Once more on the question of value-price equivalence (I'm trying to clarify
something I said before, and also to understand Alan's position). After this
I'll shut up for a while.
1. In the "conventional" v = vA + L [henceforth, equation (1)] approach
values are defined independently of prices in this sense: the ingredients of
the calculation of values are technical I-O coefficients and direct
embodied-labour coefficients, with no price magnitudes in sight. It is
therefore an open (empirical) question to what degree values, so defined,
correspond to the prices at which commodities are actually traded. [It is
also an open (theoretical) question whether there exists a mapping from
values, so defined, onto the hypothetical set of prices ("prices of
production") ensuring an equalized rate of profit, which preserves certain
aggregate equalities. I mean "open" in the sense that the answer is not
implicit in the way value is defined -- but Sraffa's followers have shown
that for the pair of aggregate equalities {sum of prices of production = sum
of values, sum of profits expressed in terms of prices of production = sum of
surplus value} there exists, in general, no such mapping.]
2. On the approach expressed by the equation v = pA + L [henceforth, equation
(2)], values are _not_ defined independently of prices. Prices form part of
the data required for the calculation of values. Nonetheless, the approach
does not impose the assumption that prices and values are equivalent. It
leaves open the question of the "closeness of fit" between v and p. It also
-- and this is the thing I find puzzling -- leaves open the question of the
"closeness of fit" between the v's (so defined) and the labour-time socially
necessary for the production of the various commodities! The v's of
equation (2) hang somewhere "between" market prices and socially-necessary
labour-contents. In fact, it's clear that the v's of this theory equal
socially-necessary labour-contents [i.e. the v's of equation (1)] only if
v = p.
Who is assuming what? I take it that Alan's idea is this: If one _starts
out_ with v = pA + L, then one obtains the conventional v = vA + L only if
one imposes the condition p = v. This is what he means when he says that
equation (1) "assumes value-price equivalence", isn't it? But my rejoinder
is that there is no necessity to _start_ from v = pA + L. The conventional
equation can be said to "assume value-price equivalence" only in a frame of
reference in which v is _already defined_ as per equation (2); and it is the
appropriate definition of value that is in dispute. If I _define_ value as
in equation (1) then, as stated above, value-price equivalence becomes an
open empirical matter. Furthermore, if I define value in this way I reach
the following judgment: equation (2) is correct only in the special case of
value-price equivalence! Given v = vA + L, the equation v = pA + L holds if
and only if p = v. The situation is quite symmetrical.
Allin Cottrell
Department of Economics
Wake Forest University