From: Allin Cottrell (cottrell@WFU.EDU)
Date: Sun Jun 04 2006 - 21:46:45 EDT
On Sat, 3 Jun 2006, Ian Wright wrote:
> Paul wrote
>> However, I am not convinced that this use of 'socially
>> necessary' corresponds at all to that which Marx meant
>> and as such I dont see that it resolves the transformation
>> problem...
>
> There are two different issues here:
> (i) whether the Sraffian or real-cost definition of labour-value more
> closely corresponds to what Marx meant, and
> (ii) which of the two definitions of labour-value is theoretically correct.
Agreed. On (i) I think it's worth emphasizing the following:
(1) Insofar as Marx was an economist (of course he was more than
that), his theory was firmly based on Ricardo's.
(2) Marx was highly critical of Ricardo on various points. In
respect of most of the specifically economic arguments at issue, I
believe Marx was wrong and Ricardo right, but that's another story
and I won't get into it here. My main point is that
(3) where Marx does _not_ explicitly criticize Ricardo, I think we
can take it that he agreed with R, and took his points as read;
and
(4) Ricardo provides a very clear verbal account of what one means
by the labour required "directly and indirectly" to produce a
given commodity (i.e. the vertically integrated labour
requirement). This is found in Chapter 1, Section III of the
Principles, under the heading "Not only the labour applied
immediately to commodities affect their value, but the labour also
which is bestowed on the implements, tools, and buildings, with
which such labour is assisted."
Ricardo includes (besides the direct labour) all the labour that
goes into the means of production (materials, productive equipment
of various kinds). There is no mention in this context of labour
going into the production of goods consumed by the capitalists who
finance the production of the commodity in question. (Such goods
are "inputs" into the "production" of the wellbeing of
capitalists; but this is, implicitly, a quite separate issue from
what is required to produce the commodity at issue.)
The whole historical discussion in Ricardo and Marx is
pre-Leontief, so there's no question of a precise mathematical
specification of "total labour content", but the formula
v = l(I-A)^{-1}
has always seemed to me the "natural" mathematization of what
Ricardo and Marx had in mind. (On a side point, Ian, you ascribe
this formula to the neo-Ricardian critics of Marx, but it has also
been assumed to be correct by many defenders of Marx -- Morishima
comes to mind, but there are many others.)
In respect of Ricardo, this judgement is strengthened by the way
he formulates the problem posed for the labour theory of value by
the (assumed) equalization of the rate of profit -- "The principle
that the quantity of labour bestowed on the production of
commodities regulates their relative value, considerably modified
by the employment of machinery and other fixed and durable
capital." Ricardo goes on to argue that the "modification" in
question is second-order and does not seriously compromise the
explanatory power of the labour theory of value. It is
nonetheless clearly posed as a "modification" of the principle --
a factor other than "the quantity of labour bestowed" that plays a
role in governing exchange value, rather than a sophisticated
recomputation of "labour bestowed".
> The real-cost definition of labour value certainly departs from
> Marx regarding the non-proportionality of prices and labour
> values. But only in the special case of self-replacing
> equilibrium.
Just trying to get your position right here: Am I right in
thinking that, on your view:
* Marx has prices non-proportional to values so long as the rate
of profit is equalized and the distribution of organic
composition is non-degenerate, but
* You have such non-proportionality only in case the system is
outside of self-replacing equilibrium?
> The Sraffian definition of labour value certainly departs from
> Marx regarding the aggregrate identities and the logical
> priority of labour values over prices. The existence of the
> transformation problem is substantial evidence that the Sraffian
> definition of labour value is not what Marx meant.
I think there's a significant ellipsis here. Sraffa's "definition
of value" cannot really be said to "depart from Marx" regarding
the aggregate identities, since those aggregate identities were
(supposedly) theorems, not axioms, in Marx. That is, Marx
believed that, starting from a certain definition of value, one
could deduce that the "aggregate identities" (total price equals
total value; total profit equals total surplus value) held. It is
surely possible that Sraffa's definition of value is compatible
with Marx's, but that Marx was wrong with regard to the
deductions.
Morever, Marx clearly felt that a "transformation problem"
existed. He thought he'd solved it (more or less -- modulo his
own comments on the risk of going wrong if one didn't transform
the inputs), but he definitely recognized a problem.
> With regards (ii):
> Regardless of (i) I believe that the real-cost definition of labour
> value is the correct one on theoretical grounds. The basic reason is
> simple: the modern, linear production definition of labour value is
> intended to "vertically integrate" over the cost structure in order to
> count all the direct and indirect labour required to produce
> commodities. The Sraffian definition of labour value does not
> vertically integrate over labour cost of money-capital.
This seems to be the key point. I'm having a lot of difficulty
with "labour cost of money-capital" as something distinct from the
labour embodied in the means of production. At first sight it
seems like a confusion of levels of analysis: money is just not a
input to production.
Ian, you present your argument as, primarily, a critique of
Sraffian economics. But I think that what you're arguing here has
to be classed as a substantial critique of Ricardo and Marx too,
and I think it might be clearer if you were to argue it that way.
True, your argument has the effect of, in a certain sense,
"vindicating" Marx over the transformation problem. Yet it
achieves this effect only by discarding the conception of the
labour time "required to produce" a commodity which he inherited
from Ricardo and which he gave every impression of upholding. (In
this particular respect I see it as running parallel to the TSS or
"Temporal Single System" approach, much discussed on this list in
years past.)
> Paul wrote
>> However, I think the real insights of his theory probably
>> apply to the problems of socialist planning.
>
> Maybe you could expand on this?
I would also welcome that. I see how Samuelson's "modified
values" (using a technology matrix augmented by the growth rate of
the working population and labour-saving technological change)
could have a role in planning. If Ian's argument is equivalent,
OK. But it seems to me that Ian's argument is not equivalent,
since it explcitly references capitalist consumption, and the rate
of profit may well be substantially greater than Samuelson's
growth factor.
Allin Cottrell
This archive was generated by hypermail 2.1.5 : Fri Jun 30 2006 - 00:00:03 EDT