Re: Reply to Michael on productive labour & services

jurriaan bendien (
Tue, 20 Jan 1998 19:39:21 +0100

Michael writes:
> The question of the extent of real subordination, that may well be
> related to the nature of the use-value (although not in a way that
> can be mapped simply on the physical product/services distinction),
> seems a thin reed upon which to base this apparently fundamental
> distinction.

This seems to be merely a point-scoring remark by Michael. The point being
made is that some labour activities by nature, or contingently, do not lend
themselves to being transformed into capitalist commodity production, i.e.
be fully subsumed by the specifically capitalist mode of production. And I
would would place services such as hairdressing, cleaning, teaching,
prostitution, security guards etc. in that category. They can be
transformed into commercial or wage labour, but their product is not a
commodity other than perhaps some exceptional circumstances such as we
might imagine.

It is a matter of degree, related also to such
> contingent matters as the managerial mechanisms in play, the degree
> of competition in output and labour markets, etc, etc. In terms of the
> matter at hand, none of these factors will show up in NI accounts ...
> .
I don't see what managerial mechanisms and competition have to do with it,
but it is true that instances of expenditures on productive labour may be
difficult to extract from accounting information.

Jurriaan wrote:

> > In one passage, Marx writes very explicitly "for labour to be
designated as
> > productive, qualities are required which are utterly disconnected with
> > specific content of the labour, with its particular utility or the
> > use-value in which the labour is objectified. Hence labour with the
> > content can be either productive or unproductive" (p. 1044). However
> > this quote it does not logically follow that Marx means that the
> > content of the labour itself is always irrelevant.

Michael replies:

> In strict formal logic, we cannot make this as a deductive inference.
> Nevertheless, it is strongly suggested by this passage.

No. What Marx is referring to is the social form of the labour, the social
relations under which it is carried out, quite irrespective of the specific
content of the labour.

Michael writes:

It states
> quite clearly that characteristics quite unconnected with the
> use-value of the product are *required* to characterize productive
> labour. Thus use-value characteristics are at most necessary, but not
> sufficient.


Michael writes:

But then the last sentence rules out any necessity of
> use-value difference either. So use-value difference is neither
> sufficient nor necessary.
No it does not. Marx says a given kind of concrete labour may be productive
or unproductive, depending on the social form under which it is carried
out. That is, labour is not productive because of the kind of concrete
labour it is, but in virtue of the social form under which it is carried
out. But this is logically quite compatible with saying there are labour
activities which, in virtue of their specific nature, cannot assume the
specifically capitalist social form, i.e. which cannot be converted into
commodity production, although maybe they can be performed as wage-labour.

Jurriaan writes:

> >It merely means that the specific content of the concrete labour
> >(the specific use-value produced) must be related to the objective
> >of capitalist production for surplus-value.

Michael replies:

> There is in the passage Jurriaan has cited, *nothing* to support this
> assertion. On reflexion what does 'related to ... etc.' mean in this
> context?

Well this is this is the drift of, or an implication of, Marx's discussion
in the text. "Related to" means that whether or not labour is productive
for Marx is to be judged by whether the labour produces surplus-value.
This is the general criterion. However now we have to look (as Marx begins
to do) at particular types of concrete labour and their social form and
ask: does this labour produce surplus-value or not, and why ? Some forms
of labour are simply faux frais of production, a cost which must be met,
but which doesn't generate surplus-value, and therefore must be kept as low
as possible.
> Jurriaan goes on:
> >
> > In a subsequent passage, Marx says:
> >
> > "On the whole, types of work that are consumed as services and not in
> > products separable from the worker and hence not capable of existing as
> > commodities independently of him, but which are yet capable of being
> > directly exploited in capitalist terms, are of microscopic significance
> > when compared with the mass of capitalist production. They may
therefore be
> > entirely neglected, and can be dealt with under the category of
> > that is not at the same time productive labour" (p. 1044-45)
> > Of course we would say today that this type of work is not of
> > significance" any more, and it does need to be analysed.
Michael replies:

> This statement certainly asserts that services tend ('on the whole')
> not to be commodities,

Wrong. Marx means "on the whole these types of work are of microscopic
significance". But Marx does continue in this passage the distinction he
makes between production of services and commodity production.

and suggests that not all wage-labour need be
> productive. But it doesn't give any inkling why this should be the
> case, except in the statement 'and hence not capable of existing as
> commodities independently of him [the wage-labourer]'. But why does
> this characteristic of services stop them being commodities?

FROM SERVICES. Because of that quality, you can mass-produce commodities
but you cannot mass-produce services, at most you can train up lots of
people to perform the service.

Michael says:

> that most participants in this discussion appear to have have conceded
> that services may be commodities. (They are, of course, at liberty to
> reconsider their positions.)

Well if I always go with the majority, if I always go with the flow, I will
never get anywhere with understanding what Marx was on about. To obtain a
correct view certainly requires democratic discussion. But the truth
cannot be decided by democratic vote. To borrow an example from Alan
Freeman's physical science analogies, for many years the majority believed
the sun orbited the earth and the majority was wrong.

Jurriaan writes:
> And then:
> > And Marx explicitly defines a concept of a "service":
> >
> > "In general, we may say that service is merely an expression for the
> > particular use-value of labour where the latter is useful not as an
> > article, but as an activity. Do ut facias, facio ut facias, facio ut
> > do ut des [I give so that you may do, I do so that you may do, I do so
> > you may give, I give so that you may give] - all these are
> > formulae for the same situation, whereas in capitalist production the
do ut
> > facias expresses a highly specific relationship between material wealth
> > living labour". (P. 1047).
> This again seems to be quite definite that *no* services can be
capitalist commodities.

Correct. However, I am not denying that other passages can be quoted which
do suggest in some sense that services may be commodities. Therefore as I
have said Marx's concept of productive labour is not fully worked out and
it is not fully consistent.

This is what I cannot get my head
> round. Note however, Alan, that here Marx is characterising services
> *as use-values*. This does not sit easily with your notion that that
> labour which is unproduictive is so becuase it does *not* produce a
> use-value.

Well Alan sometimes uses the notion of use-value a bit loosely. So does
Marx. But for national accounts analysis we need to tighten up the
definitions, and that is one of the use-values of that exercise - a sort of
intellectual discipline.

> Let's try another tack. Marx's comments elsewhere often indicate
> that it is the organic whole of all the wage-labour socially
> necessary for the production of a commodity that produces value in
> the labour process producing the relevant commodity. So, *I* infer,
> the in-house service wage-labour of the security guard or maintenance
> worker, as part of the organic labourer, is potentially productive.
But if these services are contracted out, and produced as vendible
> services, as Marx does not permit these to be commodities, this
> worker ceases to be productive. This seems to me to be directly
> contrary to the intuition that employment by your 'own' capitalist
> facilitates both the impact of market forces, and the real
> subordination to ensure that workers actually work longer than
> required to reproduce the value of their wage.
In this respect, I refer Michael to a passage from Capitral Volume 2:

"If we have a function which, although in and for itself unproductive, is
nevertheless a necessary moment of reproduction, then when this is
transformed, through a division of labour, from the secondary activity of
many into the exclusive activity of a few, into their special business,
this does not change the character of the function itself" (Penguin ed.,
p. 209).

> Jurriaan writes:

> In the case of "non-material" production, Marx distinguishes between the
> > production of commodities separable from the producer (books, works of
> > and production where the product is not separable from the act of
> > (e.g. teaching). He says in these cases the capitalist mode of
> > is possible only on a limited scale, a formal subordination to capital
> > take place but not a real one. The reason is in fact the specific
> > of the labour involved.

Michael responds:
> I don't doubt Marx does say this. But what is it about teaching type
> services that resists commodification in a way qualitatively
> different from other sorts of wage labour? Surely at most it can only
> be a matter of degree?

The force of "resistance" in this case is the fact that teaching by nature
usually presumes an instantaneous and unmediated interaction between pupil
and teacher, i.e. the production and consumption of teaching coincides (in
the class room). However, it is nowadays technically possible to impose
the commodity form on teaching more and more - we can substitute TV
programmes, computer programmes, through which we begin to mass produce
teaching as a commodity. The "teaching factory" Marx jokes about as a case
of formal subsumption by capital is becoming a reality.

Michael writes:

> Many of you will know that I argue that the
> commodification of all labour-power is subject to the toing and
> froing of class struggle in a way that makes it problematical to
> simply assert that labour-power *is* a commodity like any other. I'm
> not sure if anyone on this list shares this view of labour power as
> an extremely 'peculiar' commodity. But it emphasises the notion that
> the difficulty of formally subordinating service-producing labour as
> opposed to material-product-producing labour is, at best, a matter
> of degree - insufficient to bear the weight that some seem to want to
> place upon the productive/unproductive labour distinction.

I haven't argued anything inconsistent with this view. The substantive
difference between us is that I argue that

(1) for Marx (commercialised) services are not commodities - concrete
labour is not a commodity, only labour power is a commodity.
(2) if they are not commodities, the production of services is not
commodity production.
(3) if production of services is not commodity production, it cannot
qualify for Marx as capitalistically productive labour.

> Thanks for finding those quotes, Jurriaan.

Thanks to you Michael for responding to this post.


Jurriaan Bendien.