[OPE-L:2095] Re: Kliman-McGlone interpretation of the transformation problem

akliman@acl.nyit.edu (akliman@acl.nyit.edu)
Tue, 7 May 1996 00:14:53 -0700

[ show plain text ]

A reply to Riccardo's thoughtful questions in ope-l 2092. It was
a post "addressed" to Alan, but my name is being taken in vain ;)
so I'll take a stab at it.

Riccardo is right that if prices happen to be the stationary, simultaneist
prices, then all the TSS numbers (except for those dependent on the
"numeraire") are the same as the simultaneist/Sraffian numbers. And
it is also true that if profit rates continue to be equal period after
period, and there is no change in technology period after period, and
there is no change in the real wage period after period, and there is no
change in the amount of labor extracted from workers (length/intensity)
period after period, and if we hold some other relevant things constant
that matter like money capital tied up in production, [whew, take breath]
... THEN the TSS interpretation does imply that the simultaneist numbers
would prevail.

In other words, prices will converge to stationarity, the profit rate will
converge to the Sraffian rate, etc.

But I don't accept that this scenario is "equilibrium" in any meaningful
sense. The usual condition for equilibrium is understood to be a uniform
rate of profit. But as Ted and I show, a uniform profit rate, supplies
= demands, balanced reproduction are all possible even without stationary
prices (I mean by profit rate the rate of return on actual investement
here). And I know of no real economic process that even TENDS to produce
stationary prices in the same sense that capital mobility tends to level
out profit rates. (For instance, Okishio/Roemer *assume* stationary
prices even though they are considering technical change! They fail to
show, and it is almost impossible to believe, that if we keep getting
technical changes, adjustment to stationary prices will occur).

I also agree with the implicit point Riccardo seems to be making, which
is that under a stationary price "equilibrium," value theory is wholly
irrelevant. Evenbody in this case gets the same answers for all the
things that matter (realrtive prices, profit rate, etc.); the whole
wave of "transformation problem" solutions, as Samuelson rightly noted,
are really irrelevant (in the stationary framework)--a stationary price
model is one in which only RELATIVE prices matter, not the "normalization
condition" that gets used to create a price/value relation in the
aggregate. So yes, if prices were never to change, value theory would
be irrelevant (at least "quantitative" value theory).

But this doesn't mean that the TSS interpretation, or Marx, are WRONG
if prices happen to be stationary, as Riccardo at one point seems to
imply. Accept for the moment that our interpretation is right, and that
Marx is factually right. Then the profit rate *is* s/(c+v). If prices
happen to be stationary for some long period, etc., then the profit rate
will *happen to equal* the Sraffian/simultaneist rate (r). So in this
special instance s/(c+v) happens to equal r. Thus, the simultaneist
model is right in one particular special case, while Marx and the TSS
interpretation are right in this special case AND in all other cases.
(Given the assumptions of the 2d sentence of this paragraph.)

Of course, I don't think that equilibrium describes the real world, but I
do not reject equilibrium modeling/reasoning on that basis. David
Laibman, rightly in my view, characterizes such a tack as "crude empiricism."
Please see my EEA paper (1996) for a full discssion of the real reasons
I think a temporal conception of valuation is necessary (Riccardo has
this paper; I'll be glad to send it to anyone who doesn't; I think it
is also available from econ-value). Alan's view might be rather different
in this regard, also Mino's view, etc.

Finally, Riccardo's discussion of TSS input prices, the value of constant
capital, the value of variable capital, etc. is more or less correct. I
could quibble over some of the formulations, but the basic idea, that the
C and V depend on the PRICES, not values, of the means of production and
subsistence, is correct. I do NOT agree however, that this means they
differ from the labor embodied. I think the physicalist conception of
labor embodied is quite different from Marx's. I do NOT think that
for Marx, labor embodied can be translated univocally as "'labor needed
to reproduce the commodity under current conditions." Embodiment is a real
(though fetishistic and nonphysical) process in which living labor becomes
dead labor, "congealed labor," etc. Also, value is transferred from
means of production to the product, which takes TIME (this has been much
discussed, especially between Fred and I). ... Also in this regard,
Marx, even from the very outset (Ch. 8 of Vol. I of _Capital_, never
IDENTIFIES the value of constant and variable capital with the value of
means of production and subsistence (or labor-power). C and V are
DEFINED as SUMS of VALUE ADVANCED, and so MP and LP are, as he says 2 pages
from the end, simply different *forms of existence assumed* by the capital
when it enters production; they are not the capital itself (it has 2 forms
at this point, money, and physical). So, quite obviously, the constant
and variable capital advanced can differ, although the amount and value
of the MP and MS remain the same, if they are bought at changing prices.
And in general, the value of the capital is the value laid out for MP and
LP (which depend on the prices of MP and MS, and the quantities purchased)
and not the "value of" the MP and LP.

Yes, this does guarantee that there's no logical incoherence in Marx's
transformation, etc. And again, Marx's value theory--and ALL value
theory--is "redundant" if prices are always stationary. But otherwise
Marx's theory is neither redundant nor tautologically true.

If we have the price equation

P(t+1) = P(t)*(A+bL)(1+r[t])

(the usual Sraffian equation, except that input and output prices differ),
then, if we take the input prices as historically given (as Marx does),
there is ONE remaining unknown, r[t]. MARX "solves" the system by saying
the profit rate is determined in production, before outputs go to market,
so that r[t] = s[t]/{c[t]+v[t]}. But this is not tautological.
Neoclassicists will generally say r is zero, at least in equilibrium. And
there can be other theories as well. All will yield *different* predictions
for prices. But what is simply untenable, as a general theory, is to
wish value theory away by forcing input prices to equal output prices.

Andrew Kliman