Below is the first draft of a paper I will be submitting for discussion at the
International Working Group in Value Theory mini-conference at the upcoming
EEA Convention. It is basically a response to Professor Foley's paper. It is
too long to transmit whole, so I am sending it in parts. Part I is below.
Part II is being sent together with it. I hope to be able to re-format the
rest later today or tomorrow, and then post it. Comments are of course
welcome.
Andrew Kliman
A CONTRIBUTION TO THE ONGOING INQUIRY INTO THE EXISTENCE OF MARX'S MARXISM
In commemoration of the 150th anniversary of the _Communist Manifesto_, which
put forth the vision of "an association, in which the free development of each
is the condition for the free development of all."
Andrew J. Kliman, March 18, 1997
First Draft. Not for circulation outside of ope-l.
Professor Duncan Foley's "Recent Developments in the Labor Theory of Value"
argues that the temporal single-system (TSS) interpretation of the
"quantitative" dimension of Marx's value theory "seems counter to the Marxist
interpretation of the labor theory of value" (1997, p. 28).[1] It is held to
be an inauthentic interpretation because it does not equate living labor
extracted to the "value of the net product," and because it treats dead and
living labor as homogeneous. Following the methodological lead of Bortkiewicz
(1952), the author claims to correct these errors and, on this basis,
dismisses my claim to have refuted the Okishio theorem.
This paper will respond to these criticisms. It will suggest that value added
as Marx used the term is not the same as the value of the net product, and it
will show that the contrary view implies either that surplus-labor is not the
source of all profit, or that commodity values cannot change in the aggregate.
It will argue that, whether or not the TSS interpretation is authentic, TSS
critiques of the Okishio theorem have refuted it. It will suggest that
Professor Foley's conception of dead and living labor as heterogeneous (labors
of different "vintages" produce different amounts of value) has a marked
advantage over traditional simultaneism (interpretations in which the prices
and values of inputs and outputs are determined simultaneously): it arrives
at "Sraffian" conclusions, but avoids the internal inconsistency of
traditional simultaneism. Yet the vintage labor notion, too, will be shown to
imply that commodities' values are constant, and that the monetary expression
of *value* is transformed into a monetary expression of *use-value*. After
presenting textual evidence showing that Marx held that commodities' values
were variable, and that the use-value productivity of labor does not affect
the value produced by it, the paper will conclude with some comments on the
significance of the new value controversy. I will suggest that the various
different versions of Marxian value theory should all have the right to
contend in the public space, and that unless it is proven to be
self-contradictory, this right should extend to Marx's own value theory (which
may or may not be authentically represented by the TSS interpretation).
Value Added vs. "Value of the Net Product"
In his critique of my refutation of the Okishio theorem (Kliman 1996a),
Professor Foley (p. 28) notes that my representation of value determination in
Marx's theory does not equate the "value of the net product" (VNP), evidently
measured here in labor-time, with the living labor expended. He charges that
this "seems counter to the Marxist interpretation of the labor theory of
value," evidently because it permits new value to be created other than by the
extraction of living labor in capitalist production.
It is correct that my equation does not equate the VNP to living labor, but
incorrect that it departs from the conception that living labor is the source
of all value added. The reason is that, according to the temporal
single-system (TSS) interpretation of the quantitative dimension of Marx's
value theory, the category of value added *as it appears in his work* differs
from the VNP. Value added is the difference between the value of output and
the actual amount of value used up in producing it, while the VNP is the
difference between the value of output and the end-of-period cost of replacing
the inputs that were used up. When prices are constant, the two are equal;
otherwise they are not. Thus in my representation of value determination in
Marx's theory, living labor was indeed the source of all new value, the value
added, as Marx himself held. But since value added is unequal to the VNP when
prices change, so is living labor.
Professor Foley's paper uses the concepts of value added and VNP
interchangeably: "the expenditure of living labor adds money value to the
inputs …. The value added in production [is] the net domestic (or national)
product. … the monetary expression of labor [is] the ratio of the net
domestic product at current prices to the living productive labor expended [p.
18]." He evidently believes that this accords with Marx, for he claims that
my distinction between value added and VNP is contrary to "the Marxist"
theory.
Yet Professor Foley seems either to have changed his mind or to be uncertain.
In Foley (1982, p. 38), he provided a general definition of value added (and
of the VNP!) as "the money value of the commodities consumed plus the money
value of the change in stocks." It is this very concept that he now
criticizes TSS work for adopting.[2] He now contends that it holds only in "a
model economy with stationary technology" (p. 27), which implies that his 1982
definitions refer only to a special case, so that it is impossible in general
directly "to operationalize the concepts of the labor theory of value in terms
of the income statements of capitalist firms" (Foley 1982, p. 42).
Likewise, in Foley (1986 p. 44, emphases added) he not only provides what I
consider to be a correct interpretation of Marx's concept of value added but
also justifies it well:
"The capitalist advances capital both to buy labor-power and to buy nonlabor
means of production …. The value of nonlabor means of production *appears
unchanged* in the price of the finished commodity. … Marx calls the capital
advanced for nonlabor means of production *constant capital*, because it *does
not expand* in the process of production ….
"For example, suppose that in a certain year an average capitalist firm
*spent* $100 million on nonlabor inputs to production … and $50 million was
spent on the wages of production workers. If the firm sold its finished
commodities for $200 million, we would view $100 million of that total price
as a *recovery of the costs* of nonlabor inputs, or constant capital, $50
million as the equivalent of the wages paid, or variable capital, and $50
million as surplus value. The *value added* would be $200 million less $100
million *purchased* inputs, or $100 million. Thus Marx expresses the total
price of commodities as:
c + v + s"
This is correct. This is indeed how Marx reasons, without regard to whether
prices are changing, and without regard to how economists would later
construct the national income and product accounts on the basis of Keynesian
concepts. This passage is incompatible with Professor Foley's new critique of
my refutation of the Okishio theorem and with the identification of value
added and the VNP. If, for instance, prices fall during the period, the
end-of-period replacement prices of the nonlabor inputs may total $90 million
instead of $100 million, so that the VNP is $200 million - $90 million = $110
million. It is also incompatible with all simultaneist interpretations of
Marx's value theory, interpretations which hold that the prices (values) of
inputs and outputs are determined simultaneously, or that they must be equal.
It is the infamous TSS interpretation.[3]
Rather than working with replacement prices, Professor Foley here understands
constant capital as the amount actually "spent," used to "purchase" nonlabor
inputs, an amount "advanced." This sum of value "does not expand" or contract
but reappears "unchanged" in the price of the output. Note, moreover, his
wonderfully succinct justification of this concept: "$100 million of that
total price [is] a recovery of the costs of nonlabor inputs, or constant
capital."
Precisely. When commodities are sold at value, capitalists exactly recover
the full costs incurred in their production, including the portion of cost for
which no equivalent was paid, surplus-value. Thus the surplus-value is the
difference between value and the paid portion of costs that has been recovered
through sale. "The excess of the total value of the product over the sum of
the values of its constituent elements is the excess of the capital which has
been valorized over *the value of the capital originally the advanced*" (Marx
1977, p. 317, emphasis added). It is otherwise with the replacement cost of
nonlabor inputs. This sum does not recover the costs of nonlabor inputs if
values have changed. If values have fallen, for instance, the value of the
capital needed to replace them is less than the sum of value originally
advanced for them, so that the portion of the sale price of outputs that
represents replacement cost is insufficient to recover the costs that were
incurred in acquiring them.
The Quintessence of Simultaneism
The opposite holds when values, or prices, or both, are rising, and it is
instructive to examine this case. Let us assume that the figures above are
economy-wide figures, that 2 million hours of productive living labor were
extracted during the year, and that, at the end of the year, $150 million
would be required to replace the nonlabor inputs used up. Thus, their cost
has risen 50%, on average, due to a nominal rise in the monetary expression of
their value, or due to declining productivity and thus a rise in their real
value, or due to some combination of the two. This is an extreme example, but
clearly within the experience of many countries. The VNP in money terms
equals $200 million minus $150 million, or $50 million, and dividing this
figure by the living labor, we obtain what Professor Foley considers to be the
MEL: $50 million/2 million hrs. = $25/hr. Hence, the $50 million spent on
wages represents 2 million labor-hours. The workers' work of 2 million
labor-hours exactly replaced this sum. No surplus-labor was extracted. From
where has the $50 million in surplus-value come? I don't know, but it
clearly has not come from the exploitation of other people at the point of
production. **The replacement cost interpretation of constant capital and its
companion, the concept of "value added" as identical to the VNP, thus imply
that surplus-value can be created by means other than the extraction of
surplus-labor in capitalist production.** One who was less tired than I of
being labeled orthodox and dogmatic might wonder whether these notions may
perhaps be the ones that run "counter to the Marxist interpretation of the
labor theory of value." [4]
It is of course possible that the $50 million of monetary surplus-value in
this example is due entirely to a purely *nominal* rise in values. In other
words, were the monetary expression of value to have remained constant
throughout the year, it is possible that the total money value of the output
would have been $150 million instead of $200 million, so that the real
surplus-value equals zero. This would make sense of the computations above,
since the surplus-labor extracted was zero.
Yet if even a bit of the amount by which the total money value of output
exceeds $150 million is due to a *real* rise in commodity values, brought
about by declining use-value productivity of labor, then the computations
above lead to a discrepancy between surplus-labor and the real magnitude of
surplus-value. Hence, the case of a purely nominal rise in values is the
*only* case in which the New Interpretation concepts and procedures (as
expressed in Professor Foley's latest paper) can be relied on to yield a
surplus-labor figure that accords with the real surplus-value. And thus,
every calculation made on the basis of the equation of the value of constant
capital with the replacement cost of means of production, and the equation of
value added with the VNP, implicitly but necessarily presupposes that *all*
increases and decreases in the aggregate money price of a constant "basket" of
the outputs represent purely nominal increases or decreases in commodity
values. Putting the same thing more simply, these concepts imply that a
constant "basket" of the outputs always has the same *real* value, so that, on
average at least, **the unit values of commodities can never rise or fall, no
matter how much the amount of labor needed to produce them changes**.
This is the quintessence of the simultaneism: a "labor theory of value" in
which value is not determined by labor-time. Changes in aggregate value must
mirror perfectly changes in the amount of use-value. The contradiction
between value and use-value that we encounter on the first page of *Capital*,
and follow through to its culmination in the law of the tendential fall in the
profit rate, has been made to vanish. Later in this paper, we shall study
another instance in which the constancy of unit values, a heretofore
well-hidden presupposition of simultaneism, has reared its head in Professor
Foley's new critique of the TSS interpretation.
Notes to Part I:
1. Subsequent references to this work will be indicated only by pages numbers
enclosed in parentheses.
2. Actually, I know of no proponent of the TSS interpretation who interprets
value added in this manner, except when the monetary expression of value is
constant. If the latter is the case or, equivalently, when value is measured
in labor-time, then the following is generally accepted: in the absence of
fixed capital, the difference between total value and the sum of value
advanced for means of production is the value added by living labor, and if
uses of value are limited to consumption and investment, this sum will be
divided between value consumed and value the change in value of stocks.
3. Subject to the caveat that the monetary expression of value hasn't changed
during the year. See note 2.
4. Professor Foley of course does not state that the equality of living
labor with the value of output net of expenditures on nonlabor inputs is
counter to "the Marxist" theory. Rather, he writes (p. 28) that I attribute
"the changes in the value of inventories and fixed capital due to price
change[s] to the living labor expended," which does sound like a quite
unseemly thing to do. Yet the two statements are equivalent.