THANK YOU, YOUR PAYMENT WAS SUCCESSFUL.

YOU HAVEN'T YET CHOSEN A PASSWORD; ×

CLICK HERE TO DO SO

NLR I/84, March–April 1974

This article was purchased by [email protected]

GEOFF HODGSON

THE THEORY OF THE

FALLING R ATE OF PROFI T

‘Others apart sat on a hill retired,

In thoughts more elevate, and reasoned high
Of providence, foreknowledge, will, and fate,
Fixed fate, free will, foreknowledge absolute,
And found no end, in wandering mazes lost’.

Milton, Paradise Lost.

M
a r x u n c o v e r e d m a n y causes of capitalist eco-
nomic crisis. 1 It has been traditional, however, to
place his theory of the tendency of the rate of profit
to fall in the centre of the Marxian analysis and cri-
tique of capitalism. Marx’s main exposition appears in the first and
third volumes of Capital. 2 The theory attempts to show that there is
an inbuilt tendency for the capitalist system to stagnate or fall into
crisis, as a result of the falling rate of profit. But Marx did not expect
the rate of profit to decline in a persistent and uninterrupted man-
ner; certain ‘counteracting influences’ would periodically halt the
downward slide. Despite this qualification, the theory has been re-
garded, by most Marxists, as the backbone of revolutionary Marxism.
According to this view its refutation or removal would lead to re-
formism in theory and practice. In this regrettable context we shall
attempt to refute the theory of the falling rate of profit. In addition
we shall argue that revolutionary Marxism is not damaged by the sur-
gical removal of the theory from the theoretical system. On the con-
trary, it becomes possible to extricate completely the fatalistic and
mechanistic interpretations of Marxism that have gained prevalence
amongst both its supporters and its hostile critics.

I The Theory

The General Rate of Profit

The existence of separate capitalist firms creates a tendency for the

rate of profit to be equalized between firms. The more competitive
the situation the more pronounced is this tendency. Capitalist com-
petition, therefore, leads to the formation of a general rate of profit in
the economy. This tendency is even present under monopoly capital-
ism, as capitalism is inconceivable without some degree of competi-
tion and separation between firms. 3
With increasing competition
and inter-dependence we have no reason to suppose that this ten-
dency is dead today.
Marx’s analysis of the falling rate of profit proceeds from this essen-
tial feature of capitalist production. At a given level of abstraction it is
justified to ignore the various frictions and barriers that prevent the
rapid formation of an equilibrium general rate of profit. Marx starts
from the rate of profit in value terms in each firm, i.e. surplus value
divided by the value of the total capital invested. He then treats the
whole economy as a ‘single capital’ 4 and equates the general with the
average rate of profit. Hence, in Marx’s view, the general rate of profit
is the total surplus value in the economy divided by the total value of
capital invested.

Two points are evident here. Firstly, no reason is given to identify the
general with the average rate of profit. Secondly, Marx’s general rate
of profit is a ratio between value amounts, i.e. amounts of socially
necessary labour time. It is not a ratio between prices. Some Marxists
and non-Marxists, such as the Ricardian economist L. v. Bortkiewicz,
have criticized this formulation of the general rate of profit on the
grounds that there is no reason to assume that the rate of profit in
value terms will tend to be equalized. The rate of return on capital ad-
vanced is calculated in terms of prices, as capitalists are not aware of,
or disposed towards, any embodied labour calculation. The general
rate of profit is the ratio between profit and the price of capital in-
vested, as this is the actual rate of profit that is equalized between
firms in the real world. This point of contention relates to the well-
known transformation problem. Several articles exist on this topic
and it is not appropriate to discuss it here. 5

Despite this connection between the transformation problem and the

question of the falling rate of profit it is possible to deal with the lat-
ter without invoking a rejection of Marx’s solution to the transforma-
tion problem. Our critique of the falling rate of profit theory in
Section II is directed at Marx’s formula for the general rate of profit,
as in certain circumstances this coincides with the correct formula
adduced by Bortkiewicz and others—when prices are proportional to
values, for instance. Hence we can avoid the intricacies of the trans-
formation problem, at the cost of a lack of completeness in our
argument.

The following mathematical symbols shall be adopted:

y = net output in value terms.

This is the magnitude of the socially necessary living labour time ex-
pended in the economy in one year. It is part of the value of the
output.

v = variable capital.

The working class receives a number of wage goods in a year. The

amount of socially necessary labour time embodied in these goods is
the variable capital.

s = surplus value.

The workers are compelled to work for the capitalists and produce an
amount of extra or surplus value over and above the value of the
wage goods they receive in return. In other words s is expropriated by
the capitalist class. Obviously, by definition: v + s = y.

c = constant capital flow.

This is the value of the raw materials used up, plus the depreciation
of the means of production, in value terms. Like v and s, c is a flow
variable.

k = constant capital stock (i.e. fixed capital).

Normally certain means of production will remain at the end of a

production or turnover period, and these will have a value k. This
value is not part of the value of the social product that is exchanged
on the market, unless the capitalists sell their machinery. The value
of the goods that are produced in one year is c + v + s.
t = time period of production (i.e. turnover period).

The above variable is not familiar in Marxian literature. It refers to

the length of the period of time that is required to produce, transport
and sell a particular good. In order to simplify our presentation we
shall assume that t is the same for all goods, and that wages are paid
at the start of the time period of production. These may seem to be
extreme assumptions, but our arguments are not invalidated if t is
different for all commodities. On the contrary, our position is rein-
forced in the heterogeneous case.

The capitalist spends his investment funds on three basic types of

commodity: firstly labour power, secondly raw materials and expendi-
ture to cover depreciation, and thirdly fixed capital goods. Their re-
spective values are v, c and k. Now it is important to note that c and v
are flows, i.e. they refer to an amount of labour time per year, whereas
k is a stock item, i.e. it is just a congealed aggregate of labour time, it
is not a rate or flow. The amount k corresponds to the fixed capital
that is required to set up production. But the whole of c and v need
not be advanced at first, if t is less than unity. It is necessary to set up
production for only one time period of production t. At the end of
this period the extra funds that are realized can be thrown into
circulation.

If we assume that the rate of growth in the economy is small then the
amount of c and v advanced will be t(c + v). Otherwise this will be the
average amount of c and v advanced in a year. Hence the total capital
invested has at least an approximate value of

k + t(c + v)

This appears reasonable if the units of the amounts k, c and v are in-
spected. As c and v are amounts of labour time per year they have to
be multiplied by an amount of time, in this case t years, to make
their addition to the stock variable k sensible. Hence the general rate
of profit, according to Marx’s definition, is given by the equation:

where p is the rate of profit in value terms. This expression is so un-

familiar that its basis in Marx’s writings may be contested. In particu-
lar it has been traditional for Marxists to ignore k in their formula-
tion. However, apart from the occasional assumption that k is zero,
Marx repeatedly asserts 6 that the rate of profit must include k. In the
real world the capitalists calculate the rate of profit in terms of total
capital invested. It is quite inadmissible for Marxists to continue to
ignore constant capital stock. Secondly, the introduction of t is novel.
A close inspection of Capital, however, will indicate that the above
formula corresponds to the one implied by Marx and Engels. 7 The
formula will appear more familiar if t is assumed to be unity:

Marx’s Formulation of the Theory

Marx’s exposition of the theory of the falling rate of profit in the third
volume of Capital commences with a numerical example. 8
He as-
sumes that

s = v = 100.

Also Marx implicitly assumes that t is unity. He examines the effect

of a gradual increase in the total amount of constant capital (k + c).
Using Equation 2 we get the following table:
These numerical examples can be generalized in the following man-
ner. Dividing top and bottom of the fraction in Equation 2 by v we get

Marx calls the fraction s/v the rate of surplus value. Now if the latter
is constant and the fraction

increases, as in the above example, then the rate of profit will fall. Or,
in Marx’s words: ‘the gradual growth of constant capital in relation to
variable capital must necessarily lead to a gradual fall of the general
rate of profit, so long as the rate of surplus value, or intensity of ex-
ploitation of labour by capital, remain the same’. 9

Marx’s justification for assuming that the fraction (k + c)/v increases

is supposedly based on a number of related arguments that appear in
various parts of Capital. In one place he sees the increase as a result
of the decrease of v, due to productivity increases. 10
Elsewhere he
sees the increase as resulting from the accumulation of capital. 11 We
shall discuss these arguments at a later stage.
Critics often attack the assumption of a constant rate of surplus value
(s/v). 12
It is argued that rises in productivity, causing a fall in v, will
also lead to a rise in the rate of surplus value. This may compensate
for any rise in (k + c)/v and the rate of profit may not fall. It has been
pointed out that Marx was aware of this difficulty and he attempted to
deal with it. 13
Marx suggested that surplus value per man per day
could rise, but up to a certain limit only. 14 This limit is provided by
the number of hours in a day. But that does not define a limit for s/v.
Increases in productivity may still bring down v, and there is no theo-
retical upper limit for the rate of surplus value. As it turns out, Marx
had a valid but somewhat latent point, which must be extracted by a
reformulation of the theory.

To complete the exposition of the theory of the falling rate of profit it

remains to show that the theory can be reduced to the hypothesis of a
tendency for the organic composition of capital to rise. Marx failed to
give a formal demonstration of this point. This partly stems from a
slight clumsiness in the definition of the basic mathematical ratios in
Capital. Taking account of mathematical considerations, some new
definitions make the formal demonstration simple and short. This
formal demonstration is found in the Appendix. Here we shall con-
fine ourselves to a short verbal summary.

The demonstration rests upon a convenient redefinition of the basic

Marxian ratios. It is possible to abstract from changes in the degree
of exploitation by expressing each ratio in terms of the net output (y).
This does not mean that any variable is assumed constant, or that any
variable is regarded as an exclusive function of net output. It is sim-
ply a method of focusing attention on the determinants of the rate of
profit that do not directly relate to changes in the degree of exploita-
tion, using Marx’s formula for the rate of profit. Evidently Marx at-
tempted to abstract from the degree of exploitation in his exposition
of the theory in Capital.
By means of simple algebra it is easy to show that the rate of profit
cannot exceed an upper bound. This upper bound, or maximum rate
of profit, is equal to the value of net output (y) divided by the value of
the constant capital stock (k). Hence, whatever the degree of exploita-
tion, the rate of profit cannot exceed the magnitude of y/k. The frac-
tion k/y is dubbed the organic composition of capital, as it is argued
that it is quite close to Marx’s category, and it best displays the essen-
tial meaning. So if q is the organic composition of capital then

The theory of the failing rate of profit is thus reduced to the question
of the rise or fall in organic composition of capital. For if q rises then
the maximum rate of profit will fall with it, and the actual rate of
profit will fall if all other variables, including the degree of exploita-
tion, remain constant.

II A Critique

Technical Change and the Organic Composition of Capital

Our attention must now shift to the validity of the supposition that
the organic composition of capital will rise. P. M. Sweezy has made
the following point: ‘In physical terms it is certainly true that the
amount of machinery and materials per worker has tended to grow at
a very rapid rate for at least a century and a half. But the organic com-
position of capital is a value expression; and because of steadily rising
labour productivity, the growth in the volume of machinery and ma-
terials per worker must not be regarded as an index of the change in
the organic composition of capital. Actually the general impression
of the rapidity of the growth of the organic composition of capital
seems to be considerably exaggerated.’ 15

Elsewhere 16 Sweezy argues that Marx’s insistence on an increasing

organic composition of capital stems from the fact that Marx was wit-
nessing the transition from hand labour to mechanized production.
Today we have an already mechanized economy where the problem
for the capitalist is to minimise his expenditure on both means of
production and labour power, whilst increasing his productivity. We
have no reason to suppose that the fall in the organic composition of
capital has continued after the transition from extensive to intensive
mechanization.

M. Blaug 17
and others have focused attention on the possibility of
‘capital saving’ innovations, and their role in lowering, or preventing
a rapid rise, in the organic composition of capital. Whilst a Marxist
would object to the use of the word ‘capital’ in this context, such in-
novations deserve examination. They could fall into two overlapping
classes: those that lead to a reduction in the organic composition of
capital by reducing the value of constant capital stock relative to net
output, and those that lead to a similar reduction in the rate of con-
stant capital flow. Examples of the first class include the more effi-
cient use of machinery and buildings. In the second class is included
the more efficient use of raw materials. The existence of these inno-
vations undermines any notion of the tendency of the organic com-
position of capital to rise.

However, D. S. Yaffe 18 has argued that these innovations cannot be

given a great deal of significance; they must be shown to necessarily
recur. In reply we must ask why innovations causing an increase in
the organic composition of capital necessarily recur? If the physical
and value aspects of accumulation are separated then there is no rea-
son to suppose that technical change will have any particular bias in
the long run. In the chapter on ‘Counteracting Influences’ in the
third volume of Capital, Marx devoted a section to the ‘cheapening of
the elements of constant capital’. He wrote: ‘In isolated cases the
mass of the elements of constant capital may even increase, while its
value remains the same, or falls’. 19

Just as Marx gives an inadequate explanation of a tendency for the or-

ganic composition to rise, he merely asserts that the reduction in the
value of constant capital is an isolated case. Perhaps there is no less
justification to assert that the reduction in the value of elements of
constant capital is the underlying tendency, and the increase in the or-
ganic composition of capital is a counteracting influence or isolated
case?

We are drawn to an agnostic conclusion on the trend of the organic

composition of capital. But it does little justice to Marx, or the
Marxian tradition, to leave matters there. Some commentators have
detected certain theories of technical change in Marx. Such theories
profess to demonstrate that the organic composition of capital will
rise as a result of the process of capital accumulation. It must be ad-
mitted that these theories are quite convincing at first sight: so con-
vincing that similar arguments can be found in mainstream bour-
geois economic theory.

Technical Change and the Concept of Capital

A recent and forceful presentation of Marx’s theory of technical

change has been offered by Yaffe: ‘Marx regarded it as an incontro-
vertable fact, as a self evident or a tautological proposition that the or-
ganic composition of capital should rise. . . . The compulsion to em-
ploy machinery, under capitalist production and to increase by these
means the productivity of labour is expressed in reality by competi-
tion and the consequent need to reduce the cost of production. But
this is not its explanation which must be deduced, in terms of Marx’s
method, from the concept of capital itself. The concept of capital is a
contradictory one. On the one side we have capital as “value in
process” as value attempting to expand itself without limit and on the
other side we have the working population, the limited basis of this
expansion. Capital, therefore, must, on the one hand, try and make it-
self as independent as possible of that basis in its process of self ex-
pansion; it attempts to reduce the necessary labour time to a mini-
mum by increasing the productivity of labour. On the other hand it
needs to increase the basis of its expansion, that is the labour power
available for exploitation; that means to increase simultaneously the
working population. . . . The dialectical solution to this contradiction
. . . is to increase the scale of production through the replacing of liv-
ing labour by objectified (dead) labour in the form of machinery . . .
What we have tried to show from an examination of the concept of
capital is the necessity of increasing the social division of labour,
through the application of machinery and therefore, of replacing on
an increasing scale living labour by objectified (dead) labour. It fol-
lows from this that both the technical composition of capital and the or-
ganic composition of capital must increase in the process of capitalist
production although the latter will not increase as quickly as the for-
mer due to increases in the productivity of labour.’ 20

Yaffe produces a number of quotations from Marx which indicate

that Marx had a similar type of explanation in mind. Also Yaffe gives
an unfortunate pseudo-mathematical ‘proof’ of the necessity of an
increasing organic composition of capital, which assumes what he is
attempting to prove. 21 To criticize Yaffe’s argument we first turn to
the question of productivity increases.

The Problem of Productivity Increases

Yaffe mentions increases in the productivity of labour only. But in re-
ality such increases are more problematic. For instance, as was men-
tioned above, an existing machine can be utilized more efficiently
with the same amount of labour, which means that the amount of
machinery per unit of output is reduced as a result. Technical change
often takes the form of replacing one machine by another different
one. In which case we cannot talk about an increase or decrease in
the mass of machinery, in an economic sense, as we are talking about
heterogeneous objects. And it is quite possible that less embodied
labour in the form of machinery will be required per unit of output.

The increase of productivity is certainly a hallmark of capitalism. As

a result there will be a tendency to reduce the amount of living
labour required for every item of output. But we have no reason to
suppose that the labour embodied in machinery per unit of output
will decrease at a slower rate. The notion that productivity increases
are associated with increases in the organic composition of capital is
without foundation.

The Nature of Capital Accumulation

The second erroneous notion that appears in the quotation from

Yaffe is that the accumulation of capital is, for practical purposes,
synonomous with the accumulation of dead labour, i.e. constant capi-
tal. An accomplished Marxist like Yaffe is, of course, aware that capi-
tal is not just a thing but a social relation. Nevertheless, the habit of
confusing social relations with things is a fundamental, albeit dis-
guised, error found in the canons of over-zealous interpreters of
Marx. Exactly the same error is found in neoclassical economics: the
dominant school of bourgeois economics.

A. Bhaduri, a Marxist critic of neoclassical economics, indicated the

significance of the distinction between the concepts of capital as a
thing and capital as a social relation in an important essay produced
during the recent capital theory controversy. 22 He wrote: ‘It must be
granted that Marx himself was unable to indicate the logical implica-
tions of his understanding of the role of “capital” for the formulation
of a theory of distribution between profits and wages in a capitalist
economy. In the view of the present writer this is precisely what the
recent controversies on capital theory do: they lay bare the logical
weakness of treating capital merely as an instrument of production
in developing a theory of distribution in a capitalist economy.’ 23

These remarks apply, with equal force, to the subject of capital accu-
mulation. After the capital theory controversy the neoclassical model
of economic growth, which is discussed below, now lies in ruins.

The accumulation of capital, therefore, cannot be simply reduced to

the accumulation of homogeneous embodied labour. This error has
continually re-occurred in the Marxian tradition. It is not uncommon
for Marxists to treat reproduction schemes as if they reflect money
prices, or even the physical scale of production, whereas these
schemes are in value terms only. In the historic debates that were
generated by the publication of Rosa Luxemburg’s The Accumulation
of Capital in 1913, Otto Bauer 24 and others made the same error.
Bauer ignored the problems uncovered by Luxemburg by concentrat-
ing exclusively on the accumulation of embodied labour values.
Luxemburg on the other hand compounded this confusion by mis-
taking the accumulation of capital for the accumulation of money,
and an increasing social product measured in price terms. 25

In fact accumulation involves all these aspects, but is not reducible to

any one of them; capital accumulation is not just the accumulation of
things, or the augmentation of single quantities. Fundamentally, the
accumulation of capital is the reproduction of capitalist social relations
on an extended scale. It involves the extension of these relations over
all other subordinate modes of production, which become destroyed
or subsumed by capitalism, and the intensification of these relations,
when, for instance, the means of production become monopolized by
fewer capitalists.

Capital Accumulation and Employment

Another argument, quite similar to the one used by Yaffe, is some-

times brought up to defend the falling rate of profit theory. It is ar-
gued that as capitalism expands to the extent that unemployment
falls, wages tend to rise as a result of the more favourable situation of
the working class. As a result, it is argued, capitalists tend to reduce
the size of their labour force and ‘substitute’ constant capital for
labour power. Hence the organic composition of capital will tend to
rise. To be complete this theory must also argue that the process is
not reversed, with a fall in the organic composition of capital, when
wages are low during a recession. Otherwise no overall trend could
be deduced.

There is a grain of truth in this theory. Wages do tend to oscillate in

this manner. Capitalists often lay off workers when the wage bill is
too high. In these circumstances they are likely to ‘rationalize’ pro-
duction and invest in new plant and equipment. But we have no rea-
son to suppose that the value of their constant capital will increase as
a result. What happens when full employment is reached and the
capitalists still strive to accumulate? They cannot enlarge their labour
power, so perhaps they are forced to increase constant capital, and
thereby increase the organic composition of capital? This argument
is unsound because it either assumes that accumulation necessarily
involves an increase in the value of constant capital, which we have
argued to be false, or it assumes that the capitalists consciously strive
to augment the value of their capital. On the contrary, the capitalists
are not aware of their embodied labour values, or inclined to find out.
Perhaps they will strive to increase the mass of machinery employed,
but this bears no necessary relation to its value.

Of course we do not argue that capitalism operates according to the

subjective plans of the capitalists. The overall dynamic of the system
is a result of a complex interaction of forces, and capitalism retains
an anarchic character. But we cannot mechanically divorce the ac-
tions of powerful individuals from the objective course of events, or
regard the former as completely ‘determined’ by the ‘economic base’,
which is conceived as a sort of separate machine devoid of individu-
als and the force of ideas. The basis of analysis, in any field of scien-
tific enquiry, cannot be reduced to either the whole alone, or to the
constituent parts by themselves.

The Concept of Capital and the Materialist Method

The tendency for the organic composition of capital to rise cannot be

justifiably derived from the ‘concept of capital’ in a purely a priori
manner. It is a mere tautology to start from the definition of capital as
‘self expanding value’, add the correct notion of the limited size of the
pool of living labour power, and triumphantly conclude that the or-
ganic composition of capital will rise. This method of reasoning ‘ex-
plains’ social reality from a pre-defined idea; it does not explain ideas,
including the concepts of political economy, from social practice. To
derive the notion of a rising organic composition from the concept of
capital in this manner is to collapse into Hegelian idealism. A
Marxist, however, must proceed from the real processes and relations
of the capitalist mode of production. On this basis the Marxist can ex-
plain how both Marxist and bourgeois categories are produced within
capitalism. The former have a real concrete existence, the latter have
an existence as a ‘false consciousness.’
The Marxian method involves initial abstraction from a multitude of
empirical phenomena. However, Marxian concepts such as the com-
modity, capital, and abstract labour are not just ideas, they are real
under capitalism. In contrast, bourgeois economics starts to ‘explain’
reality from ahistorical ideas such as utility and human nature.
Correct economic categories are only abstract expressions of real so-
cial relations, and only remain true as long as these relations exist.

Marx, himself, tried to derive the law of the falling rate of profit from
the concept of capital in several passages in the Grundrisse. However,
this idealistic method of reasoning receives little prominence in
Capital. In the opinion of this writer this observation testifies that the
earlier work was still written under the influence of Hegelian modes
of thought. In Capital Hegelian idealism is almost completely
eliminated.

An Agnostic Conclusion

There seems to be no a priori reason for the organic composition of

capital to rise. This conclusion rests on a rigorous separation of three
aspects of capitalist production: the physical aspect, the price aspect,
and the value aspect. The relations between these aspects and the
whole partly determine the dynamic behaviour of the capitalist sys-
tem. Only by such a rigorous separation can capital be conceived as a
social relation, rather than a homogeneous ‘thing’.

We do not need to elaborate the point that vulgar bourgeois economy

confuses the different aspects of capitalist production. Neoclassical
economics elevates the physical aspect of capital to the detriment of
all others. We have ‘marginal productivity’, ‘factor substitution’: capi-
tal as a thing par excellence. But the point needs to be emphasized
that some Marxists have committed a very similar mistake in trying
to defend the falling rate of profit theory. They have confused the
value aspect with the physical aspect (and in the case of the transfor-
mation problem prices are confused with values). By reducing capital
to a mere value, capital is implicitly regarded as a homogeneous
‘thing’.

For these reasons the recent attack on neoclassical economics, in the

capital controversy, is a significant event for Marxism. A brief and
unsystematic account is given here, as Marxists cannot remain silent
in the face of the theoretical conclusions. A correct interpretation of
capital theory can lead to a forceful re-establishment of the concept of
capital as a social relation, if certain Ricardian pitfalls are avoided.
We are led to abandon the theory of the falling rate of profit, and
along with it all vulgar notions of capital and capital accumulation.
Marxism can vengefully escape from its isolation, and develop in the
process.

III The Impact of the Capital Controversy

Most of the conclusions of the capital debate stem from rigorous and
logical arguments applied to a situation where heterogeneous capital
goods exist. 26
There are consequences for the theories of price, dis-
tribution, and capital accumulation. Here, of course, we are primarily
concerned with the latter.

The Concept of Dated Labour

It is a mistake to simply analyse the system in terms of just two types

of labour time, i.e. living labour and dead labour embodied in com-
modities. In most cases we cannot usefully aggregate all embodied
labour from the past into one homogeneous whole. The date at
which a past labour input is required to produce a commodity is cru-
cial. Nearly all goods are produced with both living labour and means
of production. The means of production are, in turn, products of liv-
ing labour and means of production in a previous time period.
Hence the labour embodied in a commodity can be split into a long
series of dated labour inputs 27 diminishing into the past. Each of the
terms in this series has an independent significance in determining
such variables as the rate of profit. Marx drew a distinction between
dead labour and living labour, so the dated labour series is an exten-
sion of Marx’s distinction from two to many time epithets.

It is possible to regard all technical innovation as labour saving in

some sense. But the crucial point is that we need to regard amounts
if labour from different time periods as qualitatively different. Similar
remarks apply to the notion of constant capital saving innovation.
Technical advance cannot be regarded as a two-dimensional matter;
it is a changing multi-dimensional pattern of dated labour inputs, all
of which apply to the technical conditions of the present.

In the cases where constant capital stock is included the genealogy of

the dated labour inputs must also be considered. This means that the
history of a dated amount of labour, the time of its origin, and when
it was transferred to and from fixed and circulating elements is
significant.

The capitalist has a persistent tendency to ‘measure’ all things with a

single quantitative index: their price. To paraphrase Oscar Wilde: the
capitalist knows the price of everything but the value of nothing. This
reduction of all things to their price is a significant feature of capital-
ist accounting, which is inseparable from capitalism. But these quan-
titative indices are not independent of the rate of profit, interest rates,
and the social relations between classes. As these alter, even by the
slightest amount, so will the book value of the bourgeois world. The
consequence of heterogeneity is that there is no independent mea-
sure of capitalist wealth.

D. M. Nuti concluded an important essay on capital theory with the

following words on the bourgeois notion of the ‘value of capital’: ‘The
ideological role of the “value of capital” is that of breaking the direct
actual link between the time pattern of output in which any technol-
ogy can be resolved, and establishing instead a relation between cur-
rent output and current labour. To this purpose the current “value of
the capital stock” is needed; a mythical conceptual construction in
which the past and the future of the economy are telescoped into the
present.’ 28
This criticism can be also applied to the habit of measur-
ing constant capital in terms of a single amount of embodied labour.

The Solovian Growth Model

The similarity between the bourgeois concept of capital and the

crude ‘embodied labour’ conception is reflected in the similarity be-
tween the falling rate of profit theory and the neoclassical growth
model. The neoclassical economist R. M. Solow published his
Contribution to the Theory of Economic Growth 29
in 1956. It is a
macroeconomic model involving the neoclassical idea of a ‘produc-
tion function’. Two inputs, dubbed ‘factors of production’, i.e. ‘capital’
and ‘labour’, combine together in production to create a net output.
This output is represented as a mathematical function of the inputs.
Solow discusses a number of such production functions. He makes
the simplifying assumption of constant returns to scale, i.e. output
per man does not depend upon the size of the plant, just the relative
proportions of ‘capital’ and ‘labour’. This allows him to represent the
production function by a two-dimensional graph, examples of which
are shown below:
Solow shows that in many cases there is an equilibrating process
which allows ‘output per man’ and ‘capital per man’ to converge to a
fixed level, and full employment is achieved. But this allows for no
technical progress. It would seem reasonable to assume, along with
Solow, that technical progress can be represented by an ‘expanding’
production function of the type shown in figure II.

At first the production function is represented by the curve PF1.

Later it moves up to PF2, and later still it has moved to PF3. Hence
‘output per man’ increases even if the amount of ‘capital per man’
stays constant, as a result of technical progress. And now the discus-
sion of Solow’s equilibrating process leads to the conclusion that ‘the
capital-labour ratio never reaches equilibrium but grows forever’. 30

If we ignore the ideologically bound terminology and the monstrous

presumption that full employment can be maintained automatically
under capitalism, then the similarities with the falling rate of profit
theory are evident. In both instances we have the presumption that
we can measure constant capital independently of all other economic
conditions. Solow assumes that in the majority of cases output per
man will increase as ‘capital per man’ increases. Orthodox Marxists
such as Yaffe can write: ‘The increase in the means of production per
worker . . . is not merely a technical premiss . . . It is the expression
in general terms of the only way the productivity of labour can rise
under capitalist production . . .’. 31 We have the conception of a partic-
ular type of technical progress which can lead Solow and some
Marxists to a similar conclusion. As cited above, Yaffe writes: ‘It fol-
lows that both the technical composition of capital and the organic
composition of capital must increase in the process of capitalist pro-
duction.’ Finally, the notion is shared that an increase in the organic
composition of capital, or the amount of ‘capital per man’ will lead to
a fall in the rate of profit.

The Attack on the Neoclassical Aggregate Production

Function

One of the first shots in the battle was fired by Joan Robinson in
1953. 32 She contested the complacency of the neoclassicals who as-
sumed that the ‘amount of capital’ can be readily measured. After
twenty years of debate the aggregate ‘capital and labour’ production
function lies in ruins. One of the latest and more important blows
was delivered by P. Garegnani. 33 From the premiss of heterogeneous
capital goods he developed several feasible ‘production functions’, de-
pending on given feasible technical conditions of production. These
bear no relation to the ‘well behaved’ neoclassical production func-
tion. Four of Garegnani’s examples are shown in figure III.

It is clear from these examples that increased ‘capital per man’ (Q) is
related to ‘output per man’ (O) in no simple or consistent way. There
is no basis, therefore, for asserting that increased productivity is generally
associated with an increased organic composition of capital. Also
Garegnani shows that there is no simple inverse relation between Q
and the rate of profit. The notion that the march of productivity leads
to a general fall in the rate of profit is completely shattered.

If we try to introduce a notion of technical progress into these pro-

duction functions we do not get the simple Solovian result that Q
‘grows forever’. Far from it. Technical progress bears no simple or
necessary relation to Q, or the organic composition of capital.

The arguments of Garegnani, Sraffa and others are systematic and

logical. Their destructive power is rooted in these qualities. Marxists
have no reason to abandon these arguments, but they must be sup-
plemented by a critique of the fashionable Ricardian interpretations
of capital theory. However, this cannot be done by aping the argu-
ments of neoclassical economists which have been proved so inde-
fensible. Neither can the matter be resolved by a simple reiteration of
Marx.
We now turn to an examination of the empirical data for the United
States which suggest that there has been no consistent rise in the or-
ganic composition of capital. The evidence does suggest a rise in the
organic composition of capital up to about 1920 with the general
spread of mechanization–after that date innovations seem to have led
to constant capital saving improvements and a consequent decline in
the organic composition of capital.

IV A Survey of Empirical Data for the United States

Data alone cannot decisively refute a theory. But that does not mean
that empirical tests have no status in Marxism. Marxian categories
are not just ideas, they correspond to real relations and parameters in
the capitalist system. The rejection of an empiricist dualism between
theory and reality does not necessarily mean a rejection of empirical
data. On the contrary, Marx’s method, that is the ‘method of rising
from the abstract to the concrete is the only way in which thought ap-
propriates the concrete, reproduces it as the concrete in the mind.’ 34

In addition, the concrete ‘appears in the process of thinking . . . as a

process of concentration, as a result, not as a point of departure, even
though it is the point of departure in reality and hence also the point
of departure for observation and conception.’ 35
It would not be rash,
therefore, to assert that the Marxian method of analysis involves em-
pirical verification. 36 This is in contrast to some Marxists who define
their categories in such a way as to have no contact with empirical
reality.

The evidence that is relevant to an examination of the theory of the

falling rate of profit is not the actual profit rate, or the share of profits
in national income, but data concerning the organic composition of
capital and related expressions. In Britain, for example, Glyn and
Sutcliffe have argued that there has been a fall in the rate of profit
due to a falling share of profits in the national income. But that does
not, in any way, endorse Marx’s theory, which stems from the hypoth-
esis of a rising organic composition of capital.

Unfortunately there are few empirical studies of the organic compo-

sition of capital. 37
The author is not aware of any other major study
other than the ones carried out by Gillman and Mage. Both of these
studies apply to the United States. The former is concerned with the
organic composition of capital in the manufacturing sector, the latter
is concerned with the economy as a whole. There have been many
criticisms of these statistics, and in the opinion of the present writer
both sets do not show real value ratios, i.e. ratios between amounts of
socially necessary labour time. However, the data are reproduced
here for the information of the reader. The data are expressed in
terms of the definition of the organic composition of capital that is
found in this article (i.e. q = k/(v + s)). Mage’s data were already ex-
pressed in this form, but Gillman’s had to be calculated from the sta-
tistics he provides for k, v, and s.

No pronounced upward trend in the organic composition of capital is

evident in Mage’s figures in table 1. The high figures for 1930, 1935
and 1940 are partly a result of the Great Depression, when net out-
put (v + s) was low and a great deal of constant capital stock was
unutilised. If these figures are excluded the slight upward trend is
even less significant.
Some startling facts are apparent in table ii. Firstly, it appears that
the organic composition of capital in the manufacturing sector is
much less than in the economy as a whole. Perhaps this can be ex-
plained by the high productivity of the industries that produce capital
goods for that sector. Secondly, after a clear rise in the organic com-
position of capital from 1880 to 1921, there is a tendency for its mag-
nitude to decline after the latter date. Discounting the high figures in
the years of severe depression, the organic composition of capital was
about 1·3 in the boom period in the 1920s, and this figure is not ri-
valled after the Second World War, at least up to 1952.

We now turn to the data provided by bourgeois economists. The ratio

that is analogous with the organic composition of capital, according
to our definition, is the ‘capital-output’ ratio. This is the ratio be-
tween the price of constant capital stock and the price of output. The
capital-output ratio is related to the rate of profit in the following
manner:

The latter result illustrates the analogy between the capital-output ra-
tio and the organic composition of capital. These two ratios are not
identical but they have a similar status within two respective account-
ing systems, one in terms of prices the other in terms of values. In
fact the capital-output ratio is more relevant for a direct calculation of
the rate of profit in real terms. The operative rate of profit, upon
which the capitalists base their investment decisions, is a ratio be-
tween price amounts, not a ratio between values. It is possible for the
organic composition of capital to rise whilst the capital-output ratio
falls, but the capitalist is unaware of the former, which does not nec-
essarily effect the real rate of profit, or the investment decision.

Once again, this does not mean that the economy operates entirely in
accord with the subjective wishes of the capitalists. But these subjec-
tive wishes are part of the objective reality, and any investigation into
the dynamics of the capitalist system must show the basis on which
capitalists make decisions to invest. To ‘explain’ the workings of the
capitalist system without any reference to appearances, or the ideas
that motivate the capitalist, is to raise the ‘economy’ to the status of a
heavenly machine grinding out the destiny of capitalist society.
Marxists, like high priests, alone are aware of the god-like power of
the machine. Hence this ‘materialist’ attempt to understand capital-
ism collapses into an idealism; society is divided into two parts, one
of which is superior to society. The result is that Marxism has no
contact with empirical data, and no possible basis for a fruitful dia-
logue with bourgeois social science.

One of the most extensive studies of the capital-output ratio in the

us has been carried out by S. Kuznets. His data for the economy as a
whole is presented in table iii.
The figures in table iii provide a remarkable resemblance to Mage’s
data in table i. There is no marked upward trend in the capital-out-
put ratio, and a slight downward trend is evident after 1909–18 if we
disregard the inflated figure for the depression years of 1929–38.
Even if we include the figure for the years 1929–38 statistical analysis
shows that the overall upward trend in the capital-output ratio is very
slight indeed. The trend line shows a rise of only 0·0086 per year.
On this basis the trend reaches the magnitude of 4·3 in the year
2000. But the extent of the variance of the actual figures from the
trend allows us to make no such prediction from the statistics.

Kuznets regards the figures in table iv as rough approximations only.

The earlier figures are larger than those provided by Gillman for the
manufacturing sector, but a similar pattern is evident. The figures
show a rise before 1922, but the figure for 1948 indicates a fall in the
capital-output ratio after the former date.
The figures in the first three parts of table v, where individual indus-
tries are considered, are the ratios between capital and gross output,
so they are not strictly comparable with the capital-net output ratios,
which are larger for a given industry. Most of the industries show a
slight overall decline in the capital output ratio over time. Petroleum
and Natural Gas shows a very rapid rise from 1870 to 1919, and an
even more rapid fall after the latter date. Steam Railways, Electricity
Supply, and Telephones all show a very marked fall over the whole
period.

The figures for agriculture are especially interesting as they show the
effects of mechanization in that sector. The upward trend in the price
ratio of buildings and equipment to net farm income reflects the
process of increasing mechanization. But this does not create an
overall rise in the total capital-net income ratios, including or exclud-
ing land. It appears that machinery has replaced power animals and
other live-stock along with savings in the use of other agricultural ma-
terials. These two simultaneous processes have led to a slight fall in
the capital-net income ratios.

In conclusion, most of these figures do not give empirical backing to

the hypothesis of a rising capital-output ratio. Most of the figures
show a rise up to about the year 1920 and a general fall after that
date. A similar pattern is evident in Gillman’s data. The period up to
1920 was characterized by an extensive accumulation of capital, i.e. a
rise in the mass of machinery, a spreading of mechanization, an ac-
cumulation of values, and the general extension of capitalist relations
of production in the us. The years after 1920 could be regarded as
years of intensive innovation in an already mechanized economy,
punctuated by crises such as the Great Depression. More attention
was shifted to constant capital saving improvements, and the more
efficient utilization of existing plant and machinery, in the home
economy. After 1920, of course, the us was in a stronger position to
compete for the title of leading imperialist nation. Perhaps its in-
creasing imperialist role was partly a result of the ending of the pe-
riod of extensive capitalist accumulation at home? At the moment
the theory of imperialism is too primitive to supply us with an
answer.

V The Theory and Revolutionary Marxism

In this section we shall discuss the political and methodological im-

plications of the so-called law of the falling tendency of the rate of
profit. The rejection of the law has profound implications. Some
would argue that such a rejection constitutes a victory for reformism.
On the contrary such an antithesis is based upon a faulty problem-
atic. Having rejected the theory of the falling rate of profit it becomes
necessary, therefore, to assault the problematic within which the fall-
ing rate of profit becomes the soul of revolutionary Marxism.

The Law and its Counteracting Influences

It is commonplace to assert that society is not a laboratory. It is clear

that laws of social development cannot be isolated from their coun-
teracting influences. In contrast the physical scientist attempts, with
some success, to isolate the phenomenon under investigation and
determine its inner laws, without the clutter of extraneous
influences.

With this point in mind, interpretations of the law of the falling rate
of profit can be grouped into three classes: the law as a manifest ten-
dency, the law as a concomitant force, and the law as an ultimate ten-
dency. In the first conception the law is regarded as an evident and
persistent force; counteracting forces just retard the fall in the rate of
profit, they do not annull its clear downward trend. The second con-
ception is less decisive: the law is regarded as one force amongst
many. The outcome of this multitude of interacting influences is not
necessarily a fall in the rate of profit. Finally we may regard the law as
an ultimate tendency, which can be checked by counter-acting influ-
ences. Consequently a fall in the rate of profit may not be evident for
long periods of time, it appears ultimately sometime in the future.

Perhaps it is easy to dismiss the conception of the law as a manifest

tendency; few Marxists adhere to this conception today. But this may
be explained by the facts that a persistent fall in the rate of profit or
rise in the organic composition of capital are not clearly evident in
the 20th century. In contrast, Adam Smith and Ricardo were much
bothered by the fall in the rate of profit which was evident in the 18th
and early 19th centuries. Marx’s theory was, at least in part, an at-
tempt to solve this riddle. Today, however, with no consensus in eco-
nomic circles, and in view of the evidence of Gillman, Mage and
Kuznets, few would deny that the ‘counteracting influences’ have be-
come prominent for many decades.

The Law as a Concomitant Force

The second conception is practically a polar opposite of the first; in-

stead of necessity we have indeterminancy in the long run. Over
thirty years ago M. Dobb put forward an interpretation of Marx’s law
which seems close to the notion of the law as a concomitant force:
‘There is often a tendency . . . to give Marx’s view of this matter a too
mechanistic twist, depicting it as though it relied on the forecast of
profit falling in a continuous downward curve until it reached a point
at which the system would come to an abrupt stop, like an engine
with insufficient pressure of steam behind the piston. The true inter-
pretation would seem to be that Marx saw tendency and counter-ten-
dency as elements of conflict out of which the general movement of
the system emerged . . .’. 46

In the second section of this present essay we have contested the idea
of a necessary fall in the rate of profit on theoretical grounds. The no-
tion of the law as a concomitant force, with an indeterminate out-
come, could seem to be in accord with our theoretical position. Such
an interpretation would be false. It is not justified to describe forces
bringing down the rate of profit as ‘tendencies’ whereas forces acting
in the opposite direction are seen to be mere ‘counteracting influ-
ences’. Such an arbitrary designation of conceptual status could be re-
versed. In which case the counteracting influences would become
‘law’ and the law of the falling rate of profit would collapse–by a mere
change of terminology.
In a reaction against mechanistic Marxism the notion of the law as a
concomitant force does not completely escape from the mechanistic
problematic. The agnosticism of this position could be reduced to a
lack of knowledge of the laws of motion of the ‘economy’. Further dis-
covery might reveal laws which act to bring down the rate of profit.
To escape from this problematic we need to reject the notion of the
economy as a machine. We shall return to this problem at a later
stage.

The Law as an Ultimate Tendency

This is, no doubt, the most widespread conception of the law. It, it-
self, has two variants: some regard the ‘underlying’ fall in the rate of
profit as being superimposed by periodic fluctuations, others regard
the ‘ultimate’ fall in the rate of profit as an ‘inevitable’ process which
is to become pronounced sometime in the future. In the latter case
the ‘periodic fluctuation’ spans an epoch. At least for the purposes of
this discussion these two variants are essentially similar. Within this
conception of an ultimate tendency we have, in a sense, a synthesis
of the first two conceptions: conjunctural indeterminacy but ‘in the
last instance’ the force of necessity.

Here the law runs the gauntlet of counteracting influences. It is in

constant danger of being thrown back to its starting point. But in the
long run it triumphs: not in the shape of rich empirical experience,
but in the idea of its ‘ultimate’ victory. The ‘last instance’ is never an-
nounced by the sound of trumpets and the collapse of the citadel of
profit. It is prophesied, but its coming is unrecognizable. Its status as
an ‘ultimate’ law faces the perennial challenge of another periodic
upswing in the rate of profit, which would lead us to the conclusion
that there is at least one more ‘last instance’ to come. As Althusser
has aptly remarked in a different context: ‘From the first moment to
the last, the lonely hour of the “last instance” never comes.’ 47
The law as an ultimate tendency can never be identified with empiri-
cal experience: for fear of the tyranny of facts. The history of the capi-
talist mode of production becomes a dualist combination of rational
forces and empirical surroundings. 48 The law finally comes to rest in
the realm of pure reason: it explains the demise of capitalism, but the
law of the falling rate of profit is never revealed as an ultimate ten-
dency in the realm of appearance. As in Hegel: ‘what is rational is
real and what is real is rational’. From Marxism we are led back to
Hegelian rationalism.

This unconscious return to Hegel accounts for the more or less gen-
eral failure to Marxists to develop both theories of the relation be-
tween political economy and particular social formations, and serious
research to provide the empirical backing for such studies. Political
economy has tended to become a seance with the spirit of a weird
‘economic machine’ which never appears in view. Its ‘laws’ are identi-
fied, its mechanics become known, or rather they are already known,
even before they become manifest. History submits to our Principia
Economica.

The Role of Marxian Political Economy

It may be argued that the previous theoretical position applies to all

tendential laws of an ‘ultimate’ character. That argument is indeed
correct. Marxism is more or less rid of the ‘law’ of the absolute im-
misseration of the proletariat, even its origin in Marx is doubtful.
Efforts are being made to purge the movement of all notions of a
breakdown theory. It is now opportune to reject the law of the falling
tendency of the rate of profit.

L. Colletti 49
and others have pointed out that the works of Marx and
Engels have been interpreted in a mechanistic manner by most of
the deans of orthodox Marxism for nearly a hundred years. It is the
view of the present author that the renaissance of interventionist and
unmechanistic Marxism must reject tendential laws of an ‘ultimate’
character. It has become commonplace to identify the source of these
mechanistic distortions of Marxism in some of the works of Engels.
However, some of the blame must also fall on Marx. His Preface to a
Contribution to the Critique of Political Economy can be, and has been,
interpreted in a crassly mechanistic fashion, although its real mean-
ing is somewhat ambiguous or obscure. In the preface to the first
German edition of the first volume of Capital Marx talks of ‘laws . . .
working with iron necessity towards inevitable results’. 50
Clearly
Marx, and especially Engels, were never completely free of the out-
look of Newtonian mechanics or Darwinian biology.

The version of Marxism that was given prominence by the leading

theoreticians of the Second International, such as Kautsky and
Plekhanov, rests on a vulgar notion of the ‘economy’, which is seen as
one isolated ‘factor’, emptied of all effective social and historical con-
tent. The ‘economy’ runs on like a machine, prior to any real human
intervention or mediation. 51
Whereas in Marx we can find countless
references to his notion of the ‘social relations of production’ which
embraces both the production of things and the production of ideas:
material production and the reproduction of social relations.

Unmechanistic interventionist Marxism cannot proceed, therefore,

from a pure analysis of the ‘economy’, and then embellish this fabric
with sociological and political ‘detail’. 52 These ‘factors’ cannot be me-
chanically isolated. The categories of Marxian political economy are
at once economic, sociological and political. Consider, for example,
the concept of labour power as a commodity. It involves the existence
of separate ‘sociological’ classes between which purchase and sale can
take place, a legal framework within which a labour contract can exist,
and an existence of a state which can protect capitalist social rela-
tions, as well as the more obvious ‘economic’ connotations. It is only
within this unity of the ‘political’ the ‘economic’ and the ‘sociological’
that we can recognize the significance and prominence of production
in Marxism.

Marxian political economy has traditionally been the fount of predic-

tion in the shape of ‘economic perspectives’ for socialist organiza-
tions. The empiricist duality between theory and phenomena has
been transformed into a de facto separation between theory and prac-
tice. The role of theory is mere prediction: to assure the movement of
the ‘inevitability’ of socialism, to herald the next crisis which is ‘just
round the corner’. Theory, in short, is a commentary on the workings
of the mythical economic machine. Practice, on the other hand, is in-
volvement in economic struggle as an acknowledged cog of the
machine.

This does not mean that in Marxism there is no room for a theory of
crisis. On the contrary it remains central. But in a concrete and con-
junctural rather than a rationalistic sense. In other words the theory
of crisis becomes investigation into the causes of concrete socioeco-
nomic convulsions, from the point of view of practical revolutionary
struggle, rather than the idea which becomes the rational basis for
‘revolutionary socialism’.

Rationalistic theory is a basis for a quasi-religious fanaticism: the

idea that despite isolation and defeat the objective force of events will
ensure that victory is inevitable. This fanatical aspect of mechanical
determinism was persistently attacked by Gramsci. 53 It is no accident
that Gramsci’s studies of the specific ideological, structural, national
and international circumstances of revolutionary intervention are
largely unparalleled in the history of Marxism.

The structural basis of capitalist crisis lies in the conflict between the
tendency to socialize the forces of production, and the capitalist rela-
tions of production with private appropriation. With an analysis of
the specific manifestations of this contradiction the re-occurrence of
crises under capitalism, in one form or another, is assured. But these
manifestations are concrete, as Marx attempted to demonstrate
throughout Capital. This does not mean that there are no general
laws of capitalist development, but such laws are real social facts,
rather than just rationalistic ideal arguments in the heads of
economists.

The role of Marxian political economy is not pure contemplative pre-

diction. As Gramsci notes: ‘Anybody who makes a prediction has in
fact a “programme” for whose victory he is working and his predic-
tion is precisely an element contributing to that victory. This does not
mean that prediction need always be arbitrary and gratuitous, or sim-
ply tendentious. Indeed one might say that only to the extent to
which the objective aspect of prediction is linked to a programme
does it acquire its objectivity . . . therefore if one excludes all volun-
tarist elements, or if it is only other people’s wills whose intervention
one reckons as an objective element in the general interplay of
forces, one mutilates reality itself. . . . Hence to believe that one par-
ticular conception of the world, and of life generally, in itself pos-
sesses a superior predictive capacity is a crudely fatuous and superfi-
cial error.’ 54

Marxian political economy has a practico-theoretical role, not just as

‘theoretical practice’, but as an active tool of intervention in the revo-
lutionary class struggle. 55
Through a ‘concrete analysis of concrete
conditions’ Marxism can become fused with revolutionary organiza-
tion, agitation and propaganda. This involves a total analysis of the
socio-economic formation in both the generic and the specific sense.
The basis for this revival of interventionist Marxism is, possibly,
found in Lenin’s methodology of revolutionary practice, Gramsci’s
studies of social formations and the role of ideology, and Trotsky’s
analysis of particular revolutionary conjunctures. Its slogan can be
taken from Connolly: ‘The only true prophets are those that carve out
the future they announce’. To that end we must bury the last iron law
of Marxian political economy–the law of the falling tendency of the
rate of profit.

APPENDIX: Reduction of the Theory of the Falling

Tendency of the Rate of Profit to the Question of

the Organic Composition of Capital.

Some New Definitions

The problem of the boundless upper limit of the fractions k/v or s/v
can be dealt with by a simple redefinition of the basic Marxian ratios.
We shall let

Marx explicitly rejected the definition of the rate of exploitation that

is formulated above. 1 He wrote: ‘The habit of representing surplus
value and value of labour power as fractions of the value created–a
habit that originates in the capitalist mode of production itself. . .–
conceals the very transaction that characterizes capital, namely the
exchange of variable capital for living labour power, and the conse-
quent exclusion of the labourer from the product. Instead of the real
fact, we have the false semblance of an association, in which labour
and capitalist divide the product in proportion to the different ele-
ments which they respectively contribute towards its formation’. 2

Marx makes two points here, but they have little to do with the differ-
ent formulations for the degree of exploitation. Surely the exchange
of variable capital (v) for living labour (s + v) would be characterized,
for instance, by the fraction v/(s + v)? But this is nothing but i-e, ac-
cording to our definition.

Furthermore, the use of either definition does not in itself guard

against the ‘false semblance of an association’ where capitalists and
workers receive their ‘fair shares’. The attack on this notion must be
added, in both cases, by remarking that surplus value is produced by
the workers under conditions of compulsion, and it must be ex-
plained that a fair distribution of the social product is inconceivable
under capitalism. Finally, both definitions do not say anything about
the length of the working day, this must be built into our analysis
elsewhere.

The advantage of our definition is that the magnitude of e now

ranges from zero to one, instead of zero to infinity. Hence e is now
the fraction of the net output produced by the workers that is expro-
priated by the capitalists. i-e is the remaining fraction that is left to
the workers.

We now turn to the organic composition of capital. Again this differs

from Marx’s definition. Here it is defined as the amount of constant
capital stock per man-hour. Marx often assumed that the rate of ex-
ploitation was constant. In this case there is a direct relation between
the two definitions. The advantage of our definition is that q is now
independent of any changes in the degree of exploitation, at least in
the short run.

The term ‘rate of constant capital flow’ is a new one, applied to the
ratio between constant capital flow and net output. Our rigorous dis-
tinction between stocks and flows is reflected in the two definitions
of q and r. Once again, r is independent of any short run changes in
e.

A Re-Formulation of the Theory

We may divide numerator and denominator by the net output(y):

It is clear from Equation 3 that if e falls, whilst q, r and t are constant,

then p will fall. Also if t falls, whilst e, q and r are constant, then p will
rise; this is the phenomenon of the reduction in the turnover period,
discussed in the third volume of Capital. 3

The advantage of our definitions is now evident. The fraction e has a

maximum value of unity, hence with given values of q, r and t the
rate of profit must have a maximum value. This is easily derived by
giving ‘e’ the value of one in Equation 3:

where Pmax is the maximum rate of profit.

According to Marx, two of the historical trends of capitalist accumu-

lation are a reduction in t and an increase of q in relation to r. Hence
the term tr is very small in relation to q, and the maximum rate of
profit approximates to 1/q. Perhaps the existence of these trends is
questionable, but in any case Pmax must be less than 1/q.
If Pmax falls this does not necessarily mean that p will fall. However
Pmax does provide a barrier to accumulation and growth, and its fall
is just as significant as a fall in the immediate rate of profit. The
problem of the rate of profit is thus reduced to the postulated rise in
the organic composition of capital. It is clear that we may now disre-
gard fluctuations in the rate of exploitation (e). Some economists,
such as H. D. Dickenson 4 and S. H. Mage 5 have adopted more com-
plicated mathematical proofs to reach the same conclusion.

1
I am indebted to Ian Steedman, Andrew Glyn, Bob Rowthorn, and
Norman Geras for making valuable comments on sections of the text.
Responsibility for the final product is, of course, entirely mine.

2
Capital, Volume 1, Chapter 25, Sections 2 and 3; and Volume 3, Part 3. See
also Grundrisse, London 1973, pages 386–98 and 745–58. Hereafter we shall
refer to the Moscow 1961–2 edition of Capital, and the Pelican Marx Library
1973 edition of the Grundrisse.

3
Grundrisse, pp. 413–14.

4
Capital, Volume 3, p. 153.

5
Capital, Volume 3, Part 2. P. Sweezy, The Theory of Capitalist Development,
New York 1968, Chapter 7. L. v. Bortkiewicz, Value and Price in the Marxian
System, in International Economic Papers, Volume 2, London 1952. See also
the articles by I. Steedman, D. Yaffe, and G. Hodgson in the Bulletin of the
Conference of Socialist Economists (BCSE), Autumn 1973.

6
Capital, Volume 3, pp. 47, 70–6, 111, 222 and 224.

7
Capital, Volume 2, part 2, and Volume 3, Chapter 4.

8
Capital, Volume 3, p. 207.

9
Ibid., p. 208.

10
Ibid., p. 208–9.
11
Theories of Surplus Value, Part 2, Moscow 1969, pp. 415–16.

12
J. Robinson, An Essay on Marxian Economics, London 1966, p. 36–40.

13
R. L. Meek, Economics and Ideology and Other Essays, London 1967, pp.
131–5.

14
Capital, Volume 1, p. 305.

15
P. M. Sweezy, op. cit., p. 103.

16
P. M. Sweezy, Some Problems in the Theory of Capital Accumulation, in
BCSE Autumn 1973, pp. 28–9.

17
M. Blaug, Technical Change and Marxian Economics, in Marx and Modern
Economics, edited by D. Horowitz, London 1968.

18
D. S. Yaffe, The Marxian Theory of Crisis, Capital and the State, in
Economy and Society, May 1973, pp. 197–8. Also in the BCSE, Winter 1972.

19
Capital, Volume 3, p. 231.

20
D. S. Yaffe, op. cit., pp. 195–6. (Certain references to Marx have been
omitted in our quotation.)

21
Ibid., page 197. Yaffe assumes that ct+1-ct is less than vt-vt+1, where t and
t+1 denote successive time periods. He also implicitly assumes that ct+1-ct
is positive. He concludes that ct+1/vt+1 is greater than ct/vt, which is, of
course, obvious from his assumptions.

22
A. Bhaduri, On the Significance of Recent Controversies on Capital Theory: A
Marxian View, in Capital and Growth, edited by G. C. Harcourt and N. F.
Laing, London 1971.

23
Ibid., p. 254.

24
O. Bauer, Die Akkumulation des Kapitals, in Neue Zeit 1913, No. 24, pp.
831–8 and 862–74.

25
R. Luxemburg, The Accumulation of Capital–An Anti-Critique, in
Imperialism and the Accumulation of Capital, edited by K. Tarbuck, London
1972. See also, R. Luxemburg, The Accumulation of Capital, London 1963,
especially chapters 4–9.

26
For a Survey see G. C. Harcourt, Some Cambridge Controversies in the
Theory of Capital, Cambridge 1972.

27
P. Sraffa, The Production of Commodities by Means of Commodities,
Cambridge U.P. 1973, Chapter 6. See also I. Steedman, Marx on the Rate of
Profit, in BCSE Winter 1972.

28
D. M. Nuti, Capitalism, Socialism and Steady Growth, in Capital and
Growth, London 1971, p. 338.

29
R. Solow, reprinted in Growth Economics, edited by A. Sen, London 1970,
pp. 161–92.

30
Ibid., p. 183.

31
D. Yaffe, op. cit., p. 194 (our emphasis).

32
J. Robinson, The Production Function and the Theory of Capital, in Capital
and Growth, pp. 47–64.

33
P. Garegnani, Heterogeneous Capital, the Production Function and the
Theory of Distribution, in A Critique of Economic Theory, edited by E. K. Hunt
and J. G. Schwartz, London 1972, pp. 245–91.

34
Grundrisse, p. 101.

35
Ibid., p. 101.

36
Note the relevance of this issue to the debate on the transformation prob-
lem (cf. footnote 5, especially the BCSE articles).

37
M. Cogoy, in an article entitled ‘The Fall of the Rate of Profit and the
Theory of Accumulation’, which appears in the BCSE, Winter 1973, argues
that the data of ‘capital per employed worker’ gives support to the notion of
a rising organic composition of capital’ (op. cit., p. 62). He produces some
data which indicates a clear rise in the amount of ‘capital per employed
worker’ in the us in the last century. However, this data does not take ac-
count of the massive rise in the productivity of labour in this period. The
amounts of ‘capital per worker’ would have to be divided by some index of
labour productivity before they gained any resemblance to the value of con-
stant capital per worker. In the case of the ‘capital-output’ ratio the relation-
ship is analogous but much more simple:

Output per employed worker is an index of labour productivity. In conclu-

sion, the amount of ‘capital per employed worker’ bears no direct inverse
relation to the rate of profit, unless it is deflated by some index of labour
productivity. Secondly, Cogoy quotes some figures which indicate that the
capital goods sector (Dept. 1) has grown much more rapidly than the con-
sumer goods sector (Dept. 2). Like Mandel, he draws the erroneous conclu-
sion that this necessarily means that the organic composition of capital has
risen. Such an assertion must be based on the conviction that Dept. 1 has a
higher organic composition than Dept. 2. There is no basis for this convic-
tion, as the values, i.e. the amounts of socially necessary labour time re-
quired to produce the respective goods have to be taken into account. The
indications are that the organic composition is much lower in the capital
goods sector, as a result of high productivity. Kuznets’ data considered be-
low justifies a similar conclusion for the ‘capital-output’ ratios. If a
Bortkiewicz-type formula for the rate of profit, which is superior to the one
produced by Marx, is inspected, it will be seen that the maximum rate of
profit depends on the organic composition in Dept. 1 only. This gives an ad-
ditional refutation of the Cogoy-Mandel thesis, even if the organic composi-
tion is lower in Dept. 2.

46
M. Dobb, Political Economy and Capitalism, London, 1940, p. 110.

47
L. Althusser, For Marx, London, 1969, p. 113.

48
See G. Stedman Jones, Engels and the End of Classical German Philosophy,
in New Left Review 79, May-June 1973.

49
L. Colletti, From Rousseau to Lenin, London 1972, and Marxism and Hegel,
London 1973.

50
Capital, Volume 1, p. 8.

51
See L. Colletti, From Rousseau to Lenin, pp. 62–72.
52
Ibid., pp. 10–17.

53
A. Gramsci, Prison Notebooks, London 1971, pp. 167–8, 336–7, 342–3.

54
Ibid., p. 171.

55
A. Glyn and R. B. Sutcliffe’s British Capitalism, Workers and the Profits
Squeeze, London 1972, is a pioneering example of ‘interventionist’ Marxian
political economy. Despite its weaknesses it has had an important effect on
the revolutionary socialist movement.

1
Capital Vol. 1 Chapter 18.

2
Ibid., p. 533.

3
Capital Vol. 3, Chapter 4 (written by Engels).

4
H. D. Dickenson ‘The Falling Tendency of the Rate of Profit In Marxian
Economics’ Review of Economic Studies, February 1957 pp. 120–30.

5
S. H. Mage The Law of the Falling Tendency of the Rate of Profit, Columbia
University Ph.D. Thesis.