[OPE-L:7255] [OPE-L:780] Re: TSS and the Okishio Theorem

Andrew Kliman (Andrew_Kliman@email.msn.com)
Sat, 27 Mar 1999 12:51:46 -0500

A reply to OPE-L 742.

I very much appreciated John's response (OPE-L 751). Some of the
following repeats what he wrote.

Duncan writes: "1. In an economy with long-lived fixed capital,
labor-saving technical change will lower the value of existing
capital by reducing the prices of
commodities, and also possibly by reducing the cost of replacement

I'm afraid I don't fully agree with this. Duncan may think the
following is a "semantic difference," but semantic differences mark
conceptual differences. Thus, keeping firmly in mind the
contradiction between value and use-value in Marx's theory, I would
say that the existing capital-value advanced is not lowered by
technical change, at least not directly. What is lowered is the
value of the material elements acquired by means of that advance,
i.e., the value of "capital goods."

I concur with John that this is synonomous with a fall in the cost
of replacement "capital goods." Let me also note that prices may
not in fact fall. This means that the MELT rises. Yet because a
rising MELT indicates inflation in the monetary expression of value,
it remains true that the *value* of "capital goods" is reduced.

Duncan: "This "devaluation" of existing capital reduces the wealth
of capitalists, and, if they have financed their investments by
borrowing, can threaten them with insolvency, thus creating
disruptions in the circuits of credit and the financing of

I agree with this, though I think John was more precise in restating
"wealth of capitalists" as the "value of the capital they possess."
I also concur with his point that "I don't think you have to move to
borrowing to see how devaluation can
be a problem for capitalists."

Duncan: "2. Labor-saving technical change, given a constant real
wage, raises the
profit rate on new investments. I don't think there's actually any
disagreement on this point, either. In the TSS literature this point
is acknowledged through the analysis of the "commodity" profit rate

I strongly disagree with this. It follows from the TSS
interpretation of Marx's value theory that the profit rate can fall
(as a result of labor-saving technical change, given a constant real
wage) EVEN IN THE ABSENCE OF FIXED CAPITAL. Note that, when there's
no fixed capital, the profit rate on new investments is identical to
the general profit rate. The general rate can fall. Hence, the
profit rate on new investments can fall.

The "'commodity' profit rate" (material rate, physical rate) is NOT
the actual, realized profit rate, even on new investments. The
former is essentially the same as the simultaneist (replacement
cost) profit rate. They differ from the actual rate, in part
because input and output prices differ. Moreover, the time paths
can diverge. Imagine, for instance, that, in a one-sector economy,
the input-output ratio falls continually. (Real wages are included
in input.) The "commodity" profit rate, identical here to the
simultaneist rate, thus rises continually. Yet if the ratio of the
output price to the input price is falling at an ever-increasing
rate, and the ratio is sufficiently small, then the actual profit
rate can fall continually.

Assume, for instance, a one-good economy, without fixed capital or
wages, in which all output (X) is reinvested as new input (K), and
output = input grows according to

X(t) = K(t) = m + n*exp(at),

a, m, and n are positive constants. Assume that the time-interval
between input and output is 1. The commodity profit rate =
simultaneist rate is then

CPR(t,t+1) = X(t+1)/K(t) - 1 = (exp(a) - 1)*n*exp(at)/(m +

Its initial level is CPR(0,1) = (exp(a) - 1)*n/(m + n) and, as time
proceeds, it rises monotonically and approaches (exp(a) - 1).

According to the TSS interpretation of Marx's theory, the unit value
(V) of the commodity, in labor-time terms, is given as

V(t+1)*X(t+1) = V(t)*K(t) + L(t)

where L is living labor. Since X(t) = K(t), X(t+1) = K(t+1), we

V(t+1)*K(t+1) = V(t)*K(t) + L(t).

Denoting V(t)*K(t) as C(t), this becomes

C(t+1) = C(t) + L(t).

Assume that L grows according to

L(t) = L(0)*exp(bt)

where b is a constant smaller than a. Then the solution to C(t+1) =
C(t) + L(t) is

C(t) = H + (exp(b) - 1)^(-1)*L(0)*exp(bt); H = C(0) - (exp(b) -

The labor-time profit rate is

LTPR(t,t+1) = L(t)/C(t) = L(0)*exp(bt)/(H + (exp(b) -

Its initial level is LTPR(0,1) = L(0)/(H + (exp(b) - 1)^(-1)*L(0)).
As time proceeds, it approaches (exp(b) - 1) < (exp(a) - 1). Assume
that b > 0 and that H < 0. Then LTPR falls continually.

Now, factor into the above situation a small amount of fixed
capital, small enough that it does not alter the dynamics. Then, it
follows that the profit rate is still falling because it is falling
on new investments. Also, if the new investments (labor-saving
innovations) are the cause of the declining prices, it follows that
the general rate of profit is falling *because* of labor-saving
innovation. The profit rate falls, not because labor becomes less
productive, but because it becomes more productive.

Matters may well appear to be the opposite. The firms that have
made the new investments may receive a higher profit rate than those
that have not. The profit rate received by the former may indeed
rise continually, despite the decline in the general rate. This is
the source of the capital-fetish, i.e., the notion that capital is
productive, i.e., the theory that rising productivity enhances
profitability. (As his defense of this theory, a prominent Marxian
economist has recently stated that we should just ask any capitalist
what the facts are!) Yet because it is the new investments that are
lowering the general rate, it follows that the more productive
capitals are increasing their profitability *at the expense of* the
less productive. Surplus-value is being transferred. That is all.
The phenomenon of enhanced (individual) profitability due to
productivity growth is merely a form of appearance of its opposite.

Duncan: "3. In real history, real wages aren't constant, but rise
at about the same
rate as labor productivity, giving rise to a roughly constant value
of labor-power."

The real history of the last quarter-century seems to indicate
otherwise. Real wages have fallen in the U.S. despite rising
productivity. With a few exceptions, the decline in real wages in
the Third World seems to be even more drastic. Since the tendential
fall in the profit rate is continually being negated by means of
crises, it doesn't seem useful to try to analyze the profit rate in
any time-frame longer than this.

Duncan: "Marx analyzes the falling rate of profit on the basis of
the hypothesis of a constant value of labor-power, not a constant
real wage. When the value of labor power is constant, technical
change that is labor-saving and capital-using will lower the rate of
profit on new investment."

Again, I have to disagree with this. The first sentence is a quite
partial, one-sided account. Marx sometimes holds the rate of
surplus-value constant in order to illustrate the law of the falling
profit rate. It is not true that the law *depends* on a constant
rate of surplus-value.

He says the profit rate must fall even if workers were to live on
air. Elsewhere, he says that the profit rate must fall even if the
full 24 hours of the day were to be appropriated by capital. Still
elsewhere he says that nothing is more absurd than to explain the
fall in the profit rate on the basis of rising wages.

A correct interpretation must correspond to the original theory. (I
don't think there's really any disagreement about this.) In the
original theory, the profit rate falls *because* productivity rises,
and it would fall even if wages were zero. (I don't think there's
disagreement here either.)

It is true that, when the value of labor power is constant,
technical change that is labor-saving and capital-using will lower
the rate of profit on new investment. Yet, given a zero real wage,
such technical changes would not be viable. I agree with John that
this sort of technical change is Rube Goldberg-like, and not part of
Marx's law of the tendential fall in the profit rate. I also agree
with Robert Brenner that to predicate a fall in the profit rate upon
it is Malthusian, and that Marx was fiercely anti-Malthusian. Thus,
while one can derive a fall in the profit rate on this basis, this
fails to vindicate *Marx's own* law of the falling profit rate or to
refute the Okishio theorem.

Given that one theory holds that productivity increases tend to
raise the profit rate, while the other holds that they tend to lower
it, I don't think there can be much "common ground of agreement
about the substantive issues." The fair thing to do, it seems to
me, is to give Marx's own theory a place at the table alongside
opposing (Marxist and non-Marxist) theories.

Andrew Kliman