LONG LTV reply to Steve

Mon Aug 8 12:35:39 MDT 1994

Steve (and whoever else is still interested):

I appreciated the thorough explanation of your position on the
use-value/exchange-value dialectic and the LTV.

I agree that the use-value/exchange-value dialectic is invaluable
(so to speak) for marxist analysis -- indispensable for dealing
with issues of distribution such as realization of surplus,
cyclical crises, and others as well.  (But what do you mean by
the "transformation problem?)

I also agree with you emphatically when you argue that the idea
that the Tendency of the Rate of Profit to Fall necessarily
spells the self-destruction of capitalism and the inevitability
of socialism is wishful-thinking, a kind of shibboleth we should
learn to do without.  The TRPF is for the forseeable future so
severely counteracted a tendency that to count on it to produce
significant change would be foolish (if not suicidal).

But I *don't* agree that the use-value/exchange-value dialectic
and the LTV are incompatible.  Let me try to explain why:

1)  You raise the "transfer of value" question, but the use-
value/exchange-value dialectic applies as clearly to machines as
to labor power (albeit in different ways).

You are right, I think, to say that strictly speaking in the
passages you quote, Marx means that it is exchange-value and not
use-value that is being transferred.  The use-value of a machine
is simply its ability to transform material; *that* ability is
not transferred to buyers in the products processed by the
machine; it lasts the life-time of the machine, and the vanishes
abruptly.  (To suggest that some abstract use-value is
transmitted by the machine to products and thence to the buyers
is, I agree, a kind of occultism.)

If the machine's use-value is not "transmitted" to the product,
then its exchange-value is.  Now its exchange-value is based
(like every other commodity) on its cost (average socially
necessary) of production -- which has nothing to do with its
resulting use-value.  The accounting practice of "depreciation"
simply attempts to estimate how long a machine will last, and
distribute its exchange-value (the exchange-value it transfers to
its products) over that life-time; hence Marx's insistence that
"however useful a given... machine,... it cannot under any
circumstance add to the value of the product more than #150
[i.e., its cost of production]."   When he says "its use-value
has been completely consumed, and therefore its exchange-value
completely transferred to the product"** (Ibid, p. 197), he is
not expressing an equivalence, much less a cause -- but merely
asserting that, if depreciation were to actually correspond to
the actual value-transfer (instead of being an estimate), the day
the machine broke down for the last time would be the day the
last penny of the initial cost of the machine would be included
in the valuation of products.

What this analysis suggests, and I want to underscore, is that
exchange-value and use-value are in some sense incommensurable:
no matter how useful or productive a machine is in the physical
sense, it is still only worth (and only transfers) its original
cost of production.  To go one step further: the whole idea of
trying to assess the "transfer of value" from machine to product
derives from the status of the machine qua capital, and not from
its properties qua machine: depreciation is merely part of an
accounting system that tries to monetarize the return on
investment over the estimated life-time of the machine; the
physical productivity of the machine is, for better and for
worse, a different matter.  (Incidentally, an interesting eco-
marxism derives from this very incommensurability.)

So I would take the phrase from the Grundrisse that you quote

>It also has to be postulated (which was not done above) that
>**the use-value of the machine significantly (sic) greater than
>its value**; i.e. that its devaluation in the service of
>production is not proportional to its increasing effect on
>production." (p. 383. Emphasis added.)

to mean simply that the physical use-value of a machine greatly
exceeds the depreciation ("devaluation in the service of
production") assigned according to its original exchange-value.
Hence the immense productivity of capitalism, but also the fact
that this productivity is potentially independent of the social
relations that make machines into capital under capitalism.

2)  You treat *all* commodities on the model of labor-power,
which is a very peculiar one; what is valid for commodities in
general is valid for the commodity labor-power, but the reverse
is *not* true.

You claim that

>What this paragraph [Marx on depreciation] asserts is that the
>exchange-value of a machine sets an upper bound to its use-
>value; yet, earlier, Marx used the proposition that under the
>"eternal laws" of exchange in capitalism, the two are
>"completely different magnitudes": "The past labor that is
>embodied in the labor power, and the living labor that it can
>call into action; the daily cost of maintaining it, and its
>daily expenditure in work, are two totally different things..."

>"The daily cost of maintaining" a machine is its depreciation,
>and this is--over its lifetime--equivalent to its exchange-
>value. "its daily expenditure in work" is its value
>contribution, and in Marx's dialectic, this is its use-value. IF
>GENERAL ("The seller of labor power, like the seller of any
>other commodity, realizes its exchange-value, and parts with its
>use-value"), then in the case of machinery too, there should be
>a difference.

I have capitalized what I take to be a wrong move here: you have
in effect extrapolated "backwards" the peculiar properties of
labor-power as a commodity to *all* commodities, in this case to
machines as commodities.  But "the daily cost of maintaining a
machine" is an accounting fiction: it does not correspond in the
least to the "necessary labor" component of a worker's daily wage
(which he *must* spend on goods to live to work another day).
Similarly, a machine's "daily expenditure in work" is indeed its
use-value, but machines don't have the peculiarity that labor-
power does of producing more value than they consume -- because
machines don't consume value, by which I mean that they don't
purchase goods on the retail market the way workers do.  Labor-
power's use-value exceeds its exchange-value (or more precisely:
the exchange-value eventually realized on the use-value of labor-
power exceeds the exchange-value initially expended on that
labor-power).  There *is* a difference between the use-value
and exchange-value of machines; but it's not the *same*
difference as that between the use-value and exchange-value of

(I would therefore disagree with your statement that "in the case
of the inputs to production, use-value and exchange-value are
both objective and quantifiable" -- because (i) the use-value of
labor-power is never a mere a "input to production" under
capitalism; rather it is determinable only through the complete
dialectic of production->consumption/reproduction->production;
and because (ii) the use-value of labor-power is based on the
*difference* between necessary and surplus labor, and that
difference is a historical judgment, not a quantifiable fact.)

The passage you quote from Sweezy helps target one source of

>"It is clear that surplus-value cannot arise from the mere
>process of circulation of commodities... It seems equally
>obvious that the materials entering into the productive process
>cannot be a source of surplus-value. The value which the
>materials have at the outset is transferred to the product at
>the conclusion, but there is no reason to assume that they
>possess an occult power to expand their value... From the
>standpoint of value there is no reason to assume that either
>materials or machinery can ultimately transfer to the product
>more than they themselves contain. This leaves only one
>possibility, namely that labour-power must be the source of
>surplus-value." (Sweezy, Theory of Capitalist Development, pp.

The problem is that Sweezy talks (mistakenly) about the source of
surplus-value in terms of production inputs and outputs
("entering into the productive process... at the outset" vs "at
the conclusion"), whereas surplus-value derives from the
difference between necessary and surplus labor -- **and this
difference arises not from the production process alone but only
from complete cycles of production--consumption/reproduction--

You conclude from the Sweezy quote that

>all attempts to show that machinery is not productive of net
>value boil down to attributing "occult" powers to labor-power--
>in the sense that value becomes something peculiarly pertaining
>to labor, but at the same time not physical (since to be
>physically linked brings in the Sraffian critique of the
>transformation problem)--or reduce to the kind of negative
>methodology presumed above by Sweezy, and criticised by Bohm-

I agree that there is (or should be) nothing occult (or
"ontological") about labor-power as the source of surplus-value;
but there is in fact nothing "occult" about labor-power in my
view *precisely because* there is *nothing physical* about
"productive" labor-power, in the sense Marx defines it for
capitalism: whatever the quality/quantity of goods labor-power
produces, it is only "productive" when it produces surplus-value,
and it does so only because in the complete p->c/rp->p cycle, the
exchange-value realized on that labor-power at the "end" of "the"
cycle exceeds the exchange-value paid for it at the "beginning"
of "the" cycle.  This is, I would add, the only context in which
the use-value and the exchange-value of labor-power become
comparable. (To put the point another way: Marx has no interest
in the notion of "net value," only in the production of surplus

This is *not*, whatever Bohm-Bawerk may think, a "negative proof"
of the source of surplus-value in labor-power: it is rather a
positive, dialectical one -- dialectical in the sense that the
entire *movement* through the p->c/r->p cycle is required to
distinguish labor-power as the source of surplus-value.

In conclusion

I'm sorry if I sound like a broken record in my insistence on the
"critical" or "dialectical" nature of Marx's theory of the LTV.
But as I see it, there is no incompatibility of the use-
value/exchange-value dialectic and what I have called the
"critical" LTV.

I have found our exchange very stimulating, and if you have
published any articles along the lines of what you've said here,
I would appreciate full citations for the opportunity to examine
your arguments in greater detail.

     Gene Holland


More information about the Marxism mailing list