that if my econometric estimates were drawn from a time
series which was essentially ergodic during the relevant period, and if
the future would remain in the same ergodic circumstances, then a 95 yen
rate, etc.  But of course, Bergsten merely assumed ergodicity.
	On the day I was testifying only myself and Les Thurow argued
that the 95-yen scenario would not wipe out even a significant portion of
the trade deficit. How could I make such a prediction in a nonergodic
world? It was a matter of looking at the actual figures and then using
some fundamental Keynes general theory.  For example, I noted that the
value of US net imports of petroleum products equaled approximately
55% of the total foreign deficit and hence would not be affected by  an
exchange rate decline since petroleum products were in dollars. I did
this for other major groups as well -- noting which imports came from
countries that were likely to peg their currency to the US dollar and
hence would not be affected by dollar devaluation.
Then I imposed the Marshall -Lerner condition in order to see what kind
of elasticity the remainder of the commodity and service groups would
have to have in order to close the balance of payments deficit. I also
then used Thirlwall's Law to see what income elasticities might be
required. I then made a judgment on the basis of the hypothetical
price and income elasticities that were required under the Marshall-Lerner
condition and Thirlwall's Law to eliminate the goods and services
payments imbalance and concluded there was no way that a 95 yen-dollar
could do it the next 4-5 years (the time frame Bergsten was using).
   Thus I could warn that permitting devaluation (or even encouraging i.e.,
talking down the dollar) was more likely to create problems than resolve
them. That is  good policy advice -- as the uncertain future as it became
the past proved. And it certainly wasn't policy nihilism! (Nor is my IMCU
proposal which is an institution that has certain rules that kick in to
act as balancing wheel when socially undesirable events occur. -- That,
in my view, is good theory and good policy design for a nonergodic world.)

Any more questions about policy design Barkley?
And now to get another cup of coffee.  Paul D.