A response to Ajit's ope-l 5001.
He writes:
"As i said in response to Andrew, his prices are abstract too because in
general prices for a commodity when bought in bulk is lower than prices for
the same commodity when bought in small quantity."
I did not address this point in my response to Ajit, because I wanted to focus
on the difference between the determination of value by labor-time and the
determination of value by corn. But let me say now that
(a) Ajit has drawn attention to an insufficiency in some of Ted's and my
presentations. This insufficiency is due largely to a failure to anticipate
that people would misunderstand, and partly to a desire to keep the
presentation concrete by, for instance, refuting Bortkiewicz without having to
explain all of Marx's value theory.
(b) Yet, in my interpretation, C and V refer to actual costs, not abstract
ones. Yet there may be a difference between actual expenditures and actual
costs, because only socially necessary labor-time counts as value. There is
no presumption, nor any requirement, that a single price prevails.
Imagine, for instance, a two-producer sector, in which A buys 2500 yards of
cloth at a price of $2 each, and B buys 200 yards of cloth at a price of $2.27
each. Ignore other inputs and, in order simply to focus on price differences,
imagine that they have the same technology: A extracts, say, 500 labor-hours
and produces 50 widgets, and B extracts 40 labor-hours and produces 4 widgets.
Also imagine that $1 = 1 labor-hour.
Then the individual value of A's total output is 2*2500 C + 500 V+S = 5500.
The individual value of B's total output is 2.27*200 C + 40 V+S = 494. The
total social value of widgets is 5500+494 = 5994, the total output is 50+4 =
54, and the social unit value is 5994/54 = 111.
For the sector as a whole, C = 2*2500 + 2.27*200 = 5454, and the C per widget
is 5454/54 = 101. This is the sum transferred to the widget, the actual
social COST, i.e., the socially necessary labor-time needed to produce a
widget, net of the value added. So, through production, A transfers and
preserves a value of 101*50 = 5050, which is 50 more than A's actual
expenditure (2*2500). B transfers and preserves a value of 101*4 = 404, which
is 50 less than B's actual expenditure (2.27*200). So, via intrasectoral
competition, value has been distributed from B to A, with no gain or loss in
the aggregate.
Thus, even though C does not correspond immediately to actual expenditures, it
is an actual cost, not an imaginary amount computed by postulating conditions
contrary to fact, as simultaneist costs are.
(c) Ajit's ope-l 4912 makes the related point: "when you take the input
prices as given and work out the output prices as prices of production by
equalizing the rate of profit, there is no guarantee that these prices would
be the actual prices.
Most likely, the actual prices would diverge from your prices of production,
... Thus your prices again are abstract prices and not actual or real
prices." I agree with this. But I'm not sure of Ajit's point here. I do
not consider the Marx's account of the transformation of commodity values into
production prices to be an algorithm for predicting actual prices. The point
was instead to show how one coherently can maintain that competition and the
market do not negate the determination of value by labor-time -- they have no
effect on commodity value and surplus-value in the aggregate. I have never
said that the actual market prices will equal or even be close to production
prices. It makes no difference: whatever the actual prices may be, the
actual aggregate price is C+V+S and the actual aggregate profit is S.
Moreover, although the production prices are certinly not actual prices,
neither are they imaginary amounts computed by postulating conditions contrary
to fact, as simultaneist equilibrium prices are. They are computable by
appropriately "averaging" actual price data.
Andrew Kliman