Re: [OPE] The Labour Theory of Value: a Marginal Analysis

From: Ian Wright <>
Date: Tue Sep 16 2008 - 17:39:03 EDT

Hi Duccio

I'm glad you shared your thoughts on Hagendorf's work here.

> Assuming perfect competitive equilibrium conditions, Hagendorf defines
> the concept of marginal labour value (i.e. the quantity of labour
> embodied in a unit increment of output) as the inverse of the marginal
> productivity of labour (the increase of output due to a marginal
> increase of labour) which sets the uniform wage rate. But he seemingly
> confuses the quantity of labour embodied in a unit increment of output
> with the quantity of labour commanded by this increment, as he obtains
> it by deviding the money value of output by the wage rate.

Perhaps one way of understanding Hagendorf better is to first restrict
his statements to simple commodity production in which there is no
profit "on stock" (to use Adam Smith's term).

In this case, in equilibrium, labor-embodied is identical to
labor-commanded because prices are proportional to labor-values.
Everyone agrees with this, including neoclassicals, Sraffians, etc.

Do you think Hagendorf's analysis is entirely correct and consistent in
this special case? I haven't checked in detail, but my guess is "yes".

You state that, in the more general case, in which there are profits in
stock, then Hagendorf is "confused" when he freely translates between
labor-embodied and labor-commanded. But I think you might be
misunderstanding Hagendorf here. As far as I understand it, he is saying
that labor-values *should* be as he defines them, in which case it is
possible to freely translate between the two.

In general, but not in specifics, I agree with Hagendorf here, because
the "standard" definition of labor-value, in my view, is a correct
measure of labor-values only in the case of simple commodity production.
The correct measure in the more general case requires "nonstandard"
labor-values. If we adopt this (correct) definition then prices of
production are proportional to labor-values

So the claim that it is "confused" to translate between labor-embodied
and labor-commanded begs the question since it relies on the standard
definition of labor-values. But this standard definition is itself being

We all know that the standard definition has generated a whole series of
paradoxes, problems and confusions for the labor theory of value,
notably the problem of an invariable measure of value, the
transformation problem, the redundancy critique, and so on. This is
easily explained using the "nonstandard" approach: the standard
definition does not measure labor-values under conditions of capitalist

> Further, he
> regards the marginal labour values of commodities as 'socially necessary
> labour'. On these grounds he affirms that marginal analysis applied to
> labour values eliminates any quantitative difference between labour
> commanded and labour embodied.
> He then introduces some questionable neoclassical static equilibrium/
> /assumptions and on their behalf he claims to have shown that in long
> run competitive equilibrium prices of production must be proportional to
> marginal labour values, reckoned as vertically integrated additional
> amounts of dated quantities of labour extending back in time and
> required to produce an additional unit of output (an odd marginalist
> reinterpretation of Pasinetti’s vertically integrated labour
> coefficients). He concludes that modern economics is not only consistent
> with the classical definition of the labour theory of value, but that it
> is built on it. And that Marxian exploitation is easily explained
> in such framework.
> As noticed by Jan Wright, Hagendorf - disregarding the large theoretical
> evidence originally supplied by Dmitriev and Bortkiewicz and later
> reproposed and extended by Sraffa - seems to consider both the Marxian
> ‘law of value’ (the 'pure' labour theory of value, which has
> an ‘objective’ nature) and the marginalist theory of value (a theory of
> ‘subjective’ nature, based on utility and scarcity) as logically tenable
> and easily reconcilable.

I think this is likely too, although I hesitate, at this stage, to be as
definitive as Hagendorf. But I think he is onto something very important.

> As far as I understand, in his paper Hagendorf refers to ‘real’,
> standard labour costs. Not to Jan’s ‘nonstandard labour values’, which
> are the result of a virtual production process, where capitalist
> consumption is reckoned as an additional social cost of production (in a
> counter-factual theoretical framework, where capitalists are assumed to
> consume their entire non-invested income during the replacement
> process). I do not see why Jan says that Hagendorf’s alleged marginalist
> approach to the labour theory of value “has an important relation to the
> difference between standard and nonstandard labour values”, a subject
> matter which he is particularly interested in.

Nonstandard labor-values (NSLVs) are not the result of a virtual
production process. Just like standard labor-values (SLVs) they are also
a property of the current production structure.

I do, however, explain the difference between SLVs and NSLVs in terms of
a counterfactual "process of replacement", which is simply a sequential
interpretation of the process of vertical integration. But both SLVs and
NSLVs could be interpreted in other ways (e.g., as employment
multipliers, or indices of productivity etc.).

For example, the electrostatic potential energy of a charged particle in
a field is defined as the work that must be done to move it from an
infinite distance away to its present location. This is one
interpretation that physicists use to understand potential energy.
However, it does not imply that the particle was actually moved through
an infinite distance.

Similarly, the labor-value of a commodity can be defined as the total
amount of labor that must be supplied to replace it "from scratch",
given the current production structure (the analogy of the "field"). But
this does not imply that there really was a process of commodity
production that starting from labor alone in pre-civilized times.

One subtle difference between SLVs and NSLVs is how this "process of
replacement" is constructed. In the standard approach, *both* workers
and capitalists abstain from consumption during this "process of
replacement"; in the nonstandard approach *only* workers abstain. So now
we have two definitions of labor-values, whereas before we only had one.
Which one is correct? What is the relationship between them? Once you
start asking these questions I feel that many things that were once
obscure and intractable become transparent and soluble.

Best wishes,
ope mailing list
Received on Tue Sep 16 17:41:12 2008

This archive was generated by hypermail 2.1.8 : Wed Dec 03 2008 - 15:12:31 EST