The Economics of Welfare

Part II, Chapter XXI

PUBLIC CONTROL OF MONOPOLY

§ 1. IN the course of Chapters IX., X. and XI., reference was frequently made to devices, by which the State could interfere, where self-interest, acting through simple competition, failed to make the national dividend as large as it might be made. Apart from governmental operation of industries into which the right amount of resources would not otherwise be turned, and apart also from penal legislation in extreme cases, these devices were fiscal in character and consisted in the concession of bounties or the imposition of taxes. Where self-interest works, not through simple competition, but through monopoly, fiscal intervention evidently ceases to be effective. A bounty might, indeed, be so contrived as to prevent restrictions of output below what is socially desirable, but only at the cost of enabling the monopolist to add to his already excessive profits a large ransom from the State. In the present chapter, therefore, I propose to consider what methods are available under conditions of monopoly. For simplicity of exposition, I shall ignore the qualifications set out in Chapter XI. and proceed as though simple competition might still be believed, as it was believed by some of the more rigid followers of the classical economists, to make the national dividend a maximum. The State, then, contemplating a monopoly or the possibility of a monopoly in some industry, may be supposed to contrast the dividend under it with the dividend under simple competition. Its problem will be, not to make things perfect, but to make them as good as they would be if monopolistic power were not at work.

§ 2. In industries where monopolistic power is liable to be introduced through the development of combinations, it is open to the State, if it chooses, to aim at preventing monopoly power from arising, or, if it has arisen, at destroying it. The original Federal Anti-Trust Law (1890) of the United States, commonly known as the Sherman Act, was overtly directed against actions “in restraint of trade or commerce among the several States,” but was interpreted in the earlier decisions of the Supreme Court as an Act banning all combinations large enough to possess a substantial element of monopolistic power. Thus Justice Harlan’s judgment in the Northern Securities Case, 1904, asserted that “to vitiate a combination, such as the Act of Congress condemns, it need not be shown that the combination in fact results, or will result, in a total suppression of trade or in a complete monopoly, but it is only essential to show that by its necessary operation it tends to restrain interstate or international trade or commerce, or tends to create a monopoly in such trade or commerce and to deprive the public of the advantages that flow from free competition.”
*1 The Clayton Law of 1914, while making no further provision as regards combinations that had already been formed, follows the line of this interpretation as regards the formation of new combinations in the future. It lays it down, not only that no person shall be a director of more than one large bank or large corporation, but also that no corporation shall acquire—acquirements already made are not affected—the whole or part of the stock of any other corporation, when the effect of such acquisition may be substantially to lessen competition, or to restrain commerce in any section of the community, or to tend to the creation of a monopoly in any line of commerce. This general policy—trust prohibition and trust breaking—seems, however, to be open to three serious objections.

First, it is a policy exceedingly difficult to enforce in an effective manner. The legislature and the courts may succeed in getting rid of certain forms of combination, but the result will often be merely the appearance of other forms—possibly of forms, which, as would happen if an
informal price-fixing agreement took the place of a complete amalgamation, sacrifice the merits, without getting rid of the demerits, of those which preceded them. The declaration of the Supreme Court of the United States, that the granting of a power of attorney to common trustees by a number of companies was
ultra vires, led, in some industries, to the purchase of a majority of stock in each of the companies by the said trustees, and, in others, to the substitution of a holding company for a trust. Governmental attacks on holding companies can easily be met either by complete consolidation, if this is not also made illegal, or by dissolution into separate companies, each subject to the same controlling interest. The Austrian law against Kartels likely to injure the revenue abolished Kartels possessed of a central office; but only with the result of substituting informal understandings. The British Committee on Railway Agreements and Amalgamations (1911) summed up the situation thus: “While Parliament may enact that this must be done and that must be prohibited, past experience shows that even Parliament appears to be powerless to prevent two parties, either by agreement or without formal agreement, from abstaining from a course of action, namely, active competition, that neither party desires to take. Parliament can, of course, refuse to sanction Bills authorising the amalgamation or working union of two or more railway companies, and may provide that certain classes of agreement shall be invalid or even illegal. But it cannot prevent railway companies [and, of course, the same thing is true of industrial companies] coming to understandings with each other to adopt a common course of action, or to cease from active competition.”
*2 The recent policy of the United States Government and Courts, in forcing the dissolution of monopolistic companies into their constituent parts and providing at the same time various regulations to prevent these from becoming subject to a common control, may, indeed, for a time be more effective, and, even though it does not succeed in stimulating real competition among men formerly colleagues, yet may, by its harassing
effect, make the task of forming new combinations less attractive. It was stated, for example, by Professor Durand in 1914, that no new combinations had been formed since the Government began to bring suits under the Sherman Act.
*3 This, however, is no longer true. In the Memorandum on American Combinations furnished by the Board of Trade to the Committee on Industry and Trade in 1927, the following conclusion is reached. “Only a very partial success has been achieved in preventing the growth of combination or the lessening of competition. The experience of the administration of the Sherman Act between 1890 and 1914 seems to have been repeated. When one form of combination is attacked and declared illegal, the lawyers advising the corporation evolve a new form, which, even if an objection by the Federal Trade Commission is eventually upheld, takes a considerable period to upset.”
*4 Thus it may still be claimed, as the teaching of experience as a whole, that laws aimed directly at “maintaining competition” have very small prospect of succeeding in their purpose.

There is a second serious objection to this policy. The root idea lying behind it is that competition implies a condition of things in which the value of the marginal social net product of investment in the businesses affected is about equal to the value of the marginal social net product elsewhere. But, passing by the qualifications to this view set out in Chapters IX. and XI., we have to note that the competition, from which the above good result may be expected, is “simple competition,” whereas the competition, to which laws against combination lead, will very probably be monopolistic competition, namely, competition among a
few competitors. With railway combinations this result is certain; for the number of railways plying between any two given centres is necessarily very small. With industrial combinations the issue at first sight seems more doubtful, since there is no such necessary limitation in the number of industrial concerns of any given type. When,
however, we reflect that combinations can rarely be organised except in industries where, as a matter of fact, the number of leading firms is small, the force of this objection is much reduced. Among industrial combinations, as well as among railway combinations, dissolution is much more likely to lead to monopolistic than to simple competition. It has been shown, however, in Chapter XV., that monopolistic competition does not tend to bring about an output of such a magnitude that the value of the marginal social net product of investment in the industry affected is equal to that prevailing elsewhere. On the contrary, the output is indeterminate. When the competitors hope to destroy or to absorb one another, we may get “cut-throat competition,” under which production is carried so far as to involve absolute loss; and the chance of this is made greater by the desire of one giant business to win even a barren victory over another. In short, even if the conditions were such that laws for “maintaining competition” could really prevent combination, they would still be unable to secure the establishment of competition in that sense in which alone it can be expected to evolve the level of prices and the rate of output which is most advantageous from the standpoint of the national dividend.

Even now, however, the case against the policy we are considering is not exhausted. There remains a third objection. Combination is not the parent of monopoly only, but also, very often, of incidental benefits. Thus, as was observed in Chapter XIV., a combination, which is large relatively to the market in which it trades, has more inducement than a small single seller to adopt a policy of developing demand among potential customers, since it may reckon on receiving a larger proportion of the gain resulting from any investments which it may make with this object. In addition to this, a large combination will often enjoy certain economies of production, which, if the Government were to adopt a policy of maintaining active competition, would fail to emerge. No doubt, some of those forms of Kartel agreement, under which a proportion of the market is guaranteed to the several members, since they tend to conserve weak firms, which competition would “naturally” destroy, not only fail to yield economies,
but actually yield diseconomies.
*5 It should be observed, however, that pooling agreements do not necessarily act in this way. The British Committee on Trusts (1919), for example, reports that in a great many associations there is an arrangement under which firms, on producing less than their quota, receive from the pool 5 per cent in value upon the amount of their deficiency. It was urged by some witnesses that this arrangement had the effect of driving weak firms out of the industry by the economical method of pensioning, instead of the more costly method of fighting them.
*6 Against this, indeed, we have to set the fact that the money to provide the pensions has to be obtained by some kind of tax on firms that exceed their quota—a necessary discouragement to them. Moreover, some forms of pool, by making the profits of each member severally depend on the efficiency of all collectively, may lead to relaxed energy and enterprise. But, on the other hand, in all combinations that involve any measure of common management, savings of the kind referred to in Chapter XIV. are bound to accrue in greater or less degree.
*7 In a peculiar industry like the telephones, where the actual thing supplied to A is improved if B draws his supply from the same agency, the advantage is especially great. It may also be considerable in more ordinary industries.
Inter alia, weak or badly situated plants are apt to be shut down
much more quickly than they would be under competition; while, among those that remain, the purposive force of “comparative cost accounting” may be expected to stimulate the energy of managers more strongly than the blind force of market rivalry could ever do.
*8

We must, indeed, be on our guard against exaggerating the importance of these economies. For, if by combination we mean existing combinations, it is necessary to recollect that, since the magnitude of the unit of control is determined by monopolistic considerations as well as by considerations of structural and other economies, this unit is often larger than the unit of maximum efficiency. If we mean only such combinations as it would be profitable to form
de novo, were the exercise of monopolistic power wholly excluded, combination will, indeed, evolve the unit of maximum
immediate efficiency, but that unit will very likely prove too large when ultimate indirect effects, as well as immediate effects, are taken into account. For this there are two reasons. The first is that a producer controlling the main part of any industry, in considering the wisdom of adopting any mechanical improvement, is tempted to take account, not merely of the direct positive yield to be expected from capital invested in that improvement, but also of the indirect negative yield in lessening the returns to his existing plant. But, if he does this, he will, as was shown on pp. 190-92, be holding back from improvements which it is to the interest of the national dividend that he should adopt. A monopoly makes no proper use—at all events is under temptation to make insufficient use—of that invaluable agent of progress, the scrap heap.
*9 The second reason is that indicated in Chapter X., namely, that large combinations, by lessening the opportunities for training in the entrepreneur function, which are available when men who have done well in one company can be passed on to more responsible work in another, and by weakening the stimulus to keenness and efficiency, which is afforded by the rivalry of
competing concerns, tend indirectly to prevent the average level of business ability from rising as high as it might otherwise do.

The qualifications which these considerations suggest are of great importance. They tell strongly against the claim made by Professor Clark, when he writes: “A nearly ideal condition would be that in which, in every department of industry, there should be one great corporation, working without friction and with enormous economy, and
compelled to give to the public the full benefit of that economy.”*10 Nevertheless, there can be little doubt that,
in some circumstances, the combination of competing institutions into “Trusts” and consolidations that dominate the market does involve, even from a long-period point of view, considerable net economies.
*11 These economies
may be so great that the favourable effect produced by them on the output of the monopolised commodity exceeds the unfavourable effect produced on this output by the exercise of monopolistic power. Attempts to determine, by a comparison of prices, or of “margins” between prices and the cost of materials, before and after the formation of any combination, whether or not this has actually been so, are inevitably baffled by our inability to allow for changes in manufacturing methods and in the utilisation of by-products, or to gauge accurately the—probably abnormal—price conditions that ruled immediately before the combination was formed.
*12 Analysis, however, enables us to say that combination is likely, on the whole, to diminish the output of the commodity affected by it and to raise its price, unless the associated economies are
so large that, had they been introduced without monopolisation, they would have raised output to about double its former amount.
*13 Economies so large as this are improbable, and I do not, therefore, claim that to prevent combination in any department of industry would often make the output of that department smaller than it might have been. But—and this is the essential point—the effect of combination on the output of the commodity affected by it is not the same thing as its effect on the national dividend. For suppose combination to bring about economies which enable the same output as before to be produced by the use of half the former quantity of productive resources, and suppose that, in consequence of monopolistic action, no more than this output is in fact produced. The released productive resources will not, in general, be idle, but will be occupied in adding to the output of other commodities. Hence, in this case, the output of no commodity is diminished and the output of some commodities is increased, which obviously implies that the size of the national dividend is increased. It follows that to prevent combination would sometimes injure the dividend, even though it enabled the output of the commodity immediately affected to be larger than it would have been. This point need not, however, be laboured further. For it is in any even certain that to prevent combination, when combination carries with it any net economies, must be more injurious to the dividend than to allow combination and to prevent the exercise of monopolistic power.

§ 3. A second line of policy which it is open to the State to pursue is as follows. Instead of endeavouring, by obstructing combination, to prevent industrial concerns from becoming possessed of monopolistic power, it may seek, by conserving
potential rather than actual competition, to make it to their interest to leave that power unexercised; the idea, of course, being that, if they expect new competitors to come into the field, should output be restricted and prices be raised high enough to yield abnormal profits, they will have no inducement to charge more than “reasonable” prices. The policy,
to which this line of thought leads, is that of penalising the use of “clubbing” devices, whose repute might otherwise drive potential competitors away. Among these devices the two principal are cut-throat competition, as described in Chapter XV., and various forms of boycott, namely, the exercise of pressure upon third parties not to purchase services from, or sell services to, a rival seller on terms as favourable as they would have offered to him if left to themselves.
*14

§ 4. It is obvious that the weapon of cut-throat competition, or, as it is sometimes called, “destructive dumping,” when practised by a business already large enough to monopolise any department of industry, must prove overwhelmingly powerful against newcomers. The monopolist necessarily possesses immense resources, and these can be poured out, in almost unlimited quantities, for the destruction of a new, and presumably much less wealthy, intruder. This is especially clear when a monopolist, dealing in many markets or in many lines of goods, has to do with a competitor dealing only in a few; for in these conditions the competitor can be destroyed by a cut, made either openly or through a bogus independent company,
*15 that affects only a small part of the monopolist’s business. An extreme example of this kind of cut is given in the statement of certain opponents of the Standard Oil Trust, “that persons are engaged to follow the waggons of competitors to learn who their customers are, and that then they make lower offers to those customers; and it is still further asserted that at times the employés in the offices of rivals are bribed to disclose their business to the Standard Oil Company.”
*16 It is needless to emphasise
the immense power of this weapon. “After two or three attempts to compete with Jay Gould’s telegraph line from New York to Philadelphia had been frustrated by a lowering of rates to a merely nominal price, the notoriety of this terrible weapon sufficed to check further attempts at competition.”
*17

§ 5. The weapon of boycott has a narrower range than that of cut-throat competition. It is worked through a refusal to deal, except on specially onerous terms, with any one who also deals elsewhere. When the worsened terms attached by a dominating seller to dealings with himself are more injurious to the client than the loss of that client’s other dealings, the monopolist can force the client to boycott his rivals. In order that he may be able to do this, the goods or services that he offers for sale must be rendered, by nature or by art, non-transferable;
*18 for it is impossible to hurt a customer by refusing to sell to him, if he is able to purchase through a middleman the goods which are refused to him by the monopolist. Hence, when nature does not cause non-transferability, there must be stringent conditions about re-sales in the contracts between the monopolist and any intermediary agents, if such there are, who intervene between him and the ultimate consumers. But non-transferability is not sufficient by itself. It is necessary, further, that the rival producer’s possible supply
to one recalcitrant consumer at current prices shall be very small. Usually, of course, though any one seller’s output is likely to be small relatively to the total consumption of the market, it is many times as large as that of any representative single consumer. Where this is so, recalcitrant consumers can successfully counter a refusal to sell on the part of the monopolist by purchasing all that they want from outside competitors and leaving to non-recalcitrants the whole output of the monopolist. This consideration is not, however, entirely fatal to the weapon of boycott, because in many industries, though by no means in all,
*19 producers deal with their customers
indirectly through wholesalers or further manufacturers or transporters, who purchase individually a considerable mass of products. When intermediaries of this kind are present, effective boycott may become practicable.

First, a boycott can be forced when the commodities or services supplied by the monopolist consist, not in a single kind of good, but in several goods, and when, among these several goods, there are one or more for which the demand is very urgent, and of which the monopolist has, through patents or reputation (e.g. brands of tobacco) or otherwise, exclusive control. A good example is furnished by the boot and shoe trade, in which certain firms control important patents. The patented machines are not sold, but are let out on lease, under “conditions which debar manufacturers from employing these machines save and except in conjunction with other machines supplied by the same controlling owners… one of the conditions being that the latest machines must not be used for goods which have, in any other process of manufacturing, been touched by machines supplied by other makers.”
*20 This kind of boycott is also illustrated by the “factors’ agreement,” which makers of popular proprietary goods sometimes secure from retailers.

Secondly, a boycott can be forced where it is important for purchasers—here, as before, the purchasers are, in general, manufacturers—to be able to get the service that they need immediately the need arises, and where an ordinary supplier, though producing much more service in the aggregate than any single purchaser wants, may not be producing more than such a purchaser wants at some definite single moment. This condition is realised in the transport of goods which are so perishable, or for which the demand is so instant, that transport, to be of use, must be available at the moment when it is asked for. It is in the transport by sea of goods of this kind that the method of boycott has been most fully elaborated.
The transport of goods, which are in fairly steady demand and which have no need of speedy delivery, can be arranged for by purchasers, if they wish, wholly through tramp steamers; but the transport of urgent goods cannot be so arranged for, because tramps and small lines cannot guarantee regular sailings.
*21 Hence it comes to be practicable for shipping rings to force a boycott against independent lines. They usually accomplish this through “deferred rebates.”
*22 Of these there are two degrees. When the Royal Commission on Shipping Rings published their report (1909), in the West
African Shipping Conference and in all the Conferences engaged in the trade with India and the Far East, the rebates were paid to exporting merchants only, on condition that these merchants had not been interested in any shipment by rival carriers, but there was no requirement that the forwarding agent, through whom the merchant might have acted, should have dealt exclusively with the Conference in respect of the goods of his other clients.
*23 In the South American Conferences, however, “the form of claim for rebates has, in the case of goods shipped through a forwarding agent, to be signed by such agent as well as by the principal, and, if the forwarding agent has not conformed to the conditions of the rebate circular in all his shipments for all his clients, claims to rebates are invalidated.”
*24

Thirdly, a boycott can be forced when the intermediary, whom a monopolist wishes to use as his instrument, is, not a manufacturer or a wholesaler purchasing that rival’s goods, but a railway company conveying them. When an alternative route for his own goods is available, the monopolist, by threatening the railway with the withdrawal of his custom, is sometimes in a position to compel it to charge differential rates against his rival. In a boycott engineered by the Oil Trust it is even asserted that the railways were compelled to hand over a part of the extra charges levied on their rivals to the executive of the Trust.
*25 When the boycotting concern actually owns the agency of transport, its power in this matter is, of course, still greater.
*26 A boycott of this kind
may also be operated through banks, pressure being exerted upon them to refuse loans to a rival producer.

§ 6. Attempts to prevent the use of cut-throat competition,
i.e. destructive dumping, by legal enactment are confronted with the difficulty of evasion. The American Industrial Commission recommended that “cutting prices in any locality below those which prevail generally, for the purpose of destroying local competition,” should be made an offence. Any person damaged was to have the right to sue for penalties, and officers were required to prosecute offenders.
*27 It is plain, however, that, even when it is possible, as it is with public service corporations, to insist that tariff rates shall be regularly published, evasion may be practised through unpublished discounts and rebates to particular customers; nor, since discovery is unlikely, will the enactment of heavy penalties against breaches of the law necessarily secure obedience thereto.
*28 Where destructive dumping is threatened, not by public service corporations, but by industrialists engaged in the manufacture of many commodities at different places, the enforcement of regular published rates is impracticable. Hence the problem confronting the legislator demands the unravelling of still more tangled knots. Where the form of destructive dumping which is employed is that of price-cutting limited to the local market of a particular competitor or group of competitors, the offence is at least definite, though, especially when worked through a bogus independent company, it may be extraordinarily difficult to detect. Against destructive dumping of this kind
operated by foreigners the Governments of Canada (1904) and South Africa (1914) have endeavoured to guard their citizens by anti-dumping legislation, providing that, when goods are exported to them at prices, which, exclusive of freight charges and so on, are substantially below
the contemporary prices ruling at home, these goods shall be subjected to a special import duty equivalent to the difference between the home and foreign prices.
*29 This legislation, however, hits, not merely destructive dumping in the sense here defined, but also (1) the clearing of surplus stock on a foreign market at less than home prices in periods of depression and (2) the permanent selling abroad at the world price, by a foreign monopolistic producer, of goods for which at home he is able to charge monopoly prices. The policy of discouraging the former of these two practices is one whose merits are open to debate, but clearly there is nothing to be said for discouraging the second—except, indeed, the least tenable of the things that can be said in favour of all-round protection. The United States Government, wishing to direct its legislative blows against destructive dumping exclusively, included in the Federal Revenue Act of 1916 the following modified version of the Canadian anti-dumping law. In Section 801 of the Act it is enacted: “That it shall be unlawful, for any person importing or assisting in importing any articles from any foreign country into the United States, commonly and systematically to import, sell or cause to be imported or sold such articles within the United States at a price substantially less than the actual market value or wholesale price of such articles, at the time of exporting to the United States, in the principal market of the country of their production, or of other foreign countries to which they are commonly exported, after adding to such market value or wholesale price freight, duty and other charges and expenses necessarily incident to the importation and sale thereof in the United States. Provided, that such act or acts be done with the intent of destroying or injuring an industry in the United States or of preventing the establishment of an industry in the United States or of
restraining or monopolising any part of the trade or commerce in such articles in the United States.” Offences against this clause were penalised, not, as in Canada, by a special duty, but by a fine. Under the Emergency Tariff Act of 1921 and the final Act of 1922 the reference to intention was omitted, and the imposition of a special duty was authorised whenever an efficiently conducted industry of the United States was being, or was likely to be, injured by importation at less than the price ruling in the principal home market of the exporting country
plus f.o.b. costs. In the United Kingdom the Safeguarding of Industries Act 1921 provided for the imposition of special duties on goods that are being imported from any foreign country and sold in the United Kingdom at a price,
inclusive of freight charges, less than 95 per cent of the wholesale price at the works charged to consumers in that country; provided that, by reason of the importation, employment in any industry in the United Kingdom is being, or is likely to be, seriously affected. An Act of the same general character was passed in Australia in 1921, though under that Act discretion is allowed to the Executive to refrain from action if it so chooses. We need not concern ourselves here with the difficult problems which legislation of this kind provides for the officials charged with ascertaining the relevant facts—including, as these did under the Federal Revenue Act of 1916, the motive actuating foreign sellers—and for those who have to detect and prevent the evasive use of nominally independent agents, on whose account goods may be imported at full price, thereafter to be sold at less than was paid for them. The main point for our present purpose is that, when, as in the conditions which this legislation contemplates, destructive dumping is attempted by inter-local price discriminations, the task of preventing it is
relatively easy, because there is something definite to go on. When, however, as often happens in domestic trade, we have to do with cuts made on
all sales of a particular line of goods, the offence is not definite; for, clearly, not all cuts are destructive dumping, and it is difficult to distinguish among them the innocent from the guilty. One authoritative writer proposes
as a test that, “if the price of the particular grade of goods were first put down and then put up again, and if rivals were crushed in the interval, this would be evidence that the purpose of the cut was illegitimate.”
*30 Such a test has been attempted in the American Mann-Elkins Railway Law of 1910, which provides that, “when a railway reduces rates between competitive points, it shall not be permitted to increase the rates on the cessation of the competition, unless it can satisfy the Commission that the conditions are changed otherwise than by the mere elimination of water competition.”
*31 There is a similar provision in the American Shipping Act of 1916 as regards shipping charges in interstate trade. But this test cannot be pushed very vigorously; for, if it were, any firm, which lowered prices in a time of depression or for purposes of experiment, might find itself precluded from afterwards raising them again, should any other firm in the same line meanwhile have failed.

Similar difficulties stand in the way of effective legislation against boycotts. It is true that such legislation has been widely attempted. The United States (under the Clayton Law), Australia and New Zealand all prohibit, under penalty, any person from making the act of sale, or the terms of sale, of anything conditional on the buyer not using or dealing in the goods of any competitor. On similar lines the United States Federal Revenue Law of September 1916 “imposes a double duty upon goods imported under agreement that the importer or others shall use those goods exclusively.”
*32 Yet again, the American Shipping Act of 1916 makes deferred rebates illegal. It is obvious, however, that, when the condition or agreement is made between a manufacturer and a dealer, both of whom profit by it, the difficulty of preventing evasion must be very great. When the boycott is worked; not through a wholesaler, but through a railway company, the difficulty is still
greater. American law has long endeavoured to prevent railway discriminations favourable to the large Trusts. But: “A partisan of the Trust said to me: ‘The Pennsylvania Railroad could not refuse the cars of a competitor of the Standard Oil Company, but nothing could hinder it from side-tracking them.'”
*33 “A consignment note acknowledges the receipt of 70 barrels of flour; 65 only are shipped, and the railway company pays damages for the loss of the five non-existent barrels.” Except when long notice of alterations is required by law, rates may be changed suddenly, secret notice being given to the favoured shipper and no information to others; and so forth. It is true that the Attorney-General of the United States declared in 1903, after the passing of the Elkins law: “The giving and taking of railroad rebates is now prohibited by a law capable of effective enforcement against corporations as well as against individuals.”
*34 This view, however, appears to have been unduly optimistic. The Interstate Commerce Commission reported, as to the conditions in 1908, that many shippers still enjoy illegal advantages. “Thus the rebate, as an evil in transportation, even since amendment of the law in 1906-10, while under control, is still far from being eradicated. Favouritism lurks in every covert, assuming almost every hue and form. Practices, which outwardly appear to be necessary and legitimate, have been shown to conceal special favours of a substantial sort.”
*35 The boycott engineered through railway companies thus dies hard. It is said, however, that, in the United States, the Transportation Act of 1920, which created a system of Federal supervision over railways, has finally put an end to it.
*36

These considerations make it clear that a policy of legal prohibition against the exercise of clubbing methods cannot easily be rendered proof against evasion. It should not be forgotten, however, that laws, which
could be evaded if people took sufficient pains, as a matter of fact are often not evaded. For the mere passage of a law reacts on public opinion and
throws on the side of the practice upheld by law the strong forces of respectability and inertia. Hence we may reasonably expect that laws of this character, if carefully prepared, would, at all events, partially succeed in their immediate purpose. It is, therefore, of great interest to observe that Section 5 of the United States Federal Trade Commission Act of 1914 declares “that unfair methods of competitive commerce are hereby declared unlawful,” and establishes a Federal Trade Commission to take proceedings to enforce this declaration whenever it appears to it to be in the public interest to do so. Section 14 of the Clayton Act provides further that, whenever a corporation violates any of the penal provisions of any of the anti-trust laws, “such violation shall be deemed to be also that of the individual directors, officers, or agents of such corporation as shall have authorized, ordered, or done any of the acts constituting in whole or in part such violation.” Upon conviction any director, officer or agent is subject to a fine not exceeding 5000 dollars, or to imprisonment not exceeding one year, or to both, in the discretion of the Court.

§ 7. Granted that clubbing methods can be, in some measure, prevented, we turn to the further question, how far their prevention would avail to maintain potential competition. Professor Clark appears to hold that it would avail completely for this purpose. “In so far,” he writes, “as legitimate rivalry in production is concerned, it is safe enough to build a new mill.” In reality, however, even when clubbing methods are excluded, other obstacles to the full maintenance of competition are still present. First, when the unit firm normal to any industry is very large, the heavy capital expenditure required to start a new firm will check the ardour of aspirants. Furthermore, it should be noticed, in this connection, that in many industries the size of the normal unit firm has recently been increasing. For example, the output of the English paper industry between 1841 and 1903 rose from 43,000 to 773,080 tons, but the number of firms fell from 500 to 282;
*37 and a like development has taken place in the raw iron industry. Secondly, the inducement
to new competitors to spring up is smaller, the greater are the productive economies which concentration on the part of the monopolistic seller has involved. For, if great economy has been brought about by concentration, a potential competitor will know that the monopolistic seller, by simply abandoning some of his monopoly revenue, can, without suffering any positive loss, easily undersell him. Thirdly, the obstacles in the way of new competition are further enlarged, when a policy of secrecy as to costs and profits makes it difficult for outsiders to guess at what rate the monopolistic seller
could sell, if he, were to content himself with the normal gains of competitive industry. Fourthly, a combination, by extensive advertising or a distinctive trade mark, may have established a sort of monopoly of reputation, which it would require heavy advertising expenditure on the part of any would-be rival to break down. It may, indeed, be suggested that, even so, the combination would be kept in check by fear of a rival concern being started with a view to forcing the combination to buy it out. But there is less in this than there might appear to be at first sight. For, if a rival did succeed in this policy, the increase in the combination’s capital might well be so large that the rate of profit available per unit of capital would turn out too small to make the venture worth the rival’s while.
*38 This consideration would, of course, tend strongly to hold him back. Thus attempts to maintain potential competition by preventing the employment of clubbing devices can at best be only partially successful, and are, therefore, very imperfect means of restraining bodies that possess monopolistic power from making use of that power. This is true even of industry proper. In some departments of production—roughly those covered by public utility concerns—the evident wastefulness of competition makes it practically certain that the public authorities will not permit it, and so exempts monopolists from any check which the fear of it might otherwise exercise upon them.

§ 8. The inadequacy, as a method of control, of preventing combination, which means maintaining actual competition, and of anti-clubbing legislation, which means maintaining
potential competition, leads forward naturally to the suggestion of direct methods. The position, which is relevant to industrial, no less than to railway, monopolies, is well put by the Departmental Committee on Railway Agreements and Amalgamations (1911) with special reference to the latter class. They write: “To sum up, we are strongly of opinion that, in so far as protection is required from any of the consequences which may be associated with railway co-operation, such protection should, in the main, be afforded by general legislation dealing with the consequences as such, independently of whether they occur as the result of agreement or not. Such a method would afford a much more extensive protection than the regulation of agreements. It would protect the public in the case of understandings as well as agreements…. It would not tend to introduce a confusing distinction between what a company might reasonably do under an agreement and what it might reasonably do if no agreement existed.”
*39 If this method could be employed with perfect accuracy, there would, of course, be no need for
any accompanying indirect methods of the kind we have so far been discussing. In practice, however, the policy of dealing directly with the consequences of monopolistic power is, as will presently appear, exposed to very great difficulties. Furthermore, since in most forms it must almost necessarily work on the basis of some standard of reasonable earnings, deduced from the circumstances of other industries in which competition is available, these difficulties would become enormous, if attempts to maintain potential competition were abandoned altogether and resort had universally to direct methods. Consequently, it would seem that the policy of maintaining potential competition should be pursued everywhere vigorously, and that direct methods of dealing with the consequences should be employed, not instead of, but in addition to it.

§ 9. At first sight it seems obvious that direct dealing with the consequences of monopolistic power means, and can only mean, some kind of direct interference on the part of the public authority. In the main, of course, this
is what it does mean. But it is of some theoretical interest to note a possible alternative line of policy which was advocated, as regards shipping, by the Royal Commission on Shipping Rings. The Commission recommended, in effect, that the State should encourage the formation, over against a monopolistic seller, of a combination of buyers possessing also monopolistic powers. It was hoped that the combination of buyers might be able to neutralize attempts on the part of the seller to charge monopoly prices. This plan was advocated as a partial remedy for the evils that have arisen out of the conference system. Analytically, the plan is a weak one, because what the creation of the second monopolist does is, not to bring prices to the natural, or competitive, point, but to render them indeterminate over a considerable range, within which that point lies. No doubt, the position of the purchasers is made better than it would be if combination among them were absent; and there is reason to hope that prices and output will approach more nearly to what, from the standpoint of the national dividend, is desirable than they would do under those conditions. But the chance that the bargain between the two combinations will lie in the near neighbourhood of that proper to simple competition is not very large. This difficulty would exist even though the monopoly created to stand against the sellers were a monopoly of ultimate consumers. In practice, however, ultimate consumers are scarcely ever in a position to combine in this way. The only persons who can so combine are middlemen between the ultimate consumers and the monopolistic seller,—middlemen who are not particularly concerned to fight for the consumers’ interests.
*40 If they combine, the goods in which they deal will have to pass through the hands of two monopolistic combinations, instead of one, before they reach the ultimate consumers. The effect upon the price which those people will then have to pay is economically indeterminate. The price may be less than it would have been if the middlemen had not combined, but it may, on the other hand, be greater. In any event, it and, with it, the quantity of service accessible to the
ultimate consumers, are likely to be exceedingly unstable.
*41 These considerations show that there is a serious flaw in the Commissioners’ policy.
*42

§ 10. Interference on the part of the public authority does not necessarily mean orders about the terms of sale. It may well be that anti-social practices by powerful corporations might be substantially restrained by publicity alone. An important part of the task assigned to the Federal Trade Commission of the United States is to make investigations and to publish reports. The British
Committee on Trusts (1919) recommended that the Board of Trade should obtain and publish in an annual report information about the development of organizations having for their purpose the regulation of prices or output so far as they tended to the creation of monopolies or to the restraint of trade; and should investigate complaints regarding the action of such organizations. Powers would, of course, need to be taken to compel the officers of the organisations affected to produce their books and to answer questions. The fear of an adverse report issued by an investigating body in which public opinion had confidence might often turn the scale against gross abuses of monopoly power. There is, in short, little doubt that the weapon of publicity can accomplish something of importance: but it can hardly accomplish all that is required.

§ 11. We turn, therefore, to interference by the public authority with the terms of sale—a method which
may be necessary even in industrial enterprises, when the “remedies” considered so far prove inadequate, and which, apart from public operation, is certainly necessary in public utility concerns. Analytically, the problem may be stated as follows. Assuming that the output proper to simple competition (allowing, of course, for any economies in production which a combination may have introduced) is also the output most advantageous to the national dividend, we need so to regulate things that that output will be forthcoming. In industries operating under
conditions of increasing supply price, this type of regulation cannot be accomplished by the machinery of price control alone. For, if the price be fixed by the State at the level proper to competitive conditions,
i.e. at such a level that, if competitive conditions prevailed, the output would be adjusted to yield normal profits, it will pay a monopolist to produce less than this output. By reducing output he will, under conditions of increasing supply price, also diminish the supply price, thus obtaining a monopoly gain measured by the difference between the regulated selling price multiplied by the output and the supply price multiplied by the output. It will be to his interest to control his output in such a way as to make this monopoly gain as large as possible. According to the form of the demand and supply schedules, the resultant output may be greater or less than it would have been under unregulated monopoly; but, in any event, it is certain to be less than the output proper to competition, at which the Government is aiming. This difficulty, however, is only present in industries operating under conditions of the increasing supply price. When constant supply price or decreasing supply price prevails, it will not pay a monopolist, when price is fixed at the competitive level, to reduce output below the competitive output; for he would not secure any diminution in his costs by doing this. Consequently, if the Government can succeed in fixing prices at the competitive level, it will also indirectly secure competitive output.
*43 As a matter of practice, concerns (whether industrial combinations or public utility corporations), which it is necessary to regulate because they tend towards monopoly, are rarely of a kind which we should expect to be subject to increasing supply price. In the main, therefore, control means control over price.

§ 12. When this has been said, there inevitably comes to mind the sort of control over price which was exercised during the Great War, and some account of which was given in Chapter XII. It is very important, however, to realize that what we are now concerned with is fundamentally different from that. In controlling monopoly, it is required to prevent the monopolist from charging high prices, because, by so doing, he will
reduce output below the level at which he could put it with normal profits to himself. As explained above, under conditions of constant supply price or decreasing supply price, the fixing of maximum prices at the rate corresponding to the “competitive” output will in fact cause that output to be forthcoming. There is no question of the maximum price being associated with an output for which the demand price that the public are prepared to pay exceeds that price.
*44 But in the war problem, as was clearly brought out in Chapter XII., the whole point of intervention was to fix a maximum price below the demand price that the public would be prepared, at need, to pay for such quantity of the commodity as was forthcoming. This is the reason why, at the maximum price, there was always a greater quantity demanded than could be supplied, and why, therefore, it was necessary to prevent accidental inequities in distribution by rationing all consumers to purchases smaller than many of them would have wished to make. This, too, is the reason why it was not sufficient to fix prices as from the producer only. Because the demand price was bigger than the price which the Government wished to allow, to have limited,
e.g., shipping freights, without also limiting the price of the things brought in at the limited freights, would merely have enabled the intermediaries between the ship and the consumer to take the whole benefit for themselves. It was necessary, therefore, not merely to fix maximum prices to the original producer, but also to fix maximum additions that might be put on to these prices by the various persons through whose hands (whether as further manufacturers or as retailers) the controlled commodities would afterwards pass. In the regulation of monopoly charges there is, of course, no need for any of these secondary arrangements.

§ 13. We may now proceed to investigate this form of price control directly. One way in which it may be exercised is, as it were, negative. It may take the form of general provisions against “unreasonable” conduct, leaving the definition of what is, in fact, unreasonable to the decision of a
Commission or of the Courts. This way was, in substance, followed, for proposed
changes of rates, in the work of the English Railway Commissioners prior to 1921, and of the American Railway Commissioners prior to the passage of the Hepburn law. The Commissioners had to decide whether any proposed increase of rates was reasonable, and to permit or forbid it accordingly. Thus their task was
comparatively light. They had not to regulate all prices always, but only to intervene against specially unreasonable prices; and, furthermore, the knowledge of their existence was likely to serve indirectly as a check against the setting-up of unreasonable prices.
*45 The negative way is also followed in certain franchises, which permit municipalities to take over the business of a licensed corporation
at a proper price—an ambiguous phrase—should the corporation fail to “operate and develop it in compliance with reasonable public requirements.”
*46 It is followed again in the Canadian Industrial Combines Investigation Act (1910). Provision is made for determining whether, with regard to any article, on the subject of which complaint has been made, “there exists any combine to promote unduly the advantage of the manufacturers or dealers at the expense of the consumers by fixing the price higher than is proper”; and, if the charge is established, penalties are provided. In New Zealand the Act of 1910 applies the same test. “Any person commits an offence, who, either as principal or agent, sells or supplies, or offers for sale or supply, any goods at a price which is unreasonably high, if that price has been in any manner, directly or indirectly, determined, controlled, or influenced by any commercial trust, of which that person or his principal (if any) is or has been a member.” The Russian Criminal Code of 1903 had a similar proviso: “A merchant or manufacturer, who increases the price of victuals or other articles of prime necessity in an extraordinary degree in accord with other merchants or manufacturers dealing in the same articles, shall be punished with imprisonment.”
*47 In all these
rules excessive prices are forbidden, but no attempt is made actually to fix prices by decree. The other, positive, way, in which control may be exercised, consists in the authoritative determination of definite maximum rates of charge or minimum provision of service. This way is illustrated by the terms of the charters usually accorded to companies operating public utility services under lease from city governments and by the power, conferred on the Interstate Commerce Commission by the Hepburn law of 1906, to “determine and prescribe” maximum rates for railway, telephone, and other services of communication.

§ 14. Whether the negative or the positive way of regulation is followed, some sort of sanction to make the law effective must be provided. This can be done in a variety of ways. Sometimes the penalty for breach is a direct money fine. Sometimes, in protected countries, for example in Brazil,
*48 it consists in the withdrawal of duties on competing foreign goods. The Canadian Industrial Combines Investigation Act of 1910 provides for both sorts of penalty. If a statutory Commission “finds that there is a Combine, the Government may either lower or repeal the duties, and, in addition, impose a fine of 1000 dollars a day on those who continue in their evil courses after the judgment of the Board has been officially published.”
*49 Another interesting form of sanction is provided, as against the owners of vessels which violate any of the American anti-trust laws, by a clause in the Panama Canal Act of 1912 forbidding the use of the canal to their ships.
*50 Sometimes, the sanction consists in the threat of governmental competition. Thus, in connection with the 1892 agreement, by which the Post-Office took over the National Telephone Company’s trunk lines, the Chancellor of the Exchequer hinted that the State, while securing its right to compete, would not be likely to exercise that right if the Company acted reasonably.
*51 Sometimes, again, the sanction consists in the threat of State purchase, on terms either fixed beforehand or to be decided by arbitration, of the whole
of the plant of the regulated business. Sometimes, finally,—and this, in effect, is what seems to be contemplated under the authoritative interpretation of the Sherman Act, as given by the United States Supreme Court in the Standard Oil Case (1911)—combinations, whose price (and other) policy is found to be reasonable, may be left undisturbed, but combinations which use their power to the injury of the public may be dissolved by order of the Court.
*52

§ 15. Though, however, many sanctions, some of them of great force, are available when breaches of the law are detected, it is necessary to add that, whether the negative or the positive method of control is adopted, it is exceedingly difficult to prevent people from escaping these sanctions by evasion. Thus in the pre-war period our railway companies, in effect, raised their rates without applying for the consent of the railway commissioners. Charges for rent of sidings and so forth were created; the number of articles which the companies refuse to carry at owner’s risk, unless packed to their satisfaction, was increased; rebates were withdrawn; and other such devices were employed.
*53 But the kind of evasion which it is hardest to deal with is that which meets price regulation by manipulating quality. To prevent this it is essential to couple with rules about maximum price further rules about minimum quality. But in some things, such as the comfort and punctuality of a tramway service, or the sanitary condition of slaughter-houses and sewers, it is difficult to
define a minimum of quality. When there are a number of different grades of quality, all of which have to be distinguished from one another and subjected to a separate maximum price—different grades, even of simple things like tea, and, still more, of complicated things like hats—the difficulty of effective definition is enormous. It is easy to sell a lower-grade thing at a higher-grade price. In other things, such as water supply, gas supply, milk supply and house accommodation, where tests of quality are available to give a basis for definition, it may
be difficult to
detect departures from the stipulated minimum. Something can, no doubt, be done by an elaborate system of inspection, like that developed in support of the Adulteration of Food and Drugs Act, but the openings for evasion are, in any event, likely to be considerable.

§ 16. Even, however, if this difficulty could be completely overcome, the most formidable obstacle in the way of direct control would still remain. It is necessary to determine what prices shall be regarded as unreasonable, and, when the positive method of fixing maxima is adopted, what the maxima shall be. As was explained at the beginning of this chapter, the goal aimed at is the competitive price,
i.e. the price which would have been arrived at if other things had been the same but the output had been that proper to simple competition instead of that proper to monopoly. In what way is this price to be ascertained by the controlling authority? It is conceivable that some reader, thinking loosely upon recent experience, may claim that competitive prices could be determined directly from the recorded expenses of converting the raw material used into finished goods. Plainly, however, in order to get the
full conversion costs, we need to know how much should be added to the cost of material and labour for the share due, for the article we are studying, to the standing charges of the business. Given a decision about that, we can, indeed, by conversion cost accounting—the technique of which was greatly developed during the war—determine the proper price for any particular product or group of joint products;
*54 but to proceed in the reverse direction is impossible. The calculation of conversion costs is a necessary step towards any practical scheme of price regulation. But it is a subordinate step.

§ 17. It seems clear that our problem can only be solved by some reference to a “normal” rate of return on investment. We know that, had the investment proper to competition been made, a price, which, on the then profitable output, would
have yielded a normal return on this investment, would be the price we require. Unfortunately, however, the investment that actually has been made is not likely to be equal to that proper to competition. If the monopoly has been started
ab initio, so to speak, it will be less than that proper to competition, and, if the monopoly is a result of a combination of unduly numerous concerns engaged in cut-throat competition, it will be greater than this. Clearly then the price we require is not the price that, with the output profitable in respect of the actual investment, would yield a normal return, unless it so happens that that price is equal to the price which would have yielded a normal return on the amount of investment proper to competition, had that amount been provided. This condition implies that the commodity with whose production we are concerned is one that, from a long period point of view, is produced under conditions of constant supply price. With a monopoly that has been started
ab initio, if the commodity is produced under conditions of increasing supply price, the price that yields a normal return on actual investment will be too high; if it is produced under conditions of decreasing supply price, too low. In fact, as has already been observed, there is little likelihood of commodities subject to increasing supply price coming under monopolistic control. If then we calculate our “proper price,” for control purposes, by reckoning what price would yield a normal return were the output most profitable at that price produced,
*55 the figure we reach will, with a monopoly that has been started
ab initio, probably be somewhat too high. With a monopoly that is the aftermath of excessive investment and cut-throat competition, it will, on the other hand, probably be too low. There is not, however, so far as I can see, any other way in which a figure can be calculated at all.
*56

§ 18. If then,
faute de mieux, we decide to make use of this way, it becomes necessary to determine what rate of profit, in any particular enterprise, the price of whose product has to be regulated, may rightly be considered normal. At first sight it might be thought that this issue can be settled fairly easily. Will not normal profit be such profit that, when allowance is made for earnings of management (as in joint stock companies is done automatically), what is left provides interest at the ordinary rate on the capital of the concern? This plausible suggestion is, however, easily shown to be very far from adequate. Let us, to begin with, suppose that the ordinary rate of interest really does correspond in all businesses to normal profits. We have still to determine what the capital is on which this ordinary rate is to be paid. Clearly we cannot interpret it as the market value of the concern, because, the market value of a business being simply the present value of its anticipated earnings, these earnings
must yield the ordinary rate of interest on it, allowing for the particular risks involved, whatever sum they amount to. Indeed, if we were to take existing market value as our basis, since this depends on what people believe that the system of rate regulation will be, we should come perilously near to circular reasoning. Capital value, therefore, for rate control purposes, is something quite different from capital value for, say, taxation purposes. It must mean, in some sense, that capital which has in fact been invested in the business in the past. But this is not at all easy to calculate. When the sums of money invested in any concern include commission paid to a promoter for accomplishing a fusion, the advantage of which is expected to consist in the power to exact monopoly charges from the public, this commission ought not, it would seem, to be counted, except in so far as the fusion has also brought about increased productive efficiency. That this is an exceedingly important point becomes apparent when we learn that, according to high officials in some of the industrial combinations of the United States, “the cost of organisation, including the pay of the promoter and financier, amounts often to from 20 to 40 per cent of the total amount of stock issued.”
*57

The same difficulty has to be faced as regards that part of the capital expenditure which has been employed in buying up existing concerns at a price enhanced by the hope that combination will make monopolistic action practicable. Apart from these difficult items, what we want to ascertain is the original capital expenditure, whether employed in physical construction, parliamentary costs, the purchase of patent rights or the upbuilding of a connection by advertisement, together with such later expenditure, in excess of the repairs and renewals required to keep the original capital intact, as has not been taken out again in earnings, allowance being made for the different dates of the various investments.
*58 For new businesses, to be established in the future, it would be easy enough to secure by law that information about all these items should be made available. But for businesses already long established it may be impossible to get this information. For example, similar expenditures on good – will and so forth, which one concern may have charged to capital, another will have treated as current expenditure, in such a way that it cannot practically be distinguished. In view of these difficulties some roundabout way of approximating to the truth may have to be employed. Obviously the nominal capital is quite useless for this purpose. It may have been watered and manipulated in ways that completely disguise the real facts. The market value of the capital we have already shown to be inappropriate. It is usual, therefore, to make use either of the estimated “cost of reproduction” of the concern’s plant—which may be very misleading if the relevant prices have changed substantially since the original investment was made—or of a value ascertained by direct physical valuation of the plant—the amount of which will, of course, depend on the principles in accordance with which the valuation is made—; and then to make some more or less arbitrary allowance for costs of promotion, investments to build up good-will, patent rights, and so on. These
data are not wanted for themselves, but are supposed to enable a rough estimate to be made of the actual capital investment, when this is not directly ascertainable.
To develop the difficulties of this process is outside my present purpose.
*59

§ 19. There remains a more fundamental complication. Up to this point it has been tacitly assumed that the capital invested in any concern is properly and unambiguously represented by the money invested in it. In actual fact, however, a real investment of 1000 days’ labour may be measured by £200 if it happens to be made in one year, and by £400 or, in conceivable circumstances, by £400,000 if it happens to be made in another. In periods during which currencies are violently unstable, as in Russia and Germany after the war, this sort of difficulty inevitably becomes prominent. It is plain that real investment, and not money investment, is the fundamental thing, and that, therefore, in strictness, when general prices have changed, money investments ought, for our present purpose, to be written up or down so as to allow for this fact. This means revising all the records of past years and multiplying the money investment of each year by the ratio between the index number of general prices for that year and for the present time. In view of the acknowledged imperfections of existing index numbers, a device of this kind could hardly hope to win sufficient acceptance to make it practicable. When, as often happens, a substantial part of the investment in a concern has been made in the form of bonds or debentures, on which a fixed money interest is contracted for whatever happens to prices, it is open to a further serious objection. To allow a doubled gross money return to offset a doubling of prices would involve compensating the stockholders for the bondholders’ losses as well as for their own and leaving the bondholders to bear their fate unaided.
*60 None the less, to ignore altogether large and rapid changes in general prices, such as have resulted from the Great War, would be to accept and act upon a serious falsification of the facts. These considerations, taken in conjunction with those set out in the preceding paragraph, suffice to show that to determine what the capital of a concern
is, on which “ordinary” interest is to be allowed, is not an easy task.

§ 20. But this is not all. It is not true that the normal “competitive” profits of any enterprise are the profits that would yield the “ordinary” rate of interest on the capital that has actually been invested in that enterprise. For the establishment of different enterprises involves both different degrees of risk and different initial periods of development, during which no return at all is likely to be obtained; and appropriate compensation under these heads must be made to those investors—the only ones with whom the State can deal—whose enterprises turn out successfully, and who, therefore, must be paid enough to balance the losses of those who have failed.
*61 This circumstance need not, indeed, be responsible for large practical difficulties in industries in which production has attained more or less of a routine character, but in all industries in an experimental stage it is of dominant importance.
*62 Furthermore, even if there were no risks, we could not regard as proper prices which would yield the ordinary rate of interest in all circumstances, but only prices which would yield that rate, if the management, and, indeed, the actual organisation of the original investment also, were conducted with “ordinary” ability; and this is a vague and difficult conception. As Professor Taussig pertinently observes: “Every one knows that fortunes are made in industries strictly competitive, and are to be ascribed to unusual business capacity…. When a monopoly or quasi-monopoly secures high returns, how are we to separate the part attributable to monopoly from the part attributable to excellence in management?”
*63 To allow the same rate of return to companies which invest their
capital wastefully as to those which invest it well plainly makes against economical production. Incidentally, if there were two competing combinations to be dealt with, it would logically require forcing the better managed one to charge lower prices than the other, an arrangement which would not only have awkward consequences at the moment but would effectively discourage good management. In this connection it should be noted that to extend combination further, so long as extension involves economies, is a form of good management, and a form that would be discouraged if prices were so regulated that no advantage were allowed to accrue to those who had brought it about. Finally, when a plant has been built to fit an expected future demand much in excess of the present demand, it would plainly be unreasonable to sanction rates high enough to yield a full return on the whole investment before that future demand has developed.
*64 In view of these complications, and of the necessary limitations of its knowledge—for, as a rule, the controllers are bound to be much behind the controlled in technical experience—a public authority is almost certain either to exact too easy terms from the concerns it is seeking to control, and so to leave them with the power of simple monopoly, or to exact too hard terms, and so, though not permitting monopoly exaction to them, nevertheless to prevent the development of their industry to the point proper to simple competition. The British Tramways Act of 1870 appears to have failed in the latter way, and to have been responsible for prolonged delay in the development of electric traction in this country.

§ 21. It is evident that the difficulties, which are involved in determining what scale of return should be regarded as normal in any particular productive enterprise, complicate alike the negative way of control, under which the Legislature simply condemns unreasonable prices, leaving the Courts to decide whether any given price scheme is in fact unreasonable, and the positive way, which lays down definite price maxima. Plainly, however, they complicate the positive way more seriously than the other. An ordinary industrial concern produces a great number of different varieties of goods, the raw materials
for which are continually altering in cost, and the distinctive character of the finished product continually being modified. For any outside authority to draw up a schedule of permitted charges for a concern of this sort would be a hopeless task. On the other hand, for a trained Commission or judicial body equipped, like the American Federal Trade Commission, to make full inquiries, it would not be impossible to decide in a broad way whether, taking one product with another and one time with another, some selected large combination—the Standard Oil Company, the United Steel Corporation, or another—was charging prices calculated to yield to it more than the return deemed in the circumstances to be reasonable. For industrial concerns in normal times no attempt has ever been made to go beyond this negative way, and it does not seem, at all events in the present condition of economic knowledge and governmental competence, that any such attempt either can, or should, be made. Imperfect as the results to be hoped for from the negative way are, they are better than would be got from a blundering struggle after the other. In public utility concerns, on the other hand, the excess difficulty of the positive over the negative way is slight. As a general rule, the service provided by these concerns is single and relatively simple—gas, water, electricity, transport of passengers. Not many separate prices—railway freight rates are, of course, a very important exception—have, therefore, to be fixed. Further, the demand is generally unaffected by fashion, and equipment plays so large a part in the cost that changes in the price of raw material do not very greatly matter. Finally, even if these things were otherwise, the nature of the goods sold and the convenience of customers make it very desirable that the prices charged should not undergo frequent change. In these concerns, therefore, the positive way of control by fixing maximum prices has generally been adopted.

§ 22. When this is done, it becomes imperative to seek out the best means of guarding against the two opposite sorts of error, undue laxness and undue harshness, to which, as was shown in § 20, all forms of regulation are in practice
inclined. For this purpose one device sometimes recommended is to put up the licence to operate certain public utility services to a kind of auction. This plan allows the persons most interested themselves to present estimates of terms which they would reckon profitable. It has been described thus: “According to the best plan now in vogue, the City sells the franchise for constructing the works to the company, which bids to furnish water at the lowest rates under definitely specified conditions, the franchise being sometimes perpetual, but often granting to the City at some future date an option for the purchase of the works.” Since, however, in many cities, the companies capable of making tenders will be very few in number, and since, furthermore, their own estimates must be largely tentative, the adoption of this device is not incompatible with large errors. The likelihood of error is made greater by the fact that the conditions of most industries are continually changing, in such wise that the scheme of price-regulation, which is proper at one time, necessarily becomes improper at another.

§ 23. A further effort at limiting the range of error can be made through arrangements under which the regulations imposed are submitted to periodical revision. Franchises “cannot be fixed, or justly fixed, for all time, owing to rapidly changing conditions.”
*65 With the growth of improvements and so on, it may well happen that a maximum price designed to imitate competitive conditions will, after a while, stand above the price that an unrestricted monopolist would find it profitable to charge, and will, therefore, be altogether ineffective. “The public should retain in all cases an interest in the growth and profits of the future.”
*66 A provision for periodic revision in a franchise may, however, by creating uncertainty, restrict investment in the industry concerned to an extent that is injurious to the national dividend. Further, if the revision is to occur at fixed intervals, it may tempt companies, shortly before the close of one of these intervals, to hold back important developments till after the revision has taken place, lest a large part of their fruits
should be taken away in the form of lowered prices;
*67 and this difficulty cannot be wholly overcome by clauses stipulating for the introduction of such technical improvements as are, from time to time, invented elsewhere. One way of meeting these dangers is to hedge round the revising body with conditions designed to defend the company’s interest. For example, the Railway Act of 1844 provided that, if dividends exceeded 10 per cent on the paid-up capital after twenty-one years from the sanctioning of the lines, the Lords of the Treasury might revise tolls, fares, etc., on the condition that they guaranteed a 10 per cent dividend for the next twenty-one years. Another way is to make the revision period so far distant from the date at which an undertaking is initiated that the effect upon investment due to the anticipation of it will be very small. It is evident that, just in so far as either of these lines of defence is adopted, the efficacy of revision, as a means of lessening the gap between actual regulation and ideal regulation, is diminished. But, if regulation is to be attempted at all, the retention by the State of revising powers in some form is absolutely essential. It would seem that this could be provided for without imposing a serious check either on investment or on enterprise, if the principles on which the revision would proceed were clearly laid down and understood. The revisers might be instructed at each revision period to fix a price—or, when they have to do with joint products, several adjusted prices—sufficiently high to continue to the company a fair rate on their total real investment, account being taken of the fact that the capital turned into it in the first instance was probably subject to great risk, while that added subsequently needs a less reward under this head. They might be instructed further, in deciding what constitutes a fair return, to consider generally the quality of management that has been displayed, fixing prices to yield higher returns when the management has been good than when it has been indifferent or bad. No doubt, the technical difficulty of this kind of revision would be exceedingly great, but it
would not be nearly so great as that of the initial regulation. It is not unreasonable to suppose that a class of official might eventually be evolved whose decision on such matters, when founded on adequate comparative statistics, would at once deserve and command the confidence of would-be investors. Such investors would have the consolation of knowing that, while, on the one hand, the price of their product was liable to enforced reduction, on the other hand, if costs of material and labour went against them, it might be raised in their favour.

§ 24. Yet another device designed to limit the range of error remains. In all ordinary industries many variations in the costs of material and so forth occur
within the successive revision periods. If the guidance of simple competition is to be followed, such variations should be accompanied by variations in the prices charged to consumers. No doubt, where, as in railway service, the technical inconvenience of constantly changing prices would be very great, it may, on the whole, be best not to follow this guidance for short-period movements; but such cases are probably rare. Attempts are sometimes made by controlling authorities to organise the required price variations by some sort of self-adjusting arrangement. A crude method, which has been applied to some gas companies in this country, is to fix a
maximum dividend. If the competence of the management remains constant, this implies that, when costs fall beyond a certain point, the prices charged to consumers must also fall. This method, however, has the grave disadvantage that it is likely, so soon as it becomes operative, to cut away the normal motives for skill and care in management and for avoidance of waste. A less crude method is to lay down a standard of earnings, always to allow prices to be charged high enough to yield this standard, and to provide further that, when this standard is passed, a defined proportion of the balance shall be used to reduce prices, while the remainder is left to augment earnings. In the South Metropolitan Gas Company’s Act of 1920 there is an arrangement of this kind, three-quarters of whatever balance there may be over standard earnings being allotted to consumers. Railway rates in this country are now
governed on a somewhat similar plan. The Railways Act, 1921, lays down for each amalgamated company a
standard income, based on the earnings of 1913 with allowances for new investments, and so forth. If experience shows that the rates of charge fixed by the Rates Tribunal yield, or could, with efficient and economical management, have yielded, an income greater than the standard income, the Rates Tribunal are instructed to reduce the rates of charge “so as to effect the reduction of the net income of the company in subsequent years to an extent equivalent to 80 per cent of such excess”; and, if the net income actually yielded turns out to be less than the standard income together with appropriate allowances for new capital, the tribunal shall increase the rates of charge to bring it up to this amount, provided that the deficiency is not due to lack of efficiency or economy in the management. Plainly, under this arrangement the discouragement to skill and economy will be smaller than it would be under a maximum dividend plan
after the maximum has been reached; but, in view of the extreme difficulty that any tribunal must find in adjudicating as to efficiency and economy in management, it is bound to be greater than it would be under that plan
before the maximum has been reached. Yet a third method is that of connecting changes in the dividend paid to shareholders during any licence period with changes in selling price by means of a sliding scale. Illustrations of this method are furnished in a number of English Acts of Parliament dealing with gas companies. One pre-war Act, for example, fixed a standard price of 3s. 9d. per thousand cubic feet, and provided that, for every penny put on or off that price, the company might, when there are reductions, and must, when there were increases, move the dividend up or down a quarter per cent. Another illustration is furnished by the Act governing the Lancashire Power Company, which furnishes electricity in bulk. This Act “provides for a dividend of 8 per cent and an additional 1.25 per cent reduction in price for every 0.25 per cent increase of dividend above 8 per cent, in respect of every 5 per cent charged below the maximum price allowed by the Act.”
*68 Sliding scales of this kind—which,
if they are to be effective, must, of course, be combined with Government control over the issue of new capital by the companies concerned—are, like sliding scales of wages, not substitutes for, but complements to, a system of periodic revision of the licence terms; for, if they were treated as permanent arrangements, all improvements and discoveries that reduced cost of production, whether made by the concerns themselves or by others, would steadily and continuously enhance profits. They are not easily organised for new companies, because the appropriate standards of price and dividend cannot be determined till some experience has been gained of the working of a concern. But it is feasible, and before the war it was the practice of the Board of Trade in dealing with gas companies, to fix a simple maximum price at first and to reserve power to substitute a sliding scale after the lapse of a certain interval.
*69 These scale arrangements, like the other methods discussed above, are open to the objection that they push prices up, not only when the costs of raw materials and labour rise, but equally when the profits of a company are reduced by incompetent management. In spite, however, of these difficulties, sliding scales—and the same thing may be said of the standard earnings plan—may be expected, when carefully constructed, to make possible a nearer approach to the system of prices proper to simple competition than would be possible under any plan that, over the intervals between revision periods, fixed prices rigidly. Moreover, the danger of discouraging competent management may be met to some extent by provisions, such as those embodied in the British railway law, under which the controlling authority is allowed to veto an increase of charges when falling earnings appear to be the result, not of natural causes, but of incompetence.

§ 25. It should be added that arrangements of the kind I have been describing, though they may make fair provision for adjusting charges within the revision periods to variations in the cost of raw materials, and so forth, are extremely ill-fitted to cope with variations in demand; for, whereas, if simple
competition is to be followed, upward movements of demand—we are here, of course, only concerned with short-period fluctuations—should be associated with upward movements of price, under these arrangements they will be associated with downward movements of price. Demand variations, moreover, may be very important, and may call for large associated price variations. With given demand variations, the extent of these should be especially great in industries where the part played by supplementary costs—which are not reduced proportionately when output is reduced—is large relatively to the part played by prime costs: and supplementary costs are, in fact, very important in the generality of industries. It might be possible to take account of variations in demand, as well as of variations in cost, by a scale system that should link up permitted changes in price with changes in the volume of service supplied, instead of with changes in earnings. So far as I am aware, however, no self-acting system of this kind is as yet anywhere in operation.

§ 26. There remains another difficulty of a different order. The main part of what has been said so far has tacitly assumed that, in framing our control policy, we start with a clear table. For industrial monopolies that come into being after the general lines of our policy have been fixed, and for public utility corporations upon which conditions are imposed at the time when the original franchise is granted, this is, of course, true. But, in so far as we have to do with monopolistic concerns over which at present control is either not exercised at all or exercised in a very imperfect manner, the case is different. To bring these concerns now under a system of price regulation of the type that we are contemplating would, in many instances, involve a large reduction in their income and in the capital value of their shares. So far as original shareholders or persons who have inherited from them are concerned, this does not greatly matter. The fact that these persons have made abnormal profits in the past is no reason why they should be allowed to do so in the future. But the position is different with recent purchasers of shares, whose purchase price has been regulated by the conditions ruling before control, or the strict form of control here contemplated,
was seriously thought about. Such persons may perhaps be getting now, say 8 per cent on their money, and control may knock this down to 5 per cent, reducing the value of their capital by one-third or even one-half. To make regulations that will strike with cruel severity on arbitrarily selected groups of perfectly innocent persons is not a thing to be lightly undertaken. There are limits to the right of the State to ride rough-shod over legitimate expectations. And yet to refrain from control that ought to be imposed, because we neglected our duty in not imposing it before, is to enslave ourselves to past mistakes. Surrender to the “widows and orphans” argument means, in substance, abandonment of reform. No perfect solution of this conflict can be hoped for. But it would seem a reasonable compromise, and one adequately careful of vested interests, to provide that, when the sudden introduction of a full measure of price control on the principles indicated above would greatly depress values, this control should only be introduced after an interval of notice, and then by gradual steps.

§ 27. Even, however, if this somewhat special difficulty be left out of account, the preceding general review makes it evident that, under any form of State control over private monopoly—and it should be noticed that, though the examples cited have to do only with special kinds of private monopoly, the argument refers to all kinds—a considerable gap between the ideal and the actual is likely to remain. The method of control, whether positive or negative, is, in short, an exceedingly imperfect means of approximating industry towards the price level and output proper to simple competition. Moreover, it is apt to prove a costly method. As Professor Durand observes: “Government regulation of prices and profits of private concerns always involves a large element of waste, of duplication of energy and cost. It means that two sets of persons are concerning themselves with the same work. The managers and employees of the corporation must study cost accounting and conditions of demand in determining price policy. The officers and employees of the Government must follow and do it all over again. Moreover, the fact that these two sets of persons have different motives in approaching
their work means friction and litigation, and these spell further expense. To superimpose a vast governmental machinery upon the vast machinery of private business is an extravagance, which should be avoided if it is possible to do so.”
*70 This consideration is one that ought not to be ignored. The expense involved in public supervision should be debited against the system of private enterprise in monopolistic industries before the real efficiency of this system is brought into comparison with the rival system of public enterprise.