[OPE-L:7511] [OPE-L:1048] Re: Marx's Concept of Prices of Production

John Ernst (ernst@PIPELINE.COM)
Sun, 27 Jun 1999 13:34:12 -0400

RE: Fred's 1047
Hi Fred,

You wrote:

I have argued that Andrew and Ted's interpretation of Marx's concept of
prices of production is a misinterpretation because Andrew and Ted's
prices of production change every period, even though productivity and the
real wage remain constant, whereas Marx's prices of production change if
and ONLY IF productivity or the real wage changes.

My comment: Before we get into the "over how many periods thing", let's
clear up one thing. You continue to impute to Marx the notion that
real wage changes are one of two possible causes of changes in prices of
production. This is simply wrong. In Chapter 11, Marx briefly
discusses cases in which changes in the value of the means of subsistence
can cause the prices of production to change. It's clear that the
commodities whose values fall are not necessarily those sold at prices
of production; they seem to be products of sectors in which rent is
earned. The surplus value produced in those sectors does not enter
into the equalization of the profit rate.

The implication of all this is clear. The rate of profit used to
determine prices of production is not the same as the one which
has a tendency to fall nor is it the one which Marx calls "the
rate of gross profit." (Chapter 50, V 3) If one is to claim that
Marx's prices of production are equilibrium prices, one would
have to incorporate that "rate of gross profit" and its relation to
the rate of profit that is equalized in the transformation procedure.
Without that, it's unclear how those long-run equilibrium prices are
even equilibrium prices.

You wrote:

John's first defense of Andrew and Ted's interpretation was to argue that,
although Andrew and Ted's prices of production do not change as a result
of a change of productivity during the periods analyzed, they do change as
a "delayed effect" of a change of productivity prior to the periods
analyzed. I have argued that there is no textual evidence to support this
interpretation and that it contradicts Marx's logical method of first
determining prices of production and then analyzing the causes of changes
of prices of production. Nothing is said in this most recent post about
this first argument.

My comment: Let me see if I can be a bit clearer.

1. Given that Ted and Andrew are starting with inputs whose prices
are the "actual values", then at the end of the first period we have
outputs which are prices of production. At this point, I'm not sure
if you have any objection to their effort. But perhaps you do.
Clearly the unit output prices which are prices of production differ
from the unit input prices which are actual values.

If the purpose of Ted and Andrew's effort were simply to demonstrate
Marx's transformation, in my judgment they could have stopped right
here. But to answer Bortkiewitz they assume that material structure
of production does not change as reproduction takes place.

2. Given that the outputs of that first period are used as inputs
in the next period, the input prices of that next period will differ
from the input prices of the first period. In the second period,
there is generally a difference in the unit input and unit output
prices. Here then we finally reach the point where the effect of
a wage change on prices of production can be examined.

3. Generally, in each period that follows save the one in which input
prices equal output prices, the wage does change. In no way does
this represent Marx's discussion of Chapter 11 where the prices of
the means of subsistence are assumed constant. Yet I see no reason
why we cannot consider what happens as we move from period 1 to n.


Now if the input prices of that initial period are assumed to be equal
to the actual values, then we can ask what went on in the prior period.
Clearly, the rates of profit were not equal in that prior period. It's
difficult me to imagine how these unequal rates of profit came about
without any change in productivity.

On the other hand, if those initial input prices are equal to not only
the actual values but also the prices of production of the previous
period, then it is obvious that as we begin the initial period a change
in productivity has taken place. In either of these two cases,

(1) initial inputs are actual values but not prices of production


(2) initial inputs are actual values and prices of production,

it's difficult to imagine that some change in productivity has not
taken place. That is, in (1) you would somehow have a prior period
with unequal rates of profit. How that came about without any changes
in productivity is unimaginable. In (2) the change is clear.

You wrote:

In this post, John presents a new defense of Andrew and Ted's
interpretation. The new defense is to argue that Marx's statements that
prices of production change if and only if productivity or the real wage
changes APPLY ONLY TO ONE PERIOD of time. These statements DO NOT APPLY
TO MORE THAN ONE PERIOD of time. The implication seems to be that, when
we consider more than one period of time, there is a THIRD CAUSE of
changes of prices of production: because input prices are not equal to
output prices, the equalization of profit rates in the next period
requires that prices of production change again in the next period, even
though productivity and real wages remain constant, as in Andrew and Ted's


I do not think that this is a reasonable interpretation of Marx's concept
of prices of production, for the following reasons:

1. Marx never said anything remotely like this. In his discussions of
causes of changes of prices of production, Marx never once mentioned this
possible third cause. Marx never said anything like: "These are the two
causes of changes of prices of production WITHIN A GIVEN PERIOD, but OVER
MORE THAN ONE PERIOD, there is a THIRD CAUSE of changes of prices of
production: the inequality between input prices and output prices." This
third possible cause of changes of prices of production is completely
without any explicit textual support.

My comment:

Since Marx never used the concept of simple reproduction nor that of
expanded reproduction in his discussion of prices of production, it
is hardly surprising that we find no "explicit textual support." He
did however assume that the effects of the changes are not immediate.

If the effects of a wage change that changes the rate of profit take
two years to be felt in all sectors and I am examining the economy on
yearly basis, then you seem to be saying I'm out of luck for that first
year. There are no prices of production at that point in time.

Marx's discussion in Chapter 11 dealt with one period. My
question to you has been and continues to be -- "What do you think
happens in the next period?" Let's go along with Marx and look at the
case where the wage increases and the prices of the means of subsistence
are constant. Would not the output prices of production of other
commodities change? Would not these be the input prices of the next
period? Why would we not see still further price changes in
that next period?

I suspect that the reason these questions have remained unanswered
is that you claim that a period is the length of time necessary
to bring about a set of prices that are equilibrium prices. Hence,
the next period is irrelevant to your argument. But in Chapter
11, Marx focused upon a wage change. In most of his examples,
no changes in value occur. Given that wage change and the
initial prices of production, do not the prices of production
of the outputs generally differ from those inputs?

You wrote:

2. In several of Marx's discussions of causes of changes of prices of
production, Marx explicitly discussed changes OVER MORE THAN ONE PERIOD of
time. These passage make it clear that Marx's discussions of the causes
of changes of prices of production apply BOTH to changes within a given
period AND also to changes over several periods of time, not just to the

3.1 In a discussion from Chapter 9 of Vol. 3, Marx stated:

For all the great changes that constantly occur in the actual
rates of profit in particular spheres of production (as we shall later
show), a genuine change in the general rate of profit, one not
simply brought about by exceptional economic events, is the
final outcome of a whole series of protracted oscillations, which
require a good deal of time before they are consolidated and
balanced out to produce a change in the general rate. In all
periods shorter than this, therefore, ... a change in prices
of production IS ALWAYS TO BE EXPLAINED prima facie by an actual
change in commodity values, i.e. by a change in the total sum of
labor-time needed to produce the commodities. (p. 213;
emphasis added)

Nothing is said here about this statement only applying to one period of
time. Indeed, the discussion is about a fairly long time horizon. Marx
does not explicitly identify periods, but "a good deal of time" seems to
imply more than one period. During this "good deal of time," a change
of prices of production IS ALWAYS TO BE EXPLAINED by changes in
productivity. After this "good deal of time", changes of prices of
production are also changed by changes in the rate of profit, in
addition to changes of productivity in the production of the given
commodity. But never in this discussion of changes of prices of
production over a fairly long time horizon did Marx say anything about a
possible third cause of these changes (because input prices are not equal
to output prices).

My comment: As I first read the passage you cite, I thought Marx was
simply contradicting himself. That is, could not a wage change be
responsible for changes in prices of production? I kept reading and
found that Marx added to the remark you cite:

"Alternatively, the general rate of profit can change, with the value
of the commodities remaining constant, if the level of exploitation of
labour changes." (p. 266-67; note that in my text the passage you
cite is on p 266 not 213)

This then takes us back to the same dilemma. Do prices of production
exist? If so, when? What you seem to be saying is that until
the system reaches an equilibrium there are no prices of production.
If I want to examine the capitalist system on an annual basis, computing
prices of production would be less than useless since within every
year there would generally be some changes in productivity and, perhaps,
in wages. Yet, for you prices of production only exist when the system is
in equilibrium. Since this never really occurs, prices of production
never exist.

This same result can be seen when we examine your idea of short-run
and long-run equilibrium prices. You'd agree that what you call
the short-run equilibrium prices may well be the same as those
that Ted and Andrew compute for any one of periods they depict.
You argue that these are not Marx's prices of production since Marx's
only exist as long-run equilibrium prices. In itself this seems
but a pedantic criticism of Ted and Andrew. But there is something
more to this or none of us would bother with it. We could call
their short-run prices of production -- pre-prices of production --
and refer to the long-run prices of production as the real prices
of production. The definitional differences concerning adherence
to Marx disappear.

But then what? How in time does one move from those pre-prices of
production to prices of production? For Ted and Andrew, the answer
is simple; the system does not change in material terms from period
to period. Insofar as they are simply dealing with criticisms of
Marx, this will suffice. But we know that this is but a mere
hypothetical case and clearly unrealistic. Thus, we still face
the problem of moving from those pre-prices of production to prices
of production in a system which depicts capitalism as a system that
does more than simply reproduce. Let's note that in reality the
set of prices that exists at the beginning of a production period
will not be the same set at the end. Hence, in reality all we would
ever see or would be able to compute are pre-prices of production.
We could only reach those long-run equilibrium prices by assuming either
simple reproduction or expanded reproduction with no technical
change. Without such assumptions the pre-prices of production would
never lead to long-run equilibrium prices of production.

Thus, Marx's prices of production seen as long-run equilibrium prices are
rendered useless as they never exist anymore than long-run equilibrium
prices exist. Your critique of Ted and Andrew becomes the dismissal
of Marx. Or does it? It seems to me you need to show how those
long-run equilibrium prices come into existence period by period.

You wrote:

3.2 From a discussion from Chapter 10 of Vol. 3:

In whatever way prices [of production] are determined, the
following is the result:
(1) The law of value governs their MOVEMENT in so
far as reduction or increase in the labor-time needed for their
production makes the price of production rise or fall.
(p. 280; emphasis added)

It is not made explicit here, but the MOVEMENT in prices of production
would seem to be MOVEMENT OVER TIME. In any case, certainly nothing is
said in this Chapter 10 discussion about a distinction between changes of
prices of production within one period and changes of prices of production
over several periods, nor anything about input prices not equal to output
prices as a possible third cause of changes of prices of production.

My comment: I'd agree with Marx in the passage you cite. Let's say
that the period of time that you have in mind to bring about a change in
prices of production is 10 years. In the years, from the initial
change in productivity to end of the 10 years, are there no prices of
production? Of course, if within those 10 years another change in
productivity in some sector takes place, then we would have to wait
for, say, another 10 years for the changes to alter the prices of
production. We could go on with this process but I think it is obvious
that the delay in arriving at those equilibrium prices is endless.

You wrote:

3.3 From Chapter 50 of Vol. 3 ("Illusions Created by Competition"):

Market prices rise above these governing production prices
or fall below them, but these fluctuations balance each other
out. If one compiles price lists OVER A PROLONGED PERIOD,
and ignores those cases in which the actual value of a commodity
alters as a result of changes in labor productivity, as well as cases
in which the production process is disturbed by natural or
social disasters, it is surprising both how narrow the limits of
these divergences are and how regularly they are balanced out.
(p. 1000; emphasis added)

This analysis has to do with "a prolonged period" and yet the only cause
of a change of prices of production that is mentioned is a change of
prices as another possible cause of changes of prices of production over
this "prolonged period."

My comment: I note that in your comments Marx's "prolonged period
of time" becomes simply a "prolonged period." Let's look at
what happens in time and not simply in that "prolonged period." If
we assume that Marx was looking at prices listed on a monthly or yearly
basis, he says that save for changes in the "actual values" of
commodities and abstracting from periods of crisis, individual
prices are fairly stable. I have no doubt this was true.

But what does this prolonged period have to with periods of
production and reproduction? Within the prolonged period there
are obviously many periods. For example, we could look at prices
on an annual basis. Given that we begin with a system in which
the rates of profit are more or less equal and with a system
in which absolute rent exists, then changes in productivity in
some sectors have, at most, a muted effect on prices in other
sectors. This is especially true

(1) if the output prices of any one period become the input prices
of the next period.


(2) if the value of the money commodity is changing at the same time
Marx observes those stable prices.

The stability of the prices to which Marx refers leads you to believe
that he is looking at "the center of gravity" of the prices for those
commodities and hence the prices of production. In looking at these, say,
annual prices he knows that changes in productivity are taking place
in sectors other than those that directly produce the commodities whose
prices are stable. Hence, not all prices are prices of production; only
some of them are. Is this a long-run equilibrium?

For Ted and Andrew, a long-run equilibrium is reached only because
they assume simple reproduction in material terms. If they were
to jump from their starting point to that equilibrium, I think you'd
have no objection. Indeed, if we cut out all the period business,
all you need for *the* transformation is that final period in which
input prices before and after transformation are equal to output
prices. Note that this is the only point in time at which equilibrium
prices exist. Somehow the prices at which the inputs were purchased
were already equal to the prices of production of the outputs. Yet,
in the passage you cite, Marx is busy looking at prices and deviations
from prices of production over time; he notes how little they change.
Is Marx looking at equilibrium prices as well as deviations from
those prices? Or, are we now talking about partial equilibriums?

You wrote:

4. Therefore, I conclude that John's second defense of Andrew and Ted's
interpretation is no more successful than the first attempt. Andrew
and Ted's short-run equilibrium prices of production, that change every
period, even though productivity and the real wage remain constant, is a
misinterpretation of Marx's concept of prices of production, which are
long-run equilibrium prices that change if and ONLY if productivity or the
real wage changes.

My comment: See above for more clarification on that first attempt.
Again and again, we see the notions of "short-run" and "long-run" creeping
into your view of Marx. These are concepts that contain no time
dimension at all. Further, as neoclassical terms they "abstract from"
changes in technique as well. Hence we never see what your views
are concerning the transformation process or the process of changing
prices of production in anything resembling real time. Instead we leap
from one set of prices to another as we move from period to period.
The periods in question are not periods of reproduction but are
simply seen as the "long-run." The connection between any two
consecutive periods in time is not simply ignored; it does not exist.
Each and every period simply presents us with a new set of equations
from which those long-run equilibrium prices are derived. Or we
could simply say that the inputs are already transformed
and somehow their prices are equal to those of the outputs.

Yet it is the criticism of Marx that placed the notion of simple
reproduction within Marx's transformation procedure. Rather than
ignore it; Ted and Andrew attempt a direct answer. If you
consider the various sets of prices of production they derive
as short-run and claim that Marx's are long-run, how do *you*
connect the short-run and the long-run? They have one way. It may
or may not be correct. To show that it is incorrect you need to
do more than quote Marx's efforts that focus on either a single
period or on a "prolonged period of time" in which technical change
takes place. Put simply, to better understand your criticism of
their efforts we need further explanation of why we must use
the neoclassical concepts of long-run and short-run to understand
the notion of prices of production. I'm not saying that
the connection itself must be shown by using reproduction schemes.
Anything that works will do.

As matters now stand, it is easy to see why your view is often taken
to be that of a Sraffian. That is, if the given sums of money
capital are used to purchase the inputs, what are or were the prices
of those inputs and how would they differ from Sraffian prices?
My bet is that the relative prices of their system are the same
as yours. Indeed, given that the Sraffian prices are unique equilibrium
prices, this has to be the case. With the right choice of the
numeraire, their prices become yours.