Both Allin and Fred have asked for the citation on Torrens. It is TSV III,
pp. 77-79. Anyone using the Collected Works, etc. can find this near the
beginning of Marx's discussion of the disintegation of the Ricardian
school.
I don't have the ability to re-post my posts--maybe Jerry or someone else
can do so? I think all my discussions of Torrens came under the subject
heading "the 'transformation problem'"--which it actually has nothing
to do with, since it's a one-sector model.
Allin suggests that the inability of simultaneously determined ("replacement
cost") representations of Marx's value theory to replicate the conclusions
he reached in his critique of Torrens (they all give Torrens's results, not
Marx's), should not be too worrisome. Why not? Because this is only a
one-sector model.
Response: Sure, but no matter how many sectors you have, the simultaneous
models (of value as well as price determination) eliminate intrinsic
value--the labor-time needed to produce commodities--as a *functional
determinant* of value and price aggregates. The simplest way to see this
is to imagine all compositions of capital are identical. Then the
conclusions are exactly those of a one-commodity model.
Now, you can make it look like labor-time matters to the determination of
value by letting compositions of capital differ, but it is surely a
strange kind of "Marxian" value theory that REQUIRES price-value deviations
in order for labor-time to influence values! Moreover, even in this case,
simultaneous (or "replacement cost") valuation is still really incompatible
with the determination of value by labor-time. How do you think Okishio
manages to "disprove" Marx's law of the FRP? Analytically, what goes on is
that one commodity is made a numeraire and, by postulate, its value never
changes--no matter how much the labor-time needed to produce it changes.
The only "values" that can change (or prices) are *exchange*-values,
relative prices. The possibilities for a decline in total value when
labor extraction falls are all but eliminated.
For instance, imagine that a commodity's value in terms of a numeriare
falls by 3% and the intrinsic value of the numeraire falls by 17%. The
fall in the commodity's intrinsic value is then about 20%, but the
simultaneity postulate wipes away 17 0n one swoop, because the numeraire's
"value" in terms of the numeriare doesn't change. Just as in the corn
model case, a commodity, not labor, has been made the substance of value.
Another simple way of understanding all this is: when one stipulates that
the value (or price) of a commodity must be the same at two different
moments of time (whether in reality or as a "definition" of costant
capital as the "replacement cost" of means of production), no matter
how much the labor-time needed to produce it has changed during this
expanse of time, then one has stipulated that labor-time is irrelevant
to the determination of its value!
The "Marxists" have all worked with a "defintion" of value that itself
invalidates Marx's whole value theory. From this follows *all* the
alleged quantitative internal inconsistencies of his theory. In other
words, what contradicts his theoretical conclusions is not his own
theory as such, but the standard *interpretation* of that theory.
Andrew Kliman