Keen's thesis

James Miller jamiller at
Sun Oct 22 15:23:08 MDT 1995


   This is a continuation of my previous commentary on Steve Keen's
thesis: "The Demise of the Labor Theory of Value."
   One general comment: Keen spends a lot of time reviewing the
ideas of various Marxists: Hilferding, Rosdolsky, Sweezy, Meek, etc.,
whose analyses of the relation between use value and exchange
value have little or no bearing on the particular criticism he has of
Marx. This series of selections does in fact deal with questions
surrounding the relationship of use value to exchange value in
Marx's work. But none of these writers raises any kind of serious
challenge to Marx's labor theory of value. Steve, on the other hand,
has launched a frontal assault on this central axiom of Marxism.
Rosdolsky, Meek, etc., were trying to supplement Marxism with
additional insights. Steve, on the other hand, aims to prove Marx
false at the very core.
   One of Steve's claims, repeated several times, is that there is
a contradiction in Marx's writing. At one point (when considering
simple circulation) Marx is supposed to have claimed that use
value and value are "unrelated," and at another point (when
considering the circuit of capital) that they are interrelated.
   Some examples of Steve's assertions (from Chapter: Use-
Value in Marx's Economics, Section: Capital, Subsection: The
Second Strand): "The lack of a relation between labor's
quantitative use value and its quantitative exchange value
means that the two will normally be different, and this difference
was the source of surplus value. ...The quantitative difference
between the use value and exchange value of labor power is
the direct consequence of Marx's initial classical proposition that
use value and exchange value are unrelated."
   Steve's feeling that Marx had stated initially that the two were
"unrelated" results from a misunderstanding of Marx's conception
of the relatedness of use value and exchange value. Marx never
said they were "unrelated," even in simple circulation. One has to
pay closer attention to what Marx said.
   Steve quotes Marx in _CPE_, p. 20, "use value in this indifference
to the nature of its economic destination, i.e. use-value as such
lies outside the sphere of political economy. It falls within the sphere
of the latter only in so far as it forms its own economic destination."
Here we note that Marx sees a specific type of relation between use
value and exchange value. Note that he speaks of "use value as
such," not just "use value."
   In _Grundrisse_, Marx said that, "as we have already seen in
several instances, nothing is therefore more erroneous than to
assert that the distinction between use value and exchange value,
which falls outside the characteristic economic form in simple
circulation, to the extent that it is realized there, falls outside it
in general." (p. 646) Marx never maintained that use value and
exchange value were "unrelated," but of course he defined these
two categories in such a way that students of his work would be
able to make the distinction, and, on this basis, understand how
they can be interrelated.
   The topics that can be discussed under the heading of "specific
ways in which use value and exchange value interact" include
"the use value of labor power," "the use value of the means of
production" and "the use value of the precious metals," to name
a few. And the specific ways in which use value affects exchange
value become more critical in capitalist production than before,
but these processes develop without obliterating the original
distinction between use value and exchange value. There is no
need to belabor this point. In any case, after arguing that Marx
initially held the view that use value and exchange value were
"unrelated," Steve showed how Marx held that the two are in fact
related in specific ways, and provided copious documentation of
Marx's views on the subject.
   But Steve's point has nothing to do with Marx's views on these
specific relationships. What Steve does is to ignore the analyses of
Marx on the subject under consideration, and advance his own claim
that the labor theory of value is incorrect.
   Leading up to his major conclusions (Chapter: The Sources of Value,
Section: Introduction), Steve points to the capital-labor relation, the
hiring of labor power by the capitalist, and calls attention to the fact
that the equivalent of the value of the means of subsistence could
be produced by the wage worker in a given portion of the working
day, say 6 hours. "However," Steve says, "the use value of that hired
labor was that it could be put to work producing commodities (which
could later be sold to realise exchange value). This use value
manifested itself as a number of hours of labor, the length of the
working day (say 12 hours). There is an obvious difference between
these two amounts, consonant with the general rule that use value
and exchange value 'bear no relation to each other', so that when both
are quantitative they will normally be different. This difference between
the laborer's exchange value and his use value was a source of
surplus value, which is the foundation of profit."
   Note Steve's use of the so-called "general rule" that "use value and
exchange value bear no relation to each other." In the first place, this
"general rule" doesn't exist. Steve himself has already gone to great
lengths to prove that this "general rule" doesn't exist. But if there were
such a "general rule," it would make no sense to say that "when both
are quantitative they will normally be different." What Steve means is
that the qualitative difference between use value and exchange value
has been eliminated, and they can now be measured according to a
common scale. In other words, what was once considered
incommensurate is now regarded as commensurate. And there has
to be a quantitative difference between the two formerly
incommensurate elements-- and why? Because of the former
incommensurability itself--the "general rule," a "general rule" that
Steve has already repudiated.
   On the face of it, the logic doesn't work. Even if it were true that
the incommensurability between use value and exchange value had
been overcome, it would not follow that there would be a quantitative
difference between them. Reference to the "general rule" that they
"bear no relation to each other" does not give any information as to
whether there would be any quantitative difference, or how much it
would be.
   But, of course, the main problem with Steve's attempt at a logical
proof of the incorrectness of the labor theory of value is not so much
the inadequacy of the logic itself. The problem is that his procedure
bears no relation to the reality of the world. He does not define use
value and exchange value scientifically. He does not treat them as
categories which reflect actually existing material and social
processes. He treats them as abstractions which can be
manipulated at will independently of the real world.
   Every day in the real world, millions of wage laborers go to work
and create exchange value for the purpose of the enrichment of
the capitalists. This is the world that Marx, and Marxists, try to
analyze. We begin with this world of capitalist social relations, and
try to explain it. Steve has forgotten about this world, and has come
to the conclusion that "use value" creates value. Not labor.
   Use value and exchange value are incommensurable, as Marx
explained in _Capital_. As for the use value of labor power, it
is no different. Marx pointed out, "the use of a commodity belongs
to its purchaser, and the seller of labor power, by giving his labor,
does no more, in reality, than part with the use value he has sold.
>From the instant he steps into the workshop, the use value of his
labor power and therefore also its use, which is labor, belongs to
the capitalist. By the purchase of labor power, the capitalist
incorporates labor, as a living agent of fermentation, into the
lifeless constituents of the product, which also belong to him."
(_Capital_, V. I, p. 292, 1977 Vintage ed.)
   When the capitalist purchases labor power, he is not interested
in the use value of labor power, as such; rather he is interested
in labor, which is not synonymous with the use value of labor
power, but is its function.
   The function of a use value is not the same as the use value
itself. If I buy an automobile, it is not because I want an
automobile as such. Rather, I am interested in the function of
the automobile, which is transportation. I buy it so that I may
be able to go from point A to point B. The car is the use value,
which I have no choice but to utilize, and to consume, if I wish
to enjoy the function of that use value: transportation.
   Likewise, the use value of labor power cannot be confused
with the function of that use value: labor. Nor can use value
and its function be measured in the same units. An automobile
can be measured in terms of weight, volume, chemical
composition of its components, etc. Its function (transportation)
however, is measured in miles. The two categories of the
use value, and the function of the use value, are conceptually
and mathematically incommensurable.
   The use value of labor power is defined as the capacity to
create exchange value. But this exchange value, created in the
labor process, cannot exist apart from the use value in which it
is materialized, i.e. the product of labor. Thus the capitalist has
no choice but to assess the fitness, or suitability, of any labor
power by measuring (at least in a provisional way) the skills,
discipline and knowledge of the worker in a particular line of work.
Since there cannot be any production apart from the production
of particular use values, likewise there cannot be any workers
apart from those who produce these particular use values.
   Thus these workers have to be prepared, by their physical
attributes, their temperament, their knowledge and experience,
to be able to produce specific kinds of commodities under the
specific cirumstances prevailing in the trade or industry at the
   When the capitalist hires labor power, he takes all these
specific potentials into account, together with more general
capabilities, such as language skills, strength, literacy, health,
etc., in assessing, or measuring, the fitness, or suitability, of
the worker for the work to be performed. This is how the
capitalist evaluates the use value of labor power.
   This assessment process implemented by the capitalist
is undertaken so as to assure that the worker will be able
to produce specific use values in the most efficient manner
possible. Only in this way will the capaitalist be able to tell
whether the worker will be able to work quickly and efficiently,
with a minimum of mistakes and wasted motion, so that the
output of use values from the worker's labor will be as high
as possible. Only in this way can the capitalist assure that
the exchange value garnered from the worker's labor will
be maximized.
   The assessment, or measuring, of the use value of labor
power is thus a physical, mental and technical process. As
such, the "units" of calculation of the use value of labor
power are incommensurable with the units of calculation
of the labor itself (hours). Just as the automobile cannot
be measured in the same units as transportation, so the
use value of labor power cannot be measured in the same
units as its function (labor).
   If this analysis is correct, then there is no point in
commenting on the rest of Steve's thesis, since his
premise is disproved, and anything he may derive from
this premise is thus discounted in advance.
   All that remains is to respond to the arguments Steve
made in his post of Oct. 20, responding to my first
criticism of his thesis.
   Steve quotes Marx from the criticism of Wagner, that
"surplus value itself is derived from a specific and
exclusive use-value of labor power." Steve then goes on:
"Marx then on numerous occasions states that this
use-value of labor power is a quantitative thing, whereas
in general--when it is irrelevant to political economy--
use-value is qualitative."
   It is not true that use values as such cannot be
quantified. It is just that they are quantified in units which
are incommensurable with value units. On the second page
of _Capital_, Chap. I, Marx says, "when treating of use-value,
we always assume to be dealing with definite quantities,
such as dozens of watches, yards of linen, or tons of iron."
   However, when dealing with the world of commodites, we
do divide it up, in the first place, into two counterposed
realities: exchange value, which is homogeneous throughout
this world, and is purely quantitative, on the one hand; and
use value, which immediately presents itself as divided into
many qualitative features, corresponding to the uses to which
these manifold products are put, on the other. But this, of
course, does not imply that the use values are not quantified.
Each is quantified according to its own physical nature.
   Steve continues: "Marx's genius was to see that, in the
case of production, the use-value of inputs was quantitative
--but that the general rule of the incommensurability of use-
value and exchange value continued to apply."
   We have seen that the use-value of inputs are indeed
quantitative, as outlined above. But in the case of labor
power, the qualities that it possesses as a use value are
measured in units that are still incommensurable with
value units; just as much as in the case of iron, whose
use value is measured in tons, while its exchange value
is measured in value terms (hours of labor time, or units
of money).
   Finally, Steve says that, "in the case of commodities
consumed as part of the circuit of commodites, C--M--C,
this incommensurability simply means that, to Marx, the
use-value of a given commodity plays no role in
determining the exchange-value of that same commodity.
But in the case of commodities consumed as part of the
circuit of capital, M--C--M, that incommensurability is
between two quantitative things--and hence all you can
say is that they will be different."
   There are two areas of concern expressed in this quote.
One of them deals with differences in the relation between
use value and exchange value which arise in capitalist
production that are not present in simple commodity
production. It would take a long time to fully develop this
theme, and I won't do that work now. It's not necessary
for the purposes of this discussion anyway. But these
differences are real, although not as absolutely so as
Steve conceives them. Even in simple commodity
production, instances can be found wherein use value
impinges upon exchange value, altering it, but this
analysis lies beyond the scope of this discussion.
   However, the problem lies in Steve's phrase, "that
incommensurability is between two quantitative things."
As I have indicated, Steve fails to scientifically
conceptualize the "incommensurability" he purports
to discuss, because he fails to distinguish the use value
of labor power from its function: labor.
   One final comment: If the tone of my criticism of
Steve's thesis appears to be somewhat sharp at times,
that is not because I bear any animosity towards Steve.
In so far as I can judge him from his posts on this list,
he appears to be remarkably congenial and is certainly
not lacking in scholarly qualities.
   But I am forced to conclude that Steve's attack on
Marx's theory of value cannot be disentangled from the
bourgeois campaigns to discredit Marx; specifically,
those strains of anti-Marxism which originate from within
the Marxist tradition. Thus I feel that there is a class
tension between Steve and myself, and it is that which
accounts for what may appear to be elements of
sharpness in my critique.

Jim Miller

     --- from list marxism at ---


More information about the Marxism mailing list