Jim Devine initiated this exchange with the claim that my argument
concerning the relative significance of Walrasian-type general
equilibrium models and Marx's labor theory of value contained a
"contradiction." When I negated this claim by referring to a
distinction made explicitly in my original post, Jim dropped the
issue cold--nowhere in his most recent post does he pursue the claim
that my argument contains a contradiction (and properly so, since
there never was one).
Now his central criticism is that I don't understand Marx's
labor theory of value. Perhaps so, but on closer reading it turns
out that what he is really saying is that I don't understand *Jim
Devine's* interpretation of Marx's labor theory of value. This may
be true (and I may not be alone in this condition), but that is a
very different statement, as Allin C.'s recent post confirms.
Rather than copying Jim's entire missive, I'll summarize the
debate leading up to each of his major points, print his statement
verbatim (subject to some editing for brevity; his comments will be
prefaced by >), and then take it from there.
1) In response to Jim's suggestion that my argument contained a
"contradiction", I noted the distinction, explicit in my earlier
post, which voided his claim--specifically, that Marx's LTV, unlike
GE conditions, has no coherent logical grounding (e.g., the first
premise in Marx's Vol I, ch. 1 argument is not in general true, a
point on which I see Steve Keen has recently preempted me).
Note that, even if this position is wrong, it eliminates any
possibility of "contradiction" in my (limited) support for one
analytical system over the other.
To this Jim responds:
> Gil, please don't reject a contradiction this way: you respond
> to the "on the one hand" before you get to the "on the other hand."
> I was specifically dealing with your rejection of the LTV on the
> basis that folks don't act following signals sent by values (a
> rejection that Marx would accept). I paid attention to your
> contributions, but I
> happen to reject your dismissal of Marx's law of value because
> I don't think you understand it very well. But that was a different
> debate. I didn't feel like repeating that discussion.
These considerations are utterly irrelevant to Jim's initial
claim that my argument was contradictory. The latter is a claim
about the logical *structure* of my argument. Here Jim is taking
issue with the *substance* of my argument. Even if he's right on the
substance, there is still no contradiction. Jim does not pursue this
claim further.
1a) In passing, I suggested an equivalence between Sraffian-style
prices of production and a type of GE price structure. To this Jim
responds:
> prices of production are not GE prices, since (among other things),
> labor-power markets are not in equilibrium.
I gave my reason for considering prices of production a version
of GE prices in my discussion on the net with Ajit S. (to which I
must return sometime...)--that being, the conditions imposed in the
Sraffian system must be considered utterly arbitrary, and thus not
worth considering, *unless* one interprets them as (minimalist)
general equilibrium conditions. Jim does not respond to this, so I
maintain the point. [I owe Ajit more argument, but that's a separate
post.]
2) In response to my discussion of why I consider Marx's labor
values epiphenomenal at best, Jim offers the following tutorial.
> So I'll repeat the discussion:
> As I've said before, I think that the whole Ricardian problem-
> atic concerning so-called "labor values" is misplaced or misled
> if it aims to be a description of Marx's Law of Value. Among other
> things, I try to separate the Ricardian "labor theory of value"
> (oft criticized by Steedman, Roemer, etc.) from Marx's "law of
> value." ...
>... I think Marx's theory makes sense
> and gets us beyond the Steedman-Sraffa and Roemer critiques
> (and the valid critiques of Steedman by Farjoun et al). The MTV
> (as I understand it) is a completely different vision from the
> Ricardian project of the LTV as a way of calculating prices on
> the basis of supposedly pre-existing "labor values." Different
> goals, different tools, different results.
> I think Allin's discussion of labor values is interesting (and even
> valid) but -- to repeat myself for at least the third time in this
> e-missive -- it's not Marx.....
> In my view, as I've stated over these wires several times, at the
> time that a commodity is sold, it has both a price and a value.
> Thus, price cannot be derived from value and value cannot be
> derived from price. In my reading of Marx, (exchange)values form an
> alternative accounting framework to prices. While prices reflect
> the fetishized realm of appearances that is the basis for indivi-
> dual actions, values are a tool that Marx uses in order to dig
> through the level of appearances in order to reveal the class
> nature of capitalist society. Among other things, Marx *assumes*
> that values are proportional to prices, which amounts to assuming
> that the average or "representative" firm and labor process
> is the image of the macrocosm of class relations. (it's also
> a tactic involving assuming "bourgeois" standards of equal
> exchange, in order to show that capitalism doesn't live up
> to its own standards.)
This is an interesting and possibly valid approach to justifying an
approach similar to Marx's labor theory of value, but there are at
least two problems with it:
a) As Allin C. has already suggested, and contrary to Jim's
representation, this is not Marx's understanding of the LTV, at least
as he develops it in Volumes I and III of _Capital_. First, he does
not merely "*assume*" a relationship between prices and values. As I
stated in my original post, Marx thinks of value as the equivalent of
"long-run equilibrium" prices in a classical sense, i.e. the value
toward which prices tend when the various "short-run" effects on
market price cancel out. Two direct quotes:
Volume I, Ch. 5 (p. 261 in the Penguin Books edition): "It is true
that commodities may be sold at prices which diverge from their
values, but this divergence appears as an infringement of the laws
governing the exchange of commodities. In its pure form, the
exchange of commodities is an exchange of equivalents..."
In addition, Marx cites two authorities, Mercier de la Riviere
and Le Trosne, in support of this claim. There is no need to cite
authorities in support of a mere "assumption".
Volume III, Ch. 21 (p. 478 in the Penguin Books edition): "Similarly
with wages. If supply and demand coincide, their effect ceases, and
wages are equal to the value of labour-power."
It seems clear that Marx is not simply assuming an equation of
price and value in the "pure" or "long-run" case, contrary to Jim's
suggestion. Allin's recent post buttresses this conclusion.
b) If there is no necessary relationship between prices and values,
as Jim suggests, it is hard to see on what basis other than mere
tautology Jim can claim that values "reveal the class nature of
capitalist society." How, necessarily, if there is *no* _a priori_
relationship between price and value, as Jim suggests?
Furthermore, John Roemer has shown that there are no analytical
grounds for such an unqualified claim. Whatever you think of the
relevance of his analysis to Marxism, he does show that logically,
values provide neither a necessary nor a sufficient basis for
understanding the "class basis of a capitalist society". Not
necessary, because Roemer (and for that matter, on different grounds,
Erik Olin Wright) has shown that one can account for class
structure in non-value theoretic terms; not sufficient, because for
example M-- M' arises in his Credit Market Island (and thus producers
are exploited) precisely *because* values diverge from prices. Marx
invalidly discounts this possibility, on value-theoretic grounds, in
Ch. 5 of Volume I.
Next, Jim says:
> Gil asserts that values have nothing to do with anything.
>Gee, production can happen without labor? please tell me how.
The equivocation fallacy in this statement is so obvious as to not
require (or deserve) further comment.
> importantly, the MTV is organically linked with Marx's materialist
> conception of history, which I view as summed up by his aphorism
> that "people make history but not exactly as they please." (For
> a good summary of this theory (though it's a bit mechanistic at
> times, since it's a critique of German idealism), see THE GERMAN
> IDEOLOGY.)* The bottom line of the MTV is that we live in a
> complex network of social relations, one created by people and
> one that shapes and limits the actions of people in a dynamic
> and historical process. (The bottom line of Walrasian GE, on
> the other hand, is that we don't.)
This statement seems doubly inappropriate. On one hand, contrary to
Jim's suggestion, Walrasian GE is absolutely consistent with the
notion that "we live in a complex network of social relations, one
created by people and one that shapes and limits the actions of
people..." It abstracts from and simplifies that reality, but
then.....On the other hand, Marx's value "accounting framework" is no
less an abstraction from and simplification of capitalist reality
than is Walrasian GE.
But, Jim insists,
> *It's important to remember that Marx rejected the academic
> distinctions that we cling to. Economics is not separable
> from history, or from sociology, or from philosophy...
Perhaps not, but Marx's *value theory* is certainly separable
from his history, his sociology, and his philosophy, even if grounded
in them.
3) I argued that GE models provide a "first-pass" view of reality by
imposing a minimum standard of mutually compatible behavior, while
the LTV doesn't have even this _a priori_ standard going for it. To
this Jim responds:
> I don't think GE reveals much at all. At best it says something
> about the realm of appearances-- and I am being charitable, since
> the market system that GE claims to describe or give us a handle
> on describing is very different from the totally abstract and idealized
> economic utopia that is GE. As I've said before, I like to read
> utopian novels. Debreu's book should be put on the same shelf
> with Bellamy's LOOKING BACKWARD. While Bellamy presents a totally
> unrealistic picture of a statized economy, Debreu presents a similar
> very nice picture of an atomized economy. Both utopias say something
> about empirical reality by way of contrast. Both are strictly
> speaking normative works. (The normative overlay of GE theory
> explains why Roemer very quickly abandoned positive economics and
> got into normative economics. Smart guy.)
>
As a general thing I have no major objection to these statements,
since my defense of GE models was a limited one from the beginning.
However, 2 points are worth noting.
a) There is no important sense in which the condition of GE is
necessarily "utopian". Roemer has shown that it is typically
exploitive under capitalist property rights. Furthermore, in the
presence of certain forms of market failure (nonverifiability of
effort, say, or negative externalities) GE is not efficient. It may
be an abstraction, but contrary to Jim's suggestion, it is not
necessarily a utopian one.
b) The common use of "computable GE models" suggests, contrary to
Jim's representation, that they have an application to empirical
reality beyond that "by way of contrast".
Jim continues:
>
> More importantly,
> I get the feeling from the above that Gil has a very slippery
> meaning for GE theory. Whereas most economists see the GE model
> a model which actually represents a "first pass" description of
> the way the world works, here Gil describes it as only being
> a minimum standard of socially consistent behaviors. That is,
> while others see GE as the whole Walras-Debreu-Arrow-Hahn-etc.
> system, here Gil describes it as simply being Walras' Law
> (the sum of all excess demands equals zero) and such propositions
> that prices tend to equalize within markets and rates of profit
> tend to equalize between markets.
No, this is an inaccurate representation of my position. *I*
said that GE may provide a "first-pass" description of empirical
reality. My *justification* for arguing the superiority of GE over
an LTV approach was that the former, unlike the latter, was based on
some consistent logical ground, in this case the "minimum standard of
socially consistent behaviors." And by the way, that "rates of
profit tend to equalize between markets" is *not* a necessary result
of the Walrasian GE framework, contrary to Jim's suggestion. That's
one of the problems with the Sraffian system.
> If this is true, then GE is a theory that was developed by the
> Physiocrats. If this is true, we don't need all of the mathematical
> baggage that Debreu et al have introduced. If this is true,
> then Marx's reproduction schemes for simple reproduction
> (toward the end of vol. II of CAPITAL) are GE.
Yes, Jim has re-discovered a point I've been making in the discussion
with Ajit: the classical systems, and for that matter Marx's
reproduction scheme and the Sraffian system, should be considered as
limited versions of a GE model, because they impose conditions which
must be considered as *utterly arbitrary*, and thus not worthing
bothering with, outside of an interpretation of "long-run
equilibrium".
However, it doesn't follow that in that case "we don't need all
of the mathematical baggage", because as I've also argued, and we've
seen an example of above, the minimalist conditions of these other
systems often don't make sense. Walrasian GE, in other words, has
more content than, say, Sraffian GE. That's why Roemer's critique of
mainstream theory says everything that Sraffa's critique can say, and
more (such as Roemer's Isomorphism and Class Correspondence Theorems).
Well, this is more than long enough. I'll finish with the last part
of Jim's comments tomorrow. Gil [[EMAIL PROTECTED]]
