[OPE-L:1457] Re: Temporality vs simultaneity

Bruce Robert (broberts@usm.maine.edu)
Tue, 12 Mar 1996 06:33:46 -0800

[ show plain text ]

A reply to Massimo"s comments (#1422) on my previous post (written before
his recent exchanges with Riccardo that just came up on my screen, but
hopefully still relevant).

Massimo says:

>****IF***** we accept Bruce's notion and (consequent assumptions)
>of temporality, then Bruce is absolutely right when he says that the
>sequential example I used converges to the Simultaneous Solutions. However,
> is there any reason to accept Bruce's assumption? Bruce's notion of
> (and I am afraid not only his) is not only that after period two comes
> period three (obvious enough), but that while there was some qualitative
> change between period one and two, no such a change occurs between
> two and three. That is, Bruce's notion of temporality is a iterative notion.
> I want to stress the fact that THE TA AND SA solutions converge
> ONLY if THIS assumption is made, which is at the basis of Bruce's
> criticism. Fine, but in principle between period 2 and 3 anything can happen.

I stated no assumption about what happens after period 2 because I made no
assumption about it. What I did, from my own perspective, was to respond
to Massimo's stated intent in his original post (#1357): "I want to find
out what are the insights of the SA and TA in studying the effect of a
reduction in working time." To answer *that* question, no reference is
needed to the myriad other changes that might occur as time passes. I
grant the obvious--that in capitalist economies nothing stands still--but I
also think that introducing other changes simply obfuscates the answer to
the question that presumably was the focus of the discussion. It's
perfectly legitimate to pose as a general problem the modeling of multiple
and overlapping dynamic responses to multiple and overlapping changes, but
if you say you're interested in the effect of *one* change, as Massimo
initially said he was, then all the other things that might happen in
addition are beside the point.


> The point I am trying to make is that whatever we are saying is happening
> between period 2 and 3 we are FORCING A CLOSURE - for the sake
> of the argument - on what would otherwise be an open process. So Bruce's
> criticism is based on a particular closure.

Concerning "closure," it depends what we mean. Yes, there's closure in one
sense and one only: if you want to see the effect of the change in
worktime, you look at *it*, and you don't worry about all the other effects
that may proceed from other causes. If that's closure, then so be
it--that's what theoretical analysis does (and there are any number of
examples of Marx doing just that--"closing out" other changes to focus on
the consequences of one which is given discursive priority: absolute
surplus value, relative surplus value, etc.). But in no sense does that
theoretical tactic commit one to some sort of grand "closure" that denies
the reality or possibility of other events.


>2. To repeat the same thing: I am not "forcing" any "one-period lag"
>in evaluating capital. I am simply evaluating it in that period.

I don't know how to interpret this except to say that Massimo must be using
language differently than I do. By definition, the TA equations impose a
one-period lag in evaluating capital. That's what the time subscripts do
(erase them and you've got a simultaneous model). It never occurred to me
that this was something that could be argued about.


> I refuse the
> argument that "the results in period 2 are only a *partial slice* of the full
> effects of the change being analysed." In fact, the "full" effect depends on
> 2 things: 1. passing of time; 2. what happen to the conditions of production
>in that time. 2. depends on specific assumptions. Assuming that 2. is defined
>by a "nothing happen argument", does not make Bruce or anybody else entitled
>to say that THEREFORE they (any simultaneists) have shown the full effect
> of sequentialism is = to simultaneous solutions.

Here is the heart of our difference. Invoking the inevitability of other
changes doesn't get rid of the effects of the first one, it just adds more
to the problem of analysis. But the initial problem remains: what happens
to values, surplus values, exchange relations, and profit in response to
the change Massimo posed. The TA equations guarantee that the capital that
figures into the calculation of period 2 production prices *can't change*;
therefore, no matter how much other stuff happens in the interim or later
on, a one-period analysis is incomplete, unless one is somehow prepared to
deny that the initial event has any effect at all on the magnitude of
capital. To state my view again: this is a case that I think Marx would
regard as failed generation of relative surplus value. Productivity rises,
which *could* raise the rate of exploitation and the rate of profit, but
the workers who still have jobs get real wages that rise in proportion to
the productivity gain. So there is no rise in r, but no decline
either--what does happen, though, is that values and production prices
fall, and the value of capital falls along with them, because with higher
productivity the good represents less labor-time. (An aside to John Ernst:
this is why I'm not particularly troubled by the seeming "disappearance"
of constant capital. Constant capital doesn't have to disappear in money
terms--that depends on what's happening to the monetary expression of
labor--but it sure does have to represent less labor-time after a change
like this, and the TA equations just slow that effect down through time.)

I don't think this case has anything to do with a falling r a la vol. 3
(c/v doesn't rise here even in the TA calculation). Does Massimo really
want to say that capital is less profitable or that workers are less
exploitad here? Do others who favor this approach? If so, I'd like to see
a case made for it, without introducing other changes as a defense. I
don't believe I made any sort of "nothing happens" argument--I simply
stated what the TA equations themselves imply through time--but from my
perspective, Massimo's reply here sounds like "all kinds of stuff happens,
so (after period 2) the initial event has no consequences that Marxism can
say anything about."

There's another question I'd like to ask the TSS or TA proponents: if
changes through time are the heart of the matter, why is a sequence of
time-subscripted numbers based on a *uniform r* of any particular interest?
As I read Marx, the period-to-period dynamic adjustments in a capitalist
economy are crucially and unavoidably the result of responses to
continuously unequal profit rates, since market prices (the ones that
actually are there to be seen in any one period) are *never* equal to
prices of production, except by fluke, or (maybe) on average. If we're
going to take time seriously, why should every period have a uniform r?

I ask this because an SSS solution, at least as I understand it, is not at
all in conflict with this, since it is not a solution for prices in any
particular time period. Marx, as I read him, is no more interested in a
Garegnani-type "long-period equilibrium" than I am. The simultaneous
equation production prices I favor are a "structural" abstraction, not a
temporal one--they define, in labor-time terms, the exchange ratios
consistent with capitalist equivalent exchange on the basis of current
conditions, so that when current conditions change, the effect (if any) on
production prices and values is necessarily instantaneous. All the lags in
adjustment are then part of the time path of market prices and realized
profit rates, which is a worthwhile subject for investigation, but a
different and necessarily concrete question. Do you really think that
there's any set of general equations we can write down, without explicit
demand functions, to describe actual period-to-period adjustments?


Bruce B. Roberts
Department of Economics
University of Southern Maine
Portland ME 04104-9300
(O) 207-780-5503
(H) 207-772-7047
fax 207-780-5507-------------------------------------------------