[OPE-L:1135] Re: does price affect value

Duncan K Foley (dkf2@columbia.edu)
Tue, 20 Feb 1996 05:31:49 -0800

[ show plain text ]

Please forgive me for the length of this posting, but the issues seemed
important, and I wanted to be as explicit as possible.

I'd like to thank Andrew for sending me copies of two of his papers with
Ted McGlone. Their "A New Interpretation of Marx's Value Theory" (August,
1995) is a very clear and complete statement of their position on the
transformation problem and related issues. (It also includes some
discussion of the falling rate of profit and the Okishio Theorem, which I
won't comment on in this posting.)

In the early 1980s Gerard Dumenil and I (Alain Lipietz's paper was based
on Dumenil's work, which Lipietz cites) pointed out that the only way in
general to maintain the quantitative equivalence of surplus value and
unpaid labor time in the Marxian framework was to define the value of
labor-power as the product of the wage and what I called the "value of
money", which is the ratio of the total living (productive) labor time
expended over a period (say, a year) to the value added in the commodity
system in the same period, or, equivalently, as the wage share in value
added. This definition of the value of labor-power is consistent with any
theory of the determination of the value of labor-power, including the
view that the value of labor-power is determined by a given bundle of
workers' consumption. The value of money is a coefficient with dimension
labor time/$ (or other money unit) that permits us to translate between
labor time measures and money measures. We also noted that if you
multiply the money value of constant capital (assuming for the sake of
simplicity a unit turnover time, so that the flow of constant capital is
also the stock of capital tied up) by the value of money you get an
expression in labor time units (say, hours) which need not represent the
historical labor time embodied in the elements of constant capital nor
the labor time that would be required to reproduce the elements of
constant capital.

Kliman and McGlone (KM) implicitly adopt this conception of the value of
money (or its inverse, the "monetary expression of value") as the
coefficient that translates between money expressions and labor-time
expressions, as is clear from the fact that in their scheme the value
added, variable capital, and surplus value are conserved. I think it
would be desirable to acknowledge and emphasize this logical connection
between the Dumenil-Foley and Kliman-McGlone (and others) interpretations
of Marx's theory of value, not least on pedagogical grounds, since the
apparent multiplication of interpretations is confusing and disheartening
to students, and it would make them happier to realize that these two
approaches share a substantial common ground.

KM go a step further, and propose to define (or interpret) references to
constant capital measured in labor-time with the product of the money
price of the elements of constant capital by this same value of money.
This interpretation of constant capital then preserves the equality of
total (gross) value produced in money and labor-time terms and the
equality of the rate of profit calculated in labor-time and in money.

As far as logic goes, this is a consistent point of view, and a genuine
generalization of the case where market prices happen, for whatever
reason, to be proportional to embodied labor coefficients, and technology
is stationary, since in that case their constant capital will be equal to
the labor-time embodied in the elements of constant capital, and to the
labor-time required to reproduce the elements of constant capital as well.

The issue of "simultaneity", however, seems to be somewhat misleading in
this context, since their proposed interpretation could be applied
whether prices are stationary or not. They could claim reasonably that
the applicability of their interpretation to the case of non-stationary
prices is a formal advantage to it, but the basic idea seems to be quite
independent of assumptions about the stationarity of prices.

KM also state a case for regarding their interpretation as consistent
with Marx's ideas on this matter. Dumenil and I put forward similar
arguments in relation to the value of money conception which KM have also
adopted as part of their interpretation, and ran into similar questions
from critics. At this point in my own intellectual development, I rather
doubt that this issue can be resolved satisfactorily. The difficulty lies
in the fact that ch 9 of Volume III of Capital is a draft. It is very
hard for us to know whether Marx viewed the schemas that he put forward
there as a complete expression of his views or not. It seems to me quite
possible that Marx was only dimly aware of full ramifications of the
problems of valuation of constant capital when he wrote this chapter, and
assumed that they could be ironed out without actually thinking the
issues through as thoroughly as later commentators have done. A point in
KM's favor, as they assert, is that their interpretation coincides with
Marx's actual calculations in this chapter; there is, however,
substantial evidence from the chapter itself and other important texts
that Marx interpreted "value" as "labor embodied" which this
interpretation does not support (and no interpretation can, given the
mathematical constraints of the problem). These same criticisms, I must
say, apply to the interpretation of the value of labor power as the wage
multiplied by the value of money, though there are texts where Marx
pretty clearly adopts this point of view, too.

It seems to me, however, that the more important questions involve the
relation of Marx's value theory to the broad issues of political economy,
in particular his discovery that the source of profit lies in unpaid
labor, and his construction of a coherent theoretical edifice to explain
the phenomena of capitalist production consistently with this discovery,
for example, in the circuit of capital. Interpretations that use the
value of money in the sense of the ratio of living productive labor time
to value added allow us to analyse the real data of capitalist economies
directly in this framework. Where the "traditional" (Bortkiewicz, Seton,
Morishima, Roemer) interpretation gets off the track, it seems to me, is
in maintaining the embodied labor coefficient interpretation of "value"
and as a result throwing out a good deal of Marx's political economic