Let's look at the annual working of social capital — that is, of the total capital, of which each individual capital is only a fragment. A fragment's movement is its own movement, and at the same time part of the movement of the whole. Let's look at this working in its result: the mass of commodities society turns out over the year. Looking at it this way must show how the reproduction process of social capital actually runs, what marks it off from the reproduction process of an individual capital, and what the two share.

The year's product contains two kinds of parts: the parts that replace capital — social reproduction — and the parts that fall to the consumption fund, what gets eaten, worn and lived on by workers and capitalists alike. So it contains both productive consumption and individual consumption.

It equally contains the reproduction — that is, the upkeep — of the capitalist class and of the working class. And because it contains that, it also contains the reproduction of the capitalist character of the whole production process.

The shape of circuit we have to work with is obvious, and consumption necessarily plays a part in it: the starting point, C´ = C + c — the commodity capital — holds the constant and variable capital-value together with the surplus-value. So its movement covers both individual consumption and productive consumption together.

In the circuits M-C...P...C'-M' and P...C'-M'-C...P, it is the movement of capital that forms the starting point and the end point of the circuit. That does also take in consumption, since the commodity — the product — has to be sold. But once the sale is taken as already done, what happens to that commodity afterward makes no difference to the movement of an individual capital.

With the movement of C'...C', by contrast, the conditions of social reproduction become visible precisely here, because this circuit forces us to show what becomes of every part of the value of this total product C'. So here the whole reproduction process includes the process of consumption carried by circulation just as much as it includes the reproduction process of capital itself.

For what we're doing now, the reproduction process has to be looked at from two angles together: how the value of each part of C' gets replaced, and how its material gets replaced too. We can no longer settle, as we could when we were analysing the value of an individual capital's product, for simply assuming that the individual capitalist turns the parts of his capital into money by selling his commodity-product, and then turns that money back into productive capital by buying the elements of production on the market. Those elements of production, as far as they are physical things, are themselves just as much a part of the social capital as the individual finished product that gets exchanged for them and replaced by them. On the other hand, the part of the social commodity-product that the worker consumes by spending his wage, and the capitalist consumes by spending the surplus-value — its movement is not just one integrating piece of the movement of the whole product. It is bound up with the movement of the individual capitals, and its course can't be explained by just assuming it happens.

Here is the question as it stands right in front of us: how does the annual product replace, in value, the capital used up in production — and how does this replacement interweave with the capitalists consuming the surplus-value and the workers consuming their wages?

So for now this is about reproduction on the same scale as before — simple reproduction. It also assumes not just that products exchange at their values, but that no revolution in value — no sudden change in what the things themselves are worth — hits the components of productive capital.

Where prices diverge from values, that fact cannot affect the movement of social capital as we're tracing it. The same total masses of products still exchange against each other as before, even though the individual capitalists involved end up in value-relations that would no longer be proportional to what each of them advanced or to the mass of surplus-value each of them produced on their own.

As for revolutions in value: where they are general and spread evenly, they change nothing in the relations between the value-parts of the year's total product. Where instead they hit only some branches of production and not others, they show up as disturbances. First, a disturbance can only be understood as such by treating it as a deviation from value-relations that would otherwise have stayed constant. Second, once the law is established that one value-part of the annual product replaces constant capital and another replaces variable capital, a revolution in the value of either the constant or the variable part would change nothing in that law — it would only change the relative size of the value-parts playing the one role or the other, because other values would have stepped into the place of the original ones.

As long as we were looking at capital's production of value and its product-value one capital at a time, the physical shape of the commodity-product made no difference at all to the analysis — whether it was, say, machines, or corn, or mirrors. It was always just an example; any branch of production whatsoever could serve the illustration equally well. What we were dealing with was the immediate production process itself, which at every point presents itself simply as the process of one individual capital. As far as the reproduction of capital went, it was enough to assume that, somewhere within circulation, the part of the commodity-product that represents capital-value finds the chance to turn back into its elements of production and so back into its shape as productive capital — just as it was enough to assume that the worker and the capitalist find, on the market, the commodities on which they spend the wage and the surplus-value. That merely formal way of presenting things no longer suffices once we're considering the total social capital and its product-value. Turning one part of the product-value back into capital, and letting another part go into the individual consumption of the capitalist class and of the working class — this is a movement inside the very product-value that the total capital has resulted in. And this movement is not just a replacement of value; it is a replacement of material too. So it is conditioned just as much by how the value-components of the social product relate to each other as by their use-value, their material shape.

Simple reproduction on an unchanging scale looks like an abstraction, and for two reasons. On one hand, on capitalist ground, having no accumulation at all — no reproduction on an expanded scale — is itself a strange assumption to make. On the other hand, the conditions under which production happens do not stay exactly the same from year to year (even though staying the same is exactly what we are assuming here).

What we're assuming is this: a social capital of a given value delivers, this year as last, the same mass of commodity-values and satisfies the same amount of need, even though the forms the commodities take may change in the process.

And yet, wherever accumulation does happen, simple reproduction always forms a part of it — so it can be looked at on its own, and it is a real factor of accumulation.

The value of the year's product can fall while the mass of use-values stays the same; the value can stay the same while the mass of use-values falls; value and the mass of reproduced use-values can both fall together. All of this just comes down to reproduction happening either under more favourable circumstances than before, or under harder ones — and harder circumstances can end up as an incomplete, a deficient, reproduction. All of this can only touch the quantitative side of the different elements of reproduction. It does not touch the role they play — as capital being reproduced, or as revenue being reproduced — in the process as a whole.

Engels notes the source: this section is in the main taken from Marx's Manuscript II, while the schema that follows comes from the later Manuscript VIII.

The whole product of society — and so the whole of its production — splits into two great departments:

I. Means of production — goods whose form is such that they must enter productive consumption, or at least can enter it.

II. Means of consumption — goods whose form lets them enter the individual consumption of the capitalist class and the working class.

Within each department, all the different branches of production belonging to it count as one single great branch — one branch for means of production, the other for means of consumption. All the capital used in each of these two branches forms its own great department of the total social capital.

In each department, capital splits into two parts:

1. Variable capital. Looked at by value, this equals the value of the social labour-power used in that branch of production — that is, the sum of the wages paid for it. Looked at materially, it consists of the labour-power itself at work: the living labour that this capital-value sets in motion.

2. Constant capital — the value of all the means of production used to produce in that branch. This in turn splits into fixed capital (machines, tools, buildings, draught animals, and so on) and circulating constant capital (materials used up in production: raw materials, auxiliary materials, semi-finished goods, and so on).

The value of the whole annual product that this capital produces in each of the two departments splits into two parts. One part is the constant capital c — capital used up in production whose value is merely carried over onto the product, not newly added. The other part is the value added by the year's labour as a whole. This second part splits again: into the replacement of the variable capital v laid out, and the excess over that, which forms the surplus-value s. So just like the value of any single commodity, the value of the whole annual product of each department splits into c + v + s.

The value-part c, which stands for the constant capital used up in production, does not match the value of all the constant capital used in production.

The materials are used up completely, so their whole value passes onto the product. Of the fixed capital, only a part is used up completely, so only that part's whole value passes onto the product. The rest of the fixed capital — machines, buildings, and so on — goes on existing and working just as before, only with its value reduced by the year's wear and tear. For the purpose of valuing this year's product, we are leaving that still-working part out of account altogether. It is a piece of capital-value standing beside the newly produced commodity-value, not inside it.

This already came up when we looked at the value of the product of an individual capital (Volume 1, Chapter VI). But here we must, for now, set that treatment aside. There, we saw that the value fixed capital loses through wear passes onto the commodity-product made during the period of wear — and that it makes no difference whether part of this fixed capital is replaced in kind out of that transferred value during that time, or not.

Here, by contrast, looking at the total social product and its value, we are forced — at least for now — to leave out the value that wear on fixed capital transfers to the annual product during the year, but only insofar as this fixed capital has not also been replaced in kind during the year. We will take the point up separately in a later section of this chapter.

For our study of simple reproduction, let's take the following schema as our basis, where c = constant capital, v = variable capital, and s = surplus-value, with the rate of surplus-value s/v assumed at 100%. The figures may stand for millions of marks, francs, or pounds sterling.

To sum up, the year's total commodity-product:

Total value = 9,000 — and by our assumption, this excludes the fixed capital that goes on functioning in its own natural form.

Now, if we look at the exchanges required for simple reproduction — where the whole of the surplus-value is consumed unproductively — and set aside for now the circulation of money that carries them out, three major footholds present themselves right from the start.

1. The 500v — the workers' wages — and the 500s — the surplus-value of department II's capitalists — must be spent on means of consumption. But their value exists in means of consumption worth 1,000, which sit in the hands of department II's own capitalists: 500 replacing what they advanced, and 500s representing their surplus-value. So the wages and surplus-value of department II are exchanged, within department II itself, against department II's own product. With that, (500v + 500s) II = 1,000 in means of consumption drops out of the total product.

2. Department I's 1,000v + 1,000s must likewise be spent on means of consumption — that is, on the product of department II. So it must be exchanged against the constant-capital part of that product still remaining, 2,000c, which is equal to it in amount. In return, department II receives an equal sum of means of production — product of department I — embodying the value of I's 1,000v + 1,000s. With that, 2,000 IIc and (1,000v + 1,000s) I drop out of the reckoning.

3. There remains 4,000 Ic. This is made up of means of production that only department I itself can use up, serving to replace the constant capital it has consumed. It is disposed of by mutual exchange among department I's individual capitalists — just as the (500v + 500s) II was disposed of by exchange between the workers and the capitalists of department II, and between those capitalists among themselves.

These three points are given only, for now, to help understand what follows.

Engels notes that from this point the text returns to Marx's Manuscript VIII.

Let's start with the big exchange between the two classes. Department I holds 1,000v+1,000s in value — value that currently sits, in the hands of the people who made it, as means of production. This exchanges against 2,000 IIc: value that exists as means of consumption. Through this, capitalist class II converts its constant capital — worth 2,000 — back out of the form of means of consumption and into the form of means of production for making means of consumption. In that form it can work again as a factor in the labour process and function as constant capital-value. At the same time, this realizes, in means of consumption, both the equivalent for labour-power in department I (1,000 Iv) and the surplus-value of the capitalists in department I (1,000 Is). Both are converted out of their natural form as means of production into a natural form in which they can be consumed as revenue.

This exchange between the two classes only happens by way of a circulation of money — and that same circulation, in mediating the exchange, is exactly what makes it hard to see clearly what's going on. But it matters decisively, because the variable part of capital must keep turning up again in money form: as money-capital that then converts into labour-power. In every line of business running at once anywhere in society — whether it belongs to department I or department II — variable capital must be advanced in money. The capitalist buys labour-power before it enters the production process, but he only pays for it at agreed dates, after it has already been used up producing use-values. Like the rest of the value of the product, the part of that value which is merely the equivalent of the money he spent paying for labour-power — the part representing variable capital-value — also belongs to him. And in that very part of the value, the worker has already handed him the equivalent of his wage. But it is the reconversion of the commodity into money — its sale — that gives the capitalist his variable capital back in money form, so that he can advance it again to buy labour-power.

In department I, the capitalist class as a whole has paid the workers £1,000 (I say pounds sterling just to mark that this is value in money form) = 1,000v, for the part of the value of product I that already existed as the v-part — that is, for the means of production the workers made. The workers take this £1,000 and buy means of consumption of the same value from the capitalists in department II, and in doing so turn one half of department II's constant capital into money. The capitalists in department II, in turn, use this £1,000 to buy means of production worth 1,000 from the capitalists in department I. This turns the variable capital-value of 1,000v — which, for department I, existed as part of their product in the natural form of means of production — back into money. It can now function again, in the hands of the capitalists in department I, as money-capital that converts into labour-power, the most essential element of productive capital. This is the route by which their variable capital flows back to them in money form, as a result of realizing part of their commodity-capital.

As for the money needed to exchange the surplus-value part of department I's commodity-capital against the second half of department II's constant-capital part — that can be advanced in various ways.

In reality this circulation is made up of a countless mass of individual purchases and sales between individual capitalists of both departments. But in every case the money must come from these capitalists themselves, since we have already accounted separately for the money the workers throw into circulation. Sometimes a capitalist in department II might buy means of production from a capitalist in department I out of the money-capital he holds alongside his productive capital; sometimes, the other way round, a capitalist in department I might buy means of consumption from a capitalist in department II out of a money-fund set aside for personal spending, not for capital. Certain reserves of money — whether for advancing capital or for spending revenue — must in every case be assumed to sit in the capitalist's hands alongside his productive capital; the earlier parts of this volume established that.

Let's assume — the exact proportion doesn't matter for our purpose — that half this money is advanced by the capitalists of II to replace their constant capital by buying means of production, and the other half is spent by the capitalists of I on consumption. Then: department II advances £500 and uses it to buy means of production from I. Together with the £1,000 that came earlier from the workers of I, this replaces three-quarters of its constant capital in kind. Department I uses this same £500 to buy means of consumption from II, completing the circuit commodity → money → commodity (C-M-C) for half the surplus-value part of its commodity-capital — that part of its product is now realized as a fund of consumption. Through this second step, the £500 flows back into department II's hands as money-capital held alongside its productive capital.

On the other side, for the other half of the surplus-value part of its commodity-capital — still sitting with it unsold — department I lays out, in advance of selling it, £500 to buy means of consumption from II. With this same £500, II buys means of production from I, and so replaces its whole constant capital in kind (1,000 + 500 + 500 = 2,000), while I has now realized its entire surplus-value in means of consumption.

In total, £4,000 worth of commodities would have changed hands here, carried by a circulation of £2,000 in money — and that £2,000 comes out only because the whole year's product is being presented as if exchanged all at once, in a few large lots. What matters is only this: department II not only converts its constant capital — reproduced as means of consumption — back into the form of means of production, but also gets back the £500 it advanced into circulation to buy means of production. And in the same way, department I not only holds its variable capital again in money form — reproduced as means of production — as money-capital directly convertible once more into labour-power, but also gets back the £500 it laid out in advance, before selling the surplus-value part of its capital, to buy means of consumption. That £500 flows back to department I, though, not because it was spent, but because of the sale that followed — the sale of the part of its commodity-product carrying half its surplus-value.

In both cases, something more than the obvious is going on. Department II doesn't only convert its constant capital back from product-form into the natural form of means of production — the only form in which it can function as capital at all. And department I doesn't only convert its variable-capital part into money form, and the surplus-value part of its means of production into a form it can consume as revenue. Beyond that: the £500 of money-capital that II advanced to buy means of production flows back to it — even though it advanced that money before it had sold the matching part of its constant capital, the part sitting there as means of consumption. And the £500 that I laid out in advance to buy means of consumption flows back to it too. This money flows back to each of them only because each threw an extra £500 into circulation beyond the value of their own commodities — II beyond its constant capital existing in commodity-form, I beyond its surplus-value existing in commodity-form. In the end they have paid each other in full through the exchange of their respective commodity-equivalents. The money that each threw into circulation, over and above the value of their own commodities, as the means for this exchange, comes back out of circulation to each of them, in proportion to how much each put in. Neither of them is one iota richer for it. Department II had a constant capital of 2,000 in the form of means of consumption, plus £500 in money; it now has 2,000 in means of production and £500 in money — just as before. Department I likewise has, just as before, a surplus-value of 1,000 — now turned from means of production into a fund of consumption — plus £500 in money, just as before. The general rule follows: of the money that industrial capitalists throw into circulation to carry their own commodities round — whether on account of the constant value-part of the commodity, or of the surplus-value in the commodities to the extent that it is spent as revenue — exactly as much flows back into the hands of each capitalist as he advanced for that money circulation.

Now, as for how class I's variable capital turns back into money: once the capitalists of I have laid it out as wages, it exists for them, at first, only in the commodity-form the workers handed them in return. They paid this out to the workers, in money, as the price of their labour-power. In doing so, they paid for the part of their commodity-product's value equal to that variable capital laid out in money — and that is what makes them the owners of this part of the product too. But the workers department I employs are not buyers of the means of production they themselves have just made; they are buyers of the means of consumption that department II produces. So the variable capital I advanced in money to pay for labour-power does not flow straight back to the capitalists of I. Instead, through the workers' purchases, it passes into the hands of the capitalist producers of the goods that workers need and can get — that is, into the hands of the capitalists of II. And only once II uses that money to buy means of production — only by this detour — does it flow back into the hands of the capitalists of I.

What follows from this is that, under simple reproduction, the value-sum v+s of commodity-capital I — and so too the corresponding proportional part of department I's total commodity-product — must equal the constant capital IIc marked off as the corresponding proportional part of the total commodity-product of class II. In other words: I(v+m) = IIc.

Two components of Department II's product value are still to be examined: v (wages) and s (surplus-value — the German writes it m, for Mehrwert). Looking at them has nothing to do with the biggest question occupying us here — whether the split of value into c + v + s, true of each individual capitalist's product, also holds for the value of the whole year's product, even though at that scale it shows up in a different guise. That question gets answered elsewhere: through the exchange of Department I's wages-plus-surplus, I(v+m), against Department II's constant capital, IIc, and through an examination — saved for later — of how Department I's own constant capital, Ic, gets reproduced out of Department I's own year's product.

Department II's v+s exists physically as consumption goods. The variable capital capitalists advance to pay for labour-power has to be spent by the workers mostly on things to consume. And s, on the assumption of simple reproduction, actually does get spent as revenue on consumption goods. So at first glance it is clear enough: with the wages capitalists II pay them, the workers of Department II buy back part of their own product — as much of it as the money value of their wages will cover.

This is how capitalist class II turns the money capital it advanced for labour-power back into money. It is exactly as if it had paid its workers in mere tokens standing for value. Once the workers cash in these tokens by buying part of the commodity product they made — a product that belongs to the capitalists — the tokens flow back into the capitalists' hands, except that here the token does not just represent value: being gold or silver, it actually carries that value in its own body. We will look more closely later at this kind of reflux of variable capital advanced in money form, in the process where the working class appears as buyer and the capitalist class as seller. Here, though, a different point needs discussing about this same reflux of variable capital back to its starting point.

Department II's yearly output comes from all sorts of different industries. But looking at what they produce, these industries fall into two broad groups:

a) Necessities. These are goods that go into the working class's consumption; and so far as they are necessary means of subsistence, they also form part of what the capitalist class consumes — though the capitalists' version is often of a different quality and value from the workers'. For our purposes we can lump this whole group under one heading: necessities. It makes no difference whether a given product — tobacco, say — is something the body actually needs. It is enough that people are in the habit of treating it as one.

b) Luxuries. These only enter the capitalist class's consumption — they can only be bought with spent surplus-value, which never falls into a worker's hands.

With necessities, it's clear enough: the variable capital advanced to produce this category of goods must flow straight back, in money form, to the part of capitalist class II that produces them — the capitalists of IIa. They sell these goods to their own workers for the same amount the workers were paid in wages. This reflux runs directly to the whole of subdivision IIa, no matter how many transactions between capitalists in the various industries involved are needed to spread that returning variable capital among them in the right proportions. These are just circulation processes, and the money that circulates in them comes directly from what the workers spend.

Subdivision IIb works differently. The whole value-product we're dealing with here, IIb's v+s, takes the physical form of luxury articles — goods the working class can no more buy than it can buy the machinery and materials that Department I's own wage-value, Iv, happens to exist as, even though these luxury goods, like those means of production, are products of these very workers. So the reflux that returns the variable capital advanced in this subdivision to its capitalists in money form cannot happen directly. It has to travel by a detour — the same as with Iv.

Let's take the same example as before for the whole of class II: v = 500, s = 500. But now suppose the variable capital and the surplus-value that matches it are split up as follows:

Subdivision a: necessities. v = 400, s = 400. That gives a mass of necessities worth 400v + 400s = 800 — written IIa(400v + 400s).

Subdivision b: luxuries, worth 100v + 100s = 200 — written IIb(100v + 100s).

The workers of IIb were paid 100 for their labour-power — say, £100 in money. With it they buy £100 worth of necessities from the capitalists of IIa. Those capitalists then use this same £100 to buy £100 worth of IIb's goods — luxuries — which is how the variable capital of the IIb capitalists flows back to them in money form.

In IIa, 400v has already come back into the capitalists' hands as money, through the exchange with their own workers. Beyond that, a quarter of the part of their product that represents surplus-value has been handed over to the workers of IIb, and in exchange IIa has received 100v worth of IIb's luxury goods.

Now suppose — and this is an assumption, not something we've found to be true — that the capitalists of IIa and IIb split their revenue spending between necessities and luxuries in the same proportion: say, both spend 3/5 on necessities and 2/5 on luxuries. On that assumption, the capitalists of subclass IIa lay out their surplus-value revenue of 400s as follows: 3/5, or 240, on their own product, necessities; and 2/5, or 160, on luxuries. The capitalists of subclass IIb divide their surplus-value of 100s the same way: 3/5, or 60, on necessities, and 2/5, or 40, on luxuries — this last amount produced and exchanged within their own subclass.

The 160 worth of luxuries that IIa's surplus-value obtains comes to the capitalists of IIa as follows. Of IIa's 400 in surplus-value, we already saw that 100 — in the form of necessities — was exchanged for an equal amount of IIb's variable capital, existing as luxuries; and a further 60 in necessities was exchanged for 60 of IIb's surplus-value, also in luxuries. Here, then, is the full reckoning:

1. The 400v of subdivision a gets eaten up by the workers of IIa — it forms part of their own product, necessities, and they buy it from the capitalist producers of their own subdivision. This brings those capitalists back £400 in money: the same 400 in variable capital they had paid out as wages to these very workers. With it, they can buy labour-power all over again.

2. Part of the 400s belonging to a — the part equal to 100v of b, that is, a quarter of a's surplus-value — gets realized in luxury articles as follows. The workers of b were paid 100 in wages by the capitalists of their own subdivision, b. With this they buy a quarter of a's surplus-value, that is, goods that consist of necessities. The capitalists of a then use this same money to buy, at the same value, luxury articles worth 100v of b — half of the whole luxury output. This is how the variable capital of the capitalists of b flows back to them in money form, letting them start their reproduction over again by buying labour-power anew — but only because the whole of class II's constant capital has, by this point, already been replaced through the exchange of I(v+m) against IIc. So the labour-power of the luxury workers can be sold again only because the part of their own product created as the equivalent of their wage gets drawn by the capitalists of IIa into their own consumption fund and used up there. (The same holds for the sale of labour-power under step 1: since IIc — the thing I(v+m) is exchanged against — consists of both luxuries and necessities, what gets renewed through I(v+m) supplies the means of production for both luxury goods and necessities alike.)

3. Now we come to the exchange between a and b, so far as it is only an exchange between the capitalists of the two subdivisions. What we've covered so far has already accounted for the variable capital (400v) and part of the surplus-value (100s) in a, and the variable capital (100v) in b. We also assumed, as the average ratio of capitalist revenue-spending in both classes, 2/5 on luxuries and 3/5 on necessities. So beyond the 100 already spent on luxuries, the whole of subclass a still has 60 left over for luxuries, and, in the same ratio, subclass b has 40.

So IIa's surplus-value splits into 240 for necessities and 160 for luxuries: 240 + 160 = 400s for IIa.

IIb's surplus-value splits into 60 for necessities and 40 for luxuries: 60 + 40 = 100s for IIb. This class consumes the last 40 — two-fifths of its surplus-value — straight out of its own product. It gets the 60 worth of necessities by exchanging 60 of its surplus product for 60s of a.

So for the whole of capitalist class II — where v + s exists as necessities in subdivision a, and as luxuries in b — we have:

IIa(400v + 400s) + IIb(100v + 100s) = 1,000. Through this whole movement, that gets realized as: 500v(a+b) — realized in 400v(a) and 100s(a) — plus 500s(a+b) — realized in 300s(a), 100v(b), and 100s(b) — totalling 1,000.

Looking at a and b separately, here is how each realizes its value:

To keep things simple, let's hold the same ratio between variable and constant capital across the board — though nothing forces us to. Then 400v in branch a comes with a constant capital of 1,600, and 100v in branch b comes with a constant capital of 400. This splits department II into its two branches, a and b, as follows:

Accordingly, of the 2,000 IIc in means of consumption that get exchanged against 2,000 I(v+s), 1,600 turn into means of production for necessary means of subsistence, and 400 into means of production for luxury goods.

The 2,000 I(v+s) would then itself break down into (800v+800s)I for a — 1,600 worth of means of production for necessary means of subsistence — and (200v+200s)I for b — 400 worth of means of production for luxury goods.

A large part — not just the actual instruments of labour but also the raw and auxiliary materials and so on — is the same for both branches. But when it comes to how the different value-parts of the whole product I(v+s) get exchanged, this split into a and b makes no difference at all. Both the 800 Iv above and the 200 Iv are realized the same way: wages get spent on 1,000 IIc worth of consumption goods, so the money capital laid out for this comes back distributed evenly among the capitalist producers of I, replacing each one's advanced variable capital in money in proportion to their share. On the other side, realizing the 1,000 Is works the same way: the capitalists again draw evenly — in proportion to the size of their surplus-value — on the whole second half of IIc, the 1,000 made up of 600 IIa and 400 IIb in consumption goods. So those who replace the constant capital of IIa:

What's arbitrary here — in both I and II — is the ratio of variable to constant capital, and likewise the fact that this ratio is the same across I and II and their sub-branches. That sameness is assumed purely to keep things simple; assuming different ratios instead would change absolutely nothing about the conditions of the problem or its solution. But what does follow as a necessary result, on the assumption of simple reproduction, is:

1. That the new value-product of a year's labour, created in the natural form of means of production (splitting into v+s), must equal the constant capital-value c of the product-value made by the rest of the year's labour, reproduced in the form of means of consumption. If it were less than IIc, department II could not fully replace its constant capital; if it were greater, a surplus would be left over unused. Either way, the assumption of simple reproduction would be violated.

2. That for the annual product reproduced in the form of means of consumption, the variable capital v advanced in money form can only be realized — for its recipients, insofar as they are luxury workers — in the part of the necessary means of subsistence that embodies, in its first shape, the surplus-value of the capitalist producers of those necessities. In other words, the v laid out in luxury production equals a corresponding part, by value, of the s produced in the form of necessary means of subsistence — and so must be smaller than that whole s, namely (IIa)s. Only by realizing that v in this part of s does the money form of the advanced variable capital flow back to the capitalist producers of luxury articles. This is exactly the same kind of phenomenon as the realization of I(v+s) in IIc — except that here, (IIb)v is realized in a part of (IIa)s equal to it in value. These relations stay qualitatively decisive for every distribution of the annual total product, as far as that product genuinely enters the process of annual reproduction mediated by circulation. I(v+s) can only be realized in IIc, just as IIc, in its function as part of productive capital, can only be renewed through this realization; in the same way, (IIb)v can only be realized in a part of (IIa)s, and only in this way is (IIb)v converted back into its form as money capital. This holds, of course, only to the extent that all of this is genuinely a result of the reproduction process itself — that is, only so long as, for instance, the capitalists of IIb are not raising money capital for v some other way, through credit. Quantitatively, though, the exchanges of the different parts of the annual product can only take place in the proportions set out above so long as the scale and value-ratios of production stay stationary, and so long as these strict ratios are not altered by foreign trade.

Suppose one said, in Adam Smith's manner, that I(v+s) resolves into IIc and IIc resolves into I(v+s) — or, as he more often and even more absurdly puts it, that I(v+s) forms components of the price (or value — he says "value in exchange")

of IIc, and IIc forms the whole component of the value of I(v+s) — then, just as well, one could and would have to say that (IIb)v resolves into (IIa)s, or (IIa)s into (IIb)v, or that (IIb)v forms a component of the surplus-value of IIa, and vice versa: surplus-value would then resolve into wages, that is, into variable capital, and variable capital would form a "component" of surplus-value. This absurdity is in fact found in Adam Smith, because for him wages are determined by the value of the necessary means of subsistence, while the value of those very commodities is in turn determined by the value of the wages (variable capital) and surplus-value contained in them. He is so absorbed in the fragments into which the value-product of a working day breaks down on a capitalist basis — namely into v and s — that he completely forgets: in simple commodity exchange it makes no difference at all whether the equivalents, existing in different natural forms, consist of paid or unpaid labour, since in both cases they cost the same amount of labour to produce. It likewise makes no difference whether A's commodity is a means of production and B's a means of consumption, or whether, after the sale, one commodity goes on to function as a component of capital while the other enters the consumption fund and gets consumed as revenue, according to Adam. What the individual buyer does with his commodity plays no part in the exchange of commodities, in the sphere of circulation, and does not touch the commodity's value. None of this changes just because, in analysing the circulation of the annual total social product, the specific use each part of that product is put to — the moment of its consumption — has to be taken into account.

None of this — the exchange of (IIb)v for an equal-value part of (IIa)s established above, nor the further exchanges between (IIa)s and (IIb)s — assumes that the individual capitalists of IIa and IIb, or their two classes taken as wholes, split their surplus-value between necessary consumption goods and luxury goods in the same proportion. One capitalist may spend more on the one kind of consumption, another more on the other.

On the ground of simple reproduction, all that is assumed is that a sum of value equal to the whole of the surplus-value gets realized in the consumption fund. That fixes the total, then, and nothing else about how it is spent. Within each department, one capitalist may spend more on a, another more on b — but this can offset itself across the group, so that the capitalist classes a and b, taken as wholes, each take the same share of both.

The value-ratios — the proportional share of the two kinds of producers, a and b, in the total value of product II, and hence also a determinate quantitative ratio between the branches of production that supply those products — are, however, necessarily given in every concrete case. Only the particular ratio used here as an example is hypothetical; assume a different one, and nothing about the qualitative relations changes — only the quantitative figures would change. But should some circumstance bring about a real change in the proportional size of a and b, the conditions of simple reproduction would change correspondingly too.

From the fact that (IIb)v is realized in an equivalent part of (IIa)s, it follows that as the luxury share of the annual product grows — as a rising share of labour-power gets absorbed into luxury production — the reconversion of the variable capital advanced in (IIb)v back into money capital, so that it can function again as the money form of variable capital, and with it the existence and reproduction of the part of the working class employed in IIb — their supply of necessary means of subsistence — comes to depend, in that same proportion, on the capitalist class's extravagance: on their spending a substantial part of their surplus-value on luxury articles.

Every crisis momentarily reduces luxury consumption. It slows down and delays the reconversion of (IIb)v into money capital, allows it only partially, and so throws part of the luxury workers onto the street — while, by the same token, it also brings the sale of necessary means of consumption to a standstill and cuts it back. This is quite apart from the unproductive workers dismissed at the same time, who receive part of the capitalists' luxury spending in payment for their services (these workers are themselves, to that extent, a luxury article), and who take a very large part, in particular, in the consumption of necessary means of subsistence too. The reverse happens in a period of prosperity, especially during its speculative bloom — when, for other reasons as well, the relative value of money expressed in commodities falls (without any real change in value elsewhere), so that the price of commodities rises independently of their own value. Not only does the consumption of necessary means of subsistence rise; the working class — whose whole reserve army has now become actively employed — also gets a momentary share in the consumption of luxury articles otherwise closed to it, and besides that, in the class of necessary consumption articles which otherwise, for the most part, form "necessary" means of consumption only for the capitalist class — which in turn drives prices up further.

It is a pure tautology to say that crises arise from a shortage of consumption that can pay, or of consumers who can pay. The capitalist system knows no kind of consumption except paying consumption — apart from the pauper's kind, or the thief's. That commodities can't be sold means nothing more than that no buyers able to pay were found for them — that is, no consumers (whether the commodities are ultimately bought for productive or for individual consumption). But suppose one tries to give this tautology the appearance of a deeper explanation by saying that the working class receives too small a share of its own product, and that the trouble would be fixed as soon as it received a larger share — that is, as soon as wages rise. Then the only thing to point out is this: crises are, every single time, prepared precisely by a period in which wages rise generally and the working class really does get a larger share of the part of the annual product meant for consumption. On the logic of these knights of sound and "simple" (!) common sense, that period ought instead to banish the crisis. So it seems that capitalist production contains conditions, independent of anyone's good or bad will, that allow that relative prosperity of the working class only for a moment — and always only as the storm-petrel heralding a crisis.

We saw earlier how the proportional relation between the production of necessary means of consumption and the production of luxury goods determined the split of II(v+s) between IIa and IIb — and so also the split of IIc between (IIa)c and (IIb)c. This relation reaches down to the very root of the character and the quantitative proportions of production, and is an essential, determining factor in how the whole of it takes shape.

In substance, simple reproduction is directed toward consumption as its purpose, even though the individual capitalists' driving motive appears to be the grabbing of surplus-value. But the surplus-value — whatever its proportional size — is ultimately meant, here, to serve only the capitalist's own individual consumption.

Insofar as simple reproduction is a part — and the most significant part — of every annual reproduction on an expanded scale too, consumption as the aim persists there as well, alongside and in opposition to the motive of getting rich for its own sake. In reality the matter looks more tangled, because the others who take a cut of the loot — of the capitalist's surplus-value, people like landlords and lenders — show up as consumers in their own right, apparently nothing to do with him.

Up to this point, the exchanges between the different classes of producers have followed this pattern:

So that settles the circulation of 2,000 IIc, which is exchanged against I(1,000v+1,000s).

Setting 4,000 Ic aside for now, what's left is the circulation of v+s inside class II itself. II(v+m) splits between the two subclasses, IIa and IIb, like this:

The 400v of subclass a circulates entirely inside that subclass: the workers paid with it buy back, from their own employers the IIa capitalists, the very means of subsistence they themselves produced.

The capitalists of both subclasses spend their surplus-value in the same proportion: three-fifths on necessary means of subsistence from IIa, two-fifths on luxuries from IIb. That means three-fifths of subclass a's surplus-value — 240 — is consumed inside IIa itself, and likewise two-fifths of subclass b's surplus-value, already sitting there as luxuries, is consumed inside IIb itself.

That leaves the following still to be exchanged between IIa and IIb:

On IIa's side there's 160 of surplus-value; on IIb's side, 100v plus 60 of surplus-value. These two match up exactly. The workers of IIb take the 100 they were paid in wages and buy necessary means of subsistence worth 100 from IIa. The capitalists of IIb spend three-fifths of their surplus-value — 60 — buying their own necessary means of subsistence from IIa too. That gives the capitalists of IIa the money they need to lay out the other two-fifths of their surplus-value — 160 — on the luxury goods IIb produces: 100 replacing the wages IIb paid its workers, plus 60. Set out as a schema, this reads:

The figures in brackets are the ones that never leave their own subclass — they circulate and get consumed there alone.

When money-capital advanced as wages flows straight back to the capitalist who laid it out, that only happens for the capitalists of subclass IIa, the ones producing necessary means of subsistence — and even this is just one special case, shaped by particular conditions, of a general law already stated: money that commodity-producers put into circulation comes back to them, as long as commodity circulation runs its normal course.

One thing follows from this in passing. Suppose a money-capitalist stands behind the commodity-producer — someone who advances money-capital, in the strict sense (capital-value in money form), to the industrial capitalist. Then the real point where that money flows back to is this money-capitalist's own pocket. In this way, even though the money passes more or less through every hand along the way, the bulk of the circulating money belongs to the division of money-capital that is organized and concentrated in the form of banks and the like. The way this division advances its capital determines that the money must keep coming back to it in money form in the end — even though that return is itself carried out through the industrial capital turning back into money-capital.

Commodity circulation always needs two things: commodities put into circulation, and money put into circulation. Circulation doesn't grind to a halt the way direct exchange of products does, when a use-value simply changes hands. Money doesn't vanish just because it eventually falls out of one commodity's chain of transformations — it always lands on some new spot in circulation that a commodity has just vacated.

Take the circulation between IIc and I(v+m): we assumed 500 pounds in money gets advanced by II to carry it out. Across the countless separate transactions that make up circulation between whole classes of producers, sometimes one side, sometimes the other, is the one to act first as buyer — the one who puts money into circulation. Setting aside individual circumstances, that alone follows from the different production periods, and so the different turnover times, of the different capitals involved. So: II buys means of production from I for 500 pounds; I in turn buys means of consumption from II for 500 pounds; the money flows back to II. II is not enriched one bit by getting this money back. It first put 500 pounds of money into circulation and drew out commodities of the same value; then it sold commodities for 500 pounds and drew money of the same value back out. That's how the 500 pounds return to it. Looked at as a whole, II has put into circulation 500 pounds in money plus 500 pounds in commodities — 1,000 pounds total — and has drawn out of circulation 500 pounds in commodities plus 500 pounds in money. To exchange 500 pounds of I's commodities against 500 pounds of II's commodities, circulation needs only 500 pounds in money: whoever advances the money to buy someone else's commodity gets it back when selling their own. Had I instead bought first from II for 500 pounds and only later sold to II for 500 pounds, the 500 pounds would have flowed back to I, not to II.

In class I, the money laid out in wages — the variable capital advanced in money form — doesn't come straight back in that same form; it comes back indirectly, by a roundabout route. In II it's different: the 500 pounds in wages flows straight back from the workers to the capitalists. That direct return always happens wherever buying and selling between the same two parties keeps repeating, so the same two people are constantly facing each other, now as buyer, now as seller. Here's how: the capitalist in II pays for labour-power in money. That act — for him, simply money-capital turning into productive capital — is what makes him an industrial capitalist facing a wage-labourer. But then the worker, who a moment ago was the seller, the one dealing in their own labour-power, turns around and becomes the buyer, the one holding money, facing the capitalist as seller of goods. That's how the money laid out in wages flows back to the capitalist. So long as the sale of these goods isn't some kind of swindle, but a straight exchange of equal values in goods and money, this is not a process that enriches the capitalist. He doesn't pay the worker twice — once in money, once in goods. His money simply comes back to him the moment the worker spends it on his goods.

Money-capital turned into variable capital — that is, the money advanced in wages — plays a leading role in money circulation as such. Here's why: workers have to live from hand to mouth, so they can't extend the industrial capitalists any real credit. That means variable capital has to be advanced in money simultaneously at countless different points scattered across society, on short fixed terms — weekly, say — repeating at fairly quick intervals, whatever the turnover periods of capital happen to be in this or that branch of industry. (The shorter these intervals, the smaller the total sum of money this channel needs to throw into circulation at any one moment.) In every capitalist country, the money-capital advanced this way makes up a decisively large share of total circulation — all the more so because, before it flows back to its starting point, the same money travels through all sorts of other channels, serving as the means of circulation for a huge number of unrelated transactions along the way.

Now let's look at the circulation between I(v+m) and IIc from a different angle.

The capitalists of I advance 1,000 pounds to pay wages. With it, the workers buy 1,000 pounds' worth of means of subsistence from the capitalists of II, and II turns around and buys means of production worth the same money from the capitalists of I. That brings I's variable capital, in money form, back to it, while II has converted half of its constant capital back from commodity-capital into productive capital. II then advances a further 500 pounds to buy more means of production from I; I spends that money on means of consumption from II; so the 500 pounds flows back to II. II advances it again, to convert the last quarter of its constant capital — still sitting there as commodities — back into its productive, natural form. This money flows back to I once more, and is used again to buy the same amount of means of consumption from II; so the 500 pounds flows back to II a second time. II's capitalists now hold, just as before, 500 pounds in money and 2,000 pounds of constant capital — except this constant capital has now been freshly converted from commodity-capital back into productive capital. With only 1,500 pounds in money, a mass of commodities worth 5,000 pounds has been circulated. Here is how: (1) I pays the workers 1,000 pounds for labour-power, worth the same in commodities; (2) the workers use that same 1,000 pounds to buy means of subsistence from II; (3) II uses the same money to buy means of production from I, which restores I's 1,000 pounds of variable capital in money form; (4) II buys means of production from I for 500 pounds; (5) I uses that same 500 pounds to buy means of consumption from II; (6) II uses that same 500 pounds to buy means of production from I; (7) I uses that same 500 pounds to buy means of subsistence from II. In the end, 500 pounds has flowed back to II beyond the 2,000 pounds in commodities it threw into circulation — and for that 500 pounds, circulation did not take any commodity-equivalent away from II.

Set out step by step, the exchange runs like this:

I pays 1,000 pounds in money for labour-power — a commodity worth 1,000 pounds.

The workers spend that 1,000 pounds in wages buying means of consumption from II — again, a commodity worth 1,000 pounds.

With that same 1,000 pounds it just took in, II buys means of production from I of equal value — again, a commodity worth 1,000 pounds.

With that, the 1,000 pounds has flowed back to I as the money-form of its variable capital.

II buys means of production from I for 500 pounds — a commodity worth 500 pounds.

I uses that same 500 pounds to buy means of consumption from II — a commodity worth 500 pounds.

II uses that same 500 pounds to buy means of production from I — a commodity worth 500 pounds.

I uses that same 500 pounds to buy means of consumption from II — a commodity worth 500 pounds.

Total value of commodities exchanged: 5,000 pounds.

The 500 pounds that II advanced to make its purchase has flowed back to it.

The result is this:

I now holds 1,000 pounds of variable capital in money form — the same sum it originally put into circulation. On top of that, I has spent 1,000 pounds on its own personal consumption, paid for out of its own commodity-product: that is, it spent the money it took in from selling 1,000 pounds' worth of means of production.

Meanwhile, the thing that variable capital in money form has to turn into — labour-power itself — has been kept alive and renewed by that consumption. It exists again as the one thing its owners have to sell if they want to go on living. So the relationship between wage-labourers and capitalists has been reproduced right along with it.

Second: II's constant capital has been replaced in its actual physical form, and the 500 pounds II advanced to circulation has flowed back to it.

For the workers of I, this whole circuit is the simple C-M-C: they sell a commodity (their labour-power), get money (the 1,000 pounds that is I's variable capital in money form), and use it to buy a commodity (necessary means of subsistence worth 1,000 pounds). That same 1,000 pounds is what turns into money — to the same value — the constant capital of II that exists in the form of commodities, namely means of subsistence.

For the capitalists of II, the process is C-M: part of their commodity-product turns into money, and out of that money it turns back into components of productive capital — specifically, part of the means of production they need.

When the capitalists of II advance that 500 pounds in money to buy the remaining part of their means of production, they're anticipating — getting in money form ahead of time — the value of the part of their own constant capital that is still sitting there as a commodity, as means of consumption, waiting to be sold. In this act, money (II's) turns into a piece of productive capital, while the commodity (I's) goes through its own conversion into money. But that money, for I, isn't a piece of its capital-value at all — it's monetized surplus-value, and it gets spent purely on means of consumption.

In the circuit M-C...P...C'-M', one capitalist's first move, M-C, is another capitalist's last move, C'-M' (or part of it). And it makes no difference at all to commodity circulation itself whether that commodity — the one that turns money into productive capital for the buyer — represents, for its seller, a piece of constant capital, a piece of variable capital, or surplus-value.

Look now at class I's own v+s: in money terms, it draws more out of circulation than it put in. First, its 1,000 pounds of variable capital comes back to it. Second, it sells means of production for 500 pounds (step 4 above) — that monetizes half its surplus-value. Then it sells means of production for another 500 pounds (step 6) — the second half of its surplus-value — and with that, the whole of its surplus-value has been pulled out of circulation in money form. Step by step: variable capital turned back into money, 1,000 pounds; half the surplus-value monetized, 500 pounds; the other half, 500 pounds; total monetized: 1,000v+1,000s = 2,000 pounds. So although I threw only 1,000 pounds into circulation — setting aside, for now, the exchanges that will later account for the reproduction of Ic — it has drawn out twice that amount. Of course, the monetized surplus-value doesn't stay in I's hands: it immediately passes into someone else's (II's), the moment I spends that money on means of consumption. And here is the point: the capitalists of I have drawn out in money no more value than they threw in as commodities. That this value happens to be surplus-value — that it cost the capitalists nothing to produce — changes absolutely nothing about the value of those commodities themselves. As far as the exchange of values within commodity circulation goes, it makes no difference whatsoever. The monetized form of surplus-value is, naturally, just as fleeting as every other form the advanced capital passes through along the way. It lasts only as long as the gap between commodity I turning into money and that money then turning into commodity II.

Had we assumed shorter turnover times — or, thinking of it simply as commodity circulation, a faster number of rounds for the circulating money — then even less money would be enough to circulate the same commodity-values. Given how many successive exchanges there are, the sum of money needed is always fixed by the total sum of prices — or of values — of the commodities in circulation. What share of that total value is surplus-value and what share is capital-value makes no difference to this at all.

Suppose, in our example, that I paid wages four times a year instead of once: 4×250=1,000. Then 250 pounds in money would be enough to handle the circulation of Iv against half of IIc, and the circulation between I's variable capital and its labour-power. In the same way, if the circulation between Is and IIc also happened in four rounds instead of two, only 250 pounds would be needed for that as well. Altogether, that's a sum of money — a money-capital — of just 500 pounds circulating 5,000 pounds' worth of commodities. And the surplus-value would then be monetized not twice, in two halves, but four times, in four quarters.

Suppose that instead of department II, in exchange 4, it is department I who buys — laying out £500 in money on means of consumption of the same value. Then in exchange 5, department II buys means of production with that same £500. In exchange 6, department I buys means of consumption with that same £500. In exchange 7, department II buys means of production with that same £500. So the £500 ends up back with department I, just as earlier it ended up back with department II. Here the surplus-value is turned into money by money that its own capitalist producers spend on their own private consumption — money that stands for revenue anticipated in advance, income drawn ahead of time against the surplus-value still sitting unsold in their commodities. But the surplus-value is not turned into money by the £500 coming back. Besides the £1,000 worth of commodities that represent department I's variable capital, department I had, at the end of exchange 4, thrown an extra £500 in money into circulation — money thrown in on top, not, as far as we know, proceeds from a commodity sold. If that money flows back to department I, all department I has gotten back is its own extra money — it has not turned its surplus-value into money. Department I's surplus-value is turned into money only by selling the commodities that embody it, and only for as long as the money that sale brings in has not been spent again on means of consumption.

Department I buys means of consumption from department II using that extra £500. It has now spent this money, and gotten an equivalent for it in department II's commodities. The money flows back to department I for the first time when department II turns around and buys £500 worth of commodities from department I. So the money flows back as the equivalent of the commodity department I sold — but that commodity cost department I nothing, so it counts as surplus-value for department I. This means the very money department I threw into circulation is what turns its own surplus-value into money. The same happens at its second purchase (no. 6): department I again gets its equivalent in department II's commodities. Now suppose department II does not go on, at no. 7, to buy means of production from department I. Then department I would in fact have paid out £1,000 for means of consumption — consuming its whole surplus-value as revenue: £500 of it in department I's own commodities, £500 in money. But it would still be sitting on £500 worth of unsold means-of-production commodities, and would have parted with £500 in money without getting it back.

Department II, meanwhile, would have converted three-quarters of its constant capital back out of commodity form — goods sitting for sale — into productive form, means of production actually in use. But one quarter would still sit as money-capital — £500 of idle money, money that has stopped functioning and is simply waiting. If this went on any longer, department II would have to cut back the scale of its reproduction by a quarter.

But the £500 worth of means of production that department I is left holding is not surplus-value sitting in commodity form. It stands in for the £500 in money that department I had advanced, on top of its £1,000 of surplus-value in commodity form. As money, that £500 is always realizable; as a commodity, it is for the moment unsellable.

One thing is clear: simple reproduction — where every element of productive capital in both department II and department I must be replaced — stays possible here only if the 500 golden birds fly back to department I, the department that first sent them flying.

Suppose a capitalist — here we are looking only at industrial capitalists, who also stand in for all the rest — spends money on means of consumption. For him, that money is simply gone, spent for good. If it ever comes back to him, that can only happen to the extent that he fishes it back out of circulation in exchange for commodities — that is, through his commodity-capital. Just as the value of his whole year's commodity output splits into constant capital-value, variable capital-value, and surplus-value, so does the value of each single commodity within it. So turning any one of those commodities into money is, at the same time, turning some portion of the surplus-value contained in the whole output into money. So it is, in this case, literally true that the capitalist himself threw the money into circulation — by spending it on means of consumption — and that this very money is what turns his surplus-value into money, that is, realizes it. Of course, these need not be the identical coins; it is a matter of an amount of hard cash equal to (or an equal share of) what he threw into circulation to cover his personal needs.

In practice, a capitalist advances money against his own future surplus-value in two different ways. If a business has only just opened this year, it takes a good while — a few months, at best — before the capitalist can pay for his own personal consumption out of the business's own earnings. But he does not put his consumption on hold even for a moment. He advances himself money — whether from his own pocket or borrowed from someone else's makes no difference here — against a surplus-value he has yet to capture. In doing so he also supplies the circulating money that will later realize that surplus-value. If, on the other hand, the business has already been running steadily for some time, then payments and receipts fall on different dates spread through the year. But one thing never stops: the capitalist's own consumption, which he anticipates in advance and sizes according to a fixed proportion of his usual or expected income. With every batch of commodities sold, part of the year's surplus-value also gets realized. But suppose that, over the whole year, only enough of the commodity produced were sold to replace the constant and variable capital-value it contains — or suppose prices fell so far that selling the entire year's output realized nothing but the advanced capital-value it contains. Then the anticipatory character of the money spent against future surplus-value would show through clearly. If our capitalist goes bankrupt, his creditors and the court examine whether his anticipated personal spending stood in proper proportion to the size of his business and to the surplus-value income that business normally brings in.

But looked at from the standpoint of the whole capitalist class, the claim that it must itself throw into circulation the money that realizes its own surplus-value (and also keeps its capital, both constant and variable, circulating) is not only not paradoxical — it is the necessary condition of the entire mechanism. Because there are only two classes here: the working class, which has nothing at its disposal but its labour-power, and the capitalist class, which holds the monopoly of society's means of production and of its money as well. The paradox would only arise if the working class had to be the ones advancing, out of their own resources, the money needed to realize the surplus-value sitting in the commodities. The individual capitalist, for his part, only ever makes this advance in the form of acting as a buyer: spending money to purchase means of consumption, or advancing money to purchase elements of his productive capital, whether labour-power or means of production. He only ever gives the money away in exchange for an equivalent. He advances money to circulation in exactly the same way he advances it commodities. In both cases, he is the starting point of that circulation.

What actually happens is obscured by two things.

First: merchant capital — whose starting form is always money, since the merchant as such produces no "product" or "commodity" of his own — and money-capital appear, within industrial capital's circulation process, as special objects that a distinct kind of capitalist manipulates.

Second: surplus-value — which must always land, in the first instance, in the hands of the industrial capitalist — then splits into different categories, whose bearers appear alongside the industrial capitalist: the landowner (drawing ground-rent), the moneylender (drawing interest), and so on, along with the government and its officials, rentiers, and the rest. These figures appear, facing the industrial capitalist, as buyers — and to that extent as the ones who turn his commodities into money. They too throw their proportional share of "money" into circulation, and he receives it from them. What always gets forgotten, in all this, is where they originally got that money from — and keep getting it from, again and again.

Engels notes that from this point the text is taken from Marx's Manuscript II.

One thing is still left to examine: department I's constant capital, 4,000 Ic. This value equals the value that reappears in department I's commodity-product — the value of the means of production used up in producing that mass of commodities. This reappearing value was not produced within department I's own production process. It entered that process a year earlier, as a given constant value already attached to its means of production. That value now sits in the whole part of commodity-mass I that department II has not absorbed — and the value of that part, remaining in the hands of the capitalists of department I, comes to two-thirds of the value of their entire annual commodity-product. For an individual capitalist producing one particular means of production, we could say this: he sells his commodity-product and turns it into money. In turning it into money, he also turns the constant value-part of his product back into money. With that money he then buys back his means of production from other sellers — or turns the constant value-part of his product into a natural form in which it can serve again as productive constant capital. Now, though, that assumption becomes impossible. The capitalist class of department I comprises the whole body of capitalists who produce means of production. And the 4,000 worth of commodity-product left in their hands is a part of the social product that cannot be exchanged for any other part — because no other part of the year's product is left to exchange it for. Apart from this 4,000, everything else has already been accounted for: one part has been absorbed into the social fund of consumption, and another part has to replace department II's constant capital, which has already handed over everything it has to offer in exchange with department I.

The difficulty resolves quite simply once you notice something: department I's whole commodity-product, in its natural form, consists of means of production — that is, of the very material stuff that makes up constant capital. The same thing we saw before with department II shows up here too, just from a different angle. There, in department II, the whole commodity-product consisted of means of consumption; one part of it — the part measured by the wages plus surplus-value contained in it — could be consumed by its own producers. Here, in department I, the whole commodity-product consists of means of production: buildings, machinery, vessels, raw and auxiliary materials, and so on. One part of it — the part that replaces the constant capital used up in this sphere — can therefore, in its natural form, go straight back into service as a piece of productive capital. Wherever it does enter circulation, that circulation stays inside class I. So: in department II, a part of the commodity-product is consumed in kind, individually, by its own producers; in department I, a part of the product is consumed in kind, productively, by its capitalist producers.

In the part of commodity-product I that equals 4,000c, the constant capital-value used up in this category reappears — and it reappears in a natural form that lets it go straight back into service as productive constant capital.

Compare department II: there, of its 3,000 commodity-product, the part whose value equals wages plus surplus-value (=1,000) goes directly into the personal consumption of II's capitalists and workers. But the constant capital-value of that same commodity-product (=2,000) cannot go back into the productive consumption of II's capitalists — it has to be replaced through exchange with I.

In department I, though, it works the other way. Of its 6,000 commodity-product, the part whose value equals wages plus surplus-value (=2,000) does not go into the individual consumption of its own producers, and given its natural form, it cannot: it must first be exchanged with department II.

The constant value-part of this same product, 4,000, is the reverse case: in its natural form, taking the whole capitalist class of department I together, it can go straight back into service as their constant capital.

In other words: the whole product of department I consists of use-values that, in their natural form, under capitalist production, can serve only as elements of constant capital. So of this 6,000-value product, one-third (2,000) replaces the constant capital of department II, and the remaining two-thirds replace the constant capital of department I itself.

Department I's constant capital is made up of a mass of different capital-groups, each invested in one of the various branches that produce means of production — so much in ironworks, so much in coal mines, and so on. Each of these capital-groups — each of these social group-capitals — is itself made up of a larger or smaller mass of individual capitals, each functioning on its own. Start from the top: society's total capital — say 7,500 (which could stand for millions) — splits into these different capital-groups. The social capital of 7,500 breaks into particular parts, each one invested in a particular branch of production. The part of the social capital-value invested in each particular branch consists, in its natural form, partly of the means of production belonging to that branch, and partly of the labour-power needed to run it and suited to the job — labour-power shaped in different ways by the division of labour, depending on the specific kind of work each particular branch requires. The part of the social capital invested in each particular branch, in turn, consists of the sum of the individual capitals invested in it, each functioning independently. This holds, of course, for both departments — for I just as much as for II.

Now take the constant capital-value that reappears in the shape of department I's commodity-product. Part of it goes straight back — as a means of production — into the very branch of production (or even the individual business) that it came out of as a product: grain back into growing grain, coal back into mining coal, iron in the shape of machines back into making iron, and so on.

But insofar as the separate products making up department I's constant capital-value do not go straight back into their own particular or individual sphere of production, they simply change places. They pass, in their natural form, into a different sphere of production within department I, while the products of other spheres within department I replace them in kind. It is nothing more than these products swapping locations. They all go back in as factors replacing constant capital in I — just in a different group of I than the one they left. Where exchange happens here, between the individual capitalists of I, it is an exchange of one natural form of constant capital for another — one kind of means of production for other kinds of means of production. It is an exchange among the different individual constant-capital parts of I themselves. Wherever the products do not serve directly as means of production in their own branch, they are moved from where they were produced to somewhere else, and in that way replace one another reciprocally. Put another way — similar to what happened with surplus-value in department II — each capitalist in I draws the means of production he needs out of this mass of commodities, in proportion to his share of ownership in this 4,000 of constant capital. If production were organized socially instead of capitalistically, it is clear that these products of department I would still, just as constantly, be distributed among this department's branches of production for the sake of reproduction: one part would stay directly in the sphere of production it came out of as a product, while another part would be moved to other places of production — so that a constant back-and-forth would take place between the different production sites of this department.

So the total value of the year's means of consumption equals: the variable capital of department II that the year reproduces, plus the new surplus-value department II produces — together, the whole value department II produces in the year — plus the variable capital of department I that the year reproduces, plus the new surplus-value department I produces — together, the whole value department I produces in the year.

So, assuming simple reproduction, the total value of the year's means of consumption equals the year's value product — the whole value society's labour produces in the year. And this must be so: under simple reproduction, the whole of that value gets consumed.

The whole social working day splits into two parts. First, necessary labour: over the year it creates a value of 1,500v. Second, surplus labour: it creates an extra value, a surplus-value, of 1,500s. These add up to 3,000 — the same as the value of the year's means of consumption, 3,000. So the total value of the year's means of consumption equals the total value the whole social working day produces in the year: the value of society's variable capital plus society's surplus-value — the whole year's new product.

But we already know that even though these two totals match in size, that does not mean the whole value of department II's goods — the means of consumption — was actually produced in that department. The two totals match because the constant-capital value that reappears in department II equals the value newly produced under department I — its variable capital plus surplus-value. That is why I(v+m) can buy the part of II's product that, for its own producers in department II, represents constant capital. This also shows why, although for the capitalists of department II the value of their product still splits into c + v + s, viewed socially that same value can be resolved into just v + s. But this only holds because IIc here equals I(v+m), and these two portions of the social product swap their physical forms when they're exchanged for each other. After the exchange, IIc exists again as means of production, while I(v+m) now exists as means of consumption.

It is this very fact that led Adam Smith to claim the value of the annual product resolves entirely into v + s. That claim holds, first, only for the part of the annual product made up of means of consumption. And second, it does not hold in the sense that this whole value is produced in department II, so that department II's product-value equals the variable capital II advanced plus the surplus-value II produces. It holds only in the sense that II(c+v+m) = II(v+m) + I(v+m) — only because IIc equals I(v+m).

It also follows:

The whole social working day — the labour the entire working class spends over the year — splits, like any single day's labour, into just two parts: necessary labour and surplus labour. So the value it produces also splits into just two parts: variable capital value (the part the worker uses to buy their own means of subsistence) and surplus-value (the part the capitalist can spend on their own consumption). Even so, viewed socially, part of the social working day is spent exclusively on producing fresh constant capital — products destined only to serve, in the labour process, as means of production, and so, in the accompanying valorization process, as constant capital. On our assumption, the whole social working day comes to a money value of 3,000, of which only a third — 1,000 — is produced in department II, the department that produces means of consumption, the goods in which the whole of society's variable capital value and surplus-value is finally realized. So, on this assumption, two thirds of the social working day go into producing new constant capital. From the standpoint of the individual capitalists and workers of department I, these two thirds merely serve to produce variable capital value plus surplus-value — exactly like the last third of the social working day in department II. Even so, viewed socially — and equally viewed in terms of the product's use-value — these two thirds of the social working day produce nothing but replacement for constant capital that is being used up in productive consumption. Even viewed individually, these two thirds of the working day do produce a total value equal, for their own producers, only to variable capital value plus surplus-value. But they produce no use-values of the kind wages or surplus-value could actually be spent on: their product is a means of production.

First, notice this: no part of the social working day, in either department, goes to producing the value of the constant capital already at work — already functioning — in these two great spheres of production. What they produce is only additional value: 2,000 I(v+m) plus 1,000 II(v+m), on top of the constant capital value of 4,000 Ic plus 2,000 IIc. The new value produced in the shape of means of production is not yet constant capital. It is only destined to function as constant capital in future.

Department II's whole product — the means of consumption — considered concretely, by use-value, in its physical form, is the product of the third of the social working day that department II performed. It is the product of labour in its concrete form — weaving, baking, and so on — the labour actually employed in that department, insofar as that labour functions as the active element of the labour process. But the constant part of this product's value is different. It only reappears in a new use-value, a new physical form — the form of means of consumption — whereas before it existed in the form of means of production. Its value has simply been carried over, by the labour process, from its old physical form into its new one. This part of the product's value — two thirds of it, 2,000 — was not produced in this year's valorization process in department II.

Just as, from the standpoint of the labour process, department II's product is the result of newly active living labour together with its own given, presupposed means of production — the objective conditions in which that labour realizes itself — so, from the standpoint of the valorization process, the value of department II's product, 3,000, is made up of two parts. One is new value, produced by the newly-added third of the social working day: 500v + 500s = 1,000. The other is a constant value, in which two thirds of a past social working day — one that elapsed before this year's production process in department II — is objectified. This part of the product's value shows up as part of the product itself: it exists in a quantity of means of consumption worth 2,000, equal to two thirds of a social working day. That is the new use-form in which it reappears. So when part of the means of consumption — 2,000 IIc — is exchanged for means of production I(1,000v + 1,000s), what is really being exchanged is two thirds of a total working day that forms no part of this year's labour but elapsed before this year, against two thirds of this year's own, newly-added working day. Two thirds of this year's social working day could not be used to produce constant capital and, at the same time, form variable capital value plus surplus-value for their own producers — unless they were exchanged against a portion of the value of the year's consumed means of consumption, a portion in which two thirds of a working day spent and realized before this year, not within it, was lodged. It is an exchange of two thirds of this year's working day against two thirds of a working day spent before this year — an exchange between this year's labour-time and last year's. This, then, solves the riddle: why can the value-product of the whole social working day resolve into variable capital value plus surplus-value, even though two thirds of that working day was not spent producing things in which variable capital or surplus-value can be realized, but rather producing means of production to replace the capital used up during the year? The explanation is simply this: two thirds of department II's product-value — the two thirds in which the capitalists and workers of department I realize the variable capital value plus surplus-value they produced, and which make up two thirds of the whole year's product-value — considered by value, are the product of two thirds of a social working day that elapsed before this year.

Take the sum of the social product of departments I and II together — means of production and means of consumption. Considered concretely, by use-value, in physical form, this whole is indeed the product of this year's labour. But only in the sense that this labour counts as useful, concrete labour — not in the sense that it counts as an expenditure of labour-power, as value-forming labour. And even that first sense holds only because the means of production were turned into new product — this year's product — by the living labour added to them, working on them. The reverse is equally true: this year's labour could not have turned itself into a product without means of production independent of it — without instruments of labour and materials to work on.

About the whole product's value of 9,000, and the categories it splits into — working this out is no harder than working out the value of an individual capital's product. In fact it's exactly the same task.

The whole year's social product here contains three social working days — each one standing for a full year of society's combined labour. Each of these working days is worth 3,000. So the value of the total product is three times 3,000, which is 9,000.

Some of this labour, though, had already been spent before the one-year production process we're looking at even began: in department I, 4/3 of a working day (worth 4,000), and in department II, 2/3 of a working day (worth 2,000). Together that's two whole social working days from the past, worth 6,000. That is why 4,000 Ic plus 2,000 IIc equals 6,000c — the value of the means of production reappearing in the total product, the constant capital value.

Now take the labour newly added this year. In department I, a third of the social working day is necessary labour — labour that replaces the 1,000 of variable capital and pays for the labour used in department I. In department II, a sixth of the social working day is likewise necessary labour, worth 500. So 1,000 Iv plus 500 IIv equals 1,500v. That is the value of half of this year's newly added social working day — the half made up of necessary labour.

Finally, in department I a third of the whole working day, worth 1,000, is surplus labour; in department II a sixth of the day, worth 500, is also surplus labour. Together these make up the other half of this year's newly added working day. So the total surplus-value produced is 1,000 Is plus 500 IIs, which is 1,500s.

So:

So the difficulty does not lie in working out the value of the social product itself. It arises when we compare the value-parts of the social product with its physical, material parts.

The constant part of the value — the part that merely reappears — equals the value of the portion of the product made up of means of production, and it is embodied in that portion.

The new value produced this year — v plus m — equals the value of the portion of the product made up of means of consumption, and it is embodied in that portion.

Apart from exceptions that don't matter here, means of production and means of consumption are completely different kinds of goods. They have completely different natural forms, completely different use-forms — so they are also products of completely different kinds of concrete labour. The labour that uses machines to produce food is nothing like the labour that builds those machines.

This creates the appearance of a puzzle. The whole year's social working day, worth 3,000, seems to be spent entirely on producing means of consumption worth 3,000 — and no constant value reappears in them, since this 3,000 (1,500v + 1,500s) resolves into nothing but variable capital and surplus-value. Meanwhile, the constant capital value of 6,000 reappears in a completely different kind of product, the means of production — even though no part of the social working day seems to have gone into producing these new products at all. The whole working day seems to consist only of the kinds of labour that end up in means of consumption, not in means of production.

But the puzzle is already solved. The value-product of the year's labour equals the value of department II's product, the total value of the newly produced means of consumption. But that product-value is bigger than the part of the year's labour actually spent producing means of consumption — bigger by two thirds of itself, because only a third of the year's labour went into producing them. Two thirds of this year's labour was spent producing means of production — that is, in department I.

The value-product created during that time in department I — equal to the variable capital value plus surplus-value produced there — equals the constant capital value of II that reappears in the means of consumption. So the two can be exchanged for each other and replace each other in kind. The total value of department II's means of consumption is therefore equal to the sum of the new value-product of I and II together — II(c+v+m) = I(v+m) + II(v+m) — that is, equal to the sum of the new value this year's labour produced in the form of wages plus surplus-value.

On the other hand, the total value of the means of production (department I) equals the sum of the constant capital value that reappears in the form of means of production (I) plus the constant capital value that reappears in the form of means of consumption (II) — in other words, it equals the whole constant capital value that reappears in the total social product. This total value equals the value of 4/3 of a working day that had already passed, before this production process, in department I, plus 2/3 of a working day that had already passed in department II — together, two whole working days.

So the difficulty with the social yearly product comes from this: the constant part of its value shows up in a completely different kind of product — means of production — than the new value (v+s) added to it, which shows up in means of consumption. This creates the appearance that, in terms of value, two-thirds of the product used up in the year has reappeared in a new form, as a new product, without society spending any labour at all to produce it. That never happens with an individual capital. Every individual capitalist applies one particular kind of concrete labour, which turns its own particular means of production into a product. Say the capitalist is a machine-builder. The constant capital spent during the year is 6,000c, the variable capital 1,500v, the surplus-value 1,500s; the product is 9,000 — say, 18 machines, each worth 500. The whole product here takes the same form throughout: machines. (If he made several kinds, each would be reckoned separately.) The whole commodity-product is the product of the labour spent during the year in machine-building — the same kind of concrete labour, combined with the same means of production. So the different parts of the product's value show up in the very same natural form: 6,000c is contained in 12 machines, 1,500v in 3 machines, 1,500s in 3 machines. Now here's a subtlety worth catching: the 12 machines are worth 6,000c, but not because those particular 12 machines are simply made of labour spent before this year's machine-building and not used up in it. The value of the means of production for 18 machines has not simply turned itself into 12 machines. Rather, the value of these 12 machines — itself made up of 4,000c + 1,000v + 1,000s, the same mix as any of the 18 — happens to add up to the same total as all the constant capital value spread across the 18 machines. So the machine-builder must sell 12 of his 18 machines in order to replace the constant capital he spent — the constant capital he needs to make 18 new machines. The case would be inexplicable, on the other hand, if the labour applied consisted purely of machine-building, yet its result turned out to be: on one side, 6 machines worth 1,500v + 1,500s, and on the other side, iron, copper, screws, belts and so on worth 6,000c — that is, the means of production for the machines in their own natural form, which the individual machine-building capitalist, as everyone knows, does not produce himself but must replace through the circulation process. And yet, at first glance, this is exactly the senseless way the reproduction of the social yearly product seems to proceed.

The product of an individual capital — that is, of any fragment of the social capital that functions on its own, with a life of its own — can take any natural form whatever. The only condition is that it actually has a use-form, a use-value, marking it fit to circulate in the world of commodities. Whether it can go back as a means of production into the very same process it came out of — whether, in other words, the part of its product-value that represents the constant capital has a natural form in which it can actually function again as constant capital — is entirely indifferent and a matter of chance. If it cannot, this part of the product's value is turned back, through sale and purchase, into the form of its material elements of production, and the constant capital is thereby reproduced in a natural form fit to function.

It is different with the product of the total social capital. All the material elements of reproduction must, in their natural form, themselves form parts of this product. The constant capital used up can be replaced by the total production only to the extent that the whole reappearing constant capital value shows up in the product in the natural form of new means of production that can actually function as constant capital. Assuming simple reproduction, the value of the part of the product made up of means of production must therefore equal the constant value-part of the social capital.

Further: seen individually, the newly added labour produces, within the capitalist's product-value, only his variable capital plus surplus-value — while the constant part of the value is carried over onto the product by the concrete character of that same newly added labour.

Seen socially, the picture is different. The part of the social working day that produces means of production adds new value to them and also carries over onto them the value of the means of production used up in making them — but what it produces is nothing but new constant capital, meant to replace the constant capital used up in the form of the old means of production, both the constant capital consumed in department I and in department II. It produces only product meant to fall into productive consumption. So the whole value of this product is only value that can function again as constant capital, that can only buy back constant capital in its natural form — value that, seen socially, resolves into neither variable capital nor surplus-value. The part of the social working day that produces means of consumption, on the other hand, produces no part of the social replacement capital at all. It produces only products whose natural form is meant to realize the value of the variable capital and the surplus-value of both department I and department II.

When we speak of the social point of view — when we look at the whole social product, which includes both the reproduction of the social capital and individual consumption — we must not fall into the manner Proudhon copied from bourgeois economics: treating capitalist society en bloc, as one totality, as if it thereby lost its specific, historically economic character. Just the opposite. What we are dealing with then is the total capitalist. The total capital appears as the joint-stock capital of all the individual capitalists put together. This joint-stock company has one thing in common with many other joint-stock companies: each shareholder knows what he puts in, but not what he draws out.

The whole social product for the year is worth 9,000: 6,000c+1,500v+1,500s. Put differently, 6,000 of that value reproduces the value of the means of production, and 3,000 reproduces the value of the means of consumption. So the value of society's revenue — wages plus surplus-value, v+s — comes to only a third of the whole product's value. That third is the most that everyone together, workers and capitalists alike, can draw out of the total social product and add to their own consumption. The other 6,000 — two-thirds of the product's value — is the value of the constant capital, and it must be replaced in kind. Means of production to that same amount have to go back into the production fund. This is exactly what Storch recognizes as necessary, without being able to prove it:

Storch put it this way: the value of a year's product splits into capitals on one side and profits on the other, and each of these two parts regularly buys back whatever the nation needs — the capital part to keep the nation's capital going, the profit part to renew what people consume. The products that make up a nation's capital, he added, cannot be consumed at all.

A. Smith is the one who set up this extraordinary dogma, still believed today — and not only in the form already met, that the whole value of the social product resolves into revenue, wages plus surplus-value, or as he puts it, wages plus profit (interest) plus rent. He set it up in an even more popular form too: that consumers, in the end, must pay producers the whole value of the product. This is still, today, one of the best-attested commonplaces — one of the supposed eternal truths — of political economy. The illustration runs like this: take some article, linen shirts say. First, the spinner of linen yarn has to pay the flax-grower the whole value of the flax: flax-seed, manure, feed for the draught animals, and so on, plus the share of the flax-grower's fixed capital — buildings, farm tools — that this crop uses up; the wages paid in growing the flax; the surplus-value, profit and rent, sitting inside the flax; and finally the freight from the field to the spinning-mill. Then the weaver has to pay the spinner back not just that price of the flax, but also the share of machinery, buildings and the rest of the spinner's fixed capital that gets passed on, plus all the materials used up in spinning, the spinners' wages, their surplus-value, and so on. The same continues with the bleacher, then the cost of carrying the finished linen, and finally the shirt-maker, who has now paid the whole price run up by every earlier producer — producers who, between them, supplied nothing but his raw material. In the shirt-maker's own hands, more value is added again: partly the constant capital used up as tools and materials in making the shirts, partly the labour spent there, which adds the shirt-workers' wages plus the shirt-maker's surplus-value. Say the whole batch of shirts finally costs £100, and that is society's whole outlay on shirts for the year. The people who buy the shirts pay that £100 — which is the value of every means of production that went into the shirts, plus the wages and surplus-value of the flax-grower, the spinner, the weaver, the bleacher, the shirt-maker, and everyone who carried the goods along the way. All of this is completely true. It is exactly what any child can see. But then the claim goes further: so it is with the value of every other commodity. It should say instead: so it is with the value of every means of consumption — with the value of the share of the social product that goes into the consumption fund, the share of the social product's value that can be spent as revenue at all. The sum of value of all these goods is indeed equal to the value of every means of production used up in making them, plus the value the labour just added — wages plus surplus-value. So all consumers together can pay this whole sum — because although each single commodity's value is made of c+v+s, the total value of everything that goes into the consumption fund can, at most, only equal the share of the social product's value that resolves into v+s: equal, that is, to the value a year's labour has added to the means of production it found already there, and not to the value of that constant capital itself. But as for the value of the constant capital itself — we have already seen it gets replaced out of the social mass of products in two ways. First, through exchange between the capitalists of department II, who make means of consumption, and the capitalists of department I, who make the means of production for them. This is where the phrase comes from, that what is capital for one is revenue for another. But that is not how it actually stands. The 2,000 IIc, sitting in means of consumption worth 2,000, is constant capital value for the capitalist class of department II. They cannot consume it themselves, even though, in its natural form, the product must be consumed by somebody. On the other side, 2,000 I(v+m) is the wages plus surplus-value produced by the capitalists and workers of department I. It exists in the natural form of means of production — things whose own value cannot be consumed. So here we have a sum of value of 4,000, of which, before the exchange as after it, one half only ever replaces constant capital and the other half only ever forms revenue. Second, though: the constant capital of department I is replaced in kind — partly through exchange among the capitalists of department I themselves, partly through each individual business replacing its own in kind.

The claim that the whole year's product-value must, in the end, be paid by consumers would only be true if 'consumers' were made to cover two quite different kinds: individual consumers and productive consumers. But to say that part of the product must be consumed productively means nothing more than that it has to function as capital — it cannot be used up as revenue.

Suppose we split the value of the whole product, 9,000, into 6,000c+1,500v+1,500s, and look at the 3,000 (v+s) purely as revenue. Then, the other way round from before, variable capital seems to disappear, and capital, looked at socially, seems to consist of constant capital alone. Because what first appeared as 1,500v has, on this view, dissolved into a piece of society's revenue — wages, the revenue of the working class — making its character as capital appear to vanish. This is exactly the conclusion Ramsay draws. For him, capital, looked at socially, consists only of fixed capital — but by 'fixed capital' he means constant capital: the mass of value sitting in means of production, whether those means of production are instruments of labour or material — raw material, semi-finished goods, auxiliary materials, and so on. He calls variable capital 'circulating' instead:

Ramsay wrote: circulating capital is nothing but the food and other necessities advanced to workers before their labour's product is finished. Fixed capital alone — not circulating capital — is, properly speaking, a source of national wealth. Circulating capital is not directly involved in production at all, and is not even essential to it; it is merely a convenience made necessary by the wretched poverty of the mass of the people. Fixed capital alone counts, from a national point of view, as an element of the cost of production.

Ramsay explains more closely what he means by fixed capital — which is to say, constant capital:

Ramsay wrote: what matters is the length of time some portion of the product of that labour — meaning labour spent on making a commodity — has existed as fixed capital: that is, in a form which, although it helps to bring the future commodity into being, does not support any workers.

Ramsay's definitions show, once again, the damage Adam Smith did: in his hands, the distinction between constant and variable capital gets drowned in the distinction between fixed and circulating capital. Ramsay's 'fixed capital' is just his name for constant capital — the instruments of labour — and his 'circulating capital' is just his name for variable capital — the means of subsistence. Swapping the names does not clear up the confusion. Both of Ramsay's capitals are simply commodities of a given value, and neither one can produce surplus-value any more than the other.

Engels notes that from this point the text is taken from Marx's Manuscript VIII.

The whole of this year's reproduction — the whole product of this year — is the product of this year's useful labour. But the value of that whole product is bigger than the part of its value in which this year's labour, as labour-power spent during the year, is embodied. The value product of this year — the value newly created in commodity form during the year — is smaller than the product-value: the total value of the whole mass of commodities made over the whole year. Take the total value of the year's product and subtract the value that this year's current labour added to it: what remains is not value that was really reproduced. It is only value that reappears in a new form of existence — value carried over onto this year's product from value that already existed before it. Depending on how long the constant-capital components lasted that took part in this year's social labour process, that value may be of an earlier or a later date; it may come from a means of production that came into being last year, or in some earlier year. Whatever the case, it is value carried over from previous years' means of production onto the product of the current year.

Now take our schema. After the exchange of the elements we have looked at so far — between department I and department II, and within department II — we have:

The value newly produced during the year lies only in the v and the s. So the sum of this year's value product equals the sum of v + s: 2,000 I(v+s) + 1,000 II(v+s) = 3,000. Every other part of this year's product-value is only transferred value — value carried over from earlier means of production used up in this year's production. Beyond that value of 3,000, this year's current labour has produced no value at all. That 3,000 is its whole annual value product.

Now, as we saw, the 2,000 I(v+s) replace department II's 2,000 IIc in the natural form of means of production. So two-thirds of the year's labour, spent in department I, have newly produced the constant capital of department II — both its whole value and its natural form. Socially considered, then, two-thirds of the labour spent during the year has created new constant-capital value, realized in the natural form that suits department II. So the greater part of society's annual labour has gone into producing new constant capital — capital-value existing in means of production — to replace the constant-capital value spent in producing consumption goods. What distinguishes capitalist society from the savage here is not, as Senior thinks, that it is the savage's special privilege and peculiarity to spend his labour for a certain time without getting any fruits from it that can be turned into revenue — that is, into consumption goods. The difference lies here instead:

a) Capitalist society spends more of its available yearly labour producing means of production — that is, constant capital — value that cannot be resolved into revenue, whether as wages or as surplus-value, but can only function as capital.

b) When the savage makes bows, arrows, stone hammers, axes, baskets and so on, he knows perfectly well that he has not spent that time making consumption goods — that all he has done is cover his need for means of production, and nothing more. Besides, the savage commits a serious economic sin through his complete indifference to how much time a thing costs: sometimes, as one anthropologist reports, he spends a whole month making a single arrow.

There is a common idea that some political economists use to shrug off the real theoretical difficulty — that is, to avoid actually understanding how things really connect: that what is capital for one person is revenue for another, and the other way round. This idea is partly right. But stated as a general rule, it becomes completely wrong — it then contains a total misunderstanding of the whole process by which things change hands in the course of annual reproduction, and so also a misunderstanding of the real basis for the part of it that is right.

We can now set out the actual relations that this partly-right idea rests on — and in doing so, the mistaken way of understanding those relations will show itself too.

1. Variable capital functions as capital in the capitalist's hands, and functions as revenue in the wage-worker's hands.

Variable capital exists, at first, in the capitalist's hands as money-capital; it functions as money-capital in that he uses it to buy labour-power. As long as it stays in his hands in money form, it is nothing but a given value existing in money form — a constant quantity, not a variable one. It is only potentially variable capital, simply because it is capable of being converted into labour-power. It becomes really variable capital only once it sheds its money form — once it has been converted into labour-power, and that labour-power is functioning as a component of productive capital in the capitalist process.

The same money that first functioned, for the capitalist, as the money-form of variable capital, now functions in the worker's hands as the money-form of his wages, which he converts into means of subsistence — that is, as the money-form of the revenue he draws from constantly repeated sales of his labour-power.

All we have here is the simple fact that the buyer's money — the capitalist's — passes out of his hands into the hands of the seller, here the seller of labour-power, the worker. It is not the variable capital that functions twice over, as capital for the capitalist and as revenue for the worker. It is the same money: money that, in the capitalist's hands, first exists as the money-form of his variable capital, and so only as potentially variable capital, and that, once the capitalist has converted it into labour-power, serves in the worker's hands as the equivalent for labour-power sold. But that the same money serves one use in the seller's hands and a different use in the buyer's hands — that belongs to every purchase and sale of commodities whatever.

Apologist economists get this wrong, and the mistake shows up most clearly if we look only at the bare act of circulation — M-C, money turning into labour-power, on the buyer's side, the capitalist; and C-M, the commodity labour-power turning into money, on the seller's side, the worker — and set aside, for now, what happens next. They say: here the same money brings two capitals into being. The buyer, the capitalist, converts his money-capital into living labour-power, which he incorporates into his productive capital. The seller, the worker, meanwhile converts his commodity, labour-power, into money, which he spends as revenue — and that is exactly what lets him keep selling his labour-power again and again, and so keep himself alive. So, on this view, his labour-power is itself his 'capital in commodity form', the constant source of his revenue. In fact labour-power is his asset — one that keeps renewing and reproducing itself — not his capital. It is the one commodity he can and must keep selling in order to live, and it works as capital (variable capital) only once it is in the buyer's, the capitalist's, hands. That a man is constantly forced to keep selling his labour-power — that is, to keep selling himself — to a third person proves, according to those economists, that he is a capitalist, because he constantly has a 'commodity' (himself) to sell. On this reasoning even the slave becomes a capitalist, even though he is sold once and for all, as a commodity, by a third party — because this commodity, the labouring slave, is by its very nature such that its buyer not only makes it work anew every day, but also gives it the means of subsistence that let it go on working again and again. (Other writers have made this comparison too.)

In the exchange of 1,000 Iv + 1,000 Is against 2,000 IIc, then, what is constant capital for one side (2,000 IIc) is variable capital and surplus-value — revenue, in other words — for the other side. And what is variable capital and surplus-value for one side (2,000 I(v+m)) — revenue, in other words — becomes constant capital for the other side.

Let's look first at the exchange of Iv against IIc — starting from the worker's standpoint.

The whole body of workers in department I have sold their labour-power to the whole body of capitalists in department I for 1,000; they receive this value paid out to them in money, as wages. With this money they buy means of consumption from department II, to the same value. Capitalist II stands opposite them purely as a seller of commodities, nothing more — even where, as with the 500 IIv exchange discussed earlier, a worker happens to buy from his own capitalist. The circulation their commodity — labour-power — passes through is the simple form aimed only at satisfying needs, at consumption: commodity (labour-power) — money — commodity (means of consumption, commodity II). The result of this circuit is that the worker has kept himself in being as labour-power for capitalist I; and to go on keeping himself in being as labour-power, he must keep repeating this same process. His wage is realized in means of consumption — it is spent as revenue, and, taking the working class as a whole, it is spent as revenue over and over, without end.

The whole commodity-product of department II is made up of means of consumption — things meant to go into yearly consumption, meant to realize somebody's revenue. Here, that somebody is the whole body of workers in department I. But for the whole body of capitalists in department II, part of that same commodity-product — worth 2,000 — is something else: it is the constant capital-value of their productive capital, now sitting in commodity-form. It has to be converted back out of that commodity-form into its natural form, so it can go back to work as the constant part of productive capital. So far, what capitalist II has achieved is this: by selling to worker I, he has turned half (1,000) of his constant capital-value — currently sitting in commodity-form as means of consumption — back into money-form. It was not variable capital Iv that bought this first half of constant capital IIc. What happened is that the money which had functioned for I as money-capital, in the purchase of labour-power, passed into the hands of the seller of that labour-power — for whom it is not capital at all but revenue in money-form, meant to be spent buying means of consumption. That same money — the 1,000 that has now flowed to capitalist II from the workers of I — cannot, on II's side, function as a constant element of his productive capital. It is still only the money-form of his commodity-capital, still waiting to be turned into the fixed or circulating pieces of constant capital. So II takes this money, realized from the workers of I who bought his goods, and uses it to buy 1,000 worth of means of production from I. That renews half the total value of constant capital II, in the natural form it needs to function again as an element of productive capital II. The circuit here was: means of consumption worth 1,000 — money worth 1,000 — means of production worth 1,000.

But this movement — commodity, money, commodity — is a movement of capital here. The commodity, sold to the workers, turns into money, and that money is converted into means of production: a re-conversion from commodity-form back into the material elements that make up that commodity. On the other side: just as capitalist II, facing I, acts only as a buyer of commodities, capitalist I, facing II, acts here only as a seller of commodities. I originally used 1,000 in money — money meant to function as variable capital — to buy labour-power worth 1,000. So I received an equivalent for the 1,000v he had paid out in money-form. That money now belongs to the worker, who spends it buying from II. I can only get this money back — the money that has now landed in II's till — by fishing it back out again, through selling goods to the same value.

At first I had a definite sum of money, 1,000, meant to function as the variable part of his capital; it functions as such by being exchanged for labour-power to the same value. But as the result of the production process, the worker has delivered to him a mass of commodities (means of production) worth 6,000, of which one-sixth — 1,000 — is, by value, an equivalent of the variable capital-part he had advanced in money. The variable capital-value functions as variable capital now, in its commodity-form, no more than it did before in its money-form: it can only function as variable capital once it has actually been exchanged for living labour-power, and only for as long as that labour-power is at work in the production process. As money, the variable capital-value was only potential variable capital. But it was in a form directly convertible into labour-power. As a commodity, this same variable capital-value is now only a potential money-value; it is turned back into its original money-form only once the commodity is sold — here, once II buys 1,000 worth of goods from I. The circulation movement here is: 1,000v in money — labour-power worth 1,000 — 1,000 in commodities (the equivalent of the variable capital) — 1,000 in money again. That is: money — commodity ... commodity — money — in other words, money — labour-power ... commodity — money. The production process that falls between the two commodity-stages does not itself belong to the sphere of circulation; it does not appear in the exchange of the different elements of the year's reproduction against one another — even though that exchange includes the reproduction of every element of productive capital, both its constant part and its variable part, labour-power. Everyone carrying this exchange appears only as a buyer or a seller, or as both: the workers appear in it only as buyers of commodities; the capitalists appear alternately as buyers and sellers; and, within certain limits, sometimes only as buyers of commodities, sometimes only as sellers of commodities.

The result: I once again holds the variable part of his capital's value in money-form — the only form it can be directly converted into labour-power from, that is, the only form in which it can actually be advanced as the variable element of his productive capital. On the other side, before the worker can appear again as a buyer of commodities, he must first appear again as a seller of commodities — as a seller of his labour-power.

With the variable capital of category II (500 IIv), the circulation process between the capitalists and the workers of the same branch of production takes an unmediated form — so long as we look at it as running directly between the whole body of capitalists II and the whole body of workers II.

The whole body of capitalists II advances 500v to buy labour-power worth the same amount; here the capitalist is the buyer, the worker the seller. Then the worker, with the money he got for his labour-power, appears as a buyer of part of the very commodities he himself produced. Here, then, the capitalist is the seller. The worker has given the capitalist back the money he was paid for his labour-power, in the form of part of the produced commodity-capital II — namely 500v worth of goods. Before the worker spends it, the capitalist holds that same 500v in commodity-form, where before buying labour-power he had held it in money-form. The worker, for his part, has realized the value of his labour-power in money, and now realizes that money again by spending it — as revenue, to cover his own consumption — buying part of the very means of consumption he produced. This is an exchange of the worker's revenue, in money, against the 500v portion of goods that he himself reproduced in commodity-form for the capitalist. That is how this money returns to capitalist II as the money-form of his variable capital. An equivalent amount of revenue-value, in money-form, here replaces variable capital-value that had been sitting in commodity-form.

The capitalist does not get richer by taking back, through selling the worker an equivalent mass of goods, the very money he paid the worker to buy his labour-power. He would in fact be paying the worker twice over if he first paid him 500 to buy his labour-power and then, on top of that, handed him for nothing the 500 worth of goods he had made the worker produce. Conversely, if all the worker had produced for him was a bare equivalent in goods — 500 — matching the 500 price of his labour-power, the capitalist would stand, after the operation, at exactly the same point as before it. But the worker has in fact reproduced a product worth 3,000. He has restored the constant value-part of the product — the value of the means of production used up in it, = 2,000 — by converting them into a new product. And beyond that given value, he has added a further value of 1,000 (v+s). (The notion that the capitalist enriches himself — in the sense of gaining surplus-value — through this reflux of the 500 in money is Destutt de Tracy's; it is examined at length below, in Section XIII of this chapter.)

Through this purchase of means of consumption worth 500 by worker II, the value of 500 IIv — which capitalist II had, a moment ago, only in commodity-form — flows back to him in money, in the very form in which he originally advanced it. The immediate result of the transaction, as with any sale of commodities, is simply the conversion of a given value out of commodity-form into money-form. And the reflux of money to its starting point that this brings about is nothing special either. Had capitalist II instead bought goods worth 500 in money from capitalist I, and then sold goods worth 500 to I in turn, 500 in money would equally have flowed back to him. That 500 in money would only have served to circulate a mass of commodities worth 1,000, and — by the general law already established — would have flowed back to whoever had thrown that money into circulation to circulate this mass of commodities.

But the 500 in money that has flowed back to capitalist II is, at the same time, renewed potential variable capital in money-form. Why is that? Money — and so money-capital too — is only potential variable capital because, and to the extent that, it can be converted into labour-power. The return of £500 to capitalist II is accompanied by the return of labour-power II to the market. Both returns, at opposite poles, are conditioned by one and the same process — which is also why the 500 reappears not just as money, but as variable capital in money-form. The money = 500 flows back to capitalist II because he has sold means of consumption worth 500 to worker II — in other words, because the worker has spent his wage, and by doing so has kept himself and his family, and with them his own labour-power, in being. To go on living, and to be able to appear again as a buyer of commodities, he must sell his labour-power afresh. So the return of the 500 in money to capitalist II is, at the same time, the return — or rather, the continued availability — of labour-power as a commodity that the 500 can buy, and so also the return of the 500 as potential variable capital.

For category IIb, which produces luxury goods, their variable capital — (IIb)v — works the same way as Iv does. The money that renews their variable capital in money-form for capitalists IIb flows to them by a detour, through the hands of capitalists IIa. But even so, it makes a difference whether the workers buy their means of subsistence directly from the capitalist producers they sold their labour-power to, or whether they buy from a different category of capitalists, so that the money only flows back to the first group by a roundabout route. The working class lives from hand to mouth, so it buys as long as it can buy. It is different for the capitalist — take, for instance, the exchange of 1,000 IIc against 1,000 Iv. The capitalist does not live from hand to mouth: what drives him is getting the greatest possible return on his capital. So if circumstances of any kind make it seem more advantageous to capitalist II to hold at least part of his money for a while, rather than immediately renewing his constant capital, then the reflux of the 1,000 IIc (in money) to I is delayed — and with it, the restoration of 1,000v in money-form. Capitalist I can then only keep working on the same scale if he has reserve money available: reserve capital in money is needed in general, so that production can carry on without interruption regardless of whether the variable capital-value flows back faster or slower.

When we examine the exchange between the different elements of this year's ongoing reproduction, we are also examining the result of last year's labour — the labour of a year already closed out. The production process that resulted in this year's product lies behind us; it is past, absorbed into its product — and so, even more, is the circulation process that precedes or runs alongside production: the exchange of potential into actual variable capital, that is, the purchase and sale of labour-power. The labour market forms no part of the commodity market we have before us here. By this point the worker has already not only sold his labour-power, but delivered — beyond the surplus-value — an equivalent of the price of his labour-power in commodity-form; he, meanwhile, has his wage in his pocket and figures in this exchange only as a buyer of commodities (means of consumption). On the other hand, the year's product must contain every element of reproduction — it must restore every element of productive capital, and above all its most important element, variable capital. And we have indeed seen what the exchange yields, with respect to variable capital: as a buyer of commodities, by spending his wage and consuming the goods he buys, the worker maintains and reproduces his labour-power — the one commodity he has to sell. Just as the money the capitalist advanced to buy this labour-power flows back to him, so too does the labour-power itself, as the commodity that money can buy, flow back onto the labour market. As a result — here, specifically, for the case of 1,000 Iv — we get: 1,000v in money on the side of the capitalists of I, facing labour-power worth 1,000 on the side of the workers of I, so that the whole reproduction process of I can start over again. This is one result of the exchange process.

On the other hand, the spending of the wages of the workers of I has taken 1,000 worth of means of consumption off II's hands, turning it from commodity-form into money-form. Out of that money-form, II has converted it back into the natural form of his constant capital, by buying goods worth 1,000v from I — and this is how I's variable capital-value flows back to him in money-form.

The variable capital of I passes through three transformations — transformations that, in the exchange of the year's product, either do not appear at all, or appear only by hint.

1. The first form: 1,000 Iv in money, exchanged for labour-power to the same value. This exchange does not itself appear in the exchange of commodities between I and II — but its result does: the working class of I confronts the commodity-seller II holding 1,000 in money, just as the working class of II confronts the seller of the 500 IIv commodities holding 500 in money.

2. The second form — the only one in which variable capital really varies, really functions as variable, the one where value-creating power stands in for the given, fixed value that was exchanged for it — belongs entirely to the production process that now lies behind us.

3. The third form — the one in which variable capital has proved itself as such, in the result of the production process — is the year's value-product: for I, this is 1,000v + 1,000s = 2,000 I(v+m). In place of its original value of 1,000 in money, a value twice as large — 2,000 — has appeared, in commodity-form. So the variable capital-value of 1,000 in commodities makes up only half of the value-product that variable capital, as an element of productive capital, has created. The 1,000 Iv in commodities is the exact equivalent of the part of total capital originally advanced by I as 1,000v in money — the part meant to function as variable. But in commodity-form, it is only potential money (it becomes actual money only once it is sold), and so it is even less directly variable money-capital. It finally becomes that through the sale of the 1,000 Iv commodities to IIc, and through the prompt reappearance of labour-power as a purchasable commodity — as the material into which the 1,000v in money can be converted.

Through all these transformations, capitalist I holds the variable capital in his hands the whole time: first, as money-capital; then, as an element of his productive capital; later still, as a value-part of his commodity-capital, that is, as commodity-value; and finally, again as money, facing once more the labour-power it can be converted into. During the labour process, the capitalist holds the variable capital in his hands as labour-power actively at work, creating value — but not yet as a value of a given, fixed size. Since he only ever pays the worker after that worker's labour-power has already been at work for some shorter or longer stretch of time, he already holds in his hands — before he pays — both the replacement-value that labour-power has created for itself and the surplus-value on top of it.

Since variable capital always stays, in one form or another, in the capitalist's hands, it cannot in any way be said to turn into revenue for anybody. The 1,000 Iv in commodity-form is converted into money, rather, through its sale to II — for whom it replaces, in kind, half of his constant capital.

What dissolves into revenue is not the variable capital of I, the 1,000v in money. That money stopped functioning as the money-form of I's variable capital the moment it was converted into labour-power — just as the money of any other seller of commodities stops representing anything of his the moment he has converted it into some seller's commodity. The transactions that this money — now received as wages — goes through in the hands of the working class are not transactions of variable capital at all, but transactions of the value of their labour-power, now turned into money. It is exactly the same as with the exchange of the value-product the worker has created (2,000 I(v+m)): that exchange is only the exchange of a commodity belonging to the capitalist, something that is none of the worker's business. But the capitalist — and still more his theoretical spokesman, the political economist — finds it hard to shake off the notion that the money paid out to the worker is somehow still the capitalist's own money. If the capitalist happens to be a gold producer, then the variable value-part — that is, the equivalent, in commodity-form, that replaces for him the purchase-price of labour — appears directly in money-form itself. It can then go straight back to functioning as variable money-capital, with no detour through a reflux at all. As for the worker in II — setting the luxury worker aside — the 500v exists as goods meant for the worker's own consumption, goods that the worker, taken as a whole body of workers, buys straight back from the very body of capitalists he sold his labour-power to. The variable value-part of capital II, in its natural form, consists of means of consumption, meant for the most part to be eaten up by the working class. But it is not the variable capital that gets spent by the worker in this form — it is his wage, his own money — and it is precisely by realizing itself in these means of consumption that this money restores the variable capital of 500 IIv for the capitalist, back in money-form. Variable capital IIv is reproduced in means of consumption, just as constant capital 2,000 IIc is reproduced in them; neither one dissolves into revenue any more than the other does. What dissolves into revenue, in both cases, is the wage.

That the spending of wages as revenue restores, in one case, 1,000 IIc, and by the same roundabout route 1,000 Iv, and likewise 500 IIv — restoring, that is, both constant and variable capital (variable capital partly through a direct reflux, partly through an indirect one) once again as money-capital — is an important fact about the exchange of the year's product.

Showing how the year's reproduction turns over runs into one big difficulty. Take the simplest form the thing appears in, and we get:

The sum above finally breaks down into:

= 9,000. Part of the constant capital's value — specifically, the part made up of actual means of labour, a distinct group within the means of production — has passed from those means of labour onto the product, the commodity. The means of labour themselves keep working as part of the productive capital, still in their old physical shape. What passes over is only their wear: the value they lose bit by bit as they keep functioning over some period. That lost value reappears as a value-component of the commodities made with them — it moves from the instrument of labour to the product of labour. So for the year's reproduction, only those parts of fixed capital that last longer than a year are in question here at all. If something dies out completely within the year, it has to be replaced and renewed in full by that year's reproduction — the point at issue does not concern it. But with machines and other longer-lasting kinds of fixed capital, it can happen, and often does, that certain component parts have to be replaced outright within the year, even though the building or machine as a whole is long-lived. Those component parts belong to the same category as the elements of fixed capital that need replacing within the year.

This value-element in the commodities must never be confused with repair costs. When the commodity is sold, this value-element is turned into money just like the others — but its difference from the other value-elements only shows up after that conversion into money. Raw materials and auxiliary materials used up in production must be replaced in kind, or the reproduction of the commodities cannot even begin — the production process could not go on continuously; the labour-power spent on them must likewise be replaced by fresh labour-power. So the money that comes from selling the commodity must constantly be turned back into these elements of productive capital, out of money form and into commodity form. It makes no difference that, say, raw and auxiliary materials get bought in bigger batches at certain intervals, forming stocks — so that for a while these means of production don't need to be bought again, and, as long as the stock lasts, the money coming in from the sale of the commodities, so far as it is meant for this purpose, can pile up. This part of the constant capital then appears, for the time being, as money-capital suspended in its active function. It is not revenue-capital — it is productive capital, suspended in money form. Renewal of the means of production must go on all the time, though the form this renewal takes, as far as circulation is concerned, can vary. The new purchase — the circulation operation by which they are renewed and replaced — can happen at longer intervals: then one large outlay of money at once, matched by a corresponding stock of the means of production; or it can happen in short, quick succession: then small doses of spending following one another rapidly, matched by small stocks. None of this changes anything about the matter itself. The same holds for labour-power: where production runs continuously at the same scale all year, the labour-power used up is constantly replaced by new; where labour is seasonal, or applied in different amounts at different times, as in agriculture, labour-power is bought correspondingly — sometimes in smaller, sometimes in larger quantities. By contrast, the money that comes from selling the commodity, so far as it monetizes the part of the commodity's value equal to the wear of fixed capital, is not converted back into the component of productive capital whose loss of value it replaces. It settles down alongside the productive capital and stays in money form. This money deposit repeats itself, again and again, until the reproduction period — made up of a greater or smaller number of years — has run its course; and throughout that period the fixed element of constant capital keeps functioning in the production process in its old physical form. Once that fixed element — buildings, machinery, and so on — has lived out its life and can no longer function in the production process, its value stands alongside it, fully replaced in money: the sum of the money deposits, the values that the fixed capital gradually passed onto the commodities it helped produce, and that turned into money form when those commodities were sold. This money then serves to replace the fixed capital, or parts of it, since its different parts have different lifespans, in kind, and so actually renews this component of the productive capital. This money is thus the money-form of part of the value of the constant capital — its fixed part. This forming of a hoard is therefore itself a moment of the capitalist reproduction process: the reproduction and storing-up, in money form, of the value of fixed capital or its individual parts, until the time when the fixed capital has lived out its life, has consequently given up its whole value to the commodities produced, and must now be replaced in kind. But this money only loses its hoard-form, and so only actively re-enters capital's reproduction process as carried by circulation, once it is turned back into new elements of fixed capital to replace the ones that have died out.

Just as simple commodity circulation is not the same thing as plain exchange of products, the turnover of the year's commodity product cannot be resolved into a plain, unmediated, mutual exchange of its various parts either. Money plays a specific role in this, a role that shows up above all in the way the value of fixed capital gets reproduced. (It remains to be examined afterward how this would look different, supposing production were held in common and did not take the form of commodity production.)

Let's go back to the basic schema. For department II we had: 2,000c+500v+500s. All the means of consumption produced over the year add up here to a value of 3,000; and each of the different kinds of commodity making up that total value breaks down, value-wise, in the same proportions: ⅔c+⅙v+⅙m, or as percentages, 66⅔%c+16⅔%v+16⅔%m. The different kinds of commodity in department II may contain constant capital in different proportions from one another; the fixed part of that constant capital may differ between them too; so may the lifespan of the fixed parts of capital, and therefore the yearly wear, or the share of value each transfers, pro rata, to the commodities it helps produce. None of that matters here. As far as the social reproduction process goes, what's at stake is only the turnover between department II and department I. Department II and department I face each other here only in their social mass-proportions; so the proportional size of the value-part c of department II's commodity-product — which is all that matters for the question now being dealt with — is the average ratio once every branch of production classed under II is added together.

Every one of these kinds of commodity — and for the most part they are the very same kinds of commodity — whose total value is entered under 2,000c+500v+500s, breaks down evenly, value for value, into 66⅔%c+16⅔%v+16⅔%m. This holds for every 100 units of the commodities counted under c, just as much as for those under v or under m.

The commodities in which the 2,000c is embodied can themselves be broken down again by value into:

1. 1,333⅓c+333⅓v+333⅓s = 2,000c. Likewise, the 500v breaks down into:

2. 333⅓c+83⅓v+83⅓s = 500v. And finally, the 500s breaks down into:

3. 333⅓c+83⅓v+83⅓s = 500s.

Let's now add up the c-portions from 1, 2, and 3: 1,333⅓c+333⅓c+333⅓c = 2,000. Do the same for the v-portions — 333⅓v+83⅓v+83⅓v = 500 — and likewise for the m-portions. Adding it all together gives the same total value of 3,000 as before.

So the entire constant-capital value contained in department II's mass of commodities, worth 3,000, is contained in the 2,000c — and neither the 500v nor the 500s contains a single atom of it. The same holds, each in its own place, for v and for m.

In other words: the whole quota of department II's mass of commodities that represents constant-capital value, and can therefore be turned into something else — whether into its natural form or into its money form — exists in the 2,000c. So everything to do with the turnover of the constant value of department II's commodities is confined to the movement of 2,000 IIc alone; and this turnover can only be carried out against department I's 1,000v+1,000s.

In the same way, everything to do with the turnover of the constant-capital value belonging to department I must be confined, in our examination, to the 4,000 Ic.

Let's start by taking:

If we take this schema, the exchange of these commodities — 2,000 worth from IIc — against commodities of the same value from department I would require that the whole of that 2,000 IIc gets converted back, in kind, into the physical things department I produces for constant capital II. But the commodity-value of 2,000 in which that capital exists contains an element for the loss of value of fixed capital, and that element cannot be replaced right away, in kind. It has to be turned into money instead — money that piles up bit by bit, as a total sum, until the time comes due to renew the fixed capital in its physical form. Every year is a death-year for some fixed capital: capital that has to be replaced in this business or that, in this branch of industry or that. Within one and the same individual capital, first one part of the fixed capital has to be replaced, then another, since its different parts wear out at different rates. When we look at annual reproduction — even on a simple scale, leaving accumulation aside — we are not starting from nothing. This is one year among many in an ongoing flow; it is not the first year capitalist production was ever born. So the different capitals invested across the many branches of department II are all of different ages. And just as, every year, people working in these branches die off, so every year masses of fixed capital reach the end of their working life and have to be renewed in kind out of an accumulated fund of money. To that extent, the exchange of 2,000 IIc against 2,000 I(v+m) includes converting 2,000 IIc out of its commodity-form — as means of consumption — into physical things that are not just raw and auxiliary materials, but equally the physical stuff of fixed capital: machines, tools, buildings, and so on. So the wear-and-tear that has to be replaced in money, inside the value of 2,000 IIc, is by no means proportional to the whole extent of the fixed capital actually in use, since only part of it needs replacing in kind each year. But that itself assumes that, in earlier years, the money needed for this replacement had already piled up in the hands of department II's capitalists. And this same assumption holds just as much for the current year as it is taken to hold for the earlier ones.

In the exchange between I (1,000v + 1,000s) and 2,000 IIc, notice first that the value-sum I(v+m) contains no constant-capital element at all — so no element for wear-and-tear that needs replacing, no value that a fixed part of constant capital has transferred onto the commodities whose physical form is v+s. That element does exist in IIc, though, and it is precisely part of this value owed to fixed capital that cannot turn straight from money into physical form — it has to stay as money for the time being. So a difficulty appears at once in the exchange of I (1,000v + 1,000s) against 2,000 IIc: the means of production from I, whose physical form holds that 2,000 (v+s), must be exchanged at their full value of 2,000 for an equivalent in means of consumption from II. But the means of consumption 2,000 IIc cannot be exchanged at their full value for means of production from I — because a proportional part of their value, equal to the wear-and-tear that has to be replaced, must first settle down as money, and within the current annual period we're considering, that money does not go back into circulation. But the money that turns this wear-and-tear element into cash — the part locked inside the value of 2,000 IIc — can only come from I. II cannot pay itself; it gets paid by selling its own goods. And since, on our assumption, I(v+m) buys the whole 2,000 IIc, class I must, through this very purchase, turn that wear-and-tear into money for II. But money advanced into circulation must, by the law established earlier, flow back to the capitalist producer who later throws an equal quantity of commodities into circulation. Clearly, when I buys IIc, it cannot hand II both 2,000 in goods and a surplus sum of money on top, once and for all, without that money coming back to I through the exchange itself — otherwise I would be buying IIc's goods above their value. If II really does exchange its 2,000c for I's 1,000v + 1,000s, then it has nothing further to claim from I, and the money that circulates during this exchange flows back to whichever side threw it into circulation — that is, to whichever acted first as buyer. But in that case II would have converted the whole value of its commodity-capital back into the physical form of means of production, while our assumption is that a proportional part of it, after being sold, does not get converted back out of money into the physical form of II's fixed capital — not within the current year. So a money balance could only flow to II if II sold 2,000 worth to I but bought less than 2,000 from I — say, only 1,800. Then I would have to make up the difference with 200 in money, and that money would not flow back to I, because I would not have withdrawn it from circulation again by throwing in a further 200 worth of goods. In that case we would have a money fund for II to cover its fixed-capital wear-and-tear — but on the other side, on I's side, we would have an overproduction of means of production worth 200. And with that, the whole basis of the schema would have dissolved: reproduction on an unchanging scale, which assumes complete proportionality between the different branches of production. One difficulty would only have been removed by a much worse one.

This problem has difficulties all its own, and no political economist has ever dealt with it before. So let's go through, one by one, every possible — or at least seemingly possible — solution, or rather every possible way of posing the problem itself.

First, we just assumed that II sells 2,000 worth to I but buys only 1,800 worth of goods from I. Inside the value of 2,000 IIc, 200 was locked up for wear-and-tear replacement — money that has to be hoarded. So the value of 2,000 IIc splits into 1,800, to be exchanged for means of production from I, and 200 for wear-replacement, to be held as money once the 2,000c has been sold to I. Or in terms of value: 2,000 IIc = 1,800c + 200c(d), where d stands for déchet — wear-and-tear.

We would then need to look at the exchange:

I buys, with the £1,000 that flowed to the workers as wages for their labour-power, means of consumption worth 1,000 from IIc. II then buys, with that same £1,000, means of production worth 1,000 from Iv. This brings the capitalists of I their variable capital back in money-form, so next year they can buy labour-power of the same value again — that is, replace the variable part of their productive capital in kind. Next, II advances a further £400 and buys means of production from Is, and Is buys with that same £400 means of consumption from IIc. The £400 that II advanced into circulation has thus flowed back to the capitalists of II — but only as payment for goods sold. I then advances a further £400 and buys means of consumption; II buys means of production worth £400 from I, and with that the £400 streams back to I. So far, the account stands as follows:

I throws into circulation, in goods: 1,000v + 800s. I also throws into circulation, in money: £1,000 as wages, and £400 for exchange with II.

Once the exchange is complete, I has: 1,000v in money-form, 800s converted into 800 worth of IIc means of consumption, and £400 in money.

II throws into circulation 1,800c in goods (means of consumption) and £400 in money. Once the exchange is complete, it has: 1,800 worth of goods from I (means of production) and £400 in money.

What's left standing now is this: on I's side, 200s still sitting in means of production; on II's side, 200c(d) still sitting in means of consumption.

On our assumption, I uses £200 to buy the means of consumption c(d), worth 200. But II holds onto that £200, because 200c(d) stands for wear-and-tear — it cannot be turned straight back into means of production. So the 200 Is cannot be sold: a fifth of the surplus-value that has to be replaced cannot be realized — it cannot pass out of its physical form as means of production into the form of means of consumption.

That the 200 Is can't be sold does not just contradict the assumption of reproduction on a simple scale. In itself it is not even a hypothesis that explains how 200c(d) gets turned into money — it amounts, rather, to saying that this cannot be explained at all. Since there is no way to show how 200c(d) is supposed to become money, it simply gets assumed that I does II the favour of monetizing it — precisely because I itself is unable to monetize its own remaining 200s. Treating this as a normal operation of the exchange mechanism is exactly the same as assuming that £200 rains down from heaven every year, like clockwork, to turn that 200c(d) into money.

The absurdity of a hypothesis like that isn't obvious right away, though, when Is doesn't show up in its raw original shape — as part of the value of means of production, part of the value of goods that their capitalist producers must realize as money by selling them — but instead turns up in the hands of people who merely share in that surplus-value: as ground-rent, say, in the hands of landowners, or as interest in the hands of money-lenders. But if the part of the goods' surplus-value that the industrial capitalist has to hand over as ground-rent or interest to these other co-owners of the surplus-value cannot, in the long run, be realized by selling the goods themselves, then the payment of rent or interest comes to an end too — so landowners or interest-receivers, by spending their income, cannot serve as some deus ex machina that monetizes whatever part of the annual reproduction needs it. The same holds for the spending of all the so-called unproductive workers — state officials, doctors, lawyers, and so on — and whatever else, under the name of "the general public," does "service" for political economists by explaining away what they cannot otherwise explain.

Nor does it help to bring in the merchant as a middleman, in place of direct exchange between I and II — the two great departments of capitalist producers themselves — and let his "money" carry us past every difficulty. In the case before us, for example, the 200 Is must, in the end, finally be sold to the industrial capitalists of II. It may pass through the hands of a whole chain of merchants, but the last one in that chain finds himself, on the same hypothesis, in exactly the position the industrial capitalists of I were in at the start: unable to sell the 200 Is to II. And the sum he has sunk into buying it cannot start that same process over again with I.

This whole survey of failed solutions makes clear, quite apart from our real purpose here, how necessary it is to examine the reproduction process in its most basic form, with every obscuring middleman stripped away. Only that lets us get rid of the false evasions that give the appearance of a "scientific" explanation, once the social reproduction process is made the object of analysis straightaway in its tangled, concrete form.

The law is this: under the normal course of reproduction — whether on a simple or an expanded scale — the money a capitalist producer advances into circulation must flow back to its starting point, and it makes no difference whether that money is the producer's own or borrowed. This law, then, rules out once and for all the hypothesis that 200 IIc(d) could be turned into money by money advanced by I.

Once we set aside the case we just looked at, the only possibilities left are ones where — besides replacing the wear-and-tear portion in money — the completely worn-out fixed capital must also actually be replaced in kind.

We had assumed earlier:

(a) That the £1,000 paid out by department I as wages gets spent by the workers on IIc goods of the same value — that is, they use it to buy means of consumption.

That the £1,000 here is advanced by I in money is simply a statement of fact. The capitalists must pay wages in money; the workers then spend this money on means of subsistence, and it serves the sellers of those goods in turn as circulating medium for turning their constant capital from commodity-capital back into productive capital. The money passes through many hands along the way — shopkeepers, landlords, tax collectors, unproductive workers such as doctors, whom the worker himself needs — so only part of it flows directly from the hands of I's workers into the hands of capitalist class II. This flow may run more or less unevenly, which is why the capitalists may need an extra money reserve. None of this matters for the basic form we are considering here.

(b) We had also assumed that at one point I advances a further £400 in money to buy from II — money that flows back to I — just as at another point II advances £400 to buy from I — money that flows back to II. This assumption has to be made, since the alternative — that only class I, or only class II, one-sidedly advances the money circulation needs — would be arbitrary. Now, the previous section showed that it is absurd to suppose I throws in extra money to turn 200 of IIc(d) into money. That seems to leave only an even more absurd-looking supposition: that II itself throws into circulation the money that turns into cash the part of its commodity-value which has to replace the wear of fixed capital. Take an example. The value that Mr. X's spinning machine loses in production reappears as part of the value of the yarn. What his machine loses in value on one side is supposed to pile up as money in his hands on the other. Say X buys £200 of cotton from Y, advancing £200 in money into circulation; Y then buys yarn from X with that same £200, and X now treats this £200 as his fund for replacing the wear on his spinning machine. But this would mean nothing more than X, quite apart from his production and its sale, setting aside £200 to pay himself back for the machine's loss of value — that is, on top of the £200 his machine actually loses in value, he would have to put in yet another £200 out of his own pocket every year, just so that he could eventually afford a new machine.

But the absurdity is only apparent. Class II is made up of capitalists whose fixed capital stands at quite different points in its cycle of renewal. For some of them the moment has arrived when it must be replaced wholly in kind. For others that moment is still more or less distant — and what all the members of this latter group have in common is that their fixed capital is not actually being renewed yet: it is not being replaced in kind by a new machine of the same sort, but its value is instead being gradually accumulated in money. The first group stands — wholly, or partly, it makes no difference here — exactly where it stood when the business was founded, when it came to market with money capital in order to convert part of it into constant capital, fixed and circulating, and part of it into labour-power, into variable capital. Just as then, it now again has to advance this money capital into circulation — the value of its fixed constant capital just as much as that of its circulating and variable capital.

So suppose that of the £400 which capitalist class II throws into circulation to trade with I, half comes from those capitalists in II who must renew not only the circulating means of production they buy with their commodities, but also their fixed capital in kind, paid for with their money — while the other half comes from capitalists in II who use their money only to replace in kind the circulating part of their constant capital, without yet renewing their fixed capital in kind. On this assumption there is nothing contradictory at all in the £400 that flows back — flowing back as soon as I spends it on means of consumption — now being divided differently between these two groups within II. The money flows back to class II, but not into the same hands: it is redistributed within the class, passing from one part of it to the other.

One group within II — besides the portion of means of production its commodities have already paid for — has converted £200 in money into new fixed-capital elements in kind. Just as at the founding of the business, this money it laid out only flows back gradually, over a run of years, as the wear-and-tear portion built into the value of the commodities this fixed capital will go on to produce.

The other group within II, by contrast, has not received any commodities from I for its £200; instead, I pays this group with the very money the first group used to buy its fixed-capital elements. So one group within II now holds its fixed-capital value again in renewed, physical form; the other is still in the process of accumulating that value in money form, ready for when it eventually replaces its own fixed capital in kind.

The starting point, after the exchanges already carried out, is the remainder still left to be traded on each side: 400 in surplus-value for I, and 400 in constant capital for II.

Suppose II advances £400 in money to trade this remaining £800 worth of commodities. One half of that £400 — £200 — must, whatever else happens, be laid out by the part of IIc that has been accumulating £200 in money as wear-value, and that now has to turn this money back into the physical form of its fixed capital.

Just as the value of II's commodity-capital, like I's, splits into constant capital value, variable capital value, and surplus-value, each of which can itself be represented by its own proportional slice of the commodities themselves, so too within the constant-capital value there is a further split: a part not yet due to be converted into the physical form of fixed capital, but still, for now, to be gradually hoarded as money. A given quantity of commodities from II — here, half of the remainder, £200 — is nothing more than the carrier of this wear-value, which has to be turned into money through the exchange. (The group within II that renews its fixed capital in kind may already have realized part of its wear-value through the wear-and-tear component of the whole mass of goods, of which only this remainder is still under discussion — but £200 in money still remains for it to realize.)

As for the second half of that £400 — the other £200 — which II throws into circulation in this remaining transaction, it is used to buy circulating elements of constant capital from I. This £200 may be put into circulation by either group within II, or only by the group that is not renewing its fixed-capital component in kind.

With this £400, then, I parts with two lots of goods: first, £200 worth consisting only of elements of fixed capital; second, £200 worth that merely replaces the physical elements of the circulating part of II's constant capital. I has now sold the whole of its annual output that was destined for II — but the value of a fifth of that output, £400, now sits in I's hands as money. This money, though, is surplus-value turned into cash, and it must be spent as revenue on means of consumption. So I uses the £400 to buy up the whole £400 of commodity-value still held by II. The money thus flows back to II, since it is used to take II's goods off its hands.

Let us now take three cases. We will call the group of capitalists in II that replaces fixed capital in kind "Part 1", and the group that is accumulating the wear-value of fixed capital in money form "Part 2". The three cases are these: (a) Of the £400 still outstanding in commodities under II, a share for Part 1 and a share for Part 2 — say, half each — still has to be used to replace certain portions of the circulating part of constant capital. (b) Part 1 has already sold the whole of its commodities, so Part 2 still has £400 left to sell. (c) Part 2 has sold everything except the £200 that carries wear-value.

This gives us the following breakdowns.

(a) Of the £400 worth of goods still in II's hands, Part 1 holds £100 and Part 2 holds £300 — of which £200 represents wear. Now, of the £400 in money that I sends back to take up II's goods, Part 1 originally laid out £300 of it: £200 in money, for which it drew fixed-capital elements in kind from I, and £100 in money to carry out its ordinary trade with I. Part 2, meanwhile, advanced only a quarter of the £400 — £100 — likewise to carry out its trade with I.

So of the £400 in money, Part 1 advanced £300 and Part 2 advanced £100.

But of this £400, what flows back is this:

To Part 1: £100 comes back — only a third of the money it advanced. But for the other two-thirds it now holds renewed fixed capital worth £200. For this fixed-capital element worth £200 it handed over money to I, but supplied no commodity in return. With respect to this portion, Part 1 stands toward I only as a buyer, never afterward as a seller too. So this money cannot flow back to Part 1 — if it did, I would have given Part 1 the fixed-capital elements as a gift. With respect to the last third of the money it advanced, Part 1 first appeared only as a buyer of circulating elements of its constant capital. With that same money, I then buys from Part 1 the rest of its commodity, worth £100. So this money does flow back to Part 1 — because right after acting as a buyer, it turns around and acts as a seller of commodities. If the money did not flow back, then II's Part 1 would have given I, for £100 worth of commodities, first £100 in money and then another £100 worth of commodities on top — in other words, would have given away its commodity as a gift.

To Part 2, by contrast, which laid out only £100 in money, £300 in money flows back: £100, because it first threw £100 into circulation as a buyer and gets this back as a seller; £200, because with respect to this portion it acts only as a seller of goods worth £200, never as a buyer. So this money cannot flow back to I. The wear of fixed capital is thus settled by the money that II's Part 1 threw into circulation to buy fixed-capital elements — but this money reaches the hands of Part 2 not as Part 1's money, but as money belonging to class I.

(b) On this assumption, the remainder of IIc is divided so that Part 1 holds £200 in money and Part 2 holds £400 in commodities.

Part 1 has sold all its commodities, but its £200 in money is simply the transformed shape of the fixed component of its constant capital, which it still has to renew in kind. So here it appears only as a buyer, and receives, in place of its money, goods from I consisting of physical elements of fixed capital of the same value. Part 2, at most — assuming I advances no money of its own for the trade between I and II — only has £200 to throw into circulation, since for half of its commodity-value it is only a seller to I, never a buyer from I.

£400 flows back to Part 2 out of circulation: £200, because it advanced this as a buyer and gets it back as a seller of £200 worth of goods; £200, because it sells goods worth £200 to I without drawing any equivalent commodity back from I in return. (c) Part 1 holds £200 in money and £200 worth of constant-capital goods; Part 2 holds £200 worth of constant-capital goods carrying wear-value.

On this assumption, Part 2 has no money at all to advance, since toward I it no longer acts as a buyer in any way, only as a seller — so it simply has to wait until I buys from it.

Part 1 advances £400 in money: £200 for ordinary trade with I, and £200 purely as a buyer from I. With this second £200 it buys the fixed-capital elements.

I uses £200 in money to buy £200 worth of goods from Part 1, so that the £200 Part 1 advanced for this trade flows back to it. And I uses the other £200 — which it likewise received from Part 1 — to buy £200 worth of goods from Part 2, so that Part 2's fixed-capital wear comes down to it in money.

Nothing about the outcome in case (c) would change if, instead of II's Part 1, it were class I that advances the £200 to set the existing goods in motion. Suppose I first buys £200 worth of goods from II's Part 2 — which, by assumption, has only this remainder left to sell. Then this £200 does not flow back to I, since Part 2 does not turn around and act as a buyer. But Part 1 of II then still has £200 in money to spend as a buyer, and also still has £200 worth of goods of its own to trade — £400 in all to exchange with I. £200 in money then flows back to I from Part 1 of II. If I lays this out again to buy the £200 of goods from Part 1, it flows back to I once more, as soon as Part 1 buys the second half of I's £400 worth of goods.

Part 1 laid out its £200 in money purely as a buyer of fixed-capital elements, so this £200 does not flow back to it; instead it serves to turn Part 2's remaining £200 of goods into money. Meanwhile the £200 I laid out for trading purposes has flowed back to I — not by way of Part 2, but by way of Part 1. For its £400 worth of goods, I has received back an equivalent worth £400; and the £200 in money I advanced to circulate the whole £800 of goods has likewise come back to it. So everything is in order.

The difficulty we ran into was over one exchange in particular — department I's 1,000v + 1,000s against department II's 2,000c, set out just below.

That whole difficulty has now been narrowed down to a smaller one: exchanging only what is left over on each side — the remnants set out next.

In department II, part 1, £200 of commodities gets exchanged for £200 of Is (commodities). And every coin that circulates between I and II in this £400 exchange of commodities flows back to whoever advanced it — I or II. So this money, as far as the exchange between I and II goes, is in fact no element of the problem we're dealing with here. Put another way: suppose that in the exchange between £200 of Is (commodities) and £200 of IIc (the commodities of II, part 1), money functions as a means of payment rather than a means of purchase — and so not as a "medium of circulation" in the strictest sense. Then it's clear, since £200 Is and £200 IIc (part 1) are commodities of equal value, that means of production worth £200 are exchanging against means of consumption worth £200. Money here functions only ideally: no money actually has to be thrown into circulation to settle a balance on either side. The problem only comes out in its pure form once we strike out the commodity £200 Is and its equivalent, the commodity £200 IIc (part 1), on both I's side and II's side.

Once we take away these two equal-value amounts of commodities (from I and from II), which cancel each other out, what's left is the residue of the exchange — the part where the problem shows up in its pure form, namely:

Here it's clear: with its £200 in money, II part 1 buys the £200 Is that make up components of its fixed capital. That renews II part 1's fixed capital in kind, and it turns I's surplus-value of £200 from commodity-form — means of production, specifically elements of fixed capital — into money-form. With that money, I buys means of consumption from II part 2. The result for II is: part 1 has renewed a fixed component of its constant capital in kind, and part 2 has had another component — one that stands in for wear and tear of fixed capital — turned into money. This goes on year after year, until that second component also needs renewing in kind.

The precondition here is plainly this: the fixed component of II's constant capital that gets reconverted into money at its full value, and so must be renewed in kind every year (part 1), must equal the annual wear of the other fixed component of II's constant capital — the one still working on in its old natural form, whose wear, the loss of value it passes onto the commodities it helps produce, has first to be made good in money. Such a balance would then appear as a law of reproduction on an unchanging scale. In other words: in class I, which produces means of production, the proportional division of labour must stay unchanged, insofar as I supplies, on one side, the circulating components and, on the other, the fixed components of department II's constant capital.

Before we look at this more closely, we first need to see what happens when the residue of IIc(1) isn't equal to the residue of IIc(2) — it can be bigger or smaller. Let's take the two cases one at a time.

Here IIc(1) uses its £200 in money to buy the £200 of Is commodities, and I uses that same money to buy the £200 IIc(2) commodities — the fixed-capital component that has to be turned into money. That component is now money. But £20 of IIc(1), still sitting there as money, can't be converted back into fixed capital in kind.

This drawback looks fixable if we set the residue of Is not at £200 but at £220 — so that of department I's £2,000, only £1,780 is accounted for by the earlier exchange, instead of £1,800. In that case, then:

IIc, part 1, uses its £220 in money to buy the £220 Is, and I then uses £200 of that to buy the £200 IIc(2) in commodities. But then £20 is left over in money on I's side — a piece of surplus-value that I can only hold as money, not spend on means of consumption. The difficulty hasn't gone away; it's just moved, from IIc (part 1) to Is.

Now let's assume the opposite: that IIc, part 1, is smaller than IIc (part 2). So:

II (part 1) uses its £180 in money to buy £180 of commodities Is. I uses that same money to buy an equal value of commodities from II (part 2) — £180 of IIc(2). That leaves £20 of Is unsold on one side, and likewise £20 of IIc(2) on the other: £40 worth of commodities that can't be turned into money.

It wouldn't help us to set I's residue at £180 instead. Then I would have no surplus left over, true — but as before, a surplus of £20 in IIc (part 2) would remain unsold, unable to be turned into money.

In the first case — where II(1) is bigger than II(2) — a surplus stays on IIc(1)'s side, in money, unable to be converted back into fixed capital. Or, if we set the residue of Is equal to IIc(1), that same surplus stays instead on Is's side, in money, unable to be converted into means of consumption.

In the second case — where IIc(1) is smaller than IIc(2) — a shortfall in money remains on the side of £200 Is and IIc(2), matched by an equal surplus of commodities on both sides. Or, if we set the residue of Is equal to IIc(1), the shortfall in money and the surplus in commodities both sit on IIc(2)'s side.

Let's always set the residue of Is equal to IIc(1) — since orders determine production, and it makes no difference to reproduction whether I produces more fixed-capital components this year and more circulating-capital components of department II's constant capital next year. On that basis: in the first case, Is could be converted back into means of consumption only if I used it to buy part of II's surplus-value — meaning that surplus-value, instead of being consumed, would have to be hoarded by II as money. In the second case, the only remedy would be for I itself to spend the money — that is, the hypothesis we already rejected.

If IIc(1) is bigger than IIc(2), importing foreign commodities is needed to realize the money surplus sitting in Is. If IIc(1) is smaller than IIc(2), the reverse: exporting commodity II (means of consumption) is needed to realize the wear-and-tear portion of IIc that's tied up in means of production. Either way, foreign trade is necessary.

Suppose, for the sake of studying reproduction on an unchanging scale, that we assume the productivity of every branch of industry — and so the proportional value-relations of their commodity-products — stays constant. Even so, the last two cases we looked at, where IIc(1) is bigger or smaller than IIc(2), would still matter for production on an expanded scale, where they can arise as a matter of necessity.

On the replacement of fixed capital, one general point needs making. Suppose everything else stays the same — not just the scale of production but, in particular, the productive power of labour too. Now suppose that this year a bigger share of department II's fixed capital (the part making means of consumption) dies off than died the year before, so a bigger share has to be replaced in kind. Then the share that is, for now, only being made good in money — the part still dying, not yet dead, whose value keeps getting replaced in cash until its day of death arrives — must shrink in the same proportion. That follows because, by assumption, the total value of the fixed capital at work in department II stays the same. This carries two consequences. First: if a bigger share of department I's output-in-commodities consists of fixed-capital elements for IIc, then a correspondingly smaller share consists of circulating elements for IIc — because department I's total output for IIc is unchanged: what one part gains, the other loses. But department II's total output must also stay the same size. How can that be, when its raw materials, semi-finished goods, and auxiliary materials — the circulating elements of its constant capital — have shrunk? Second: a bigger share of department II's fixed capital, once restored in money-form, now flows over to department I, to be turned back from money into its natural form. So more money flows to I than the money already circulating between I and II for ordinary buying and selling — money that isn't mediating an exchange of commodity for commodity, but showing up only on one side, as pure means of purchase. At the same time, the mass of commodities from IIc that carries the value-replacement for wear and tear would have shrunk in proportion — the mass of goods from II that has to be turned into money rather than exchanged for goods from I. So more money would flow from II to I as pure purchasing power, and there would be less commodity from II for I to buy with it. A bigger share of department I's surplus-value sitting in commodity-form — since I's variable-capital portion is already converted into commodity from II — could not be converted into commodity from II at all, and would sit stuck in money-form.

The opposite case — where in some year less of department II's fixed capital dies off and needs replacing in kind, while the part merely wearing down is correspondingly larger — doesn't need to be worked through separately here.

And so a crisis would be here — a crisis of production — despite reproduction going on at an unchanged scale.

In a word: take simple reproduction with everything else held constant — in particular, the productive power of labour, the total scale, and the intensity of labour all unchanged. Suppose no constant proportion is assumed between the fixed capital that is dying off (and so must be renewed) and the fixed capital that goes on working in its old natural form (merely adding value to the product to replace its wear). Then in one case, the mass of circulating components needing reproduction would stay the same, while the mass of fixed components needing reproduction would have grown. Department I's total output would then have to grow — or else, quite apart from any question of money, there would be a deficit in reproduction.

In the other case: suppose the proportional size of the department II fixed capital needing renewal in kind decreases, so that — in the same ratio — the portion of department II's fixed capital that now needs replacing only in money increases. Then the mass of circulating components of department II's constant capital that department I reproduces would stay unchanged, while the mass of fixed components needing reproduction would have shrunk. So either department I's total output decreases — or else there is a surplus (the mirror image of the deficit before), a surplus that cannot be turned into money.

True, in the first case, the same labour could — given rising productive power, a bigger workforce, or greater intensity — turn out a larger product, and the deficit could be covered that way. But such a shift could not happen without moving labour and capital out of one branch of department I's production and into another, and every such move would cause momentary disruptions. And second — insofar as it is the extension or intensification of labour that is doing the work — department I would have to exchange more value for less value from department II, so department I's product would be depreciated.

Conversely, in the second case, department I has to contract its production — which spells crisis for the workers and capitalists employed there — or else it turns out a surplus, which again spells crisis. Surpluses like this are no evil in themselves — quite the opposite, they are an advantage. It is only in capitalist production that they become an evil.

Foreign trade could help out in both cases: in the first, by turning the commodity of department I that is stuck in money-form into means of consumption; in the second, by selling off the surplus abroad as commodity. But foreign trade — except where it simply replaces elements, value for value — does not remove these contradictions. It only shifts them onto a wider stage and gives them more room to play out.

Once the capitalist form of reproduction has been done away with, the matter comes down to this: the size of the portion of fixed capital that dies off each year — and so must be replaced in kind (here, the fixed capital at work making means of consumption) — varies from one year to the next. If in one year it is very large — above the average death-rate, the way it is with people — then the following year it is bound to be correspondingly smaller. But the mass of raw materials, semi-finished goods, and auxiliary materials needed each year to produce the means of consumption — everything else assumed constant — does not shrink for that reason. So the total output of means of production would have to grow in one year and shrink in the next. The only fix for this is to keep producing somewhat more than is immediately needed, on an ongoing basis: on one hand, a certain quantity of fixed capital produced beyond what is directly needed; on the other hand — and especially — a stock of raw material and the like that goes beyond the immediate yearly requirement (this holds above all for the means of subsistence). Overproduction of this kind is exactly what it looks like when society has control over the material means of its own reproduction. Within capitalist society, though, it is an anarchic element.

This example of fixed capital — under reproduction at an unchanged scale — makes the point sharply. Disproportion between the production of fixed and circulating capital is one of the economists' favourite explanations for crises. That such a disproportion can and must arise from the mere upkeep of fixed capital is something new to them. That it can and must arise even on the assumption of an ideal, normal production — at simple reproduction of the social capital already at work — is new to them too.

One thing has been left out so far: the yearly production of gold and silver. As mere material for luxury goods, gilding, and so on, they wouldn't need any special mention here, any more than any other product would. But they play an important role as money material - and so as potential money. To keep things simple, we'll consider only gold as money material here.

Older estimates put the world's total yearly gold output at 800,000-900,000 pounds - around 1,100 or 1,250 million marks. But one more careful estimate, covering the average of the years 1871-75, puts it at only 170,675 kilograms, worth around 476 million marks. Of that, Australia supplied about 167 million marks' worth, the United States 166 million, and Russia 93 million. The rest was spread across various countries, each contributing less than 10 million marks. Yearly silver production over the same period came to just under 2 million kilograms, worth 354½ million marks. Of that, in round numbers, Mexico supplied 108 million, the United States 102 million, South America 67 million, Germany 26 million, and so on.

Among countries where capitalist production predominates, only the United States produces both gold and silver. The capitalist countries of Europe get almost all their gold, and by far the largest part of their silver, from Australia, the United States, Mexico, South America, and Russia.

But we are going to relocate the gold mines into the very country of capitalist production whose yearly reproduction we are analysing here, for the following reason:

Capitalist production never exists at all without foreign trade. But once we assume normal yearly reproduction on a given scale, we have already assumed that foreign trade only replaces home-produced goods with goods of a different use-form or natural form, without touching the value ratios - including the ratio in which the two categories, means of production and means of consumption, exchange against each other, and including the ratios of constant capital, variable capital, and surplus-value into which the value of each category's product breaks down. Bringing foreign trade into the analysis of the annually reproduced product-value can therefore only cause confusion, without adding anything new either to the problem or to its solution. It must be left out of account entirely. So gold, too, must be treated here as a direct element of annual reproduction, not as a commodity brought in from outside through exchange.

Gold production, like metal production generally, belongs to department I, the category covering the production of means of production. Let's assume the yearly gold product is worth 30 (for convenience - actually far too high compared with the figures in our scheme). Suppose this value breaks down into 20c+5v+5s. The 20c has to be exchanged against other elements of Ic, which we'll consider later. But the 5v+5s have to be exchanged against elements of IIc - that is, against means of consumption.

As for the 5v: every gold-producing business begins by buying labour-power - not with gold it has produced itself, but with a portion of the money already circulating in the country. The workers spend this 5v buying means of consumption from department II, and department II then uses that money to buy means of production from department I. Say department II buys 2 worth of gold from department I as raw material (part of its constant capital); then 2v flows back to the gold producers in department I, in money that already belonged to circulation before this. If department II buys no further material from department I, department I can still buy from department II - by throwing its own gold into circulation as money, since gold can buy any commodity. The only difference is that here department I appears not as a seller but only as a buyer. The gold-diggers of department I can always sell their goods: their product is always already in directly exchangeable form.

Suppose a yarn spinner pays 5v to his workers; setting aside surplus-value, they hand him back a product - yarn - worth 5. The workers spend that 5 buying from department II, which in turn spends 5 in money buying yarn from department I - so the 5v flows back to the spinner in money. But in the case we're considering, the gold producer - call him Ig - advances 5v in money to his workers, money that already belonged to circulation beforehand. The workers spend it on means of subsistence, but of that 5, only 2 flows back to Ig from department II. Even so, Ig can start the reproduction process afresh just as well as the spinner can: his workers have delivered him 5 in gold, of which he has sold 2, and he still holds 3 in gold. All he needs to do is mint it or turn it into banknotes, and his whole variable capital is back in his hands in money form directly - without needing department II as a go-between at all.

Even in this first round of yearly reproduction, though, a change has already taken place in the mass of money that actually or potentially belongs to circulation. We assumed that department II bought 2 of this money from Ig as material, and that the remaining 3 was laid out again by Ig within department II as the money-form of variable capital. So out of the money supplied by this new gold production, 3 has stayed within department II and not flowed back to department I. By assumption, department II has now met its need for gold material. That 3 remains in its hands as a gold hoard.

This extra 3 in gold cannot become part of department II's constant capital, and department II already had enough money-capital before this to buy labour-power. Except for covering wear and tear, this additional 3 has no job to do within IIc, set against the portion of goods it was exchanged for — it could only help cover wear and tear if the first part of IIc happened to be smaller than the second, and that would be pure coincidence.

On the other hand, except again for the wear-and-tear element, the whole of IIc's commodity-product must be converted into means of production from department I. So this money must be moved entirely out of IIc and into IIs — department II's surplus-value — whether that surplus-value consists of necessities or of luxuries; and a matching amount of goods-value must move the other way, from IIs into IIc.

The result: part of the surplus-value gets stored up as a money-hoard.

In the second year of reproduction, if the same proportion of the yearly gold output continues to be used up as material, then again 2 will flow back to Ig, and the 3 will be replaced in kind - that is, once again turn into a hoard sitting in department II, and so on.

Now consider variable capital in general. Like any other capitalist, Ig constantly has to advance this capital in money to buy labour. But when it comes to this v, it is not Ig himself but his workers who buy from department II - so it can never happen that Ig himself appears as the buyer here, throwing gold into circulation on his own initiative rather than on department II's. Still, insofar as department II buys material from him - because it has to convert its constant capital IIc into gold material - part of (Ig)'s v flows back to him from department II, exactly as it does for the other capitalists in department I. And insofar as that doesn't happen, he replaces his v in gold directly out of his own product. But to the extent that the v he advanced in money does not flow back from department II, part of the money already circulating there - money that flowed to it from department I and was never sent back - turns into a hoard, and correspondingly, part of department II's surplus-value goes unspent on means of consumption. Since new gold mines are constantly being opened, or old ones reopened, a certain proportion of the money Ig has to lay out as v is always drawn from the money mass that already existed before this new gold production. This money gets thrown into department II by way of Ig's workers, and to the extent it doesn't come back from department II to Ig, it becomes there an element of hoard-formation.

Now for (Ig)'s s: here Ig can always appear as a buyer. It throws its s into circulation as gold and draws out means of consumption from IIc in return. Part of this gold gets used up as material, and so functions as a real element of the constant part, c, of department II's productive capital; and to the extent that this isn't the case, it becomes, once again, an element of hoard-formation - the part of IIc that stays behind in money form. This shows - and this is even leaving aside the case of Ic, which we'll come to later - that even in simple reproduction, where accumulation in the strict sense of the word (that is, reproduction on an expanded scale) is excluded, the accumulation of money, or hoard-formation, is nevertheless necessarily included. And because this repeats afresh every year, it explains the very assumption we started from in looking at capitalist production: that at the beginning of reproduction, a mass of money corresponding to the turnover of goods is already sitting in the hands of the capitalist classes of departments I and II. This build-up of hoards happens even after subtracting the gold that gets lost through the wear and tear of circulating money.

Naturally, the older capitalist production gets, the bigger the mass of money piled up everywhere, and so the smaller the share that each year's new gold output adds to that mass — even though the gold added in any one year can still be large in absolute terms. Here we want to come back once more, in general terms, to the objection raised against Tooke: how can every capitalist draw a surplus-value out of the annual product in money — that is, take more money out of circulation than he puts in — when, in the end, the capitalist class itself has to be seen as the very source that puts money into circulation in the first place?

Here is the answer already given in chapter 17, pulled together once more.

The only condition actually needed here is this: that there be enough money in existence to circulate the various parts of the annual mass of reproduced goods. Whether part of the value of these goods is surplus-value or not makes no difference to that condition at all. Suppose the whole product belonged to the workers themselves, so that their extra labour was extra labour for themselves and not for any capitalist. The mass of commodity-value in circulation would be exactly the same, and, other things equal, it would need exactly the same mass of money to circulate it. So in both cases the only real question is: where does the money come from to circulate this whole mass of commodity-value? It is never: where does the money come from to turn the surplus-value into money?

Still, to return to it once more: every single commodity is made up of c + v + s. So circulating the whole mass of commodities needs, on one side, a certain sum of money to circulate the capital c + v, and on the other side a separate sum of money to circulate the capitalists' revenue, the surplus-value m. Just as for one capitalist, so for the whole class: the money laid out as capital is a different money from the money spent as revenue. Where does that second money come from? Simply from this: part of the money sitting in the hands of the capitalist class — and, broadly speaking, part of the whole mass of money in society — circulates the capitalists' revenue. We already saw earlier how a capitalist setting up a new business gets back, once the business is running, the very money he spent keeping himself in food and other necessities, now returning to him as money that turns his surplus-value into money. But speaking generally, the whole difficulty has two sources:

First: suppose we look only at the circulation and turnover of capital, treating the capitalist purely as capital personified — not as someone who consumes and enjoys life. Seen this way, he is constantly throwing surplus-value into circulation as part of his commodity-capital. But we never see money sitting in his hands as revenue; we never see him throwing money into circulation to spend on consuming that surplus-value.

Second: when the capitalist class throws a sum of money into circulation in the form of revenue, it looks as if it were paying an equivalent for that part of the annual total product — as if that part stopped being surplus-value. But the surplus-product that embodies the surplus-value costs the capitalist class nothing at all. As a class, it owns and enjoys that product for free, and no amount of money circulation changes that. All that money circulation changes is this: instead of consuming his surplus-product exactly as it comes — which mostly isn't even possible — each capitalist draws out of the whole social stock of annual surplus-product whatever goods he wants, up to the value of the surplus-value he has appropriated, and takes them out of the general market. But the mechanism of circulation shows that when the capitalist class throws money into circulation to spend as revenue, it also draws that very money back out of circulation again — so it can start the same process over and over. In other words, considered as a class, the capitalists go on holding the very sum of money needed to turn the surplus-value into money. So when a capitalist withdraws goods from the market in the form of surplus-value that cost him nothing, and at the same time gets back the money he paid for those goods, then plainly he has taken the goods out of circulation without giving anything in return. They cost him nothing, even though he handed over money for them. Say I buy goods with a pound, and the seller hands that pound straight back to me as payment for surplus-product that cost me nothing — then clearly I got the goods for free. Doing this over and over changes nothing: I keep withdrawing goods and keep holding the pound, even though each time I let go of it for a moment to get the goods. The capitalist keeps getting this money back, as the turning into money of a surplus-value that cost him nothing.

We already saw that in Smith's account the whole value of the social product dissolves into revenue — into v + s — which means the constant capital-value is set at zero. From that it follows, necessarily, that the money needed to circulate the annual revenue must also be enough to circulate the whole annual product. In our example: the money needed to circulate 3,000 worth of means of consumption would have to be enough to circulate the whole year's product, worth 9,000. This is indeed Smith's view, and Tooke repeats it. This false picture of the ratio between the money needed to turn revenue into money and the money that circulates the whole social product is bound to follow once the different material and value elements of the annual total product, and the way they are reproduced and replaced each year, go unexamined and get pictured thoughtlessly. It has therefore already been refuted.

Let's hear Smith and Tooke in their own words.

Smith writes, in Book II, chapter 2:

“The circulation of every country can be split into two parts: the circulation between dealers, and the circulation between dealers and consumers. Even though the same pieces of money — paper or metal — might sometimes be used in one of these circulations and sometimes in the other, both go on side by side all the time, and each of them needs a certain mass of money of one kind or another to keep going. The value of the goods circulating among the various dealers can never exceed the value of the goods circulating between dealers and consumers, because whatever the dealers buy must in the end be sold to consumers. Since circulation among dealers happens wholesale, it generally needs a fairly large sum for each single transaction. Circulation between dealers and consumers, by contrast, mostly happens retail and often needs only very small sums of money — sometimes a shilling, or even half a penny, is enough. But small sums circulate far faster than large ones... So although the annual purchases of all consumers are worth at least” {that “at least” is a nice touch!} “as much as those of all the dealers, they can usually be settled with a far smaller mass of money,” and so on.

Commenting on this passage of Smith's, Tooke writes (An Inquiry into the Currency Principle, London 1844, pp. 34–36, in extracts):

“There can be no doubt that the distinction drawn here is correct in substance... The exchange between dealers and consumers also includes the payment of wages, which form the main resource (the principal means) of consumers... All transactions from dealer to dealer — that is, every sale starting from the producer or importer, through every stage of manufacturing and intermediate processing, down to the retailer or the export merchant — can be resolved into movements of capital transfer. But capital transfers do not necessarily require, and in the great mass of transactions do not actually involve, any real handing-over of banknotes or coin — I mean an actual, not a fictitious handing-over — at the moment of transfer... The total volume of transactions between dealer and dealer must, in the end, be determined and limited by the volume of transactions between dealers and consumers.”

If that last sentence stood on its own, one might think Tooke was just pointing out that some relation holds between dealer-to-dealer transactions and dealer-to-consumer transactions — in other words, between the value of the whole annual revenue and the value of the capital that produces it. But that is not the case. He explicitly signs on to Smith's view. So there is no need for a separate critique of his theory of circulation — the general one already covers it.

Every industrial capital, when it starts up, throws money into circulation all at once for the whole of its fixed component — money it only draws back out gradually, over a run of years, by selling its annual output. So at first it puts more money into circulation than it takes out. This happens again every time the whole capital gets renewed in kind; it happens every year for some number of businesses that need to renew their fixed capital in kind; and it happens bit by bit with every repair, every partial renewal of fixed capital. So wherever more money is drawn out of circulation than is put in on one side, the opposite is happening on the other.

In every branch of industry where the production period — as distinct from the labour period — runs long, the capitalist producers keep throwing money into circulation the whole time it runs: partly to pay for the labour-power they employ, partly to buy the means of production they use up. This pulls means of production straight out of the market, and pulls means of consumption out too — partly at one remove, through the workers spending their wages, partly directly, through the capitalists themselves, who go on consuming as usual. And all this happens without these capitalists putting an equivalent amount of goods back onto the market at the same time. Throughout this period, the money they put into circulation serves to turn commodity-value — including the surplus-value inside it — into money. This factor becomes very important once capitalist production is fully developed, in long-drawn-out ventures run by joint-stock companies and the like: building railways, canals, docks, major city construction, iron shipbuilding, large-scale land drainage, and so on.

Other capitalists, leaving aside their outlay on fixed capital, draw more money out of circulation than they put in when they buy labour-power and circulating capital. Gold- and silver-producing capitalists work the opposite way. Apart from the precious metal that serves them as raw material, they only ever throw money into circulation — and only ever take goods out of it. Their constant capital (except the part that wears out), most of their variable capital, and their whole surplus-value (except for whatever hoard piles up in their own hands) all get thrown into circulation as money.

On one hand, all kinds of things circulate as commodities that weren't produced within the year at all — land, houses, and so on — and also products whose production period stretches over more than a year: cattle, timber, wine, and the like. For these and other cases, it matters to keep in mind that, besides the sum of money needed for immediate circulation, there is always a certain amount sitting idle, doing no work, which can spring into action the moment something calls for it. And the value of such products often circulates bit by bit and gradually too — like the value of a house, paid off piece by piece through years of rent.

On the other hand, not every movement within the reproduction process runs through money circulation at all. The whole process of production itself, once its elements have been bought, has nothing more to do with money. And so does all the product that the producer goes straight back to consuming himself — whether for his own use or productively — which includes paying rural workers in kind rather than in money.

So the mass of money that circulates the annual product is already there in society — built up gradually over time. It is not part of this year's newly produced value, except perhaps for the gold that replaces worn-out coins.

This whole account assumes that only precious-metal money is circulating, and, within that, the simplest form of cash purchase and sale — even though, on the basis of plain metallic circulation, money can also serve as a means of payment, and historically really has done so, and on that basis a credit system, with certain sides of how it works, has grown up.

This assumption isn't made only for reasons of method — though the weight of those reasons shows up in the fact that Tooke and his school, and their opponents alike, kept being forced, whenever they argued about banknote circulation, to fall back on the hypothesis of purely metallic circulation. They were forced into this after the fact, and then did it only superficially — necessarily so, because on their approach the starting point only ever plays the role of an incidental point in the analysis.

But simply looking, in the plainest way, at money circulation as it actually takes shape by itself — and here that circulation is a built-in part of the annual reproduction process — shows the following:

Assume capitalist production is fully developed — that is, the wage-labour system rules. Then money-capital plainly plays a leading role, as the form in which variable capital gets advanced. As the wage-labour system spreads, every product turns into a commodity, so — with a few important exceptions — absolutely all of it must pass through a stage of becoming money as part of its movement. The mass of money in circulation must be enough to turn all these goods into money, and the largest part of that mass is supplied as wages: money that industrial capitalists advance, as the money-form of variable capital, to pay for labour-power, and that in the workers' hands — for the great bulk of it — only ever functions as a means of circulation, a means of buying things. This is the complete opposite of a natural economy, the kind that prevails under every system of bondage, serfdom included, and even more so among more or less primitive communities — whether or not those communities are mixed up with relations of bondage or slavery.

Under slavery, the money-capital spent buying labour-power plays a different role: it is the money-form of fixed capital, only replaced gradually, once the slave's working life is over. That is why, among the Athenians, the profit a slave-owner made — whether directly, by putting his slave to industrial use, or indirectly, by hiring him out to other users, say for work in the mines — was reckoned simply as interest, plus repayment of the capital, on the money-capital he had advanced. It is exactly how, under capitalist production, an industrial capitalist counts part of his surplus-value, plus the wear on his fixed capital, as interest and replacement for that fixed capital — and exactly the rule, too, for capitalists who rent out fixed capital like houses or machines. Ordinary household slaves, whether doing necessary work or serving as pure luxury display, don't belong here — they correspond to our servant class. But even the slave system — wherever it was the dominant form of productive labour, in agriculture, manufacturing, shipping, and so on, as in the developed states of Greece and in Rome — kept one foot in natural economy. The slave market itself was constantly restocked with fresh labour-power through war, piracy, and the like, and that plunder was not something money circulation brought about at all — it was the direct seizure of other people's labour-power by naked physical force. Even in the United States, once the borderland between the wage-labour states of the North and the slave states of the South turned into a slave-breeding region supplying the South — so that the slave put up for sale had himself become part of the annual reproduction process — even that wasn't enough for long, and the African slave trade kept being pushed as far as it could go, just to keep the market supplied.

Consider all the ways money naturally flows out and flows back under capitalist production as the annual product changes hands. Fixed capital gets advanced all at once, for its whole value — and then that value is drawn back out of circulation only gradually, spread out over years. So fixed capital gets rebuilt in money-form bit by bit, year by year, through a kind of hoard-building. And this hoard-building is essentially a different thing in kind from the hoard-building that runs alongside it and comes from each year's new gold production. Add to this: money has to be advanced for different lengths of time depending on how long each commodity's production period runs, so it has to keep being hoarded up again and again beforehand, before it can be drawn back out of circulation by selling the goods. The length of that advance also varies simply because production sites sit at different distances from their markets. And the size and timing of the money flowing back varies too, depending on the level — the relative size — of production-stocks in different businesses, and among the different capitalists within the same line of business, which in turn sets the dates on which they buy the elements of their constant capital. All of this happens within a single year of reproduction. None of these naturally-occurring movements needs anything more than being noticed and found striking through experience, for them to give rise, quite systematically, both to the mechanical devices of the credit system and to the actual fishing-up of the loanable capital that is sitting around available.

On top of this comes another difference: between businesses whose production, all else normal, runs on continuously at the same scale, and businesses that employ labour-power in very different amounts depending on the time of year — agriculture, for instance.

Engels notes that the closing section on Destutt de Tracy is taken from Marx's Manuscript II.

Take Destutt de Tracy as an example of the muddled, self-important carelessness political economists bring to the question of social reproduction — this "great logician" whom even Ricardo took seriously, calling him "a very distinguished writer" (Principles, p. 333).

This distinguished writer offers the following account of the whole process of social reproduction and circulation:

"People will ask me how these industrial entrepreneurs make such large profits, and from whom they can draw them. My answer is that they do it by selling everything they produce for more than it cost them to produce it — and that they sell it, first, to each other, for that whole part of their consumption spent on meeting their own needs, which they pay for out of part of their profits;"

"second, to the wage-workers — both the ones they themselves employ and the ones employed by the idle capitalists — from whom they get back, by this route, the whole of the wages they paid out, except perhaps for a few small savings;"

"and third, to the idle capitalists, who pay them out of the part of their revenue that they have not already handed over to the wage-workers they employ directly — so that the whole rent the industrialists pay out each year flows back to them by one or another of these routes." (Destutt de Tracy, Traité de la volonté et de ses effets, Paris 1826, p. 239.)

So, on this first count, the capitalists get richer by overcharging each other when they trade among themselves the part of the surplus-value spent on their own consumption, or consumed as revenue. Say that part comes to £400. Because each of them marks up what he sells to the others by a quarter, that same £400 turns into about £500. But since everyone does the same thing to everyone else, the net result is exactly as if they had all traded at the true value — except that circulating £400 worth of goods now takes £500 in money. That looks less like a way of getting richer than a way of getting poorer: they have to keep a large part of their whole wealth sitting idle and unproductive, in the useless form of extra circulating money. Strip away the general, nominal rise in prices, and the capitalist class as a whole still has only £400 worth of goods to divide up among themselves for their own consumption — they have simply given themselves the mutual pleasure of moving £400 worth of goods with £500 worth of money.

And that is quite apart from the fact that "a part of their profits" — and so a stock of goods in which profit is already represented — is simply assumed here. But it is exactly where this profit comes from that Destutt is supposed to be explaining to us. How much money is needed to circulate it is a distinctly secondary question. The mass of goods in which the profit is represented seems to arise from the capitalists not merely selling this mass of goods to each other — already a fine and profound thought — but all overcharging each other in the process. So now we know one source of the capitalists' enrichment. It comes down to the old joke that great poverty comes from great poverty — say it in French and it sounds like a discovery.

These same capitalists, Destutt goes on, further sell "to the wage-workers — both the ones they themselves employ and the ones employed by the idle capitalists — from whom they get back, in this way, the whole of their wages, except for their small savings."

The reflux of the money-capital — the very form in which the capitalists had advanced wages to the worker — back into the capitalists' hands makes up, according to Herr Destutt, this second source of their enrichment.

So say the capitalist class pays workers £100 in wages, and those same workers then buy back, from that very capitalist class, goods worth that same £100 — so that the £100 the capitalists advanced to buy labour-power flows back to them when they sell the workers £100 worth of goods: on this telling, the capitalists get richer by it. From the standpoint of ordinary common sense, it looks as though this procedure leaves the capitalists back in possession of the £100 they had before it started. At the beginning of the procedure they hold £100 in money. With that £100 they buy labour-power. With that same £100, the labour they have bought produces goods worth — so far as we know — £100. By selling those £100 worth of goods to the workers, the capitalists get their £100 back in money. So the capitalists again have £100 in money, while the workers have £100 worth of goods — which they themselves produced. How the capitalists are supposed to get richer by this is not clear. Had the £100 not flowed back to them, they would first have had to pay the workers £100 in money for their labour, and second have had to hand them the product of that labour — £100 worth of means of consumption — for nothing. So the reflux could explain, at most, why the capitalists end up no poorer for the operation. It could never explain why they become richer through it.

There is, admittedly, a different question: how the capitalists come to have the £100 in the first place, and why the workers, instead of producing goods on their own account, are forced to exchange their labour-power for it. But that is something a thinker of Destutt's calibre simply takes for granted.

Destutt himself is not entirely satisfied with this reflux story. After all, he had not told us that one gets rich by paying out £100 in money and then taking in £100 again — that is, not by the mere reflux of £100, which only shows why the £100 is not lost. He had told us that the capitalists enrich themselves "by selling everything they produce for more than it cost them to buy."

So in their dealings with the workers too, the capitalists must be getting richer by selling to them too dear. Splendid!

"They pay out wages ... and it all flows back to them through the spending of all these people, who pay the capitalists more for the products than those products cost the capitalists by means of this very wage." (p. 240.)

So: the capitalists pay the workers £100 in wages, and then sell the workers their own product for £120, so that not only does the £100 flow back to them but they gain another £20 besides? That is impossible. The workers can only pay with the money they received as wages. If they get £100 in wages from the capitalists, they can only buy £100 worth, not £120. So it cannot work this way. But there is another way. The workers buy goods from the capitalists for £100, but in fact receive only £80 worth of goods. They are unquestionably cheated of £20. And the capitalist has unquestionably enriched himself by £20 — because he has in fact paid for labour-power 20% below its value, or made a 20% deduction from the nominal wage by a roundabout route.

The capitalist class would reach the same result if it simply paid the workers only £80 in wages to begin with, and then actually delivered £80 worth of goods for that £80. Looking at the whole class, this seems to be the normal way — since, according to Destutt himself, the working class must receive "sufficient wages" (p. 219), wages that must at least suffice to maintain their existence and their capacity to work, "to obtain for themselves the barest subsistence" (p. 180). If the workers do not receive these sufficient wages, then — on Destutt's own account — this is "the death of industry" (p. 208): so, it seems, no way for the capitalists to enrich themselves. But whatever level of wages the capitalist class pays the working class, those wages have some definite value — say £80. So if the capitalist class pays the workers £80, it owes them £80 worth of goods for that £80, and the reflux of the £80 does not enrich it. If instead it pays them £100 in money and then sells them, for that £100, goods worth only £80, then it has paid them 25% more than their normal wage in money, and delivered them 25% less in goods.

In other words: the fund from which the capitalist class draws its profit at all would be formed by a deduction from the normal wage — by paying for labour-power below its value, that is, below the value of the means of subsistence necessary for the worker's normal reproduction as a wage-worker. So if the normal wage were paid — which, according to Destutt, is what should happen — there would be no fund of profit at all, neither for the industrialists nor for the idle capitalists.

So Herr Destutt would have had to reduce the whole secret of how the capitalist class enriches itself to just this: a deduction from wages.

The other funds of surplus-value — the ones Destutt lists under 1 and 3 — would then not exist.

In all countries, then, where the workers' money wage is reduced to the value of the means of consumption needed for their subsistence as a class, there would be no consumption fund and no accumulation fund for the capitalists — hence no fund for the capitalist class's own existence at all — hence no capitalist class. And this, according to Destutt, would be the case in all the rich, developed countries of old civilization, since here "in our long-established societies, the fund out of which wages are paid ... is an almost constant magnitude" (p. 202).

Even where wages are docked, the capitalists' enrichment does not come from first paying the worker £100 in money and then delivering him £80 worth of goods for that £100 — in effect circulating £80 worth of goods with a sum of money, £100, that is a quarter too large. It comes from the fact that the capitalist appropriates from the worker's product, besides the surplus-value — the part of the product in which surplus-value is represented — a further 25% of the part of the product that should have fallen to the worker in the form of wages. On Destutt's own silly way of putting it, the capitalist class would gain absolutely nothing. It pays out £100 in wages and gives the worker back, out of his own product, £80 worth of goods for that £100. But for the next round of the very same operation, it must again advance £100. So all it does is give itself the useless pleasure of advancing £100 in money and delivering £80 worth of goods for it, instead of advancing £80 in money and delivering £80 worth of goods for it. That is: it constantly and pointlessly advances, to circulate its variable capital, a money-capital a quarter too large — a rather peculiar method of getting rich.

3. Finally, the capitalist class sells to the idle capitalists, who pay for it with the part of their revenue that they have not already handed over to the wage-workers they employ directly — so that the whole rent they pay the idle capitalists each year flows back to them by one route or another.

We saw earlier that the industrial capitalists pay for the whole of their own consumption, the part meant to satisfy their own needs, out of a portion of their profits.

Say their profits are £200. They spend £100 of it, for instance, on their own personal consumption. But the other half, £100, is not theirs — it belongs to the idle capitalists, that is, the landowners and the capitalists who lend at interest. So they owe this group £100 in money. Now say that of this money, the idle capitalists need £80 for their own consumption and £20 to pay servants. So they use the £80 to buy means of consumption from the industrial capitalists. That sends £80 in money flowing back to the industrial capitalists — while £80 worth of product leaves their hands — which is four-fifths of the £100 they had paid the idle capitalists as rent, interest, and so on. Then the servant class, the direct wage-workers of the idle capitalists, have received £20 from their employers. They too use it to buy £20 worth of means of consumption from the industrial capitalists. That sends £20 in money flowing back to them — while £20 worth of product leaves their hands — the last fifth of the £100 in money paid to the idle capitalists as rent, interest, and so on.

By the end of the transaction, the £100 in money that the industrial capitalists had handed over to the idle capitalists as rent, interest, and so on has flowed back to them — while half of their surplus product, worth £100, has passed out of their hands into the consumption fund of the idle capitalists.

For the question at hand, it turns out to be quite unnecessary to bring in at all how the £100 is split between the idle capitalists and their own direct wage-workers. The matter is simple: their rent, their interest — in short, their share of the £200 surplus-value — is paid to them by the industrial capitalists in money, £100. With this £100 they buy, directly or indirectly, means of consumption from the industrial capitalists. So they pay back £100 in money, and take away £100 worth of means of consumption.

With that, the £100 in money the industrial capitalists paid to the idle capitalists has flowed back to them. But is this reflux of money, as Destutt gushes, a way for the industrial capitalists to get richer? Before the transaction they held a sum of value worth £200: £100 in money and £100 in means of consumption. After the transaction they hold only half of that original sum. They have the £100 in money again, but they have lost the £100 in means of consumption, which have passed into the hands of the idle capitalists. So they are £100 poorer, not £100 richer. Suppose that, instead of taking this roundabout route — first paying out £100 in money, then getting that same £100 back in payment for £100 worth of means of consumption — they had simply paid the rent, interest, and so on directly, in the natural form of their product. Then no £100 in money would have flowed back to them out of circulation at all, because they would never have thrown £100 in money into circulation in the first place. Paid this way, in kind, the matter would simply have looked like this: of the surplus product worth £200, they kept half for themselves and gave the other half away, for nothing, to the idle capitalists. Not even Destutt could have felt tempted to call that a way of getting richer.

The land and the capital that the industrial capitalists borrow from the idle capitalists, and for which they must pay them part of the surplus-value as ground-rent, interest, and so on, were of course profitable to them: they were one of the conditions for producing the product at all, including the part of the product that forms the surplus product, the part in which the surplus-value takes shape. But this profit comes from using the borrowed land and capital, not from the price paid for it. That price is, on the contrary, a deduction from it. Otherwise one would have to claim that the industrial capitalists would become not richer but poorer if they could keep the other half of the surplus-value for themselves instead of giving it away. But that is the confusion you fall into when you lump together circulation phenomena, like the reflux of money, with the distribution of the product — a distribution that such circulation phenomena only mediate.

And yet this same Destutt is sharp enough to observe:

'Where do the revenues of these idle people come from? Do they not come from the rent that is paid to them, out of profit, by those who put the idle people's capital to work — that is, by those who use the idle people's funds to pay for labour that produces more than it costs — in a word, by the industrialists? So it is to the industrialists that one must always go back, to find the source of all wealth. They are the ones who, in reality, feed the wage-workers employed by the idle people.'

So now, paying this rent and so on is a cut taken out of the industrialists' profit. A moment ago, it was supposed to be a way for them to get richer.

But our Destutt still has one consolation left. These upstanding industrialists treat the idle capitalists the way they treat each other, and the way they treat the workers: they overcharge them on every sale, say by 20%. Now there are two possibilities. Either the idle capitalists have money of their own besides the £100 they get every year from the industrialists, or they don't. In the first case, the industrialists sell them £100 worth of goods at a price of, say, £120. So when they sell their goods, not only does the £100 they paid the idle capitalists flow back to them, but an extra £20 besides — and that £20 really is new value for them. How does the sum work out? They gave away £100 worth of goods for nothing, because the £100 in money used to pay for part of it was their own money to begin with — so their own goods have been paid for with their own money. That is a loss of £100. But on top of that they took in £20 from selling above value. £20 gain plus £100 loss makes £80 loss — it never turns into a plus, it stays a minus. Cheating the idle capitalists this way has reduced the industrialists' loss, but it has not turned that loss of wealth into a way of getting richer. This method, though, cannot go on for long, since the idle capitalists cannot possibly keep paying out £120 a year in money if they only take in £100 a year.

So the other method: the industrialists sell goods worth £80 for the £100 in money the idle capitalists paid them. In this case, they are still giving away £80 for nothing, in the form of rent, interest, and so on, just as before. Through this cheating they have reduced the tribute paid to the idle capitalists, but it still exists all the same — and on that very same theory, that prices depend on the seller's good will, the idle capitalists are just as able to demand £120 in rent, interest, and so on for their land and capital in future, instead of the £100 they got before.

This brilliant piece of reasoning is entirely worthy of the profound thinker who, on the one hand, copies from Adam Smith that 'labour is the source of all wealth,' that the industrial capitalists 'use their capital to pay for labour that reproduces it with a profit' — and who, on the other hand, concludes that these same industrial capitalists 'feed everyone else, are the sole ones who increase the public wealth, and create all our means of enjoyment,' that it is not the capitalists who are fed by the workers but the workers who are fed by the capitalists — and for the brilliant reason that the money the workers are paid with does not stay in their hands, but keeps flowing back to the capitalists in payment for the very goods the workers produced.

'They only receive with one hand and give back with the other. Their consumption must therefore be regarded as produced by those who pay their wages.'

After this exhaustive account of social reproduction and consumption, as mediated by the circulation of money, Destutt goes on:

'That is what rounds off this perpetual-motion machine of wealth — a movement which, although poorly understood' (poorly understood, certainly! Marx breaks in here) 'has rightly been called circulation; for it truly is a cycle, always returning to its point of departure. That point is the one where production takes place.'

Destutt, that very distinguished writer, a member of the Institut de France and of the Philosophical Society of Philadelphia, and indeed something of a luminary among the vulgar economists, finally asks the reader to admire the wonderful clarity with which he has laid out the course of the social process, the flood of light he has poured over the subject — and is even condescending enough to let the reader know where all this light comes from. This has to be given in the original:

'One will notice, I hope, how consistent this way of looking at the consumption of our wealth is with everything we have said about its production and its distribution, and at the same time what clarity it spreads over the whole course of society. Where do this consistency and this clarity come from?'

'From the fact that we have hit upon the truth. It recalls the effect of those mirrors in which objects are pictured clearly, in their true proportions, when you stand at the right vantage point — and in which everything looks confused and blurred when you are too close or too far away.'

Now that is bourgeois cretinism in all its blissful glory!