The Economics of Welfare
§ 1. IN Chapter IX. § 3 attention was drawn to a type of divergence between private and social net product additional to the two types that were studied in that chapter. This type of divergence arises when a part of the effect of employing a unit of resources in any occupation consists of something, which, instead of coming in the first instance to the person who invests the unit, comes instead, in the first instance, to other persons engaged in the occupation. To simplify the study of these divergences I shall imagine that there exists an archetypal industry, in which the values of the marginal private net product and of the marginal social net product of investment are both equal to one another and also stand at a sort of central level representative of industries in general.*11 In any actual industry conducted under conditions of simple competition—that is to say, conditions such that each seller produces as much as he can at the ruling market price, and does not restrict his output in the hope of causing that price to rise—investment and output must be carried to a point at which the value of the marginal private net product of investment there conforms to the central value. It follows that the value of the marginal social net product of investment in the industry can only diverge from this central value if and in so far as it diverges from the value of the marginal private net product. The present chapter is concerned exclusively with conditions of simple competition.
§ 2. Let us place ourselves in imagination in a country where the flow of resources coming annually into being has to be distributed regularly among a variety of occupations. It is assumed that, when any given quantity of resources is devoted to a given occupation, the concrete form assumed by these resources—their distribution, for example, into many or few individual firms and so forth—is the most economical form available (from the standpoint of the period of time relevant to our problem) for that quantity of new resources; and that, when a slightly greater quantity is devoted to the occupation, the concrete form assumed is the most economical form available (from the same standpoint) for that quantity. When this assumption is made, it is plain that, if one unit is added to the resources of any given sort*12 that normally flow into any one occupation, that unit will yield the same net product as each of the other units in the flow. All the units are interchangeable in this sense. But, none the less, the presence of an extra unit may alter the output of the other units, in such wise that the addition made to aggregate output is either more or less than proportionate to the difference made to the quantity of resources invested. In so far as the other units belong to the investor of the given unit and the difference made to the output of the other units comes in the first instance to him, it enters into the private net product as well as into the social net product of the extra unit. But, in so far as the other units belong to people other than the investor of the given unit, the difference made to the output of these units enters into the social net product, but not into the private net product of the given unit. The two sorts of marginal net product in the particular industry, and hence their values, therefore, differ. Since then investment under competitive conditions is carried to the point at which the value of the marginal private net product of the resources placed there is equal to the central value, the value of the social net product in that industry must diverge from the central value; and the national dividend is not maximised.*13
§ 3. This statement has now to be brought into connection with the familiar economic concept of increasing, constant and diminishing returns, or, as some prefer to say, decreasing, constant and increasing costs. As a preliminary it will be convenient to provide ourselves with an appropriate terminology. The expressions cited above, used in the present connection,*14 are designed to describe certain relations between the output of a commodity and the expenses, measured in money, that are incurred in producing it. Diminishing and increasing returns mean diminishing and increasing yields of commodity per unit of money expenses as the output of the commodity increases; increasing and decreasing costs mean increasing and decreasing money expenses per unit of the commodity as the output increases. The two sets of terms are thus, so to speak, reciprocals of one another. Both alike, however, are open to an objection. It is not clear, on the face of things, whether the returns per unit of money expenses, or expenses per unit of returns, to which they refer, are average or marginal returns or expenses; and, when diminishing returns (or increasing costs) hold for some amounts of output and increasing returns (or decreasing costs) for other amounts, there must, it would seem, be certain amounts in respect of which marginal returns are increasing (or marginal expenses decreasing) while average returns are decreasing (or average expenses increasing), and vice versa. It is, therefore, best, as I think, to surrender both the above forms of expression, and to distinguish industries according as they conform to conditions of increasing, constant or decreasing supply price. It will be shown in Appendix III. that, in competitive industries, the supply price of any quantity of output is equal both to the marginal expenses and to the average expenses of what I call the equilibrium firm engaged in the industry; and that it is a concept free from ambiguity. That discussion need not be repeated here. For the reader not interested in refinements of analysis it is sufficient to say that my laws of increasing, constant and decreasing supply price correspond, for practical purposes, to what are ordinarily known as the laws of diminishing, constant and increasing returns, or increasing, constant and decreasing costs.*15 It is, of course, with long-period or "normal," not with any form of short-period supply price, that we here have to do.
§ 4. The relations which these laws express between variations in supply price and variations in output are not necessarily the relations which do subsist between these things in history, but the relations which would subsist subject to the condition other things being equal. In real life, with the general advance of knowledge, new methods of production are being continually introduced and new technical appliances invented. Some of these changes are due to factors which would operate even though the scale of output of the industry remained constant. Others are the result of the changes in the scale of output, being called out in response to changes in demand. Of course, in practice, it may often be impossible to say whether a particular invention in, say, the process of steel manufacture is or is not due to changes in the scale of output. Logically, however, the distinction is quite clear. For the present purpose changes not due to changes in the scale of output are definitely ruled out of consideration. Therefore an industry may display continually falling supply price through a long series of years, and yet may not be operating under conditions of decreasing supply price*16 as understood here. In like manner, when, for example, the coal seams of a country are being gradually worked out, an industry may display continually rising supply and yet may not be operating under conditions of increasing supply price.*17 An industry is said to conform to increasing, constant or decreasing supply price, when, apart from changes in technique or other inventions not due to changes in the scale of output, increase of output would be associated, as the case may be, with increasing, constant or decreasing supply prices.
§ 5. Attention must next be called to a distinction, which, for the present purpose, is fundamental. When we speak without qualification of laws of increasing, constant or decreasing supply price, we have in view the relation between variations in the output of a commodity and variations in the supply price per unit from the standpoint of the industry producing the commodity. These variations are not always or necessarily the same as the variations in the supply price per unit of the commodity from the standpoint of the community. Consider an industry which purchases from others factors of production only. When, with a given measure of increase in the output of anything, the money expenses per unit of output incurred by the equilibrium firm increase, because for each unit of output it has to buy, at the same price as before, greater quantities of one or another factor, the two sorts of variation are identical. But, when the expenses per unit increase because it has to pay for the factors of production which it employs a higher money price, the extra payment that it makes is offset by an equal and opposite extra payment which the owners of the several factors of production receive. From the point of view of the community as a whole no extra expense per unit of output is incurred. In like manner, when the expenses per unit to the equilibrium firm in an industry decrease because it pays for the factors it employs a lower money price, from the point of view of the community as a whole there is no saving of expenses per unit of output. This matter is examined in more detail in Appendix III. We say, then, that an industry conforms to the law of increasing, constant or decreasing supply price simpliciter, when it so conforms from the standpoint of the industry under review, i.e. when the variations in supply price, calculated without any allowance for the transfer elements distinguished above, that are associated with increases of output, are positive, nil or negative respectively. We say that it conforms to the law of increasing, constant or decreasing supply price from the standpoint of the community, when these variations, corrected so as to eliminate transfer elements, are positive, nil or negative.*18
§ 6. Conditions of decreasing supply price from the standpoint of the community are clearly possible in a material as well as in a formal sense. For, when the scale of an industry increases, this change often leads to changes in the internal structure and methods of working of the firms engaged in it, or in the proportions in which the several factors of production are employed, of a kind which would lower the cost of production per unit, even though the prices per unit of all the factors of production employed were unaltered. Thus many writers have called attention to the fact that, when an industry is on a small scale, the firms belonging to it all engage in producing a number of different types or varieties of their commodity. They are, more or less, firms of all work. There is not a sufficiently wide or assured market to allow of close specialisation. As, however, the general demand grows, it becomes more and more worth while for firms to specialise on particular types. Thus Sir Sydney Chapman has observed that the relatively large scale of the cotton industry in England is associated, not only with specialisation between the processes of spinning and weaving, but with further specialisation between firms spinning fine counts and those spinning coarse counts. In contrast to this: "The range of work undertaken by the typical factory in Germany is far greater than that undertaken by the typical factory in England. Hence naturally the skill of the operatives is less in Germany; more time is wasted and factory organisation is less perfect."*19 The increased specialisation of its component firms made possible by an enlargement in an industry as a whole often involves a large reduction in costs. This reduction might, so far as pure theory goes, be accompanied by no change, or even by a decrease, in the size of the typical firm. In practice it is likely to be accompanied by some increase in this size. Thus Marshall writes: "An increase in the aggregate volume of the production of anything will generally increase the size, and, therefore, the internal economies possessed by (such) a representative firm."*20 This, however, is a secondary consideration. The essential point is that an increase in the scale to which an industry is producing frequently alters—in general diminishes—the average (and marginal) costs of the equilibrium firm contained in it, whether or not it also alters its size. There is, then, no difficulty in seeing that the law of decreasing supply price from the standpoint of the community is not merely formally possible, but is likely to be followed in practice by many manufacturing industries.
§ 7. Conditions of increasing supply price from the standpoint of the community are, however, in different case. We consider, as before, an industry which purchases for its own use factors of production, but no other ingredients. At the worst the equilibrium firm is in a position to maintain its original scale of output: and it will, in fact, do this unless some other scale now makes possible a lower cost per unit of output. Hence, in order that the law of increasing supply price from the standpoint of the community may prevail, conditions must be such that a mere increase in the output of the industry as a whole would cause the average expenses of the equilibrium firm—which are equal, as is shown in Appendix III., to the supply price of the industry—to increase, even though that firm continued to produce exactly the same output as before and paid exactly the same prices for the factors of production employed by it. It is possible, no doubt, that diseconomies of this character—injuries to the efficiency of old firms brought about by the mere existence of new ones—may occur. But their occurrence is, on the face of things, highly improbable; and that they should occur in sufficient measure to outweigh the factors, described in the preceding section, that make for decreasing supply price is more improbable still. In general, therefore, we may conclude that an industry, whose purchases embrace only factors of production, cannot conform to the law of increasing supply price from the standpoint of the community, i.e. when variations in supply price are expressed in such a way that transfer elements are eliminated.
§ 8. When an industry, besides purchasing ultimate factors of production, purchases also materials, machinery and so on, the matter is less simple. The price variations in these things, if such occur, do not now necessarily represent only transfer expenditure. They do not do so, for example, when an increase in the size of the cotton industry enables textile machinery to be produced with more help from specialisation and standardisation, and, therefore, more cheaply. The fall in price of textile machinery brought about in this way when the cotton industry expands is relevant to the law of supply price in that industry, not merely from the standpoint of the industry, but also from that of the community. Since, however, what is sold to one industry, other than factors of production, must be the product of another industry, it follows from the discussion of the preceding section that an increase in the output (and, therefore, in the demand for materials and machinery) in a given industry cannot involve an increase in the price of the things it buys, except by causing an increase in the price of the factors of production that make them. Therefore, in this complex case, no less than in the simpler one, conditions of increasing supply price from the standpoint of the community are excluded. From the cosmopolitan point of view they are excluded absolutely. From the point of view of a particular country purchasing materials from abroad they may, however, be present. For, though, if the price of imported materials rises as a consequence of an increase in the scale of the given industry, this can only be because a transfer is made to the owners of the factors that help to make them, these owners belong to other countries, and, therefore, so far as the one country is concerned, the transfer does not cancel itself.
§ 9. We may now return to the argument of § 2. In that argument, it will have been noticed, no reference was made to the prices of the several factors of production that are at work. Quantities of resources in a physical sense were related directly to quantities of output, so what was said was complete. Now, provided that (small) variations in output in a given industry do not involve variations in the prices of any factor of production, variations in the quantities of the several factors employed may, with propriety, be measured by the variations in the amount of money which is expended by the industry in purchasing them for its use. In these conditions the argument of § 2 can be translated into an inverse form thus. The marginal private net product of the equilibrium firm is equal to the average net product of the equilibrium firm per unit cost; and is thus the reciprocal of the supply price of the product. The marginal social net product, on the other hand, is the reciprocal of the marginal supply price of the product to the industry, i.e. of the difference made to the total money expenses of the industry by adding a small increment of output. Hence to say that the marginal private net product of investment in any industry is greater (or less) than the marginal social net product is the same thing as to say that the supply price is less (or greater) than the marginal supply price to the industry. This fact, taken in conjunction with the argument of Appendix III. §§ 16-17, implies that in a many-firm industry the value of the marginal private net product of any quantity of investment is greater than, equal to or less than the value of the marginal social net product, according as the industry conforms to conditions of increasing, constant or decreasing supply price from the standpoint of the industry—which in the conditions supposed is the same as the standpoint of the community. When, in a competitive industry, variations in output involve variations in the price of some of the factors of production employed, so that the rates of change in supply price, regarded from the two standpoints, differ, an extension of the above argument shows that the value of the marginal private net product of investment is greater than, equal to or less than the value of the marginal social net product according as the industry conforms to conditions of increasing, constant or decreasing supply price from the standpoint, not of the industry, but of the community.*21
§ 10. It remains to inquire in this last case what, if any, light can be thrown on the relation between the values of the marginal social and marginal private net products, if we know of an industry merely that it conforms to one or other of the laws of increasing, constant or decreasing supply price simpliciter (i.e. from the standpoint of the industry). The conclusions that flow from the preceding analysis are then as follows: First, it is on the face of things very improbable that an increase in the output of any commodity will cause a fall in the aggregate money price that would have to be expended to secure the same quantities of all the factors collectively as were employed before the increase.*22 Therefore, when conditions of decreasing supply price simpliciter rule, conditions of decreasing supply price from the standpoint of the community must, in general, also rule, and, therefore, under competitive conditions, the marginal private net product of investment in the industry will, in general, fall short of the marginal social net product. When, however, conditions of increasing supply price simpliciter rule, it need not happen that conditions of increasing supply price from the standpoint of the community also rule: indeed, except in the special circumstances described in § 8, this cannot happen. Therefore the law of increasing supply price simpliciter does not imply that, under competitive conditions, the marginal private net product of investment in the industry exceeds the marginal social net product; on the contrary it may fall short of it. Hence, while, with rare exceptions, simple competition always causes too little investment to be made in industries of decreasing supply price (simpliciter), it does not always, or even generally, cause too much to be made in industries of increasing supply price (simpliciter). On the contrary, in a number of those industries it may cause too little investment to be made. British agriculture, for example, though obviously conforming to conditions of increasing supply price (simpliciter), may well be an industry of decreasing supply price from the standpoint of the community, and as such, in danger of suffering from a shortage of investment.
§ 11. If the amount of investment in any industry was carried exactly to the point at which the value of the marginal social net product there is equal to the central value of marginal social net products, the national dividend, so far as that industry is concerned, would be maximised. Disregarding the possibility of multiple maximum positions, I propose, for convenience, to call the investment that would then be made in the industry the ideal investment and the output that would be obtained the ideal output. Under conditions of simple competition, if in any industry the value of the marginal social net product of investment is greater than the value of the marginal private net product, this implies that the output obtained is less than the ideal output: if the value of the marginal social net product is less than the value of the marginal private net product, this implies that the output obtained is greater than the ideal output. It follows that, under conditions of simple competition, for every industry in which the value of the marginal social net product is greater than that of the marginal private net product, there will be certain rates of bounty, the granting of which by the State would modify output in such a way as to make the value of the marginal social net product there more nearly equal to the value of the marginal social net product of resources in general, thus—provided that the funds for the bounty can be raised by a mere transfer that does not inflict any indirect injury on production—increasing the size of the national dividend and the sum of economic welfare; and there will be one rate of bounty, the granting of which would have the optimum effect in this respect. In like manner, for every industry in which the value of the marginal social net product is less than that of the marginal private net product, there will be certain rates of tax, the imposition of which by the State would increase the size of the national dividend and increase economic welfare; and one rate of tax, which would have the optimum effect in this respect. These conclusions, taken in conjunction with what has been said in the preceding paragraphs, create a presumption in favour of State bounties to industries in which conditions of decreasing supply price simpliciter are operating, and of State taxes upon industries in which conditions of increasing supply price from the standpoint of the community are operating. They do not, of course, create a presumption in favour of fiscal interference with industries selected at haphazard or operated through rates of bounty or tax so selected. It is true that particular drugs consumed in particular quantities at particular times may cure diseases; but it is no less true that the consumption of drugs in general in a miscellaneous manner is highly injurious to health.
§ 12. Moreover, it may be well to make explicit a further consideration. When it was urged above that in certain industries a wrong amount of resources is being invested because the value of the marginal social net product there differs from the value of the marginal private net product, it was tacitly assumed that in the main body of industries these two values are equal, and, therefore, that there is scope for increasing the national dividend by shifting resources between the particular industry under review and this body of industries. If in all industries the values of marginal social and marginal private net product differed to exactly the same extend, the optimum distribution of resources would always be attained, and there would be, on these lines, no case for fiscal interference. It would still be possible, however, to defend a system of bounties to industries in general, the funds for which should be collected by some kind of lump-sum taxation, by arguing that the sum total of effort and waiting devoted to industry could be increased with advantage to economic welfare. Moreover, even when attention is confined to the distribution of resources among the several industries, what has been said does not imply that the mere prevalence in all industries of some degree of decreasing supply price from the standpoint of the community would rule out fiscal interference. It would still be possible—at all events in theory—to increase the national dividend by shifting resources from industries in which the law of decreasing supply price acted only weakly to industries in which it acted strongly.
§ 13. For completeness—though strictly this matter lies outside our formal limits—attention may be called, in parenthesis, to a type of reaction analogous to that discussed in § 12 of Chapter IX. The investment of an increment of resources in an industry, besides yielding a product to the purchasers, not reflected in the investor's profit, by altering costs of production, may yield a further product of a like kind by altering the amount of satisfaction which the purchasers derive from a given quantity of their purchases. This form of indirect product may be either positive or negative. Among commodities, the desire for which is partly a desire to possess what other people possess, the creation of the 1000th unit adds to aggregate satisfaction more satisfaction than it carries itself, because it makes every unit of the commodity more common. Top-hats are examples. Among commodities, the desire for which is partly a desire to possess what other people do not possess, the creation of the 1000th unit adds to aggregate satisfaction less satisfaction than it carries itself, because it makes every unit of the commodity more common. Diamonds are examples.*23 Among industries whose products are desired for their own sake, and not as means to any form of distinction, the creation of the 1000th unit adds to aggregate satisfaction exactly as much satisfaction as it carries itself. On the basis of this analysis, inferences analogous to those set out in § 11 are easily obtained. For every industry, the desire for whose products is enhanced if they become less common, there must be certain rates of tax, the levy of which on the industry would increase economic welfare; and for every industry, the desire for whose products is enhanced if they become more common, there must be certain rates of bounty, the imposition of which would produce a like effect. But there is reason to believe that the great bulk of ordinary commodities consumed by the mass of the population are desired almost entirely for their own sake, and not as a means to any form of distinction. The sphere of usefulness that could belong, even under a perfectly wise and perfectly virtuous Government, to these fiscal devices is, therefore, probably smaller than it might appear to be at first sight.
§ 14. These results, like the companion results which will be established presently in connection with monopoly, are results in pure theory. Attempts to develop and expand them are sometimes frowned upon on the ground that they cannot be applied to practice. For, it is argued, though we may be able to say that the size of the national dividend and the sum of economic welfare would be increased by granting bounties to industries falling into one category and by imposing taxes upon those falling into another category, we are not able to say to which of our categories the various actual industries of real life belong. In other words, it is maintained that the economic boxes and sub-boxes, labelled increasing, constant and decreasing supply price (whether simpliciter or from the standpoint of the community) and so on, are empty boxes and, therefore, useless except as toys. This conclusion does not, however, appear to be well grounded. Even though we should be for ever unable to fill these boxes and sub-boxes, the labour involved in studying them would not be thrown away. By means of it we are enabled to see, for example, what conditions are implicitly assumed when it is stated that the imposition of a tax or the introduction of monopolistic policy will have such and such consequences. We are thus put in a position to detect and expose sophistical dogmatism. It is better to know exactly what facts are required to make the answering of a question possible, even though those facts are unattainable, than to rest in a fog of vague and credulous opinion. But this is not all. Difficult as it must necessarily be to classify industries into the categories which analysis has distinguished, we need not yet conclude that it is impossible. Statistical technique, by itself, in spite of the growing volume and improving quality of the material available, will not enable us to accomplish this; for statistics refer only to the past. But able business men with a detailed realistic knowledge of the conditions of their several industries should be able to provide economists with the raw material for rough probable judgments. Economists unaided cannot fill their empty boxes because they lack the necessary realistic knowledge; and business men unaided cannot fill them because they do not know where or what the boxes are. With collaboration, however, it is not unreasonable to hope that some measure of success may eventually be achieved. At least the effort is worth making. It is premature, in impatience at the present shortage of straw, to scrap our brickmaking machinery. It is the better part to advertise abroad the urgent need for straw, and to call for students to produce it.*24
Notes for this chapter
It is not necessary to suppose that this central value is actually attained in any industry; it is rather to be conceived as the level which would be attained under conditions of simple competition in an industry of constant supply price.
In view of the definition of an increment of investment given in Chapter IX. § 2, we must not speak of units being added to the resources without qualification that flow into an occupation.
In an industry, for whose product the demand has an elasticity equal to unity, the amount of resources devoted to production will obviously be the same whether, the social net product of any rth increment of investment being given, this social net product of that increment is equal to or greater than the private net product. Therefore, when there is an excess of social over private net product in respect of the marginal unit of resources invested, the effect of the existence of this excess, in conditions of competition, is that consumers obtain for nothing exactly this excess: and the effect of there being an excess of (the given) social above private net product in respect of the other units of invested resources is that they obtain for nothing the aggregate of all these excesses. In industries, however, in which the elasticity of demand is not equal to unity, the existence of an excess of (the given) social above private net product of various quantities of investment causes the quantity of investment made to be different from what it would otherwise have been. This implies that the effect on the consumers' material estate, due to the existence of an excess of social over private net product in respect of the unit of investment which is marginal when there is such an excess, is not simply the amount of that excess. In like manner the effect on their material estate of the sum of all the excesses of the units of resources that are in fact employed is not simply the sum of those excesses. It follows, of course, that the effect on their satisfaction (as expressed in money) is not measured, as it is in the case of unitary demand elasticity, by the excess of their aggregate demand price for the quantity of product they do secure over their aggregate demand price for the quantity that they would have secured had (the given) social net product throughout been equal to private net product. These considerations render it inappropriate, except when the elasticity of demand is unitary, to speak of the excess of marginal social net product over marginal private net product as equivalent, in conditions of competition, to what "accrues" to the consumer in consequence of there being such an excess. For this reason I have in the present text modified the phraseology employed in the corresponding passages of the third edition.
For another use cf. post, Part IV. Ch. III.
Professor Cannan has objected to the use of the term "law" in connection with diminishing and increasing returns as defined above, on the ground that, whereas in some industries diminishing, and in others increasing, returns prevail, a scientific law is a statement that holds true in all, and not only in some, circumstances (Wealth, p. 70). It might be answered that in fact this is only true of the most general laws of physics. Biologists, for example, regularly speak of Mendel's law of inheritance, without any implication that all inheritance obeys this law. But in any event the point is a verbal one.
Geometrically the continuous fall in costs would be represented by a lowering of the whole supply curve.
In this case the extent to which the seams are worked out at any time is, of course, a result of the scale of output that ruled in the past; but this leaves my distinction intact.
This conception, though mathematically simple, needs to be handled carefully when translated into ordinary language. An industry conforms to conditions of decreasing, constant or increasing supply price simpliciter according as the rate of increase, from the standpoint of the industry, of the supply price (associated with increasing output) is negative, nil or positive: it conforms to conditions of decreasing, constant or increasing supply price from the standpoint of the community according as the rate of increase from the standpoint of the community is negative, nil or positive. The rate of increase from the standpoint of the industry of the supply price is a differential, the integral of which is the supply price. The rate of increase from the standpoint of the community of the supply price is also a differential, but the integral which corresponds to it has no real significance. The rate of change from the standpoint of the community in the supply price does not mean the rate of change in the supply price from the standpoint of the community. There is no separate supply price from the standpoint of the community: there is only one supply price from all standpoints.
Work and Wages, vol. i. p. 166.
Principles of Economics, p. 318.
Cf. Appendix III. § 17.
It is not, it should be understood, impossible that this should happen. For example, if an industry employed only a small proportion of the total supplies of two factors of production and nearly the whole supply of a third, and if an increase in the scale of output caused new methods to be introduced which led to an absolute decrease in the amount of this third factor that was wanted, the prices of the first two factors would be practically unchanged, while that of the third would fall substantially. As a result a units of the first plus b units of the second plus c units of the third might cost less money than before. Plainly, however, such a combination of circumstances is not likely to occur.
It should be added that, when commonness or rareness is an element in the esteem in which a person holds a thing, it is often not general commonness or general rareness alone, but, in many instances, both commonness among one set of people and also rareness among another set. As Mr. McDougall writes of the followers of fashion: "Each victim is moved not only by the prestige of those whom he imitates but also by the desire to be different from the mass who have not yet adopted the fashion" (Social Psychology, p. 336). This aspect of the matter cannot, however, be pursued here.
Cf. a papor by Dr. Clapham on "Empty Economic Boxes" in the Economic Journal for September 1922, and a reply by the present writer in the December issue of the same journal.
Part II, Chapter XII
End of Notes
Return to top