Let us suppose the demand for corn is rising, and that it could be satisfied only in one of three ways: by successive capital investments with deficient productivity on the rent-bearing lands; by an additional capital investment on land A, likewise with declining productivity; or by investing capital in new land of poorer quality than A.

Let B stand in for the rent-bearing soils.

The extra capital investment calls for the market price to rise above the production price of £3 per qr that has governed so far, if the extra quarter is to be produced on B. (That 1 qr may stand for a million qrs, and each acre for a million acres.) On C and D, the soils that carry the highest rent, extra product may also appear, though only with declining surplus productivity; that one quarter from B, however, is assumed to be necessary if demand is to be met.

Now suppose that quarter can be raised more cheaply by putting extra capital into B than by putting the same extra capital into A, or than by moving down to a soil worse still, A-1 — which, say, can produce the quarter only at £4, while the extra capital on A could already produce it at £3 3/4 per qr. Then it is B's extra capital that would govern the market price.

Suppose A has produced 1 qr at £3, as before. B too, as before, has produced 3 1/2 qrs in all, at an individual production price of £6 in all. Now if another quarter on B required an extra £4 of production costs (profit included), while on A it could be produced for £3 3/4, then obviously it would be produced on A, not on B. So let us suppose instead that it can be produced on B for an extra £3 1/2 in production costs.

Then £3 1/2 would rule as the price for the whole output. B's product is now 4 1/2 qrs, and it would sell for £15 3/4. The production costs to be deducted are £6 for the first 3 1/2 qrs and £3 1/2 for the last quarter — £9 1/2 in all. What remains as surplus profit, available for rent, is £6 1/4. Previously it had been only £4 1/2.

On this supposition, A's acre would also yield £1/2 in rent. But the £3 1/2 production price would be governed not by A, the worst land, but by B, the better land.

Of course, this assumes no new land of A's quality, as favourably placed as the land already under cultivation, can be had. What would be needed instead is a second investment of capital on the stretch of A already farmed, though at greater production costs, or the bringing in of still poorer land, A-1. Once successive capital investments bring differential rent II into operation, better land can set the limits on how far the production price rises; and then the worst land, the basis of differential rent I, can bear rent as well. Under simple differential rent, then, all cultivated land would bear rent.

We would then have the two tables below. Here the price of production means capital advanced plus 20 per cent profit: on each £2 1/2 of capital, £1/2 profit, £3 altogether.

That is where things stand before B gets the new £3 1/2 of capital, which brings in only 1 qr. Once that investment is made, the picture is this:

This is again not quite correctly calculated. The farmer of B pays for his 4 1/2 qrs, first, £9 1/2 in production costs and, second, £4 1/2 in rent, together £14; the average per qr = £3 1/9. This average over his whole output thereby becomes the price that governs the market. On that basis the rent on A would come to £1/9 instead of £1/2, while the rent on B would remain £4 1/2 as before: 4 1/2 qrs at £3 1/9 = £14, from which £9 1/2 in production costs is deducted, leaving £4 1/2 as surplus profit. So even though the figures have to be altered, the example shows how, by way of differential rent II, the better land that already bears rent can govern the price, and how in this way all land, even land that was previously rentless, can be turned into rent-bearing land. — F. E.

Corn rent has to rise as soon as the governing production price of corn rises — that is, as soon as either the price of a quarter of corn on the governing soil rises, or the governing capital investment on one of the soil types rises. It is the same as if every soil type had become less fertile and, with a new capital investment of, say, £2 1/2, each of them produced only 5/7 qr instead of 1 qr. Whatever extra corn they produce with the same capital investment turns into surplus product — the form in which surplus profit, and therefore rent, is represented. Suppose the profit rate stayed the same. Then the farmer can buy less corn with his profit. And the profit rate can stay the same if wages do not rise: either because wages are pressed down to the physical minimum, below the normal value of labour-power; or because the other things the worker consumes, the ones manufacture supplies, have become relatively cheaper; or because the working day is lengthened or made more intense, so that the profit rate in the non-agricultural branches of production — which is what governs agricultural profit — has stayed the same, if it has not risen; or because, in agriculture, the same capital is laid out but with more constant and less variable capital.

So we have now looked at the first way rent can arise on A, the poorest land cultivated up to now, without any still worse land being brought under cultivation. It comes from the gap between A's own individual production price, which governed until now, and the new, higher production price — the price at which the last extra capital, working with weaker productivity on better soil, supplies the extra product that is needed.

If the extra product had had to come from land A-1, which can supply a quarter only at £4, the rent per acre on A would have risen to £1. But in that case A-1 would have taken A's place as the worst cultivated land, and A would have entered the series of rent-bearing soil types as its lowest member. Differential rent I would have changed. So the situation is not one that differential rent II deals with — that is the rent that comes from capital put into the same piece of land again and again, with each successive round differently productive.

But on land A a differential rent can still come into being, in two further ways.

The first way: the price stays where it is — any given price, and that price may itself be lower than the one before — and the extra capital put into A brings surplus productivity. On the face of it, that must always be so up to a certain point, and especially on the worst land.

The second way is the opposite: the productivity of one capital investment after another on land A falls.

In both cases we assume that the state of demand requires the extra production.

From the standpoint of differential rent, though, a peculiar difficulty arises here. It comes from the law worked out earlier: what always decides is the individual average production price of a quarter across the whole production — or across the whole capital laid out. With land A, though, there is no production price given outside it, as there is with the better soils, to restrict the equalization of the individual production price with the general one when new capital is invested. For A's individual production price is exactly the general production price that governs the market.

Suppose the following.

(1) Suppose the successive investments get more productive. Then on one acre of A, £5 of capital advanced — £6 of production costs, with the profit already counted in — can produce 3 qrs instead of 2. The first investment of £2 1/2 gives 1 qr, the second gives 2 qrs. Here £6 of production costs yields 3 qrs, so the quarter costs £2 on average. If the 3 qrs are sold at £2, A still bears no rent; what has changed is only the basis of differential rent II. £2 has now become the governing production price instead of £3; £2 1/2 of capital now produces on average 1 1/2 qrs instead of 1 qr on the poorest land, and this is now the official yield for every higher type of soil when £2 1/2 is invested. From now on, part of what used to be their surplus product counts towards building their necessary product — just as part of their surplus profit counts towards building up the average profit.

Now count it the way we do on the better soils. There the average calculation changes nothing in the absolute surplus, because for those soils the general production price is given as a barrier that capital investment runs into. On A, then, the quarter from the first investment costs £3, and each of the two quarters from the second costs only £1 1/2. So a corn rent of 1 qr and a money rent of £3 would arise on A, while the 3 qrs would still be sold at the old price — £9 for the lot.

Suppose a third investment of £2 1/2 follows, with the same fertility as the second. Then 5 qrs would be produced for £9 of production costs. If A's own average production price stayed the regulating one, the quarter would now have to sell at £1 4/5. The average price would have fallen again — not because the third investment was newly more fertile, but only because a new investment had been added with the same extra fertility as the second. Investments like these raise the rent on rent-bearing soils. On A they would do the opposite: higher but constant fertility, added investment after investment, would lower the production price and with it — other things unchanged — the differential rent on all the other soils, in proportion.

If instead the first investment, which produces 1 qr at £3 of production costs, stayed the measure on its own, the 5 qrs would sell for £15, and the differential rent of the later investments on A would come to £6. Adding more capital to the acre of A, in whatever form, would be an improvement here, and the extra capital would also have made the original part of the capital more productive. It would be nonsense to say that a third of the capital produced 1 qr and the other two thirds produced 4 qrs. £9 per acre would always produce 5 qrs, while £3 would produce only 1 qr.

Whether a rent arose here or not — a surplus profit — would depend entirely on circumstances. Normally, the regulating production price would have to fall.

That is what happens when this improved but more expensive cultivation of A is adopted only because the better soils adopt it too — a general revolution in agriculture. Now, when we speak of the natural fertility of A, we are assuming that it is worked with £6 or £9 instead of £3. This would hold especially if the majority of the cultivated acres of A, the ones that supply the bulk of the country's corn, were put under the new method.

But if the improvement took in only a small part of A's area at first, that better-worked part would yield a surplus profit, and the landowner would be quick to turn it wholly or partly into rent and fix it as rent. So if demand kept pace with the growing supply, then as the whole area of A was gradually brought under the new method, rent could form bit by bit on all land of quality A, and the surplus productivity be confiscated in whole or in part, according to market conditions. Once more capital is laid out, fixing the surplus profit of that extra outlay as rent could in this way block A's production price from levelling to the average price of its product.

In that case it would again be — as we saw earlier on the better lands, when the additional capitals had declining productivity — the conversion of surplus profit into ground-rent, that is, the intervention of landed property, which would raise the production price, whereas otherwise differential rent would be nothing but the consequence of the differences between the individual and the general production price. For land A it would stop the two prices from coinciding, since it would stop A's average production price from regulating; so it would keep a production price higher than necessary in being, and create rent by doing so. The same result could be produced, or could persist, even with free corn imports from abroad: farmers would be forced to turn land that could compete in corn-growing without bearing rent, at the price set from outside, to another use, say pasture, so that only rent-bearing land would be used for corn — only land whose own average production price per quarter was lower than the price set from outside.

On the whole, it is to be assumed that in the given case the production price will fall — but not as far as its average price. It will stand higher than that, and yet below the production price of the worst cultivated land A, so that competition from new land of A is restricted.

(2) The productivity of the extra capitals is falling. Suppose land A-1 can produce the additional quarter only at £4, while land A can produce it at £3 3/4 — cheaper, but £3/4 dearer than the quarter that A's first capital investment produced. In that case the two quarters produced on A would together cost £6 3/4, so the average price per quarter would be £3 3/8. The production price would go up — but only by £3/8. If instead the extra capital went into new land that produced at £3 3/4, it would go up by a further £3/8, to £3 3/4, and that would raise all the other differential rents in proportion.

With the larger capital investment, A's production price of £3 3/8 per quarter would then be levelled to its own average production price, and that would be the governing price. So A would pay no rent, because it would earn no surplus profit.

But if that quarter from the second capital investment were sold at £3 3/4, land A would now yield a rent of £3/4, and every acre of A would bear it — including the acres where no extra capital was invested, which still produce their quarter at £3.

As long as untilled stretches of A are still there, the price could rise to £3 3/4 only for a time. Competition from new stretches of A would hold the production price down at £3 until all the A land whose favourable situation lets it produce a quarter for less than £3 3/4 had been used up. That is what we would assume, even though a landowner who has one acre paying rent will not let another acre go to a tenant rent-free.

It would again depend on how general the second capital investment has become on the available A land. The question is whether the production price is equalized with the average price, or whether the individual production price of the second investment, £3 3/4, becomes the governing one. This second outcome happens only if the landowner gains the time to turn into rent the surplus profit that would be earned while demand is still being met at £3 3/4 a quarter.

On the declining productivity of the soil when capital is invested in it again and again, Liebig is the one to consult. We have seen that as the surplus productivity of successive capital investments falls away, the rent per acre always rises, so long as the price of production stays the same — and that it can rise even when that price is falling.

But there is a general point to note here.

From the standpoint of the capitalist mode of production, products always become relatively dearer when, to get the same product, an outlay has to be made, something has to be paid for that was not paid for before.

For when we speak of replacing the capital used up in production, we mean only the replacement of values that were represented in particular means of production. Natural elements that enter production as agents and cost nothing do not enter it as components of capital, whatever role they may play there. They enter as a free natural power of capital — that is, as a free natural productive power of labour. On the basis of the capitalist mode of production, though, this power presents itself as a productive power of capital, like every productive power.

So if a natural power of this kind, which originally costs nothing, goes into production, it does not count in the determination of prices as long as the product supplied with its help is enough for demand. But if, as development proceeds, a greater product has to be supplied than can be produced with the help of this natural power — so that this additional product must be produced without its help, or with human assistance, human labour — then a new, additional element enters capital. Relatively more capital outlay is therefore needed to get the same product. With all other circumstances unchanged, production becomes dearer.

From a notebook, ‘Begun mid-February 1876’.

So-called permanent improvements change the physical character of the soil, and part of its chemical character as well, through operations that cost a capital outlay and can be regarded as capital worked into the soil. Almost all of them come down to one thing: giving one particular patch of land — soil in one definite, limited spot — qualities that some other patch, in another spot and often quite close by, has by nature. One plot is level already, while another has to be levelled; one drains itself, while another has to be drained; one has a deep layer of topsoil by nature, while in another it has to be dug deeper; one clay soil already carries the right amount of sand, while in another that mix has to be created; one meadow is irrigated or covered with layers of silt by nature, while another has to be made so by labour — or, to use the language of bourgeois economics, by capital.

Now here is a genuinely comical theory: on the one plot, whose comparative advantages were acquired, rent is interest; on the other, which has those advantages by nature, it is not. (When the thing is actually carried through, though, it is distorted into this: because rent really does coincide with interest in the one case, it has to be called interest in the other cases too — where that is positively not so — and lied into interest.)

But once the investment has been made, the land bears rent not because capital was laid out on it, but because the investment has turned it into a more productive field of investment than it was before. Suppose every plot of land in a country needed this investment. Then every plot that has not yet had it has to pass through that stage first, and the rent borne by the land already furnished with the investment — the interest it yields in the given case — is just as much a differential rent as if it had that advantage by nature and the other plot had to acquire it artificially.

This rent too, the kind that can be resolved into interest, becomes pure differential rent as soon as the capital laid out has been amortized. Otherwise the same capital would have to exist twice over, as capital.

It is one of the most laughable spectacles. Every one of Ricardo's opponents fights the idea that labour alone determines value. But when differential rent comes out of differences in land, they insist that here nature, not labour, is made the thing that determines value. At the same time they claim that same determining role for the land's position — or even, and still more, for the interest on capital put into the soil to cultivate it.

The same labour produces the same value in the product made in a given time. But the size or quantity of that product — and so too the portion of value that falls to any equal part of it — depends, for a given quantity of labour, solely on the quantity of the product. And that in turn depends on how productive the given quantity of labour is, not on how large that quantity of labour is. Whether this productivity is owed to nature or to society makes no difference at all. Only when the productivity itself costs labour — that is, capital — does it add a new component to the costs of production. With bare nature that is not the case.