A reply to Paul C's latest comments:
(1) Of course Torrens wasn't assuming spring and autumn occur simultaenously.
Nor was *he* assuming simultaneous valuation. He had a use-value theory
of value (at least here; he wasn't consistent). In other words, he
treated corn as the substance of its own value, measured profit in corn.
(2) I am totally perplexed by Paul's statement that Torrens was assuming
unchanged agricultural conditions. Is this more than conjecture? I haven't
looked up the original quote, but Marx's discussion of it makes it hard
to believe that this was Torrens' assumption. Marx attacked him for
conflating use-value and value here.
(3) My point about simultaneous valuation was that "Marxists," who
supposedly do not take corn to be the substance of value, actually do
make it the substance of value when *they* "define" values to be simul-
taneously determined. If the input and output values of corn are
stipulated to be equal, one gets Torrens's profit rate, no matter what
changes in labor extraction, wages, etc. occur.
If v is the vector of unit values, A the input/output matrix, and L the
vector of living labor coefficients, most "Marxists" have DEFINED
values as
v = vA + L ==> v = L(I-A)**-1
THIS is simultaneous valuation. It equates the value at harvest time
(the LHS v) with the value at sowing time (the RHS v). Now Paul makes
the EXTREMELY important point that this makes no sense: the corn at
sowing time and at harvest time "do not exist simultaneously," so why
should their values? RIGHT ON! This is one of the 2 key points that
people like Mino, Alan, John, Ted McGlone, Eduardo Maldonado-Filho,
Werner de Haan, Paolo Giussani, evidently Manuel Perez, and myself
(have I forgotten anyone?) have been trying to make for a long time.
Since this is one of only 2 key points, if we can get everyone to
agree with this, we're halfway there! So, what do you think, folks?
Now, I agree that if one specifies that one is assuming unchanged
production conditions (and some other stuff), that values will
adjust to v over time. My point is that v is not a DEFINITION
of unit values, or "the" vector of unit values; that the unit
values are not solely a function of A and L. The Torrens example
shows that this is inconsistent with Marx's theory, and my discussion
with Paul indicates that it is perhaps a good thing that it is
inconsistent with Marx's theory, because it doesn't make too much
sense.
Now, we could have a whole discussion about equilibrium methodology and
stuff, which is the direction that Duncan seems to have been taking
my comments on the "transformation problem." (Notice that the
"transformation problem" is not at issue here--we have a 1-sector model.)
I'm not too interested in getting into that kind of thing at the moment.
Rather, what I want to stress is that almost EVERYONE has "defined"
both values and prices of production in Marx's theory as simultaneously
determined, stationary, or whatever one calls it. Without this
postulate taken as a definition (and not a special-case assumption), all
of the "proofs" of Marx's internal inconsistency with respect to the
falling rate of profit, the value/production price transformation, etc.
collapse. As Paul notes, in another EXTREMELY important comment:
"conditions of production can vary from year to year, making the value
rate of profit different from the rate of profit expressed in a single
use-value."
My only quibble with this formulation is that what matters is not the
"expression" of value, but its magnitude. One can always express
simultaneously determined "values" in labor-time or money, but the rate
of profit remains determined by the ratio of corn produced to corn
advanced. The point is that changing conditions--especially changes
in the amount of labor needed for production can make the value rate
of profit (and the "price" rate of profit, given that Marx's transformation
is correct just as he formulated it) differ from the rate of profit
DETERMINED by use-value relations.
BTW, all the stationary price models are ultimately reducible to ones
in which intrinsic value, value determined by labor-time, is eliminated.
One gets one's relative prices and solves the equations by *assuming*
that the price of the n-th commodity (the numeraire) never changes, or
that any changes are economically irrelevant, no matter how much the
albor-time needed for its production has changed. The basic logic is
that of the Torrens-Dmitirev-Sraffa corn model.
Almost all the "Marxists" have adopted this stuff and developed more and
more elaborate conceptions on its basis. But it is completely inconsistent
with the determination of value by labor-time.
Andrew Kliman