Comments on this discussion provoked by Massimo's numerical comparison:
1. I found Massimo's example cogent and reveraling of the differences
between temporal and simultaneous valuation. The advice given by the
temporal consultant to the labor bureaucrat accords very closely with
the "advice" I would give--if by "advice" we mean "giving him a piece
of my mind."
2. It is now clear that Massimo was NOT concluding anything from a one-
period example alone. He also had a very cogent discussion of the
"feedback" effects, falling rate of profit, automation, etc. Thus, I
disagree with Alan when Alan wrote that massimo opened himself up to
Bruce's critique.
3. I agree with Bruce that if one wants to study the effect of a change
in labor extraction, one should study the *immediate* effect by excluding
all factors which change in response. BUT AMONG THESE OTHER FACTORS ARE
CHANGES IN PRICES; THEY SHOULD ALSO BE HELD CONSTANT. If one wants to
look at the dynamic generated by the initial change, one should in principle
not exclude any process, and one should certainly not "priviledge" price
changes to the exclusion of everything else. I.e., one should not hold
everything else constant and iterate prices ad infinitum (and ad nauseum)
as if that had any meaning at all. When one "iterates" a temporal valuation
model, the rate of profit is changing, s/v is changing, etc. Capital will
react to this in some way, no? And as it reacts, we will have what is to
the simultaneist another "perturbation" from the equilibrium (or whatever
--I'm not interested in playing word games) position, even before the
adjustment to the "center of gravity" (or whatever ...) has occurred. So
much for "convergence."
BTW, I've never ignored or tried to hide "convergence." I've always tried
to note the conditions under which convergence will occur.
BTW, Massimo's original post also noted this: he spoke about *continuous*
technological change. As the work of John E., Alan F. and myself has
shown, when technological change is *continuous*, then the adjustment to
the static equilibrium (or whatever ...) result of the Okishio theorem
never takes place; so that if value is determined by labor-time, the profit
rate can fall even when Okishio's rate must rise.
What we're again hearing from the simultaneists is static equilibrium
reasoning, as Massimo noted, in which all changes are one-time-only
perturbations--EXCEPT for the iteration ad infinitum of prices. Hmm.
I wonder why price changes are treated differently. Might it be in
order to reinstate a physicalist conception of valuation instead of
determination of value by labor -time?
Let me also note that when Marx said that changes in the wage rate do not
affect (output) values although they do affect output prices, he was doing
exactly what Bruce thinks he (Bruce) was recommending--holding everything
constant except for the wage rate change--including input prices. This
is well known to Wolff, Callari, and Roberts, who note that their model
differs from Marx in this regard! But then they try to reconcile their
model with Marx, by saying he abstracted from 2d-order effects--yes,
because they take place in the NEXT period. The change in this period's
output prices affects NEXT period's input prices, etc. So what WCR do
is to revalue the inputs retroactively, it seems, which makes sense
to me only when I have my head on backwards. I do agree that the change
in the wage rate can affect SUBSEQUENT values, etc. and that it is proper
to take these 2d order effects into account. But, as I've noted, then one
must look at *all* the relevant changes and reactions brought about by the
wage change, not just revalue the prices.
4. Thus, contrary to what Bruce says, it is not the TSS interpretation
that imposes *lags* on capital valuation. Rather, the simultaneists
impose *leads* on capital valuation--value it retroactively. To quote
Bruce: "It never occurred to me that this was something that could be
argued about." Corn enters production at time 0 with a value of 5; at
harvest time, the same amount of the same type of corn is worth 4. The
simultaneists say its input value in this period was 4. The TSS
interpretation says its input value was 5. Thus, the TSS interpretation
treats price as coincident, not lagged; the simultaneists treat price with
a lead.
5. In repsonse to Bruce's question: yes, if the value of wages remains
constant, and workers do less work, they are less exploited. And in
Marx's theory, workers are paid for their labor-power when they "sell" it
in the market, BEFORE production, even if they are actually paid after-
wards. In other words, the value of the wage is fixed by contract or
whatever when the worker sells herself. Thus, a change in the value of
means of subsistence caused by and occurring after a fall in the working
day cannot affect the value of wages already paid (or due the worker).
6. Again in response to Bruce: I know of no TSS proponent who says that
every period should have a uniform profit rate. Quite the opposite.
What we do say is that a uniform profit rate and stationary prices are
two entirely *different* things. I tend to work with uniform profit
rate models as examples precisely so that no noe can dismiss the results
as "disequilibrium dynamics" simply because input and output prices
differ. Of course, when one is examining production prices, uniform
profitability must be assumed.
Not at all incidentally, Marx's Vol. III, Ch. 9 discussion assumes a
period of one year, so it isn't a "long-period" adjustment model.
Moreover, Marx very clearly distinguishes between the FORMATION of
a general profit rate and the EQUALIZATION of the profit rate, even by
putting them into separate chapters. What's my point: the formation
of production prices does NOT require, or presume, that the profit
rate is actually equal. Production prices are the ideal average
result of competition
7. To Riccardo: value is indeed redundant in the simultaneous models,
not because of any "transformation problem," but because of simultaneity
itself. E.g., in a 1-sector model, value is still redundant under
simultaneous valuation, since the profit rate and all (= 0) relative
prices are determined independent of value. But in the TSS interpretation,
value is not redundant at all, even quantitatively.
Even if we assume uniform profitability and given input prices, in the
TSS interpretation (as in Marx) we still cannot determine output prices,
because the profit rate must still be determined. And the minute you
give up stationary prices as an *axiom*, the profit rate cannot be
determined within the "price system." Marx and the TSS interpretation
say it is determined as a result of production, before commodities go
to market. Hence, value is no longer redundant, because the surplus-value
extracted in capitalist production is needed in order to determine the
output prices and the level of the profit rate. Forget sentimentality.
8. Finally, to Bruce, a question: if, as you now say, you consider your
prices a "structural 'abstraction'" why do you object to the TSS
interpretation of Marx's value theory as a representation of the
temporal process of valuation, i.e., not an abstraction, but the value
relations that actually take place?
This is my last post (temporally) prior to the EEA --
Andrew Kliman