Re: [OPE-L] price of production/supply price/value

From: Fred Moseley (fmoseley@MTHOLYOKE.EDU)
Date: Sun Jan 29 2006 - 10:13:59 EST

Hi Rakesh and Ian (and others),

I have been trying to find the time to participate in this interesting
discussion about the transformation problem, the MELT, and related issues.
I have some comments below in reply to Rakssh and Ian, and another comment
will follow in reply to Rakesh.

Thanks for the discussion and the stimulation.  I look forward to further

On Thu, 26 Jan 2006, Rakesh Bhandari wrote:

> > Ian wrote:
> >
> > I think this direction is in general right, but to nail down the
> > argument requires more. I too think that Marx's starting point is
> > counterfactual.
> >
> > Marx's chapter 9, Vol. 3 has an interesting argumentative structure.
> > It seems to go like this (although more than happy to be corrected by
> > others more well-versed in this issue):
> >
> > (i) Counterfactually, assume (a) the rates of surplus-value in sectors
> > are uniform and also (b) assume that prices exchange at values.
> >
> > (ii) Deduce from these assumptions that rates of profit are
> > non-uniform, which contradicts the principle of uniform returns to
> > capital invested.
> >
> > (iii) Conclude, then, that (a) and (b) are false; that is rates of
> > surplus-value are not uniform and there are price/value divergences.
> Ian and Jurriaan,
> the question I am getting at is simple: how can we speak of price/value
> divergences if price of production is itself a form of value? If price
> of production is a form of value, then how can a form of value diverge
> from value? Donkeys may be a form of animal, but
> we don't speak of donkeys diverging from animals.
> There seems to be a lot terminological confusion in Marx.
> What I am proposing is that we call simple price one form of value and price
> of production another form of value (neither of course is market
> price, so we can
> still speak of a divergence between value and price;
> Yours, Rakesh

Rakesh, I think you are on the right track here, but I don't think there
is terminological confusion in Marx on this issue, although it is
complicated.  Value does not just mean labor-time. Marx explains in
Chapter 1 that value has three aspects or dimensions:  substance,
magnitude, and form of appearance.  The substance of value is abstract
labor and the magnitude of value is SNLT.  And the forms of appearance of
value are money, price, price of production, etc.

So when Marx says that there is a divergence between value and price of
production, he means by value the simple form of appearance of value, or
the simple price of commodities, as derived in Section 3 of Chapter 1,
proportional to SNLT, but not SNLT itself.  In fact, when Marx discusses
the "value" of commodities in the rest of Capital, without further
qualification, he almost always means the form of appearance of value, or
the simple price of commodities, as evidenced by the numerical examples in
terms of money. (e.g. the "value" of 20 lbs. of  yarn in Chapter 7 is 30
shillings).  Price of production is a more complex form of appearance of
value, which also takes into account the equalization of profit rates.

So there is no confusion when Marx speaks about a divergence between the
value and the price of production of commodities. By "value" he means the
form of appearance of value, or the simple price of commodities, which I
think is what you are suggesting.

I hope this helps.  What do you think?

Ian again:

> > Surplus-value is redistributed to capitalists according to capital
> > invested, rather than the number of workers they employ in their
> > sector.
> >
> > The problem with this argument is step (iii): you can't conclude that
> > *both* (a) and (b) are false from the contradiction in (ii). Of
> > course, the fact that Marx doesn't transform cost prices, and lacks a
> > general equilibrium framework, is cause for further doubt regarding
> > the precise details of his argument, at least if we do the Bortkiewicz
> > thing and abstract from the dynamic process of the formation of the
> > general rate of profit.
> >
> > The problem with simply specifying a MELT to translate between prices
> > of production and values, as you are proposing, is that, according to
> > the neo-Ricardian critique, not all Marx's aggregates (total profit =
> > total surplus-value, total price = total value, and rate of profit =
> > value rate of profit) will hold. According to this formalisation of
> > the problem, there is no MELT that can do this, including Foley's. So
> > the quantitative connection between value and price accounting is
> > missing, and Marx's thesis of conservation of value in price fails.
> >
> > So that's why I think the argument requires more, even if I agree with
> > the direction you are taking it.

Ian, it is not exactly true that the rate of surplus-value changes.  The
transformation of surplus-value into average profit as a result of the
equalization of profit rates does not change the rate of surplus-value.
But it does mean that the ratio of average profit to variable capital for
each industry is different from the rate of surplus-value (i.e. the ratio
of surplus-value to variable capital).  Perhaps this is only a
terminological issue, but I think it is important to maintain the
distinction between the surplus-value produced in each industry and the
average profit received in that industry.

Marx's theory has an entirely different analytical framework than a
Sraffian general equilibrium framework.  Marx's analytical framework is
the circulation of money capital:

        M - C ... P ... C' - M'

This is very important, and I think not sufficiently emphasized and

The Sraffian general equilibrium framework takes as given the physical
quantities of inputs and outputs.  Marx's circulation of money capital
takes as given the initial quantity of money capital M advanced to
purchase means of production and labor-power, prior to production.  The
main question of Marx's theory is:  how is the initial given M transformed
into (M + dM) as a result of production?

In terms of circulating capital only, the initial given M is the cost
price.  According to Marx's theory, the cost price does not change in the
transformation of values into prices of production, because the SAME cost
price is taken as given in the determination of both values and prices of
production - the actual cost of producing the commodities.  It follows
from this interpretation that all of Marx's aggregate equalities are
always true simultaneously (they are not conditional equalities whose
satisfaction depends on the compositions of capital in different

This is obviously a big subject, which I can't do justice to here.
Attached is a recent working paper of mine (a "sympathetic critique of the
new interpretation"), which explains more fully my interpretation of
Marx's analytical framework and its relation to the transformation

As for the MELT, this is another difficult and important question, as you
emphasize.  I argue that, with commodity money, the MELT is determined by
the inverse of the value of gold (i.e. labor-time contained in a unit of
gold) (MELT = 1/Lg), and that the MELT does not change as a result of the
transformation of values into prices of production.  It follows from this
interpretation that Marx's laws of conservation of value and surplus-value
are maintained.  I have a recent paper on this subject ("Money and the
Transformation Problem") in a new book edited by me entitled "Marx's
Theory of Money:  Modern Appraisals" (Palgrave 2005) (it's a good
collection, I think, with chapters by the usual members of the ISMT, plus
Foley, Lapavitsas, Itoh, de Brunhoff and others).  I have also attached a
copy of this paper.

The case of non-commodity money (presumably capitalism today) is different
and more complicated.  I also have a recent working paper on this subject
(also attached), in which I extrapolate from Marx's treatment of fiat
money in his time, the end result of which is that the MELT =  (MV / L).
That is, the amount of money new-value produced by each hour of SNLT is
determined by the general relation in the economy as a whole between the
total quantity of money in circulation (adjusted for velocity) and the
total quantity of labor-time that has to be represented as money.

I would of course be very interested in any comments you and others might
wish to make about any of the above or the attached papers.


This archive was generated by hypermail 2.1.5 : Mon Jan 30 2006 - 00:00:02 EST