Principles of Economics by Alfred Marshall (1890)

Book Five: General Relations of Demand, Supply and Value

Chapter 11, Marginal Costs in Relation to Urban Value

1. The last three chapters examined the relation in which cost of production stands to the income derived from the ownership of the "original powers" of land and other free gifts of nature, and also to that which is directly due to the investment of private capital. There is a third class, holding an intermediate position between these two, which consists of those incomes, or rather those parts of incomes which are the indirect result of the general progress of society, rather than the direct result of the investment of capital and labour by individuals for the sake of gain. This class has to be studied now, with special reference to the value of urban sites.

We have already noted that, though nature nearly always gives a less than proportionate return, when measured by the amount of the produce raised, to increasing applications of capital and labour in the cultivation of land; yet, on the other hand, if the more intensive cultivation is the result of the growth of a non-agricultural population in the neighbourhood, this very concourse of people is likely to raise the value of produce. We have seen how this influence opposes, and usually outweighs the action of the law of diminishing return when the produce is measured according to its value to the producer and not according to its amount; the cultivator gets good markets in which to supply his wants, as well as good markets in which to sell, he buys more cheaply while he sells more dearly, and the conveniences and enjoyments of social life are ever being brought more within his reach. (1)

Again, we have seen how the economies which result from a high industrial organization (2) often depend only to a small extent on the resources of individual firms. Those internal economies which each establishment has to arrange for itself are frequently very small as compared with those external economies which result from the general progress of the industrial environment; the situation of a business nearly always plays a great part in determining the extent to which it can avail itself of external economies; and the situation value which a site derives from the grow of a a rich and active population close to it, or from the opening up of railways and other good means of communication with existing markets, is the most striking of all the influences which changes in the industrial environment exert on cost of production.

If in any industry, whether agricultural or not, two producers have equal facilities in all respects, except that one has a more convenient situation than the other, and can buy or sell in the same markets with less cost of carriage, the differential advantage which his situation gives him is the aggregate of the excess charges for cost of carriage to which his rival is put. And we may suppose that other advantages of situation, such for instance as the near access to a labour market specially adapted to his trade, can be translated in like manner into money values. When this is done, and all are added together we have the money value of the advantages of situation which the first business has over the second: and this becomes its special situation value, if the second has no situation value and its site is reckoned merely at agricultural value. The extra income which can be earned on the more favoured site gives rise to what may be called a special situation rent: and the aggregate site value of any piece of building land is that which it would have if cleared of buildings and sold in a free market. The "annual site value" — to use a convenient, though not strictly correct form of speaking — is the income which that price would yield at the current rate of interest. It obviously exceeds the special situation value, merely by agricultural value; which is often an almost negligible quantity in comparison. (3)

2. It is obvious that the greater part of situation value is "public value." There are however exceptional cases, which call for notice. Sometimes the settlement of a whole town, or even district is planned on business principles, and carried out as an investment at the expense and risk of a single person or company. The movement may be partly due to philanthropic or religious motives, but its financial basis will in any case be found in the fact that the concourse of numbers is itself a cause of increased economic efficiency. Under ordinary circumstances the chief gains arising from this efficiency would accrue to those who are already in possession of the place: but the chief hopes of commercial success, by those who undertake to colonize a new district or build a new town, are usually founded on securing these gains for themselves.

When, for instance, Mr Salt and Mr Pullman determined to take their factories into the country and to found Saltaire and Pullman City, they foresaw that the land, which they could purchase at its value for agricultural purposes, would obtain the special situation value which town property derives from the immediate neighbourhood of a dense population. And similar considerations have influenced those, who, having fixed upon a site adapted by nature to become a favourite watering-place, have bought the land and spent large sums in developing its resources: they have been willing to wait long for any net income from their investment in the hope that ultimately their land would derive a high situation value from the concourse of people attracted to it. (4)

In all such cases the yearly income derived from the land (or at all events that part of it which is in excess of the agricultural rent) is for many purposes to be regarded as profits rather than rent. And this is equally true, whether the land is that on which the factory itself at Saltaire or Pullman City is built, or that which affords a high "ground-rent" as the site of a shop or store, whose situation will enable it to do a brisk trade with those who work in the factory. For in such cases great risks have to be run; and in all undertakings in which there are risks of great losses, there must also be hopes of great gains. The normal expenses of production of a commodity must include payment for the ventures required for producing it, sufficient to cause those who are on the margin of doubt whether to venture or not, to regard the probable net amount of their gains net, that is, after deducting the probable amount of their losses as compensating them for their trouble and their outlay. And that the gains resulting from such ventures are not much more than sufficient for this purpose is shown by the fact that they are not as yet very common. They are however likely to be more frequent in those industries which are in the hands of very powerful corporations. A large railway company, for instance, can found a Crewe or a New Swindon for manufacturing railway plant without running any great risk. (5)

Some what similar instances are those of a group of landowners who combine to make a railway, the net traffic receipts of which are not expected to pay any considerable interest on the capital invested in making it; but which will greatly raise the value of their land. In such cases part of the increase of their incomes as landowners ought to be regarded as profits on capital which they have invested in the improvement of their land: though the capital has gone towards making a railway instead of being applied directly to their own property.

Other cases of like nature are main drainage schemes, and other plans for improving the general condition of agricultural or town property, in so far as they are carried out by the landowners at their own expense, whether by private agreement or by the levying of special rates on themselves. Similar cases again are found in the investment of capital by a nation in building up its own social and political organization as well as in promoting the education of the people and in developing its sources of material wealth.

Thus that improvement of the environment, which adds to the value of land and of other free gifts of nature, is in a good many cases partly due to the deliberate investment of capital by the owners of the land for the purpose of raising its value; and therefore a portion of the consequent increase of income may be regarded as profits when we are considering long periods. But in many cases it is not so; and any increase in the net income derived from the free gifts of nature which was not brought about by, and did not supply the direct motive to, any special outlay on the part of the landowners, is to be regarded as rent for all purposes.

Cases somewhat analogous to these arise when the owner of a score or more of acres in the neighbourhood of a growing town "develops" them for building. He probably lays out the roads, decides where houses are to be continuous, and where detached; and prescribes the general style of architecture, and perhaps the minimum expenditure on each house; for the beauty of each adds to the general value of all. This collective value, thus created by him, is of the nature of public value; and it is dependent, for the greater part, on that dormant public value, which the site as a whole derived from the growth of a prosperous town in its neighbourhood. But yet that share of it which results from his forethought, constructive faculty and outlay, is to be regarded as the reward of business enterprise, rather than as the appropriation of public value by a private person.

These exceptional cases must be reckoned with. But the general rule holds that the amount and character of the building put upon each plot of land is, in the main (subject to the local building bylaws), that from which the most profitable results are anticipated, with little or no reference to its reaction on the situation value of the neighbourhood. In other words the site value of the plot is governed by causes which are mostly beyond the control of him who determines what buildings shall be put on it: and he adjusts his expenditure on it to his estimates of the income to be derived from various descriptions of buildings on it.

3. The owner of building land sometimes builds on it himself: sometimes he sells it outright: very often he lets it at a fixed ground-rent for ninety-nine years, after which the land and the buildings on it (which by covenant must be kept in good repair) revert to his successor in title. Let us consider what governs the value at which he can sell the land and the ground-rent at which he can let it.

The capitalized value of any plot of land is the actuarial "discounted" value of all the net incomes which it is likely to afford, allowance being made on the one hand for all incidental expenses, including those of collecting the rents, and on the other for its mineral wealth, its capabilities of development for any kind of business, and its advantages, material, social and aesthetic, for the purposes of residence. The money equivalent of that social status and those other personal gratifications which the ownership of land affords, does not appear in the returns of the money income derived from it, but does enter into its capital money value. (6)

Next let us consider what governs the "ground-rent" which the owner can obtain for a plot which he lets on, say, a ninety-nine years' building lease. The present discounted value of all the fixed money payments under that lease tends to be equal to the present capital value of the land; after deducting, firstly, for the obligation to return the land with the buildings on it to the successor in title of the present owner at the end of the lease, and secondly for the possible inconvenience of any restrictions on the use of the land contained in the lease. In consequence of these deductions the ground-rent would be rather less than the "annual site value" of the land, if that site value were expected to remain fixed throughout. But in fact the site value is expected to rise in consequence of the growth of population, and other causes: and therefore the ground-rent is generally a little above the annual site value at the beginning of the lease, and much below it towards the end. (7)

Among the estimated outgoings on account of any building, which have to be deducted from its estimated gross yield before deciding what is the value of the privilege of erecting it on any given plot of land, are the taxes (central and local) which may be expected to be levied on the property, and to be paid by the owner of the property. But this raises difficult side-issues, which are postponed to Appendix G.

4. Let us revert to the fact that the law of diminishing return applies to the use of land for the purposes of living and working on it in all trades. (8) Of course in the trade of building, as in agriculture, it is possible to apply capital too thinly. Just as a homesteader may find that he can raise more produce by cultivating only a half of the 160 acres allotted to him than by spreading his labour over the whole, so even when ground has scarcely any value, a very low house may be dear in proportion to its accommodation. But, as in agriculture, there is a certain application of capital and labour to the acre which gives the highest return, and further applications after this give a less return, so it is in building. The amount of capital per acre which gives the maximum return varies in agriculture with the nature of the crops, with the state of the arts of production, and with the character of the markets to be supplied; and similarly in building, the capital per square foot which would give the maximum return, if the site had no scarcity value, varies with the purpose for which the building is wanted. But when the site has a scarcity value, it is worth while to go on applying capital beyond this maximum rather than pay the extra cost of land required for extending the site. In places where the value of land is high, each square foot is made to yield perhaps twice the accommodation, at more than twice the cost, that it would be made to give, if used for similar purposes where the value of land is low.

We may apply the phrase the margin of building to that accommodation which it is only just worth while to get from a given site, and which would not be got from it if land were less scarce. To fix the ideas, we may suppose this accommodation to be given by the top floor of the building. (9)

By erecting this floor, instead of spreading the building over more ground, a saving in the cost of land is effected, which just compensates for the extra expense and inconvenience of the plan. The accommodation given by this floor, when allowance has been made for its incidental disadvantages, is only just enough to be worth what it costs without allowing anything for the rent of land; and the expenses of production of the things raised on this floor, if it is part of a factory, are just covered by their price; there is no surplus for the rent of land. The expenses of production of manufactures may then be reckoned as those of the goods which are made on the margin of building, so as to pay no rent for land. That is to say the rent of the land does not enter into that set of expenses at the margin at which the action of the forces of demand and supply in governing value may be most clearly seen.

Suppose, for instance, that a person is planning a hotel or a factory; and considering how much land to take for the purpose. If land is cheap he will take much of it; if it is dear he will take less and build high. Suppose him to calculate the expenses of building and working his establishment with frontages of 100 and 110 feet respectively, in ways equally convenient on the whole to himself, his customers and employees, and therefore equally profitable to himself. Let him find that the difference between the two plans, after capitalizing future expenditure, shows an advantage of 500 in favour of the larger area; he will then be inclined to take the larger if the land is to be got at less than 50 per foot of frontage, but not otherwise; and 50 will be the marginal value of land to him. He might have reached this result by calculating the increased value of the business that could be done with the same outlay in other respects on the larger site as compared with the smaller, or again by building on less expensive ground instead of in a more favourable situation. But, by whatever route he makes his calculation, its character is similar to that by which he decides whether it is worth his while to buy business plant of any other kind: and he regards the net income (allowance being made for depreciation) which he expects to get from either investment as standing in the same general relation to his business; and if the advantages of the situation are such, that all the land available on it can find employments of different kinds in each of which its marginal use is represented by a capital value of 50 per foot of frontage, then that will be the current value of the land.

5. This assumes that the competition for land for various uses will cause building in each locality and for each use to be carried up to that margin, at which it is no longer profitable to apply any more capital to the same site. As the demand for residential and business accommodation in a district increases, it becomes worth while to pay a higher and higher price for land, in order to avoid the expense and inconvenience of forcing more accommodation from the same ground area.

For instance, if the value of land in, say, Leeds rises because of the increased competition for it by shops, ware. then a woollen manufacturer finding houses, iron works, etc., his expenses of production increased, may move to another town or into the country; and thus leave the land on which he used to work to be built over with shops and warehouses, for which a town situation is more valuable than it is for factories. For he may think that the saving in the cost of land that he will make by moving into the country, together with other advantages of the change, will more than counterbalance its disadvantages. In a discussion as to whether it was worth while to do so, the rental value of the site of his factory would be reckoned among the expenses of production of his cloth; and rightly.

But we have to go behind that fact. The general relations of demand and supply cause production to be carried up to a margin at which the expenses of production (nothing being entered for rent) are so high that people are willing to pay a high value for additional land in order to avoid the inconvenience and expense of crowding their work on to a narrow site. These causes govern site value; and site value is therefore not properly regarded as governing marginal costs.

Thus the industrial demand for land is in all respects parallel to the agricultural. The expenses of production of oats are increased by the fact that land, which could yield good crops of oats, is in great demand for growing other crops that enable it to yield a higher rent: and in the same way the printing-presses, which may be seen at work in London some sixty feet above the ground, could afford to do their work a little cheaper if the demand for ground for other uses did not push the margin of building up so high. Again a hop-grower may find that on account of the high rent which he pays for his land, the price of his hops will not cover their expenses of production where he is, and he may abandon hop-growing, or seek other land for it; while the land that he leaves may perhaps be let to a market-gardener. After a while the demand for land in the neighbourhood may again become so great that the aggregate price which the market-gardener obtains for his produce will not pay its expenses of production, including rent; and so he in his turn makes room for, say, a building company.

In each case the rising demand for land alters the margin to which it is profitable to carry the intensive use of land: the costs at this margin indicate the action of those fundamental causes which govern the value of the land. And at the same time they are themselves those costs to which the general conditions of demand and supply compel value to conform: and therefore it is right for our purpose to go straight to them; though any such inquiry would be irrelevant to the Purposes of a private balance sheet.

6. The demand for exceptionally valuable urban land comes from traders of various kinds, wholesale and retail, more than from manufacturers; and it may be worth while to say something here as to the very interesting features of demand that are peculiar to their case.

If two factories in the same branch of trade have equal outputs they are sure to have nearly equal floor space. But there is no close relation between the size of trading establishments and their turn-overs. Plenty of space is for them a matter of convenience and a source of extra profit. It is not physically indispensable; but the larger their space, the greater the stock which they can keep on hand, and the greater the advantage to which they can display specimens of it; and especially is this the case in trades that are subject to changes of taste and fashion. In such trades the dealers exert themselves to collect within a comparatively small space representatives of all the best ideas that are in vogue, and still more of those that are likely soon to be so; and the higher the rental values of their sites the more prompt they must be in getting rid, even at a loss, of such things as are a little behind the time and do not improve the general character of their stocks. If the locality is one in which customers are more likely to be tempted by a well-chosen stock than by low prices, the traders will charge prices that give a high rate of profit on a comparatively small turn-over: but, if not, they will charge low prices and try to force a large business in proportion to their capital and the size of their premises; just as in some neighbourhoods the market-gardener finds it best to gather his peas young when they are full of flavour, and in others to let them grow till they weigh heavily in the scales. Whichever plan the traders follow, there will be some conveniences which they are in doubt whether it is worth while to offer to the public; since they calculate that the extra sales gained by such conveniences are only just remunerative, and do not contribute any surplus towards rent. The goods which they sell in consequence of these conveniences, are goods into whose expenses of marketing rent does not enter any more than it does into those of the peas which the market-gardener only just finds it worth his while to produce.

Prices are low in some very highly-rented shops, because their doors are passed by great numbers of people who cannot afford to pay high prices for the gratification of their fancy; and the shopkeeper knows that he must sell cheaply, or not sell at all. He has to be content with a low rate of profit each time he turns over his capital. But, as the wants of his customers are simple, he need not keep a large stock of goods; and he can turn over his capital many times a year. So his annual net profits are very great, and he is willing to pay a very high rent for the situation in which they can be earned. On the other hand, prices are very high in some of the quiet streets in the fashionable parts of London and in many villages; because in the one case customers must be attracted by a very choice stock, which can only be sold slowly; and in the other the aggregate turn-over is very small indeed. In neither place can the trader make profits that will enable him to pay as high a rent as those of some cheap but bustling shops in the East end of London.

It is however true that, if without any increase in traffic such as brings extra custom, a situation becomes more valuable for purposes other than shopkeeping; then only those shopkeepers will be able to pay their way who can manage to secure a large custom relatively to the prices which they charge and the class of business which they do. There will therefore be a smaller supply of shopkeepers in all trades for which the demand has not increased: and those who remain, will be able to charge a higher price than before, without offering any greater conveniences and attractions to their customers. The rise of ground values in the district will thus be an indication of a scarcity of space which, other things being equal, will raise the prices of retail goods; just in the same way as the rise of agricultural rents in any district will indicate a scarcity of land which will raise the marginal expenses of production, and therefore the price of any particular crop.

7. The rent of a house (or other building) is a composite rent, of which one part belongs to the site and the other to the buildings themselves. The relations between these two are rather intricate, and may be deferred to Appendix G. A few words may however be said here as to composite rents in general. At starting there may appear to be some contradiction in the statement that a thing is yielding at the same time two rents: for its rent is in some sense a residual income after deducting the expenses of working it; and there cannot be two residues in regard to the same process of working and the same resulting revenue. But when the thing is composite each of its parts may be capable of being so worked as to yield a surplus of revenue over the expenses of working it. The corresponding rents can always be distinguished analytically, and sometimes they can be separated commercially. (10)

For instance, the rent of a flour-mill worked by water includes the rent of the site on which it is built, and the rent of the water power which it uses. Suppose that it is contemplated to build a mill in a place where there is a limited water power which could be applied equally well on any one of many sites; then the rent of the water power together with the site selected for it is the sum of two rents; which are respectively the equivalent of the differential advantages which possession of the site gives for production of any kind, and which the ownership of the water power gives for working a mill on any of the sites. And these two rents, whether they happen to be owned by the same person or not, can be clearly distinguished and separately estimated both in theory and in practice.

But this cannot be done if there are no other sites on which a mill can be built: and in that case, should the water power and the site belong to different persons, there is nothing but "higgling and bargaining" to settle how much of the excess of the value of the two together over that which the site has for other purposes shall go to the owner of the latter. And even if there were other sites at which the water power could be applied, but not with equal efficiency, there would still be no means of deciding how the owners of the site and the water power should share the excess of the producer's surplus which they got by acting together, over the sum of that which the site would yield for some other purpose, and of that which the water power would yield if applied elsewhere. The mill would probably not be put up till an agreement had been made for the supply of water power for a term of years: but at the end of that term similar difficulties would arise as to the division of the aggregate producer's surplus afforded by the water power and the site with the mill on it.

Difficulties of this kind are continually arising with regard to attempts by partial monopolists, such as railway, gas, water and electrical companies, to raise their charges on the consumer who has adapted his business arrangements to make use of their services, and perhaps laid down at his own expense a costly plant for the purpose. For instance, at Pittsburgh when manufacturers had just put up furnaces to be worked by natural gas instead of coal, the price of the gas was suddenly doubled. And the history of mines affords many instances of difficulties of this kind with neighbouring landowners as to rights of way, etc., and with the owners of neighbouring cottages, railways and docks. (11)


1. See IV, III, section 6.

2. See IV, X-XIII.

3. If we suppose that two farms, which sell in the same market, return severally to equal applications of capital and labour amounts of produce, the first of which exceeds the second by the extra cost of carrying its produce to market, then the rent of the two farms will be the same. (The capital and labour applied to the two farms are here supposed to be reduced to the same money measure, or which comes to the same thing, the two farms are supposed to have equally good access to markets in which to buy.) Again, if we suppose that two mineral springs A and B supplying exactly the same water are capable of being worked each to an unlimited extent at a constant money cost of production; this cost being, sav twopence a bottle at A whatever the amount produced by it, and twopence half penny at B; then those places to which the cost of carriage per bottle from B is a half-penny less than from A, will be the neutral zone for their competition. (If the cost of carriage be proportional to the distance, this neutral zone is a hyperbola of which A and B are foci.) A can undersell B for all places on A's side of it, and vice versa; and each of them will be able to derive a monopoly rent from the sale of its produce within its own area. This is a type of a great many fanciful, but not uninstructive, problems which readily suggest themselves. Compare von Thünen's brilliant researches in Der isolierte Staat.

4. Cases of this kind are of course most frequent in new countries. But they are not very rare in old countries: Saltburn is a conspicuous instance; awhile a more recent instance of exceptional interest is furnished by Letchworth Garden City.

5. Governments have great facilities for carrying out schemes of this kind, especially in the matter of choosing new sites for garrison towns, arsenals, and establishments for the manufacture of the materials of war. In comparisons of the expenses of production by Government and by private firms, the sites of the Government works are often reckoned only at their agricultural value. But such a plan is misleading. A private firm has either to pay heavy annual charges on account of its site, or to run very heavy risks if it tries to make a town for itself. And therefore in order to prove that Government management is for general purposes as efficient and economical as private management, a full charge ought to be made in the balance-sheets of Government factories for the town-value of their sites. In those exceptional branches of production for which a Government can found a manufacturing town without incurring the risks that a private firm would incur in a similar case, that point of advantage may fairly be reckoned as an argument for Governments undertaking those particular businesses.

6. The value of agricultural land is commonly expressed as a certain number of times the current money rental, or in other words a certain "number of years' purchase" of that rental: and other things being equal it will be the higher, the more important these direct gratifications are, as well as the greater the chance that they and the money income afforded by the land will rise. The number of years' purchase would be increased also by an expected fall either in the future normal rate of interest or in the purchasing power of money.

The discounted value of a very distant rise in the value of land is much less than is commonly supposed. For instance, if we take interest at five per cent (and higher rates prevailed during the Middle Ages), 1 invested at compound interest would amount to about 17,000 in 200 years, and 40,000,000,000 in 500 years. Therefore an expenditure by the State of 1 in securing to itself the reversion of a rise in the value of land which came into operation now for the first time would have been a bad investment, unless the value of that rise now exceeded 17,000, if the payment was made 200 years ago; if 500 years ago to 40,000,000,000. This assumes that it would have been possible to invest a sum of this dimension at five per cent: which of course it would not.

7. A few site values have fallen in districts which have been deserted by fashion or trade. But on the other hand annual site values have risen to be many times as great as the ground rents in the case of land which was leased when it had no special situation value, but has since become a chief centre of fashion, or of trade: and all the more if the lease was granted in the first half of the eighteenth century, when gold was scarce and the incomes of all classes of the people, measured in money, were very low The present discounted value of the return of property to the ground landlord a hundred years hence, which will then be worth 1000, is less perhaps than is commonly supposed; though the error is not so great as in the case of anticipations ranging over many hundred years, which were discussed in a recent note: if interest be taken at three per cent. it is about 50; if at five per cent, as was the rule three or four generations ago, it is but 8.

8. See IV, III, section 7.

9. Houses built in flats are often provided with a lift which is run at the expense of the owner of the house, and in such cases, at all events in America, the top floor sometimes lets for a higher rent than any other. If the site is very valuable and the law does not limit the height of his house in the interest of his neighbours, he may build very high: but at last he will reach the margin of building. At last he will find that the extra expenses for foundations and thick walls, and for his lift, together with some resulting depreciation of the lower floors, make him stand to lose more than he gains by adding one more floor; the extra accommodation which it only just answers his purpose to supply is then to be regarded as at the margin of building, even though the gross rent be greater for the higher floors than for the lower.

But in England bylaws restrain an individual from building so high as to deprive his near neighbours of air and light. In the course of time those who build high will be forced to have a good deal of free space about their buildings; and this will render very high buildings unprofitable.

10. It will be borne in mind that if a house is not appropriate to its site, its aggregate rent will not exceed its site rent by the full building rent which the house would command on an appropriate site. Similar limitations apply to most composite rents.

11. The relations between the interests of different classes of workers in the same business and in the same trade, have some affinity to the subject of composite rents. See below VI, VIII, sections 9, 10.

Contents | next chapter | Political Economy Archive