Front Page Titles (by Subject) BOOK V, CHAPTER XIII: THEORY OF CHANGES OF NORMAL DEMAND AND SUPPLY IN RELATION TO THE DOCTRINE OF MAXIMUM SATISFACTION. - Principles of Economics (8th ed.)
The Online Library of Liberty
A project of Liberty Fund, Inc.
BOOK V, CHAPTER XIII: THEORY OF CHANGES OF NORMAL DEMAND AND SUPPLY IN RELATION TO THE DOCTRINE OF MAXIMUM SATISFACTION. - Alfred Marshall, Principles of Economics (8th ed.) 
Principles of Economics (London: Macmillan and Co. 8th ed. 1920).
About Liberty Fund:
Liberty Fund, Inc. is a private, educational foundation established to encourage the study of the ideal of a society of free and responsible individuals.
The text is in the public domain.
Fair use statement:
This material is put online to further the educational goals of Liberty Fund, Inc. Unless otherwise stated in the Copyright Information section above, this material may be used freely for educational and academic purposes. It may not be used in any way for profit.
BOOK V, CHAPTER XIII
THEORY OF CHANGES OF NORMAL DEMAND AND SUPPLY IN RELATION TO THE DOCTRINE OF MAXIMUM SATISFACTION.
§ 1. In earlier chapters of this Book, and especially in chapter XII., we have considered gradual changes in the adjustment of demand and supply. But any great and lasting change in fashion; any substantive new invention; any diminution of population by war or pestilence; or the development or dwindling away of a source of supply of the commodity in question, or of a raw material used in it, or of another commodity which is a rival and possible substitute for it:—such a change as any of these may cause the prices set against any given annual (or daily) consumption and production of the commodity to cease to be its normal demand and supply prices for that volume of consumption and production; or, in other words, they may render it necessary to make out a new demand schedule or a new supply schedule, or both of them. We proceed to study the problems thus suggested.
An increase of normal demand for a commodity involves an increase in the price at which each several amount can find purchasers; or, which is the same thing, an increase of the quantity which can find purchasers at any price. This increase of demand may be caused by the commodity's coming more into fashion, by the opening out of a new use for it or of new markets for it, by the permanent falling off in the supply of some commodity for which it can be used as a substitute, by a permanent increase in the wealth and general purchasing power of the community, and so on. Changes in the opposite direction will cause a falling off in demand and a sinking of the demand prices. Similarly an increase of normal supply means an increase of the amounts that can be supplied at each several price, and a diminution of the price at which each separate amount can be supplied115 . This change may be caused by the opening up of a new source of supply, whether by improved means of transport or in any other way, by an advance in the arts of production, such as the invention of a new process or of new machinery, or again, by the granting of a bounty on production. Conversely, a diminution of normal supply (or a raising of the supply schedule) may be caused by the closing up of a new source of supply or by the imposition of a tax.
§ 2. We have, then, to regard the effects of an increase of normal demand from three points of view, according as the commodity in question obeys the law of constant or of diminishing or of increasing return: that is, its supply price is practically constant for all amounts, or increases or diminishes with an increase in the amount produced.
In the first case an increase of demand simply increases the amount produced without altering its price; for the normal price of a commodity which obeys the law of constant return is determined absolutely by its expenses of production: demand has no influence in the matter beyond this, that the thing will not be produced at all unless there is some demand for it at this fixed price.
If the commodity obeys the law of diminishing return an increase of demand for it raises its price and causes more of it to be produced; but not so much more as if it obeyed the law of constant return.
On the other hand, if the commodity obeys the law of increasing return, an increase of demand causes much more of it to be produced,—more than if the commodity obeyed the law of constant return,—and at the same time lowers its price. If, for instance, a thousand things of a certain kind have been produced and sold weekly at a price of 10s., while the supply price for two thousand weekly would be only 9s., a small rate of increase in normal demand may gradually cause this to become the normal price; since we are considering periods long enough for the full normal action of the causes that determine supply to work itself out. The converse holds in each case should normal demand fall off instead of increasing116 .
The argument of this section has been thought by some writers to lend support to the claim that a Protective duty on manufactured imports in general increases the home market for those imports; and, by calling into play the Law of Increasing Return, ultimately lowers their price to the home consumer. Such a result may indeed ultimately be reached by a wisely chosen system of "Protection to nascent industries" in a new country; where manufactures, like young children, have a power of rapid growth. But even there the policy is apt to be wrenched from its proper uses, to the enrichment of particular interests: for those industries which can send the greatest number of votes to the poll, are those which are already on so large a scale, that a further increase would bring very few new economies. And of course the industries in a country so long familiar with machinery as England is, have generally passed the stage at which they can derive much real help from such Protection: while Protection to any one industry nearly always tends to narrow the markets, especially the foreign markets, for other industries. These few remarks show that the question is complex: they do not pretend to reach further than that.
§ 3. We have seen that an increase in normal demand, while leading in every case to an increased production, will in some cases raise and in others lower prices. But now we are to see that increased facilities for supply (causing the supply schedule to be lowered) will always lower the normal price at the same time that it leads to an increase in the amount produced. For so long as the normal demand remains unchanged an increased supply can be sold only at a diminished price; but the fall of price consequent on a given increase of supply will be much greater in some cases than in others. It will be small if the commodity obeys the law of diminishing return; because then the difficulties attendant on an increased production will tend to counteract the new facilities of supply. On the other hand, if the commodity obeys the law of increasing return, the increased production will bring with it increased facilities, which will co-operate with those arising from the change in the general conditions of supply; and the two together will enable a great increase in production and consequent fall in price to be attained before the fall of the supply price is overtaken by the fall of the demand price. If it happens that the demand is very elastic, then a small increase in the facilities of normal supply, such as a new invention, a new application of machinery, the opening up of new and cheaper sources of supply, the taking off a tax or granting a bounty, may cause an enormous increase of production and fall of price117 .
If we take account of the circumstances of composite and joint supply and demand discussed in chapter VI., we have suggested to us an almost endless variety of problems which can be worked out by the methods adopted in these two chapters.
§ 4. We may now consider the effects which a change in the conditions of supply may exert on consumers' surplus or rent. For brevity of language a tax may be taken as representative of those changes which may cause a general increase, and a bounty as representative of those which may cause a general diminution in the normal supply price for each several amount of the commodity.
Firstly, if the commodity is one, the production of which obeys the law of constant return, so that the supply price is the same for all amounts of the commodity, consumers' surplus will be diminished by more than the increased payments to the producer; and therefore, in the special case of a tax, by more than the gross receipts of the State. For on that part of the consumption of the commodity, which is maintained, the consumer loses what the State receives: and on that part of the consumption which is destroyed by the rise in price, the consumers' surplus is destroyed; and of course there is no payment for it to the producer or to the State118 . Conversely, the gain of consumers' surplus caused by a bounty on a commodity that obeys the law of constant return, is less than the bounty itself. For on that part of the consumption which existed before the bounty, consumers' surplus is increased by just the amount of the bounty; while on the new consumption that is caused by the bounty, the gain of the consumers' surplus is less than the bounty119 .
If however the commodity obeys the law of diminishing return; a tax by raising its price, and diminishing its consumption, will lower its expenses of production other than the tax: and the result will be to raise the supply price by something less than the full amount of the tax. In this case the gross receipts from the tax may be greater than the resulting loss of consumers' surplus, and they will be greater if the law of diminishing return acts so sharply that a small diminution of consumption causes a great falling-off in the expenses of production other than the tax120 .
On the other hand, a bounty on a commodity which obeys the law of diminishing return will lead to increased production, and will extend the margin of cultivation to places and conditions in which the expenses of production, exclusive of the bounty, are greater than before. Thus it will lower the price to the consumer and increase consumers' surplus less than if it were given for the production of a commodity which obeyed the law of constant return. In that case the increase of consumers' surplus was seen to be less than the direct cost of the bounty to the State; and therefore in this case it is much less121 .
By similar reasoning it may be shown that a tax on a commodity which obeys the law of increasing return is more injurious to the consumer than if levied on one which obeys the law of constant return. For it lessens the demand and therefore the output. It thus probably increases the expenses of manufacture somewhat: sends up the price by more than the amount of the tax; and finally diminishes consumers' surplus by much more than the total payments which it brings in to the exchequer122 . On the other hand, a bounty on such a commodity causes so great a fall in its price to the consumer, that the consequent increase of consumers' surplus may exceed the total payments made by the State to the producers; and certainly will do so in case the law of increasing return acts at all sharply123 .
These results are suggestive of some principles of taxation which require careful attention in any study of financial policy; when it will be necessary to take account of the expenses of collecting a tax and of administering a bounty, and of the many indirect effects, some economic and some moral, which a tax or a bounty is likely to produce. But these partial results are well adapted for our immediate purpose of examining a little more closely than we have done hitherto the general doctrine that a position of (stable) equilibrium of demand and supply is a position also of maximum satisfaction: and there is one abstract and trenchant form of that doctrine which has had much vogue, especially since the time of Bastiat's Economic Harmonics, and which falls within the narrow range of the present discussion.
§ 5. There is indeed one interpretation of the doctrine according to which every position of equilibrium of demand and supply may fairly be regarded as a position of maximum satisfaction124 . For it is true that so long as the demand price is in excess of the supply price, exchanges can be effected at prices which give a surplus of satisfaction to buyer or to seller or to both. The marginal utility of what he receives is greater than that of what he gives up, to at least one of the two parties; while the other, if he does not gain by the exchange, yet does not lose by it. So far then every step in the exchange increases the aggregate satisfaction of the two parties. But when equilibrium has been reached, demand price being now equal to supply price, there is no room for any such surplus: the marginal utility of what each receives no longer exceeds that of what he gives up in exchange: and when the production increases beyond the equilibrium amount, the demand price being now less than the supply price, no terms can be arranged which will be acceptable to the buyer, and will not involve a loss to the seller.
It is true then that a position of equilibrium of demand and supply is a position of maximum satisfaction in this limited sense, that the aggregate satisfaction of the two parties concerned increases until that position is reached; and that any production beyond the equilibrium amount could not be permanently maintained so long as buyers and sellers acted freely as individuals, each in his own interest.
But occasionally it is stated, and very often it is implied, that a position of equilibrium of demand and supply is one of maximum aggregate satisfaction in the full sense of the term: that is, that an increase of production beyond the equilibrium level would directly (i.e. independently of the difficulties of arranging for it, and of any indirect evils it might cause) diminish the aggregate satisfaction of both parties. The doctrine so interpreted is not universally true.
In the first place it assumes that all differences in wealth between the different parties concerned may be neglected, and that the satisfaction which is rated at a shilling by any one of them, may be taken as equal to one that is rated at a shilling by any other. Now it is obvious that, if the producers were as a class very much poorer than the consumers, the aggregate satisfaction might be increased by a stinting of supply when it would cause a great rise in demand price (i.e. when the demand is inelastic); and that if the consumers were as a class much poorer than the producers, the aggregate satisfaction might be increased by extending the production beyond the equilibrium amount and selling the commodity at a loss125 .
This point however may well be left for future consideration. It is in fact only a special case of the broad proposition that the aggregate satisfaction can primâ facie be increased by the distribution, whether voluntarily or compulsorily, of some of the property of the rich among the poor; and it is reasonable that the bearings of this proposition should be set aside during the first stages of an inquiry into existing economic conditions. This assumption therefore may be properly made, provided only it is not allowed to slip out of sight.
But in the second place the doctrine of maximum satisfaction assumes that every fall in the price which producers receive for the commodity, involves a corresponding loss to them; and this is not true of a fall in price which results from improvements in industrial organization. When a commodity obeys the law of increasing return, an increase in its production beyond equilibrium point may cause the supply price to fall much; and though the demand price for the increased amount may be reduced even more, so that the production would result in some loss to the producers, yet this loss may be very much less than that money value of the gain to purchasers which is represented by the increase of consumers' surplus.
In the case then of commodities with regard to which the law of increasing return acts at all sharply, or in other words, for which the normal supply price diminishes rapidly as the amount produced increases, the direct expense of a bounty sufficient to call forth a greatly increased supply at a much lower price, would be much less than the consequent increase of consumers' surplus. And if a general agreement could be obtained among consumers, terms might be arranged which would make such action amply remunerative to the producers, at the same time that they left a large balance of advantage to the consumers126 .
§ 6. One simple plan would be the levying of a tax by the community on their own incomes, or on the production of goods which obey the law of diminishing return, and devoting the tax to a bounty on the production of those goods with regard to which the law of increasing return acts sharply. But before deciding on such a course they would have to take account of considerations, which are not within the scope of the general theory now before us, but are yet of great practical importance. They would have to reckon up the direct and indirect costs of collecting a tax and administering a bounty; the difficulty of securing that the burdens of the tax and the benefits of the bounty were equitably distributed; the openings for fraud and corruption; and the danger that in the trade which had got a bounty and in other trades which hoped to get one, people would divert their energies from managing their own businesses to managing those persons who control the bounties.
Besides these semi-ethical questions there will arise others of a strictly economic nature, relating to the effects which any particular tax or bounty may exert on the interests of landlords, urban or agricultural, who own land adapted for the production of the commodity in question. These are questions which must not be overlooked; but they differ so much in their detail that they cannot fitly be discussed here127 .
§ 7. Enough has been said to indicate the character of the second great limitation which has to be introduced into the doctrine that the maximum satisfaction is generally to be attained by encouraging each individual to spend his own resources in that way which suits him best. It is clear that if he spends his income in such a way as to increase the demand for the services of the poor and to increase their incomes, he adds something more to the total happiness than if he adds an equal amount to the incomes of the rich, because the happiness which an additional shilling brings to a poor man is much greater than that which it brings to a rich one; and that he does good by buying things the production of which raises, in preference to things the production of which lowers the character of those who make them128 . But further, even if we assume that a shilling's worth of happiness is of equal importance to whomsoever it comes, and that every shilling's worth of consumers' surplus is of equal importance from whatever commodity it is derived, we have to admit that the manner in which a person spends his income is a matter of direct economic concern to the community. For in so far as he spends it on things which obey the law of diminishing return, he makes those things more difficult to be obtained by his neighbours, and thus lowers the real purchasing power of their incomes; while in so far as he spends it on things which obey the law of increasing return, he makes those things more easy of attainment to others, and thus increases the real purchasing power of their incomes.
Again, it is commonly argued that an equal ad valorem tax levied on all economic commodities (material and immaterial), or which is the same thing a tax on expenditure, is primâ facie the best tax; because it does not divert the expenditure of individuals out of its natural channels: we have now seen that this argument is invalid. But ignoring for the time the fact that the direct economic effect of a tax or a bounty never constitutes the whole, and very often not even the chief part of the considerations which have to be weighed before deciding to adopt it, we have found:—firstly, that a tax on expenditure generally causes a greater destruction of consumers' surplus than one levied exclusively on commodities as to which there is but little room for the economies of production on a large scale, and which obey the law of diminishing return; and secondly, that it might even be for the advantage of the community that the government should levy taxes on commodities which obey the law of diminishing return, and devote part of the proceeds to bounties on commodities which obey the law of increasing return.
These conclusions, it will be observed, do not by themselves afford a valid ground for government interference. But they show that much remains to be done, by a careful collection of the statistics of demand and supply, and a scientific interpretation of their results, in order to discover what are the limits of the work that society can with advantage do towards turning the economic actions of individuals into those channels in which they will add the most to the sum total of happiness129 .
[115.]A rise or fall of the demand or supply prices involves of course a rise or fall of the demand or supply curve.
[116.]Diagrams are of especial aid in enabling us to comprehend clearly the problems of this chapter.
[117.]All this can be most clearly seen by the aid of diagrams, and indeed there are some parts of the problem which cannot be satisfactorily treated without their aid. The three figures 27, 28, 29 represent the three cases of constant and diminishing and increasing returns, respectively.
[118.]This is most clearly seen by aid of a diagram. SS', the old constant return supply curve, cuts DD' the demand curve in A: DSA is the consumers' surplus. Afterwards a tax Ss being imposed the new equilibrium is found at a, and consumers' surplus is Dsa. The gross tax is only the rectangle sSKa, that is, a tax at the rate of Ss on an amount sa of the commodity. And this falls short of the loss of consumers' surplus by the area aKA. The net loss aKA is small or great, other things being equal, as aA is or is not inclined steeply. Thus it is smallest for those commodities the demand for which is most inelastic, that is, for necessaries. If therefore a given aggregate taxation has to be levied ruthlessly from any class it will cause less loss of consumers' surplus if levied on necessaries than if levied on comforts; though of course the consumption of luxuries and in a less degree of comforts indicates ability to bear taxation.
[119.]If we now regard ss' as the old supply curve which is lowered to the position SS' by the granting of a bounty, we find the gain of consumers' surplus to be sSAa. But the bounty paid is Ss on an amount SA, which is represented by the rectangle sSAL: and this exceeds the gain of consumers' surplus by the area aLA.
[120.]Let the old supply curve be SS' fig. 31, and let the imposition of a tax raise it to ss'; let A and a be the old and new positions of equilibrium, and let straight lines be drawn through them parallel to Ox and Oy, as in the figure. Then the tax being levied, as shown by the figure, at the rate of aE on each unit; and Oh, that is, CK units, being produced in the new position of equilibrium, the gross receipts of the tax will be cFEa, and the loss of consumers' surplus will be cCAa; that is, the gross receipts from the tax will be greater or less than the loss of consumers' surplus as CFEK is greater or less than aKA; and in the figure as it stands it is much greater. If SS' had been so drawn as to indicate only very slight action of the law of diminishing return, that is, if it had been nearly horizontal in the neighbourhood of A, then EK would have been very small; and CFEK would have become less than aKA.
[121.]To illustrate this case we may take ss' in fig. 31 to be the position of the supply curve before the granting of the bounty, and SS' to be its position afterwards. Thus a was the old equilibrium point, and A is the point to which the equilibrium moves when the bounty is awarded. The increase of consumers' surplus is only cCAa, while the payments made by the State under the bounty are, as shown by the figure, at the rate of AT on each unit of the commodity; and as in the new position of equilibrium there are produced OH, that is, CA units, they amount altogether to RCAT which includes and is necessarily greater than the increase of consumers' surplus.
[122.]Thus taking SS' in fig. 32 to be the old position of the supply curve, and ss' its position after the tax, A to be the old and a the new positions of equilibrium, we have, as in the case of fig. 31, the total tax represented by cFEa, and the loss of consumers' surplus by cCAa; the former being always less than the latter.
[123.]To illustrate this case we may take ss' in fig. 32 to be the position of the supply curve before the granting of the bounty, and SS' to be its position afterwards. Then, as in the case of fig. 31, the increase of consumers' surplus is represented by cCAa, while the direct payments made by the State under the bounty are represented by RCAT. As the figure is drawn, the former is much larger than the latter. But it is true that if we had drawn ss' so as to indicate a very slight action of the law of increasing return, that is, if it had been very nearly horizontal in the neighbourhood of a, the bounty would have increased relatively to the gain of consumers' surplus; and the case would have differed but little from that of a bounty on a commodity which obeys the law of constant return, represented in fig. 30.
[124.]Compare V. I. 1. Unstable equilibrium may now be left out of account.
[125.]In this illustration one of the two things exchanged is general purchasing power; but of course the argument would hold if a poor population of pearl divers were dependent for food on a rich population who took pearls in exchange.
[126.]Though not of great practical importance, the case of multiple positions of (stable) equilibrium offers a good illustration of the error involved in the doctrine of maximum satisfaction when stated as a universal truth. For the position in which a small amount is produced and is sold at a high price would be the first to be reached, and when reached would be regarded according to that doctrine as that which gave the absolute maximum of aggregate satisfaction. But another position of equilibrium corresponding to a larger production and a lower price would be equally satisfactory to the producers, and would be much more satisfactory to the consumers; the excess of consumers' surplus in the second case over the first would represent the increase in aggregate satisfaction.
[127.]The incidence of a tax on agricultural produce will be discussed later on by the aid of diagrams similar to those used to represent the fertility of land (see IV. III.). Landlords' rent absorbs a share of the aggregate selling price of almost all commodities: but it is most prominent in the case of those which obey the law of diminishing return; and an assumption of no extreme violence will enable fig. 33 (a reproduction of 31) to represent roughly the leading features of the problem.
[128.]Compare III. VI.
[129.]It is remarkable that Malthus, Political Economy, ch. III. § 9, argued that, though the difficulties thrown in the way of importing foreign corn during the great war turned capital from the more profitable employment of manufacture to the less profitable employment of agriculture, yet if we take account of the consequent increase of agricultural rent, we may conclude that the new channel may have been one of "higher national, though not higher individual profits." In this no doubt he was right; but he overlooked the far more important injury inflicted on the public by the consequent rise in the price of corn, and the consequent destruction of consumers' surplus. Senior takes account of the interests of the consumer in his study of the different effects of increased demand on the one hand and of taxation on the other in the case of agricultural and manufactured produce (Political Economy, pp. 118-123). Advocates of Protection in countries which export raw produce have made use of arguments tending in the same direction as those given in this Chapter; and similar arguments are now used, especially in America (as for instance by Mr H. C. Adams), in support of the active participation of the State in industries which conform to the law of increasing return. The graphic method has been applied, in a manner somewhat similar to that adopted in the present Chapter, by Dupuit in 1844; and, independently, by Fleeming Jenkin (Edinburgh Philosophical Transactions) in 1871.