The Economics of Welfare

Part II, Chapter XXII

PUBLIC OPERATION OF INDUSTRIES

§ 2. It must be clearly understood that the issue here raised concerns public operation, not public ownership.
Public ownership by itself, and apart from any distributional change that may have come about if the ownership has been acquired without the payment of full compensation, means very little. Suppose, for example, that a municipality raises a loan of a million pounds in order to establish an electric supply works, interest to be paid at 5 per cent and the principal to be paid off in fifty years through a sinking fund. The legal position will be that the municipality owns the works from the moment they are built, subject to what is, in effect, a mortgage to the fundholders. If a private syndicate had put up money, built the works, and loaned them to the municipality on terms involving exactly the same charges to the municipality and providing for the works to pass into its possession after fifty years, the syndicate would, during those fifty years, be the owner. But, granted that the municipality was free to do what it chose with the works, whether by altering them or adding to them, the real position would be exactly the same on this as on the other plan. The distinction between public ownership and private ownership would be a mere technicality of no practical effect. In like manner, if a public authority lends a million pounds to a private concern at 5 per cent perpetual interest to enable it to build an electric works, the real position is exactly the same as if it built the works itself and let them to a private concern at a perpetual interest of equal amount: but under the former plan the private concern, and under the latter the public body, is technically owner. There is a difference in form, but identity in substance. Between public operation and private operation, on the other hand, there is always and necessarily a fundamental difference of substance.

§ 3. In view of the many technical difficulties, to which attention has been called in the preceding chapter, in the way of the effective exercise of public control over private industry, the case for public operation, at all events in industries with a tendency towards monopoly, is, from the point of view of a right distribution of national resources among different occupations, a very strong one. It remains very strong in spite of the fact that, as is alleged to have happened with Government railways in certain democratically governed States,
it may be perverted to satisfy local and sectional, or even personal, ends;
*71 for this danger has been substantially lessened by the invention of extra-parliamentary “commissions,” as described in Chapter XX., for working public enterprises. But the comparative effect of public control and public operation upon the right distribution of national resources among different occupations is not the only thing we have to consider in making our choice between them. Other things besides this are involved, just as other things were involved in our comparison between voluntary Purchasers’ Associations and ordinary commercial businesses. We are not entitled to assume without argument that the economies of production will be the same under public operation and private operation. It may be that public operation is less economical than private operation, even when private firms are subject to public control. If this is so, the disadvantages of public operation as regards economical production have to be balanced against its advantages as regards the distribution of resources among different occupations. Hence, before any real answer to our question can be attempted, it is necessary to undertake some comparison of public with private operation from the standpoint of productive efficiency.

§ 4. It will be well at the outset to clear out of the way two arguments drawn from the experience of the war, which are based on a loose use of the term efficiency and are not really relevant.

First, it has been argued as follows: “If the individualist principle is the right thing, then it was manifestly absurd in war time to do what the Government did, for example, in taking over the railways. If divided railway control was efficient, why interfere with it; why not carry on as usual? What was there in the way of moving trains and men that was not the proper business of railway companies, and why, then, were they ‘interfered with’? If it becomes obviously necessary to mobilise railways in war to move some hundreds of thousands or millions of men, why is it not necessary to mobilise railways in peace to move to the best advantage nearly three hundred million tons of coal in a year—the
coal which is the very life-blood of British industry?”
*72 This reasoning assumes that the State took over the railways in war time in order to render them technically more efficient. In fact it took them over in order to ensure that the Government should have full command over their lines and equipment, and should not have to do without services it needed on account of conflicting claims from private persons. Normally railways, like all other concerns that sell their output for money, allocate that output in accordance with the effective monetary demand of their various customers. In war time it was obviously necessary to scrap effective monetary demand as the directing factor in the distribution of railway services among rival customers. The fact that this was done with general agreement is no proof that any one considered railways to be technically less efficient,
i.e. to require a greater real cost to obtain a given result, under private than under public management.

Secondly, an analogous argument has been built up to prove that the establishment of national munition factories enabled the Government to obtain its supplies enormously more cheaply than it could have done, and was in fact doing, from private firms. But the circumstances of the war were such that the private sellers of munitions, faced with an unlimited Government demand, were able to exact prices very greatly in excess of their own cost of production. Such a state of affairs does, indeed, provide a strong argument for national action, but the fact that a national shell-works can produce shells at a less cost than the price that a private works can force the Government to pay is no proof that it is technically more efficient. Technical efficiency concerns real costs of production, not sale prices fixed under conditions of shortage or conditions of monopoly. I do not here raise the question whether, in fact, cost of production was less in Government than in private shell factories. Whether it was so or not, it certainly cannot be proved to have been so by a comparison of the costs of production in Government factories with the selling price of private factories. This argument, like the preceding, therefore, falls to the ground.

§ 5. Another negative proposition of a general character may be set down. This is that attempts to conduct such a comparison by reference to statistics are foredoomed to failure. No doubt, if it could be shown that,
other conditions being the same, a given output was, in general, obtained at greater, or at less, real cost under public than under private management, genuine evidence about the relative efficiency of the two forms of organisation would be obtained. But in real life this is impracticable. In the first place, the quality of services, which are called by the same name, varies enormously in different places, and it is almost impossible properly to allow for these variations. “Our street cars,” say the Reporters of the American Civic Federation, “run faster, carry more strap passengers, and kill and injure more people than the street cars, public or private, of any other country. Our people seem to like this, but the English would not.”
*73 How can differences of this sort possibly be taken into account? Again, the conditions of production in different places are utterly different. “In Syracuse (U.S.A.) the water flows to the city by gravity; in Indianapolis it must be pumped.”
*74 “To compare a private corporation within the limits of a great city, where an immense supply is furnished, and where special conditions of non-interference with adjoining property rights are to be met, with some municipal plant in a suburban town, upon a basis of the relative amount of supplies and labour required per unit of electrical energy, would obviously be unfair to both contestants. Nor is it possible to compare in this manner two lighting-stations having approximately the same yearly output, and which are similarly located with reference to adjoining interests, but are situated, the one in the North and the other in the South, for the reason that the daily period of service will vary in these two localities on account of variation in the hours of darkness. For the same reason we cannot compare the summer service of one station with the winter service of another, even though we should attempt to reduce them both to a common basis by obtaining the amount of human effort employed per unit of electrical
energy.”
*75 In short, arguments from statistics, even apart from the pitfalls with which unwary inquirers are confronted in the interpretation of municipal accounts,
*76 are, in this field, almost entirely valueless. This remark is of general application. But, in view of the exceptional psychological conditions of war time and the temporary use by government of a large number of able men normally engaged in private business, as well as of the fact that the commodities produced in government factories during the war were for State use and not for the market, it has very special relevance to arguments drawn from the experiences of the war period.

§ 6. Statistical evidence being thus inadequate, it is necessary to proceed—again as in our study of voluntary Purchasers’ Associations—by way of general considerations. Let us begin by comparing public operation with
uncontrolled private operation. There is general agreement that, when conditions are such as to allow of small scale production by private businesses, the personal interest of the head of the business in its success provides a stimulus to efficiency that is lacking in both joint stock private concerns and in public concerns. Over a large field of industry, however, the practical choice is, not between private businesses and public concerns, but between joint stock companies and public concerns. Here the initiative, freedom and interest of the captain of industry working his own comparatively small business cannot be had in any event. The issue is a different and more evenly balanced one. The discussion of it may well be started with an observation of the Committee of the American Civic Federation: “There are no particular reasons why the financial results from private or public operation should be different if the conditions are the same.”
*77 The reason for this remark, of course, is that, whether a service is provided by a private company or by a public governmental authority, the actual running of the business must be similar. An expert staff must be appointed, controlled in a general way, in the one case by a committee of
directors chosen by the shareholders, in the other case by a committee, a commission, a council,
a ministerial department, or an
ad hoc body like the Port of London Authority, to represent the public. Managing power, as a whole, may be conceived as distributed among electors, directors—or committee, or whatever the controlling authority is—and staff. There seem no general
a priori grounds for holding, without reference to the special nature of the controlling organisations evolved under them, that either public or private management is likely to prove technically the more efficient.

§ 7. In some matters of slight, but not negligible, importance experience suggests that the public authority has an advantage. This advantage is analogous to one found in productive co-operation. It is that, for a given sum of money, a more efficient engineer or manager can be obtained than will be forthcoming under private management, for the reason that the position of a public servant is at once attractive in itself and also makes appeal to altruistic motives; or, alternatively, that an engineer or manager of given efficiency can be obtained for a smaller sum of money. This advantage, it must be clearly understood, is a real advantage, and not a kind of bounty obtained at the expense of the engineer or manager; for there is created a new value in the extra satisfaction which the said engineer or manager derives from the fact of serving the public. The difference between what a man of given ability would have been willing to work for in a private company and what he does work for in a State department is, in effect, extra product due to the adoption of the public form of industrial organisation. This difference is not, of course, equivalent to the difference between the earnings of the head of a State department and those of the head of a private business, because in the earnings of the latter there is generally included a return for “waiting” and “uncertainty-bearing,”—services which in the public departments are provided by the tax-payer. It is fallacious to take the excess of the income of an American railway king over that of the administrator of the Prussian State railways in pre-war days as a measure of the comparative wastefulness of private enterprise. Still there is,
pro tanto, an advantage on the side
of public operation in the fact that good technical experts under it cost less.

§ 8. A more important matter is the business capacity of the authority above the technical managers which determines general policy. In municipal undertakings this authority is generally a committee of the town council—a body whose members are elected for their political, rather than for their commercial, qualifications, and are also more liable than the directors of a company to lose their seats at short intervals. There is the further difficulty that the employees of a municipal enterprise may play an important part in electing councillors. This may lead some councillors to interfere for political reasons with the disciplinary and other discretionary powers of the higher officials. It has even been suggested that in some towns city engineers have been hindered by the council from introducing labour-saving devices, by which the employment of some of the councillors’ electors would be threatened.
*78

In national undertakings run by a government department the higher authority is a body of civil servants technically subordinate to a political head responsible to Parliament. Through this political head, pressure of various sorts, some of it probably anti-social in character, can be exercised on the running of the undertaking. But, even if this does not happen, the civil service organisation, no doubt excellent for the purposes for which it is primarily designed, is apt to cramp efficiency. When important decisions have to be taken there is a tradition of method in government offices that makes for delay, hesitancy and immobility.
*79 Thus Mr. Justice
Sankey, in his report on the coal mining industry, speaks of “the present Civil Service system of selection and promotion by length of service, of grades of servants, of minuting copies and reports from one servant to another, and of salaries and pensions.” In pure routine work this system may do no harm, but, when enterprise and quick decisions are necessary, it cannot fail to prove hampering both in the choice of the most suitable man for a given work and in the actual carrying on of work. It should be noted that this consideration is very much less important in concerns producing exclusively, as, in war time especially, certain concerns do, for government; for work of this kind must be done on order, and there is no scope for that forecasting of the market in which government departments are commonly supposed to be inferior to private businesses or joint stock companies. Thus Professor Lehfeldt has well observed: “Anyone—government, company or individual—who can take over the whole output of a factory may well be justified in setting one up for himself; but the ordinary factory has to sell its products and to find customers: that is quite a different matter.”
*80 In this connection the early history of telegraphic communication is interesting. The semaphore system of optical telegraphy invented at the end of the eighteenth century was taken up by the French Government for military use, confined to that use, and worked by the government exclusively. In 1845 the French Government, in a like spirit and for a like purpose, started an electric telegraph. “The public authorities felt a need of their own, and, finding no one else to supply it for them, set to work to supply it for themselves…. Public ownership of telegraphs in the beginning was not, strictly speaking, a manifestation of the spirit of business enterprise. It was simply a branch of the public administration, forced upon the government by the lack of private enterprise.”
*81 When an industry is chiefly engaged in producing things for the consumption of the general public, the call for prevision and constructive speculation is, of course, much greater, and, therefore, the defects of civil service methods correspondingly more marked.

It is coming, however, more and more to be realised that the public operation of national undertakings does not necessarily imply that these undertakings are run by a government department organised on civil service principles. The Port of London Authority is a special authority working quite differently from a government department. Canada has created the Canadian National Railways Company, with the government as the only shareholder, but with business directors appointed like ordinary directors and endowed with complete freedom of management.
*82 The proposals of the Sankey Commission also aimed at setting up an authority for coal mining, which, though national, should nevertheless be run in the main on non-bureaucratic and non-political lines. Consumers’ representatives were to be associated with the management, as is already done in the national telephone service of Switzerland.
*83 A comparison of a body of this type with the directorate of a joint stock company might well work out more favourably than a comparison of say, the body ruling the Post Office with such a directorate. Plainly, it is impossible to generalise on this matter of business competence, apart from a knowledge of the detailed organisation under which it is proposed that particular public undertakings should be run.

§ 9. So far governmental operation has been set against
uncontrolled private or joint stock operation. In practice, however, as has already been made clear, where public operation is a live issue, the alternative is
controlled private or joint stock operation. Control must hamper that initiative which is the chief merit of private enterprise, and the extent to which it hampers it will be greater the more far-reaching is the control. If it goes so far as to settle the things to be produced and the method of production, it will hamper initiative greatly. If, on the other hand, it does not extend beyond fixing a maximum price with a liberal margin, or even
fixing a sliding scale of profits and prices in combination, as under some gas company charters, it will, of course, hamper initiative much less. It is, thus, not possible to compare the technical efficiency of government operation and of controlled private operation in general terms, because controlled private operation may mean any one of a great number of different things—just, indeed, as, on the showing of preceding sections, government operation itself may also do. The only broad inference to which we are entitled is that, as between government operation and controlled private operation, a comparison of technical efficiency is likely to be somewhat more favourable to government operation than it would be as against uncontrolled private operation.

§ 10. This somewhat impotent conclusion does not, however, exhaust the discussion. There remain three important groups of considerations which, when we look beyond mere technical competence, tend to suggest that public operation is likely—not, of course, always, but as a general rule—to be inferior, from the standpoint of the dividend, to public control. The first of these has to do with the fact that, not only different producers within the same industry, but also different producers in apparently disconnected industries, are often, in reality, rivals. No doubt, an industry can be imagined which is monopolistic in the widest possible sense, in such wise that not only are there no competing firms within it, but also there are no competing industries outside it. There is some reason to believe that the service of supplying a modern city with water is monopolistic in this sense. It would be possible, by combining together a number of industries that are now separate, to create other monopolies of the same sort. For example, the various means of communication, such as omnibuses, trams, motor cars and carriages, might all conceivably be brought together under one hand. The same thing might conceivably be done with all the means of providing artificial light or all the means of providing power. But such arrangements are quite out of relation to actual facts. As things are at present, I should doubt if any industry, except that of water supply, can properly be regarded as monopolistic in the wide sense here taken. Now, the
interest of the national dividend requires that, where a number of establishments, whether in the same industry or in different industries, are competing for the supply of some public need, that one which can supply it most efficiently shall oust the others. But, when any enterprise is operated by a public authority, it is likely to be maintained by artificial support, even though it is less efficient than its rivals. The reason is that persons in control of such an enterprise, being naturally anxious to make that enterprise a success, tend to identify the good of the whole with the good of their own department. Hence a government authority embarked on a business is almost certain, if it prove commercially weak, to employ unfair weapons from its non-commercial armoury, the use of which will maintain it more or less permanently in being, despite the fact that its productive methods are more costly than those of its rivals. These unfair methods are of two sorts, according as they are directed primarily to defend the government enterprise or to obstruct its competitors.

Defensive non-commercial methods consist, in the main, in the conscious or unconscious practice of devices for securing a differential bounty from the general public. A government authority, which is engaged partly in business and partly in rendering general unremunerated services, may charge expenses that really belong to the business against the other part of its work. A very glaring example is the practice of the London County Council in writing down the value of land purchased for workmen’s dwellings to the value which it has, not in the general market, but as ear-marked for this particular purpose. Again, municipal tramway accounts may be given a false appearance of prosperity by the device of charging expenditure upon roads, which is properly attributable to them, to the general road account.
*84 A like device is adopted in a milder form when a municipality fails to set aside a special fund to balance the advantage it possesses over private enterprise in being able to borrow money on easier terms. “A municipality can float bonds at a lower rate of interest than a private company, since the whole assessable
property of the town is generally liable for the payment of interest and principal, while the company can give security only on the works.”
*85 This ability on the part of a municipality is thus due, in the main, simply to the fact that it is able to force upon the ratepayers an obligation to pay its bondholders even if the enterprise fails, while a private company has, by the offer of higher pay, to obtain debenture holders who are prepared, in the event of failure, to lose their money. Except in so far as the fear of failure, and, therefore, the extra compensation asked for by debenture holders, is due to public ignorance of facts which are more readily ascertainable in connection with municipalities than with companies
*86—to that extent municipal operation effects a small real saving—the social cost of the municipality’s cheap loan is the same as that of the company’s relatively dear loan. If the two enterprises are to compete fairly, the municipality ought to transfer to the rates the bulk of its gain from better credit, before balancing the accounts of its business. If it does not do this, it is, in effect, assisting that business by a contribution from the general public. In so far as the lower terms on which it can engage managers and engineers are due to the fact that the shouldering of risks by the ratepayers safeguards them against the possibility of their employers going bankrupt, it is doing the same thing a second time, unless it transfers the gain made under this head also to the rates. Of course, if a municipality managed undertaking,
on account of superior efficiency, is less likely to make a loss than a corresponding private concern, there is a real gain. But, since, in any event, there is the gilt-edged guarantee of the ratepayers, that gain is not reflected in the better terms on which the municipality can borrow.

Aggressive non-commercial methods are made possible by the fact that public authorities, besides operating their own
enterprises, are often also endowed with powers of control over other enterprises. When they are in this position, there is a grave danger that the public authorities may be tempted to use their powers of control in such a way as to obstruct and injure rivals. An Education Authority, for example, which both runs schools of its own and makes regulations for the running of rival schools, is under strong temptation. So is an authority which at once builds houses and frames building bye-laws; and so also are municipalities operating gas-lighting or tramways and controlling electric-lighting or motor omnibuses. Among the methods of aggression open to them perhaps the simplest is that of making the conditions about sinking funds, under which their own establishments work, more favourable than the conditions about purchase at the end of the lease, which are imposed upon private companies. A public authority, which provides a sinking fund to extinguish the capital debt of its enterprise, as well as a fund to cover depreciation and obsolescence, is, in effect, taxing its present citizens for the benefit of posterity.
*87 In like manner, a public authority, which confers a franchise on a private company upon condition that the company’s plant shall pass to itself at the end of the lease, either free of charge or at “cost of replacement,” is imposing a similar tax. It is readily seen that the terms of sinking funds and franchises respectively
can be so arranged that the burden under the sinking fund is the smaller, and, therefore, that private operation suffers, as against the rival system, a differential injury.

There are, however, grosser forms of aggression than the above. It is notorious that those municipalities which operated their own gas-plant vigorously obstructed, by the exercise of their veto and in other ways, the development of electric-lighting companies. Again: “Since 1898 the desire to protect the local municipal electric light plants has been permitted to impede the spread of the so-called electricity-in-bulk generating and distributing companies.”
*88 In like manner,
the central government, in order to protect its telegraph monopoly, has placed administrative obstacles in the way of other means of electrical communication. In 1884 the Postmaster-General declined to allow the National Telephone Company to receive or deliver a written message at any of its offices, and, in defending this course, said: “It would make, I am afraid, a serious hole in the telegraph revenue, if written messages were allowed to be sent.”
*89 In like manner, in Norway, when (in 1881) a company sought a licence to establish a long-distance telephone between Drammen and Christiania, the Government made it a condition that the company should guarantee to make good “all losses occasioned to its (the Government’s) telegraph lines between the two cities”; and similar compensations were required of other telephone promoters.
*90 Finally, in the charter granted to the Marconi Wireless Company in 1906, permitting the transmission of wireless messages between the United Kingdom and North America, it was specially provided that permission would not be granted for messages to or from any European country except Italy, the purpose being to safeguard the interests of the cables owned by the British and Continental governments.
*91

The use of defensive and aggressive weapons of an “unfair” uncommercial character by public authorities operating enterprises brings it about, as already explained, that an enterprise run by them is often maintained in existence, despite the fact that the end served by it would be served more cheaply by a rival enterprise. It is necessary to note, in conclusion, that the use of these methods tends to extrude economically superior rivals even more effectively than it appears to do at first sight. For it acts, not only directly, but also indirectly through anticipation. It not only drives out of the market existing competitors, but checks the entry of new ones. When a man contemplating a philanthropic enterprise is given to understand that, should his experiment succeed, a public authority will enter the field he has proved fruitful, he
does—or should—rejoice. But, when a man engaged in a business enterprise is given to understand this, the end he is pursuing is not, like the philanthropist’s, furthered. It is, on the contrary, thwarted, and his energies are, therefore, diverted from the undertaking. An effect of this kind is claimed to have resulted from municipal experiments in house-building. These considerations, when they have relevance, evidently strengthen the probability that the operation of industries by public authorities will be injurious to productive efficiency; and they are bound to have some degree of relevance except in industries that are monopolistic in the widest possible sense.

§ 11. I pass to a second consideration. This has to do with the fact that the working of any industrial enterprise involves some degree of uncertainty. As will be explained at length in Appendix I., the exposing of money to uncertainty is a definite factor of production, which makes output larger than it would become without it. In the long run willingness to expose £100 to an equal chance of becoming £160 or of becoming £50 is bound to increase the national dividend. If willingness to expose money to uncertainty on the part of people in control of industry is “artificially” restricted, enterprise and adventure that make for industrial progress, and, therewith, for production, will be hampered. Furthermore, the injury thus wrought is very much larger than appears at first sight. For, since any experiment with an untried process
may fail, a diminished willingness to expose money to uncertainty implies a restriction of experiment, and, hence, a diminution in the inducement to enterprising persons to make useful inventions. No doubt, there is reason to believe that, with the growing dependence of industry upon non-commercial science, this consideration has become less important than it used to be. Dr. Mertz has well observed: “The great inventions of former ages were made in countries whose practical life, industry and commerce were most advanced; but the great inventions of the last fifty years in chemistry, electricity and the science of heat have been made in the scientific laboratory; the former were stimulated by practical wants; the latter themselves produced new functional requirements, and created new spheres
of labour, industry and commerce.”
*92 It still remains true, however, that, though the fundamental discoveries are often non-commercial, yet the application of them through “inventions,” in the earlier stages before the inventions have been proved by experience, generally requires a commercial stimulus. Anything which restricts unduly willingness to make ventures in any industry must still, therefore, threaten heavy loss. The point I have to urge is that a public body engaged in industrial operations is
likely to restrict unduly this willingness.

The defence of this proposition rests on the following reasons. First, public authorities recognise that hostility to government on the part of the people is an evil, and they also recognise that an unsuccessful State speculation, “if it involves repudiation or oppressive taxation for years to come, produces a popular revulsion and deep-seated distrust of government itself in all its branches.” Secondly, the persons at any time in control of a public authority, when that authority is dependent on the party system, cannot but know that “failure would give their political opponents too good an opportunity to ride into power.”
*93 Thirdly, these persons are partly able to perceive that, if people are
compelled to expose resources to uncertainty in proportion to the rateable value of their houses, more real sacrifice will be involved than if the same aggregate of resources to be exposed to the same scheme of uncertainty were obtained, by way of voluntary contributions, in proportion to the attractive force exercised upon the several contributors by the prospective profits. Finally, and this is really the most fundamental point, if inventors must appeal to government officials, they are confronted, as it were, with the average daring of the community, whereas, if they are free to appeal to private enterprise, they can select a group of supporters from persons above this average. As Leroy Beaulieu well wrote: “A man of initiative will always find, among the forty million inhabitants of a country,
some audacious persons who will believe in him, will follow him, will make their fortunes with him or will ruin themselves with him. He would waste
his time in trying to convince those hierarchical bureaus which are the heavy and necessary thought-organs and action-organs of a State.”
*94 It follows that, in general, while the hope of gain operates more strongly on private enterprise than on the public authority, the fear of loss operates more strongly on the public authority. Of course, this is not true in war time. Then, as recent experience has shown, governments will authorise experiments in new types of destructive apparatus regardless of cost. But the fact that it is not true in war time is no argument against its truth in times of peace. Just as experience shows it to be untrue in war, so also it shows it to be true in normal conditions. Public authorities are, in general, less willing than private concerns to take risks, or, to put it technically, to provide the factor uncertainty-bearing. A good illustration of this tendency is afforded by the conduct of the British Government in regard to the working of the telephone trunk lines after they had been taken over by the Post Office in 1892. “The Treasury compelled the Post Office to adopt the policy of refusing to make any extensions of doubtful prospect, unless private persons, or the local authority interested, should guarantee ‘a specific revenue per year, fixed with reference to the estimated cost of working and maintaining a given mileage of trunk-line wire.'”
*95 The opinion of Sir George Gibb may be cited in evidence that this proceeding is representative of the general attitude of public authorities. He wrote: “Whatever may be thought as to the respective merits of private and public ownership, it cannot be denied that private enterprise does take more risk than any government is likely to do except under pressure of military necessities.”
*96 Marshall brings out very clearly the effect upon inventions implied in this unwillingness of public bodies to bear uncertainty: “It is notorious that, though departments of central and municipal governments employ many thousands of highly-paid servants in engineering and other progressive industries, very few inventions of any importance are made by them; and nearly all of these few are the work of men like
Sir W. H. Preece, who had been thoroughly trained in free enterprise before they entered Government service. Government creates scarcely anything…. A Government could print a good edition of Shakespeare’s works, but it could not have got them written…. The carcase of municipal electric works belongs to the officials, the genius belongs to free enterprise.”
*97 Again, the Reporter of the American Civic Federation writes: “The Assistant Secretary of the Board of Trade, Mr. Pelham, told the Committee [of the Civic Federation] that they did not encourage the trying of new inventions, or the trying of systems in any way experimental, by municipalities. They waited for these to be proven out by private companies. Progress is all with the companies.”
*98 Moreover, at present, the comparatively small number of undertakings which are operated by public authorities stand in a
milieu where private enterprise is dominant, and where most of the constituents of governing persons are working under private enterprise. In these circumstances public enterprise may be keyed up to a degree of daring which it would not attain if, instead of being the exception, it became the rule.
*99

Now, it is evident that the effect of a restriction of the willingness to take risks, and, therewith, of the stimulus to invention, upon the economies of production will vary in importance in different industries, according to the extent of the speculative element involved in them. Hence it follows that the relative inefficiency of public operation, as compared with private operation, is very large in highly speculative undertakings, and dwindles to nothing in respect of those where the speculative element is practically non-existent. This idea is sometimes crystallised in an attempt to group industries into two divisions, the speculative and the non-speculative, after the manner in which trustees distinguish between speculative securities and investment securities. This grouping, it is sometimes suggested, can be adequately worked out by setting on the one side new industries in an experimental stage, and on the other industries that are
already tried and known. Thus a recent writer has put in the former category “airship construction, wireless telegraphy, ornamental and luxury trades, the production of single special machines and special transport arrangements, the erection of big and difficult buildings and the like,” and in the latter “coal mines, the manufacture of steel, cement, locomotives, telephones, electric cables, motors, and so forth.”
*100 Again, Sir George Gibb distinguishes, from this point of view, the railway industry at an early, and at a mature, age. “As regards the age of construction, at all events, England has derived incalculable benefit from the fact that the railway system has been made by private enterprise. But the problem of working the railway system after it has been constructed is, I admit, essentially different from the problem of securing its construction.”
*101 In like manner, Professor Commons, writing in 1904, while he approved of the establishment of city electric-lighting plants at that time, considered that “those cities which entered upon municipal electric lighting eight or ten years ago are open to criticism.” “Private parties,” he holds, “should be encouraged to push forward in all the untrod fields.”
*102 The distinction thus insisted on has, no doubt, considerable importance. Two points, however, should be noticed. First, an industry, which is old-established at one place, may need new construction at another, and the conditions of construction there may be such that a large speculative element remains. For example, though the industry of water supply is an old one, different towns have to be supplied from sources situated so differently, and along routes of such varying character, that little guidance for one town can be drawn from the experience of others. Secondly, no industry is likely to be so far established that experimentation—which involves speculation—as to improved methods is undesirable. In some measure all industries, in which possibilities of development remain, demand readiness to take risks if further inventions are to be made, and are, therefore, liable to be hampered by anything that obstructs this readiness. It would, therefore, be an error to
suppose that the relatively uneconomic character of public operation, due to the circumstances discussed in this section, is significant only for new industries. It probably has some appreciable significance in regard to nearly all industries, though, of course, its importance is greatest in regard to those in an experimental stage.

§ 12. I pass to a third consideration. The relative inferiority of public operation, due to the interference which it causes with the most economical combination of the different factors of production—for that is, in effect, what obstacles in the way of people’s readiness to take risks, or to brave uncertainty, implies—is paralleled, in many industries, by a further inferiority due to interference with the most economical size of business unit. Practically speaking, public undertakings can only be operated by groups of people united into some form of political organisation. But it is highly improbable that the areas of control most economical for the working of any industry will correspond in size with the areas covered by the public authorities existing in a modern State, since these are set up with regard to quite other considerations than the efficient running of industries. Consequently, in general, it must happen either that special public authorities are created for the express purpose of running certain industries or that the size of the units of control in these industries is altered to fit the scope of existing public authorities. For very large enterprises having a scope midway between that of the central government and that of the relevant local authority, experience shows that special public bodies, adapted to this scope, can be, and have been, created. We are familiar, for example, with the various harbour trusts and dock trusts, with the London Water Board and the Port of London Authority. Another device is that of joint boards of management representing two or more local authorities. “In England and Wales, during the year 1907-8, there were twenty-five joint boards or committees for the supply of water, two for the supply of water and gas, and one for the supply of electricity and the management of a tramway undertaking.”
*103 Though, however, for very
large businesses, the creation of special public bodies is admittedly a practicable policy, it is not always likely to be adopted. The dauger, that, under public operation, local authorities inadequate in area will become the agents of that operation, is especially great in industries originally adapted to the area covered by these agents, but afterwards fitted, as a result of new inventions, for larger areas. In former times the areas of management most suitable for the industries of water supply, gas lighting and electric power supply were approximately coincident with the several municipal areas. But, since the advent of certain modern discoveries, the areas, which might be expected to prove economically most efficient, are often much larger than municipal areas. Thus, “with horse traction the limit of each local authority was, roughly, the limit of commercial working. With electric traction the parish became a mere item in a comprehensive system, which might extend over a whole county.”
*104 Again, with the improvement in methods of distribution for electricity in bulk, the most economical area for the supply of electricity has come to extend over thousands of square miles. Even in the supply of water, now that the needs of large towns are satisfied by the tapping of distant lakes, there may be economy in a joint organisation for supplying a number of towns along the route that the pipes must follow. Indeed, it would seem that gas lighting is the only one of the public utility industries for which the most economical area of management at the present time does not exceed the municipal area. These changes in the area proper to management have not, however, in general, been followed by the transference of the public utility industries to new public authorities created
ad hoc; for the task of ousting the municipalities is opposed by an immense amount of friction, and is, therefore, little likely to be successfully undertaken. Hence, in practice, public operation often implies that industries, whose most economical area of management is intermediate between the areas representative of the central authority and of local authorities respectively,
will, in fact, be worked by local authorities; and this, of course, implies a reduction of the unit of management below what is economically best.
*105 In enterprises whose most economical area of management is smaller than that covered by the smallest existing type of public authority, the creation of new authorities for the special purpose of running them cannot even be said to be practicable. If such industries are to be taken over by any public authority, this authority can hardly be other than one of the authorities that already exist for other purposes. Consequently, in these industries public operation, not merely in general, but practically always, implies the introduction of a scale of management larger than is economically most efficient.

§ 13. Now, if it were the fact that under private enterprise all industries would always evolve the most economical unit of management, it would follow that public operation could not, in this respect, be superior, and would, in general, be greatly inferior, to private operation. In industries normally conducted under conditions of simple competition, such as the industries of baking, milk-supply, house-building or farming, we may fairly presume that private enterprise will, for the most part, evolve the most efficient size of unit. But, where any element of monopoly is present, we may by no means presume this. The most economical unit may be prevented from realising itself through friction, or through the hindrance imposed by popular dislike of large amalgamations, or in other ways. The probability that it will be so prevented is especially great in an industry whose normal condition is, not that of simple monopoly, but that of monopolistic competition. Here, as was pointed out in Chapter IX., there are large wastes due to competitive advertisement and so forth, which
centralisation under a single management might remove. Of railways, for example, Sir George Gibb wrote some years ago: “Each railway company works for its own route. The result is that unnecessary mileage is run, and train loads are lessened…. If those responsible for the handling and carriage of railway traffic could work with a single eye to economical results, and in all cases forward traffic by the routes which yielded the best working results, great economies could undoubtedly be effected.”
*106 This statement was borne out by the experience of the joint working of British railways during the war; though it must be remembered that the character of war-time traffic, with its large train loads of munitions and troops, was exceptionally favourable to economical working. Like economies are sometimes obtainable from the combination, not of different firms engaged in the same occupation, but of different occupations. There is probably an economy in the co-ordination under one hand of the various industries that utilise the public streets. “Water mains may be laid before streets are paved, thus saving the damage and expense of tearing up good pavement to lay water pipes.”
*107 In like manner, it may well be held that important economies would result if the work of treating disease could be brought, by means of a State medical service, into direct connection with the work of preventing disease that is now undertaken by the Public Health authorities. Though, therefore, indirect evil consequences of the kind discussed in Chapter X. § 4 may emerge, and though also certain vertical combinations,
e.g. between a particular coal-mine and a particular ironworks, that would yield structural economies, may be impeded, it is at least possible that, in enterprises of this sort, public operation, instead of hindering, might actually foster the growth of the most economical unit of management.
*108

§ 14. So far of generalities. When the practical issue is
raised whether a particular class of enterprise could, with greater advantage to the national dividend, be publicly controlled or publicly operated, it will be necessary, in order to reach a satisfactory conclusion, to take into account both the comparative effects which the two forms of management are likely to have on productive efficiency and the comparative ease with which whatever regulation the public interest may require can be applied under them. In industries closely associated with the public health, where reliable quality is essential and where inspection cannot easily be made thorough, public operation may be desirable, even though the probable alternative is competitive, and not monopolistic, production. Thus there is much to be said for the public provision of slaughter-houses, to which, as in Germany, all butchers are compelled to resort, and for the public provision of milk for the use of young children. The Reporters to the American Civic Federation are of opinion that “undertakings in which the sanitary motive largely enters should be operated by the public.”
*109 On the other hand, industries, in which the typical producing unit is small, and private firms, rather than joint stock companies, are dominant, are hardly ever suitable for management by public authorities. Apart from a few special exceptions, the proposal for public operation is a live one only in industries in which the typical producing unit is large, and which, therefore, tend towards monopoly. The case for it, as against the case for public control, is strongest in industries which have been reduced more or less to routine and in which there is comparatively little scope for daring adventure. It is relatively weak in industries which are to an important extent rival to other privately operated industries, or in which the normal unit of management covers an area widely different from that covered by existing public authorities. Whether any particular monopolistic industry should be publicly operated or publicly controlled cannot be determined in a general way. Before a decision between these alternative methods is arrived at, a detailed investigation of the industry must be made, and this should be supplemented by an impartial estimate of the quality of the particular
public authority whose action is involved, as well as of the probable effect of new tasks upon its efficiency for the purpose of its primary non-industrial duties.

§ 16. This question of the purchase price leads on to
a very important consideration, with which this Part may suitably close. At first sight it might seem that, if the public has to pay the full market value for a monopolistic concern, which is charging exorbitant prices for its products, no national advantage can possible result. The monopolist simply takes in a lump sum what otherwise he would have got in an annual tribute. This way of looking at the matter is, however, mistaken. The evil of monopoly is not merely, or even mainly, that it enables one set of people to mulct another set. It is that it causes resources to be held back from a form of investment in which the value of the marginal social net product is larger than it is elsewhere, and thereby contracts the national dividend. To do away with this monopolistic policy will increase the size of the national dividend and augment economic welfare, in spite of the fact that, in order to do away with it, one part of the community has to pay a fine to another part. For this reason it is greatly preferable that the government should pay the ransom demanded by the monopolist, write off part of the purchase price, and then operate the concern on terms that would have yielded normal returns if no ransom had been necessary, than that it should allow the private monopolist to continue, by exorbitant charges, to hold back production and check the flow of resources into the enterprise. It is true that, if it does this, the government will have to borrow the purchase price from the public, and, subsequently, to levy taxation to provide interest upon it, and that the amount of this taxation will be roughly equivalent to the exaction which the monopolist would otherwise have made. We may presume, however, that the taxes imposed will either be direct, or, if indirect, will be spread over several commodities, and, therefore, will not shift industrial effort out of its normal channels nearly so far as the monopolist’s exaction would have done. This consideration must not, of course, be allowed to make those persons who have to bargain on the public’s behalf unduly pliant to the pressure of interested sellers. But it is, none the less, an important consideration. When the vested interests of new shareholders in monopolistic concerns make the government
unwilling to force prices down to the proper level through a policy of control, it constitutes a very powerful argument for a policy of purchase. If purchase is made, the natural consequence would be for the government to operate the industry itself. But, if, for any reason, it does not wish to do this, and prefers to sell or lease it to private persons, on terms which involve a money loss to itself but provide for the establishment of a proper price level, it will still have eliminated the evil of monopolistic restriction of output and indirectly benefited the national dividend.