Part One of this volume showed that the profit rate always states the rate of surplus-value as lower than it really is. And we have now seen something further: even when the rate of surplus-value is rising, it has a tendency to show up as a falling rate of profit. The two rates would be equal only if c = 0 — that is, if the whole capital were laid out on labour-power. And a falling rate of profit points to a falling rate of surplus-value under just one condition: the ratio between constant capital's value and the quantity of labour-power that sets this capital in motion has stayed unchanged — or else that quantity of labour-power has grown relative to constant capital's value.

Ricardo claims to be looking at the rate of profit. In fact he looks only at the rate of surplus-value, and even that only assuming a working day of fixed length — fixed both in its length and in how hard the work is.

A falling rate of profit and accelerated accumulation are two expressions of one and the same process — only in this respect, that both express the development of productivity. Accumulation in turn makes the profit rate fall faster, in so far as it goes together with work being concentrated on a large scale, and so with a higher composition of capital. And the falling profit rate in turn speeds up concentration and centralization: smaller capitalists are dispossessed, and the last of the direct producers — wherever there is still something to be taken from them — are expropriated. So accumulation grows faster in sheer size, although the rate of accumulation falls along with the rate of profit.

Then there is the other side. In so far as the rate at which the total capital valorizes itself — the profit rate — is what spurs capitalist production on, just as the valorization of capital is its only aim, a fall in that rate slows down how new, independent capitals are formed, and it appears as a threat to how the capitalist production process develops. It encourages overproduction, speculation and crises; it sets excess capital alongside a surplus population.

So economists who, like Ricardo, treat the capitalist way of producing as the absolute one, sense here that this way of producing throws up a barrier of its own making — and they therefore push that barrier off onto nature (in the theory of rent) instead of onto production. What matters in their horror at the falling profit rate is that they sense a barrier to the growth of the productive forces, one that the capitalist mode of production runs into, and one that has nothing to do with producing wealth as such. And this peculiar barrier is evidence of how narrow the capitalist mode of production is, and how purely historical and passing: it shows that this is not an absolute way of producing wealth, but that at a certain stage it comes into conflict with wealth's further growth.

Ricardo and his school look at industrial profit only, and they count interest inside it. But ground-rent, too, has a falling tendency in its rate, even though its total mass grows, and it may even grow as a proportion of industrial profit. (See Edward West, who had already worked out the law of ground-rent before Ricardo.)

Take the total social capital C. Write p1 for the industrial profit left after interest and ground-rent are deducted, i for interest and r for ground-rent. Then s/C = p/C = (p1 + i + r)/C = p1/C + i/C + r/C. We saw earlier that s — the total surplus-value — keeps growing as capitalist production develops, and yet s/C keeps shrinking, because C grows faster still than s. So there is no contradiction whatever in p1, i and r each being able to keep growing, while both s/C = p/C and each of p1/C, i/C and r/C get ever smaller — or in p1 growing relative to i, or r relative to p1, or r relative to both p1 and i at once. When total surplus-value or profit, s = p, rises while the profit rate s/C = p/C falls at the same time, the proportions among the parts p1, i and r into which s = p divides can shift in any way at all within the limits that the total s sets — and neither s nor s/C changes size as a result.

That p1, i and r vary against one another is just surplus-value being distributed differently under different headings. So p1/C, i/C or r/C — the rate of an individual industrial profit, the rate of interest, the ratio of rent to the total capital — can each rise against the others even while s/C, the general rate of profit, falls. The one condition is that the three of them together add up to s/C.

Say the profit rate drops from 50 per cent to 25 per cent because the composition of capital, with a rate of surplus-value of 100 per cent, changes from 50c + 50v to 75c + 25v. Then with a capital of 1,000 the profit is 500, and with a capital of 4,000 the profit is 1,000. The mass, s or p, has doubled, while p' has gone down by half. And if out of that 50 per cent there was 20 industrial profit, 10 interest and 20 rent, then p1/C was 20 per cent, i/C 10 per cent and r/C 20 per cent. If the proportions stayed the same when the rate changed to 25 per cent, then p1/C = 10 per cent, i/C = 5 per cent and r/C = 10 per cent. If instead p1/C now fell to 8 per cent and i/C to 4 per cent, then r/C would rise to 13 per cent. r's proportional size would have risen against p1 and i, and p' would still have stayed the same. On either assumption the sum of p1, i and r would have grown, since it now comes out of a capital four times as large.

Besides, it is wrong both in history and in concept to assume, as Ricardo does, that industrial profit (plus interest) originally pocketed the whole surplus-value. It is only as capitalist production advances that (1) industrial and commercial capitalists get the whole profit into their hands first, for later distribution, and (2) rent is cut down to the surplus over and above profit. On that capitalist basis rent then grows again — a portion of profit (that is, of surplus-value looked at as the product of the total capital), but not the particular portion of the product that the capitalist pockets.

Take the means of production as given — that is, enough capital already accumulated — and then the creation of surplus-value runs up against no barrier except the working population, so long as the rate of surplus-value, the degree to which labour is exploited, is also given; and against no barrier except the degree of exploitation, so long as the working population is given. And the capitalist production process consists essentially in producing surplus-value, which appears in the surplus product — the proportional share of the goods produced in which unpaid labour is objectified, made solid in them.

We must never forget that producing this surplus-value — and turning part of it back into capital, which is accumulation, is an integral part of that production — is the immediate aim of capitalist production and the motive that governs it. So capitalist production must never be depicted as what it is not: as production whose immediate purpose is enjoyment, or the making of things for the capitalist to enjoy. That would leave out its specific character altogether, the character it shows in its whole inner core.

Squeezing out this surplus-value is the immediate process of production, and that process has no barriers except the ones just named. Once the amount of surplus labour that can be extracted has been objectified in commodities, the surplus-value has been produced. But that is only the first act of the capitalist production process — the immediate production process is finished. Capital has sucked in so much unpaid labour. And as this process develops — its development showing up as the fall in the profit rate — the mass of surplus-value produced in this way swells to monstrous size.

Now comes the second act. The whole mass of commodities, the whole product — both the part that replaces constant and variable capital and the part that represents surplus-value — has to be sold. If that does not happen, or happens only in part, or only at prices below the prices of production, then the worker really has been exploited, but the capitalist gets nothing out of that exploitation as such: it may bring no realization at all of the surplus-value squeezed out, or only a partial one, and it may even mean a partial or complete loss of his capital.

What is needed for direct exploitation in production, and what is needed for it to be realized, are not identical. They come apart not only in time and place, but in concept too. The first is limited only by society's productive forces; the second by how proportionate the different branches of production are, and by society's power to consume. That power to consume is fixed neither by absolute power of production nor by absolute consuming power, but by consumption power resting on antagonistic relations of distribution, which hold the consumption of the great mass of society down to a minimum that can move only within more or less narrow limits. It is further limited by the drive to accumulate, the drive to make capital bigger and to produce surplus-value on a wider scale still. This is the law of capitalist production, given by the constant revolutions in the methods of production themselves, by the devaluation of existing capital that always goes along with them, by the general competitive struggle, and by the need to improve production and enlarge its scale — a need that is only a means of holding on, on pain of going under. So the market has to be extended again and again, until its connections and the conditions that rule them take on more and more the shape of a natural law standing independent of the producers, and become ever harder to control. The contradiction inside seeks to even itself out by pushing the field of production outward. But the more productive power develops, the more it collides with the narrow base that the relations of consumption rest on. On this contradictory basis it is no contradiction at all that too much capital sits side by side with a surplus population that keeps growing: for even though bringing the two together would raise the mass of surplus-value produced, it would sharpen to the same degree the contradiction between how that surplus-value is produced and how it is realized.

Once a particular rate of profit is given, how big the mass of profit is always depends on how big the capital advanced is. What actually gets accumulated, though, is decided by the part of that mass which is turned back into capital. And since that part equals profit minus the revenue the capitalists consume, it depends not only on how valuable that mass is, but also on how cheap the commodities are that the capitalist can buy with it — commodities that go partly into his own consumption, his revenue, and partly into his constant capital. (Wages are assumed to be given here.)

The amount of capital a worker sets in motion — and whose value his labour preserves and makes reappear in the product — is a completely different thing from the value he adds. Capital mass = 1,000, labour added = 100: capital reproduced = 1,100. Capital mass = 100, labour added = 20: capital reproduced = 120. So the first case shows a profit rate of 10 per cent, the second of 20 per cent. And yet a larger accumulation can come out of 100 than out of 20. And so the stream of capital rolls on — setting aside any devaluation of it through a rise in productive power — accumulating in proportion to the weight it already carries, not in proportion to how high the profit rate is. A high profit rate, in so far as it rests on a high rate of surplus-value, is possible when the working day is very long even though the labour is unproductive; it is possible because the workers' needs are very small, so that the average wage is very low, even though the labour is unproductive. A low wage will be matched by a lack of energy in the workers. So capital piles up only slowly here, however high the profit rate. The population stands still, and the product costs a lot of labour-time, although the wage paid to the worker is small.

The rate of profit falls, not because the worker is exploited less, but because less labour in total is applied relative to the capital employed.

When the profit rate falls while the mass of profit rises, as we have seen, then the capitalist takes a larger share of the year's product of labour under the heading of capital (replacing the capital used up), and a relatively smaller share counted as profit. Hence the fantasy of the Reverend Chalmers: that the less of the year's product the capitalists lay out as capital, the more profit they gulp down — with the Established Church helping them out here, so that much of the surplus product gets consumed instead of being turned back into capital. The parson has cause and effect the wrong way round. And in any case the mass of profit grows as more capital is laid out, even at a lower rate.

But this also brings concentration of capital with it, because production now demands capital on a massive scale. And it brings centralization too — small capitalists swallowed up by big ones, and stripped of their capital. This is the producers being cut off from the conditions of labour, raised to a second power — the same separation happening over again, one level up; for these smaller capitalists count as producers too, since their own labour still matters. The work a capitalist does is, in general, inversely proportional to how big his capital is — that is, to how far he is a capitalist. This divorce — labour's conditions standing over here, the producers standing over there — is what makes up the very concept of capital. It opens with primitive accumulation (Volume 1, Part Eight), then appears as a constant process in the accumulation and concentration of capital, and here at last shows itself as the centralization of capitals that already exist into a few hands, and the decapitalization of many — which is what expropriation has now turned into. This process would soon bring capitalist production to collapse, were it not that counteracting tendencies keep working in the opposite direction, constantly decentralizing, alongside the centripetal force.

The growth of the social productivity of labour shows itself twice over. First, in the size of the productive forces that have already been built up: in the scale of the conditions of production — measured both in value and in sheer bulk — under which new production goes ahead, and in how much productive capital has already been amassed, in absolute terms. Second, in how small a share of the total capital has to be laid out on wages — that is, how little living labour, relative to the whole, a given capital needs in order to reproduce itself and be valorized, in order to produce on a mass scale. This also presupposes that capital is concentrated.

Now look at the labour-power that gets set to work, and productivity again shows itself twice over. First, in the growth of surplus labour — in other words, in a shortening of the necessary labour-time, the time needed to reproduce labour-power. Second, in a decline in the total amount of labour-power, the number of workers, that has to be applied to set a given capital in motion.

These two movements do not merely go along together. Each one conditions the other, and both are ways in which one and the same law expresses itself. And yet they push the profit rate in opposite directions. Profit in its total mass is the same as surplus-value in its total mass; the rate of profit is s/C — surplus-value over the total capital advanced. But the size of that surplus-value total is fixed by two things: first, by its rate; second, by the mass of labour that is applied at that rate at the same time — which is to say, by how big the variable capital is. One factor, the rate of surplus-value, rises; the other factor, the number of workers, falls — relatively or absolutely.

Where the growth of productivity cuts the paid part of the labour applied, it raises surplus-value by lifting its rate. But where it cuts the total amount of labour a given capital applies, it cuts down the number you have to multiply the rate of surplus-value by to arrive at its mass. Two workers on the job 12 hours a day cannot turn out the same mass of surplus-value as 24 workers who each put in only 2 hours — even if those two could live on air and so would not have to work for themselves at all. On this front, then, making up for a smaller number of workers by squeezing the labour harder runs into limits it cannot get past. It can hold the falling profit rate back, but it cannot do away with that fall.

So with the development of the capitalist mode of production, the rate of profit falls, while the mass of profit grows, along with the growing mass of capital that is set to work. Given the rate, how much capital grows in absolute terms depends on how big it already is. But once that size is fixed, how fast it grows — the rate of its growth — depends on the profit rate. A rise in productivity (which, as noted, always goes hand in hand with a devaluation of the capital already on hand) can add to the value magnitude of capital directly in one way only: by raising the rate of profit, so that a bigger slice of the year's product gets turned back into capital. As far as the productivity of labour goes, this can only happen — it has no direct bearing on the value of the capital already there — where it either raises relative surplus-value or lowers the value of the constant capital: where it cheapens either what it takes to reproduce labour-power or the elements of constant capital. But both of these bring a devaluation of the capital already on hand with them, and both go hand in hand with variable capital shrinking relative to constant. Both of them condition the falling profit rate — and both of them slow it down. And further: in so far as a higher profit rate brings a greater demand for labour, it works towards an increase in the working population, and so in the exploitable material — which is what first turns capital into capital.

Indirectly, though, the growth of labour productivity helps to swell the capital value already on hand. It multiplies the mass and the variety of use-values in which the same exchange-value is represented — and those use-values are the material base, the physical elements of capital, the actual stuff that constant capital consists of directly and variable capital at least indirectly. The same capital, with the same labour, throws off more things that can be converted into capital, whatever their exchange-value. Things that can soak up extra labour, and so extra surplus labour too, and thereby form extra capital.

How much labour capital can command is not a matter of its value. What matters is the sheer bulk of raw and auxiliary materials, of machinery and other elements of fixed capital, of means of subsistence, out of which capital is put together — whatever all of that may be worth. As the mass of labour applied grows in this way, and the mass of surplus labour with it, the value of the reproduced capital grows too, and so does the new surplus-value added on top of it.

But these two aspects of the accumulation process are not to be seen just as lying quietly side by side, the way Ricardo treats them. They contain a contradiction, and that contradiction makes itself known in opposing tendencies and appearances. The contending agencies act at one and the same time, each against the other.

At the same time as the impulses pushing for a real growth of the working population — impulses that come from the growth of the part of society's total product that functions as capital — other agencies are at work that create a surplus population which is only relative.

At the same time as the profit rate falls, capital grows in mass — and right along with that goes a devaluation of the capital already on hand, which holds that fall back and gives the accumulation of capital value an accelerating push.

At the same time as productivity develops, the composition of capital grows higher: the variable part declines relative to the constant part.

These different influences make themselves felt now more side by side, spread out in space; now more one after the other, in time. And periodically the conflict between the contending agencies finds an outlet in crises. Crises are never more than momentary, violent solutions to the contradictions that exist — violent outbursts that put the disturbed balance back in place, for the moment.

Put in the most general terms, the contradiction comes to this. The capitalist mode of production includes a tendency to develop the productive forces absolutely — without regard to value, or to the surplus-value locked up in value, and without regard, either, to the social relations in which capitalist production goes on. At the same time, what it aims at is preserving the capital value that exists and valorizing it to the greatest possible extent — that is, making that value grow at an ever-accelerating pace. Its specific character is that it treats the existing capital value as the means for valorizing that value as much as possible. And the methods by which it does this include: a declining profit rate, the devaluation of capital already on hand, and the development of labour's productive forces at the price of the productive forces already produced.

The existing capital gets devalued periodically. This is a means built into the capitalist mode of production: it holds up the fall in the profit rate and, by forming new capital, speeds up how fast capital value accumulates. But it throws out of order the given conditions under which capital circulates and reproduces itself, and so it is attended by crises and by sudden stoppages in production.

The relative decline of variable capital against constant capital — which keeps pace with the growth of the productive forces — goads the working population into growing, while at the same time it constantly creates an artificial surplus population. Looked at in terms of value, the falling profit rate puts the brakes on the accumulation of capital — and this serves to speed up the accumulation of use-value still further. That in turn sets value accumulation going at a faster pace again.

Capitalist production never stops trying to get past the barriers built into it. But the only way it gets past them is by means that put those barriers back in its path, fresh and on a mightier scale.

Capital itself is the true barrier to capitalist production. What that means is this: capital and its self-valorization show up as the starting point and the finishing point, as the motive and the purpose of production; production is carried on only for capital, and not the other way round, with the means of production acting as mere means for the producers' society to give its life process — its whole pattern of life — an ever-widening shape.

The barriers within which the preservation and valorization of capital value can only move — and that rests on dispossessing and impoverishing the great mass of the producers — these barriers therefore keep colliding with the production methods that capital has to use for its purpose, methods that steer towards production without limit, towards production as its own end, towards an unchecked growth of the productive powers of social labour.

The means — an unrestricted development of society's productive forces — comes into constant conflict with the narrow purpose it is meant to serve: the valorization of the capital it already has. So if the capitalist mode of production is a historical means of developing the material productive power of society and of creating the world market that belongs to it, it is at the same time the standing contradiction between that historical task and the relations of production that go with it.

The lower the rate of profit falls, the more capital a single capitalist must have in hand before he can use labour productively. He needs that minimum both simply to exploit labour at all, and to ensure that the labour-time he applies is the labour-time the commodities actually need — that it does not run past the average that is socially necessary for making them. At the same time concentration grows: past a certain point, a big capital with a low rate of profit piles up faster than a small capital with a high one. And this growing concentration, in turn, once it reaches a certain height, brings on a fresh fall in the rate of profit. That drives the mass of small, scattered capitals onto the road of adventure:

Speculation, credit swindles, share swindles, crises.

The so-called plethora of capital — an over-supply of capital — always refers in the main to a plethora of that capital whose falling rate of profit is not outweighed by its mass — and that is always the fresh offshoots of capital that are just being formed — or else to the plethora which such capitals, unable to act on their own account, hand over, as credit, to those who run the big lines of business. This plethora of capital grows out of the same causes that bring forth a relative surplus population, and is therefore a phenomenon complementary to that surplus population, even though the two stand at opposite poles: unoccupied capital on one side, an unemployed working population on the other.

Overproduction of capital, not overproduction of individual commodities — though overproduction of capital always includes overproduction of commodities — therefore comes down to one thing: overaccumulation of capital. To understand what this overaccumulation is (a closer look at it follows further on), we only have to take it in its absolute form. At what point would the overproduction of capital be absolute? And we mean an overproduction that does not reach over this or that area of production, or a few important ones, but would itself be absolute in scope — one that would take in every field of production at once?

There would be absolute overproduction of capital as soon as additional capital counted as zero for the purpose of capitalist production. But what capitalist production is for is the valorization of capital — that is, the appropriation of surplus labour, the production of surplus-value, of profit. So as soon as capital had grown so far, relative to the working population, that the absolute labour-time this population supplies could not be stretched any further, and its relative surplus labour-time could not be widened either (and that second one would not be possible anyway in a case where demand for labour is so strong, so that there is a tendency for wages to rise); where, then, the grown capital turns out only the same mass of surplus-value as it did before, or even less — there would be absolute overproduction of capital; that is, the grown capital C + ΔC would produce no more profit, or even less, than capital C produced before its increase by ΔC. In both cases the general rate of profit would also drop sharply and suddenly — but this time because the composition of capital had shifted, and not through any development of the productive power of labour, but through a rise in the money value of variable capital (because wages are higher) and the corresponding fall in the ratio of surplus labour to necessary labour.

In reality it would look like this. One part of the capital would lie wholly or partly idle — because before it could be valorized at all, it would first have to push the capital already functioning out of its place — and the other part would be valorized at a lower profit rate, under the pressure of the idle or half-idle capital. It would make no difference that one part of the additional capital stepped into the place of old capital, so that the old capital took up a place inside the additional one. We would still have, on one side, the old sum of capital and, on the other, the additional one.

This time the fall in the rate of profit would go together with an absolute drop in the mass of profit, since on our assumptions the mass of labour-power applied could not be increased and the rate of surplus-value could not be raised, so the mass of surplus-value could not be increased either. And the smaller mass of profit would have to be reckoned on a larger total capital.

But suppose the employed capital went on being valorized at the old rate of profit, so that the mass of profit stayed the same — it would still be reckoned on a grown total capital, and that too contains a fall in the rate of profit. Say an overall capital of 1,000 throws off a profit of 100; if it grows to 1,500 and still throws off only 100, then in that second case a capital of 1,000 now throws off only 66⅔. The old capital's valorization would have shrunk in absolute terms. A capital of 1,000 under the new conditions would bring in no more than a capital of 666⅔ did before.

But it is clear that this devaluation of the old capital in real terms could not happen without a struggle — that ΔC could not go on working as capital without a fight. The rate of profit would not sink because of competition arising from the overproduction of capital. Just the opposite: because the fallen rate of profit and the overproduction of capital spring from the same circumstances, the competitive struggle would now set in. The slice of ΔC that lay in the hands of the old, already functioning capitalists, they would leave more or less idle, so as not to devalue their own original capital and not to squeeze the room it has to produce in; or they would use it, even at a momentary loss, to shift the idling of the additional capital onto the new intruders and onto their rivals generally.

The slice of ΔC that had landed in new hands would try to win a place for itself at the old capital's expense, and would partly manage it — forcing part of the old capital to lie idle, making it vacate its old place and itself take the place of that additional capital which is only half-employed or not employed at all.

A part of the old capital would have to lie idle under all circumstances — idle in its character as capital, in so far as it is supposed to function as capital and be valorized. Which part this idling particularly strikes would be settled by the competitive struggle. While things are going well, competition works — as we saw when the general rate of profit was equalized — as a practical freemasonry among the capitalists, so that they share the common booty jointly, in proportion to how big a stake each has put in. But once the question is no longer the division of profit, but the division of loss, each tries to shrink his own share of it as much as he can and to shove it onto the others. For the class, the loss is unavoidable. But how much of it each individual has to carry, how far he has to take part in it at all, then becomes a question of power and cunning, and competition becomes a fight between enemy brothers. The opposition between each individual capitalist's interest and the interest of the capitalist class then makes itself felt, just as before the identity of those interests asserted itself in practice through competition.

How then would this conflict even out again, and how would the relations that suit the 'healthy' movement of capitalist production be restored? How the conflict is resolved is already implied in how the conflict itself is stated. It involves capital lying idle, and even being partly destroyed — to the value of the whole additional capital ΔC, or at least of a part of it. Although, as already follows from the account of the conflict, the distribution of this loss by no means falls evenly across the individual capitals concerned — it is settled in a competitive struggle in which, depending on particular advantages or positions already won, the loss gets spread very unevenly and in very different shapes, so that some capitals lie idle, others are destroyed, others again suffer only a relative loss or a merely temporary devaluation, and so on.

Under all circumstances, though, the balance would be restored through capital lying idle and even through its destruction, on a greater or smaller scale. This would extend in part to the material substance of capital: some of the means of production — fixed capital and circulating capital — would stop functioning, would no longer act as capital; part of the production operations already begun would be shut down. Although on this side time attacks and spoils every means of production (the land excepted), what would happen here, through the stalling of function, is a far heavier real destruction of means of production. The main effect on this side, though, would be that these means of production no longer did anything as means of production — a shorter or longer destruction of their function as means of production.

The main destruction, and the one with the sharpest character, would take place in capital in so far as it has the property of value — in capital values. The part of capital value that exists only in the form of claims on future shares in surplus-value, in profit — in fact nothing but promissory notes on production, in its various forms — is devalued at once as the revenues it is reckoned on fall. Some of the ready gold and silver lies idle, doing no work as capital. Some commodities now on the market can get through circulation and reproduction only if their prices are contracted enormously, that is, only through a devaluation of the capital they represent. Fixed capital's elements lose value more or less in the same way. Add to this that certain assumed price relations are a condition of the reproduction process, so that a general fall in prices throws it into stagnation and confusion. This disruption and stalling cripples money's role as means of payment — a role that comes along with the development of capital and depends on those assumed price relations — snaps the chain of payments due on fixed dates at a hundred points, and is sharpened further by the collapse of the credit system that grew up alongside capital; and so it brings violent, acute crises, abrupt forcible devaluations and a real stalling and disruption of the reproduction process, and with that a real decline in reproduction.

At the same time, though, other forces would have been at work. The stagnation of production would have left a section of the working class idle, and so put those still employed in a position where they had to swallow a cut in wages, even below the average — a move whose effect for capital is exactly the same as if surplus-value, relative or absolute, had been raised at the average wage. The prosperous period would have encouraged marriages among the workers and cut down the number of their children who die — circumstances that, however much they may involve a real growth in population, do not put any more people to work, yet work on the ratio between workers and capital exactly as a larger number of workers actually in action would. Falling prices and the struggle between competitors, on the other side, would have spurred each capitalist to push the individual value of his total product below its general value — to make his product more cheaply than the going value allows — by using new machinery, improved methods of labour, new combinations — that is, to raise the productive power of a given quantity of labour, to lower the ratio of variable capital to constant, and so to set workers free, in short to manufacture an artificial surplus population. And the falling value of the elements of constant capital would itself be one more thing that brings a higher rate of profit with it. More constant capital would be set in motion relative to variable, but the value of that mass could have dropped. The stagnation of production that had set in would have paved the way for production to expand again later — within capitalist limits.

And so the circle would be gone round once more. Part of the capital that had been devalued by the stalling of its function would win back its old value. And beyond that, with wider conditions of production, a wider market and a higher productive power, the same faulty cycle would be run through all over again.

Even under the most extreme assumption one could make, absolute overproduction of capital is not overproduction pure and simple, and not an absolute overproduction of means of production either. Only in one respect are means of production overproduced: in so far as they have to serve as capital, and so, measured against the value that has swollen along with their swollen mass, have to bring in a return on that value — have to produce an additional value.

All the same, it would still be overproduction — because capital would then have become unable to exploit labour at the level of exploitation that the "healthy", "normal" development of the capitalist production process requires: a level at which the mass of profit at least grows along with the growing mass of capital that is applied. Such a level rules out the rate of profit sinking in step with the growth of capital, and still more rules out its sinking faster than capital grows.

Overproduction of capital never means anything else: it means overproduction of means of production — means of labour and means of subsistence — things that are able to work as capital, that is, that can be set to exploiting labour at a given degree of exploitation — a degree that is given because, once the degree of exploitation falls below a certain point, it brings on disruption and stagnation of the capitalist production process, crises, the destruction of capital. There is no contradiction in the fact that this overproduction of capital goes hand in hand with a relative surplus population, larger or smaller.

The same causes that have raised the productivity of labour, swollen the mass of commodity products, widened the markets, quickened the accumulation of capital in both mass and value, and pushed the rate of profit down — these same causes have created a relative surplus population, and keep creating one: workers the surplus capital does not take on, because the level of exploitation at which alone they could be taken on is too low, or at least because the rate of profit they would bring at the given level of exploitation is too low.

When capital is sent abroad, it is not because it could not be employed at home at all. It goes abroad because it can be employed there at a higher rate of profit. But as far as the working population that is employed is concerned, and the country in question as a whole, this capital is absolutely surplus capital. It sits there as surplus alongside the relative surplus population — one example of how the two live side by side and each conditions the other.

On the other side, the fall in the rate of profit that comes with accumulation necessarily brings forth a competitive struggle. Making up for the falling rate through a growing mass of profit works only for society's total capital, and for the large capitalists who are already fully established. New capital, operating independently, finds no such compensation waiting for it; it has to win it first. And so the fall in the rate of profit is what calls forth the competitive struggle among capitals — not the other way round. That struggle does come with a temporary rise in wages, and from that a further, temporary fall in the rate of profit.

The same thing shows up in the overproduction of commodities and the glutting of markets. Capital's purpose is not to satisfy needs but to produce profit, and it reaches that purpose only by methods that set how much is produced by the scale of production, not the other way round; so there must always be a rift between the narrow limits of consumption on a capitalist footing and a production that keeps straining past this, its own immanent barrier. Besides, capital is made up of commodities, so overproduction of capital brings overproduction of commodities with it. Hence the odd spectacle: economists who deny that too many commodities are produced will still admit that there is too much capital.

If someone says that this is not overproduction in general but disproportion between the different branches of production, that says no more than this: inside capitalist production, the proportionality of the separate branches shows itself as a process constantly coming out of disproportionality — because here the whole web of production presses on the agents of production as a blind law, instead of being a law their combined reason has grasped and thereby mastered, so that their shared process of production is brought under their common control. The same objection also demands that lands where the capitalist way of producing is not developed should consume and produce at a level that suits the lands where it is. If someone says that overproduction is only relative, that is quite right; but then the entire capitalist mode of production is itself only a relative one, and its barriers are not absolute — they are absolute for it, on its own ground. How else could demand be missing for the very goods the mass of the people go short of? And how else would that demand have to be hunted for abroad, in far-off markets, just so that the workers at home can be paid enough, on average, for what they need to live on? It is because only inside this specific, capitalist setup does the surplus product take on a form in which its owner can release it for consumption only once it has been converted back into capital for him. And if someone says, finally, that the capitalists have only to swap and eat their commodities among themselves, then the whole character of capitalist production is forgotten — forgotten that what is at stake is the valorization of capital, not its consumption. In short, every one of these objections to the glaring facts of overproduction — facts that pay no attention to the objections — comes down to saying that capitalism's barriers are not barriers to production as such, and so are not barriers to this specific, capitalist mode of production either. But the contradiction of this capitalist mode of production lies precisely in its tendency towards the absolute development of the productive forces — a development that keeps coming into conflict with the specific conditions of production in which capital moves, and can only move.

It is not too many means of subsistence that are produced for the existing population. Quite the opposite. Too few are produced for the mass of the population to be provided for decently and humanely.

Nor is it too many means of production that are produced, if the aim is to give work to the part of the population that can work. The opposite. First, too big a slice of the population is produced that in fact cannot work — that its circumstances leave dependent on exploiting other people's labour, or on kinds of work that can only count as work inside a wretched way of producing. Second, not enough means of production are produced for the whole population that can work to labour under the most productive conditions, which would cut down their absolute labour-time through the mass and effectiveness of the constant capital set to work during that time.

Periodically, though, there is too much in the way of means of labour and means of subsistence for them to keep serving as means of exploiting the workers at a given rate of profit. Too many commodities are produced for the value they hold, and the surplus-value locked up inside that value, to be realized, and turned back into new capital, as things stand under capitalist production's own arrangements for handing out the product and for consumption — that is, for the whole process to run without explosions that keep coming back.

Too much wealth is not what gets produced. But periodically, wealth is produced in excess — in the capitalist, antagonistic shapes it takes.

The barrier of the capitalist mode of production comes into view:

1. The development of the productive power of labour brings forth, in the falling rate of profit, a law that at a certain point turns most hostilely against that very development, and so has to be overcome again and again through crises.

2. It is the appropriation of unpaid labour — and the ratio of that unpaid labour to objectified labour in general, or, put in capitalist terms, the profit and the ratio of that profit to the capital applied, that is, a certain rate of profit — that settles whether production expands or is held back, instead of the relation of production to social needs, to the needs of socially developed human beings. So barriers for it arise already at a level of expansion which, under the other assumption, would look vastly inadequate. Production comes to a halt — not where the satisfaction of needs bids it halt, but where the producing and the realizing of profit bids it halt.

When the rate of profit falls, on one side capital strains and exerts itself: the individual capitalist presses the individual value of his own commodities below their social average value through better methods and so on, and so, at the given market price, makes a surplus profit. On the other side we get swindling, and a general encouragement of swindling, through frantic attempts at new production methods, new capital investments, new adventures — anything to secure some extra profit that is independent of the general average and rises above it.

The rate of profit — that is, the relative growth of capital — matters above all for all the new offshoots of capital that group themselves independently. And if the forming of capital were to fall exclusively into the hands of a few big, fully developed capitals, for which the mass of profit outweighs the rate, then the fire that keeps production alive would go out altogether. Production would fall asleep.

The rate of profit is what drives capitalist production, and nothing is produced except what can be produced at a profit, and only so far as it can. Hence the English economists' anxiety about the rate of profit falling. That the bare possibility of it troubles Ricardo shows precisely how deeply he understood the conditions of capitalist production. What he is reproached for — that he cared nothing for "human beings" and, in looking at capitalist production, had eyes only for the development of the productive forces, whatever sacrifices of human beings and of capital values that development was bought with — is precisely what is significant in him. The development of the productive forces of social labour is capital's historical task and its justification. And in doing this, capital unconsciously creates the material conditions for a higher form of production.

What troubles Ricardo is that the rate of profit — the goad of capitalist production, and the condition of accumulation as its driver — is endangered by the development of production itself. And the quantitative relation is everything here. In fact something deeper lies underneath it, which he only senses. Here, in purely economic terms — that is, seen from the bourgeois standpoint, within the limits of what capitalist reason can take in, from the standpoint of capitalist production itself — what shows itself is this mode's barrier, its relativity: that it is not an absolute but only a historical mode of production, one that belongs to a particular, limited epoch in the development of the material conditions of production.

Productivity grows very unevenly from one branch of industry to another — not just uneven in degree, but often in opposite directions. The result: the total mass of average profit — and average profit is simply surplus-value — has to sit very far below the level you would expect from the productivity reached in the most advanced branches.

That productivity moves unevenly from branch to branch, and often in opposite directions, is not explained only by competition's anarchy and the oddities of bourgeois production. Labour's productivity also depends on natural conditions, and those often yield less in the same measure as productivity rises — productivity, that is, so far as it depends on social conditions. So the different spheres move in contrary directions: progress here, regression there. Think of the plain effect of the seasons, on which the quantity of most raw materials depends; or the exhaustion of forests, coal and iron mines.

In the case of the circulating part of constant capital — raw material and so on — the mass always grows in proportion to the productivity of labour. That is not true of the fixed part: buildings, machinery, installations for lighting and heating. As machinery gets bulkier it becomes dearer absolutely, but relatively cheaper. Five workers may now turn out ten times the commodities they used to; the money laid out on fixed capital does not grow tenfold for that reason. The value of this part of constant capital does rise with the development of productivity, but nothing like in the same proportion.

We have stressed several times already that the ratio of constant capital to variable capital shows up differently in the falling rate of profit than the same ratio does, as labour's productivity develops, when you look at a single commodity and its price.

A commodity's value is fixed by the total labour-time that goes into it — past labour and living labour together. A rise in the productivity of labour consists precisely in this: living labour's share goes down and past labour's share goes up, but so that the total labour the commodity contains falls — living labour falls by more than past labour rises.

The past labour embodied in a commodity's value — the constant capital part — comes partly from the wear of fixed capital and partly from circulating constant capital that goes into the commodity whole: raw and ancillary materials. The value part coming from raw and ancillary materials must fall as productivity rises, because for those materials a rise in productivity shows itself exactly in their value having fallen. The fixed part of constant capital is different: it is precisely characteristic of rising productivity that this part grows very sharply, and with it the value part transferred to commodities as wear and tear.

So a new production method counts as a real gain in productivity only if it passes on to each commodity a smaller added value share for wear of fixed capital than the value share it saves by using less living labour — in a word, it must lower the value of that commodity. That is self-evident even where, as sometimes happens, an extra value share goes into the commodity for more or dearer raw or ancillary materials on top of the extra wear share. Every one of those added value shares has to be more than cancelled out by the fall in value that comes from using less living labour.

This shrinking of the total labour a commodity takes looks, then, like the essential mark of a rise in the productivity of labour — no matter what social conditions production goes on under. In a society where the producers run their production by a plan they draw up beforehand — indeed even in simple commodity production — the productivity of labour would still be measured by exactly this standard, without qualification. But how does it stand in capitalist production?

Take a branch of capitalist production that turns out the normal item of its commodity under these conditions: wear of fixed capital, ½ shilling per item; raw and ancillary materials, 17½ shillings; wages, 2 shillings; and with a rate of surplus-value of 100 per cent, surplus-value of 2 shillings. Total value = 22 shillings.

For simplicity, suppose capital in this branch has the average composition of social capital — the same balance of constant and variable as society's capital on average — so that this commodity's production price is the same as its value, and the capitalist's profit is the same as the surplus-value he produces. So the commodity's cost price comes to ½ + 17½ + 2 = 20 shillings, the average rate of profit is 2/20 = 10 per cent, and the production price of the item is the same as its value, 22 shillings.

Now suppose someone invents a machine that halves the living labour needed for each item, but triples the share of value coming from wear of fixed capital. Then it works out like this: wear, 1½ shillings; raw and ancillary materials, 17½ shillings, as before; wages, 1 shilling; surplus-value, 1 shilling — 21 shillings in all. The commodity has now lost 1 shilling of value; the new machine has decidedly raised the productivity of labour.

For the capitalist, though, it stands like this. His cost price is now 1½ shillings wear, raw and ancillary materials 17½ shillings, 1 shilling wages — 20 shillings altogether, the same as before. The new machine does not by itself change the profit rate, so he has to take a further 10 per cent on his cost price, which makes 2 shillings; the production price is therefore unchanged at 22 shillings — but it now lies a shilling above the value. For a society whose production runs on capitalist lines, the commodity has not got cheaper and the new machine is no better.

So the capitalist has no interest in bringing the new machine in. And since bringing it in would simply make worthless the machinery he already has, which has not yet worn out, turning it into nothing but old iron, and so would cost him a positive loss, he guards very carefully against this folly — utopian folly, as far as he is concerned.

For capital, then, the law of the increased productivity of labour does not hold unconditionally. For capital, productivity is raised not whenever living labour is saved in general, but only when more is saved on the paid portion of living labour than is added in past labour — as was briefly indicated in Volume 1, Chapter 15.

Here capitalist production falls into a fresh contradiction. Its historical mission is to unfold human labour's productivity without mercy, driving it forward in geometrical progression. It proves untrue to that mission as soon as it blocks the development of productivity, as it does here. This proves yet again that it is growing senile and increasingly outliving itself.

Under competition, the rising minimum of capital that a rise in productivity makes necessary for running an independent industrial business successfully shows up like this. Once the costlier new equipment has become general, smaller capitals are shut out of the business in future. Only at the start of mechanical invention in a sphere of production can smaller capitals still operate independently there.

On the other side, very large undertakings with an extraordinarily high ratio of constant capital — railways, say — do not bring in the average rate of profit; they bring in only part of it, an interest. Otherwise the general rate of profit would sink still lower. Here, though, a great accumulation of capital in the form of shares does find a direct field of employment.

Capital growing — that is, accumulation — brings a falling rate of profit with it only to the extent that this growth carries with it the changes in the ratio between capital's organic components that were looked at above. But now, despite the constant daily upheavals in production, first this and then that larger or smaller part of total capital goes on accumulating for certain periods on the basis of some given average ratio among those components, so that no organic change, and hence no cause of a falling rate of profit, is present in its growth.

This constant enlargement of capital, and so the expansion of production, on the basis of the old method of production — which carries on as before while new methods are already coming in next door — is once more a cause why the profit rate fails to fall in step with the growth of society's total capital.

The absolute number of workers grows even as the variable capital laid out on labour-power falls relatively — but not in every branch of production, and not evenly in the branches where it happens. In agriculture the amount of living labour can fall absolutely, not just its share.

Besides, it is capitalist production alone that requires the number of wage-workers to rise absolutely, while it falls relatively. For it, labour-power becomes superfluous as soon as it no longer has to be kept busy 12 to 15 hours a day.

A development of the productive forces that cut the number of workers absolutely — that is, that actually enabled the whole nation to get through its entire production in a smaller part of the time — would bring about revolution, because most of the population would be left without work.

Here again the specific barrier of capitalist production comes into view, and the fact that it is in no way an absolute form for developing the productive forces and creating wealth; on the contrary, it collides with that development at a certain point. In part this collision shows up in the periodic crises that come when one section of the workers, then another, is made superfluous in the work it used to do.

Its barrier is the surplus time of the workers. Whatever absolute surplus time society gains means nothing to it. For it, the development of productivity matters only in so far as it swells the working class's surplus labour-time, not in so far as it cuts the labour-time that material production in general needs; so it moves in a contradiction.

We have seen that as capital accumulates more and more, it concentrates more and more. So the power of capital grows: the social conditions of production, personified in the capitalist, set themselves up over against the actual producers. Capital shows itself more and more as a social power — the capitalist is its functionary — a power that bears no possible relation any more to what one person's labour can create, but an alienated, independent social power that stands over against society as a thing — and as the capitalist's power, exercised through that thing.

Capital grows into a general social power, while individual capitalists hold private power over the social conditions of production; the clash between the two grows ever louder and holds the ending of this relation inside it, because the same development is at the same time working the conditions of production up into general, communal, social conditions of production. This working-out comes about through the growth of the productive forces under capitalism, and through the way that growth is carried through.

No capitalist brings in a new way of producing voluntarily — however much more productive it may be, or however much it may raise the rate of surplus-value — as soon as it lowers the rate of profit. But every such new method makes commodities cheaper. So to begin with he sells them above their production price, perhaps above their value. What he pockets is the gap between what these goods cost him and the market price of the rest, which are made at higher production costs. He can do that because the labour-time society on average needs to make those goods is greater than what the new method needs: his production procedure stands above the social average.

But competition spreads it everywhere and brings it under the general law. Then the fall in the rate of profit sets in — perhaps first within this branch, then evening out with the rest — a fall entirely independent of any capitalist's will.

One more thing here: the same law prevails in spheres of production whose product enters neither directly nor indirectly into what workers consume, or into the making of the things workers live on — so too in branches where making commodities cheaper cannot raise relative surplus-value or cheapen labour-power. (A cheapening of constant capital in all these branches can of course raise the rate of profit, if the exploitation of the worker stays the same.)

As soon as the new method begins to spread, and so it is actually proved that these goods can be made more cheaply, the capitalists still working under the old conditions must sell their product under its full production price, because the value of this commodity has fallen, and they now need more labour-time to make it than is socially necessary. In a word — it shows up as the work of competition — they too have to bring in the new way of producing, in which variable capital stands lower against constant.

Everything that makes machinery cheapen the goods it turns out comes down, first, to less labour being absorbed in each single commodity, and second, to a smaller wear share of the machinery whose value goes into that commodity. The slower the machinery wears, the more commodities that wear is spread over, and the more living labour it does the work of before it must be replaced. In both cases the quantity and the value of fixed constant capital grow as against variable capital.

Jones (An Introductory Lecture on Political Economy, London, 1833, p. 50 ff.) draws the contrast like this: other things being equal, a nation's power to save out of its profits moves with the rate of profits — large when profits are high, smaller when they are low. But, Jones goes on, when the rate of profit declines, other things do not stay equal. A low rate of profit usually goes together with a rapid rate of accumulation relative to the size of the population, as in England; a high rate of profit with a slower accumulation relative to the size of the population. Examples: Poland, Russia, India, and so on.

Jones rightly stresses that even as the rate of profit falls, the inducements and the faculties for accumulating multiply. First, because relative surplus population keeps growing. Second, because as labour grows more productive, the same exchange-value stands for a greater mass of use-values — more material elements of capital. Third, because the branches of production multiply. Fourth, through the credit system, joint-stock companies and so on, and the ease they give of turning money into capital without becoming an industrial capitalist oneself. Fifth, the growth of needs and of the greed for enrichment. Sixth, the growing mass of fixed capital being invested, and so on.

There are three cardinal facts of capitalist production:

(1) The means of production are concentrated in a few hands. With that they stop looking like the property of the workers who directly use them, and become instead social powers of production — even if at first as the capitalists' own private property. These capitalists are trustees of bourgeois society, but they rake in every fruit of that trusteeship.

(2) Labour itself is organized as social labour: through cooperation, the division of labour, and the joining of labour with natural science.

In both directions, then, capitalist production cancels private property and cancels private labour as well — even if in opposing forms.

(3) The creation of the world market.

The tremendous productive power, measured against the population, that develops inside capitalist production — and, even if not to the same degree, the rise in capital values (not only in their physical material), which grow far faster than the population — stand in contradiction with the ever-narrower basis, relative to growing wealth, on which this immense productive power works, and with the conditions under which this swelling capital can be valorized — can still expand its own value. Hence the crises.