The organic composition of capital depends, at any given moment, on two things: first, the technical proportion between the labour-power and the mass of means of production applied; second, the price of those means of production. As we have seen, it has to be looked at as a percentage. A capital made up of four-fifths constant capital and one-fifth variable capital we write as 80c + 20v.
For comparing spheres, we also hold the rate of surplus-value fixed — some arbitrary rate, say 100 per cent, the same everywhere. A capital of 80c + 20v then throws off surplus-value of 20s, which works out to a rate of profit of 20 per cent on the whole capital.
How much the product is actually worth depends on how large the fixed part of the constant capital is, and on how much of it goes into the product each year as depreciation and how much does not. But that makes no difference to the rate of profit, so to keep things simple we assume the whole of the constant capital, in every case alike, passes into the year's product. We also assume that capitals in different spheres realize the same amount of surplus-value each year in proportion to the size of their variable part — for now we set aside the difference that unequal turnover times could make here. That point comes later.
Let's take five different spheres of production, each with capitals of a different organic composition invested in them, set out as follows:
In these five spheres of production, with a uniform exploitation of labour, we get very different rates of profit, matching the different organic composition of the capitals.
The capitals invested in the five spheres come to 500 in all; the surplus-value they produce comes to 110; the value of the commodities they produce comes to 610.
Suppose we treat this 500 as a single capital, of which spheres I to V are simply different parts — the way a cotton mill has different ratios of variable to constant capital in its different departments, the carding room, the roving room, the spinning room, the weaving room, and the average ratio for the whole mill has to be worked out separately. On that supposition, the average composition of the capital of 500 would be 390c + 110v, or as a percentage 78c + 22v. Each capital of 100, looked at as one-fifth of the total capital, would have this same average composition of 78c + 22v; likewise each 100 would carry an average surplus-value of 22; so the average rate of profit would be 22 per cent, and the price of each fifth of the total product made by the 500 would be 122. The product of each fifth of the capital advanced would have to sell at 122.
But to avoid drawing completely wrong conclusions, we cannot reckon all the cost prices at 100.
With 80c + 20v and a rate of surplus-value of 100 per cent, the total value of the commodity produced by capital I, a capital of 100, would be 80c + 20v + 20s = 120, if the whole of the constant capital passed into the year's product. Now this might well happen in some spheres of production. But it can hardly happen where the ratio of c to v is 4 to 1.
So when we look at the values of the commodities produced by each 100 of the various capitals, we have to bear in mind that these values will differ according to how c is split between fixed and circulating parts, and that the fixed parts of different capitals themselves depreciate faster or more slowly, adding unequal amounts of value to the product in equal periods of time. But none of this makes any difference to the rate of profit. Whether the 80c transfers a value of 80, or 50, or 5, to the year's product — whether the year's product is 80c + 20v + 20s = 120, or 50c + 20v + 20s = 90, or 5c + 20v + 20s = 45 — in every one of these cases the excess of the value of the product over its cost price is 20, and in every one of these cases, when the rate of profit is worked out, that 20 is reckoned on a capital of 100. So the rate of profit of capital I is 20 per cent in every case.
To make this still clearer, the following table lets different parts of the constant capital pass into the value of the product, for the same five capitals as before.
Looking at capitals I to V once more as a single total capital, we find that in this case too the composition of the sum of the five capitals, 500 = 390c + 110v, stays the same, so the average composition is still 78c + 22v and the average surplus-value is still 22. Distributing this surplus-value evenly over I to V gives the following commodity prices:
Taken together, the commodities are sold at 2 + 7 + 17 = 26 above their value and 8 + 18 = 26 below it, so that by spreading the surplus-value evenly — or by adding to the respective cost prices of commodities I to V the average profit of 22 per 100 of capital advanced — the divergences of price from value cancel each other out exactly: to the same extent that one part of the commodities is sold above its value, another is sold below it. It is only by selling at these prices that the rate of profit can come out the same for I to V, at 22 per cent, regardless of the different organic composition of capitals I to V.
The prices that arise when the average is taken of the different rates of profit in the different spheres of production, and this average is added to the cost prices of the different spheres of production, are the prices of production. They presuppose the existence of a general rate of profit, and that in turn presupposes that the rates of profit in each particular sphere of production, taken on its own, have already been reduced to just as many average rates. These particular rates of profit are, in each sphere of production, equal to s/C — the sphere's own surplus-value divided by its own total capital — and have to be developed out of the value of the commodity, as was done in the first Part of this volume. Without that derivation, the general rate of profit — and with it the production price of the commodity — remains a notion without sense or content.
The production price of a commodity is, then, equal to its cost price plus the profit added to it as a percentage, in line with the general rate of profit — or equal to its cost price plus the average profit.
Because capitals invested in different branches of production have different organic compositions, capitals of the same size set very different quantities of labour in motion, depending on what percentage of a given total capital is variable — so they appropriate very different quantities of surplus labour, and produce very different amounts of surplus-value. Correspondingly, the rates of profit prevailing in different branches of production start out very different.
These different rates of profit are balanced out by competition into a general rate of profit, which is the average of all these different rates. The profit that falls, at this general rate, on a capital of a given size — whatever that capital's organic composition — is called the average profit. A commodity's price of production is its cost price plus the share of the annual average profit that falls to it in proportion to its conditions of turnover — a share reckoned on the capital applied in its production (not merely the capital consumed in its production).
Let us take, for example, a capital of 500, of which 100 is fixed capital, depreciating by 10 per cent during one turnover period of the circulating capital of 400. Suppose the average profit for the length of this turnover period is 10 per cent. Then the cost price of the product made during this turnover will be: 10c for depreciation plus 400 (c + v) circulating capital = 410; and its price of production: 410 cost price plus 50 (10 per cent profit on 500) = 460.
So although capitalists in different spheres of production recover, when they sell their commodities, the capital values consumed in producing them, they do not cash in the surplus-value — and so the profit — produced in their own sphere in producing those commodities. They cash in only as much surplus-value, and so profit, as falls to them when the total surplus-value — the total profit — produced in a given period by the total capital of society across all spheres of production taken together is distributed equally over every aliquot part, every equal slice, of that total capital. Whatever its composition, every capital advanced draws, per 100, in each year or other period, the profit that falls for that period on 100 as whatever fraction of the total capital that 100 happens to be. Here, as far as profit is concerned, the different capitalists behave like mere shareholders in a joint-stock company, where the dividends are distributed evenly per 100 and so differ between the various capitalists only by the size of the capital each has put into the joint undertaking, by his proportional stake in it, by the number of his shares.
So while the part of a commodity's price that replaces the portions of capital value consumed in producing the commodities, and with which those consumed capital values have to be bought back — the cost price — is governed entirely by the outlay within the particular sphere of production, the other component of the commodity's price, the profit added onto that cost price, is not governed by the mass of profit actually produced by this particular capital in this particular sphere of production during a given time, but by the mass of profit that falls, on average during a given period, on every capital applied, as an aliquot part of the total social capital applied in production as a whole.
So when a capitalist sells a commodity at its price of production, he draws back money in proportion to the value of the capital he consumed in producing it, and he pockets profit in proportion to his capital advanced, treated simply as an aliquot part of the total social capital. His cost prices are specific to his own production. But the profit added onto that cost price is independent of his particular sphere of production — it is simply the average per 100 of the capital advanced.
Suppose the five different capital investments I to V in the previous example belonged to one man. How much variable and constant capital was consumed, per 100 of capital applied, in producing the commodities in each single investment I to V, would be given; and this portion of the value of commodities I to V would obviously form part of their price, since the price must at the very least replace the part of the capital advanced and consumed. So the cost prices would differ for each kind of commodity from I to V, and the owner would fix them differently for each. But as for the different masses of surplus-value or profit produced in I to V, the capitalist could quite well reckon them as profit on his total capital advanced, so that a fixed aliquot part fell on every 100 of capital. So the cost prices of the commodities produced in the separate investments I to V would differ; but the part of the selling price that came from the profit added per 100 of capital would be the same for all these commodities. The total price of commodities I to V would then be equal to their total value — that is, equal to the sum of the cost prices of I to V plus the sum of the surplus-value or profit produced in I to V; in other words, the money expression for the total quantity of labour, past and newly added, contained in commodities I to V.
And this is how, in society itself — considering the totality of all branches of production — the sum of the prices of production of the commodities produced is equal to the sum of their values.
It looks as if this conflicts with a fact: under capitalist production, the elements of productive capital are usually bought on the market, so their prices already contain a realized profit. On this reckoning those prices carry inside them another branch's production price together with the profit contained in it, so that the profit of one branch enters into the cost price of another.
But if we put the sum of the cost prices of all the commodities in a whole country on one side, and the sum of its profits, or surplus-values, on the other, it's clear the account has to come out right. Take a commodity A: its cost price may include the profits of B, C and D, just as A's profit may in turn enter into the cost prices of B, C, D, and so on.
So when we draw up the account, A's own profit is missing from its own cost price, and the same holds for B, C, D and the rest: each one's profit is missing from its own cost price. No one counts their own profit into their own cost price.
So if there are, say, n spheres of production, and each one makes a profit equal to p, then across all of them together the cost price comes to k - np. Looking at the whole account: wherever one sphere's profits enter into another sphere's cost price, they have already been counted once, in the total price of the final product, and they cannot appear a second time on the profit side. If they do appear on that side, it is only because the commodity itself was the final product, so its price of production never went into another commodity's cost price.
Suppose a sum equal to p, for the profits of the producers of the means of production, enters into a commodity's cost price, and a further profit equal to p1 is then added on top of that cost price. The total profit is then P = p + p1. The commodity's total cost price, once every price-part that goes to profit is stripped out, is then its own cost price minus P. Call that cost price k, and it is clear that k + P = k + p + p1.
We saw earlier, in dealing with surplus-value in Volume 1, that the product of any capital can be treated as if one part merely replaces the capital and the other expresses only surplus-value. Apply that same reckoning to the whole product of society, and corrections have to be made: taking society as a whole, the profit contained in the price of flax, say, cannot figure twice over — once as part of the price of linen, and again as the flax producer's profit.
There is no difference here between profit and surplus-value when A's surplus-value, say, enters into B's constant capital. For the value of commodities it makes no difference at all whether the labour contained in them is paid or unpaid labour. All this shows is that B pays A's surplus-value. In the account for the whole, A's surplus-value cannot count twice.
But there is a real difference. It isn't only that the price of capital B's product deviates from its value, because the surplus-value actually realized in B may be bigger or smaller than the profit added onto the price of B's products. The very same thing holds again for the commodities that make up the constant part of capital B — and, indirectly, as the workers' means of subsistence, its variable part too. Take the constant part: it is itself equal to cost price plus surplus-value, which now means cost price plus profit, and that profit can again be bigger or smaller than the surplus-value it stands in for. Take the variable capital: the average daily wage is always equal to the value the worker himself adds in the hours he has to work to produce the necessary means of subsistence — but that number of hours is itself distorted by the way the production prices of the necessary means of subsistence deviate from their values.
Still, this always resolves itself: whatever goes in too much as surplus-value in one commodity goes in too little in another, so the deviations from value lodged in commodities' production prices cancel each other out.
With the whole of capitalist production, it is only ever in a very complicated and approximate way — as an average of perpetual fluctuations that can never actually be pinned down — that the general law asserts itself as the dominant tendency.
Because the general rate of profit is formed by averaging the different profit rates on every 100 of capital advanced, over a given stretch of time — a year, say — the differences produced by different capitals having different turnover times are wiped out in that average. But those very differences do enter, and do help determine, the different profit rates of the different spheres of production whose average makes up the general rate of profit.
In the illustration just used to build the general rate of profit, every capital in every sphere was set at 100. That was done to bring out clearly the percentage difference in the profit rate, and so also the difference in the values of commodities produced by capitals of equal size. But of course the actual masses of surplus-value produced in each particular sphere depend on the size of the capitals applied, since within any given sphere the composition of capital is fixed. Even so, the particular profit rate of a single sphere is not affected by whether the capital applied is 100, or any multiple of 100, however large. The profit rate stays 10 per cent, whether the total profit is 10 on 100 or 1,000 on 10,000.
But profit rates differ from one sphere of production to another, because the ratio of variable capital to total capital differs, and very different masses of surplus-value, and so of profit, get produced accordingly. So it's clear that the average profit per 100 of the social capital — and hence the average or general rate of profit — will vary a great deal depending on the relative sizes of the capitals invested in the different spheres.
Take four capitals, A, B, C and D. Let the rate of surplus-value be 100 per cent for all of them. Out of every 100 of total capital, let the variable capital be 25 for A, 40 for B, 15 for C, 10 for D. Then out of every 100 of total capital there would fall a surplus-value, or profit, of 25 for A, 40 for B, 15 for C, 10 for D — 90 together. So if the four capitals are equal in size, the average rate of profit is 90/4 = 22½ per cent.
But if the total sizes of the capitals are instead A = 200, B = 300, C = 1,000, D = 4,000, the profits produced would be 50, 120, 150 and 400 respectively. Together, on a capital of 5,500, that is a profit of 720, or an average rate of profit of 13 1/11 per cent.
The masses of total value produced differ according to the different sizes of the total capitals advanced in A, B, C and D. So forming the general rate of profit isn't just a matter of the difference between the profit rates in the different spheres of production, of which a simple average could be taken — it's a matter of the relative weight with which those different profit rates enter into forming the average. And that depends on the proportional size of the capital invested in each particular sphere — on what aliquot part of society's total capital the capital in each sphere forms. There obviously has to be a very great difference according to whether a larger or smaller share of the total capital yields a higher or lower profit rate. And that in turn depends on how much capital sits in the spheres where variable capital is relatively large or small compared with the total capital.
It's exactly like the average rate of interest made by a moneylender who lends out different sums at different rates — say at 4, 5, 6, 7 per cent. His average rate depends entirely on how much of his capital he has lent out at each of the different rates.
The general rate of profit, then, is fixed by two factors:
(1) by the organic composition of the capitals in the different spheres of production — that is, by the different profit rates of the individual spheres;
(2) by the way society's total capital is distributed across these different spheres — that is, by the relative size of the capital invested in each particular sphere, and hence at a particular rate of profit; in other words, by the proportional share of society's total capital that each particular sphere swallows up.
In Volumes 1 and 2 we had only the values of commodities to deal with. Now, on one side, the cost price has separated off as a part of that value; on the other, the production price of the commodity has developed as a transformed form of that value.
Suppose the composition of the average social capital is 80c + 20v, and the annual rate of surplus-value, s', is 100 per cent. Then the annual average profit for a capital of 100 would be 20, and the general annual rate of profit would be 20 per cent. Whatever the cost price k of the commodities a capital of 100 produces in a year, their price of production would be k + 20.
In the spheres of production where the composition of capital is (80 - x)c + (20 + x)v, the surplus-value actually created — that is, the annual profit produced within that sphere — would be 20 + x, greater than 20, and the value of the commodities produced would be k + 20 + x, greater than k + 20, that is, greater than their price of production. In the spheres where the composition of capital is (80 + x)c + (20 - x)v, the annual surplus-value or profit would be 20 - x, less than 20, and so the value of the commodities, k + 20 - x, would be less than the price of production, which is k + 20.
Leaving aside any differences in turnover time, the production prices of commodities would equal their values only in the spheres where the composition of capital happened to be exactly 80c + 20v.
The specific development of the social productive power of labour differs in degree from one particular sphere of production to another — higher in some, lower in others. It is higher where a given quantity of labour — a given number of workers, the working day being fixed — sets a larger quantity of means of production in motion, and so where the quantity of labour needed for a given quantity of means of production is correspondingly small.
We call capitals that contain, percentage-wise, more constant and so less variable capital than the average social capital, capitals of higher composition. Conversely, capitals where the constant capital takes up a relatively smaller share and the variable a larger one, compared with the average social capital, we call capitals of lower composition. Capitals of average composition, finally, are those whose composition coincides with that of the average social capital. If the average social capital is composed, percentage-wise, of 80c + 20v, then a capital of 90c + 10v stands above the social average, one of 70c + 30v below it. In general, if the average social capital's composition is mc + nv, where m and n are constants and m + n = 100, then (m + x)c + (n - x)v represents the higher, and (m - x)c + (n + x)v the lower, composition of an individual capital or group of capitals.
How these capitals behave once the average rate of profit is established — assuming a single turnover in the year — is shown by the following table, in which I stands for the average composition, so that the average rate of profit is 20 per cent:
For the commodities produced by capital II, then, their value would be less than their price of production; for those of capital III, the price of production would be less than the value; and only for capitals like I, in branches of production whose composition happens to match the social average, would value and price of production be equal.
That said, when applying these terms to particular cases, it obviously has to be taken into account how far the ratio between c and v departs from the general average not because of a difference in technical composition at all, but merely because of a change in the value of the elements of constant capital.
The development just given does, however, bring in a modification of how commodities' cost price is determined. It was originally assumed that a commodity's cost price is equal to the value of the commodities consumed in producing it. But a commodity's price of production is, for whoever buys it, simply its cost price, and so it can enter as cost price into the forming of another commodity's price. And since the price of production can deviate from a commodity's value, a commodity's cost price — where it includes another commodity's price of production — can likewise stand above or below the part of its total value formed by the value of the means of production that went into it. It's necessary to keep this modified sense of cost price in mind, and so to remember that whenever, in some particular sphere of production, a commodity's cost price is simply equated with the value of the means of production used up in making it, an error is always possible.
For the investigation we are engaged in here, there is no need to go any further into this point. What still holds throughout is that a commodity's cost price is always less than its value. For however a commodity's cost price may deviate from the value of the means of production consumed in it, that past error makes no difference to the capitalist. The commodity's cost price is simply a given — a premise that is independent of the capitalist's own production — while the result of his production is a commodity that contains surplus-value, that is, a surplus of value over its cost price. Beyond that, the proposition that cost price is less than a commodity's value has now, in practice, turned into the proposition that cost price is less than the price of production. For society's total capital, where the price of production equals value, this comes to the same thing as the earlier proposition, that cost price is less than value. And although the proposition carries a different sense for particular spheres of production taken on their own, the underlying fact remains: taking society's total capital as a whole, the cost price of the commodities it produces is less than their value — or, here, less than the price of production, which for the total mass of commodities produced is identical with that value.
A commodity's cost price refers only to the quantity of paid labour contained in it. Its value refers to the total quantity of paid and unpaid labour contained in it. Its price of production refers to the sum of the paid labour plus a quantity of unpaid labour that is fixed for that particular sphere of production independently of the sphere itself.
The formula that a commodity's price of production is k + p — cost price plus profit — has now been pinned down more closely: p = kp', where p' is the general rate of profit, and so the price of production is k + kp'. If k = 300 and p' = 15 per cent, the price of production, k + kp', is 300 + 300 × 15/100 = 345.
The price of production of commodities in any particular sphere of production can change in size in three ways:
(1) the value of the commodities can stay the same — the same amount of dead and living labour still goes into producing them — while the price of production still changes, because the general rate of profit has changed for reasons that have nothing to do with this particular sphere.
(2) the general rate of profit can stay the same while the value changes — either because of a technical change inside this particular sphere of production itself, or because the value has changed in the commodities that go into its constant capital as formative elements.
(3) finally, both of these can happen together.
Even though the actual profit rates in each particular sphere of production keep changing a great deal, as we will see later, a real change in the general rate of profit — unless some extraordinary economic event forces one through as an exception — is the very late result of a series of oscillations spread over very long stretches of time: oscillations that take a long time before they settle down and add up to a change in the general rate.
Over any shorter period — leaving fluctuations in market prices aside entirely — a change in prices of production is, prima facie, always to be explained by a real change in the value of the commodities, that is, by a change in the total amount of labour-time needed to produce them. A mere change in the money expression of the same values obviously does not come into this at all.
On the other hand, it is clear that, taking the total social capital as a whole, the sum of the values of the commodities it produces (or, expressed in money, their price) equals the value of the constant capital plus the value of the variable capital plus the surplus-value. Taking the degree of exploitation of labour as constant, the rate of profit here can only change, with the mass of surplus-value staying the same, if either the value of the constant capital changes, or the value of the variable capital changes, or both change — so that C itself changes, and with it s/C — the surplus-value divided by that whole capital — which is the general rate of profit. So in every case, a change in the general rate of profit presupposes a change in the value of the commodities that go in as formative elements of the constant capital, or the variable capital, or both at once.
Or the general rate of profit can change while the value of the commodities stays the same, if the degree of exploitation of labour changes.
Or, with the degree of exploitation of labour staying the same, the general rate of profit can change if the total amount of labour applied changes relative to the constant capital, because of technical changes in the labour process. But changes like that must always show up in, and therefore be accompanied by, a change in the value of the commodities, since producing them would now take more or less labour than before.
We saw in the first Part that surplus-value and profit were identical in mass — that is, in total amount. But the rate of profit is different from the rate of surplus-value from the very start — which at first looks like nothing more than a different way of calculating. Yet it is also, from the very start, exactly what obscures and mystifies where surplus-value really comes from: the rate of profit can rise or fall while the rate of surplus-value stays the same, and the other way round, and it is only the rate of profit that practically interests the capitalist. Still, the difference in size was only between the rate of surplus-value and the rate of profit — not between surplus-value and profit themselves.
Because in the rate of profit the surplus-value is reckoned against the total capital and measured by it, surplus-value itself now appears to have sprung from the total capital, equally from all its parts, so that the organic distinction between constant and variable capital — the difference between the part that replaces means of production and the part that sets labour to work — is wiped out in the concept of profit. In this transformed shape, as profit, surplus-value really does deny its own origin, loses its character, becomes unrecognizable.
But so far, the difference between profit and surplus-value was only a qualitative change, a change of form. At this first stage of the transformation, a real difference in size exists only between the rate of profit and the rate of surplus-value — not yet between profit and surplus-value themselves.
It is different once a general rate of profit has been established, and through it an average profit corresponding to the size of the capital applied in each particular sphere of production.
It is now only by chance if the surplus-value actually produced in a particular sphere of production — and so its profit — coincides with the profit contained in the commodity's selling price. As a rule, profit and surplus-value, and not just their rates, are now really different magnitudes. Given a certain degree of exploitation of labour, the mass of surplus-value produced in a particular sphere of production now matters more for the total average profit of social capital — that is, for the capitalist class as a whole — than it matters directly for the capitalist within that particular branch of production. It matters to him only in so far as the quantity of surplus-value produced in his own branch has a part in setting the average profit. But this is a process that goes on behind his back: he does not see it, does not understand it, and it does not in fact interest him.
The real difference in magnitude between profit and surplus-value — not just between their rates — in the particular spheres of production now completely hides the true nature and origin of profit, not only for the capitalist, who here has a particular interest in deceiving himself, but for the worker too. With the transformation of values into prices of production, the very basis for determining value is removed from view.
Finally: in the mere transformation of surplus-value into profit, the part of the commodity's value that forms the profit already stands opposed to the other part of its value as the commodity's cost price. At that point the concept of value is already slipping away from the capitalist, because he does not have in front of him the total labour that the commodity's production costs, but only the part of that total labour which he has paid for, in the form of means of production, dead or living — so that profit appears to him as something standing outside the commodity's immanent value. This idea is now fully confirmed, cemented, set hard, because the profit added on to the cost price is, in fact, looking at the particular sphere of production, not fixed by the limits of the value-formation going on inside that sphere itself, but set quite externally against it.
This inner connection is revealed here for the first time. And previous economics, as later chapters and Volume 4 will show, either forcibly abstracted from the differences between surplus-value and profit, and between the rate of surplus-value and the rate of profit, in order to hold on to the determination of value as its basis, or else gave up the determination of value altogether, sacrificing the whole ground of scientific method, in order to hold on to those differences that stand out so plainly in appearance. This confusion among the theorists shows, better than anything else, how the practical capitalist, caught up in the fight of competition and unable to see through the appearances that struggle throws up, must be utterly incapable of seeing through the surface appearance to the inner essence and inner shape of this process.
All the laws about the rise and fall of the rate of profit developed in the first Part in fact carry the following double meaning:
(1) On the one hand they are the laws of the general rate of profit itself. Given the many different causes that, on the argument so far, make the rate of profit rise or fall, one might expect the general rate of profit to change every day. But the movement in one sphere of production cancels out the movement in another — the influences cross and paralyse each other.
We will look later at which way these fluctuations tend in the end; but they are slow. Because the fluctuations in the different spheres of production are sudden, many-sided, and of different lengths, they partly offset each other in their sequence over time, so that a fall in price follows a rise and the other way round — which keeps them local, confined to their own particular sphere of production. And finally, the various local fluctuations neutralize one another.
Within each particular sphere of production there are changes, departures from the general rate of profit, which on one hand even out again over a given stretch of time and so do not feed back into the general rate; and which on the other hand also do not feed back into it because other local fluctuations happening at the same time cancel them out. Since the general rate of profit is fixed not only by the average rate of profit in each sphere but also by how the total capital is distributed across the different particular spheres, and since that distribution keeps changing, this is itself a constant cause of change in the general rate — but a cause of change which, given how unbroken and many-sided this movement is, for the most part paralyses itself all over again.
(2) Within each sphere there is room, for a shorter or longer stretch of time, for that sphere's own rate of profit to fluctuate before the fluctuation, whether rising or falling, settles down enough to have time to affect the general rate of profit and so reach beyond merely local significance. Within those limits of space and time, then, the laws of the rate of profit developed in the first Part of this volume also hold good.
The theoretical opinion — about the first transformation of surplus-value into profit — that every part of the capital throws off profit equally, expresses a practical fact. In both cases an industrial capital throws off the same amount of profit. However it is composed, whether it sets in motion a quarter dead labour and three-quarters living labour, or three-quarters dead labour and a quarter living labour; whether in one case it draws in three times as much surplus labour, or produces three times as much surplus-value, as in the other — the profit is the same, at an equal degree of exploitation of labour, and setting aside individual differences that vanish anyway, since in both cases we only have the average composition of the whole sphere of production in front of us.
The individual capitalist (or, equally, the whole body of capitalists in any particular sphere of production), whose view is narrow, rightly believes that his profit does not come solely from the labour he employs, or that is employed in his own branch. This is quite correct for his average profit. But how far this profit is mediated by the total exploitation of labour by the total capital — that is, by all his fellow capitalists together — is a complete mystery to him, all the more so because even the bourgeois theorists, the political economists, had not yet uncovered it.
Saving on labour — not only on the labour needed to produce a given product, but also on the number of workers employed — together with a greater use of dead labour, that is, constant capital, appears to be an economically quite sound operation, and seems from the outset not to touch the general rate of profit and the average profit in any way. So how, then, could living labour be the exclusive source of profit, given that reducing the amount of labour needed for production not only does not seem to touch profit, but under certain circumstances even appears as the nearest source of increasing it — at least for the individual capitalist?
If, in a given sphere of production, the part of the cost price that represents the value of the constant capital rises or falls, that part comes in from circulation and enters the commodity's production process already enlarged or reduced from the outset. If, on the other hand, the same number of workers produces more or less in the same time — so that, with the number of workers unchanged, the amount of labour needed to produce a given quantity of commodities changes — then the part of the cost price representing the value of the variable capital may stay the same, entering the cost price of the whole product at the same total size. But each single commodity making up that whole product now carries more or less labour (paid, and therefore also unpaid), and so more or less of the outlay for that labour — a bigger or smaller slice of the wage. The total wage paid by the capitalist stays the same, but reckoned per piece of commodity it is different. So a change would occur here in this part of the commodity's cost price.
Now it makes no difference whether the cost price of the individual commodity rises or falls because of such changes in value — whether in the commodity itself or in the commodities forming its elements (or whether the cost price of the whole quantity of commodities produced by a capital of a given size rises or falls) — if the average profit is, say, 10 per cent, it stays 10 per cent; even though, looked at for the individual commodity, that 10 per cent represents a very different amount, depending on the change in size that the presupposed change in value has produced in the cost price of the individual commodity.
With variable capital — and this is the most important case, because variable capital is the source of surplus-value, and anything that hides its relation to how the capitalist gets richer mystifies the whole system — the situation looks cruder, or this is how it appears to the capitalist: a variable capital of £100, say, represents the weekly wage of 100 workers.
If these 100 workers, in a given working day, produce a weekly product of 200 pieces of commodities = 200C, then — leaving aside the part of the cost price added by the constant capital — since £100 = 200C, 1C costs £100/200 = 10 shillings. Now suppose the productive power of labour changes: it doubles, so the same number of workers produces, in the same time, twice 200C, that is, 400C, where they used to produce 200C. In this case — so far as the cost price consists simply of wages — since now £100 = 400C, 1C costs £100/400 = 5 shillings. If the productive power fell by half instead, the same labour would produce only 200C/2; and since £100 = 200C/2, 1C now costs £200/200 = £1.
Changes in the labour-time needed to produce the commodities, and so in their value, now appear, with respect to the cost price and therefore also the price of production, as nothing but a different distribution of the same wages over more or fewer commodities, depending on whether the same labour-time, for the same wages, produces more or fewer commodities.
What the capitalist — and so also the political economist — sees is that the part of the paid labour falling on each piece of the commodity changes with the productive power of labour, and with it the value of each single piece. What he does not see is that the same is true of the unpaid labour contained in each piece — all the less so, since his average profit is in fact only accidentally determined by the unpaid labour absorbed in his own sphere. Only in such a crude, conceptless form does the fact still show through: that the value of commodities is determined by the labour contained in them.
The organic composition of capital depends, at any given moment, on two things: first, the technical proportion between the labour-power and the mass of means of production applied; second, the price of those means of production. As we have seen, it has to be looked at as a percentage. A capital made up of four-fifths constant capital and one-fifth variable capital we write as 80c + 20v.
For comparing spheres, we also hold the rate of surplus-value fixed — some arbitrary rate, say 100 per cent, the same everywhere. A capital of 80c + 20v then throws off surplus-value of 20s, which works out to a rate of profit of 20 per cent on the whole capital.
How much the product is actually worth depends on how large the fixed part of the constant capital is, and on how much of it goes into the product each year as depreciation and how much does not. But that makes no difference to the rate of profit, so to keep things simple we assume the whole of the constant capital, in every case alike, passes into the year's product. We also assume that capitals in different spheres realize the same amount of surplus-value each year in proportion to the size of their variable part — for now we set aside the difference that unequal turnover times could make here. That point comes later.
Let's take five different spheres of production, each with capitals of a different organic composition invested in them, set out as follows:
In these five spheres of production, with a uniform exploitation of labour, we get very different rates of profit, matching the different organic composition of the capitals.
The capitals invested in the five spheres come to 500 in all; the surplus-value they produce comes to 110; the value of the commodities they produce comes to 610.
Suppose we treat this 500 as a single capital, of which spheres I to V are simply different parts — the way a cotton mill has different ratios of variable to constant capital in its different departments, the carding room, the roving room, the spinning room, the weaving room, and the average ratio for the whole mill has to be worked out separately. On that supposition, the average composition of the capital of 500 would be 390c + 110v, or as a percentage 78c + 22v. Each capital of 100, looked at as one-fifth of the total capital, would have this same average composition of 78c + 22v; likewise each 100 would carry an average surplus-value of 22; so the average rate of profit would be 22 per cent, and the price of each fifth of the total product made by the 500 would be 122. The product of each fifth of the capital advanced would have to sell at 122.
But to avoid drawing completely wrong conclusions, we cannot reckon all the cost prices at 100.
With 80c + 20v and a rate of surplus-value of 100 per cent, the total value of the commodity produced by capital I, a capital of 100, would be 80c + 20v + 20s = 120, if the whole of the constant capital passed into the year's product. Now this might well happen in some spheres of production. But it can hardly happen where the ratio of c to v is 4 to 1.
So when we look at the values of the commodities produced by each 100 of the various capitals, we have to bear in mind that these values will differ according to how c is split between fixed and circulating parts, and that the fixed parts of different capitals themselves depreciate faster or more slowly, adding unequal amounts of value to the product in equal periods of time. But none of this makes any difference to the rate of profit. Whether the 80c transfers a value of 80, or 50, or 5, to the year's product — whether the year's product is 80c + 20v + 20s = 120, or 50c + 20v + 20s = 90, or 5c + 20v + 20s = 45 — in every one of these cases the excess of the value of the product over its cost price is 20, and in every one of these cases, when the rate of profit is worked out, that 20 is reckoned on a capital of 100. So the rate of profit of capital I is 20 per cent in every case.
To make this still clearer, the following table lets different parts of the constant capital pass into the value of the product, for the same five capitals as before.
Looking at capitals I to V once more as a single total capital, we find that in this case too the composition of the sum of the five capitals, 500 = 390c + 110v, stays the same, so the average composition is still 78c + 22v and the average surplus-value is still 22. Distributing this surplus-value evenly over I to V gives the following commodity prices:
Taken together, the commodities are sold at 2 + 7 + 17 = 26 above their value and 8 + 18 = 26 below it, so that by spreading the surplus-value evenly — or by adding to the respective cost prices of commodities I to V the average profit of 22 per 100 of capital advanced — the divergences of price from value cancel each other out exactly: to the same extent that one part of the commodities is sold above its value, another is sold below it. It is only by selling at these prices that the rate of profit can come out the same for I to V, at 22 per cent, regardless of the different organic composition of capitals I to V.
The prices that arise when the average is taken of the different rates of profit in the different spheres of production, and this average is added to the cost prices of the different spheres of production, are the prices of production. They presuppose the existence of a general rate of profit, and that in turn presupposes that the rates of profit in each particular sphere of production, taken on its own, have already been reduced to just as many average rates. These particular rates of profit are, in each sphere of production, equal to s/C — the sphere's own surplus-value divided by its own total capital — and have to be developed out of the value of the commodity, as was done in the first Part of this volume. Without that derivation, the general rate of profit — and with it the production price of the commodity — remains a notion without sense or content.
The production price of a commodity is, then, equal to its cost price plus the profit added to it as a percentage, in line with the general rate of profit — or equal to its cost price plus the average profit.
Because capitals invested in different branches of production have different organic compositions, capitals of the same size set very different quantities of labour in motion, depending on what percentage of a given total capital is variable — so they appropriate very different quantities of surplus labour, and produce very different amounts of surplus-value. Correspondingly, the rates of profit prevailing in different branches of production start out very different.
These different rates of profit are balanced out by competition into a general rate of profit, which is the average of all these different rates. The profit that falls, at this general rate, on a capital of a given size — whatever that capital's organic composition — is called the average profit. A commodity's price of production is its cost price plus the share of the annual average profit that falls to it in proportion to its conditions of turnover — a share reckoned on the capital applied in its production (not merely the capital consumed in its production).
Let us take, for example, a capital of 500, of which 100 is fixed capital, depreciating by 10 per cent during one turnover period of the circulating capital of 400. Suppose the average profit for the length of this turnover period is 10 per cent. Then the cost price of the product made during this turnover will be: 10c for depreciation plus 400 (c + v) circulating capital = 410; and its price of production: 410 cost price plus 50 (10 per cent profit on 500) = 460.
So although capitalists in different spheres of production recover, when they sell their commodities, the capital values consumed in producing them, they do not cash in the surplus-value — and so the profit — produced in their own sphere in producing those commodities. They cash in only as much surplus-value, and so profit, as falls to them when the total surplus-value — the total profit — produced in a given period by the total capital of society across all spheres of production taken together is distributed equally over every aliquot part, every equal slice, of that total capital. Whatever its composition, every capital advanced draws, per 100, in each year or other period, the profit that falls for that period on 100 as whatever fraction of the total capital that 100 happens to be. Here, as far as profit is concerned, the different capitalists behave like mere shareholders in a joint-stock company, where the dividends are distributed evenly per 100 and so differ between the various capitalists only by the size of the capital each has put into the joint undertaking, by his proportional stake in it, by the number of his shares.
So while the part of a commodity's price that replaces the portions of capital value consumed in producing the commodities, and with which those consumed capital values have to be bought back — the cost price — is governed entirely by the outlay within the particular sphere of production, the other component of the commodity's price, the profit added onto that cost price, is not governed by the mass of profit actually produced by this particular capital in this particular sphere of production during a given time, but by the mass of profit that falls, on average during a given period, on every capital applied, as an aliquot part of the total social capital applied in production as a whole.
So when a capitalist sells a commodity at its price of production, he draws back money in proportion to the value of the capital he consumed in producing it, and he pockets profit in proportion to his capital advanced, treated simply as an aliquot part of the total social capital. His cost prices are specific to his own production. But the profit added onto that cost price is independent of his particular sphere of production — it is simply the average per 100 of the capital advanced.
Suppose the five different capital investments I to V in the previous example belonged to one man. How much variable and constant capital was consumed, per 100 of capital applied, in producing the commodities in each single investment I to V, would be given; and this portion of the value of commodities I to V would obviously form part of their price, since the price must at the very least replace the part of the capital advanced and consumed. So the cost prices would differ for each kind of commodity from I to V, and the owner would fix them differently for each. But as for the different masses of surplus-value or profit produced in I to V, the capitalist could quite well reckon them as profit on his total capital advanced, so that a fixed aliquot part fell on every 100 of capital. So the cost prices of the commodities produced in the separate investments I to V would differ; but the part of the selling price that came from the profit added per 100 of capital would be the same for all these commodities. The total price of commodities I to V would then be equal to their total value — that is, equal to the sum of the cost prices of I to V plus the sum of the surplus-value or profit produced in I to V; in other words, the money expression for the total quantity of labour, past and newly added, contained in commodities I to V.
And this is how, in society itself — considering the totality of all branches of production — the sum of the prices of production of the commodities produced is equal to the sum of their values.
It looks as if this conflicts with a fact: under capitalist production, the elements of productive capital are usually bought on the market, so their prices already contain a realized profit. On this reckoning those prices carry inside them another branch's production price together with the profit contained in it, so that the profit of one branch enters into the cost price of another.
But if we put the sum of the cost prices of all the commodities in a whole country on one side, and the sum of its profits, or surplus-values, on the other, it's clear the account has to come out right. Take a commodity A: its cost price may include the profits of B, C and D, just as A's profit may in turn enter into the cost prices of B, C, D, and so on.
So when we draw up the account, A's own profit is missing from its own cost price, and the same holds for B, C, D and the rest: each one's profit is missing from its own cost price. No one counts their own profit into their own cost price.
So if there are, say, n spheres of production, and each one makes a profit equal to p, then across all of them together the cost price comes to k - np. Looking at the whole account: wherever one sphere's profits enter into another sphere's cost price, they have already been counted once, in the total price of the final product, and they cannot appear a second time on the profit side. If they do appear on that side, it is only because the commodity itself was the final product, so its price of production never went into another commodity's cost price.
Suppose a sum equal to p, for the profits of the producers of the means of production, enters into a commodity's cost price, and a further profit equal to p1 is then added on top of that cost price. The total profit is then P = p + p1. The commodity's total cost price, once every price-part that goes to profit is stripped out, is then its own cost price minus P. Call that cost price k, and it is clear that k + P = k + p + p1.
We saw earlier, in dealing with surplus-value in Volume 1, that the product of any capital can be treated as if one part merely replaces the capital and the other expresses only surplus-value. Apply that same reckoning to the whole product of society, and corrections have to be made: taking society as a whole, the profit contained in the price of flax, say, cannot figure twice over — once as part of the price of linen, and again as the flax producer's profit.
There is no difference here between profit and surplus-value when A's surplus-value, say, enters into B's constant capital. For the value of commodities it makes no difference at all whether the labour contained in them is paid or unpaid labour. All this shows is that B pays A's surplus-value. In the account for the whole, A's surplus-value cannot count twice.
But there is a real difference. It isn't only that the price of capital B's product deviates from its value, because the surplus-value actually realized in B may be bigger or smaller than the profit added onto the price of B's products. The very same thing holds again for the commodities that make up the constant part of capital B — and, indirectly, as the workers' means of subsistence, its variable part too. Take the constant part: it is itself equal to cost price plus surplus-value, which now means cost price plus profit, and that profit can again be bigger or smaller than the surplus-value it stands in for. Take the variable capital: the average daily wage is always equal to the value the worker himself adds in the hours he has to work to produce the necessary means of subsistence — but that number of hours is itself distorted by the way the production prices of the necessary means of subsistence deviate from their values.
Still, this always resolves itself: whatever goes in too much as surplus-value in one commodity goes in too little in another, so the deviations from value lodged in commodities' production prices cancel each other out.
With the whole of capitalist production, it is only ever in a very complicated and approximate way — as an average of perpetual fluctuations that can never actually be pinned down — that the general law asserts itself as the dominant tendency.
Because the general rate of profit is formed by averaging the different profit rates on every 100 of capital advanced, over a given stretch of time — a year, say — the differences produced by different capitals having different turnover times are wiped out in that average. But those very differences do enter, and do help determine, the different profit rates of the different spheres of production whose average makes up the general rate of profit.
In the illustration just used to build the general rate of profit, every capital in every sphere was set at 100. That was done to bring out clearly the percentage difference in the profit rate, and so also the difference in the values of commodities produced by capitals of equal size. But of course the actual masses of surplus-value produced in each particular sphere depend on the size of the capitals applied, since within any given sphere the composition of capital is fixed. Even so, the particular profit rate of a single sphere is not affected by whether the capital applied is 100, or any multiple of 100, however large. The profit rate stays 10 per cent, whether the total profit is 10 on 100 or 1,000 on 10,000.
But profit rates differ from one sphere of production to another, because the ratio of variable capital to total capital differs, and very different masses of surplus-value, and so of profit, get produced accordingly. So it's clear that the average profit per 100 of the social capital — and hence the average or general rate of profit — will vary a great deal depending on the relative sizes of the capitals invested in the different spheres.
Take four capitals, A, B, C and D. Let the rate of surplus-value be 100 per cent for all of them. Out of every 100 of total capital, let the variable capital be 25 for A, 40 for B, 15 for C, 10 for D. Then out of every 100 of total capital there would fall a surplus-value, or profit, of 25 for A, 40 for B, 15 for C, 10 for D — 90 together. So if the four capitals are equal in size, the average rate of profit is 90/4 = 22½ per cent.
But if the total sizes of the capitals are instead A = 200, B = 300, C = 1,000, D = 4,000, the profits produced would be 50, 120, 150 and 400 respectively. Together, on a capital of 5,500, that is a profit of 720, or an average rate of profit of 13 1/11 per cent.
The masses of total value produced differ according to the different sizes of the total capitals advanced in A, B, C and D. So forming the general rate of profit isn't just a matter of the difference between the profit rates in the different spheres of production, of which a simple average could be taken — it's a matter of the relative weight with which those different profit rates enter into forming the average. And that depends on the proportional size of the capital invested in each particular sphere — on what aliquot part of society's total capital the capital in each sphere forms. There obviously has to be a very great difference according to whether a larger or smaller share of the total capital yields a higher or lower profit rate. And that in turn depends on how much capital sits in the spheres where variable capital is relatively large or small compared with the total capital.
It's exactly like the average rate of interest made by a moneylender who lends out different sums at different rates — say at 4, 5, 6, 7 per cent. His average rate depends entirely on how much of his capital he has lent out at each of the different rates.
The general rate of profit, then, is fixed by two factors:
(1) by the organic composition of the capitals in the different spheres of production — that is, by the different profit rates of the individual spheres;
(2) by the way society's total capital is distributed across these different spheres — that is, by the relative size of the capital invested in each particular sphere, and hence at a particular rate of profit; in other words, by the proportional share of society's total capital that each particular sphere swallows up.
In Volumes 1 and 2 we had only the values of commodities to deal with. Now, on one side, the cost price has separated off as a part of that value; on the other, the production price of the commodity has developed as a transformed form of that value.
Suppose the composition of the average social capital is 80c + 20v, and the annual rate of surplus-value, s', is 100 per cent. Then the annual average profit for a capital of 100 would be 20, and the general annual rate of profit would be 20 per cent. Whatever the cost price k of the commodities a capital of 100 produces in a year, their price of production would be k + 20.
In the spheres of production where the composition of capital is (80 - x)c + (20 + x)v, the surplus-value actually created — that is, the annual profit produced within that sphere — would be 20 + x, greater than 20, and the value of the commodities produced would be k + 20 + x, greater than k + 20, that is, greater than their price of production. In the spheres where the composition of capital is (80 + x)c + (20 - x)v, the annual surplus-value or profit would be 20 - x, less than 20, and so the value of the commodities, k + 20 - x, would be less than the price of production, which is k + 20.
Leaving aside any differences in turnover time, the production prices of commodities would equal their values only in the spheres where the composition of capital happened to be exactly 80c + 20v.
The specific development of the social productive power of labour differs in degree from one particular sphere of production to another — higher in some, lower in others. It is higher where a given quantity of labour — a given number of workers, the working day being fixed — sets a larger quantity of means of production in motion, and so where the quantity of labour needed for a given quantity of means of production is correspondingly small.
We call capitals that contain, percentage-wise, more constant and so less variable capital than the average social capital, capitals of higher composition. Conversely, capitals where the constant capital takes up a relatively smaller share and the variable a larger one, compared with the average social capital, we call capitals of lower composition. Capitals of average composition, finally, are those whose composition coincides with that of the average social capital. If the average social capital is composed, percentage-wise, of 80c + 20v, then a capital of 90c + 10v stands above the social average, one of 70c + 30v below it. In general, if the average social capital's composition is mc + nv, where m and n are constants and m + n = 100, then (m + x)c + (n - x)v represents the higher, and (m - x)c + (n + x)v the lower, composition of an individual capital or group of capitals.
How these capitals behave once the average rate of profit is established — assuming a single turnover in the year — is shown by the following table, in which I stands for the average composition, so that the average rate of profit is 20 per cent:
For the commodities produced by capital II, then, their value would be less than their price of production; for those of capital III, the price of production would be less than the value; and only for capitals like I, in branches of production whose composition happens to match the social average, would value and price of production be equal.
That said, when applying these terms to particular cases, it obviously has to be taken into account how far the ratio between c and v departs from the general average not because of a difference in technical composition at all, but merely because of a change in the value of the elements of constant capital.
The development just given does, however, bring in a modification of how commodities' cost price is determined. It was originally assumed that a commodity's cost price is equal to the value of the commodities consumed in producing it. But a commodity's price of production is, for whoever buys it, simply its cost price, and so it can enter as cost price into the forming of another commodity's price. And since the price of production can deviate from a commodity's value, a commodity's cost price — where it includes another commodity's price of production — can likewise stand above or below the part of its total value formed by the value of the means of production that went into it. It's necessary to keep this modified sense of cost price in mind, and so to remember that whenever, in some particular sphere of production, a commodity's cost price is simply equated with the value of the means of production used up in making it, an error is always possible.
For the investigation we are engaged in here, there is no need to go any further into this point. What still holds throughout is that a commodity's cost price is always less than its value. For however a commodity's cost price may deviate from the value of the means of production consumed in it, that past error makes no difference to the capitalist. The commodity's cost price is simply a given — a premise that is independent of the capitalist's own production — while the result of his production is a commodity that contains surplus-value, that is, a surplus of value over its cost price. Beyond that, the proposition that cost price is less than a commodity's value has now, in practice, turned into the proposition that cost price is less than the price of production. For society's total capital, where the price of production equals value, this comes to the same thing as the earlier proposition, that cost price is less than value. And although the proposition carries a different sense for particular spheres of production taken on their own, the underlying fact remains: taking society's total capital as a whole, the cost price of the commodities it produces is less than their value — or, here, less than the price of production, which for the total mass of commodities produced is identical with that value.
A commodity's cost price refers only to the quantity of paid labour contained in it. Its value refers to the total quantity of paid and unpaid labour contained in it. Its price of production refers to the sum of the paid labour plus a quantity of unpaid labour that is fixed for that particular sphere of production independently of the sphere itself.
The formula that a commodity's price of production is k + p — cost price plus profit — has now been pinned down more closely: p = kp', where p' is the general rate of profit, and so the price of production is k + kp'. If k = 300 and p' = 15 per cent, the price of production, k + kp', is 300 + 300 × 15/100 = 345.
The price of production of commodities in any particular sphere of production can change in size in three ways:
(1) the value of the commodities can stay the same — the same amount of dead and living labour still goes into producing them — while the price of production still changes, because the general rate of profit has changed for reasons that have nothing to do with this particular sphere.
(2) the general rate of profit can stay the same while the value changes — either because of a technical change inside this particular sphere of production itself, or because the value has changed in the commodities that go into its constant capital as formative elements.
(3) finally, both of these can happen together.
Even though the actual profit rates in each particular sphere of production keep changing a great deal, as we will see later, a real change in the general rate of profit — unless some extraordinary economic event forces one through as an exception — is the very late result of a series of oscillations spread over very long stretches of time: oscillations that take a long time before they settle down and add up to a change in the general rate.
Over any shorter period — leaving fluctuations in market prices aside entirely — a change in prices of production is, prima facie, always to be explained by a real change in the value of the commodities, that is, by a change in the total amount of labour-time needed to produce them. A mere change in the money expression of the same values obviously does not come into this at all.
On the other hand, it is clear that, taking the total social capital as a whole, the sum of the values of the commodities it produces (or, expressed in money, their price) equals the value of the constant capital plus the value of the variable capital plus the surplus-value. Taking the degree of exploitation of labour as constant, the rate of profit here can only change, with the mass of surplus-value staying the same, if either the value of the constant capital changes, or the value of the variable capital changes, or both change — so that C itself changes, and with it s/C — the surplus-value divided by that whole capital — which is the general rate of profit. So in every case, a change in the general rate of profit presupposes a change in the value of the commodities that go in as formative elements of the constant capital, or the variable capital, or both at once.
Or the general rate of profit can change while the value of the commodities stays the same, if the degree of exploitation of labour changes.
Or, with the degree of exploitation of labour staying the same, the general rate of profit can change if the total amount of labour applied changes relative to the constant capital, because of technical changes in the labour process. But changes like that must always show up in, and therefore be accompanied by, a change in the value of the commodities, since producing them would now take more or less labour than before.
We saw in the first Part that surplus-value and profit were identical in mass — that is, in total amount. But the rate of profit is different from the rate of surplus-value from the very start — which at first looks like nothing more than a different way of calculating. Yet it is also, from the very start, exactly what obscures and mystifies where surplus-value really comes from: the rate of profit can rise or fall while the rate of surplus-value stays the same, and the other way round, and it is only the rate of profit that practically interests the capitalist. Still, the difference in size was only between the rate of surplus-value and the rate of profit — not between surplus-value and profit themselves.
Because in the rate of profit the surplus-value is reckoned against the total capital and measured by it, surplus-value itself now appears to have sprung from the total capital, equally from all its parts, so that the organic distinction between constant and variable capital — the difference between the part that replaces means of production and the part that sets labour to work — is wiped out in the concept of profit. In this transformed shape, as profit, surplus-value really does deny its own origin, loses its character, becomes unrecognizable.
But so far, the difference between profit and surplus-value was only a qualitative change, a change of form. At this first stage of the transformation, a real difference in size exists only between the rate of profit and the rate of surplus-value — not yet between profit and surplus-value themselves.
It is different once a general rate of profit has been established, and through it an average profit corresponding to the size of the capital applied in each particular sphere of production.
It is now only by chance if the surplus-value actually produced in a particular sphere of production — and so its profit — coincides with the profit contained in the commodity's selling price. As a rule, profit and surplus-value, and not just their rates, are now really different magnitudes. Given a certain degree of exploitation of labour, the mass of surplus-value produced in a particular sphere of production now matters more for the total average profit of social capital — that is, for the capitalist class as a whole — than it matters directly for the capitalist within that particular branch of production. It matters to him only in so far as the quantity of surplus-value produced in his own branch has a part in setting the average profit. But this is a process that goes on behind his back: he does not see it, does not understand it, and it does not in fact interest him.
The real difference in magnitude between profit and surplus-value — not just between their rates — in the particular spheres of production now completely hides the true nature and origin of profit, not only for the capitalist, who here has a particular interest in deceiving himself, but for the worker too. With the transformation of values into prices of production, the very basis for determining value is removed from view.
Finally: in the mere transformation of surplus-value into profit, the part of the commodity's value that forms the profit already stands opposed to the other part of its value as the commodity's cost price. At that point the concept of value is already slipping away from the capitalist, because he does not have in front of him the total labour that the commodity's production costs, but only the part of that total labour which he has paid for, in the form of means of production, dead or living — so that profit appears to him as something standing outside the commodity's immanent value. This idea is now fully confirmed, cemented, set hard, because the profit added on to the cost price is, in fact, looking at the particular sphere of production, not fixed by the limits of the value-formation going on inside that sphere itself, but set quite externally against it.
This inner connection is revealed here for the first time. And previous economics, as later chapters and Volume 4 will show, either forcibly abstracted from the differences between surplus-value and profit, and between the rate of surplus-value and the rate of profit, in order to hold on to the determination of value as its basis, or else gave up the determination of value altogether, sacrificing the whole ground of scientific method, in order to hold on to those differences that stand out so plainly in appearance. This confusion among the theorists shows, better than anything else, how the practical capitalist, caught up in the fight of competition and unable to see through the appearances that struggle throws up, must be utterly incapable of seeing through the surface appearance to the inner essence and inner shape of this process.
All the laws about the rise and fall of the rate of profit developed in the first Part in fact carry the following double meaning:
(1) On the one hand they are the laws of the general rate of profit itself. Given the many different causes that, on the argument so far, make the rate of profit rise or fall, one might expect the general rate of profit to change every day. But the movement in one sphere of production cancels out the movement in another — the influences cross and paralyse each other.
We will look later at which way these fluctuations tend in the end; but they are slow. Because the fluctuations in the different spheres of production are sudden, many-sided, and of different lengths, they partly offset each other in their sequence over time, so that a fall in price follows a rise and the other way round — which keeps them local, confined to their own particular sphere of production. And finally, the various local fluctuations neutralize one another.
Within each particular sphere of production there are changes, departures from the general rate of profit, which on one hand even out again over a given stretch of time and so do not feed back into the general rate; and which on the other hand also do not feed back into it because other local fluctuations happening at the same time cancel them out. Since the general rate of profit is fixed not only by the average rate of profit in each sphere but also by how the total capital is distributed across the different particular spheres, and since that distribution keeps changing, this is itself a constant cause of change in the general rate — but a cause of change which, given how unbroken and many-sided this movement is, for the most part paralyses itself all over again.
(2) Within each sphere there is room, for a shorter or longer stretch of time, for that sphere's own rate of profit to fluctuate before the fluctuation, whether rising or falling, settles down enough to have time to affect the general rate of profit and so reach beyond merely local significance. Within those limits of space and time, then, the laws of the rate of profit developed in the first Part of this volume also hold good.
The theoretical opinion — about the first transformation of surplus-value into profit — that every part of the capital throws off profit equally, expresses a practical fact. In both cases an industrial capital throws off the same amount of profit. However it is composed, whether it sets in motion a quarter dead labour and three-quarters living labour, or three-quarters dead labour and a quarter living labour; whether in one case it draws in three times as much surplus labour, or produces three times as much surplus-value, as in the other — the profit is the same, at an equal degree of exploitation of labour, and setting aside individual differences that vanish anyway, since in both cases we only have the average composition of the whole sphere of production in front of us.
The individual capitalist (or, equally, the whole body of capitalists in any particular sphere of production), whose view is narrow, rightly believes that his profit does not come solely from the labour he employs, or that is employed in his own branch. This is quite correct for his average profit. But how far this profit is mediated by the total exploitation of labour by the total capital — that is, by all his fellow capitalists together — is a complete mystery to him, all the more so because even the bourgeois theorists, the political economists, had not yet uncovered it.
Saving on labour — not only on the labour needed to produce a given product, but also on the number of workers employed — together with a greater use of dead labour, that is, constant capital, appears to be an economically quite sound operation, and seems from the outset not to touch the general rate of profit and the average profit in any way. So how, then, could living labour be the exclusive source of profit, given that reducing the amount of labour needed for production not only does not seem to touch profit, but under certain circumstances even appears as the nearest source of increasing it — at least for the individual capitalist?
If, in a given sphere of production, the part of the cost price that represents the value of the constant capital rises or falls, that part comes in from circulation and enters the commodity's production process already enlarged or reduced from the outset. If, on the other hand, the same number of workers produces more or less in the same time — so that, with the number of workers unchanged, the amount of labour needed to produce a given quantity of commodities changes — then the part of the cost price representing the value of the variable capital may stay the same, entering the cost price of the whole product at the same total size. But each single commodity making up that whole product now carries more or less labour (paid, and therefore also unpaid), and so more or less of the outlay for that labour — a bigger or smaller slice of the wage. The total wage paid by the capitalist stays the same, but reckoned per piece of commodity it is different. So a change would occur here in this part of the commodity's cost price.
Now it makes no difference whether the cost price of the individual commodity rises or falls because of such changes in value — whether in the commodity itself or in the commodities forming its elements (or whether the cost price of the whole quantity of commodities produced by a capital of a given size rises or falls) — if the average profit is, say, 10 per cent, it stays 10 per cent; even though, looked at for the individual commodity, that 10 per cent represents a very different amount, depending on the change in size that the presupposed change in value has produced in the cost price of the individual commodity.
With variable capital — and this is the most important case, because variable capital is the source of surplus-value, and anything that hides its relation to how the capitalist gets richer mystifies the whole system — the situation looks cruder, or this is how it appears to the capitalist: a variable capital of £100, say, represents the weekly wage of 100 workers.
If these 100 workers, in a given working day, produce a weekly product of 200 pieces of commodities = 200C, then — leaving aside the part of the cost price added by the constant capital — since £100 = 200C, 1C costs £100/200 = 10 shillings. Now suppose the productive power of labour changes: it doubles, so the same number of workers produces, in the same time, twice 200C, that is, 400C, where they used to produce 200C. In this case — so far as the cost price consists simply of wages — since now £100 = 400C, 1C costs £100/400 = 5 shillings. If the productive power fell by half instead, the same labour would produce only 200C/2; and since £100 = 200C/2, 1C now costs £200/200 = £1.
Changes in the labour-time needed to produce the commodities, and so in their value, now appear, with respect to the cost price and therefore also the price of production, as nothing but a different distribution of the same wages over more or fewer commodities, depending on whether the same labour-time, for the same wages, produces more or fewer commodities.
What the capitalist — and so also the political economist — sees is that the part of the paid labour falling on each piece of the commodity changes with the productive power of labour, and with it the value of each single piece. What he does not see is that the same is true of the unpaid labour contained in each piece — all the less so, since his average profit is in fact only accidentally determined by the unpaid labour absorbed in his own sphere. Only in such a crude, conceptless form does the fact still show through: that the value of commodities is determined by the labour contained in them.