From: Ian Wright (wrighti@ACM.ORG)
Date: Thu Mar 08 2007 - 18:42:31 EST
Hi Diego
Just some short comments to communicate that I am interested in what
you have to say.
> production amounts to a
> certain quantity of money, x, and requires a certain quantity of direct
> labour, y. Then I say that z = x/y is the monetary expression of labour (or
> the productivity in money of social labour), no matter whether the magnitude
> of z is changing or not.
Is x the total price of the net product (a la Foley/Dumenil?)
> (Others may have another theory—different from the LTV indeed—and relate the
> quantities of money to the quantity of whatever commodity they want. They
> could obtain the _monetary expression of corn_ for example. They are free to
> do so, but I would reply that when I go to the movies or the doctor or to
> cut my hair, etc., or in general observe other production processes I can
> see corn nowhere, whereas I see labour everywhere when I look at a
> production process. This is why I consider the LTV a more realistic theory
> than theirs)
I agree with your conclusion, but not the reasoning. I think the
"value controversy" can in principle only be settled in dynamic models
that include technical change and labour reallocation. I hope that
Marxist economists will develop and eventually agree upon on a
canonical (mathematical or computational, not natural language) model
of capitalist dynamics, much like neoclassical economics organized
itself around the Arrow-Debreu model. As evidence of the need for
dynamics, static models leave the theory of value under-determined.
For example, in linear production theory, which is the model type used
to critique Marx's theory of value, any real-cost basis (labour, corn
or any commodity) is formally equivalent, and is therefore a candidate
for the referent of price, even when the transformation problem is
solved via nonstandard labour values. There should be a formal theory
of why the different real-cost bases cannot be equivalent, but
quantitative value theory, it seems to me, has yet to get beyond the
transformation problem in static models.
> Now, the main point in my paper
> (http://www.countdownnet.info/archivio/teoria/521.doc) is
> that although values are a definite magnitude at the macroeconomic level,
> they can be quantified in a triple way at the microeconomic level.
> In the three cases, the cost of a given commodity (an average specimen of its
> class) is the same: m·(A+B), where m are market values in the sense
> explained below, and A and B the coefficients for both material and
> _worker-subsistence_ inputs. But the surplus-labour included in the
> commodity's value can be understood and quantified in a triple, equally
> analytically useful, way.
My initial reaction is that there is only one quantity of
surplus-labour, not three, although I note your surplus-value
equalities below.
> 1. When we think of surplus-value as proportional to the new value included
> in value, we obtain _direct value_. We do so because we are interested in
> dealing here with the process of creation of values as well as with the
> process of exploitation of workers by capital.
I wonder whether standard labour values ("direct value") are only
appropriate for simple commodity production, or production without a
tendency for profits to equalize, or snapshot moments in which new
surplus-value is produced, but has not yet been distributed?
> 2. When we assume that surplus-value is proportional to the magnitude of
> value that capitalists have to advance in production, we obtain _production
> value_. We do so because we are interested in the _abstract_ process of
> competition, i.e. competition in a context where only capitalists in the
> productive sector exist.
I wonder whether this case (in which our formulae for labour-value
appear to be formally equivalent, although perhaps we have different
interpretations) such labour values are only appropriate for simple
and expanded reproduction, in which the tendency of profits to
equalize has been realized.
> 3. When we assume that surplus-value is NOT proportional to the magnitude of
> value that capitalists have to advance in production, but rather an adjusted
> surplus-value (i.e., surplus-value once deducted from it the sum of
> interest, commercial margin, ground-rent and taxes) is proportional to
> adjusted costs (where the value-equivalent of interests, commercial margin,
> ground-rent and taxes is added to the value of material-and-subsistence
> inputs), then we have the _long-run market value_. In this case we are
> interested in the process of competition in more concrete terms, where along
> with productive capitalists there also exist commercial and financial
> capitalists, the landowners and the state.
I think if we had a dynamic model your 3 cases would be special cases
of the general causal relation between prices and labour-values.
> In case 1, surplus-value included in a (vector of) commodity is m·B·g (where
> g is the rate of surplus-value). In case 2, surplus-value is m·(A+B)·r
> (where r is the general rate of profit). In case 3, surplus-value is
> m·(A+B)·R, where R is the diagonal matrix of actual long-run rates of
> profits of each sector.
>
>
>
> Now, call sw, sp and sm the surplus-value calculated according to the three
> cases. The labour theory of value states that:
>
>
> sw·x = sp·x = sm·x (where x is the column vector of output) even if at the
> individual level sw, sp and sm differ.
> costs are always the same at both the individual and the social level (i.e.
> m·(A+B) and m·(A+B)·x respectively)
> surplus-value are the same at the aggregate level [m·B·g·x = m·(A+B)·(1+r)·x
> = m·(A+B)·R·x] even if they differ at the microeconomic level [m·B·g is in
> general neither = m·(A+B)·(1+r) nor = m·(A+B)·R·x]
> the rate of profit is a quotient between two quantities of labour at the
> social level: surplus-labour in a period divided by the value-equivalent of
> constant capital advanced. That is r = [m·B·g·x] / [m·(A+B)·x].
I have not had time to study your paper, but I certainly will. Your
suggestions are interesting, particularly the surplus-value
equalities.
Best wishes,
-Ian.
This archive was generated by hypermail 2.1.5 : Sat Mar 31 2007 - 01:00:12 EDT