Introduction

of capital in its value expression tends to increase, there is an inbuilt tendency for the average rate of profit to decline in the capitalist system.30

To be sure, Marx explicitly speaks about a tendency, not an uninterrupted linear development. He stresses that there are powerful countervailing forces at work under capitalism, to neutralize or even reverse the operation of the tendency of the average rate of profit to decline. Other forces tend, at least partially, to slow down the operation of this tendency.

The most important countervailing force is the possibility for the capitalist system to increase the rate of surplus-value. Indeed, from a purely ‘technical’ point of view, it might appear that the increase in the rate of surplus-value could indefinitely compensate for the increase in the organic composition of capital. If we change the determination of the rate of profit s/c + ν by dividing both the numerator and the denominator by v, we get the formula pr’ = s/ν/c/ν + 1. In other words, the rate of profit is directly proportional to the rate of surplus-value s/ν and inversely proportional to the organic composition of capital c/ν. If the rate of surplus-value increased in the same proportion as the organic composition of capital, the rate of profit would cease to decline.

However, a moment’s reflection will show that such a proportional increase in the rate of surplus-value and the organic composition of capital is impossible in the long run. Theoretically, the organic composition of capital can rise to infinity. That is what it would be in fully automated production, from which living labour would be totally excluded.³¹ But the rate of surplus-value cannot rise to infinity. As long as living wage-labour is employed, no level of productivity (including that of fully automated factories) is imaginable in which workers reproduce the equivalent of all the consumer goods they need to reconstitute their labour-power in a couple of minutes’ or even a couple of seconds’ work. Indeed, the higher the existing level of productivity of labour and the higher the socially recognized average wage (real wage), the harder it becomes to increase the rate of surplus-value substantially, without seriously lowering real wages – which, besides provoking a sharp social and political crisis, would create a tremendous problem of overproduction (for the mass of use-values, including in the wage-goods department, increases even more quickly than productivity of labour and accumulation of capital).32

Furthermore, once we near complete automation, s – which is not a proportion but an absolute mass – starts to decline rapidly together with v, as the number of wage-earners and the total number of labour-hours diminish steeply. Indeed, in a fully automated economy, surplus-value would disappear altogether, as living-labour inputs in the process of production would have disappeared. So it would be absurd to consider formally a ‘rate of surplus-value’ 0/0, when surplus-value itself would no longer exist.

Other countervailing forces enumerated by Marx include: the cheapening of elements of constant capital (both raw materials and machinery) which obviously, by slowing down the growth of c/ν, simultaneously slows down the decline of the rate of profit; the quickened turnover of capital, since the annual mass of profit is a function of the number of production cycles which an identical circulating money capital can perform (this turnover is, in turn, a function both of a quickened circulation process – i.e. more rapid transport and sale of commodities – and of a shortened production process, a quicker pace of production, etc.); foreign trade, with the outflow of capital towards countries with a lower organic composition of capital; and, in general, the extension of capital investment into hitherto non-capitalistically organized branches of output, where initially the organic composition of capital is considerably lower than in traditional industry.³³ A lowering of real wages, by raising the rate of surplus-value over and above the increase which normally results from a growth of productivity of labour in the wage-goods industry (which is – or can be – accompanied by stable and even rising real wages), will likewise put a brake upon the decline of the rate of profit.

Finally, Marx does not mention in Chapter 15 of Volume 3 what he had stressed in Chapter 14: that the decline in the rate of surplus-value can be (and normally is) accompanied by a rise in the mass of surplus-value – and, therefore, in the mass of profit. While this is not, in and of itself, a countervailing factor with respect to the tendency of the rate of profit to decline, it is clearly a countervailing factor with respect to some of the economic consequences of that tendency. It is obvious that the capitalist class will not significantly lower its investments (let alone close up shop altogether) when its profits rise from $100 to $200 billion, just because these $200 billion now represent ‘only’ a 5 instead of an 11 per cent return on total capital. It will look for many ways to redress this regrettable evolution, but it will definitely not be overtaken by panic or despair.

Traditionally, Marxists (and academic economists specializing in the theory of the industrial cycle) have considered Marx’s theory of the tendency of the average rate of profit to decline within two specific – and very different – time-spans: inside the industrial (or business) cycle itself; and over the ‘secular’ time-span of the overall historical existence of the capitalist mode of production (for whose capacity or otherwise for indefinite survival it is a vital question). The ‘theory of collapse’ (Zusammen-bruchstheorie), which relates to the latter time-span, will be dealt with at the end of this Introduction. As for the correlation between the ups and downs of the rate of profit and the business cycle, there is a wide consensus today between Marxists and academic economists specializing in business-cycle studies.³⁴ There remains, however, a third, intermediary, time-span to which hitherto too little attention has been paid: that of the ‘long waves’ of capitalist development, i.e. the successive periods of quicker and slower growth of the capitalist economy as a whole.

There is overwhelming evidence that on at least three occasions – after the revolutions of 1848; around 1893; and at the beginning of the Second World War in the United States, at the end of the forties in Western Europe and Japan – there was a significant increase in the average rate of growth of capitalist production. Such an increase in the rate of growth is synonymous, from a Marxist point of view, with a stepped-up tempo of capital accumulation. And a long-term increase in the rate of capital accumulation is inconceivable, within the framework of Marxist economic theory, without a sudden and sustained upsurge instead of decline in the average rate of profit.

In order to make this real history of the capitalist mode of production comprehensible, against the background of Marx’s tendency of the rate of profit to decline, we must examine the conditions which prevailed immediately prior to these three turning-points and at the start of the ‘expansionary long waves’. In this way, we shall be able to ascertain to what extent the ‘counteracting factors’ enumerated by Marx combined in a particular way to neutralize, or even reverse, for a longer period than normally occurs at a certain stage of the industrial cycle, the tendency of the rate of profit to decline. I have sought elsewhere to demonstrate empirically that this was really the case.35 It is not necessary to repeat that demonstration, but sufficient to state that such temporary neutralization of the law (which Marx also alludes to³⁶) in no way contradicts its general validity. For the ‘expansionary long waves’ are regularly followed by ‘depressive long waves’, in which the tendency of the rate of profit to decline manifests itself in a yet stronger and more durable way than it does during the normal industrial cycle. Its actions can be delayed by countervailing factors, but only for it to reassert itself with a vengeance. That, at least, is the historical evidence to date, and it fully confirms Marx’s analysis. The only additional conclusion to be drawn is that different time-spans have to be articulated with each other, if the concrete operation over time of the tendential law is to be fully grasped.

The very operation of the law (its truth content³⁷) has been increasingly challenged during the last decades by a whole series of authors. This has partially been due to the fact that long-term stepped-up economic growth after the Second World War seemed somehow incompatible – in Marxist terms themselves – with a declining rate of profit. Hence the efforts of Gillman and others to discover new categories like ‘realization expenses’ (presumably to be deducted from surplus-value, which is thus reduced only to ‘surplus-value appropriated by productive capital’) or ‘surplus’, whose supposed growth would explain why the rate of profit as conceived by Marx stops falling, while it continues to fall if conceived otherwise.38 In the meantime, however, events since 1974–5 have caught up with this type of argument, showing that the law more than ever retains its force.

More systematic have been the efforts of the neo-Ricardian school to challenge the law’s validity, on both theoretical and empirical grounds. The main theoretical argument is the so-called Okishio theorem.³⁹ As every capitalist will introduce machinery only if this increases his rate of profit, how can increased profits for every capitalist lead to a decrease in the rate of profit for capitalists taken together?

There are, however, two flaws in this reasoning. In the first place, it is not true that every capitalist will introduce new machinery only if this increases his rate of profit. As Marx himself points out, this is certainly his voluntary inclination, but he may be forced to introduce new machinery, in order to keep his market share or even to save his firm from bankruptcy, i.e. in order to cut his cost price under the pressure of competition, in spite of the effect this decision has upon his rate of profit. In fact, it would be much more correct to say that capitalists will hesitate to introduce new machinery which cuts the amount of profit; but then, the amount (mass) of profit and the rate of profit are two quite different categories. The former may go up while the latter goes down.⁴⁰

In the second place, the argument shows an astonishing misunderstanding of the very nature of the capitalist ‘laws of motion’ of which the tendency for the average rate of profit to fall is so outstanding an example. These laws operate independently from, and in spite of, conscious decisions by individual capitalist firms. In fact, they can be said to be the objective and unforeseen effects of conscious decisions by these firms. No capitalist knows in advance what the real result of his decision to buy new machinery will be. Only when the commodities produced with the help of this new machinery have been sold, and several successive annual balance-sheets have been drawn up, will these results become known. It is, therefore, perfectly possible – indeed inevitable – that the purchase of more machinery by ‘every capitalist’ is intended to increase both his mass and his rate of profit, but that the final end-result of all these decisions will be a situation where the average rate of profit of all is actually reduced.41

As for the main empirical argument put forward by the neo-Ricardians, it states that the organic composition of capital is not rising at all over time but remaining more or less even. In other words, technical progress in the long run is neither essentially labour-saving nor essentially ‘capital-saving’, but neutral.⁴² The index of this alleged stability of the organic composition of capital is an alleged stability of the capital/output ratio over time.’

Now the capital/output ratio is definitely not identical (or parallel) to the organic composition of capital. Nor is the allegedly stable ‘wage part’ in the national income parallel (or identical) to a stable rate of surplus-value. In the case of the capital/output ratio, constant capital is mistakenly identified with fixed capital: i.e. the weight of the value of raw materials, which tends to become a growing part of the value of constant capital (and total capital), is completely eliminated from the reasoning. As for the ‘wage bill’, it mixes together variable capital, which is the payment of productive labour, with the payment of unproductive labour, which comes at least partially out of surplus-value.43 Especially given the steady growth of unproductive labour in the history of late capitalism, the distinction is statistically decisive. In addition to this, Shaikh has demonstrated that the so-called stable capital/output ratio itself should be seriously challenged, from a statistical point of view, and that it corresponds to a large extent to an imprecise or wrong use of statistical categories by bourgeois statisticians.⁴⁴ Initial detailed studies have strikingly confirmed this judgement.⁴⁵

There remains the fact that, as a result of the lack of transparency of real-value relations measured by current market prices, an empirical demonstration of the rising organic composition of capital is not easy to provide on a macro-economic basis, i.e. starting from national-income and gross-national-product statistics. But a close corollary of the organic composition of capital is the part of labour costs in total annual production costs.⁴⁶ Here we are on much more solid statistical ground, since numerous monographs allow us to examine this relation for separate branches of production over time. One would have a hard time discovering a single branch of production in which labour costs constitute a larger part of total current (annual) production costs today than they did on the eve of the Second World War, or at the beginning of the twentieth century – let alone a century or century-and-a-half ago.47 In spite of all the evident tendencies to cheapen the production of machinery and raw materials, which are as inherent in capitalism as is the tendency to cheapen the production of wage-goods, the basic trend of long-term capitalist growth and technical progress has indeed been a labour-saving one. What would the terms ‘mechanization’ and ‘growing automation’ express otherwise, if not precisely this basic trend? One of Marx’s great theoretical achievements consisted in stressing this trend at a time when it was scarcely recognized as historically decisive for the capitalist mode of production.

MARXIST THEORIES OF CRISIS

As I said earlier, Marx did not leave us a completed, fully worked-out theory of crisis. His observations on the industrial cycle and capitalist crises of overproduction are dispersed among several of his major books and a whole number of articles and letters.⁴⁸ Yet it is tempting to see the tendency of the average rate of profit to fall as Marx’s main contribution to an explanation of crises of overproduction, and several contemporary Marxist authors have indeed taken this view.⁴⁹ Is it correct?

My answer would be: yes and no. There can be no doubt about the fact that, within the framework of the industrial cycle, the ups and downs of the rate of profit are closely correlated with the ups and downs of production. But this statement, in and of itself, is not sufficient to provide a causal explanation of the crisis. It can be (and has been) misunderstood in the mechanical sense that crises are ‘caused’ by insufficient surplus-value production 50 – which does not enable capital to become sufficiently valorized; which leads to a cut-down of current investment; which leads to a reduction of employment; which in turn leads to a new and cumulative reduction of income, sales, investment, employment, etc. This process continues till the fall in employment and devalorization of capital have led to a sufficient increase in the rate of surplus-value, and sufficient decrease of the mass of capital, to enable the rate of profit to go up again – which then enables investment, employment, production, income, sales, etc. cumulatively to grow again.

In this vulgar sense, explanation of overproduction crises by the decline in the rate of profit alone is both wrong and dangerous. It is wrong, because it confuses the impossibility of valorizing additionally accumulated capital with the impossibility of valorizing all previously invested capital;⁵¹ because it identifies fluctuations in the investment decisions of capitalist firms with the fluctuations of current surplus-value production. The former, however, may continue to grow when the latter is already declining, and vice versa. The explanation’s main weakness is its concentration on the sphere of production alone, which, in the last analysis, is founded on a confusion about the very nature of the commodity and of commodity production. In the same way as Jean-Baptiste Say’s famous loi des débouchés, it assumes tacitly that there is no specific problem of value realization, only one of surplus-value production. This in turn assumes that what we have under capitalism is production for barter, not production for sale; and that somehow, at least at a macro-economic level, all value produced is automatically realized.

Marx himself explicitly refuted any such assumption. ‘But this production of surplus-value is only the first act in the capitalist production process, and its completion only brings to an end the immediate production process itself. Capital has absorbed a given amount of unpaid labour. With the development of this process as expressed in the fall in the profit rate, the mass of surplus-value thus produced swells to monstrous proportions. Now comes the second act in the process. The total mass of commodities, the total product, must be sold, both that portion which replaces constant and variable capital, and that which represents surplus-value. If this does not happen, or happens only partly, or only at prices that are less than the price of production, then although the worker is certainly exploited, his exploitation is not realized as such for the capitalist, and may even not involve any realization of the surplus-value extracted, or only a partial realization; indeed, it may even mean a partial or complete loss of his capital. The conditions for immediate exploitation and for the realization of that exploitation are not identical. Not only are they separate in time and space, they are also separate in theory. The former is restricted only by the society’s productive forces, the latter by the proportionality between the different branches of production, and by the society’s power of consumption. And this is determined neither by the absolute power of production nor by the absolute power of consumption but rather by the power of consumption within a given framework of antagonistic conditions of distribution, which reduce the consumption of the vast majority of society to a minimum level, only capable of varying within more or less narrow limits. It is further restricted by the drive for accumulation, the drive to expand capital and produce surplus-value on a larger scale’52 (my italics).

Furthermore, this vulgar theory of crises as caused by ‘insufficient production of surplus-value’ is obviously dangerous, from the point of view of defending the working class against the capitalist onslaught which always coincides with a crisis of overproduction. For the conclusion which might be drawn from such an explanation is that the crisis could be overcome and employment rise again, if only real wages were to be cut and surplus-value (profits) thereby automatically increased.53 The working class in general, and the trade unions in particular, are thereby confronted with an agonizing choice between defending real wages and fighting unemployment: i.e. they are made responsible for the loss of jobs. Needless to say, reformist proponents of class collaboration are only too ready to come forward with arguments of this kind, calling upon the workers to make the necessary sacrifices in order to ‘save jobs’ or ‘restore full employment’. Experience, however, has shown time and again that this is not borne out empirically by the real course of the industrial cycle.⁵⁴ It represents an ideological weapon designed to impose the burden of the crisis on the working class and assist an increase in the rate of surplus-value, which is one of capital’s main goals during and after a crisis. ‘Profit squeeze’ theories involve a similar danger of misuse by the capitalist side in the class struggle.⁵⁵

Many extreme proponents of the decline-in-the-rate-of-profit explanation for capitalist crisis will answer indignantly that their analysis contains a built-in reply to employers’ arguments: the decline of the rate of profit is a function of the rising organic composition of capital, which leads to over-accumulation, and not of a decline in the rate of surplus-value. Indeed, they often insist upon the fact that the rate of surplus-value continues to rise until the very eve of the crisis, but just cannot rise enough to offset the effects of the rising organic composition of capital.56 They forget, however, that the rate of profit is a function both of the organic composition of capital and of the rate of surplus-value; that, except in the case of starvation wages, i.e. where any cut in real wages would bring them below the physiological minimum (a situation which no longer exists in any industrialized country), a cut in real wages always implies a rise in surplus-value produced, hence a higher rate of profit than existed before the cut.⁵⁷ We are thus back at square one: to argue that the crisis is exclusively caused by insufficient surplus-value production is to assist the employers’ argument that it can, at least partially, be overcome by a cut in real wages.

This critique of the mechanical and one-sided explanation of crises of overproduction by the falling rate of profit alone can be extended, in a more general way, into a critique of any mono-causal explanation of crises. In the framework of Marxist economic theory, crises of overproduction are simultaneously crises of over-accumulation of capital and crises of overproduction of commodities. The former cannot be explained without pointing to the latter; the latter cannot be understood without referring to the former. This means that the crisis can be overcome only if there occurs simultaneously a rise in the rate of profit and an expansion of the market, a fact which disarms both the employers’ and the reformists’ arguments.

There are three main variants of mono-causal interpretation of Marx’s theory of crisis:⁵⁸

1. The pure disproportionality theory. This sees as the basic cause of the industrial cycle and the ensuing crisis, capitalist anarchy of production: the fact that, under conditions of capitalist market economy, capitalist investment decisions cannot spontaneously lead to ‘equilibrium conditions’ – the correct proportion of value fractions produced and money flows generated in department I and department II, which Marx defined in Volume 2 of Capital. Hence the unavoidable breakdown of equilibrium and the crisis.

The main proponents of this disproportionality theory of crisis were the Russian ‘legal’ Marxist Mikhail Tugan-Baranovsky and the Austro-Marxist Rudolf Hilferding. Nikolai Bukharin was strongly influenced by similar ideas.59 The conclusions of the theory are obvious. If, through the growth of monopolies (a ‘general cartel’, as Hilferding called it), capitalists could ‘organize’ investment among themselves, there would be no crises of overproduction. There would, indeed, be capitalism without crises.⁶⁰ As Roman Rosdolsky has pointed out, however, these theoreticians overlook the fact that the disproportion between production and consumption – the tendency of capitalism to develop productive forces in an unrestricted way, while it imposes strict limits upon consumption by the mass of people⁶¹ – is inherent to capitalism, and independent from the disproportional development of department I and department II due to capitalist competition and anarchy of production (i.e. of investment decisions).⁶²

The grotesque consequences to which mono-causal disproportionality explanations of capitalist crises may lead are best exemplified by Tugan-Baranovsky himself, who seriously argued – and demonstrated ‘mathematically’ – that department I could develop completely independently from department II, to the point where the output of consumer goods would tend to fall towards zero, without such a development causing any crisis whatsoever.63

2. The pure under-consumption by the masses theory of crisis. This sees in the gap between output (or productive capacity) and mass consumption (workers’ real wages or purchasing power) the essential cause of capitalist crises of overproduction, which essentially take the form of overproduction of commodities in department II. Over-accumulation (the decline of investment) and overproduction (or over-capacity) in department I appear as a result of this overproduction (over-capacity) in the consumer goods sector.

While this theory has many non-Marxist ancestors (Thomas Malthus, Sismonde de Sismondi, the Russian Narodniks), its main proponents among Marxists have been Karl Kautsky, Rosa Luxemburg, Nathalia Moszkowska, Fritz Steinberg and Paul Sweezy.⁶⁴ Its weakness lies in its basic assumption (not always clearly understood, but at least clearly expressed, by Sweezy) that somehow there is a fixed proportion between the development of department I and the development of the productive capacity of department II. Since, simultaneously, the growth in the organic composition of capital and in the rate of surplus-value increase the purchasing power for means of production more strongly than they do the purchasing power for consumer goods, the conclusion is obvious: there will be an unsaleable residue of consumer goods.

But not only is this assumption logically unproven. It is contrary to the very nature of capitalist growth, as characterized by growing mechanization or (to borrow a correct formula from the bourgeois economist von Böhm-Bawerk) ‘roundaboutness’ of production. Capitalist growth does imply that a larger proportion of total output takes the form of means of production, although this cannot be accompanied by an absolute decline in the production of consumer goods or a stagnation in the productive capacity of department II. Once this is understood, neither the growth of c/ν nor the growth of s/ν need automatically lead to an overproduction of consumer goods. They will do so only if the fraction

grows more slowly than the fraction

But that such a development is inherent in the capitalist mode of production cannot be mathematically or logically demonstrated.

The danger in under-consumption theories (which, of course, Luxemburg completely avoided) is that they can lead to reformist conclusions, not dissimilar to the ‘harmonicist’ implications of disproportionality theories. The latter state that capitalism could avoid crisis if it ‘organized’ investment. The former tend to think that capitalism could avoid crisis if real wages were larger, or if the government distributed additional ‘purchasing power’ in the form of social security and unemployment disbursements – i.e. ‘redistributed’ national income in favour of the workers, ‘re-transformed’ a part of surplus-value into additional indirect wages.65

What these ‘solutions’ overlook is the simple fact that capitalist production is not only a production of commodities which must be sold before surplus-value can be realized and capital accumulated. It is a production for profit. Any sizable redistribution of the national income in favour of workers’ income, on the eve or in the early stages of a crisis, when the rate of profit has already been declining, means a further decline in that rate of profit through a reduction of the rate of surplus-value (this is, after all, what the ‘redistribution of national income’ is all about). Under these conditions, capitalists will not increase investment, even if sales of previously produced stocks of consumer goods go up. The depression will continue.

3. The pure over-accumulation theory, which sees the main reason for the crisis in the insufficient mass of surplus-value produced, compared to the total amount of accumulated capital. We have already dealt above with the weakness of this theory, and its dangerous implications from the point of view of the proletarian class struggle.

There is also, however, a specific demographic variant of the theory, which stresses the fact that, after long periods of capitalist prosperity, the reserve army of labour tends to disappear, and as a result real wages go up to a point where they cause a sharp decline in the rate of surplus-value and hence in the rate of profit.66 While this eventuality, the border case of what Marx calls in Chapter 15 of Volume 3 ‘absolute over-accumulation of capital’,⁶⁷ cannot be excluded from a general theoretical point of view, in the real history of capitalism – under conditions of extensive international mobility (migrations) of labour and of an even vaster potential for future migrations which exists in underdeveloped countries – any such ‘population pressure’ on capitalism seems centuries removed from us.⁶⁸ It likewise greatly underestimates capitalism’s capacity rapidly to reconstruct a reserve army of labour, by concentrating on rationalization investments which are macro-economically employment-reducing (i.e. by a medium-term increase in the average rate of growth of productivity of labour higher than the average rate of economic growth). This has been strikingly confirmed throughout the 1970s, when the total mass of unemployed in the imperialist (O.E.C.D.) countries, leaving firmly behind the ‘near full-employment’ conditions of the sixties, doubled from ten million in 1970 to twenty million in 1980, while the total number of jobs destroyed in production through technical progress was far larger even than these ten million: millions of immigrant workers from the less industrialized countries had to return to their homelands; millions of women and young people ‘dropped out of the labour market’; very many productive workers were transformed into unproductive ones.

A more sophisticated version, of this theory has been proposed by the Hungarian Marxist Ferenc Janossy, who sees in the inability of capitalism to develop enough skilled (especially highly skilled) workers an unavoidable bottleneck which pushes up real wages at the end of ‘prosperity’.69 But here again the flexibility of capital, both in speeding up skill formation (including at factory level) and in reducing the need for highly skilled labour by technological change, is greatly underestimated.

Proponents of the pure over-accumulation theory of crisis often argue that, as long as accumulation of capital proceeds smoothly, consumption by the ‘final consumers’ automatically grows, as more wage-labour is being employed (generally at increasing wages) and unproductive consumption out of surplus-value also tends to grow. Hence no glut of consumer goods can appear, as long as the decline in the rate of profit has not significantly slowed down accumulation. The first part of the assertion is correct, as far as it goes. The conclusion, however, does not follow at all. The only thing this analysis proves is the fact that consumption (i.e. realization of surplus-value in department II) grows as long as accumulation grows. But it does not prove that consumption grows in the same proportion as does the productive capacity of department II. Indeed, the combined operation of the increasing organic composition of capital in department II and the increase in the rate of surplus-value in the overall economy makes it rather probable that (at least periodically) consumption, while growing, will grow less than productive capacity in department II. In which case, a glut of consumer goods can indeed occur before accumulation has slowed down in the economy taken as a whole.

Elements of a correct theory of capitalist crisis are, of course, present in all three of the mono-causal explanations just outlined.70 They have, precisely, to be integrated with each other to furnish such a theory. The easiest way to set about such an integration, in the light of Volume 3’s basic insistence upon the tendency of the average rate of profit to fall, is by distinguishing a number of successive forms taken, over time, by the accumulation of capital.

In periods of strong upsurge of capitalist production – when business is brisk, current output is easily sold (indeed demand seems to be stronger than supply) and profits are high – there will be an ‘investment boom’ which will run rapidly into bottlenecks in both sub-sections of department I: that of machinery and equipment, and that of raw materials. Both these sub-sections of department I, by their very nature, are less flexible in adapting rapidly to demand than is department II. Hence additional investment, capital accumulation, will occur on a larger and larger scale in department I.71 More means of production have to be produced to produce additional means of production for producing additional consumer goods. Good profit expectations in addition to high profit realizations are the motivation for this boom. Hence, there is a shift of investment towards department I. An uneven development (disproportion) between department I and department II is set into motion.

At a certain point in the boom, two parallel phenomena occur more or less simultaneously. On the one hand, the additional means of production produced come into the production process only after a certain time-lag. But when they enter into that process, they increase the productive capacity in both departments by leaps and bounds. But precisely the relatively high rates of profit and investment imply that real wages and consumer-goods demand from capitalists and their hangers-on could not have developed in the same proportion as this sudden increase in productive capacity in both departments (even if output grows less rapidly in department II than in department I, and even if real wages also grow). Hence a tendency to increasing overproduction (or over-capacity), in the first place in department II.

On the other hand, the massive introduction of new means of production in both departments does not occur with old techniques, but with new up-dated techniques characterized by a basically labour-saving bias, i.e. by an increased organic composition of capital. This presses down the rate of profit, especially since under boom conditions the rate of surplus-value cannot increase in the same proportion, or even does not increase at all.72 Hence a tendency to over-accumulation: part of newly accumulated capital can no longer be invested at the average rate of profit, or is even not invested at all, pushed towards speculation, etc.⁷³

Credit expansion, for a certain time, covers the gap. But it can only postpone the crash, not avoid it. Overproduction now tends to spread from department II to department I.⁷⁴ Growing overproduction of commodities (over-capacity in a growing number of branches of industry), combined with growing over-accumulation, must of necessity lead to sharp cut-backs in productive investment. Disproportionality between the two departments now jumps from an ‘over-extension’ of department I into an ‘under-development’ of that department Investment falls more quickly than current output.

As a result of the crash – which can, but does not necessarily, take the initial form of a credit and banking crash – there is a general collapse of commodity prices (expressed in gold), together with a decline in output and employment. There is a general devalorization of capital, as a result – simultaneously – of this collapse of prices (i.e. of commodity capital), of a large number of bankruptcies, and of a decline in the value of the fixed capital and raw-material stocks of surviving firms. But this general collapse of prices is nothing but the adaptation of market prices and prices of production (through a lower average rate of profit) to the general lowering in the value of the average commodity, which is the unavoidable outcome of the general increase of investment, organic composition of capital and average productivity of labour during the previous period. Capitalists try to postpone this hour of reckoning as long as possible – whence the over-extension of credit, speculation, over-trading, etc. on the eve of the crash. But they cannot postpone it indefinitely.

The effects of the crash, for the system as a whole, are healthy, however nasty they may be for individual capitalists. General devalorization of capital is not accompanied by a proportional reduction in the mass of surplus-value produced. Or (which amounts to the same) an identical mass of surplus-value can now valorize a smaller total amount of capital. Hence the decline in the rate of profit can be stopped and even reversed. Large-scale reconstitution of the reserve army of labour, occurring during the crisis and the depression, makes possible a vigorous increase in the rate of surplus-value, not only through speed-ups but even through a cut in real wages, which in turn leads to a further rise in the rate of profit. Raw material prices generally fall more than the prices of finished goods, so part of constant capital becomes cheaper. The rise in the organic composition of capital is thereby slowed down, again pushing up the average rate of profit on industrial capital. A new cycle of stepped-up accumulation of capital, stepped-up productive investment, can now start, once stocks have become sufficiently depleted and current production sufficiently cut for demand again to outstrip supply, especially in department II.

It follows that the law of the tendency for the average rate of profit to decline is less a direct explanation for crises of overproduction properly speaking, than a revelation of the basic mechanism of the industrial cycle as such: in other words, an uncovering of the specifically capitalist, i.e. uneven, disharmonious, mode of economic growth, which unavoidably leads to successive phases of declining rates of profit, and recuperation of the rate of profit as a result, precisely, of the consequences of the previous decline. This is true at least of the way in which this law operates over the seven–ten-year time-span – leaving aside, for the moment, the memento mori it implies for capitalism in a secular perspective.

There can be little doubt that this multi-causal explanation of capitalist crisis, rather than any of the mono-causal variants, corresponds to Marx’s own conviction, at least as expressed here in Volume 3. In addition to the passage quoted on p. 40 above, three other passages can be cited which leave little room for alternative interpretations:

‘Let us conceive the whole society as composed simply of industrial capitalists and wage-labourers. Let us also leave aside those changes in price which prevent large portions of the total capital from being replaced in their average proportions, and which, in the overall context of the reproduction process as a whole, particularly as developed by credit, must recurrently bring about a situation of general stagnation. Let us likewise ignore the fraudulent businesses and speculative dealings that the credit system fosters. In this case, a crisis would be explicable only in terms of a disproportion in production between different branches and a disproportion between the consumption of the capitalists themselves and their accumulation. But as things actually are, the replacement of the capitals invested in production depends to a large extent on the consumption capacity of the non-productive classes; while the consumption capacity of the workers is restricted partly by the laws governing wages, and partly by the fact that they are employed only as long as they can be employed at a profit for the capitalist class. The ultimate reason for all real crises always remains the poverty and restricted consumption of the masses, in the face of the drive of capitalist production to develop the productive forces as if only the absolute consumption capacity of society set a limit to them.’75 (my italics)

‘Periodically, however, too much is produced in the way of means of labour and means of subsistence, too much to function as means for exploiting the workers at a given rate of profit. Too many commodities are produced for the value contained in them, and the surplus-value included in this value, to be realized under the conditions of distribution given by capitalist production, and to be transformed back into new capital, i.e. it is impossible to accomplish this process without ever recurrent explosions.’⁷⁶ (my italics)

‘The manufacturer may actually sell to the exporter, and the exporter to his foreign customer; the importer may sell his raw materials to the manufacturer, and the manufacturer sell his products to the wholesaler, etc. But at some particular imperceptible point the commodity lies unsold; or else the total stocks of producers and middlemen gradually become too high. It is precisely then that consumption is generally at flood tide, partly because one industrial capitalist sets a series of others in motion, partly because the workers these employ, being fully occupied, have more than usual to spend. The capitalists’ expenditure increases with their revenue. And besides this, there is also, as we have already seen (Volume 2, Part Three), a constant circulation between one constant capital and another (even leaving aside the accelerated accumulation) which is initially independent of individual consumption in so far as it never goes into this even though it is ultimately limited by it, for production of constant capital never takes place for its own sake, but simply because more of it is needed in those spheres of production whose products do go into individual consumption. This can continue quite happily for a good while, stimulated by prospective demand, and in these branches of industry business proceeds very briskly, as far as both merchants and industrialists are concerned. The crisis occurs as soon as the returns of those merchants who sell far afield (or who have accumulated stocks at home) become so slow and sparse that the banks press for payment for commodities bought, or bills fall due before any resale takes place.’77 (my italics)

CREDIT AND THE RATE OF INTEREST

In the same way as Volume 2 of Capital stressed the importance of previous accumulation (and presence) of money-capital, its periodic injection into circulation, and its periodic outflow from the operations of productive capital properly speaking, to make expanded reproduction (i.e. economic growth) possible for ‘capital in general’, Volume 3 stresses the key importance of credit for ‘many capitals’, i.e. for the fluctuations of the industrial cycle under conditions of competition.

The appearance of a generally known average rate of profit unavoidably leads to an equalization of the rate of interest too. Surplus-value is, first of all, split between profit for entrepreneurial capital (industrial profit, commercial profit, banking profit, and profit for agricultural entrepreneurs as distinct from passive landowners) on the one hand, and interest on the other. Through the capitalist banking system, all available money reserves (savings and non-invested surplus-value + idle money capital resulting from non-investment of part of surplus-value realized during previous cycles) are transformed into functioning capital, in other words lent to capitalist firms which are actually operating – i.e. employing wage-labour – be it in the sphere of production or in that of circulation. In this way, capitalists are able to operate with much more capital than they own personally. Capital, accumulation can take place at a much quicker pace than would be the case if each capitalist firm could practise enlarged reproduction only on the basis of the profits it had itself realized.

This constant expansion of credit, which has accompanied the whole history of the capitalist mode of production, at first sight seems to accentuate the tendency of the average rate of profit to decline.78 The total amount of profit distributed among the sum-total of capitalist firms is now lower than the sum-total of surplus-value produced, the difference being exactly the total amount of interest paid out to the passive owners of money capital (which is not to be confused with profits of banks, i.e. the average profits on their own capital, not on their deposits). But this is, of course, a false impression. The average rate of profit is the division of the total amount of surplus-value produced by the total amount of social capital. If, as a result of division of labour among capitalists or over-accumulation, part of that capital is not itself directly productive, in other words, is not engaged in the direct production of surplus-value, this does not change its nature as capital, i.e. value constantly on the look-out for an accretion of value.

Hence, according to Marx here in Volume 3, the effects of credit (like those of trade) on the tendency for the average rate of profit to decline are opposite to what at first sight appears. They in reality tend to put a brake upon that tendency, or even reverse it, as a result of three simultaneous mechanisms which they unleash:

(1) Trade and credit allow capital to rotate more rapidly, thereby increasing the number of productive cycles through which a single sum of money capital can pass in, say, one year, thereby increasing the mass of surplus-value and also the annual rate of profit (since the same amount of surplus-value is produced during each of these productive cycles, all other things remaining equal).⁷⁹ This, by the way, is why industrialists are ready to allow commercial and banking capital to share in the general distribution of entrepreneurial profit (total mass of surplus-value minus total mass of interest), although neither commercial nor banking capital produces surplus-value. Such capital does not produce surplus-value itself, but it helps industrial capital and agricultural capital produce additional surplus-value.

(2) By enlarging the scope and tempo of accumulation of capital in the productive sphere, over and above profits directly owned by industrialists and capitalist farmers, commerce and trade accelerate the concentration of capital, thereby stimulating technical progress and the production of relative surplus-value, which again counteracts the tendency for the average rate of profit to decline.

(3) By the device of joint-stock companies (corporations), credit creates a situation in which a large part of capital, owned by stockholders, is not expected to receive the average rate of profit at all, but is content with the average rate of interest only. Hence, the average rate of entrepreneurial profit is much higher than it would be if all (or the largest part) of capital were directly entrepreneurial capital, i.e. had to receive the average rate of profit.80

The greater flexibility of money capital not tied to any specific firm or branch of industry is, in turn, one of the main reasons why the equalization of the rate of profit can so easily occur and be recognized under capitalism, i.e. why social capital remains relatively mobile in spite of growing capital investment in the form of fixed, relatively immobile capital. Parallel to the reserve army of labour, these huge reserves of money capital are the preconditions for sudden, rapid phases of feverish expansion, which characterize the industrial cycle and the very nature of capitalist growth, uneven and disharmonious. Indeed, the banking system in part plays the role of a social clearing-house, through which capital is constantly being transferred from branches which face stagnating or declining overall demand, to branches which face growing overall demand not satisfied by current production (or productive capacity). The deviations of distinct rates of profits from the average are the guiding mechanism for these transfers. In that sense, Marx stresses the key role of credit in expanding the accumulation of capital to its utmost limits, while at the same time functioning as the main lever for over-speculation, over-trading and overproduction.

It follows that the credit cycle – and the ups and downs of the rate of interest – are partially desynchronized from the industrial cycle properly speaking. During the period of recovery and initial upsurge, money capital is relatively abundant; the level of self-financing of firms is high; the rate of interest is relatively low;81 and the level of entrepreneurial profit is above average. Conversely, at the peak of the boom, during the phase of over-heating and during the crash, money capital becomes scarcer and scarcer; the level of self-financing declines precipitately; demand for money capital grows constantly; and the rate of interest grows by leaps and bounds, not in spite of but as a function of the decline in the average rate of profit. Firms now borrow not to expand business but to escape bankruptcy; not in order to gain additional entrepreneurial profits, but in order to save their capital. At this precise moment of the cycle, the rate of interest, therefore, can actually be above the rate of entrepreneurial profit (which cannot, of course, ‘normally’ be the case). But when, after the crash, the crisis and depression properly speaking set in, investment declines steeply; demand for credit collapses; and the rate of interest starts to slide rapidly, which helps the rate of entrepreneurial profit slowly to pick up again.

MARX’S THEORY OF SURPLUS PROFITS

The fact that Marx’s theory of differential land rent in reality represents a special case of a more general theory of surplus profits has not hitherto been sufficiently appreciated. This is all the more strange in that Marx explicitly makes the point here in Volume 3, in several passages of Parts One and Two, and returns to the question at length in Parts Six and Seven.

The basic approach, once again, is a straightforward application of the labour theory of value. The question whether labour expended in the production of a given commodity is recognized as average socially necessary labour or not is not a simple physical matter of an actual number of labour-hours expended – of a given fraction of society’s total labour potential being used for producing a given commodity.82 It is a function of the total amount of labour expended in all the units producing that given commodity, as compared to the total amount of labour which society wishes to devote to it.⁸³ It is a function of the relation between the productivity of labour in the given productive unit and the average productivity of labour in the branch of industry as a whole.

Marx distinguishes three basic situations of current production, in relation to current social needs (not, of course, physical needs, but needs induced by commodity production and mediated through purchasing power as determined by capitalist norms of distribution – i.e. by the class structure of bourgeois society).

Case 1 concerns situations where there is a normal mobility of capital in relation to a given branch of output. Here, inflows and outflows of capital, regulated by oscillations of prices inducing oscillations of rates of profit, will normally balance out social supply and demand. In that case, equalization of the rate of profit will normally apply to the branch in question. Firms which operate at the average productivity of labour in the branch (which will be the general rule) will receive the average rate of profit Firms which operate below the average productivity of labour will receive less than the average profit, and risk being crowded out of business in situations of crisis and depression. Firms which have made technological advances, which operate at a level of productivity of labour above the average, will enjoy a temporary surplus profit, i.e. a profit over and above the average profit resulting from the difference between their individual costs of production and the average costs of production in the branch. But this surplus profit will generally disappear in periods of crisis and depression, when the new technology will become generalized throughout the branch, and the average productivity of labour (the value of the commodity) adapted to that initially higher productivity.84

Case 2 concerns branches of production characterized by structurally stagnant or declining demand: i.e. ‘outmoded’ ones, with structural overproduction. Here, only firms operating at above-average productivity of labour will receive the average rate of profit. Firms operating at average productivity of labour will receive less than the average rate of profit. Firms operating at below-average levels of productivity of labour will sell at a loss and go out of business. In general, again, when there is normal mobility of capital, such branches of industry will become ‘normalized’ (i.e. revert to Case 1) even before a general crisis of overproduction occurs, through massive closures of productive units.

But then there is also Case 3, which we might characterize as one of structurally (or institutionally) determined scarcity: i.e. the case where an influx of capital is hampered (or prevented) by natural or artificial monopolies.85 In such cases, there is a long-term preponderance of demand over supply. So the firms operating with the lowest productivity of labour in the branch still receive the average rate of profit (i.e. they determine the price of production, or the value, of the commodity produced in that branch).⁸⁶ Firms operating at a higher productivity of labour – at the average of the branch, or a fortiori at an above-average level – receive a long-term surplus profit protected by the very monopoly, i.e. by the powerful obstacle which hinders the influx of additional capital into the branch in question. This surplus profit does not even disappear in times of crisis and depression, although it will obviously be lowered in absolute terms, as a result of the fall in the average rate of profit.

These monopoly surplus profits are called differential rents. In Capital Volume 3, three such instances of differential rent are distinguished: land rent; mineral rents; and technological rents.⁸⁷ Land rent could be sub-divided into agricultural land rent and urban land rent.

Natural monopolies are determined by the fact that access to natural resources necessary for production (from a use-value point of view) is limited, and that these are not reproducible at will by capital. This applies to land as such, especially land of a given use-value (desired relative fertility, desired location); to mineral sources; to climatological preconditions for using land to produce certain specific use-values (e.g. cotton, natural rubber, tropical fruits, etc.).

Artificial monopolies are determined by limits in capital mobility related not to natural conditions but to conditions arising from the results of specific stages (forms) of accumulation of capital itself: concentration of capital (if, in order to start a new firm in a given branch of industry with minimum level of profitability, it is necessary to invest at least £500 million or $1,000 million, this is obviously an ‘obstacle to entry’ for most capitalists); monopoly rights in patents, inventions or research in certain new fields of production (or, which amounts to the same thing, qualitative advantages in the capacity to apply these); organized practices by a small number of firms dominating production in a given field, systematically resorted to in order to keep out potential competitors; and so on.

As clearly follows from this definition, natural and artificial monopolies, giving rise to surplus profits through putting a brake upon free entry of capital into branches of production where the rate of profit is higher than average, are always relative, never absolute. Land is not reproducible. But possibilities for capital investment on existing land can be vastly expanded. Furthermore, internationally, tremendous areas of potentially agricultural land are not yet exploited (in the nineteenth century, of course, these were many times greater than today). So potential agricultural land is still relatively abundant on a world scale. Capitalist technology, furthermore, can be pushed to the point where production becomes possible without the use of land. Mineral resources are finite. But synthetic production of originally natural raw materials (fibres, rubber, oil) is not finite, or at least not to anything like the same degree as natural raw materials properly speaking.

The bigger the initial capital outlays necessary for profitable production, the smaller the number of potential new competitors in a given branch of industry. But conversely, the higher the surplus profits enjoyed in these branches, the stronger the inducement for ‘many capitals’ to band together and risk the huge initial capital investments necessary to obtain a slice of the cake. The more that decisive advances in technology lead to stable surplus profits over longer periods, the stronger the pressure for potential competitors to leap ahead and bypass these advances by a new revolution in technology, etc.88 One may conclude that all monopoly surplus profits are always limited in time and, in the long run, tend to disappear, and that commodities produced in initially monopolized branches tend to be exchanged at their prices of production. Whether this ‘long run’, at least for industrial products produced in monopolized branches under monopoly capitalism (i.e. since about 1890), is the ‘long wave’ – as I hypothesized in Late Capitalism – or not, remains a subject for further investigation.89

In order fully to grasp the relative (never absolute) nature of any monopoly, whether natural or artificial – and thus the limited nature in time of any form of surplus profits under capitalism – it is necessary to reintroduce into our analysis the phenomenon of structural scarcity which was its starting-point.⁹⁰ For it is only if obstructions to capital mobility, i.e. obstacles to increases in output, create conditions under which social demand for the goods produced in that given branch of output is for long periods higher than or equal to the total amount of commodities produced (including those produced under the lowest conditions of productivity of labour, or the lowest fertility of soil in agriculture) that units of production enjoying lower costs of production will be able to realize surplus profits in the form of differential rents (differential land rents, mineral rents, technological rents).

Once, however, social demand for the goods produced in the monopolized branch of industry recedes, or stagnates, or grows more slowly than does production even under conditions of relative monopoly, differential rent will tend to be reduced and surplus profits to decline. (This does not mean, of course, that they will disappear completely, where the monopoly is natural, as long as differences in fertility, etc. still subsist and determine different unit costs on different pieces of land, in different mines, etc.) The huge increases in average productivity of agricultural labour, which have been one of the main characteristics of the development of capitalism in the twentieth century, and have indeed exceeded the rate of growth of industrial productivity of labour, have completely altered the demand/supply relation for basic foodstuffs in the advanced capitalist countries.91 The situation of structural scarcity has been transformed into a situation of structural overproduction, co-determined by the decreasing place of food expenditure in total consumers’ expenditure when real incomes rise (Engel’s Law). Not only has differential rent, therefore, been strongly contracting in these countries, but large tracts of farm land have been reconverted into pastures, while in turn large tracts of pasture have been reconverted into forests or simply waste land. Massive closures of coal pits in the nineteen-fifties, sixties and early seventies, when oil was much cheaper than coal, are a parallel development in mining, with a co-related decline of differential coal-mining rents.

But the process can also be reversed. When social demand – mediated through an increase in market prices – suddenly surges beyond output for, say, ten or twenty years, i.e. when structural scarcity reappears, a massive reappearance of differential rents occurs. This is what has happened in gold production since the collapse of the Bretton Woods system, when it became impossible for the imperialist Central Banks to maintain the gold price at $35 or $42 (35 S. D. R.) an ounce.⁹² The upsurge of the ‘free market gold price’, first to $100, then to $200, finally to more than $600 an ounce, has made many ‘marginal’ mines in South Africa (and elsewhere) profitable again, and led to a feverish development of capital investment in gold-mining. The more productive among the twenty main South African gold mines were producing gold at the end of 1979 at around $95 production costs per ounce (the single most productive mine at $64 an ounce). The less productive of these twenty mines had production costs of around $200 an ounce (with the highest single figure being $265). This situation gives a differential rent of more than $100 an ounce for the former category of mines as against the latter, once gold is selling at more than $200 + average profit: say, more than $240 or $250 an ounce.93

There is a more general reason why the capitalist mode of production produces both a tendency towards monopolization (e.g. as a result of increasing concentration and centralization of capital), and a tendency towards periodic decline of specific monopolies. This is the fact that surplus profits are deducted from the total amount of profit to be distributed among all those capitalists who participate in the equalization of the rate of profit: in other words, they tend to reduce the general cake distributed among all bourgeois except the monopolists. As there is a tendency for that average rate of profit to decline, monopolies of all kinds – including monopoly property in land – tend, therefore, to accentuate that decline. Hence, the pressure of capital to overcome natural or artificial barriers to the mobility of capital: to reduce the impact of monopolistic situations, or even try to eliminate them altogether. The outcome of this constant tug-of-war is a function of the relative strength of different layers of the ruling class. At least in the twentieth century, the pressure has been more successful with regard to absentee capitalist landlords (separate and apart from capitalist agricultural entrepreneurs) than with regard to industrial, transport or mining monopolies, although not a few cases of collapse of monopolistic surplus profits could be cited in these realms too.

This pressure remains, independently of whether one considers the surplus profits (additional surplus-value) of the monopolists to be actually produced inside the monopolized branches of output, or whether one considers them, at least in several cases, as resulting from transfers of value from non-monopolized to monopolized sectors of production. For, in both hypotheses, the mass of surplus-value to be shared out among all capitalists who do not enjoy rents is substantially lower than it would have been with a ‘perfect’ mobility of capital into all branches: in other words, their average rate of profit has been lowered. And when this accentuates a tendency which is already operating for deeper reasons, as has been indicated above, the counter-pressure will be all the more powerful.

THE SPECIFICITY OF CAPITALIST AGRICULTURE

In Volume 3 of Capital, Marx extends a notion which he had already stressed at the end of Volume 1: the key importance of private appropriation of land – the transformation of land into the private property of a given limited class of people – for the very birth, consolidation and expansion of the capitalist mode of production. This mode of production presupposes the appearance of a social class – the modern proletariat – which has no access to means of production and subsistence and is, therefore, under the economic compulsion to sell its labour-power. Means of subsistence are, in the first instance, food, which wherever access to land is free can be produced with minimal means of production. Hence, the creation of the modern proletariat hinges, to a large extent, on barring free access to land to people possessing no capital.

This process of private appropriation of land, which in Western Europe mainly took place between the fifteenth and eighteenth centuries and culminated in the sale of village ‘free’ land reserves (communal lands) unleashed by the French Revolution,94 was repeated throughout the last part of the nineteenth and the whole of the twentieth century in Eastern Europe, North and South America, the Middle East, Africa, Japan and South-East Asia. The most repulsive form of forcible separation of the original population from its fertile land reserves occurred in Eastern and Southern Africa. It is going on to this very day in countries like Brazil, Iran, the Philippines and Mexico (despite the partial achievements of the 1910–17 Revolution).

However, the interrelation between consolidation of the capitalist mode of production, the process of capital accumulation and the struggle of capital against the tendency for the rate of profit to decline is much more complex than this compulsion to transform all land into private property.

For historical reasons, the generalization of private property in land, in Western, Central and a large part of Eastern Europe as well as in Japan, took the initial form of ownership by a social class separate and apart from ‘functioning’ capitalists (i.e. capitalist farmers, entrepreneurs) properly speaking. These capitalist landowners (not to be confused with semi-feudal or feudal landlords) barred entry to their land by the capitalist class in general, unless they received a special ‘unearned’ income in the form of absolute land rent (the same rule applies, of course, to rentier-proprietors of urban land vis-à-vis capitalists engaged in the building industry). In other parts of the world, the phenomenon of private appropriation of ‘surplus’ land has involved other layers of the ruling class: sometimes foreign settlers appropriated it;95 sometimes local landowners, merchants, usurers and other sectors of the ruling class operated in the same way. There are some cases, though rather rare, of combinations in one degree or another of both processes.

But in all cases where actual ownership of the land became separated from capitalist farming, absolute land rent appeared. And as is the case with differential land rent, absolute rent is a fraction of total surplus-value produced by the sum-total of commodity-producing labour, deducted from the residue to be divided between all capitalist entrepreneurs and owners of money capital. This deduction is all the more onerous in that, contrary to differential rent, it is not open to erosion or equalization through the laws of motion of the capitalist mode of production properly speaking (competition, technical progress, increase in the organic composition of capital, concentration and centralization of capital, etc.). It thus puts a brake upon capital accumulation in agriculture. Hence, the organic drive of capital to eliminate the separation of landownership and capitalist farming: by gradually transforming landowners into entrepreneurs, and land-renting farmers into a majority of wage-earners on the one hand and a minority of landowning farmers on the other. The transformation of a situation of structural scarcity of food into one of structural plenty (latent overproduction) in most of the industrialized countries powerfully assists this process.⁹⁶ It represents a tendential disappearance of absolute rent in the imperialist countries.

Behind this process there lies an imperious long-term assertion of the law of value of a deeper kind. The source of absolute land rent is the lower organic composition of capital in agriculture as compared with industry, i.e. the higher mass of surplus-value produced by agricultural labourers as compared with industrial labourers employed by a same amount of total capital.97 The barrier of landownership separated from capitalist enterprise makes it possible for landowners to prevent this supplementary amount of surplus-value from being sucked into the general process of equalization of profit between all capitalists. Thus rent is indeed an obstacle to the full flowering of capitalist agriculture: a source of relative backwardness of agriculture compared with industry, i.e. of agricultural productivity of labour compared with industrial productivity of labour. But Marx, who himself stressed this relative backwardness, noted that it was not a fixed and final characteristic of the capitalist mode of production, but could sooner or later be overcome. But when agriculture becomes more and more industrialized, when the substitution of human labour by dead labour (machinery, fertilizers, etc.) is applied on an ever-increasing scale in that branch of production, when contemporary agro-business arises, the difference in organic composition of agricultural as compared with industrial capital tends to disappear. Consequently, the material basis for absolute land rent disappears likewise. As Robin Murray has aptly expressed it: in the same way that the formal subordination of labour to capital is transformed into a real subordination in agriculture, formal subordination of land under capitalist agriculture is transformed into real subordination of land as a material element in capitalist agricultural production.⁹⁸

The extent of this process of industrialization of agriculture can be measured by the following facts concerning the United States. Between 1915–19 and 1973–7, productivity of labour in wheat and soybean production increased tenfold, when measured by the labour-hours needed to produce 100 bushels. For maize, the increase was actually thirtyfold! Production assets – including livestock and raw materials stocked on farms, thus roughly comparable to constant capital – per farm worker increased fivefold in current dollars between 1963 and 1978. Per capita disposable income per farm worker, however, only increased less than threefold, half of which originated from sources outside farming properly speaking. Wages for hired labour barely doubled during the same period. A good index of the increase in the organic composition of capital, if there ever was one! Simultaneously, the ‘emancipation’ of capitalist agriculture from the use of land has made giant strides in animal husbandry, as exemplified above all by hog-raising, cattle-raising and by the aptly termed ‘broiler industry’. By 1972, 75 per cent of U.S. beef was raised on so-called feedlots, the largest accommodating as many as 125,000 cattle at a time.99

It should be noted that, while absolute land rent originating in the separation of landownership from capitalist farmers (differential land rent does not originate in ownership: ownership only determines who appropriates it) tends to disappear under conditions of ‘industrialized’ agriculture, it reappears in modified form as generalized mortgaging of land owned by small and medium-sized capitalist farmers – in other words, as the transfer of a significant part of surplus-value produced in agriculture to banks and finance capital.¹⁰⁰

However, as I have already emphasized, real capital movements are guided not by the average rate of profit but by deviations from that average. So while capital tends to eliminate absolute rent in the older capitalist countries, it also constantly tends to reproduce it, essentially (but not exclusively) in countries where capitalism has penetrated belatedly. There thus operates, at the level of the world economy, a kind of process of internationalization of land appropriation and creation of absolute land rent.101 Brazil offers some outstanding examples of this tendency.

Finally, since agricultural production is food production, and since food is an essential element of reproduction of labour-power – quantitatively its main element, at least in the earlier phases of development of the capitalist mode of production – there is another, contradictory, element in the relation between capitalism and agriculture. While for (real or potential) agricultural capitalists, the main problem is eliminating the dual structure of landownership and farming enterprise, for (national) capital as a whole, the main short-term problem is to ensure access to food on the cheapest possible conditions, be it through capitalist, semi-capitalist or pre-capitalist modes of production.

This means that capital as a whole has a vested interest, at least during early phases of capitalist development (which are being reproduced today in most semi-colonial countries, even those which are semi-industrialized), in maintaining a substantial part of the peasantry under conditions where it still has access to some land:¹⁰² not enough to provide a minimum basis of livelihood, but sufficient to provide part of the annual food intake of the peasant family, forcing these peasants to look for employment during part of the year. Rising capitalism, therefore, both ruthlessly suppresses free access to land through generalization of private ownership of land, and skilfully defends minifundia, i.e. small-scale parcellized subsistence farms,¹⁰³ which enable wages to be pushed below the subsistence level since this semi-proletarian sub-section of the wage-earning class produces part of its own food. The political and social function of such deliberate policies by bourgeois governments has often been pointed out. They slow down the concentration and permanent urban settlement of the proletariat; they maintain an easily manipulated electoral base, that is less easy to unionize or organize in workers’ parties; and so on. But the economic function of these policies must also be clearly acknowledged. They play an important role today in many semi-colonial countries, especially the more advanced ones. As for the direct exploitation of these miserable ‘private owners’ by capital, it takes the form not of extortion of land rent but of extortion of usury interest, the parcel owners being permanently and increasingly burdened by debt.

The overall evolution of agriculture under capitalism will be a resultant of the interaction of the five, often contradictory, tendencies just outlined. And this resultant becomes, in a certain sense, an index of the degree of maturity of capitalist development in the national economy as a whole. On a world scale, this culminates in a tragic end-result. The internationalization of absolute land rent means a growing gap between the average productivity of labour engaged in food production in the imperialist countries, on the one hand, and in the semi-colonial countries, on the other.104 Both the growing penetration of capitalism into semi-colonial agriculture (with the accompanying phenomenon of increase in commercial as against food crops) and the attempts of bourgeois governments to ‘stabilize’ parcellized subsistence farming tend to increase that gap further. The consequence is that food surpluses on a world scale tend to become increasingly concentrated in fewer and fewer countries, most of them imperialist ones.¹⁰⁵ In other words, differential land rent on the world market is accessible only to a smaller and smaller number of capitalist large-scale farmers (agro-businesses).¹⁰⁶

CAPITALISM AS A SYSTEM
AND THE BOURGEOISIE AS A CLASS

One of the outstanding features of Capital Volume 3 is the way in which Marx ties together economic analysis and social analysis at the level of the system in its totality – i.e. at a higher level than he did in Volume 1, inside the factory (the process of production properly speaking). In Chapters 48 and 51, here, he shows how the reproduction of a specific form of division of the ‘national income’ (annually produced new value) between wages on the one hand, and profits, interests and rents on the other, automatically reproduces capitalist relations of production – i.e. the basic class relations and class inequality which define the system.

It is the greatest theoretical weakness of reformism, under whatever form it appears, not to understand this basic truth. Whether wages are high or low, whether ‘indirect’ wages (social security payments) are inexistent or extensive,¹⁰⁷ they cannot upset the basic class relations and class inequality on which the capitalist mode of production is founded. Wages cannot rise to the point where they substantially lower surplus-value (profits), without setting into motion a massive ‘investment strike’ by capitalism (hence a steep decline of capital accumulation), coupled with a frantic attempt to step up the replacement of living labour by machinery – both processes acting to halt and reverse the rise in wages, through the effects of massive unemployment (and cuts in public ‘social’ expenditure). The one thing it is impossible to do with capitalists is to force them to invest or produce at a loss!

In addition, the very trend towards increased organic composition of capital, towards increased concentration of capital, towards a strong rise in the minimum requirements for founding new productive units in all branches of production, constantly consolidates monopoly ownership of the means of production by the bourgeoisie as a class, making it physically impossible for even the best-paid workers to save enough out of their wages to embark seriously upon an industrial enterprise of their own.108 While this is less true in small retail trade and small service business (or in small-scale farming, during times of acute unemployment¹⁰⁹), the overall trend is very clear. Wages tend to be spent over the whole life-span of the wage-earner. They cannot lead to any serious accumulation of capital.¹¹⁰ So wages do not just reproduce labour-power: they also reproduce a special class under permanent economic compulsion to sell its labour-power. Likewise, private appropriation of surplus-value does not just lead to accumulation of capital: it also reproduces a social class which can monopolize the means of production and, therefore, oblige the wage-earners continuously to sell their labour-power to the owners of capital; continuously to produce surplus labour, surplus-value and profits for the exclusive benefit of the latter.

To be sure, the two processes are not symmetrical. Even when real wages have a tendency to secular increase and ‘workers’ savings’ become a large-scale phenomenon, these do not free the individual wage-earner from his proletarian condition; in other words, they do not ensure him a high enough durable income (money reserve) to enable him to go into business for himself. They just represent ‘deferred consumption’, i.e. an additional insurance fund, over and above socialized ‘indirect wages’ (social security), to complement his reduced income in times of sickness, unemployment or retirement, or to defray such extra family expenditures as might be incurred for the better education or weddings of his children, etc. In addition, there exists under late capitalism a powerful incentive for the capitalist class to deprive workers of the right to dispose of these savings freely, or even to expropriate them tout court – inflation being only the mildest of the various forms of partial or total expropriation to which it resorts.111

On the other hand, the fact that all sectors of the bourgeois class have access to a fraction of the sum-total of socially produced surplus-value, even if their own capital is not directly used by themselves in surplus-value-producing endeavours, does not at all imply that this access is equal for every capitalist. Not only does the appearance of monopolies operate in the opposite direction. The law of concentration and centralization of capital acts even more powerfully to this effect. Stepped-up competition eliminates many more middle and large-scale capitalists (not to speak of petty ones) than upper layers of the wage-earning class succeed in breaking through the barrier to becoming small independent entrepreneurs in service industry, retail trade or agriculture.

The sum-total of the entire social evolution is a constant increase in that part of the population which is composed of wage-earners; a constant decline in that part which is composed of independent businessmen.¹¹² Not one of Marx’s predictions has been more thoroughly confirmed by empirical evidence (repeated claims to the contrary notwithstanding¹¹³) than that which identified a long-term trend to class polarization under capitalism. Marx was able to make that sweeping historical forecast, so strongly denied by almost all his contemporaries, because, basing himself on the laws of motion of capitalism, he understood that the division of ‘net value’ (value added) into wages and surplus-value had to lead, under the pressure of capitalist competition, to more and more wage-earners being unable to become capitalists and fewer and fewer capitalists being able to remain capitalists.

Capitalist relations of distribution, rooted in capitalist relations of production but by no means identical with them,114 constantly reproduce these relations of production. But they also reproduce the basic material preconditions of class struggle and class solidarity, both in the sphere of distribution (i.e. on the market) and in the sphere of production (in the factory):

(1) The fact that the individual worker has no economic resources on which he can fall back, that he cannot ‘wait’ till its market price (the offered wage) goes up before selling his labour-power, makes collective organization of such sales by workers – i.e. unionization and collective bargaining – a powerful inbuilt tendency under capitalism, reproducing itself universally wherever wage-labour appears.

(2) The fact that the fluctuations of the reserve army of labour, in the last analysis, regulate the fluctuations of real wages creates a strong inbuilt interest for the mass of wage-earners as such to ensure high levels of employment, in other words to demand elementary economic policies at the level of the economy as a whole which tend to limit unemployment.¹¹⁵

(3) The fact that surplus labour is the very essence of surplus-value and profit (more exactly of RIP: Rents, Interests and Profits) creates an equally strong inbuilt tendency in the working class to challenge speed-ups, reorganizations and forms of control of the labour process which tend to increase the mass of surplus labour and its degrading, de-humanizing effects upon the individual worker as well as upon whole sections of the working class.116

(4) Finally, the fact that capital can and must periodically challenge all the partial conquests of the workers, both in the sphere of distribution (increases in wages and social-security payments; free collective bargaining, trade-union rights and the unrestricted right to strike) and in the sphere of production (reduction of the working week and working day; forms of control over the rhythm of work and the organization of the labour process; union rights inside the work-place in general, etc.), especially through ruthless revolutions in technology,¹¹⁷ at least periodically teaches the most intelligent, energetic and militant parts of the working class that (to paraphrase Marx) it is not enough to fight for higher wages, it is also necessary to fight for the abolition of the wage system.¹¹⁸

Conversely, the fact that, under the capitalist mode of production, ownership of any substantial quantity of money (the starting level differing, of course, from period to period and from country to country) automatically transforms that money into money capital – which not only automatically partakes in the general distribution of total socially produced surplus-value (through acquiring the average rate of interest) but is also thus transformed potentially into additional productive capital (money capital put at the disposal of ‘functioning’ capitalists in the productive sectors) – creates a powerful class solidarity among all owners of capital in the common exploitation of all wage-earners as a class; in other words, creates the material basis of bourgeois class solidarity and class consciousness.¹¹⁹

In this sense, all capitalists have a common interest in opposing ‘excessive’ wage increases; in supporting all measures which increase the mass of profits; in supporting speed-up practices and ‘rationalization investments’; and in generalizing these throughout industry and enterprises in general.120 They have a common interest in trying to prevent the rise of militant unionism; or, when this becomes impossible, in trying to limit or curtail trade-union rights, to establish various forms of state control over trade unions, etc. – whatever their differences may be as to the tactics, forms, tempo or extent of such policies.

Likewise, the very nature of private ownership of capital and capitalist competition, through the mediation of each capitalist firm searching to maximize its own profit (i.e. striving for surplus profits over and above the average rate of profit), creates the mechanisms through which the general laws of motion of the system impose themselves. By this very fact, through elimination of the weakest capitalist firms, it ensures a temporary successful reversal of the tendency of the rate of profit to decline. Each capitalist working for his own individual interest thus, in so doing, ensures the long-term reproduction, consolidation and expansion of the capitalist system as a whole.

In the same way, the attempts of capitalists to increase the amount of surplus labour extracted from their own labour-force – by constantly striving to increase the productivity of labour, to organize mass production of an increasing number of commodities, and thereby to lower the value (expressed in gold prices) of all commodities – tend to create a collective interest of the bourgeois class in not limiting mass consumption (except in the initial stages of capitalist industrialization). This helps to counteract the difficulties of realizing the value (surplus-value) embodied in the constantly rising mountain of finished goods which inevitably accompanies enlarged reproduction and the accumulation of capital, in spite of the accompanying tendency towards increasing exploitation of productive wage-labour (towards a historically rising rate of surplus-value). This creates a basic class interest of the bourgeoisie in ‘normal’ rather than ‘abnormal’ conditions of exploitation, including whenever possible rising real wages and elementary social legislation, in order to defuse the explosive character of the class struggle. Direct repression designed to discipline the working class is used only under exceptional circumstances, in grave structural crises (whether economic, political or a combination of both).

Again, the two processes just outlined, whereby a self-conscious working class and a self-conscious bourgeois class are constituted as a direct product of the inner mechanisms of the capitalist mode of production, are not symmetrical. In spite of all the inherent segmentations of the working class – all the constantly recurring phenomena of division along craft, national, sex, generational, etc. lines – there are no inbuilt structural obstacles to the overall class solidarity of workers under capitalism. There are only different levels of consciousness, which make the conquest of that overall class solidarity more or less difficult, more or less uneven in time and space.

The same is not true of bourgeois class solidarity. In periods of prosperity, when their struggles are essentially for larger or smaller shares of an increasing mass of profits, class solidarity easily asserts itself among capitalists. In periods of crisis, however, competition has to take a much more savage form, since for each individual capitalist it is no longer a question of getting more or less profit, but one of his survival as a capitalist.121 So there are instances of acute crisis of the system in which no economic or political solidarity can assert itself among the capitalist class; in which, even in the face of the gravest collective danger for the system as a whole, sectional or individual interests will prevail over collective, class ones.¹²²

Of course, what I have just said applies to inter-capitalist competition, not to the class struggle between Capital and Labour as such, in which, by contrast, the graver the socio-political crisis, the more sharply ruling-class solidarity will assert itself. But the fundamental asymmetry of economic class solidarity within, respectively, the capital-owning and the wage-earning class has to be stressed. It is, in the last analysis, structurally connected with the basically different relations of capitalists and wage-earners towards private property and competition. Private property and competition are built into the very nature of the capitalist class. Competition among wage-earners, however, is imposed upon them from outside, not structurally inherent in the very nature of the class. On the contrary, wage-earners normally and instinctively strive towards collective cooperation and solidarity.123 Hence, to whatever extent competition among themselves is periodically reproduced, especially in times of economic crisis or after major social or political defeats, it can always be overcome by subsequent efforts to organize and to raise class consciousness assisted by the very advances of capital accumulation itself.

In Part Seven of Volume 3, Marx pays great attention to the mystifying appearance of revenues ‘produced’ by different ‘factors of production’: land, labour and capital. In our day, this mystification has been extended through the quest for growth rates or income accretions ‘produced’ by scientific progress or even by higher education.¹²⁴ In and of itself, ‘science’ produces neither value nor income. The results of scientific research, incorporated into new forms of machinery and new forms of labour organization, increase productivity of labour and thus undoubtedly contribute to the increase of material wealth. But this is something quite different from the production of value or income. What these formulas mystify is the fact that, under capitalism, private ownership of the means of production and the transformation of manual and intellectual labour – including scientifically creative labour – enable the capitalist (the capitalist firm) to incorporate into the total value produced in the course of the commodity-producing process the results of the cooperation, inventiveness and skill of all manpower employed. And this occurs essentially in the form of surplus-value, since the results in question do not directly change the reproduction costs of labour-power, which alone represent necessary labour (that part of value added which does not take the form of surplus-value). Qualities of labour thus appear as qualities separate and apart from labour: as either qualities of ‘capital’ (which is represented as a mass of things, instruments, machinery and other means of production) or qualities of ‘science’ (which is again separated from labour as some pure product of the brain).

For Marx, scientific labour is the very essence of ‘general labour’, i.e. creative labour developing new discoveries and inventions. But like collective (socialized) labour, it is indissociably related to the process of cooperation, of many manual and intellectual workers working together: ‘These savings in the use of fixed capital, as we already said, are the result of the way the conditions of labour have been applied on a large scale. In short, the way in which they serve as conditions of directly social, socialized labour, of direct cooperation within the production process. This is firstly the only condition on which mechanical and chemical discoveries can be applied without increasing the price of commodities, and this is always the sine qua non. Next, it is only with production on a large scale that we can have the economy that arises from productive consumption in common. Finally, however, it is only the experience of the combined worker that discovers and demonstrates how inventions already made can most simply be developed, how to overcome the practical frictions that arise in putting the theory into practice – its application to the production process, and so on. We must distinguish here, incidentally, between universal labour and communal labour… Universal labour is all scientific work, all discovery and invention. It is brought about partly by the cooperation of men now living, but partly also by building on earlier work. Communal labour, however, simply involves the direct cooperation of individuals.’125

THE DESTINY OF CAPITALISM

Does Capital contain a theory of the final and inevitable downfall of the capitalist mode of production? Is the answer to this query to be found in Volume 3, and specifically in Marx’s determination of the tendency for the average rate of profit to decline? Do the laws of motion of the capitalist mode of production imply that the system cannot forever survive its inner contradictions? These questions have been asked ever since Capital first appeared, by people supporting Marx’s theories as well as by his opponents. The so-called ‘collapse controversy’ has played a crucial role both in the history of Marxist theory after Marx and in the history of the international labour movement influenced by Marx’s (or Marxist) ideas.

The initial position defended by ‘orthodox’ Marxists inside the Second International was cautious but nevertheless clear: the system would in the end collapse through a general sharpening of all its internal contradictions. Engels, by and large, supported this view.126 It could undoubtedly base itself upon a number of passages from Capital (though, it is true, from Volume 1 rather than Volume 3).¹²⁷ Its main merit was to integrate the class struggle, the growth of the labour movement and of working-class consciousness, into overall perspectives regarding the final destiny of the capitalist system.

It should be stressed, however, that the question of whether capitalism can survive indefinitely or is doomed to collapse is not to be confused with the notion of its inevitable replacement by a higher form of social organization, i.e. with the inevitability of socialism. It is quite possible to postulate the inevitable collapse of capitalism without postulating the inevitable victory of socialism. Indeed, rather early in the history of revolutionary Marxism, the two were conceptually separated in a radical fashion, the destiny of capitalism being formulated in the form of a dilemma: the system cannot survive, but may give way either to socialism or to barbarism.¹²⁸

While both Marx and Engels – and especially the older Engels, faced with the tremendous and apparently irresistible rise of the modern labour movement – exhibited a robust optimism as to the future of socialism, they were always careful, when the question was posed at its most general, abstract, historical level, to reject any idea of historical inevitable sequences of social organization (modes of production). On a number of occasions, they pointed out that the passage from one mode of production to another depended upon the outcome of concrete class struggles, which might end either with the victory of the more progressive, revolutionary class, or in the mutual destruction of both the old ruling class and its revolutionary adversary and in a protracted decadence of society.

The initial position was challenged by the so-called revisionists around the German Eduard Bernstein, who denied that there was any inherent tendency for the inner contradictions of the capitalist mode of production to sharpen. They postulated, on the contrary, that these contradictions would decrease. They did not, however, conclude from this that capitalism would survive for ever, but rather believed that it would fade away gradually, so that there was no need to overthrow it by revolutionary means.129 Most of the later variants of gradualism and reformism (including, in recent years, Euro-communism) have their common roots in Bernstein’s writings, which are remarkable for the clear and consistent way in which they pose the problem¹³⁰ – the only trouble being that their predictions proved to be wrong.

Far from leading to permanent peace, capitalism has led to two world wars and risks a third one, suicidal for the whole of mankind. Far from its leading to an ever-smoother functioning of the international capitalist economy, we have witnessed the catastrophic crises of 1920–21, 1929–32 and 1938, followed, after the post-Second World War boom, by a new long slump starting in the late sixties or early seventies. And far from ever-increasing freedom and democracy, the twentieth century has seen much greater repression and far bloodier dictatorships than anything Marx, Engels or other nineteenth-century socialists ever witnessed or could have imagined in their day.

It is in this context that followers of Marx attempted to formulate in a more rigorous way the probable destiny of capitalism. Rosa Luxemburg was the first to try to elaborate, on a strictly scientific basis, a theory of inevitable collapse of the capitalist mode of production. In her The Accumulation of Capital, she tried to show that enlarged reproduction, with full realization of surplus-value produced during the process of production properly speaking, was impossible under ‘pure’ capitalism. That mode of production, therefore, had an inherent tendency to expand into a non-capitalist milieu, i.e. to gobble up the large areas of petty commodity production still surviving inside the capitalist metropolis and to expand continuously towards the non-capitalist periphery, i.e. the colonial and semi-colonial countries. This expansion – including its most radical forms: contemporary colonialism and murderous colonial wars; imperialism and imperialist wars – was indispensable for the survival of the system. If and when that non-capitalist milieu disappeared, the system would collapse, since it would be unable fully to realize surplus-value. But Luxemburg made it clear that, long before that final moment, the simple consequences of these increasingly violent forms of expansion, as well as the consequences of the gradual shrinking of the non-capitalist milieu, would sharpen the inner contradictions of the system to the point of explosion, thereby preparing its revolutionary overthrow.131

I have already discussed, in the Introduction to Volume 2 of Capital (as well as in Late Capitalism), the strengths and weaknesses of Luxemburg’s The Accumulation of Capital.¹³² Here, I only wish to deal with a methodological objection which has been raised against Luxemburg’s theory of collapse – and subsequently against a number of other such theories. Critics have alleged that, by basing the perspective of inevitable collapse of the capitalist mode of production exclusively on the system’s laws of motion, its inner economic mechanism, Luxemburg was moving back towards ‘economism’; that this was a regression from the way in which Marx and Engels themselves, and their first disciples, always integrated economic laws and movements with the class struggle, in order to arrive at overall historical projections and perspectives.¹³³

This objection, however, is unjustified. While it is true that the contemporary history of capitalism, indeed the history of any mode of production in any epoch, cannot be satisfactorily explained if the class struggle (and especially its outcome after certain decisive battles) is not treated as a partially autonomous factor, it is likewise true that the whole meaning of Marxism disappears if this partial autonomy is transformed into an absolute one. It is precisely the merit of Luxemburg, as well as of several of her subsequent antagonists in the ‘collapse, controversy’, to have related the ups and downs of the class struggle to the inner laws of motion of the system. If one were to assume that either the infinite adaptability of the capitalist system, or the political astuteness of the bourgeoisie, or the inability of the proletariat to raise its consciousness to sufficient levels (not to speak of the alleged growing ‘integration’ of the working class into bourgeois society), could, in the long run and for an undefined length of time, neutralize or reverse that system’s inner laws of motion and intrinsic contradictions, i.e. prevent them from asserting themselves, then the only scientifically correct conclusion would be that these laws of motion do not correspond to the system’s essence: in other words, that Marx was basically mistaken when he thought he had discovered that essence. (This is something different, of course, from the possibility of temporary ups and downs in the sharpening of contradictions, which are not only possible but even inevitable, as Marx himself pointed out in his treatment of the tendency for the average rate of profit to decline.)

A second attempt to produce a scientifically rigorous ‘collapse theory’ (though in the event it was less rigorous, it should be said, than Luxemburg’s) was made during and immediately after the First World War by certain leading radical Marxist economists who greatly influenced Lenin when he was drafting his Imperialism, the Highest Stage of Capitalism. The most prominent of these were the Russian Nikolai Bukharin and the Hungarian Eugen Varga.134 While avoiding any ‘mono-causal’ reduction of the problem to a single decisive factor, these authors formulated the hypothesis that capitalism had entered an irreversible period of historical decline, resulting from a combined manifestation of all its sharpened contradictions: reduction of markets; decline of. world trade; decline of the international division of labour; decline of money economy, and even a partial reversion to barter and pre-capitalist forms of production in capitalist countries; decline of material production; collapse of the credit system; absolute decline in the standard of living of the workers; recurrent wars and civil wars; recurrent revolutionary explosions and victorious socialist revolutions.

While this analysis may offer a relatively convincing description and explanation of what actually occurred in 1914 (or even 1912)—1921 and again in 1930–40 (or even in certain parts of the world in 1945–8), it gets into serious trouble once confronted with post-Second World War developments in the international capitalist economy. Tending to theoretical eclecticism, it lacks the deeper rigour needed to tie all these various developments to the basic laws of motion of the system. In particular, it avoids any discussion of the reasons why the countervailing factors, enumerated by Marx as able temporarily to neutralize the tendency for the average rate of profit to fall, would definitely cease to be effective in the epoch of capitalist decline; why the huge devalorization and destruction of capital which occurred in the 1929–32 crisis and the Second World War, coupled with a huge upsurge in the rate of surplus-value (as a result both of catastrophic working-class defeats and of a powerful increase in the productivity of labour in department II, as a result of a new technological revolution), could not lead to a new upsurge in the productive forces – inevitably ending in a new reassertion of sharpened contradictions of the system.135

One offshoot of the Bukharin-Varga theory of the irreversible decline of the capitalist system since 1914 is the concept of ‘general crisis of capitalism’, in which the emphasis has become progressively shifted from the inner laws of motion of the system towards the outside challenges it is increasingly meeting as the result of a chain of victorious socialist revolutions, which have led to a shrinking of the geographical area in which it can operate. In its initial form, the concept of a general crisis of capitalism – which originated from the victory of the October Revolution in Russia – still established an interrelation between that outside challenge and the ensuing sharpening of the system’s inner contradictions.136 But this has become less and less the case in later variants, especially the ‘state monopoly capitalism’ theory fully developed after the Second World War.

Here the ‘basic’ contradiction is clearly defined as that between the ‘socialist camp’ and the ‘capitalist camp’, and no longer as the increasingly explosive inner contradictions of the capitalist system itself. The paradox is even pushed to the point where Soviet authors seriously assert that, as a result of the ‘competition between the two systems’, capitalism is ‘condemned’ to continuous growth!¹³⁷ In this way, the theory of collapse is ‘dialectically’ turned into its very opposite: the possibility for capitalism to survive for ever. The system’s capacity to eliminate for an indefinite period the most serious effects of its inner contradictions is postulated – until such time as the economic, social and cultural superiority of the socialist camp finally asserts itself. It is hardly necessary to point out that this intellectual contortion is structurally related to the specific interests of the Soviet bureaucracy – both its attempts to maintain conditions of peaceful coexistence with international capitalism, and its concern to maintain the subordination of a large section of the international labour movement to its own diplomatic manoeuvres – and, as such, represents a typical phenomenon of ideological mystification.

A third – once again, more rigorous – attempt to theorize the inevitability of capitalism’s collapse was offered in the late twenties by the Polish Marxist Henryk Grossmann. This was essentially a generalization – one could even say an extreme extrapolation – of Marx’s law for the tendency of the average rate of profit to decline. Grossmann tried to prove that, in the long run, countervailing forces cannot prevent the law from asserting itself with increasing strength – up to the point where all accumulated capital tends to be unable to become valorized, i.e. to the point where the total mass of surplus-value cannot ensure sufficient accumulation, even if the subsistence of the capitalist class itself falls to zero.138 There are many weaknesses in this theory, which have been pointed out by a number of critics.¹³⁹ The main one is that Grossmann does not really prove that all the countervailing forces gradually lose their capacity to neutralize the declining rate of profit. He especially underestimates the effects of massive devalorization (and destruction) of capital, which has historically proven to be much larger in scope than he visualizes (his book was finished before the 1929–32 crisis unfolded to its full depth – and, of course, before the frightful destruction of the Second World War).

Therefore, Grossmann’s somewhat arbitrary numerical starting-point – the reproduction schemas which Otto Bauer worked out in his reply to Luxemburg’s The Accumulation of Capital¹⁴⁰ – leads to results which ignore the effects of devalorization cycles of capital. Such a hypothesis is untenable in the light of the real history of capitalism (which is a crisis-ridden history that has witnessed twenty-one crises of overproduction since the establishment of the world market for industrial goods). Marx explicitly points out this devalorization-of-capital function of capitalist crises in Chapter 15 of Volume 3 of Capital. Hence, one can only consider Grossmann’s successive figures as representing not annual totals but averages for seven/ten-year cycles. Thus the final collapse of the system is postponed till the twenty-second century (after thirty-seven seven/ten-year cycles). If the initial proportions between department I and department II were more realistic – and they should have been, in the light of the real history of the capitalist mode of production which, in the 1920s, had nowhere even approached a situation in which two-thirds of current production occurred in department I – the postponement of the ‘collapse’ would be even more pronounced: it would occur only after fifty or sixty cycles, i.e. after 400 or 500 years. Inadvertently, Grossmann, obsessed by his mono-causal explanation for the inevitability of collapse, was led to demonstrate precisely the opposite of what he intended: the extreme longevity rather than the final collapse of the system, as a function of its inner laws of motion.

One might be tempted to treat the Baran/Sweezy theory of the growing difficulty of ‘surplus realization’ by monopoly capitalism as either a variant of Luxemburg’s collapse theory or a fourth distinct collapse theory of its own.141 This, however, is not the case, since Baran and Sweezy, while underlining the growing difficulties for ‘surplus realization’, at the same time stress the system’s capacity to integrate the working class socially and thereby ensure its perpetuity – albeit under conditions of permanent quasi-stagnation – rather than its inevitable collapse. Like the more extreme proponents of the ‘state monopoly capitalism’ theory, these authors have to project the system’s real enemies outside the system itself: third-world peasants; marginalized super-exploited layers; and so on. But they are nowhere able to demonstrate that these social forces anywhere have a potential social and economic strength comparable to that of the modern proletariat. Since such forces are not vital to the system’s basic productive relations, they can be variously ignored, or integrated, or crushed, without making the system incapable of functioning.¹⁴² So this is not really a ‘collapse of capitalism’ theory at all.

As in the case of the mono-causal theories of crisis, there are obviously correct elements in each of the three versions of collapse theory outlined above. These have to be tied together in order to furnish a coherent theory of the inevitable collapse of capitalism, consistent with all the inner laws of motion and contradictions of that mode of production, as unfolded by Marx’s analysis in Capital.

One element in Grossmann’s analysis is important, if not decisive, as the starting-point for such a synthesis: this is the point in time when, in addition to the tendency of the rate of surplus-value to decline, the mass of surplus-value ceases to grow and begins to decline – first gradually, then permanently. This would obviously be the most serious blow to a continuous process of capitalist accumulation. Grossmann, however, fails to point out the concrete content of such an incipient decline in surplus-value production, which I have tried to specify in Late Capitalism: a level of mechanization, of semi-automation – let us say, of spreading full automation – of a growing number of branches of output, in which the total input of productive labour-hours starts to decline, hence in which total value-production declines.

This does not automatically imply an immediate decline in the absolute mass of surplus-value, since the big increase in productivity of labour inherent in ‘robotism’ can reduce necessary labour-time proportionally to the reduction of absolute value production. In the long run, however, this is impossible without more and more severe reductions even in real wages. After a certain point, moreover, it becomes physically impossible. So the extension of automation beyond a given ceiling leads, inevitably, first to a reduction in the total volume of value produced, then to a reduction in the total volume of surplus-value produced. This in turn unleashes a fourfold combined ‘collapse crisis’: a huge crisis of decline in the rate of profit; a huge crisis of realization (the increase in the productivity of labour implied by robotism expands the mass of use-values produced in an even higher ratio than it reduces real wages, and a growing proportion of these use-values becomes unsaleable); a huge social crisis;143 and a huge crisis of ‘reconversion’ (in other words, of capitalism’s capacity to adapt) through devalorization – the specific forms of capital destruction threatening not only the survival of human civilization but even the physical survival of mankind or of life on our planet.¹⁴⁴

A way out is obviously possible, via the massive transformation of ‘services’ into commodity-producing branches (which add to total value production). Indeed, it is already starting in such key services as health, education, banking and public administration. This indicates how wrong it is to speak of late capitalism as a post-industrial society.145 On the contrary, we are only now entering the age of full industrialization of a whole series of branches which have escaped that process up to now. But this only postpones the time of reckoning. For the industrialization of service sectors reproduces there, after a certain transition period, the very same processes of massive mechanization, semi-automation and full automation for which micro-processors have already, provided the necessary technical tools (the same applies, incidentally, to the process of industrialization of underdeveloped countries as a way out of the structural crisis). So it is impossible to see how capitalism can escape its final fate: economic collapse.

In addition, with the development of semi-automation and automation, a new significant reversal occurs of the revolution constantly produced by capitalism in labour organization and the actual labour process. A massive reintroduction of intellectual labour into the process of production is inevitable, alongside an at least relative decline in the extreme parcellization of labour characteristic of Taylorism. The more wage-labour is employed for supervising functions and the maintenance of delicate and costly equipment, the more its own skill, level of culture and degree of involvement in the production process becomes an indispensable element of reproduction of capital. Hence, not only are the cooperative qualities of objectively socialized labour inside the factory developed to a higher degree. The consciousness of the workers that they are able to run factories instead of capitalists or capitalist managers takes a giant leap forward. Thus the growing crisis of capitalist relations of production (both objectively and subjectively, i.e. in terms of their legitimacy in the eyes of the working class and of larger and larger sectors of the population as a whole), and the challenge which workers’ struggles pose for these, become an integral part of the system’s tendency towards collapse.

But it is evident that such a trend towards upgrading labour in productive sectors with the highest technological development must, of necessity, be accompanied by its very negation: a rise in mass unemployment, in the extent of marginalized sectors of the population, in the number of those who ‘drop out’ and of all those whom the ‘final’ development of capitalist technology expels from the process of production. This means only that the growing challenges to capitalist relations of production inside the factory are accompanied by growing challenges to all basic bourgeois relations and values in society as a whole, and these too constitute an important and periodically explosive element of the tendency of capitalism to final collapse.

As I said earlier, not necessarily of collapse in favour of a higher form of social organization or civilization. Precisely as a function of capitalism’s very degeneration, phenomena of cultural decay, of retrogression in the fields of ideology and respect for human rights, multiply alongside the uninterrupted succession of multiform crises with which that degeneration will face us (has already faced us). Barbarism, as one possible result of the collapse of the system, is a much more concrete and precise perspective today than it was in the twenties and thirties. Even the horrors of Auschwitz and Hiroshima will appear mild compared to the horrors with which a continuous decay of the system will confront mankind. Under these circumstances, the struggle for a socialist outcome takes on the significance of a struggle for the very survival of human civilization and the human race. The proletariat, as Marx has shown, unites all the objective prerequisites for successfully conducting that struggle; today, that remains truer than ever. And it has at least the potential for acquiring the subjective prerequisites too, for a victory of world socialism. Whether that potential will actually be realized will depend, in the last analysis, upon the conscious efforts of organized revolutionary Marxists, integrating themselves with the spontaneous periodic striving of the proletariat to reorganize society along socialist lines, and leading it to precise goals: the conquest of state power and radical social revolution. I see no more reason to be pessimistic today as to the outcome of that endeavour than Marx was at the time he wrote Capital.

ERNEST MANDEL

NOTE