of increasing returns. This tempts each company to increase its business at almost any cost. Moreover, it is impossible for the railroad to withdraw its investment, however unprofitable it may become. It is urged that manufacturing industries in which large fixed capital is required present substantially similar conditions in both respects. To the manufacturing establishment also, if it has large investment in plant, the fullest possible operation means the lowest interest cost per unit of output. In industries not requiring much fixed capital, it is possible for the competitor to withdraw if the business become unprofitable, thus setting a limit to the disastrous effects of competition. In industries with large fixed capital, we are told, it is impossible for any one to withdraw. Consequently, the concern which is losing most from competition will continue to cut prices, in the hope of gaining enough business to pay expenses and prevent absolute loss of the investment. Thus, it is contended, the business of all competitors often becomes unprofitable, and the temptation to combine, for restoring and maintaining prices, becomes well-nigh irresistible.

If the conditions were as serious as thus depicted, we should feel disposed not merely to give up the struggle to maintain competition in our leading industries, but even to encourage combination. Persistent loss from excessive competition is intolerable. Among wellinformed and unbiased observers, there has, therefore, developed a strong feeling in favor of permitting pooling and community of interest among railroads.

But is it true that competition in manufacturing industries tends ordinarily to such extreme lengths ? Are the conditions in any appreciable number of

industries closely similar to those in railroad transportation? This seems to me not proved.

In the first place, in most industries, the relative importance of fixed capital is much lower than with the railroads and other public service industries. The capital investment of the railroads of the United States is between four and five times as great as their gross annual revenue. The capital investment of the gas companies and of the electric light companies bears about the same ratio to their gross earnings. On the other hand, in manufacturing industries taken as a whole, the census returns show a capital investment less than the annual value of product. Even in the steel industry, which is one of exceptionally large fixed capital, the reported value of capital only slightly exceeds the annual value of output. The statistics on which these statements are based are not altogether reliable, but they do show approximately the true relations. Again, the principle of increasing returns in the case of railroads extends in large measure even to operating costs; this is seldom true of manufacturing concerns. For these reasons to reduce the output of a manufacturing plant when prices are unfavorable does not in most cases increase unit costs, including capital charges, to any such extent as in transportation. Finally, in railroad competition there are usually only a few lines involved. Each of them may readily have sufficient capacity to handle the whole competitive business.

In most manufacturing industries, on the

1 This statement with regard to manufacturing industries is based upon the census of 1909, which gives the value of capital as $18,428,000,000 and the value of products as $20,672,000,000. The latter item involves much duplication, due to the use of the products of one plant as material for another; but it is proper to compare this gross value with that of the capital. For if a manufacturing concern shuts down in order to avoid loss, it eliminates its entire cost of materials, whether strictly raw materials or the partly finished product of other manufacturing concerns.

other hand, plants are numerous. The individual plant has but a comparatively small fraction of the total capacity. Under such conditions no one plant can by price cutting expect to increase its share of the business in any such proportion as the railroad can.

For these reasons, competition of a really destructive kind is much less likely to arise in manufacturing industries than in railroad transportation. It follows that the motive for combination is less powerful in the former than in the latter. There are many manufacturing industries today in which we find neither destructive competition nor combinations in restraint of competition.

In any case, even tho the desire to suppress competition be strong among business men in manufacturing industries, it is very difficult effectively to suppress it when combination is under the ban of the law. Informal agreements and tacit understandings are far less effective in stopping competition than the more formal and definite combinations which it is proposed to prohibit. It is not easy for a group of separate concerns to act in harmony, to refrain from competition. This is particularly the case when prices are at a high level. If each concern could be sure that its competitor was maintaining prices and not seeking to get a larger proportion of the business, tacit combination might go on peacefully. But the temptation to shade prices and get business away from others is always strong. The mere unfounded suspicion that competitors are pursuing this policy often leads the business man to seek to protect his trade by lowering prices, or by other competitive measures. If once the bars are let down anywhere, the whole trade is likely to rush into the field of active competition.

The history of combinations in the past is full of efforts to make them more binding, more cohesive, to prevent more effectively the internal competition that would ever break forth. Agreements as to prices were found ineffective without systematic measures for dividing output or profits. Agreements for such pooling of business or profits broke down unless there was efficient machinery for enforcing them, backed by heavy penalties. Despite even such vigorous methods, many of the pools were not strong enough to prevent competition among their own members. It was largely for this reason that the original trust form of organization, and later the corporate combination, became popular. The difficulties which the railroads in the earlier days encountered in their efforts to suppress competition among themselves are well known; and this despite the fact that their managers knew the peculiar risk of heavy losses from rate wars. The conditions which made competition so disastrous, which offered such an incentive to combination, themselves rendered the prevention of competition peculiarly difficult. No doubt, a large majority of business men would prefer to combine with one another in order to exact high prices from the public. It has already been suggested that if combination were freely permitted, competition would very likely be eliminated in large degree. But combination under the ban of the law is a very different thing from combination with its sanction.

Those who hold that it is impossible to maintain competition as a general basis of business are called on to explain the fact that competition does exist today in a large proportion of the field of production and trade. Many as are the more formal combinations,

perhaps even more numerous the informal understandings among business concerns, there are wide fields in which real competition exists. Can any one deny, for example, that the mining of bituminous coal, or the manufacture of cotton goods, of boots and shoes, of automobiles, is conducted under essentially competitive conditions? Is not the same true of a large part of the wholesale and retail merchandizing? The advantages of combination to its members are so well known that we should expect to find competition practically eliminated everywhere, were it not for the real difficulties of eliminating it.

There is, therefore, no occasion for despair as to the suppression of trusts and pools. Monopoly is not inevitable. Competition in manufacturing and mercantile business is not so destructive as to force combination. The failure of some of the so-called trust dissolutions to restore competition is no proof that more rational and more vigorous methods of enforcing the law would also fail to do so. Competition has been restored in some cases where monopoly once reigned. In many important industries competition has never succumbed.

Hence we can consider on their merits the relative advantages of trust prohibition and trust regulation. Our choice is not foreclosed by the impracticability of the former. Is a trust régime superior from the stand- X point of efficiency to a régime of competition? Is it possible effectively to regulate the prices and profits of trusts? What would be the ultimate consequences of a policy of regulation? These questions remain to be discussed.