rily those two persons, and only incidentally, under special conditions, becomes "affected with a public interest."

The antithesis to this proposition is clearly involved in what has already been said. The most important effects of private bargains consist in the part they play in the qualitative evolution of the personal characteristics and social relations of the human race at large. Their immediate direct effects on appropriable goods and services may well be regarded as things requiring secondary attention from economists in the present stage of human history. Even to increase tangible wealth the prime requisite is to control the business cycle-clearly a community problem-and to improve the relation of labor to industry. All industry and trade is primarily affected with a public interest, and only certain phases of it may be regarded as of purely private concern.

Private contract is good, not because it deals with things of purely private concern, but because there is no sure social judgment on these matters. Society does not know what it wants, and private contract furnishes an elastic method of experiment which has many advantages over political tinkering. Its greatest present task is to remake the constitution of industry in a tentative and voluntary fashion. A society formed by "social compact" is possible in the economic realm, grounded upon a social organization developed in other ways, and this social compact is the most significant form and aspect of "private bargaining" at the present time.

Proposition 5. As a general rule, cost varies in proportion to' output: "overhead costs" which are independent of output are the exception and arise in connection with large fixed capital outlay.

As an alternative, consider the proposition that overhead costs are not merely the general rule but are universal, and that the costs that are "direct," or "variable" or "prime costs" from the standpoint of the employer are overhead costs from the point of view of society. Every laborer is his own "overhead" and has his constant costs to meet, whether he is working or not. Society has so much productive power. It cannot avoid the cost of either (1) maintaining this productive power, including the human beings who compose most of it, or (2) seeing it depreciate. The power of the entrepreneur to reduce expenses by ceasing to buy materials and reducing his working force, where this means unemployment, is for the most part a financial illusion arising out of the fact that the individual employer does not have to pay the constant costs of the

industries that make his materials, or of the labor he turns off.1 This principle has many possible applications, largely in the direction of placing the burden of labor's overhead cost on those bearing chief responsibility for the industrial cycle, and where it will have the greatest effect in stimulating people to do things to relieve the irregularities. It is one tremendously important instance in which social accounting and business accounting diverge and in which the social accounting is not yet organized and explicit.

Proposition 6. Private enterprise is necessary and efficient because people will work and sacrifice for their individual ends better than for a collective end.

For an antithesis to this, one need not deny individual selfishness. But private enterprise no longer depends on it in the old simple way. One need only remember that the dominant economic personalties today are corporations and labor unions-both types of collective personality. Even the entrepreneur function has become split up into a large group of functions, shared by different persons and even by different corporate entities-for example, the banks. The older proposition is still true, in that the employees of these corporations do not work as hard nor as faithfully as a man "working for himself." Nevertheless the advantages of large scale production have outweighed this handicap and falsified the earlier prophecies as to the limited scope of corporations in economic life.

But these associated personalities act through real persons, and in order that the corporation should act selfishly, following its corporate interests, the real persons composing it must act loyally in the interest of the collective whole. This loyalty has, indeed, become a dominant quality of the modern economic man, vitally necessary to the continued truth of "individualistic" economics. The ability of corporate organizations to stimulate loyalty, or to make loyal behavior the best policy, or both, is the vital necessity of private enterprise. We have now to compare, not private selfishness with collective loyalty, but the business versus the political methods of securing this loyalty, which is equally necessary for private or public enterprise.

Proposition 7. The rational foresight of individuals is at the basis of individualistic economics.

1 This general point has been developed by G. P. Watkins, "A Third Factor in the Variation of Productivity," Amer. Econ. Rev., v, 753, 770-77. The present writer has developed it further in a paper entitled "Some Social Aspects of Overhead Cost," Amer. Econ. Rev. xiii, Supp., 50-59, Mar., 1923.

In particular: business men know the "best combination of factors" and the marginal productive worth of different factors, and on this the efficiency of individualism depends. The question at issue is not whether people in general and business men in particular are thus intelligent, but whether an economics founded on the idea. that they are is really a sound basis for individualism.

Let us study for a moment a hypothetical society made up of individuals gifted with this degree of intelligence and foresight. Would this be peculiarly favorable to individualism? Business would be highly efficient, it is true, but there is serious question how long beings with this much foresight would continue to compete. Such beings would not let the coal business run in the present wasteful fashion. They would long since have integrated the industry to a point where wagon-mines and backyard wheelbarrow mines could not offer serious competition to the organized large-scale producers. And there would be a Federal Commission.

Perhaps in this hypothetical society employers and employees in, let us say, the building trades, might both get tired of the competitive struggle and unite to assist each other in excluding competitors. If they were very foresightful they might cover up their tracks so well that even Mr. Untermyer might have difficulty in getting evidence enough to make a case. The "best combination of factors" in this case would consist in putting some of the very best brains to work on covering up tracks, and the marginal worth of highgrade brains in this occupation would be extremely high. Since the employers would know all this, highgrade brains would be so employed and lavishly rewarded. On the other hand, the marginal worth of brains spent in detecting such conspiracies would also be very high, and government would know this (by hypothesis) and would bid against private enterprise for the same grade of brains. Thus a large part of society's ability would be devoted to this interesting game of hide-and-seek; to making and contesting valuations of properties; to checking up accounts, inspecting products to enforce standards of quality and service, and many other similar occupations; and their pay would absorb a considerable part of the national income.

Meanwhile, if people in general could choose intelligently for the future they could and would vote intelligently and select able and honest officers of government, and government departments would be very efficiently run because their heads would know the "best combination of factors," and the "marginal worth" of talent, and

would be able to get appropriations that would enable them to bid for it. Their budget system, based on knowledge of the relative marginal worth of different lines of public expenditure, would furnish something society has long needed-a system of social accounting in terms of national welfare.

Thus we should have on the one hand enormous waste in the attempt to control private industry, and on the other hand government departments quite as efficient as private concerns and able to avoid the wastes and duplications either of competition or of regulation. Under these conditions any people with such intelligent foresight as we have assumed could not fail to see the obvious advantages of national collectivism, and we might expect to see it adopted as rapidly as an efficient government bureau could handle the necessary accounting.

But suppose, on the other hand, we assume a population of little foresight, who therefore cannot safely depart far from custom, who are governed much by imitation and suggestion from their fellows, who learn only by making mistakes and being hurt, and who do not know the "best combination of factors' nor the marginal worth of different factors of production. In such an imaginary society, private production would not be very efficient, judged by absolute standards of what is technically possible. On the other hand, a considerable amount of competition would survive for a long time, perhaps indefinitely; partly because it is customary, partly because competitors do not look far ahead, and partly because they are influenced by the suggestive force of a moral code which considers monopoly as contrary to the public interest.

With competition, however, goes a weeding out of the least efficient, which means eliminating the widest departures from the "best combination of factors" and from the systems of paying factors of production according to their marginal worth to the employer, and placing them in industry on the same principle. The survivors would be those who, by luck or intuition, came nearest behaving as they would if they had known what the "best combination" was and what the marginal worth of productive factors was. The force of competition would put pressure on business to install cost accounting systems in order to lessen its ignorance of how much productive factors are worth, and to employ engineers and efficiency experts who, although they may not know the absolute "best combination," may at least be able to devise ways of improving the combination that exists. Thus there would be forces working

in the general direction of efficiency and of a rough approximation to an organization on the basis of marginal productivity.1

On the other hand, government departments would not only suffer from not knowing what the "the best combination of factors" was, but they would lack the saving grace of competitive elimination to work automatically in that direction, or to put pressure on their administrative heads to install efficient systems of accounting and research. In such a purely hypothetical society, we might expect that the accounting systems of government departments would be far inferior to those of private business; and would be, in fact, inadequate to show how much public services cost, let alone the marginal worth of different productive factors taking part in those services.

Not knowing how much factors of production were worth, government would tend to underpay the highest employees and fail to secure the highest talents in competition with private industry. Moreover, a people of this mental type would be less able to vote intelligently than to buy goods intelligently, since voting makes peculiar demands on the powers of foresight, while the buying of goods can be much less wastefully done by the method of trial and error. Most marketable commodities are renewed far oftener than is the incumbent of a political office, and of those that are not, few commodities are selected without more intelligent study of values than the average voter makes of the average candidate on the average American long ballot. Thus in this hypothetical society there would be less chance of getting able public officers than able business men. And though the absolute efficiency of private enterprise would be far less than in our first hypothetical society, that of public enterprise would suffer still more. In the second society, private enterprise would be more efficient than public.

The tendency to combination would undoubtedly appear, though restrained by the custom of competition and the moral mandate for it. The attempts of government to curb combination would be largely abortive, but the people might still prefer a moderate degree of exploitation to the inefficiency of government operation,

1 Professor J. B. Clark has, in lectures, made the point that the marginal productivity theory does not require perfect knowledge on the part of business men of what "marginal products" are; but that the force of competitive selection works in this direction automatically.

2 Cf. T. N. Carver's suggestive comparison of political and economic competition in his Principles of Political Economy, 328, and in his Essays in Social Justice, 111-125.