[OPE-L:442] Re: Gil Skillman's unwanswered question

Alan Freeman (100042.617@compuserve.com)
Tue, 7 Nov 1995 13:18:43 -0800

[ show plain text ]

Re: Gil Skillman's unwanswered question

Also Re: Chaion Lee's short question; marginal and
average value contribution of labour

In "[OPE-L:339] Re: Chaion Lee's Short Question" Gil
asked a question which, in my opinion, has not been

"why isn't the value of every commodity determined by
marginal rather than average production conditions, as in
Marx's specification?"

If I were Gil, I would feel I had every right to say that
in a discussion with the Marxists, they had no answer to
his question. I think this would be a negative outcome.

Therefore, I want to try and deal with this question in a
different way, which may help to illustrate both an
alternative procedure to the present one, and a possible

This post deals with why I think the answers so far given
are unsatisfactory. These have been, essentially, of two
different types:

One response is that either empirically, or for some
practical reason, average rather than marginal production
conditions determine the value of a commodity. I think
this is a reasonable interpretation of Steve K's
argument, but I stand to be corrected. Paul C, among
other arguments, also seems to me to make same case which
intertwines with his position that observations for
several countries by a number of different workers seem
to suggest that in practice the value and the price of a
commodity are not numerically far apart.

In summary this response says that Marx uses the average
because this is 'true' in that it corresponds to the
observed facts.

Another strand takes as given that the magnitude of value
is given by the labour time socially necessary to produce
a commodity (SNLT) and addresses the question: is it
logically coherent to maintain that this SNLT is the
average, rather than the marginal labour time required to
produce it? That is, the question addressed is not 'is it
true?' but 'is it consistent?'

This relates to a genuine controversy in Marxism. Even
accepting labour time as the measure of value, there
remain (at least) two possible definitions of the value
added by an hour of labour time which are both consistent
with the term 'socially necessary', namely:

(i) the value added by the *most efficient* worker (in
some sense) in a given sector

(ii) the value added by the *average* of workers in the
industry at a given moment

Chapter 3 of Makoto's book 'Value and Crisis' explains, I
think quite clearly, the arguments related to this
question. It is slightly more complicated than might be
supposed: for example, for Marx the marginal producer is
not necessarily the most efficient. In the case of the
understocked market (VolIII/ChX) the marginal producers
are the least efficient, because high prices give rise to
such high profits that even inefficient producers can
raise output profitably. The problem Marx was trying to
deal with is the interaction between the formation of
value and the operation of supply and demand. I do not
propose to review these arguments here though, if Makoto
is able, I think he could usefully assist us by doing so.
Maybe Dave Kristiansen, who has worked on this, should be
considered as a list member.

My point here is not to say we should settle this debate.
My point is trying to settle the debate, or perhaps rerun
it, cannot answer Gil's question.

This is because Gil does not accept that labour is the
magnitude of value. Therefore, the debate assumes what
Gil needs to have proved. Sure, *if* you accept that
labour is the measure of value, then it's an important
discussion. But Gil doesn't accept this. What he has done
is say 'hey guys, there's a hole in your argument'.

He has made no secret of the fact that he does not agree
that the magnitude of value is determined by labour time
in any way at all. I do not see, therefore, how it can
settle the argument to determine which of the two options
given above is most coherent since, in effect, Gil
rejects both solutions and is only observing that there
is a contradiction between the two. Since he doesn't
accept the premises, I can't see that any conclusion
drawn from the premises can be adequate. It seems to me
that we have to put the case for the premises also.

We might conceivably establish a consistency argument.
But the issue is very complex and many prominent Marxists
have studied it and and haven't yet reached agreement. It
is a source of ongoing work, I think very important work.
We could end up saying that Gil should wait for his answer
till the work is complete.

Moreover I think it is fair to say that Gil has not
yet been given an answer based on consistency, although
in reviewing a very large number of diverse contributions
it is always possible I have missed something.

This leaves us with the arguments that the 'average' case
is in some sense 'true' observably. I think these are
quite vulnerable and the following point should be
considered: if value *in fact* differs from price, that
is, if calculations to this effect are in fact wrong,
then by attempting to refute theoretical arguments about
the magnitude of value by immediate reference to
empirical facts, we could be digging quite a large hole
both for ourselves and for the view that labour time is
the magnitude of value.

For, suppose we refute Gil's argument the basis of
empirical observations. Then, if these observations fall,
the whole argument falls. I think the argument that price
is empirically close to value falls in this category. We
know at least *theoretically* there are many cases where
price differs quite substantially from value, and Marx
was clearly exercised to investigate them: the best known
is the influence of rent and the formation of prices of
production, but more generally he is always at pains to
emphasis that price is not only conceptually but actually
different from value.

Indeed in my view the key distinction in Marx, which
exists from the beginning of Volume I onwards, is the
distinction between value and *market price* of which
price of production is but a special case. And market
price differs from value merely through the influence of
supply and demand. Marx made this the centre of his
polemic with Proudhon and continues the theme in the

IMHO from the moment Marx realised, in about 1847, that
price *had* to differ from value in a market economy, his
main problematic was already established, namely to show,
in spite of this, that exploitation was *not* the result
of these supply-demand deviations as Proudhon maintained.
This he did in Volume I, as he explains, by abstracting
from price-value deviations and demonstrating that
exploitation still occurs.

I therefore think, incidentally, that Gil's
characterisation of Volume I as a discussion of 'pure
commodity exchange' is unfounded and derives from a
misapprehension started by Engels and turned by Sweezy
into the foundation of modern academic Marxism, that
Volume I represents a mythical society in which goods
exchange at values. Actually, in very few places in
Volume I does Marx introduce this assumption. It is
mainly to be found in Volume II. In Volume I Marx
abstracts from price-value difference, which is a
different thing: it means he confines himself to
assertions which are true whether or not prices differ
from values.

There are therefore two problems with the 'empirical'
refutation of Gil. If prices are in fact the same as
values empirically, and if we can identify the
theoretical reasons for this, then we must conclude that
Marx made a big mistake and should have ignored price-
value deviations. But if prices are not the same as
values, either empirically or theoretically, then Gil has
received no answer and can rightly announce to the world
that the there is no defence of Marx on this question [if
by 'there is no defence of Marx' we mean 'the Marxists
have not defended him']

I find both outcomes unsatisfactory, and I don't think
Gil has had an answer. What in my view is required is a
reply to Gil which identifies first the common ground
which we have and, starting from this common ground,
draws out the logical reasons why it is preferable to
identify labour time as the measure of value, and, on the
basis of this logical argument, to establish why average
rather than marginal labour time serves better as the
measure of value.

In deference to objections to long posts, I shall try to
deal with this in a subsequent posting.

In deference also I have sent the citations from OPE
referred to in this posting separately.