A reply to one issue addressed in Duncan's 3-parter, Alan's 4-parter, and
Duncan's response (ope-l 2915) of this evening. There is obviously more that
I need to and want to say, but the following can be discussed in independence
from the rest, and it is all I have time for before leaving town (and the
list) for the long weekend.
I think it is *the* single most crucial issue under discussion at this moment.
Duncan's latest post states: 'I think that Okishio and certainly Roemer do
make the equivalent of (E) explicit in their definition of the "general rate
of profit".' "(E)" stands for the following postulate (which I've edited
slightly for concision): "an input/output economy [considered] in a state
where input prices are equal to output prices and profit rates are equalised"
both before and after "a viable technique is introduced into the economy."
I *think* Duncan's statement indicates that we have come to a very important
agreement in this discussion. That is, *if* Okishio or Roemer state (E) or the
equivalent explicitly in their major presentations of the "Okishio Theorem,"
then the theorem has not been refuted in the narrow sense. However, *if* they
do not state (E) or the equivalent explicitly, then the theorem has been
refuted, even in the narrow sense.
To be perfectly clear, it is best to state the following explicitly (!):
(i) "Refutation in the narrow sense" means a demonstration that the
conclusions concerning the direction of change in "the profit rate" or "the
equilibrium profit rate" do not necessarily follow from the explicitly stated
premises of the theorem.
(ii) An "equivalent" of (E) will include some statement that equality of
input and output prices is being invoked as a postulate or premise or
assumption, or some statement that, by postulate, premise, assumption, or
definition, "profit rate," "equilibrium rate of profit," or "general profit
rate," etc. is restricted to refer only to cases in which input and output
prices are equal.
(iii) A statement such as "the equilibrium profit rate will be given by p =
p(A+bl)(1+r)" could be either a definition of r or a claim of a derived
result. Therefore, taken alone, such statements do not constitute explicit
definitions.
I am definitely prepared to accept this. It seems to me that Alan and Duncan
are as well. Yes?
If so, then this is a tremendous step forward in clarifying the issues. We
will then have a simple, unambiguous empirical test of my claim to have
refuted the theorem in the narrow sense (though John came first, of course).
Upon presenting a statement of (E), or the equivalent, made by Okishio or
Roemer, the presenter will have proven my claim to be false. I will retract
it publicly, explicitly, in print (if I'm allowed). Otherwise, a valid
counterexample to the theorem has been put forth, and this should be
acknowledged.
Andrew Kliman