Principles of Economics by Alfred Marshall (1890)

Book Five: General Relations of Demand, Supply and Value

Chapter 7, Prime and Total Cost in Relation to Joint Products.
Cost of Marketing, Insurance against Risk, Cost of Production

1. We may now return to the consideration of prime and supplementary costs, with special reference to the proper distribution of the latter between the joint products of a business.

It often happens that a thing made in one branch of a business is used as a raw material in another, and then the question of the relative profitableness of the two branches can be accurately ascertained only by an elaborate system of book-keeping by double entry; though in practice it is more common to rely on rough estimates made by an almost instinctive guess. Some of the best illustrations of this difficulty are found in agriculture, especially when the same farm combines permanent pasture and arable land worked on long rotation. (1)

Another difficult case is that of the shipowner who has to apportion the expenses of his ship between heavy goods and goods that are bulky but not heavy. He tries, as far as may be, to get a mixed cargo of both kinds; and an important element in the struggle for existence of rival ports is the disadvantage under which those ports lie which are able to offer a cargo only of bulky or only of heavy goods: while a port whose chief exports are weighty but not bulky, attracts to its neighbourhood industries which make for export goods that can be shipped from it at low freights. The Staffordshire Potteries, for example, owe part of their success to the low freights at which their goods are carried by ships sailing from the Mersey with iron and other heavy cargoes.

But there is free competition in the shipowning trade, and it has great powers of variation as regards the size and shape of ships, the routes which they take, and the whole method of trading; and thus in many ways the general principle can be applied, that the relative proportions of the joint products of a business should be so modified that the marginal expenses of production of either product should be equal to its marginal demand price. (2) Or, in other words, the amount of carrying power for each kind of cargo has a constant tendency to move towards equilibrium at a point at which the demand price for that amount in a normal state of trade is just sufficient to cover the expenses of providing it; these expenses being reckoned so as to include not only its (money) prime cost, but also all those general expenses of the business which are in the long run incurred on its account, whether directly or indirectly. (3)

In some branches of manufacture it is customary to make a first approximation to the total cost of producing any class of goods, by assuming that their share of the general expenses of the business is proportionate either to their prime cost, or to the special labour bill that is incurred in making them. Corrections can then be made to meet such cases as those of goods which require either more or less than an average share of space or light, or of the use of expensive machinery; and so on.

2. There are two elements of the general a business, the sharing of which between the different requires some special attention. They are the expense marketing and that of insurance against risk. Some kinds of goods are easily marketed; there is steady demand for them, and it is always safe to make them for stock. But for that very reason competition cuts their price "very fine," and does not allow a large margin above the direct cost of making them. Sometimes the tasks of making and selling them can be rendered almost automatic, so as to require very little to be charged on their account under the heads of the expenses of management and marketing. But in practice it is not uncommon to charge such goods with even less than the small share that would properly fall to them, and to use them as a means of obtaining and maintaining a business connection, that will facilitate the marketing of other classes of goods, the production of which cannot so well be reduced to routine; for as to these there is not so close a competition. Manufacturers, especially in trades connected with furniture and dress, and retailers in almost all trades, frequently find it best to use certain of their goods as a means of advertising others, and to charge the first with less and the second with more than their proportionate share of supplementary expenses. In the former class they put those goods which are so uniform in character and so largely consumed that nearly all purchasers know their value well, in the second those with regard to which purchasers think more of consulting their fancy than of buying at the lowest possible price.

All difficulties of this kind are much increased by that instability of supply price, which results whenever the tendency to increasing return is acting strongly. We have seen that in seeking the normal supply price in such cases we must select as representative a business which is managed with normal ability and so as to get its fair share of the economies, both internal and external, resulting from industrial organization: also that these economies, though they fluctuate with the fortunes of particular businesses, yet increase generally when the aggregate production increases. Now it is obvious that if a manufacturer makes a commodity the increased production of which would put largely increased internal economies within his reach, it is worth his while to sacrifice a great deal in order to push its sales in a new market. If he has a large capital, and the commodity is one in much demand, his expenditure for this purpose may be very great, even exceeding that which he devotes directly to the manufacture: and if, as is likely, he is pushing at the same time several other commodities, nothing more than a very rough guess can be made as to what share of this expenditure should be charged to the sales of each of them in the current year, and what share should be charged to the connection which he is endeavouring to build up for them in the future.

In fact when the production of a commodity conforms to the law of increasing return in such a way as to give a very great advantage to large producers, it is apt to fall almost entirely into the hands of a few large firms; and then the normal marginal supply price cannot be isolated on the plan just referred to, because that plan assumes the existence of a great many competitors with businesses of all sizes, some of them being young and some old, some in the ascending and some in the descending phase. The production of such a commodity really partakes in a great measure of the nature of a monopoly; and its price is likely to be so much influenced by the incidents of the campaign between rival producers, each struggling for an extension of territory, as scarcely to have a true normal level.

Economic progress is constantly offering new facilities for marketing goods at a distance: it not only lowers cost of carriage, but what is often more important, it enables producers and consumers in distant places to get in touch with one another. In spite of this, the advantages of the producer who lives on the spot are very great in many trades; they often enable him to hold his own against competitors at a distance whose methods of production are more economical. He can sell in his own neighbourhood as cheaply as they can, because though the cost of making is greater for his goods than for theirs, he escapes much of the cost which they incur for marketing. But time is on the side of the more economic methods of production; his distant competitors will gradually get a stronger footing in the place, unless he or some new man adopts their improved methods.

It remains to make a closer study of the relation in which insurance against the risks of a business stands to the supply price of any particular commodity produced in it.

3. The manufacturer and the trader commonly insure against injury by fire and loss at sea; and the premiums which they pay are among the general expenses, a share of which has to be added to the prime cost in order to determine the total cost of their goods. But no insurance can be effected against the great majority of business risks.

Even as regards losses by fire and sea, insurance companies have to allow for possible carelessness and fraud; and must therefore, independently of all allowances for their own expenses and profits, charge premiums considerably higher than the true equivalent of the risks run by the buildings or the ships of those who manage their affairs well. The injury done by fire or sea however is likely, if it occurs at all, to be so very great that it is generally worth while to pay this extra charge; partly for special trade reasons, but chiefly because the total utility of increasing wealth increases less than in proportion to its amount. But the greater part of business risks are so inseparably connected with the general management of the business that an insurance company which undertook them would really make itself responsible for the business: and in consequence every firm has to act as its own insurance office with regard to them. The charges to which it is put under this head are part of its general expenses, and a share of them has to be added to the prime cost of each of its products.

But here there are two difficulties. In some cases insurance against risk is apt to be left out of account altogether, in others it is apt to be counted twice over. Thus a large shipowner sometimes declines to insure his ships with the underwriters: and sets aside part at least of the premiums that he might have paid to them, to build up an insurance fund of his own. But he must still, when calculating the total cost of working a ship, add to its prime cost a charge on account of insurance. And he must do the same thing, in some form or other, with regard to those risks against which he could not buy an insurance policy on reasonable terms even if he wanted to. At times, for instance, some of his ships will be idle in port, or will earn only nominal freights: and to make his business remunerative in the long run he must, in some form or other, charge his successful voyages with an insurance premium to make up for his losses on those which are unsuccessful.

In general, however, he does this, not by making a formal entry in his accounts under a separate head, but by the simple plan of taking the average of successful and unsuccessful voyages together; and when that has once been done, insurance against these risks cannot be entered as a separate item in cost of production, without counting the same thing twice over. Having decided to run these risks himself, he is likely to spend a little more than the average of his competitors, in providing against their occurrence; and this extra expense enters in the ordinary way into his balance-sheet. It is really an insurance premium in another form; and therefore he must not count insurance against this part of the risk separately, for then he would be counting it twice over. (4)

When a manufacturer has taken the average of his sales of dress materials over a long time, and bases his future action on the results of his past experience, he has already allowed for the risk that the machinery will be depreciated by new inventions rendering it nearly obsolete, and for the risk that his goods will be depreciated by changes in fashion. If he were to allow separately for insurance against these risks, he would be counting the same thin twice over. (5)

4. Thus, though when we have counted up the average receipts of a risky trade, we must not make a separate full allowance for insurance against risk; though there may be something to be allowed as a charge on account of uncertainty. It is true that an adventurous occupation, such as gold mining, has special attractions for some people: the deterrent force of risks of loss in it is less than the attractive force of chances of great gain, even when the value of the latter estimated on the actuarial principle is much less than that of the former; and as Adam Smith pointed out, a risky trade, in which there is an element of romance, often becomes so overcrowded that the average earnings in it are lower than if there were no risks to be run. (6) But in the large majority of cases the influence of risk is in the opposite direction; a railway stock that is certain to pay four per cent. will sell for a higher price than one which is equally likely to pay one or seven per cent. or any intermediate amount.

Every trade then has its own peculiarities, but in most cases the evils of uncertainty count for something, though not very much: in some cases a slightly higher average price is required to induce a given outlay, if that average is the mean of widely divergent and uncertain results, than if the adventurer may reckon confidently on a return that differs but little from that average. To the average price therefore we must add a recompense for uncertainty, if that is unusually great; though if we added insurance against risk we should be counting the greater part of that twice over. (7)

5. This discussion of the risks of trade has again brought before us the fact that the value of a thing, though it tends to equal its normal (money) cost of production, does not coincide with it at any particular time, save by accident. Carey, observing this, suggested that we should speak of value in relation to (money) cost of reproduction instead of in relation to cost of production.

The suggestion has, however, no significance so far as normal values are concerned. For normal cost of production and normal cost of reproduction are convertible terms; and no real change is made by saying that the normal value of a thing tends to equal its normal (money) cost of reproduction instead of its normal (money) cost of production. The former phrase is less simple than the latter, but means the same thing.

And no valid argument for the change can be founded on the fact, which may be readily admitted, that there are some few cases in which the market value of a thing is nearer its cost of reproduction than the cost that was actually incurred in producing that particular thing. The present price of an iron ship for instance, made before the great recent improvements in the manufacture of iron, might diverge less from the cost of reproducing it, that is of producing another just like it by modern methods, than from that which was actually incurred in producing it. But the price of the old ship would be less than the cost of reproduction of the ship, because the art of designing ships has improved as fast as that of manufacturing iron; and moreover steel has displaced iron as the material of shipbuilding. It may still be urged that the price of the ship is equal to that of producing a ship, which would be equally serviceable, on a modern plan and by modern methods. But that would not be the same thing as saying that the value of the ship is equal to its cost of reproduction; and, as a matter of fact, when, as often happens, an unexpected scarcity of ships causes freights to increase very rapidly, those who are anxious to reap the harvest of profitable trade, will pay for a ship in sailing order a price much above that for which a shipbuilding firm would contract to produce another equally good and deliver it some time hence. Cost of reproduction influence on value, save when purchasers can wait for the production of new supplies.

Again, there is no connection between cost of reproduction and price in the cases of food in a beleaguered city, of quinine the supply of which has run short in a fever-stricken island, of a picture by Raphael, of a book that nobody cares to read, of an armour-clad ship of obsolete pattern, of fish when the market is glutted, of fish when the market is nearly empty, of a cracked bell, of a dress material that has gone out of fashion, or of a house in a deserted mining village.


The reader, unless already experienced in economic analysis, is recommended to omit the next seven chapters, and pass at once to Chapter XV, which contains a brief summary of this Book. It is true that the four chapters on marginal costs in relation to values, and especially Chapters VIII and IX, bear upon some difficulties which are latent in the phrase "the net product of labour"; and that this phrase is used in Book VI. But the broad explanation of it given there will suffice provisionally for most purposes; and the intricacies connected with it may be best appreciated at a somewhat advanced stage of economic studies.


1. There is scope for applications of mathematical or semi-mathematical analyses, such as are indicated in the last chapter, to some of the chief practical difficulties of book-keeping by double entry in different trades.

2. Compare VI, section 4.

3. Of course this does not apply to railway rates. For a railway company having little elasticity as to its methods of working, and often not much competition from outside, has no inducement to endeavour to adjust the charges which it makes for different kinds of traffic to their cost to itself. In fact though it may ascertain the prime cost in each case easily enough, it cannot determine accUrately what are the relative total costs of fast and slow traffic, of short and long distance traffic, of light and heavy traffic; nor again of extra traffic when its lines and its trains are crowded and when they are nearly empty.

4. Again, certain insurance companies in America take risks against fire in factories at very much less than the ordinary rates, on condition that some prescribed precautions are taken, such as providing automatic sprinklers and making the walls and floor solid. The expense incurred in these arrangements is really an insurance premium; and care must be taken not to count it twice over. A factory which undertakes its own risks against fire will have to add to the prime cost of its goods an allowance for insurance at a lower rate, if it is arranged on this plan, than if built in the ordinary way.

5. Again, when a farmer has calculated the expenses of raising any particular crop with reference to an average year, he must not count in addition insurance against the risk that the season may be bad, and the crop a failure: for in taking an average year, he has already set of the chances of exceptionally good and bad seasons against one another. When the earnings of a ferryman have been calculated on the average of a year, allowance has already been made for the risk that he may sometimes have to cross the stream with an empty boat.

6. Wealth of Nations, Book I, ch. x.

7. The evils resulting from the uncertainty involved in great business risks are well shown by von Thünen (Der isolierte Staat, II, I, p. 82).

Contents | next chapter | Political Economy Archive