Re: [OPE-L] Marx on the 'maximum rate of profit'

From: Ian Wright (wrighti@ACM.ORG)
Date: Tue Oct 31 2006 - 21:05:29 EST

Hi Paul

> What if the capitalists pay themselves a wage for their
> labour of supervision which is distinct from their profit?

I think the answer depends on one's perspective on the
productive/unproductive distinction.

A simple approach, in the Sraffian single production framework, is to
count the "labour of supervision" as standard labour, in which case
some individuals in the economy receive both wages for labour and also
returns to capital. In this case, some of the total real wage is
consumed by capitalists, and some of the total wage bill is received
by capitalists. This has no consequence for the standard analysis.

But I guess you wish to raise the issue that "labour of supervision"
is unproductive. In what follows I'll assume that you define
unproductive labour as labour that does not add new value to the

In equilibrium, if the wages of supervision are spent on consumption
goods but the labour-embodied in those goods is not counted as
reappearing in the product due to the application of "labour of
supervision", then you'll lose conservation of value in price and get
a "transformation problem".

This kind of unproductive labour approach is equivalent to considering
that the labour-embodied in the consumption goods bought by the "wages
of supervision" is not a cost of production. Labour-values, defined in
this way, implicitly assume that no such consumption goods are
consumed during the period of replacement. Hence, these labour-values
are counterfactual, in the sense they refer to the replacement costs
that obtain if "labour of supervision" is absent. This is similar to
the standard case of Sraffian labour-values, which calculate the
replacement costs that obtain if capitalist profit is absent (e.g.
simple commodity production).

If one's labour-value accounting definitions reject local conservation
of value then of necessity there is some kind of "transformation
problem". This is similar to neglecting forms of energy (e.g. heat and
sound generated by collisions) when formulating principles of
conservation of energy. And such conservations breakdowns did
initially confound the early pioneers of conservation of energy (see
Meyerson's "Identity and Reality").

Let's rewind back to the start, in the hope I can get you to look at
the TP from an entirely new perspective.

Who told us that price-value divergence is real due to the existence
of uniform profits on capital invested? It was Smith, Ricardo and
Marx. Did they provide any empirical evidence for this assertion? No.
Both Smith, Ricardo and Marx provided (more or less confused)
numerical examples that purport to demonstrate the necessity of
divergence. So, from the very beginning, the existence of this
divergence has a logical or conceptual, rather than empirical,

Only later did Marxists, such as Shaikh, armed with linear algebra,
decide to empirically measure the divergence in order to defend the
LTV from the neo-Ricardian critique. But to measure anything, one
first needs a conceptual definition of the thing to be measured.
Shaikh employs the neo-Ricardian formula for labour-value. As
predicated by the model, the Sraffian definition of labour-value does
indeed exhibit an empirical divergence from profit-equalizing prices.

But which comes first, the ontological definition or its empirical measurement?

Sraffa's labour-values do not measure the replacement costs for a
capitalist economy. But if we adopt a definition of labour-values that
correctly measures replacement costs for a capitalist economy the very
same model predicts no divergence of prices from values due to
profit-rate equalisation. And as you have shown empirically, it turns
out that prices of production (which are proportional to real-cost
labour values) are slightly better predictors of market prices
compared to Sraffian labour-values, although there's not much in it.

But the whole "transformation problem" is conceptual, rather than
empirical, from start to finish. In the context of static equilibrium
models, it is a purely epistemological problem, and has simply
functioned to discredit the idea that prices can represent amounts of

A long attempted answer to your short question ...


This archive was generated by hypermail 2.1.5 : Thu Nov 02 2006 - 00:00:03 EST