Volume 2 showed that capital's pure functions in the circulation sphere create no value and no surplus-value. These jobs are the ones the industrial capitalist has to do to get the value of his commodities realized, and then to turn that value back into the elements of production for those commodities — the jobs that mediate the changes of form of commodity capital, C'–M–C, which is to say the acts of selling and buying. And it also came out the other way round: the time these jobs take sets limits on the formation of value and surplus-value — limits that hold objectively for the commodities and subjectively for the capitalist.
None of that is changed in any way by part of this commodity capital taking on the shape of commercial capital, or by the jobs that mediate the change of form of commodity capital appearing as the special business of a special section of capitalists, or as the exclusive function of one portion of the money capital.
The change of form of commodity capital, C'–M–C, comes down to the selling and buying of commodities. If these are not operations that create value or surplus-value when the industrial capitalists themselves perform them, then they cannot possibly become such operations by being performed by other people instead. The same holds for the portion of society's total capital that must always be available as money capital so that the reproduction process is not interrupted by the circulation process but runs on continuously: if this money capital creates no value and no surplus-value, it cannot acquire those properties from being constantly thrown into circulation by another section of capitalists, doing the same jobs, instead of by the industrial capitalist.
How far commercial capital can be indirectly productive has already been hinted at, and will be gone into further later on.
Take commercial capital and strip off every other function that may be attached to it — storing goods, forwarding them, transporting them, sorting them, splitting them up into small lots. What is left is its true function: buying things in order to sell them again. Doing that, it creates neither value nor surplus-value. It only mediates their realization, and with that the actual exchange of commodities — their move out of one pair of hands into another, society's metabolism.
All the same, the circulation phase of industrial capital is a phase of the reproduction process just as much as production is. So capital that operates independently in the circulation process must yield the annual average profit just as much as capital at work in the different branches of production. If commercial capital brought in a higher average profit, in percentage terms, than industrial capital, part of industrial capital would turn into commercial capital. If it brought in a lower one, the opposite process would take place: part of commercial capital would turn into industrial capital. No kind of capital finds it easier to change its calling, its function, than commercial capital.
Since commercial capital itself produces no surplus-value, it is clear that the surplus-value falling to it in the form of average profit is only a piece of the surplus-value that productive capital as a whole produces. But now the question is: how does commercial capital pull in the share that falls to it of the surplus-value, or profit, that productive capital produces?
It is only a semblance that commercial profit consists of a mere supplement, a nominal raising of prices above the value of the commodities.
It is clear that the merchant can get his profit from nowhere but the price of the commodities he sells — and further, that the profit he makes when he sells them must be equal to the difference between what he pays for them and what he sells them for, the amount by which the sale price is above the purchase price.
It can happen that extra costs — circulation costs — go into the commodity after it is bought and before it is sold; and it can just as well not happen. If such costs do go in, then clearly the amount by which the sale price exceeds the purchase price is not all profit. To keep the investigation simple, we assume for the time being that no such costs come in.
With the industrial capitalist, the gap between what he pays for his commodities and what he sells them for is the gap between a commodity's cost price and its price of production — or, if we look at the total social capital, the gap between the value of the commodities and their cost price to the capitalists, which in turn comes down to the difference between all the labour embodied in them and the part of it that was paid for. Before the commodities an industrial capitalist has bought are thrown back on the market as things to be sold, they go through the production process, and only there is the part of their price that is later realized as profit actually produced.
With the commodity dealer it is otherwise. The commodities stay in his hands only for as long as their circulation process lasts. He merely carries on the sale the productive capitalist started, the turning of their price into money, and so he never puts them through any process in between, where they might soak up fresh surplus-value. The industrial capitalist, in circulation, only realizes a surplus-value or profit that production has already put there; the merchant, by contrast, is supposed not merely to realize his profit in circulation and through it, but to make it there in the first place. And that seems possible only if he sells above their production prices the goods the industrial capitalist sold to him at those production prices — or, taking the commodity capital as a whole, at their values; that is, if he tacks a nominal addition on to their prices and, taking the commodity capital as a whole, sells it above its value and pockets the difference between its nominal value and its real value — in a word, sells it dearer than it is.
The way this addition works is very easy to understand. Take a yard of linen at 2 shillings. If I want to make 10 per cent profit on reselling it, I have to put one-tenth on the price, and so I sell the yard at 2s. 2⅖d. The difference between its real price of production and its sale price is then 2⅖d., and on 2 shillings that is a profit of 10 per cent.
In fact I am then selling the buyer the yard at a price that is in truth the price of 1 1/10 yards. Or, what comes to the same thing: it is exactly as if the buyer got only 10/11 of a yard for his 2 shillings, and I kept 1/11 of a yard for myself. With the 2⅖d. I can in fact buy 1/11 of a yard back, reckoning the yard at 2s. 2⅖d.
So this would be nothing but a roundabout way of getting a share of the surplus-value and of the surplus product, by nominally raising the prices of commodities.
This is commercial profit realized through an addition to the price of commodities, the way it first presents itself. And in fact the whole notion that profit springs from a nominal raising of commodity prices, or from selling commodities above their value, has its origin in the outlook of commercial capital.
Looked at more closely, however, it soon turns out that this is a pure illusion — and that, if we assume the capitalist mode of production to be the dominant one, commercial profit is not realized in this way. (What is at issue here is always the average, never individual cases.)
Why do we take it that the only way the merchant can realize, say, 10 per cent on his commodities is to sell them 10 per cent above their prices of production? Because we assumed that the producer of these commodities — the industrial capitalist, who, being the personification of industrial capital, always figures to the outside world as 'the producer' — had sold them to the merchant at their price of production. If the purchase prices the merchant pays are equal to the prices of production of the commodities, and in the last analysis to their values, so that the price of production, and in the last analysis the value, is the cost price for the merchant, then what his sale price gains over what he paid for the goods — and that difference alone is where his profit comes from — has to be a gain of their commercial price over their production price, and in the last analysis the merchant sells every commodity above its value.
But why did we assume that it is at their prices of production that the industrial capitalist sells to the merchant? Or rather, what was presupposed in that assumption? That merchant's capital — and here we are dealing with it still only as commercial, commodity-dealing capital — does not enter into how the general rate of profit is formed. We had to start from this presupposition in presenting the general rate of profit: first, because merchant's capital as such did not yet exist for us at that point; and second, because the average profit, and with it the general rate of profit, first had to be developed by equalizing the profits, or surplus-values, that the industrial capitals in the different spheres of production actually produce.
With merchant's capital, by contrast, we have a capital that gets a share of the profit without having any part in producing it. So the earlier presentation now needs supplementing.
Say the total industrial capital advanced over the year is 720c + 180v = 900 (say, millions of pounds sterling), and s′ = 100 per cent. The product is then 720c + 180v + 180s. Call this product — the commodity capital produced — C. Its value, or price of production, comes to 1,080 (taking all commodities together, the two are one and the same), and the 900 yields a profit of 20 per cent. That 20 per cent is the average rate of profit, as worked out earlier: here the surplus-value is reckoned on the whole industrial capital at its average composition, not on any one capital of a particular make-up. So C is 1,080, and profit stands at 20 per cent.
Now suppose that alongside this 900 of industrial capital there is also 100 of commercial capital, which gets the same share of the profit in proportion to its size. On our assumption it is a tenth of the total capital of 1,000. So it takes a tenth of that 180 of surplus-value, and gets a profit at a rate of 18 per cent. In fact, then, the profit still to be shared among the other nine tenths of the total capital is only 162 — which is likewise 18 per cent on the 900.
So the owners of the 900 of industrial capital sell C to the merchants at 720c + 180v + 162s = 1,062. And when the merchant adds the 18 per cent average profit to his 100, he sells them at 1,062 + 18 = 1,080. That is their price of production — or, if we look at the commodity capital as a whole, their value. And yet he makes his profit only in circulation and through it, and only out of the excess of his sale price over what he paid. All the same, he is not selling above the commodities' value, not above their price of production — simply because the price the industrial capitalists sold to him at was below that value, below that price of production.
Commercial capital therefore helps fix the general rate of profit, in proportion to the share it makes up of the total capital. Say the average rate of profit in our case is 18 per cent. Take the commercial capital away — that tenth of the whole — and it would stand at 20, for that tenth is what pulls the general rate down by a tenth. This also brings in a closer, more restrictive definition of the price of production.
Price of production still means the price of the commodity = its cost (that is, what the constant and variable capital in it are worth) + the average profit on that. But this average profit is now fixed on a different basis. What fixes it is the total profit that the whole productive capital produces — but the reckoning is not made on that productive capital by itself. If it were, then with 900 as above and a profit of 180 the average rate of profit would come out at 180/900 = 20 per cent. The reckoning is made on productive and commercial capital together: so with 900 productive and 100 commercial, the average rate of profit works out at 180/1,000 = 18 per cent. So the price of production comes to k (the cost) + 18, not k + 20.
The slice of the total profit that falls to commercial capital is already counted into the average rate of profit. The real value — the real price of production — of the whole commodity capital is therefore k + p + m, where k is the cost, p the average profit, and m the commercial profit. So the price of production — what the industrial capitalist gets when he sells as an industrial capitalist — is smaller than the commodity's real price of production. Or, looking at all commodities together: the industrial capitalist class sells them for less than their value. In our case: 900 (the cost) plus 18 per cent of 900, or 900 + 162 = 1,062.
And now: in selling for 118 goods that cost him 100, the merchant does add 18 per cent; but since the goods he bought for 100 are worth 118, he is not on that account selling them above their value. We keep the expression 'price of production' in the stricter sense set out above. It is then clear that the industrial capitalist's profit is what the commodity's production price has over its cost price, and that — unlike this industrial profit — commercial profit is what the sale price has over that production price, which for the merchant is what he paid for the goods; while the real price of the commodity = its production price + the commercial profit. Industrial capital only realizes a profit that is already there in the commodity's value as surplus-value; commercial capital likewise only realizes one, and can do so only because when industrial capital sells, its price leaves part of the surplus-value — part of the total profit — not yet realized. So the merchant sells for more than he paid — but the reason is not that his selling price is above the goods' total value; it is that the price he paid was below it.
So commercial capital has a hand in equalizing surplus-value into average profit, though no hand in producing that surplus-value. Which means the general rate of profit already has inside it the deduction from surplus-value that belongs to commercial capital — a slice taken out of the industrial capitalist's profit.
From what has come before, two things follow.
First: the bigger commercial capital is in comparison with industrial capital, the lower industrial capital's rate of profit — and the smaller it is by comparison, the higher that rate.
Second: Part One showed that the rate of profit always comes out as a smaller figure than the rate of surplus-value actually is — that is, it always understates how much labour is being exploited. In the case above, for instance, 720c + 180v + 180s gives a rate of surplus-value of 100 per cent, but it shows up as a rate of profit of only 20 per cent.
That gap widens still further in so far as the average rate of profit itself — once the share falling to commercial capital is reckoned in — appears smaller again: here 18 per cent instead of 20 per cent. So the average rate of profit for the capitalist who does the exploiting directly makes the rate of profit look smaller than it really is.
Assuming that all other circumstances stay the same, commercial capital's relative size (retail traders, a hybrid species, are the exception here) moves inversely to the speed of its turnover, and so inversely to the vigour of the reproduction process as a whole.
When we follow the analysis, the general rate of profit appears to form first among the industrial capitals and their competition, and to be rectified, supplemented and modified by the intervention of commercial capital only later on. Historically, matters stand exactly the other way round. It is commercial capital that first fixes the prices of commodities, more or less by their values, and it is the sphere of circulation — the sphere that mediates the reproduction process — where a general rate of profit first takes shape. Commercial profit is what originally determines industrial profit. Only once the capitalist mode of production has pushed through, and the producer has himself become a merchant, is commercial profit cut down to its aliquot share: the portion of the total surplus-value that belongs to commercial capital as an aliquot part of the total capital taking part in society's process of reproduction.
In the supplementary equalization of profits that commercial capital's intervention brings about, two things came out. Nothing extra enters the value of the commodity on account of the money capital the merchant advances. And the addition to the price — the addition by which the merchant makes his profit — is exactly equal to the piece of the commodity's value that productive capital left out, never reckoned in, when it set the commodity's production price.
His money capital behaves just like the industrial capitalist's fixed capital: so far as it is not used up, its value forms no element of the commodity's value. In the price he pays for the commodity capital, he replaces that capital's production price with money: production price = M. His sale price, as worked out earlier, is M + ΔM, where ΔM stands for the addition to the commodity's price that the general rate of profit determines. When he sells the commodity, then, the original money capital he advanced to buy it flows back to him along with ΔM.
Here it comes out again that his money capital is nothing but the industrial capitalist's commodity capital turned into money capital, and that this can no more affect the size of the value of that commodity capital than if it had been sold straight to the final consumer instead of to the merchant. In fact it only anticipates the final consumer's payment. But that is correct only if, as we have assumed so far, the merchant has no expenses — that is, if besides the money capital he must advance to buy the commodity from the producer, he has no other capital, circulating or fixed, to advance in the process of the commodity's metamorphosis, in buying and selling. And this is not the case, as we saw when we looked at the costs of circulation (Volume 2, Chapter 6). Some of these circulation costs are costs the merchant has to reclaim from other agents of circulation; others arise directly out of his own specific business.
Whatever kind these circulation costs may be — whether they arise out of the purely commercial business as such, and so belong to the merchant's own specific circulation costs, or whether they are items arising out of later production processes that are inserted into the circulation process, such as dispatch, transport, storage and so on — they always presuppose, on the merchant's side, a capital over and above the money capital advanced to buy the goods: an additional capital advanced to buy and pay for these means of circulation. Where this cost element consists of circulating capital, the whole of it enters what the commodities sell for, as an additional element; where it consists of fixed capital, it goes in according to its wear and tear — but as an element that forms a nominal value even when it forms no real addition to the commodity's value, which is what the purely commercial circulation costs do. Circulating or fixed, though, the whole of this additional capital goes into the formation of the general rate of profit.
The purely commercial circulation costs — leaving aside the costs of dispatch, transport, storage and so on — come down to the costs needed to realize the value of the commodity: to turn it from commodity into money or from money into commodity, to mediate its exchange. Here we leave entirely out of account any production processes that carry on during the act of circulation, and out of which the commercial business can exist quite separately. In fact the actual transport industry and the dispatch of goods can be, and are, branches of industry quite distinct from trade; and the goods that are to be bought and sold may be lying in docks and other public places, with the costs arising from this charged to the merchant by third parties, in so far as he has to advance them. All of this is found in wholesale trade proper, where commercial capital appears in its purest form and least entangled with other functions. A haulier, a railway director, a shipowner — none of these is a 'merchant'. The costs we are looking at here are the costs of buying and of selling. We noted earlier that they come down to reckoning, book-keeping, marketing, correspondence and so on. The constant capital needed for this consists of office, paper, postage and so on. The other costs come down to variable capital advanced for the employment of commercial wage-labourers. (Dispatch charges, transport costs, advances of customs duties and so on can partly be seen as advanced by the merchant in the purchase of the goods, so that they enter into his purchase price.)
All of these costs are incurred not in producing the use-value of the commodities, but in realizing their value. They are pure costs of circulation. They do not enter the immediate production process, but they do enter the circulation process, and with it the overall process of reproduction.
The only part of these costs that concerns us here is the part laid out as variable capital. (Besides this, three things would need investigating: first, how the law that only necessary labour enters the value of a commodity makes itself felt in the circulation process; second, what accumulation looks like for commercial capital; third, how commercial capital functions in society's actual overall process of reproduction.)
These costs arise out of the economic form of the product as a commodity.
If the labour-time the industrial capitalists themselves lose in selling their commodities directly to one another — the same thing, said objectively, as the circulation time of the commodities — adds no value at all to those commodities, then it is clear that this labour-time does not take on a different character simply because it falls to the merchant instead of to the industrial capitalist. Turning commodity (product) into money, and money into commodity (means of production), is a necessary function of industrial capital, and so a necessary operation for the capitalist, who is in fact only personified capital, capital equipped with its own consciousness and will. But these functions do not increase value, and they do not create surplus-value either.
By performing these operations — by continuing to mediate the functions of capital in the sphere of circulation after the productive capitalist has stopped doing so — the merchant simply steps into the industrial capitalist's place. The labour-time these operations cost is spent on necessary operations in the reproduction process of capital, but it adds no value. If the merchant did not carry out these operations (and so did not spend the labour-time they call for), he would not be putting his capital to work as a circulation agent for industrial capital; he would not be carrying on the industrial capitalist's broken-off function, and so would have no claim, as a capitalist and in proportion to the capital he advanced, to a share in the mass of profit produced by the industrial capitalist class.
To take a share in the mass of surplus-value, then — to make his advance work as capital — the commercial capitalist needs to employ no wage-labourers at all. If his business and his capital are small, he may be the only worker he employs. What pays him is his slice of the profit, and that slice comes out of the gap between what he pays for the commodities and their real production price.
On the other hand, where the capital the merchant advances is small, the profit he realizes may be no larger than the wage of a better-paid skilled worker — it can even be smaller. Alongside him, in fact, work the productive capitalist's own direct commercial agents: buyers, sellers, travellers, who draw the same income or more, whether as a wage or as a claim on the profit made on each sale (commission, a percentage). In the first case the merchant collects the commercial profit himself, as an independent capitalist; in the other case the clerk — the industrial capitalist's wage-labourer — is paid a part of the profit, either in the form of a wage or in the form of a share in the profit of the industrial capitalist whose direct agent he is, and in that case his principal pockets both the industrial and the commercial profit. But in all these cases, even though the circulation agent's income may seem to him a plain wage, a payment for the work he has done, and even though, where it does not seem so, the size of his profit may amount to no more than the wage of a better-paid worker — his income springs from commercial profit alone.
This follows because his labour creates no value.
When the circulation operation drags on, it means two things for the industrial capitalist. First, he personally loses time, in so far as he is held back from performing his own function as director of the production process. Second, his product stays longer in the circulation process, in money form or commodity form — a process in which it is not valorized and in which the immediate production process is interrupted. If that is not to be interrupted, then either production has to be cut back, or additional money capital has to be advanced so that the production process keeps going on the same scale. Either way it comes to the same thing: either a smaller profit is made with the capital as it was, or additional money capital has to be advanced in order to make the same profit as before.
Nothing changes here when the merchant steps into the industrial capitalist's place. Instead of the industrial capitalist spending more time in the circulation process, the merchant spends it; instead of the industrial capitalist having to advance extra capital for circulation, the merchant advances it — or, what comes to the same thing, instead of a larger part of the industrial capital forever turning over inside the circulation process, the merchant's capital is penned up in it entirely. And instead of the industrial capitalist making a smaller profit, he now has to give up a part of his profit to the merchant outright.
As long as commercial capital stays inside the limits where it is necessary, the only difference is this: because capital's function is divided in this way, less time is spent exclusively on the circulation process, less extra capital is advanced for it, and the loss in total profit — the loss that shows up in the shape of commercial profit — is smaller than it would otherwise be. In the example above, 720c + 180v + 180s alongside a commercial capital of 100 leaves the industrial capitalist a profit of 162, or 18 per cent, and so brings about a deduction of 18. Without this separation of the function, the extra capital needed might perhaps be 200, and the industrial capitalists' total advance would then be 1,100 instead of 900 — on a surplus-value of 180, a rate of profit of only 16 4/11 per cent.
Suppose an industrial capitalist sells his own goods. He already lays out extra capital to buy new commodities before the product he has out in circulation has turned back into money. If he then lays out still more capital just to turn the value of his commodity capital into money — office costs, and the wages of the commercial employees he hires — that is certainly capital, but it is not surplus-value. It has to be replaced out of the value of the goods, because a piece of that value must be turned back into covering these circulation costs. Even so, no extra surplus-value gets formed.
Look at the total capital of society and it comes to this: a part of that capital is needed for secondary operations that lie outside the valorization process, and this part has to be reproduced over and over for that purpose. For the individual capitalist and for the whole industrial capitalist class, the profit rate is lowered by it — a result that follows from every addition of extra capital, as far as the addition is needed to set the same mass of variable capital in motion.
If the merchant now takes these extra costs of the circulation business off the industrial capitalist's hands, the profit rate still falls — only less, and by another route. What happens now is that the merchant advances more capital than he would need if these costs were not there, and the profit on that extra capital pushes the total commercial profit up. So commercial capital enters the equalization of the average rate of profit on a larger scale alongside industrial capital, and average profit drops.
Take our earlier example. If, besides the 100 of commercial capital, a further 50 were advanced for these costs, the 180 of surplus-value would now be split over a productive capital of 900 plus a commercial capital of 150, together 1,050. The average rate of profit therefore falls to 17 1/7 per cent. The industrialist lets the merchant have the goods for 900 + 154 2/7 = 1,054 2/7, and the merchant sells them for 1,130 (1,080 + 50 to cover the costs he has to replace again).
It must also be assumed that the split between commercial and industrial capital goes together with a centralization of trading costs, and so with a reduction in them.
The question now is how things stand with the commercial wage-labourers the merchant capitalist employs — here, the dealer in commodities.
From one side, a commercial worker of this kind is a wage-labourer like any other. First, in so far as his labour is bought out of the merchant's variable capital, not out of money the merchant spends as revenue — so it is bought not for a personal service, but so that the capital advanced in him valorizes itself. Second, in so far as what his labour-power is worth — and hence what he is paid — is fixed, as with all other wage-labourers, by what it costs to produce and reproduce that particular labour-power, and not by what his work turns out.
But the same difference has to hold between him and the workers industrial capital employs directly as holds between industrial capital and commercial capital, and so between the industrial capitalist and the merchant. The merchant is nothing but an agent of circulation, and an agent of circulation produces no value and no surplus-value. (The extra value he adds to goods through his expenses comes down to an addition of value that already existed — though here the question forces itself on us: how does he keep and conserve this value of his constant capital?) So the commercial workers he has doing those same jobs cannot possibly create surplus-value for him directly either.
Here, as with the productive workers, we assume that the wage is fixed by the value of labour-power — that the merchant does not get rich by a deduction from the wage, so that in working out his costs he does not enter an advance for labour he has only partly paid for; in other words, that he does not get rich by cheating his clerks and so on.
What causes trouble here is not at all the question of how these commercial wage-workers directly produce profit for the man who employs them, even though they do not directly produce surplus-value — and profit is only surplus-value in changed form. That question has in fact already been answered by the general analysis of commercial profit.
Industrial capital makes its profit by selling labour that is contained in the goods and realized there, labour it has paid no equivalent for. Commercial capital gets its profit on the same principle — paying no equivalent for what it takes: it does not pay productive capital in full for the unpaid labour contained in the goods (in the goods, in so far as the capital laid out to produce them counts as an aliquot part of the total industrial capital), and then, when it sells the goods, it has that unpaid part, still contained in them, paid over to itself.
Commercial capital's relation to surplus-value is not the same as industrial capital's. Industrial capital makes surplus-value by appropriating other people's unpaid labour directly. Commercial capital appropriates a part of that surplus-value by having that part transferred to it from industrial capital.
Only because its business is realizing values does commercial capital count, within reproduction, as capital at all — and so, as a capital doing its job, it is from the surplus-value produced by the total capital that it draws its share. How big a profit an individual merchant makes depends on how much capital he can put to work in this process, and he can put more of it into buying and selling the greater the unpaid labour of his clerks. The function that makes his money capital, the merchant himself leaves his employees to perform in large part.
His clerks' labour goes unpaid, and unpaid labour of this kind creates no surplus-value. But it does create his power to appropriate surplus-value — which, for this capital, comes out to exactly the same thing. So for this capital it is a source of profit. Without it, a commercial business could never be run on a large scale, never in a capitalist way.
The unpaid labour of the worker gives productive capital surplus-value directly. In the same way, the unpaid labour of commercial wage-workers gives commercial capital a share in that surplus-value.
The difficulty is this. The merchant's own labour-time and labour create no value, even though it wins him a share of surplus-value that has already been produced. So where does that leave the variable capital he lays out in buying commercial labour-power?
Does this variable capital count as part of his outlay, as part of the capital he has advanced? If it does not, that seems to clash with the law by which profit rates are equalized — what capitalist would put up 150 if only 100 of it counted as capital he had advanced? If it does, then it seems to clash with what commercial capital essentially is. For this kind of capital does not work as capital by setting other people's labour in motion, the way industrial capital does. It works by doing the job itself — by carrying out the buying and selling itself — and it is precisely, and only, for that reason and through that means that it shifts onto itself a piece of the surplus-value that industrial capital has produced.
(So the points still to be investigated are these: the merchant's variable capital; the law of necessary labour within circulation; how the merchant's work keeps the value of his constant capital; commercial capital's part in reproduction taken as a whole; and finally the doubling — into commodity capital and money capital on the one hand, and into commercial capital and money-dealing capital on the other.)
Suppose each merchant had only as much capital as he could personally turn over by working himself. Commercial capital would then split into endlessly many tiny pieces, and that splitting would have to grow as capitalist production advanced and productive capital turned out goods on an ever bigger scale and moved ever bigger quantities. So the mismatch between the two would keep growing. And to the same degree that capital concentrated in production, it decentralized in circulation.
The industrial capitalist's purely commercial business — and with it his purely commercial outlays — would then expand without limit, since instead of dealing with a hundred merchants he would have to deal with a thousand. Much of what is gained by commercial capital standing on its own would be lost; and on top of the purely commercial costs, the other costs of circulation would grow too — sorting, dispatch and the like. So much for industrial capital.
Now commercial capital. First the purely commercial work. Working with big numbers takes no more time than working with small ones. Ten purchases of £100 take ten times as long as a single purchase of £1,000. Writing to ten small merchants costs ten times the correspondence, the paper and the postage of writing to one big merchant. A limited division of labour in the commercial office — one keeping the books, another the cash, a third the letters, one buying, one selling, one travelling — saves labour-time on an enormous scale, so that the number of commercial workers employed in wholesale trade bears no relation at all to the comparative size of the business. The reason is that in commerce — far more than in industry — a given function eats up the same labour-time whether the scale is small or large. That is also why, historically, merchants' businesses concentrated before industrial workshops did. Then there is the outlay on constant capital. A hundred small offices cost infinitely more than one large office, a hundred small warehouses more than one large one. And transport costs, which the merchant's business carries at least as money advanced, rise along with all this fragmentation.
The commercial side of the industrial capitalist's own business would swallow more labour and more circulation costs. The same commercial capital, divided among many small merchants, would — because of that fragmentation — need far more workers to do its work; and on top of that, the same commodity capital would call for more commercial capital to turn it over.
Let us call the whole of the commercial capital that is directly laid out on buying and selling commodities B, and the variable capital laid out on paying the commercial assistants b. Then B + b is smaller than the total commercial capital B would have to be if every merchant struggled along without help — that is, if no part of it were laid out as b. But we are still not through with the difficulty.
The price at which the goods are sold has to be enough, first, to pay the average profit on B + b. That is already accounted for by the fact that B + b is a shortening of the original B — a smaller commercial capital than would be necessary without b. But this sale price must also be enough, second, to replace what he paid out in wages, his variable capital, b, itself — over and above the profit on b that now appears on top. This second thing is what makes the difficulty.
Is b a new piece of the price? Or is it only part of the profit made on B + b, which looks like wages from the commercial worker's side and like nothing but the return of his variable capital from the merchant's own side?
On that second reading, the merchant's profit on the capital B + b that he advanced would come to no more than the profit falling on B at the general rate, plus b — b being paid out as wages, but yielding no profit of its own.
In fact it comes down to finding the limits of b (in the mathematical sense). First let us pin the difficulty down exactly. Let B be the capital put straight into buying and selling goods, K the constant capital used up in that function (the material office expenses), and b the variable capital the merchant advances.
Getting B back raises no difficulty at all. To the merchant it is simply the purchase price he has realized — or, for the manufacturer, the price of production. The merchant pays that price, and when he sells again he gets B back as part of his sale price, and along with this B the profit on B, as explained earlier. Say the commodity costs £100 and the profit on it is 10 per cent: it is then sold at 110. The commodity already cost 100 beforehand; the commercial capital of 100 only adds 10 to it.
Now take K. At most it is as big as — in fact it is smaller than — the constant capital the producer himself would use up in his own buying and selling; and that would come on top of the constant capital he needs directly for production. All the same, this part too must always come back out of the price of the commodity; or, which comes to the same, a matching slice of the commodity must always be spent, and reproduced, in that form — taking society's total capital as a whole. This slice of the capital advanced would press down on the rate of profit just as much as the whole mass of constant capital invested directly in production does. In so far as the industrial capitalist hands the commercial side of his business over to the merchant, he does not have to advance this part of the capital. The merchant advances it instead. But that advance is really only nominal: the merchant does not produce the constant capital he uses up, nor does he reproduce it (his material office expenses). Their production therefore appears as a business of its own — or at least as one branch of certain industrial capitalists' business — so that these play the same role as those who supply constant capital to the people producing means of subsistence. So the merchant gets, first, this replacement, and second, the profit on it. Both cut into the industrial capitalist's profit. But thanks to the concentration and the economy that come with the division of labour, the cut is smaller than it would have been had he advanced this capital himself. The rate of profit falls by less, because less capital is advanced this way.
So far, then, the sale price comes to B + K plus the profit on the two of them. As far as we have got, this part gives no trouble. But now b enters — the variable capital the merchant has advanced.
The sale price thereby becomes B + K + b, plus the profit on B + K, plus the profit on b.
B does nothing but replace the purchase price; besides the profit on B it adds nothing to that price. K adds not only the profit on K but K itself. But K plus the profit on K — the part of the circulation costs advanced as constant capital, plus the corresponding average profit — would be bigger in the industrial capitalist's hands than in the commercial capitalist's. The drop in average profit shows up in this form: the industrial capital advanced, less B + K, is credited with the full average profit; what the average profit loses on account of B + K goes to the merchant instead, and so that deduction appears as the profit of a capital of its own — commercial capital.
But b plus the profit on b is another matter — or, in the case assumed here, since the rate of profit is taken as 10 per cent, b + 1/10 b. And here is the real difficulty.
With b the merchant buys nothing but commercial labour, on our assumption — the labour needed for the circulation of capital, C–M and M–C. But commercial labour is simply the labour that has to be done anyway for a capital to operate as commercial capital, the labour that turns commodities into money and money back into commodities. Such labour realizes values; it does not create them. And only in so far as a capital carries out these functions — that is, only in so far as a capitalist performs these operations, this labour, with his capital — does it work as commercial capital and have a hand in settling the general rate of profit, drawing its dividend out of the total profit.
In (b + the profit on b), though, it seems that first the labour gets paid (it makes no difference whether the industrial capitalist pays the merchant for his own labour or for that of a clerk the merchant pays), and second that the profit gets paid on the payment for that labour — labour the merchant himself would have had to do. So commercial capital receives both: the repayment of b, and the profit on it. And this comes about because it is paid twice over for one thing — once for the labour that lets it work as commercial capital, and once again in profit, for being a capital that functions, which is to say for doing that very labour. This, then, is the question that has to be solved.
Suppose B = 100, b = 10 and the rate of profit 10 per cent. We set K = 0, so as not to drag into the reckoning once more an item of the purchase price that has already been settled and does not belong here. The sale price would then be B + p + b + p (= B + Bp' + b + bp', where p' is the rate of profit) = 100 + 10 + 10 + 1 = 121.
But if the merchant had not laid b out as wages — b is paid only for commercial labour, for labour needed to realize the value of the commodity capital that industrial capital throws onto the market — then things would stand like this. Buying or selling B = 100 takes the merchant's time, and we shall take it that this is all the time he has. The commercial labour represented by b, or 10, if it were paid out of profit instead of as wages, would presuppose another commercial capital of 100, since 10 per cent of that comes to b, which is 10. This second B = 100 would not be added to the commodity's price, but the 10 per cent would be. So there would be two operations of 100, making 200 — buying commodities for 200 + 20 = 220.
Commercial capital is absolutely nothing but one part of the industrial capital that operates in the circulation process, in a form made independent. So every question about it has to be answered by first putting the problem in the shape it has where the phenomena peculiar to commercial capital are not yet independent, but still directly tied to industrial capital, as a branch of it. Commercial capital works in the circulation process all the time, with an office instead of a workshop. So the b we are asking about here has to be looked at first where it appears in that form — in the industrial capitalist's own office.
To begin with, this office is always vanishingly small next to the industrial workshop. For the rest it is clear: as the scale of production widens, the commercial operations that industrial capital's circulation constantly requires multiply — both selling the product that exists as commodity capital, and turning the money set free back into means of production, and keeping the accounts for the whole of it. Working out prices, keeping the books, handling the cash, writing letters — all of it belongs here. The more developed the scale of production, the greater the industrial capital's commercial operations — though by no means in proportion — and so the greater the labour and the other circulation costs that realizing value and surplus-value requires. This makes commercial wage-labourers necessary, and they make up the actual office. The outlay on them, though made in the form of wages, is a different thing from the variable capital that buys productive labour. It raises the industrial capitalist's outlays, the mass of capital he must advance, without directly adding to surplus-value. For this is money paid for labour used only to realize values that have already been created. Like any other outlay of this kind, it brings the rate of profit down, because the capital advanced grows while the surplus-value does not. If the surplus-value s stays constant while the capital advanced C grows by ΔC, then the profit rate s/C is replaced by the smaller profit rate s/(C + ΔC).
So the industrial capitalist keeps these circulation costs down as far as he can, exactly as he keeps down what he lays out on constant capital. Industrial capital therefore does not treat its commercial employees the way it treats its productive wage-labourers. The more of the latter it employs, other circumstances staying equal, the more massive the production and the greater the surplus-value or profit. Here it is the other way round. The wider the scale of production, and the greater the value, and so the surplus-value, to be realized — hence the bigger the commodity capital produced — the more do office costs grow in absolute size, though not in proportion, and the more they give occasion for a particular division of labour. How much profit is the precondition for these outlays shows itself, among other things, in the fact that as commercial salaries rise, part of the salary is often paid out as a cut of the profit, fixed at so many per cent.
It belongs to the nature of such labour that it consists only in mediating operations — part of it working out values, part realizing them, part turning the realized money back into means of production — and that its scope therefore depends on how big the values produced and to be realized are. Labour of this kind is not the cause of those magnitudes and masses; it is their consequence — unlike directly productive labour. The same goes for the other circulation costs. Where there is much to weigh, to measure, to pack and to haul, the stuff must be there in the first place. How much packing and transport work there is depends on the mass of commodities those workers handle, and never the reverse.
The commercial worker does not directly produce surplus-value. But what his labour costs is fixed by the value of his labour-power, that is, by what it costs to produce that labour-power; while the actual exercise of it — the straining, the spending of energy, the wear and tear — is no more bounded by that value than is the case with any other wage-worker. So his wage need not bear any fixed relation to how much profit he helps the capitalist realize. His cost to the capitalist and his yield to the capitalist are two different magnitudes. What he brings in comes not from creating surplus-value directly, but from helping to cut the cost of realizing surplus-value, in so far as he performs labour — part of it unpaid.
A proper commercial worker belongs among the better-paid wage-workers — one of those whose labour is skilled, above the average. But with the advance of the capitalist mode of production his wage has a tendency to fall, even measured against average labour. First, through the division of labour inside the office: only a one-sided development of the worker's capacity has to be produced, and part of what it costs to produce that capacity costs the capitalist nothing, since the worker's skill develops through the work itself, and all the faster the more one-sided it becomes with the division of labour. Second, because the groundwork — commercial know-how, languages and so on — is reproduced ever faster, more easily, more widely and more cheaply as the capitalist mode of production bends teaching methods towards the practical. The spread of popular education allows this kind of labour to be recruited from classes that used to be shut out of it and were used to a worse way of living. It also swells the supply, and with it the competition. With a few exceptions, then, the labour-power of these people loses value as capitalist production advances; their wages fall while their ability to work rises. The capitalist takes on more of these workers when he has more value and profit to realize. The growth of this labour is always a result of the growth of surplus-value, never its cause.
So a duplication takes place. On one side, the functions of commodity capital and money capital — and so, more precisely, those of commercial capital — are general forms that industrial capital necessarily takes. On the other side, particular capitals, and with them particular sets of capitalists, are occupied exclusively with these functions, and the functions accordingly turn into separate spheres in which capital is valorized.
Only under commercial capital do the commercial functions and the circulation costs take on a life of their own. The side of industrial capital that faces circulation exists not just in its steady being as commodity capital and money capital, but in the office beside the workshop as well. But it becomes autonomous for commercial capital. For commercial capital, the office is its only workshop. The share of capital used in the form of circulation costs looks much larger in the wholesale merchant's hands than in the industrialist's, because the capital that the class of industrial capitalists as a whole would have to use this way is now gathered into the hands of a few merchants — over and above the office that goes with every industrial workshop. Having taken on the circulation functions, they also shoulder the circulation costs that come with them.
To industrial capital, circulation costs appear as an expense, and an expense they are. To the merchant, those same costs appear as his profit's source — and, on the assumption of a general rate of profit, that profit is proportionate to how big those costs are. So for commercial capital, the money laid out on these circulation costs is a productive investment. And so the commercial labour it buys is, for it, directly productive too.
Volume 2 showed that capital's pure functions in the circulation sphere create no value and no surplus-value. These jobs are the ones the industrial capitalist has to do to get the value of his commodities realized, and then to turn that value back into the elements of production for those commodities — the jobs that mediate the changes of form of commodity capital, C'–M–C, which is to say the acts of selling and buying. And it also came out the other way round: the time these jobs take sets limits on the formation of value and surplus-value — limits that hold objectively for the commodities and subjectively for the capitalist.
None of that is changed in any way by part of this commodity capital taking on the shape of commercial capital, or by the jobs that mediate the change of form of commodity capital appearing as the special business of a special section of capitalists, or as the exclusive function of one portion of the money capital.
The change of form of commodity capital, C'–M–C, comes down to the selling and buying of commodities. If these are not operations that create value or surplus-value when the industrial capitalists themselves perform them, then they cannot possibly become such operations by being performed by other people instead. The same holds for the portion of society's total capital that must always be available as money capital so that the reproduction process is not interrupted by the circulation process but runs on continuously: if this money capital creates no value and no surplus-value, it cannot acquire those properties from being constantly thrown into circulation by another section of capitalists, doing the same jobs, instead of by the industrial capitalist.
How far commercial capital can be indirectly productive has already been hinted at, and will be gone into further later on.
Take commercial capital and strip off every other function that may be attached to it — storing goods, forwarding them, transporting them, sorting them, splitting them up into small lots. What is left is its true function: buying things in order to sell them again. Doing that, it creates neither value nor surplus-value. It only mediates their realization, and with that the actual exchange of commodities — their move out of one pair of hands into another, society's metabolism.
All the same, the circulation phase of industrial capital is a phase of the reproduction process just as much as production is. So capital that operates independently in the circulation process must yield the annual average profit just as much as capital at work in the different branches of production. If commercial capital brought in a higher average profit, in percentage terms, than industrial capital, part of industrial capital would turn into commercial capital. If it brought in a lower one, the opposite process would take place: part of commercial capital would turn into industrial capital. No kind of capital finds it easier to change its calling, its function, than commercial capital.
Since commercial capital itself produces no surplus-value, it is clear that the surplus-value falling to it in the form of average profit is only a piece of the surplus-value that productive capital as a whole produces. But now the question is: how does commercial capital pull in the share that falls to it of the surplus-value, or profit, that productive capital produces?
It is only a semblance that commercial profit consists of a mere supplement, a nominal raising of prices above the value of the commodities.
It is clear that the merchant can get his profit from nowhere but the price of the commodities he sells — and further, that the profit he makes when he sells them must be equal to the difference between what he pays for them and what he sells them for, the amount by which the sale price is above the purchase price.
It can happen that extra costs — circulation costs — go into the commodity after it is bought and before it is sold; and it can just as well not happen. If such costs do go in, then clearly the amount by which the sale price exceeds the purchase price is not all profit. To keep the investigation simple, we assume for the time being that no such costs come in.
With the industrial capitalist, the gap between what he pays for his commodities and what he sells them for is the gap between a commodity's cost price and its price of production — or, if we look at the total social capital, the gap between the value of the commodities and their cost price to the capitalists, which in turn comes down to the difference between all the labour embodied in them and the part of it that was paid for. Before the commodities an industrial capitalist has bought are thrown back on the market as things to be sold, they go through the production process, and only there is the part of their price that is later realized as profit actually produced.
With the commodity dealer it is otherwise. The commodities stay in his hands only for as long as their circulation process lasts. He merely carries on the sale the productive capitalist started, the turning of their price into money, and so he never puts them through any process in between, where they might soak up fresh surplus-value. The industrial capitalist, in circulation, only realizes a surplus-value or profit that production has already put there; the merchant, by contrast, is supposed not merely to realize his profit in circulation and through it, but to make it there in the first place. And that seems possible only if he sells above their production prices the goods the industrial capitalist sold to him at those production prices — or, taking the commodity capital as a whole, at their values; that is, if he tacks a nominal addition on to their prices and, taking the commodity capital as a whole, sells it above its value and pockets the difference between its nominal value and its real value — in a word, sells it dearer than it is.
The way this addition works is very easy to understand. Take a yard of linen at 2 shillings. If I want to make 10 per cent profit on reselling it, I have to put one-tenth on the price, and so I sell the yard at 2s. 2⅖d. The difference between its real price of production and its sale price is then 2⅖d., and on 2 shillings that is a profit of 10 per cent.
In fact I am then selling the buyer the yard at a price that is in truth the price of 1 1/10 yards. Or, what comes to the same thing: it is exactly as if the buyer got only 10/11 of a yard for his 2 shillings, and I kept 1/11 of a yard for myself. With the 2⅖d. I can in fact buy 1/11 of a yard back, reckoning the yard at 2s. 2⅖d.
So this would be nothing but a roundabout way of getting a share of the surplus-value and of the surplus product, by nominally raising the prices of commodities.
This is commercial profit realized through an addition to the price of commodities, the way it first presents itself. And in fact the whole notion that profit springs from a nominal raising of commodity prices, or from selling commodities above their value, has its origin in the outlook of commercial capital.
Looked at more closely, however, it soon turns out that this is a pure illusion — and that, if we assume the capitalist mode of production to be the dominant one, commercial profit is not realized in this way. (What is at issue here is always the average, never individual cases.)
Why do we take it that the only way the merchant can realize, say, 10 per cent on his commodities is to sell them 10 per cent above their prices of production? Because we assumed that the producer of these commodities — the industrial capitalist, who, being the personification of industrial capital, always figures to the outside world as 'the producer' — had sold them to the merchant at their price of production. If the purchase prices the merchant pays are equal to the prices of production of the commodities, and in the last analysis to their values, so that the price of production, and in the last analysis the value, is the cost price for the merchant, then what his sale price gains over what he paid for the goods — and that difference alone is where his profit comes from — has to be a gain of their commercial price over their production price, and in the last analysis the merchant sells every commodity above its value.
But why did we assume that it is at their prices of production that the industrial capitalist sells to the merchant? Or rather, what was presupposed in that assumption? That merchant's capital — and here we are dealing with it still only as commercial, commodity-dealing capital — does not enter into how the general rate of profit is formed. We had to start from this presupposition in presenting the general rate of profit: first, because merchant's capital as such did not yet exist for us at that point; and second, because the average profit, and with it the general rate of profit, first had to be developed by equalizing the profits, or surplus-values, that the industrial capitals in the different spheres of production actually produce.
With merchant's capital, by contrast, we have a capital that gets a share of the profit without having any part in producing it. So the earlier presentation now needs supplementing.
Say the total industrial capital advanced over the year is 720c + 180v = 900 (say, millions of pounds sterling), and s′ = 100 per cent. The product is then 720c + 180v + 180s. Call this product — the commodity capital produced — C. Its value, or price of production, comes to 1,080 (taking all commodities together, the two are one and the same), and the 900 yields a profit of 20 per cent. That 20 per cent is the average rate of profit, as worked out earlier: here the surplus-value is reckoned on the whole industrial capital at its average composition, not on any one capital of a particular make-up. So C is 1,080, and profit stands at 20 per cent.
Now suppose that alongside this 900 of industrial capital there is also 100 of commercial capital, which gets the same share of the profit in proportion to its size. On our assumption it is a tenth of the total capital of 1,000. So it takes a tenth of that 180 of surplus-value, and gets a profit at a rate of 18 per cent. In fact, then, the profit still to be shared among the other nine tenths of the total capital is only 162 — which is likewise 18 per cent on the 900.
So the owners of the 900 of industrial capital sell C to the merchants at 720c + 180v + 162s = 1,062. And when the merchant adds the 18 per cent average profit to his 100, he sells them at 1,062 + 18 = 1,080. That is their price of production — or, if we look at the commodity capital as a whole, their value. And yet he makes his profit only in circulation and through it, and only out of the excess of his sale price over what he paid. All the same, he is not selling above the commodities' value, not above their price of production — simply because the price the industrial capitalists sold to him at was below that value, below that price of production.
Commercial capital therefore helps fix the general rate of profit, in proportion to the share it makes up of the total capital. Say the average rate of profit in our case is 18 per cent. Take the commercial capital away — that tenth of the whole — and it would stand at 20, for that tenth is what pulls the general rate down by a tenth. This also brings in a closer, more restrictive definition of the price of production.
Price of production still means the price of the commodity = its cost (that is, what the constant and variable capital in it are worth) + the average profit on that. But this average profit is now fixed on a different basis. What fixes it is the total profit that the whole productive capital produces — but the reckoning is not made on that productive capital by itself. If it were, then with 900 as above and a profit of 180 the average rate of profit would come out at 180/900 = 20 per cent. The reckoning is made on productive and commercial capital together: so with 900 productive and 100 commercial, the average rate of profit works out at 180/1,000 = 18 per cent. So the price of production comes to k (the cost) + 18, not k + 20.
The slice of the total profit that falls to commercial capital is already counted into the average rate of profit. The real value — the real price of production — of the whole commodity capital is therefore k + p + m, where k is the cost, p the average profit, and m the commercial profit. So the price of production — what the industrial capitalist gets when he sells as an industrial capitalist — is smaller than the commodity's real price of production. Or, looking at all commodities together: the industrial capitalist class sells them for less than their value. In our case: 900 (the cost) plus 18 per cent of 900, or 900 + 162 = 1,062.
And now: in selling for 118 goods that cost him 100, the merchant does add 18 per cent; but since the goods he bought for 100 are worth 118, he is not on that account selling them above their value. We keep the expression 'price of production' in the stricter sense set out above. It is then clear that the industrial capitalist's profit is what the commodity's production price has over its cost price, and that — unlike this industrial profit — commercial profit is what the sale price has over that production price, which for the merchant is what he paid for the goods; while the real price of the commodity = its production price + the commercial profit. Industrial capital only realizes a profit that is already there in the commodity's value as surplus-value; commercial capital likewise only realizes one, and can do so only because when industrial capital sells, its price leaves part of the surplus-value — part of the total profit — not yet realized. So the merchant sells for more than he paid — but the reason is not that his selling price is above the goods' total value; it is that the price he paid was below it.
So commercial capital has a hand in equalizing surplus-value into average profit, though no hand in producing that surplus-value. Which means the general rate of profit already has inside it the deduction from surplus-value that belongs to commercial capital — a slice taken out of the industrial capitalist's profit.
From what has come before, two things follow.
First: the bigger commercial capital is in comparison with industrial capital, the lower industrial capital's rate of profit — and the smaller it is by comparison, the higher that rate.
Second: Part One showed that the rate of profit always comes out as a smaller figure than the rate of surplus-value actually is — that is, it always understates how much labour is being exploited. In the case above, for instance, 720c + 180v + 180s gives a rate of surplus-value of 100 per cent, but it shows up as a rate of profit of only 20 per cent.
That gap widens still further in so far as the average rate of profit itself — once the share falling to commercial capital is reckoned in — appears smaller again: here 18 per cent instead of 20 per cent. So the average rate of profit for the capitalist who does the exploiting directly makes the rate of profit look smaller than it really is.
Assuming that all other circumstances stay the same, commercial capital's relative size (retail traders, a hybrid species, are the exception here) moves inversely to the speed of its turnover, and so inversely to the vigour of the reproduction process as a whole.
When we follow the analysis, the general rate of profit appears to form first among the industrial capitals and their competition, and to be rectified, supplemented and modified by the intervention of commercial capital only later on. Historically, matters stand exactly the other way round. It is commercial capital that first fixes the prices of commodities, more or less by their values, and it is the sphere of circulation — the sphere that mediates the reproduction process — where a general rate of profit first takes shape. Commercial profit is what originally determines industrial profit. Only once the capitalist mode of production has pushed through, and the producer has himself become a merchant, is commercial profit cut down to its aliquot share: the portion of the total surplus-value that belongs to commercial capital as an aliquot part of the total capital taking part in society's process of reproduction.
In the supplementary equalization of profits that commercial capital's intervention brings about, two things came out. Nothing extra enters the value of the commodity on account of the money capital the merchant advances. And the addition to the price — the addition by which the merchant makes his profit — is exactly equal to the piece of the commodity's value that productive capital left out, never reckoned in, when it set the commodity's production price.
His money capital behaves just like the industrial capitalist's fixed capital: so far as it is not used up, its value forms no element of the commodity's value. In the price he pays for the commodity capital, he replaces that capital's production price with money: production price = M. His sale price, as worked out earlier, is M + ΔM, where ΔM stands for the addition to the commodity's price that the general rate of profit determines. When he sells the commodity, then, the original money capital he advanced to buy it flows back to him along with ΔM.
Here it comes out again that his money capital is nothing but the industrial capitalist's commodity capital turned into money capital, and that this can no more affect the size of the value of that commodity capital than if it had been sold straight to the final consumer instead of to the merchant. In fact it only anticipates the final consumer's payment. But that is correct only if, as we have assumed so far, the merchant has no expenses — that is, if besides the money capital he must advance to buy the commodity from the producer, he has no other capital, circulating or fixed, to advance in the process of the commodity's metamorphosis, in buying and selling. And this is not the case, as we saw when we looked at the costs of circulation (Volume 2, Chapter 6). Some of these circulation costs are costs the merchant has to reclaim from other agents of circulation; others arise directly out of his own specific business.
Whatever kind these circulation costs may be — whether they arise out of the purely commercial business as such, and so belong to the merchant's own specific circulation costs, or whether they are items arising out of later production processes that are inserted into the circulation process, such as dispatch, transport, storage and so on — they always presuppose, on the merchant's side, a capital over and above the money capital advanced to buy the goods: an additional capital advanced to buy and pay for these means of circulation. Where this cost element consists of circulating capital, the whole of it enters what the commodities sell for, as an additional element; where it consists of fixed capital, it goes in according to its wear and tear — but as an element that forms a nominal value even when it forms no real addition to the commodity's value, which is what the purely commercial circulation costs do. Circulating or fixed, though, the whole of this additional capital goes into the formation of the general rate of profit.
The purely commercial circulation costs — leaving aside the costs of dispatch, transport, storage and so on — come down to the costs needed to realize the value of the commodity: to turn it from commodity into money or from money into commodity, to mediate its exchange. Here we leave entirely out of account any production processes that carry on during the act of circulation, and out of which the commercial business can exist quite separately. In fact the actual transport industry and the dispatch of goods can be, and are, branches of industry quite distinct from trade; and the goods that are to be bought and sold may be lying in docks and other public places, with the costs arising from this charged to the merchant by third parties, in so far as he has to advance them. All of this is found in wholesale trade proper, where commercial capital appears in its purest form and least entangled with other functions. A haulier, a railway director, a shipowner — none of these is a 'merchant'. The costs we are looking at here are the costs of buying and of selling. We noted earlier that they come down to reckoning, book-keeping, marketing, correspondence and so on. The constant capital needed for this consists of office, paper, postage and so on. The other costs come down to variable capital advanced for the employment of commercial wage-labourers. (Dispatch charges, transport costs, advances of customs duties and so on can partly be seen as advanced by the merchant in the purchase of the goods, so that they enter into his purchase price.)
All of these costs are incurred not in producing the use-value of the commodities, but in realizing their value. They are pure costs of circulation. They do not enter the immediate production process, but they do enter the circulation process, and with it the overall process of reproduction.
The only part of these costs that concerns us here is the part laid out as variable capital. (Besides this, three things would need investigating: first, how the law that only necessary labour enters the value of a commodity makes itself felt in the circulation process; second, what accumulation looks like for commercial capital; third, how commercial capital functions in society's actual overall process of reproduction.)
These costs arise out of the economic form of the product as a commodity.
If the labour-time the industrial capitalists themselves lose in selling their commodities directly to one another — the same thing, said objectively, as the circulation time of the commodities — adds no value at all to those commodities, then it is clear that this labour-time does not take on a different character simply because it falls to the merchant instead of to the industrial capitalist. Turning commodity (product) into money, and money into commodity (means of production), is a necessary function of industrial capital, and so a necessary operation for the capitalist, who is in fact only personified capital, capital equipped with its own consciousness and will. But these functions do not increase value, and they do not create surplus-value either.
By performing these operations — by continuing to mediate the functions of capital in the sphere of circulation after the productive capitalist has stopped doing so — the merchant simply steps into the industrial capitalist's place. The labour-time these operations cost is spent on necessary operations in the reproduction process of capital, but it adds no value. If the merchant did not carry out these operations (and so did not spend the labour-time they call for), he would not be putting his capital to work as a circulation agent for industrial capital; he would not be carrying on the industrial capitalist's broken-off function, and so would have no claim, as a capitalist and in proportion to the capital he advanced, to a share in the mass of profit produced by the industrial capitalist class.
To take a share in the mass of surplus-value, then — to make his advance work as capital — the commercial capitalist needs to employ no wage-labourers at all. If his business and his capital are small, he may be the only worker he employs. What pays him is his slice of the profit, and that slice comes out of the gap between what he pays for the commodities and their real production price.
On the other hand, where the capital the merchant advances is small, the profit he realizes may be no larger than the wage of a better-paid skilled worker — it can even be smaller. Alongside him, in fact, work the productive capitalist's own direct commercial agents: buyers, sellers, travellers, who draw the same income or more, whether as a wage or as a claim on the profit made on each sale (commission, a percentage). In the first case the merchant collects the commercial profit himself, as an independent capitalist; in the other case the clerk — the industrial capitalist's wage-labourer — is paid a part of the profit, either in the form of a wage or in the form of a share in the profit of the industrial capitalist whose direct agent he is, and in that case his principal pockets both the industrial and the commercial profit. But in all these cases, even though the circulation agent's income may seem to him a plain wage, a payment for the work he has done, and even though, where it does not seem so, the size of his profit may amount to no more than the wage of a better-paid worker — his income springs from commercial profit alone.
This follows because his labour creates no value.
When the circulation operation drags on, it means two things for the industrial capitalist. First, he personally loses time, in so far as he is held back from performing his own function as director of the production process. Second, his product stays longer in the circulation process, in money form or commodity form — a process in which it is not valorized and in which the immediate production process is interrupted. If that is not to be interrupted, then either production has to be cut back, or additional money capital has to be advanced so that the production process keeps going on the same scale. Either way it comes to the same thing: either a smaller profit is made with the capital as it was, or additional money capital has to be advanced in order to make the same profit as before.
Nothing changes here when the merchant steps into the industrial capitalist's place. Instead of the industrial capitalist spending more time in the circulation process, the merchant spends it; instead of the industrial capitalist having to advance extra capital for circulation, the merchant advances it — or, what comes to the same thing, instead of a larger part of the industrial capital forever turning over inside the circulation process, the merchant's capital is penned up in it entirely. And instead of the industrial capitalist making a smaller profit, he now has to give up a part of his profit to the merchant outright.
As long as commercial capital stays inside the limits where it is necessary, the only difference is this: because capital's function is divided in this way, less time is spent exclusively on the circulation process, less extra capital is advanced for it, and the loss in total profit — the loss that shows up in the shape of commercial profit — is smaller than it would otherwise be. In the example above, 720c + 180v + 180s alongside a commercial capital of 100 leaves the industrial capitalist a profit of 162, or 18 per cent, and so brings about a deduction of 18. Without this separation of the function, the extra capital needed might perhaps be 200, and the industrial capitalists' total advance would then be 1,100 instead of 900 — on a surplus-value of 180, a rate of profit of only 16 4/11 per cent.
Suppose an industrial capitalist sells his own goods. He already lays out extra capital to buy new commodities before the product he has out in circulation has turned back into money. If he then lays out still more capital just to turn the value of his commodity capital into money — office costs, and the wages of the commercial employees he hires — that is certainly capital, but it is not surplus-value. It has to be replaced out of the value of the goods, because a piece of that value must be turned back into covering these circulation costs. Even so, no extra surplus-value gets formed.
Look at the total capital of society and it comes to this: a part of that capital is needed for secondary operations that lie outside the valorization process, and this part has to be reproduced over and over for that purpose. For the individual capitalist and for the whole industrial capitalist class, the profit rate is lowered by it — a result that follows from every addition of extra capital, as far as the addition is needed to set the same mass of variable capital in motion.
If the merchant now takes these extra costs of the circulation business off the industrial capitalist's hands, the profit rate still falls — only less, and by another route. What happens now is that the merchant advances more capital than he would need if these costs were not there, and the profit on that extra capital pushes the total commercial profit up. So commercial capital enters the equalization of the average rate of profit on a larger scale alongside industrial capital, and average profit drops.
Take our earlier example. If, besides the 100 of commercial capital, a further 50 were advanced for these costs, the 180 of surplus-value would now be split over a productive capital of 900 plus a commercial capital of 150, together 1,050. The average rate of profit therefore falls to 17 1/7 per cent. The industrialist lets the merchant have the goods for 900 + 154 2/7 = 1,054 2/7, and the merchant sells them for 1,130 (1,080 + 50 to cover the costs he has to replace again).
It must also be assumed that the split between commercial and industrial capital goes together with a centralization of trading costs, and so with a reduction in them.
The question now is how things stand with the commercial wage-labourers the merchant capitalist employs — here, the dealer in commodities.
From one side, a commercial worker of this kind is a wage-labourer like any other. First, in so far as his labour is bought out of the merchant's variable capital, not out of money the merchant spends as revenue — so it is bought not for a personal service, but so that the capital advanced in him valorizes itself. Second, in so far as what his labour-power is worth — and hence what he is paid — is fixed, as with all other wage-labourers, by what it costs to produce and reproduce that particular labour-power, and not by what his work turns out.
But the same difference has to hold between him and the workers industrial capital employs directly as holds between industrial capital and commercial capital, and so between the industrial capitalist and the merchant. The merchant is nothing but an agent of circulation, and an agent of circulation produces no value and no surplus-value. (The extra value he adds to goods through his expenses comes down to an addition of value that already existed — though here the question forces itself on us: how does he keep and conserve this value of his constant capital?) So the commercial workers he has doing those same jobs cannot possibly create surplus-value for him directly either.
Here, as with the productive workers, we assume that the wage is fixed by the value of labour-power — that the merchant does not get rich by a deduction from the wage, so that in working out his costs he does not enter an advance for labour he has only partly paid for; in other words, that he does not get rich by cheating his clerks and so on.
What causes trouble here is not at all the question of how these commercial wage-workers directly produce profit for the man who employs them, even though they do not directly produce surplus-value — and profit is only surplus-value in changed form. That question has in fact already been answered by the general analysis of commercial profit.
Industrial capital makes its profit by selling labour that is contained in the goods and realized there, labour it has paid no equivalent for. Commercial capital gets its profit on the same principle — paying no equivalent for what it takes: it does not pay productive capital in full for the unpaid labour contained in the goods (in the goods, in so far as the capital laid out to produce them counts as an aliquot part of the total industrial capital), and then, when it sells the goods, it has that unpaid part, still contained in them, paid over to itself.
Commercial capital's relation to surplus-value is not the same as industrial capital's. Industrial capital makes surplus-value by appropriating other people's unpaid labour directly. Commercial capital appropriates a part of that surplus-value by having that part transferred to it from industrial capital.
Only because its business is realizing values does commercial capital count, within reproduction, as capital at all — and so, as a capital doing its job, it is from the surplus-value produced by the total capital that it draws its share. How big a profit an individual merchant makes depends on how much capital he can put to work in this process, and he can put more of it into buying and selling the greater the unpaid labour of his clerks. The function that makes his money capital, the merchant himself leaves his employees to perform in large part.
His clerks' labour goes unpaid, and unpaid labour of this kind creates no surplus-value. But it does create his power to appropriate surplus-value — which, for this capital, comes out to exactly the same thing. So for this capital it is a source of profit. Without it, a commercial business could never be run on a large scale, never in a capitalist way.
The unpaid labour of the worker gives productive capital surplus-value directly. In the same way, the unpaid labour of commercial wage-workers gives commercial capital a share in that surplus-value.
The difficulty is this. The merchant's own labour-time and labour create no value, even though it wins him a share of surplus-value that has already been produced. So where does that leave the variable capital he lays out in buying commercial labour-power?
Does this variable capital count as part of his outlay, as part of the capital he has advanced? If it does not, that seems to clash with the law by which profit rates are equalized — what capitalist would put up 150 if only 100 of it counted as capital he had advanced? If it does, then it seems to clash with what commercial capital essentially is. For this kind of capital does not work as capital by setting other people's labour in motion, the way industrial capital does. It works by doing the job itself — by carrying out the buying and selling itself — and it is precisely, and only, for that reason and through that means that it shifts onto itself a piece of the surplus-value that industrial capital has produced.
(So the points still to be investigated are these: the merchant's variable capital; the law of necessary labour within circulation; how the merchant's work keeps the value of his constant capital; commercial capital's part in reproduction taken as a whole; and finally the doubling — into commodity capital and money capital on the one hand, and into commercial capital and money-dealing capital on the other.)
Suppose each merchant had only as much capital as he could personally turn over by working himself. Commercial capital would then split into endlessly many tiny pieces, and that splitting would have to grow as capitalist production advanced and productive capital turned out goods on an ever bigger scale and moved ever bigger quantities. So the mismatch between the two would keep growing. And to the same degree that capital concentrated in production, it decentralized in circulation.
The industrial capitalist's purely commercial business — and with it his purely commercial outlays — would then expand without limit, since instead of dealing with a hundred merchants he would have to deal with a thousand. Much of what is gained by commercial capital standing on its own would be lost; and on top of the purely commercial costs, the other costs of circulation would grow too — sorting, dispatch and the like. So much for industrial capital.
Now commercial capital. First the purely commercial work. Working with big numbers takes no more time than working with small ones. Ten purchases of £100 take ten times as long as a single purchase of £1,000. Writing to ten small merchants costs ten times the correspondence, the paper and the postage of writing to one big merchant. A limited division of labour in the commercial office — one keeping the books, another the cash, a third the letters, one buying, one selling, one travelling — saves labour-time on an enormous scale, so that the number of commercial workers employed in wholesale trade bears no relation at all to the comparative size of the business. The reason is that in commerce — far more than in industry — a given function eats up the same labour-time whether the scale is small or large. That is also why, historically, merchants' businesses concentrated before industrial workshops did. Then there is the outlay on constant capital. A hundred small offices cost infinitely more than one large office, a hundred small warehouses more than one large one. And transport costs, which the merchant's business carries at least as money advanced, rise along with all this fragmentation.
The commercial side of the industrial capitalist's own business would swallow more labour and more circulation costs. The same commercial capital, divided among many small merchants, would — because of that fragmentation — need far more workers to do its work; and on top of that, the same commodity capital would call for more commercial capital to turn it over.
Let us call the whole of the commercial capital that is directly laid out on buying and selling commodities B, and the variable capital laid out on paying the commercial assistants b. Then B + b is smaller than the total commercial capital B would have to be if every merchant struggled along without help — that is, if no part of it were laid out as b. But we are still not through with the difficulty.
The price at which the goods are sold has to be enough, first, to pay the average profit on B + b. That is already accounted for by the fact that B + b is a shortening of the original B — a smaller commercial capital than would be necessary without b. But this sale price must also be enough, second, to replace what he paid out in wages, his variable capital, b, itself — over and above the profit on b that now appears on top. This second thing is what makes the difficulty.
Is b a new piece of the price? Or is it only part of the profit made on B + b, which looks like wages from the commercial worker's side and like nothing but the return of his variable capital from the merchant's own side?
On that second reading, the merchant's profit on the capital B + b that he advanced would come to no more than the profit falling on B at the general rate, plus b — b being paid out as wages, but yielding no profit of its own.
In fact it comes down to finding the limits of b (in the mathematical sense). First let us pin the difficulty down exactly. Let B be the capital put straight into buying and selling goods, K the constant capital used up in that function (the material office expenses), and b the variable capital the merchant advances.
Getting B back raises no difficulty at all. To the merchant it is simply the purchase price he has realized — or, for the manufacturer, the price of production. The merchant pays that price, and when he sells again he gets B back as part of his sale price, and along with this B the profit on B, as explained earlier. Say the commodity costs £100 and the profit on it is 10 per cent: it is then sold at 110. The commodity already cost 100 beforehand; the commercial capital of 100 only adds 10 to it.
Now take K. At most it is as big as — in fact it is smaller than — the constant capital the producer himself would use up in his own buying and selling; and that would come on top of the constant capital he needs directly for production. All the same, this part too must always come back out of the price of the commodity; or, which comes to the same, a matching slice of the commodity must always be spent, and reproduced, in that form — taking society's total capital as a whole. This slice of the capital advanced would press down on the rate of profit just as much as the whole mass of constant capital invested directly in production does. In so far as the industrial capitalist hands the commercial side of his business over to the merchant, he does not have to advance this part of the capital. The merchant advances it instead. But that advance is really only nominal: the merchant does not produce the constant capital he uses up, nor does he reproduce it (his material office expenses). Their production therefore appears as a business of its own — or at least as one branch of certain industrial capitalists' business — so that these play the same role as those who supply constant capital to the people producing means of subsistence. So the merchant gets, first, this replacement, and second, the profit on it. Both cut into the industrial capitalist's profit. But thanks to the concentration and the economy that come with the division of labour, the cut is smaller than it would have been had he advanced this capital himself. The rate of profit falls by less, because less capital is advanced this way.
So far, then, the sale price comes to B + K plus the profit on the two of them. As far as we have got, this part gives no trouble. But now b enters — the variable capital the merchant has advanced.
The sale price thereby becomes B + K + b, plus the profit on B + K, plus the profit on b.
B does nothing but replace the purchase price; besides the profit on B it adds nothing to that price. K adds not only the profit on K but K itself. But K plus the profit on K — the part of the circulation costs advanced as constant capital, plus the corresponding average profit — would be bigger in the industrial capitalist's hands than in the commercial capitalist's. The drop in average profit shows up in this form: the industrial capital advanced, less B + K, is credited with the full average profit; what the average profit loses on account of B + K goes to the merchant instead, and so that deduction appears as the profit of a capital of its own — commercial capital.
But b plus the profit on b is another matter — or, in the case assumed here, since the rate of profit is taken as 10 per cent, b + 1/10 b. And here is the real difficulty.
With b the merchant buys nothing but commercial labour, on our assumption — the labour needed for the circulation of capital, C–M and M–C. But commercial labour is simply the labour that has to be done anyway for a capital to operate as commercial capital, the labour that turns commodities into money and money back into commodities. Such labour realizes values; it does not create them. And only in so far as a capital carries out these functions — that is, only in so far as a capitalist performs these operations, this labour, with his capital — does it work as commercial capital and have a hand in settling the general rate of profit, drawing its dividend out of the total profit.
In (b + the profit on b), though, it seems that first the labour gets paid (it makes no difference whether the industrial capitalist pays the merchant for his own labour or for that of a clerk the merchant pays), and second that the profit gets paid on the payment for that labour — labour the merchant himself would have had to do. So commercial capital receives both: the repayment of b, and the profit on it. And this comes about because it is paid twice over for one thing — once for the labour that lets it work as commercial capital, and once again in profit, for being a capital that functions, which is to say for doing that very labour. This, then, is the question that has to be solved.
Suppose B = 100, b = 10 and the rate of profit 10 per cent. We set K = 0, so as not to drag into the reckoning once more an item of the purchase price that has already been settled and does not belong here. The sale price would then be B + p + b + p (= B + Bp' + b + bp', where p' is the rate of profit) = 100 + 10 + 10 + 1 = 121.
But if the merchant had not laid b out as wages — b is paid only for commercial labour, for labour needed to realize the value of the commodity capital that industrial capital throws onto the market — then things would stand like this. Buying or selling B = 100 takes the merchant's time, and we shall take it that this is all the time he has. The commercial labour represented by b, or 10, if it were paid out of profit instead of as wages, would presuppose another commercial capital of 100, since 10 per cent of that comes to b, which is 10. This second B = 100 would not be added to the commodity's price, but the 10 per cent would be. So there would be two operations of 100, making 200 — buying commodities for 200 + 20 = 220.
Commercial capital is absolutely nothing but one part of the industrial capital that operates in the circulation process, in a form made independent. So every question about it has to be answered by first putting the problem in the shape it has where the phenomena peculiar to commercial capital are not yet independent, but still directly tied to industrial capital, as a branch of it. Commercial capital works in the circulation process all the time, with an office instead of a workshop. So the b we are asking about here has to be looked at first where it appears in that form — in the industrial capitalist's own office.
To begin with, this office is always vanishingly small next to the industrial workshop. For the rest it is clear: as the scale of production widens, the commercial operations that industrial capital's circulation constantly requires multiply — both selling the product that exists as commodity capital, and turning the money set free back into means of production, and keeping the accounts for the whole of it. Working out prices, keeping the books, handling the cash, writing letters — all of it belongs here. The more developed the scale of production, the greater the industrial capital's commercial operations — though by no means in proportion — and so the greater the labour and the other circulation costs that realizing value and surplus-value requires. This makes commercial wage-labourers necessary, and they make up the actual office. The outlay on them, though made in the form of wages, is a different thing from the variable capital that buys productive labour. It raises the industrial capitalist's outlays, the mass of capital he must advance, without directly adding to surplus-value. For this is money paid for labour used only to realize values that have already been created. Like any other outlay of this kind, it brings the rate of profit down, because the capital advanced grows while the surplus-value does not. If the surplus-value s stays constant while the capital advanced C grows by ΔC, then the profit rate s/C is replaced by the smaller profit rate s/(C + ΔC).
So the industrial capitalist keeps these circulation costs down as far as he can, exactly as he keeps down what he lays out on constant capital. Industrial capital therefore does not treat its commercial employees the way it treats its productive wage-labourers. The more of the latter it employs, other circumstances staying equal, the more massive the production and the greater the surplus-value or profit. Here it is the other way round. The wider the scale of production, and the greater the value, and so the surplus-value, to be realized — hence the bigger the commodity capital produced — the more do office costs grow in absolute size, though not in proportion, and the more they give occasion for a particular division of labour. How much profit is the precondition for these outlays shows itself, among other things, in the fact that as commercial salaries rise, part of the salary is often paid out as a cut of the profit, fixed at so many per cent.
It belongs to the nature of such labour that it consists only in mediating operations — part of it working out values, part realizing them, part turning the realized money back into means of production — and that its scope therefore depends on how big the values produced and to be realized are. Labour of this kind is not the cause of those magnitudes and masses; it is their consequence — unlike directly productive labour. The same goes for the other circulation costs. Where there is much to weigh, to measure, to pack and to haul, the stuff must be there in the first place. How much packing and transport work there is depends on the mass of commodities those workers handle, and never the reverse.
The commercial worker does not directly produce surplus-value. But what his labour costs is fixed by the value of his labour-power, that is, by what it costs to produce that labour-power; while the actual exercise of it — the straining, the spending of energy, the wear and tear — is no more bounded by that value than is the case with any other wage-worker. So his wage need not bear any fixed relation to how much profit he helps the capitalist realize. His cost to the capitalist and his yield to the capitalist are two different magnitudes. What he brings in comes not from creating surplus-value directly, but from helping to cut the cost of realizing surplus-value, in so far as he performs labour — part of it unpaid.
A proper commercial worker belongs among the better-paid wage-workers — one of those whose labour is skilled, above the average. But with the advance of the capitalist mode of production his wage has a tendency to fall, even measured against average labour. First, through the division of labour inside the office: only a one-sided development of the worker's capacity has to be produced, and part of what it costs to produce that capacity costs the capitalist nothing, since the worker's skill develops through the work itself, and all the faster the more one-sided it becomes with the division of labour. Second, because the groundwork — commercial know-how, languages and so on — is reproduced ever faster, more easily, more widely and more cheaply as the capitalist mode of production bends teaching methods towards the practical. The spread of popular education allows this kind of labour to be recruited from classes that used to be shut out of it and were used to a worse way of living. It also swells the supply, and with it the competition. With a few exceptions, then, the labour-power of these people loses value as capitalist production advances; their wages fall while their ability to work rises. The capitalist takes on more of these workers when he has more value and profit to realize. The growth of this labour is always a result of the growth of surplus-value, never its cause.
So a duplication takes place. On one side, the functions of commodity capital and money capital — and so, more precisely, those of commercial capital — are general forms that industrial capital necessarily takes. On the other side, particular capitals, and with them particular sets of capitalists, are occupied exclusively with these functions, and the functions accordingly turn into separate spheres in which capital is valorized.
Only under commercial capital do the commercial functions and the circulation costs take on a life of their own. The side of industrial capital that faces circulation exists not just in its steady being as commodity capital and money capital, but in the office beside the workshop as well. But it becomes autonomous for commercial capital. For commercial capital, the office is its only workshop. The share of capital used in the form of circulation costs looks much larger in the wholesale merchant's hands than in the industrialist's, because the capital that the class of industrial capitalists as a whole would have to use this way is now gathered into the hands of a few merchants — over and above the office that goes with every industrial workshop. Having taken on the circulation functions, they also shoulder the circulation costs that come with them.
To industrial capital, circulation costs appear as an expense, and an expense they are. To the merchant, those same costs appear as his profit's source — and, on the assumption of a general rate of profit, that profit is proportionate to how big those costs are. So for commercial capital, the money laid out on these circulation costs is a productive investment. And so the commercial labour it buys is, for it, directly productive too.