[OPE-L:6388] Re: Two rates of Profit?

John R. Ernst (ernst@PIPELINE.COM)
Tue, 31 Mar 1998 12:08:42 -0500 (EST)

>Ajit wrote:
>I found Chris Arthur's post on 'general' and 'average' rate of profit quite
>thoughtful and interesting. From economics point of view I see the matter
>this way:
>John comments:
>I, too, found it interesting. For the sake of greater clarity, I
>wonder, "Which rate of profit has a tendency to fall?" As we know,
>Marx indicated that in his discussion of the falling rate of profit,
>he includes in profit the entire mass of surplus value including

I don't understand what rent has to with this question. As far as Marx is
concerned, the falling rate of profit is discussed before rent. Rent in
Marx's scheme is nothing but a distribution of surplus value between the
capitalist class and the landlords. My superficial impression is that
Marx's theory of rent has got some serious problem and needs to be analyzed
seriously. But in any case, as far as Marx's rate of profit is concerned,
his average rate of profit is the same as the general rate of profit. So
the question of which rate of profit falls is a spurious one.

John now comments:
Of course, I agree that Marx treats the falling rate of profit prior
to his analysis of rent. However, as he discusses his falling rate
of profit, Marx notes that he does so prior to considering components
of surplus value other than industrial profit. Thus, it is this
"overall" rate of profit that has a tendency to fall and not
necessarily the industrial rate of profit. Further, as he discusses
rent, as I recall, he does state that it forms a growing portion of
surplus value as capital accumulates.

>Ajit wrote:
>Marx's surplus value and the rate of surplus value cannot be determined
>without taking the 'whole of capital' (actually only all the basic goods
>sectors) into account, since to determine surplus value you need to
>determine the necessary labor-time and that can only be done by taking all
>the basic good sectors into account. So before Marx could move to discuss
>profits and rate of profit he had to take capital as a whole. The problem
>Chris Arthur is confronting from philosophical point of view is exactly the
>transformation problem. Marx thought that the determination of surplus
>value derived from capital as a whole can be used directly to deduce the
>'average rate of profit'.
>John comments: Then why did Marx consciously exclude parts of
>surplus value and the corresponding capital from the transformation
>process? That is, why are monopolies "abstracted from"? Given that
>they are, to proceed from the average rate of profit to the general
>rate of profit is equivalent to abstracting from monopolies. Hence,
>it would seem that Marx was aware that the general rate of profit
>could not be used to "deduce the 'average rate of profit.'"

First of all Marx does not proceed from the 'average rate of profit' to the
'general rate of profit'. It is a mistake to think that Marx's 'average'
rate of profit, which is nothing but the general rate of profit, is derived
by averaging given empirical rates of profit. Not at all. He derives the
general rate of profit by taking the whole of surplus value produced and
the whole of capital invested, i.e. S/(C+V). This general rate of profit
then is taken to be the 'average' rate of profit that must prevail when the
rate of profit is equalized across sectors. I don't think that the
question of abstracting from monopoly is of much relevance here. In
understanding the transformation problem, it makes more sense to think in
terms of sectors rather than firms-- monopolies to a large extent is a more
appropriate category for a theory of firms. And as Dumenil and Levy's book
on the rate of profit has shown, the classical proposition that there is a
strong tendency for rate of profit to equalize in the long term context is
quite robust empirically. In any case, let us suppose that the existence of
monopolies in certain sectors do not allow the rate of profits to equalize
across sectors. This in itself would not change Marx's average rate of
profit, since it cannot change the total surplus value produced and the
total capital invested as calculated by Marx.

John now comments:

1. I assume that your S/(C+V) includes what Marx calls "natural
monopolies." If so, how could this rate of profit equal the
average rate of profit?

2. I think Dumenil and Levy's book raises an interesting question --
Why would the rate of profit tend to equalize? They are well
aware that capitalists invest according to anticipated rates of
returns to investment and not rates of profit.

>For now, allow me a pass on the basic vs non-basic bit.
>Ajit continues:
>However, as Bortkiewicz pointed out, this was a
>mistake. Prices of production and the rate of profit must be determined
>simultaneously. Marx's average rate of profit derived from capital as a
>whole and treated as equivalent to the 'general' rate of profit would, in
>general, be incorrect, i.e. the 'average' and the 'general' are not the
>same. But one should keep in mind that Marx, in his time, had no way of
>determining prices and the rate of profit simultaneously. Given this
>handicap, the only way he could proceed, given that Ricardo's LTV had to be
>rejected, was to sequentially go from value to 'average/general' rate of
>profit to prices of production. Of course, Sraffa becomes extremely
>relevant in thinking through this problem. Cheers, ajit sinha
>John comments:
>1. Try taking monopolies out of the transformation picture. This
> would mean, among other things, no absolute rent. If you
> hold fast to the claim that the entire amount of surplus value
> is depicted in the transformation procedure, then absolute
> rent must be ignored. If you want to analyze a society where
> there is private ownership of the means of production, absolute
> rent would seem to be a useful concept.

I don't understand what you are getting at here. By absolute rent, do you
mean ground rent? If so, then I agree that Marx's theory of ground rent is
not very satisfactory. But if you are saying that monopolies in
manufacturing sector gives rise to some kind of 'monopoly' or 'absolute'
rent, then I would need you to elaborate this idea a bit before I could
comment on it.

John now comments.

In Sraffa's work, we see only differential rent. In Marx, we see both
absolute and differential rent. If part of surplus value in your
S/(C+V) is absolute rent, then how can the surplus value in those
sectors be "shared" with the capitalists in other sectors? To be
sure, this assumes that sectors in which absolute rent is present
have a lower composition of capital than average.

>2. Let's see if what you say works when fixed capital and technical
> change occur. This is no criticism of Sraffa who recognized the
> limits of his efforts, but an objection to the claim that the
> Sraffian framework can be used as a basis for "correcting" Marx.
> Can we even execute the usual transformation procedure given
> fixed capital? Would not its durability depend upon the wage
> rate actually paid? That is, if we consider technical
> change and its effect on the economic lifetime of fixed
> capital, we are forced to take into account not the values
> but the prices of production, given they represent the
> relative prices in effect. So you do not know how long a
> machine will last without knowing prices; but to find the
> prices you need to know how long a machine will last --
> economically. This is an endless loop all too often
> broken by the heroic assumption that there is no
> technical change and that machines never become obsolete.
> Thus, in correcting Marx, we are forced to assume that
> technical change does not take place. This seems more
> like J.S. Mill's "stationary state."
> Let me be a bit more specific. If we begin the transformation
> process by assuming a structure of production in physical terms,
> how do we know how long the machines will last? Do we assume
> as Schefold does that the machines will simply wear out and
> never become obsolete before they do? If so, what is the
> justification for this assumption?

John, Take a look at Schefold's book: MR. SRAFFA ON JOINT PRODUCTION AND
OTHER ESSAYS. Particularly, Part II, chs. 18, 18a, and 18b. This might be
helpful to you. The way you pose the problem is simply intractable, since
the rate of technical change and its impact on labor productivity and
prices of production cannot be predicted, we have no theory for it yet.
Thus the situation is of total uncertainty. In this situation how can you
solve for economic depreciation of fixed capital in exact manner? Your
comment that "thus, in correcting Marx, we are forced to assume that
technical change does not take place", is simply unfair. In Marx's own
solution to the transformation problem there is no technological change.
Moreover, the assumption of no technical change has nothing to do with
"stationary state". Stationary state means that the rate of growth of
output as well as population is zero. You can always have a growing economy
with no technical change. Cheers, ajit sinha

John now comments:

The intractability may not stem from my thinking but from the problem

1. I accept what you say about the stationary state and will check out
Schefold's book. However, the key issue here is the goal of the
analysis. Are we trying to determine a set of prices only in cases
where there is no technical change? Or, are we trying to describe the
"economic law of motion of modern society"? If it is a set of
prices we seek to derive, then one could abstract from technical change
and, with a few more assumptions, carry out the derivation. To
say that Marx did this is a stretch. In other words, I do not
know where in his transformation procedure he claims there is no
technical change.

2. For Marx, capitalists are well aware that technical change
takes place as they invest. Indeed, they know that the economic
lifetime of fixed capital is less than its natural lifetime.
In what I have read of Schefold, the natural lifetime is all
we find. Yet, capitalists invest and compute their rates of
return based upon the economic lifetime of fixed capital. How
can we compute that lifetime? We need to know both the prices
of production and the rate of technical change. You're right
the problem does, indeed, become "intractable" but only if we
insist on deriving prices of production. However, if our
object is to explain the phenomenon of prices of production,
to do so we need not derive them.

The derivation of prices from values or from a structure of
production is deadly. We can easily generate cases in which
all in a given society see a falling rate of profit while the
derived price system tells them it is rising. To assume that
there is no technical change for the sake of deriving such a
price system is, at best, obscurantist. As economists, we need
a notion of prices that can be used as we take into account
technical change not one that forbids from even considering it.

Stay well,