The tendency of the rate of profit to fall 
and post-war capitalism

A critique by AG

This article explains why I disagree with one important element of our analysis of the economic crisis in Britain, and in the advanced capitalist countries as a whole. To clarify what is at issue it is worth emphasising the following points where I believe there to be basic agreement:

1. There has been a systematic and pronounced fall in the rate of profit in the UK, and to a varying extent in the other advanced countries, since the early 1960s.

2. This fall in the rate of profit flows from the accumulation of capital.

3. The fall in the rate of profit has been a fundamental contributory factor to the extreme economic difficulties of recent years (the world economic crisis) and to the specific crisis of UK capital.

4. The course taken by the economic crisis cannot be deduced simply from the fall in the rate of profit. The uneven development between the different capitalist countries, breakdown of the international monetary system, class struggle over wages and state spending, the colonial revolution, etc., all profoundly contribute to the generation of, and affect the form of, the crisis. More- over, the economic crisis itself worsens the fall in profitability through the stagnation which results.

5. In the economic crisis a prime objective of the capitalists is to restore the rate of profit by holding down real wages and by forcing up productivity through rationalisations, that is, by improving the conditions for producing surplus value. But this is only a situation for a particular (national) capital to the extent that it succeeds more than its rivals and can therefore solve the problems of realising surplus value (selling the extra surplus produced by the working class) at their expense.

My disagreement with the tendency concerns why there has been a falling trend of profitability. It is concerned, therefore, with a more precise explanation of point (2) above.

The position of the tendency, explained repeatedly in our material, is that the decline in profitability reflects the operation of Marx's famous Law of the Tendency of the Rate of Profit to fall (LTRPF for short). A typical formulation from the October 1974 IB is as follows:

"Why does capitalism inevitably tend to crisis? Most crises of capitalism are crises of overproduction... Since the capitalists produce only for profit it follows that capitalist crisis is a crisis of profitability. Too much is produced only in relation to the existing level of profitability... The general tendency of capitalist production is towards a higher organic composition of capital, that is towards a higher proportion of machinery and raw materials employed per man. But although machinery makes human labour more productive, only labour produces surplus value. There is therefore an inevitable tendency for the rate of profit to fall."

My objections are that:

A. The LTRPF as formulated by Marx is theoretically incorrect. In brief, one of the factors which Marx regarded as just an offsetting tendency, the cheapening of the elements of constant capital, could in principle operate indefinitely with sufficient intensity to prevent the organic composition from rising.

B. In actual fact the available data show conclusively that while there has been a rise in the organic composition in some countries, it has not been a systematic and pronounced trend, and even where it has taken place it has not been the explanation for declining. profitability.

These objections are justified at considerable length below, and inevitably the discussion is somewhat technical. But first it is necessary to take up two preliminary questions.

Does it really matter?

Some people may feel that it is a complicated and obscure question of no real significance. Complicated it is, and obscure as well. But insignificant it is not. Even if we can all agree on the five points I listed at the beginning, it is still important that we have as precise as possible an understanding of the forces at work. If the LTRPF is not the explanation of declining profitability then this can affect our interpretation of many detailed questions, such as the implications of new technology as a possible way out the crisis or the specific nature of the UK crisis.

The LTRPF and revolutionary Marxism

Quite the opposite line of though is that the LTRPF is such a fundamental part of "Marx's economics" that it is impossible to criticise it without abandoning historical materialism and thus revolutionary Marxism. It is actually impossible to conduct a rational discussion when one side rules out the other's argument, whether theoretical or empirical, as impermissible because the implications are unacceptable. Since it is my observation that some people hold this position, it is obviously necessary to show first of all that the LTRPF is not in fact an essential component of revolutionary Marxism.

It must be said that the LTRPF was very important for Marx himself - after all he called it the "most important law of political economy". The first general point to make is that Marxism is a method, not a set of dogmatic truths. Quite correctly we constantly repeat this as regards politics; we openly argue that Marx or Lenin or Trotsky were, in our view, wrong on certain questions; for example, Marx's general presumption was that proletarian revolution would take place first in the advanced countries. These "corrections" do not, however, in any way invalidate Marxism as the only way of understanding historical development (the fundamental role of economic development, the nature and basis of class struggle, etc.). Exactly the same applies to economics. The Marxist method in economics—piercing the veil of prices to uncover the real relations of capitalist production and exploitation, the fundamental role in capitalism of capital accumulation both in explaining development and crises—does not stand or fall on the specific application of that method by Marx himself. The LTRPF is a specific application of his method, not part of that method itself. To give a very simple analogy: adding 2 + 2 is a specific application of the method of adding, not the method itself; if somebody incorrectly concluded that the answer was 5 that would not invalidate the laws of arithmetic.

I know that one reason that some people feel that the LTRPF is somehow fundamental is that it summarises in a succinct way the historical limitations to the capitalist system, that capitalism undermines itself in the course of its own development. After all, if capital accumulation is based on profitability, and yet capital accumulation inexorably drives down the rate of profit, the self-contradictory nature of capitalism, is apparent. Now nobody I think has a crude "final breakdown theory" of capitalism, that the economic contradictions cause capitalism simply to "stop" and socialism to take over. We all accept that revolution will only take place through the conscious action of the organised workers, and that the possibilities for this are opened up through the contradictory development of capitalism rather than having to wait on some final breakdown. And yet I think many people still cling to the LTRPF as giving some underlying basis to the very important idea of Marx that "the barrier to capital is capital itself."

My answer to this is that the LTRPF is a very economistic technological interpretation of the idea that the barrier to capital is capital itself. After all, as Marx explained repeatedly, capital is not merely a bundle of physical objects (means of production like machinery and factories), but also a social relation (that is, capital is means of production operated by wage labour and owned by capitalists in competition with each other). The idea of capital as a barrier to itself does not depend on some technologically based impossibility to continue accumulating (technological in the sense that Marx says the LTRPF is the reflection under capitalism of the higher labour productivity resulting from technical progress). It may include such a technological component, but the barrier itself can just as much derive from the fact that the development of the relations between capital and labour, and between the different groups of capital (within a country or between countries), constitutes a barrier to continued development. The barrier then is the class-ridden and unplanned nature of the capitalist system.

I would suggest that this is actually a much more fundamental sense in which capital is a barrier to itself than is the LTRPF. It is interesting that in the 'Communist Manifest' there is no mention of the LTRPF; there could not have been in the first edition of course since Marx had not worked out the law. Nobody could say that Marx was not a Marxist then just because he did not understand the LTRPF. But it is striking that it is never mentioned in any of the later prefaces to the Manifesto; nor is it mentioned in the famous chapter on "Historical Tendency of Capitalism" in Volume I of Capital where again the basic ideas of the progressive and contradictory nature of the development of capitalism are described. Whilst the LTRPF is only developed in Volume III, the drafts of Volume III were already written when Volume I was published. Neither does the LTRPF rate a mention in 'Wages, Prices and Profits' nor in Engels' Socialism, Scientific and Utopian. It figures neither in Lenin's Imperialism nor in Trotsky's Curve of Capitalist Development, nor in his famous Report on the World Economic Crisis to the Third International's Third World Congress. Now, of course, some of these works are brief and deal with a variety of questions (though the LTRPF rates only one mention in Luxemburg's mammoth Accumulation of Capital). But it does suggest that our emphasis on the LTRPF is perhaps a great deal stronger than in the classic works of Marxism. Certainly the fundamental idea that the barrier to capital is capital itself, which is repeatedly discussed in these works, was not conceived of as reducible to the LTRPF.

I think that perhaps the most important reason why people believe that it is essential to maintain the LTRPF as the explanation of declining profitability is fear of the political implications of the alternative. It is easy to show that, in Marx's terminology, a fall in the rate of profit must evolve either because of rising organic composition of capital or because of a fall in the rate of exploitation (wages rising in real terms faster than the productivity of labour). People admit that in some circumstances, particularly when the demand for labour is high at the top of the boom, then real wages may rise rapidly and force the rate of profit down. But they find it impossible even to consider that rising wages could be the basic underlying factor behind the falling trend in profitability. They feel that to allow that wages play a central role in declining profitability involves accepting the argument of the employers that the workers are to blame for the crisis. But this argument should be combated by pointing out the false terms in which the debate is posed. That, after a long period of boom, workers' wages may be excessive from the point of view of the smooth functioning of the capitalist system is in no way to "blame the workers" for the malfunctioning of capitalism. After all we all accept that the capitalists' only way out of the crisis is to drive down wages (and force up productivity). In this sense everybody accepts that wages may be excessive from the point of view of the capitalists. Why then should we rule out the idea that they are excessive, because they have risen too fast? The blame surely can be placed firmly on the nature of the capitalist system which, despite all permanent waste and misused resources, reacts to rising living standards through a crisis which makes the waste and unused resources even greater. The role of rising wages in declining profitability is a subject for empirical investigation rather than for simply turning the propaganda of the capitalists on its head.

Having argued, I hope convincingly, that it is perfectly possible to question the LTRPF without abandoning Marxism, I want to consider first the theoretical objections to LTRPF and then assess the evidence as to whether in, in fact, rising organic composition of capital has been the explanation of declining profitability.

The Theory of the LTRPF

The most important criticism of the LTRPF is that it makes unjustifiable deductions about the value composition of capital from perfectly plausible assumptions about the technical composition of capital. The technical composition of capital is a physical relation between the "mass" of machinery, raw materials, etc. used and the number of workers. The substitution of machinery for labour, which Marx perfectly correctly saw as a persistent tendency of capitalist production, leads to a rising technical composition of capital. (When constant capital consists of a range of different machines, raw materials, etc. the "mass" of constant capital can only be understood in a rough and ready way—how could you measure the "mass" of two lathes plus a building? It is simpler to think of it when constant capital is just one type of machine, in which case the technical composition would show the number of machines operated by each worker.)

But part and parcel of the process of a rising technical composition is a rise in the productivity of labour. This means that the value of each commodity—the socially necessary labour time required to produce them—is falling. Thus the value of constant capital which each worker uses is subject to two contradictory processes.

(a) The rising technical composition of capital means he/she is using a greater mass of materials and machinery which, other things being equal, would mean a greater value of capital,

(b) The value of each machine, chunk of raw materials, etc. is falling, which would mean that, other things being equal, the value of constant capital which he/she operates would fall.

The movement of the value of capital depends on which of the two contradictory processes at work—the rising mass of constant capital and its devaluation—is the dominant one. If the technical composition of capital rises faster than the devaluation of capital then the value of constant capital rises per worker, and vice-versa. Contrary to Marx's assertion, it is impossible to argue for a persistent tendency for the technical composition to rise faster than capital is devalued—the outcome is indeterminate. Thus, while it is possible that over particular periods the value composition of capital will rise, it cannot be held to be a fundamental tendency of capitalist production. (See Appendix I for explanation of why analysis based on the "ratio of dead to living labour" [C/(V + S) is clearer than with the more usual measure of organic composition C/V.]

If the ratio of dead to living labour does rise over a particular period this may or may not be associated with a fall in the rate of profit. As is well known, it depends whether the rate of exploitation is increased sufficiently. What is true is that if the ratio of dead to living labour rises indefinitely then the rate of profit must fall (because of the limits to the increased rate of exploitation—see Appendix 1). But such a persistent continual rise in the rate of dead to living labour is only one out of many possibilities. It is possible that the capitalists, forced by competition to cheapen their commodities and seeking to increase as far as possible the margin of profit they earn on each commodity, might persistently introduce new techniques of production of a higher and higher ratio of dead to living labour which must eventually depress the profit rate. (See Appendix 2 on Profit Rates and Margins.) But it is equally possible, when account is taken of the cheapening of machinery due to technical progress in the capital goods industries, that the ratio of dead to living labour will not increase, so the capitalists search for cheaper commodities, and the maximum profit margin and mass of profit would not lead to a fall in the profit rate. There is no logical basis for giving one of these possibilities (falling rate of profit, reflecting rising ratio of dead to living labour) any theoretical priority at this level of abstraction and divorced from the other aspects of the accumulation process.

While Marx recognised that the cheapening of constant capital due to productivity increases tended to hold back the value of constant capital he said that this could not over-ride the increase due to rising technical composition. But this is pure assertion. It cannot be settled as a matter of logic as to which of the forces—rising technical composition or devaluation of capital—operates more strongly. Particular numerical examples based on particular rates of productivity growth in particular sectors, which can always be constructed to show the value composition of capital rising, prove nothing about a general tendency (note 1). If I am travelling from Oxford to London and you are travelling from London to Oxford whether we meet in the middle or nearer London or nearer Oxford depends entirely on our relative speeds of travel. Special cases prove nothing; it could only be argued that there will be a persistent tendency for us to meet closer to London if in principle and of necessity I had faster transport. But no factor of this sort can be brought forward to sustain Marx's argument that there must be a general tendency for the value composition of capital to increase. This leaves the question as being an empirical one. Once it is recognised that theoretically the value composition of capital may rise or may fall it is possible to assess the available evidence in a rational fashion, and it is impermissible to reject some evidence as "wrong" simply because it does not support what is regarded as theoretically necessary.

The Facts:

The fundamental question which we need to answer is whether there has been a general tendency during the post-war boom for the value composition of capital to rise or fall. The statistics presented below are from the conventional national income and capital stock figures constructed by the statistical offices of the different countries. They are not, of course, constructed with Marx's value categories—value of capital, surplus value, etc.—in mind. They are measured in terms of prices—either 'current prices' or recalculated at some set of 'constant prices'-that is what the commodities or means of production would sell for in terms of money. They are not measured in terms of Marx's values—quantities of labour time socially necessary to produce the commodities. However, it is perfectly reasonable to use these figures as indicators of trends in the value magnitudes.(note 2) The two trends would only diverge in particular cases (associated with different movements in the organic composition in different sectors) which there is no evidence to suggest actually applied. Any objection to these figures, which are without question the best available, must be based on a scientific critique of their basis, not on vague assertions as to the dangers of using bourgeois statistics.

The typical figures used to back up the LTRPF are the huge increase of fixed capital per worker, such as these shown in columns 1-3 below.

TABLE 1  

Industry 1953-72

  Source: Boltho, Economic Survey of Japan

Huge increases in the capital stock per worker are recorded (column 3).(note 2) But these are indicators of the technical composition, not the value composition. These statistics for the capital stock at constant prices are attempts to measure the volume of the capital stock (i.e. number of machines before taking account of their cheapening due to productivity growth). They do not simply get rid of the effect of inflation, but they also ignore productivity growth—the devaluation of capital, which cheapens machines. If we want to get at the value composition, i.e. what is relevant for the rate of profit which is calculated on the value of capital—not its physical volume—we have to account for this devaluation of capital. This I have done in a simple way by subtracting the growth of productivity (column 4) from the growth of the volume of capital per worker to give the growth in the value of capital per worker.(note 3) Whilst marked increases do emerge for Germany and the UK, the other countries show slight falls. These figures are not definitive, but it is most unlikely that they would fail to pick up a sustained and general rise in the value of capital per worker. At least they show clearly how grossly misleading it is for writers like Mandel to use figures for the growth of the physical stock per man (column 3) as supporting the idea of rising organic composition. Confirmation of these trends (or lack of them) comes from more accurate figures I have assembled below for the capital/output ratio, an indicator of the value composition of capital (or the ratio of dead to living labour (see Appendix 1).

TABLE 2  

Ratio of dead to living labour

  Source: National Income and Capital Stock Statistics

The picture since the early '50s would seem to be of no real trend in the USA or France, a marked downward trend in Italy and Japan (which were industrialising exceptionally rapidly) and definite upward trends in Germany and the UK.

This definitely does not add up to a general and systematic trend towards a higher composition of capital in the course of the long-boom. But if the value composition of capital failed to rise generally, perhaps the rate of profit failed to fall generally. Perhaps we do not need an alternative to LTRPF as an explanation for falling profitability if the rate of profit actually did not fall. This, however, is not the case. There can be absolutely no ambiguity whatsoever—there was a general and decisive fall in the rate of profit even before the post-1973 crisis (and with the apparent exception of France).

TABLE 3  

Rates of Profit (% before tax)

Source: Clarke and Williams; Feldstein and Summers, Deleste and Mairesse; 
Annual Report on National Income Statistics of Japan, 1979; 
estimates for Italy and Germany based on National Account and Capital Stock 
sources are very rough and cannot be compared with the other countries.

The fact that the rate of profit fell more (or rose less) than can be accounted for by the rise (or fall) in the ratio of dead to living labour means that the surplus value received by the capitalists as pre-tax profits must have fallen as a share of the value of output. (The rate of profit (P/K) is the ratio of living to dead labour (O/K), multiplied by the share of profits in the value of output (P/O), where P is total profits, K is the stock of capital and O is the value of output.)

The Profit Squeeze  

Whilst there has been no systematic and general trend in the ratio of dead to living labour, there has been a persistent and general decline in the share of profits—the infamous "profit squeeze"! Any understanding of the decline in profitability must come to terms with this phenomenon.

TABLE 4  

Manufacturing Profit Margin as a percentage of the Value of Output

  Source: National Income Statistics for Industrial Countries

It is important to be quite clear what is implied by this profits squeeze. As a matter of arithmetic a decline in the share of profits means that the cost of employing labour in real terms has risen faster than productivity (note 4). It does not necessarily mean that the workers' real take-home pay has risen faster than labour productivity because real take-home pay reflects also:

(a) the taxation (including national insurance contributions) paid out of the cost to capitalists of employing workers (the worker only gets the net wage after deduction of taxation and national insurance);

(b) the cost of things which workers buy may rise at a different rate than the price of things they produce (most obviously when the price of imported food and materials went up very sharply after 1972 workers' real take-home pay rose much more slowly than the real cost, measured in terms of the price of the things the workers produced, of employing labour) (note 5).

I believe there are really two types of explanation for the generalised profits squeeze since the early sixties and prior to the post-1973 crisis. The first type of explanation emphasises the militancy and organisation of the working class which enabled them to force up money wages and which the capitalists were unable to pass on in the form of higher prices as a result of international competition (Glyn and Sutcliffe). The argument is made less subjective when it is pointed out that the huge increase in state spending has led to a sharp rise in the burden of taxation paid by workers (and after 1972 the sharp increases in the cost of living as imported food and materials shot up) so that these large increases in real gross wages were necessary to secure much more modest increases in take-home pay (Yaffe, Bacon and Eltis). It is located much more firmly in the historical context when it is emphasised that the strengthening of workers' organisations, and increased militancy itself reflected the achievement of something approaching full employment in most of the advanced countries in the early to mid-sixties. The exhaustion, or partial exhaustion of a massive reserve army of labour—the unemployed, the refugees, the mass of semi-employed on tiny agricultural holdings and in small service sectors—which still existed in the early fifties was one of the historic achievements of the long-boom of the fifties and sixties. By 1973, for example, the share of employment in agriculture in the OECD countries had fallen to 8% from 22% in 1954; fully half of the increased number of wage-earners employed by capitalists had been recruited from the ranks of the self-employed.

Indeed, the alternative explanation of the profits squeeze, (which I now agree with) is that it is the direct economic result of the exhaustion of the reserve army of labour—the excess or over-accumulation of capital with respect to the supply of labour meant that there was excess demand for labour, employers bid up real wages in the scramble to find workers, and so the share of profit and the rate of profit fell. The militancy of workers was very important in determining the form that the crisis subsequently took, and is especially relevant to the question as to why inflationary policies cannot be easily used indefinitely to restore profitability. But fundamentally the profits squeeze would have happened anyway, simply through the mechanics of the supply and demand for labour.

It should also be noted that the sharp increase in real labour costs (gross wages) actually performs a necessary function for capital in a situation of a tight labour market. As the capitalists accumulate new and improved means of production the only way they can secure the additional labour needed to operate them is for this labour to be released from old means of production. But capitalists only stop using old means of production when they are not profitable. What makes them unprofitable is a rise in real labour costs (that is, the gross wage in ^ relation to the price of the product). So a rapid rise in real labour costs, and associated profits squeeze, performs a real function under capitalism of ensuring the transfer of labour to where it is most productive (it is precisely how the 'law of value' operates). Of course increased state spending (predominantly financed by taxation of workers) meant that much of the benefit of the high demand for labour was received by workers indirectly (through health, education, pensions, etc.) rather than in take-home pay.

The basic idea of over-accumulation in relation to the supply of labour was discussed by Marx in the first section of Chapter 24 of Capital, Vol. 1, but is then pushed to one side by his analysis of the introduction of labour-saving machinery which would tend to preserve the reserve army. But he returns to the question in the following long quotation in Vol. III:

"Over-production of capital, not of individual commodities—although over-production of capital always includes over-production of commodities—is therefore simply over-accumulation of capital. To appreciat