Re: (OPE-L) recent references on 'problem' of money commodity?

From: Fred Moseley (fmoseley@MTHOLYOKE.EDU)
Date: Sun Nov 14 2004 - 10:07:56 EST

Hi Riccardo, thanks for your message.  My responses below.

On Thu, 11 Nov 2004, Riccardo Bellofiore wrote:

> (i) I am happy that you agree with me that your interpretation rested
> on the value of money (the MELT) being given ex ante, and that this
> in Marx holds because of his theory of money as a commodity: I always
> urged you either to endorse it or to show how the same thesis could
> be maintained with non-commodity money;

Riccardo, I have been arguing for years that Marx's labor theory of value
requires that the MELT be determined independently of P, and that this is
more difficult today because money is no longer a commodity, even as the
measure of value.

As I have said many times before, Marx's LTV can be expressed abstractly
by the equation:

(1)     P   =   (MELT) L

In this theory, the MELT and the L (or SNLT) are determined independently
of P, and jointly determine P.  This is the basic logic of Marx's labor
theory of price.

With commodity money, the MELT is determined by the inverse of the
labor-time required to produce a unit of the money commodity.  I think I
have shown in my recent working paper how the MELT can be determined
independently of P with non-commodity money, as discussed in my previous

You (along with the "new interpretation") seem to think that, since money
is no longer a commodity, the MELT can no longer be determined
independently of P, and thus the MELT can no longer determine P.  P is
just taken as given, and used to determine the MELT, instead of the other
way around.

> (ii) your first attempt at the conference you are publishing went
> towards the quantity theory of money (ground: Marx's own approach
> with inconvertible paper money). Value of money being, roughly,
> MV/SNLT. But SNLT is "socially" eventually on the market, so you
> cannot have it fully ex ante, but only ideally or latently (in you is
> different since you have a kind of Says's Law.

Here we have a basic disagreement, which we have discussed many times,
including on OPEL.

You seem to argue in effect that Marx's theory of price cannot be
expressed by equation (1) above, even with commodity money, because SNLT
CANNOT BE DETERMINED INDEPENDENTLY OF P.  Therefore, it does not really
make any difference if the MELT is determined independently of P, because
SNLT cannot be determined independently of P in either case.

But then, if P is not determined according to equation (1), then how do
you think P is determined?  It seems to me that you are left with no
quantitative theory of P at all.  Rather, the P's are simply taken as

The determination of SNLT is a somewhat complicated matter, because there
are TWO DIFFERENT MEANINGS of SNLT in Marx's theory (as discussed
on OPEL in the past).  These two meanings of SNLT are complimentary, not
contradictory, but we need to clearly distinguish between them in our

The two meanings of SNLT are:

1.  SNLT1 is the social average labor-time required to produce a unit of a
commodity (or a given amount of a commodity), measured in hours of simple
average labor.  This social average labor-time is determined in
production, by the conditions of production in each industry,
independently of exchange.  This is clearly how Marx defined SNLT in
Sections 1 and 2 of Chapter 1 of Volume 1, and this definition of SNLT is
maintained throughout the three volumes of Capital.  This is what I mean
by L in equation (1) above.

In Volumes 1 and 2 of Capital, SNLT1 is assumed to determine the LONG-RUN
EQUILIBRIUM PRICES of commodities, under that assumption that S =
D.  Marx's theory of prices in Capital is almost entirely about the
long-run equilibrium prices of commodities, and is hardly at all about
actual market prices (except for a few brief discussions), which are
affected by temporary imbalances between S and D.  Marx mentioned a number
of times that the subject of market prices belongs to a "later book on
competition", at a lower level of abstraction.

2.  SNLT2 is the total labor-time required in an industry to produce the
quantity of the commodity demanded by society; i.e. the total labor-time
in an industry that is "socially necessary", in the sense that there is
demand for the commodity at the long-run equilibrium
price.  Quantitatively:

        SNLT2   =   Q (SNLT1)

where Q is the equilibrium quantity of the commodity.

Q depends on demand, and thus so does SNLT2.  Since the demand for a
commodity is not known prior to exchange, SNLT2 is not known prior to
exchange (i.e. whether or not the actual labor counted as "socially
necessary" is not known prior to exchange.  However, demand does not
affect SNLT1, which is determined solely by the average conditions of

Riccardo, when you say that "SNLT cannot be determined independently of
exchange", do you mean SNLT1 of SNLT2?  If you mean SNLT2, then I agree
completely.  However, if you mean SNLT1, then I disagree strongly, and
with lots of textual evidence to support my interpretation.  So, which
SNLT do you mean?

> I have the opposite,
> firms in  Vol. I produce for effective demand, like in Keynes; but
> this is not the leter of Marx, of course).

I am not sure what you mean here.  Do you mean that Marx did not assume
that S = D, or do you mean something else?  As discussed above, I think it
is very clear that Marx systematically assumed that S = D throughout the
three volumes of Capital, in order to explain long-run equilibrium prices
(like Smith and Ricardo before him, in this respect).  Not because he
believed in Say's Law (which he criticized severely), but because he
mainly wanted to explain the total surplus-value and long-run equilibrium
prices, and for these purposes, temporary imbalances between S and D are
irrelevant and an unnecessary complication.

> And MV is of course for
> the quantity *identity*, shared by all, = PY, so again your value of
> money is NOT independent of P. The Ps are already there.

It is true that MV = PY, by definition.  However, that does not mean that
M is "not independent of P".  The relation of causation between M and P
could go either way.  And, for Marx, the relation of causation between M
P, indirectly through the determination of the MELT) than in the case of
commodity money (P determines M).  Building on Marx's analysis of
inconvertible paper money, I argue that also in the case of modern credit
money, M determines P, indirectly through the determination of the MELT.

Riccardo, which direction of causation between M and P with modern credit
money are you suggesting?

You say below that "you, as me, GO FROM M TO P" (emphasis added).  What do
you mean by this?  Do you mean that M DETERMINES P?  If so, do you mean
indirectly through the determination of the MELT (as I do)?  If not (for
either of these questions), then what do you mean that you "go from M to

> (iii) I do not see how this (something partly quantity *theory*
> oriented) can be accepted, since rather I would like to pursue Marx's
> hints towards endogenous money in vol III;
> (iv) this is actually what I do with the help of the circuit
> approach. Now you, as me, go from M to P. I have, thanks to the
> circuitist approach (bank finance), a completely endogenous view. And
> you?

As we discussed last summer, I agree that M is at least partly endogenous,
and depends on such endogenous factors as the rate of capital accumulation
and the rate of interest.  However, I also think that the government has
at least some control over the quantity of money.  So the quantity of
money is at least partly exogenous, in this sense of limited government

It seems to me that this is a small difference between us.  Whether the
quantity of money is entirely endogenous, or partly exogenous, we both
seem to agree that with modern credit money, M determines P (i.e. we "go
from M to P).

> (v)  My view of the MELT is different from the NI strictly speaking,
> because I go back to bank finance at the beginning of the circuit,
> this in turn is linked to wage bargaining, and this again is linked
> to *expected* magnitudes (expected effective demand, expected real
> wages, expected exploitation in the labour process). This gives way
> to production and actual exploitation.

I don't understand how the MELT at the end of the circulation of capital
in determined, in your view.  I thought that you have said in previous
publications that you agree with the "new interpretation" in this
respect.  If not, then how do you determine the end-of-circuit MELT?

> (vi) you say I or the NI have no quantity theory of price. Why? There
> is a quantitative determination of prices in NI. I agree with you
> that that kind of simultaneous determination has not much to do with
> the labour *theory* of value. Exactly: the LTV has to do with the
> *constitution* of these magnitudes, that is with what is going on in
> the buying and selling of labour power and in the process of
> production.

What exactly do you mean by the "CONSTITUTION" of these magnitudes?
Is it the same as, or something different from, the theoretical
determination of these magnitudes?

> This is what *I* have stressed all the time. The point is
> usually disregarded by all Marxists who take as granted that the new
> value is the monetary representation of nothing but labour. But why?
> This is the *theory* of value. With the monetary ante-validation of
> production, and the conflictual extraction of living labour as ideal,
> I have an answer to this. This answer is that the methods of
> production from which the determination of prices starts are
> constituted by the dynamics of money as finance and exploitation as
> the use of a potentially counter-productive labour power. The
> judgement that this is not a theory is unfair. It is not *your*
> theory. It is *mine*, and it has to do with the *processual*
> constitution of what is analyzed by NI. And it is quantitative.

I don't see how what you say provides a quantitative quantity theory of
price.  What I mean by a quantitative theory of price, is something like
equation (1) above.  P is a quantity, which is determined by the product
of two other quantities, MELT and L (SNLT).

What equation (or equations) represents your quantitative theory of price?

> (vii) this is crystal clear in your approach, of course, because
> dropping the money as a commodity, as I have stressed several times
> in the ISMT, you are dropping Marx's justification for value
> exhibiting nothing but labour, the argument according to which the
> abstract labour in a  commodity being represented in the concrete
> labour producing money as a commodity.

I don't understand why you think that this point - that the concrete labor
that produces the money commodity is the form of manifestation of the
abstract labor contained in all other commodities - is a
"justification" for the labor theory of value.  This point is Marx's
"second peculiarity [of three] of the equivalent form, which is discussed
in Section 3 of Chapter 1 (pp. 149-51).

Marx's justification (or derivation) of abstract labor as the substance of
value is presented in Section 1 of Chapter 1, and has nothing to do with
commodity money.  Marx's derivation is instead based on the assumption
that the exchange of commodities is the EXCHANGE OF EQUIVALENT VALUES,
from which it follows that there must be a common property of commodities
in terms of which they are equivalents.  And then Marx argues that this
common property must be abstract labor.

To repeat, this derivation has nothing to do with commodity
money.  Commodity money is derived in Section 3 of Chapter 1 as the
"necessary form of appearance" of the quantities of abstract labor
contained in other commodities.

Riccardo, why do you think that Marx's point about the "second peculiarity
of the equivalent form" is Marx's justification for the labor theory of
value?  What about his argument in Section 1 about the exchange of

> It is this that in Marx avoids
> the contradiction between the abstract labour being given prior the
> exchange, or within exchange.

I would put it like this: commodity money enables Marx to resolve the
contradiction that the abstract labor contained in commodities MUST
that non-commodity money can also function as an objective form of
appearance of abstract labor prior to exchange, as I have discussed in my
working paper.

> Too long for the couple of comments I wanted to do. But the synthesis
> is that I do not see how you escape the circularity you want to
> escape since both MV and SNLT are not independent from what is going
> on on the commodity market and the prices there.

As I have discussed, Marx's theory does avoid the circularity, because the
MELT and SNLT are determined independently of P, and jointly determine P.

It is your interpretation that the MELT and SNLT depend on P, that
prices are "there", not mine.

> The only way out
> would be to go behind objectified labour to labour as a 'fluid',
> labour in becoming, and having M endogenous because it is bank
> initial finance. This would give you an expected MELT. But this is
> what *I* do, and has nothing to do with a different theory of the
> individual determination of prices in the simultaneous setting.

How would bank finance give you an "expected MELT"?
How would the MAGNITUDE of this "expected MELT be determined?

Thanks again.


This archive was generated by hypermail 2.1.5 : Mon Nov 15 2004 - 00:00:02 EST