Duncan
------
At a theoretical level you run into the types of puzzles Ian Steedman has
put forward: with joint production, "surplus value" measured in embodied
labor may be negative when profit is positive.
Paul
----
Steedmans examples are highly artificial, and there
is no evidence that they actually occur. There is good
reason to suppose that they will not occur, since, to
produce negative surplus value he depends on two
industries with quite different efficiencies of
production. In a real economy the high cost one would
close down, so he is portraying an extreemly
unstable situation where one of the industries is
expending more labour than socially necessary.
Duncan
------
At an empirical level, you lose the immediate connection between the
Marxian categories, say surplus value and variable capital, and aggregate
economic statistics.
I think Marx saw a much closer connection between his theoretical
categories and the surface phenomena of capitalist reality than a strict
embodied labor view of value will sustain.
Paul
----
I would certainly be loath to lose the ability to
use real economic statistics. If, however, there is a
close empirical correlation between actual prices and
embodied labour coefficients, then the error in terms
of embodied labour that arises from calculating the
rate of surplus value from money aggregates will be
small. In particular it will be small relative to
the host of other approximations one has to make in
dealing with economic statistics - the errors in the
sources, the difficulty of getting accurate classifications
of labour into productive and unproductive etc.
I think it is now pretty well established that the
correlations in question are very good, and that
the errors involved in using money quantities as
surrogates for aggregates like s/v or v/c are
tolerable.
The errors in using monetary surrogates become more
serious if we disaggregate the economy to the level
of individual industries.