[OPE-L:2383] How many periods

Alan Freeman (100042.617@compuserve.com)
Tue, 28 May 1996 02:08:24 -0700

[ show plain text ]

I'm slightly puzzled by Fred's #2368 where he re-iterates the following:

"KM argue that the transformation of values into prices of production
takes place over a number of successive periods"

In my 1905 of 23/04/96 I wrote a response which I thought would clear
this up. It seems to have escaped Fred's attention, so I'm reposting
it. Since this time Riccardo, who by no means agrees with us on other
matters, has made the same basic point in #2092.


Repost of #1905 of 23/04/96

Fred's OPE 1850, discussing the sequential treatment of transformation,
is very timely, very welcome, and very sympathetic. I am extremely
happy to respond to the points he raises and I hope this marks a
new phase in the exchange of ideas between two theoretical trends
(nondualism and sequentialism) which I have long believed belong in
a common school of thought.

The most basic and simple point of all, which is a straightforward
misunderstanding and may clear out of the way many of the
obstacles to progress in this debate, is the following point:

" the transformation of values into prices of production DOES NOT
PERIODS, which are understood to be real, historical periods. In
KM (1988), the transformation process takes place over 14 periods."

This is a straightforward misunderstanding. In all the sequential
nondualist treatments in existence (Husson/Perez 1980, Carchedi 1982,
Kliman and McGlone 1988, Freeman 1991) values are entirely transformed
into prices in every period.

The immediate result of production in each period is a new set of
output values (price of the elements of constant capital plus the value
added by direct living labour) which in general will differ from output
prices because goods are not sold for their values. These output values
are thus entirely transformed into output prices in each given period.
If the output of Department I has a value of $12bn and sells for $16bn,
and the output of Department II has a value of $20bn and sells for $16bn
then the values [ $12bn, $20bn] have been transformed into the prices
[$16bn, $16bn]. If moreover the money required to replace the used-
up constant and variable capital is $8 in each sector, then the surplus
value of 12-8=$4bn in department I is transformed into a profit of
16-8 = $8bn; conversely the surplus value of 20-8= $12bn in department II
is transformed into a profit of 16-8 = $8bn.

Total value output of the period = $32bn = total price output of the
same period

Total surplus value of the period = $16bn = total profit of the
same period

Complete transformation with both equalities satisfied. Tres simple.

The output prices resulting from each successive period *then*
become the input values of the next period, (that is, as we all agree,
the value transmitted to the product by constant capital is equal
not to the value of the elements of capital but the value of the money
paid for them) and a new transformation begins with a new (and in
general different) set of input values producing a new (and in general
different) set of output values that are transformed into a new
(and in general different) set of output prices.

It is only because the simultaneous framework imposes the
restrictive assumption that these output prices must necessarily
be equal to the input prices of the same period, that any
conceptual difficulty arises. Once this restriction is dropped
the answer becomes extremely simple and extremely general.

The confusion often arises because of the belief that the real,
disguised purpose of the Kliman/McGlone procedure is to arrive
at the equilibrium result iteratively instead of directly. I
suspect that it is because Fred takes this to be the 'real goal'
of the process that he concludes the transformation to be incomplete
until this equilibrium has been attained.

But this is just not what the procedure is about and there is
no need for it to converge at all.

It is a very common misconception to view their procedure as 'just
another way of reaching equilibrium' but if we drop the restrictive
and unnecessary assumption of constant technology (to which they
do not hold), there *is* in general no convergence and no equilibrium.
Nevertheless Marx's transformation remains fully justified and fully

In 1993 for example I showed that Marx's transformation procedure yields
the two equalities *regardless* of whether profit rates are equalised,
that is for any arbitrary output prices, and also for any arbitrary
input prices. As with Kliman and McGlone, the transformation is
complete in each period. This result is demonstrated in our book and
in my 1995 Capital and Class article. I repeated the demonstration
for the British Economy in the discussion paper I circulated at the
EEA mini-conference.

None of the approaches above 'complete' the transformation over a
long period of time. On the contrary, it is precisely because the input
prices of each period differ in general from the output prices of the
same period that the transformation *must* be completed in each
period; otherwise how can one get from one set of prices to the next
set of prices?

As regards the textual evidence: first, whatever Marx says about *value*,
I am certain that Fred will absolutely nowhere find any citation
whatsoever which even remotely suggests that Marx thought *prices*
remained constant throughout a period.

Marx recognised that input prices may differ from input values
and, we agree, showed how this can be taken into account; but nowhere
is there the remotest suggestion that this involves equating these input
prices to output prices.

It is not Marx who introduced this idea but von Bortkiewicz/Dmitriev/
Tugan. It seems to me that the boot is on the other foot; to provide the
textual evidence that the Bortkiewicz/Dmitriev/Tugan question is in any
sense a reasonable statement of the problem or bears any resemblance at
all to the question discussed by Marx. The Bortkiewicz/Dmitriev/Tugan
problem is entirely of their own making. We should leave them to it.
Of what concern is it to us? Or Marx?

There is a more complex discussion about the movement of *values*
brought by technical change which we have already engaged in to some
degree, but it is independent of whether input *prices* should equal
output *prices*.

However I am very happy to supply several passages where Marx clearly
discusses changes in value from one period to the next and distinguishes
input from output values (and prices) very clearly; for example Theories
of Surplus Value III p340-343. In fact in every case where he discusses
the effects of fluctuations in price and value resulting from changes
in the harvest, he invariably assumes:

(a) that seed corn in the second period is purchased at the
prices of the end of the first period

(b) that the value transmitted by seed corn to the product is the
value which this seed corn acquired from the harvest in which
it was produced.

But for the present I think it would be the most useful course of action
to see if we can reach agreement that the TSS procedures, from all authors
who endorse it, complete the transformation in *each* period. Hopefully
this will also remove some of Fred's concerns.