A Market Failure Interpretation of Dominance

The existence of dominant firms which, but for anticompetitive conduct, are apparently immune from the antitrust laws has several undesirable consequences. First, the existence of a dominant firm, whatever its origins, commonly results in resource misallocation — including, but not limited to monopolistic output restriction.11 Second, as Turner has cogently argued, the inability to deal effectively with established monopoly results in what may sometimes be an excessive expansion of antimerger enforcement ( 1969, p. 1213). Were the enforcement agencies and courts better able to “correct undue concentration as it appears, there would be less need to prevent mergers which present only remote possibilities of anti-competitive consequences (Turner, 1969, p. 1214). Third, for structural dominance to remain beyond the reach of section 2 produces charges that antitrust en- forcement is a charade (Galbraith, 1966, p. 794; Miller, 1968, p. 136). Whatever its merits in other respects, confidence that an American-style antitrust apparatus can monitor an enterprise system is impaired. Conspicuous, if limited, failures easily become the basis for captious contempt. Finally, dominance in business may be used as an excuse for bigness in other spheres of economic and political activity — to the possible detriment of the public (Stigler, 1952, p.123).

The antitrust enforcement agencies are not unaware of these difficulties. Substantial enforcement resources are accordingly invested in the study of dominant firm industries. Since, however, the admissible grounds for filing a complaint are limited, the main effort is to scrutinize the behavior of dominant firms to discover evidence of offensive business conduct. Considerations of reciprocity, leasing practices, and the like are seized upon to support claims that dominance has resulted from unfair business practices. That, objectively, the conduct in question could not reasonably lead to the dominance result is simply disregarded. Attention is focused instead on whether certain practices were employed and, if so, whether they may have had some, however slight, anticompetitive effect. That the process is regarded by outsiders, and even some insiders, as artificial and contrived is only to be expected.

This chapter advances the proposition that the Court’s instincts are correct in that dominance is often to be attributed to “business acumen” or “historic accident.” However, rather than treat such dominance as exempt from the coverage of section 2, it is argued that frequently it should be regarded as an actionable manifestation of market failure. Under such an interpretation, government intervention to upset this condition seems reasonable, provided only that (1) the dominant firm’s market position can be judged to be relatively secure and hence unlikely to be undone by the operation of unassisted market processes, and (2) an efficacious remedy can be devised. Contrived proof of anticompetitive conduct would thereby be made unnecessary in order to obtain relief.

1. Business Acumen 

The usual position on business acumen is that scarcity of decisionmaking skills is not responsible for dominance. One way of putting this is to state that requisite talent is available in elastic supply. The view expressed by Kaysen and Turner is typical. They assume that not only is there a sufficient supply of first-grade managerial talent to run a few hundred companies, but that such talent is relatively transferable as well. Thus, should any particular firm develop an “advantage in men and methods, rivals can and will copy the methods and hire the men away” (1959, p. 9).

Only recently has this position come under scrutiny. Marschak, in expressing concern over the ready tendency of economists to accept or employ assumptions of fungibility. has offered the following observation on the distribution of talent: “There exist almost unique, irreplaceable research workers, teachers, administrators: just as there exist unique choice locations for plants and harbors. The problem of unique or imperfectly standardized goods . . . has been indeed neglected in the textbooks” ( 1968, p. 14). Studies in entrepreneurial history would seem to bear him out: unusual men with exceptional personal force and organizing talent occasionally appear, and their influence has pervasive industry consequences (Chandler, 1966; Chandler and Salisbury, 1971).

Were superlative management the only factor responsible for differential business acumen between firms, the dominant firm that made such a showing might reasonably be exempted from divestiture. If the talent in question is not divisible, much as a scale economy of a technological sort is not, then the firm should be permitted to maintain its size because to do otherwise would result in a nontrivial loss of efficiency. The resulting monopoly power is simply an unhappy by-product of this condition.

Even in light of these considerations, however, the business acumen doctrine may sweep too broadly. For one thing, the management superiority that gave rise to dominance may refer to an earlier period in the firm’s history. Often it will have manifested itself through an organizational innovation of a significant sort. The firm may then remain dominant despite the lack of any continuing superior acuity. Displays of unusual organizing ability early in a firm’s development that have since lapsed do not obviously support the same policy conclusion reached above. The advisability of placing a time limit on a defense for dominance that relies on management excellence during an earlier period at least warrants consideration.

An even more serious difficulty with the business acumen doctrine is the relative nature of the argument.12 Dominance does not necessarily imply superiority on any absolute scale. The dominant firm may have displayed no special management expertise but extant and potential rivals, on which the responsibility for self-policing functions devolves, may have been uncommonly inept. Persistent ineptitude of this sort is an indication that the self-policing functions of rivalry have disintegrated. Such discreditable performance on the part of principal rivals during critical formative stages of an industry’s development will be referred to as default failure.

Management superiority (business acumen) may therefore take either of two forms. First, absolute superiority may have contributed to dominance — although this may have been associated with an early period in the firm’s history and may since have lapsed. And second, the advantage of the dominant firm may be attributable to the inepitude of actual and potential rivals. If, for the reasons given below, unassisted market processes cannot be expected to rectify such dominant firm outcomes in any short period of time, the possibility of eventual government intervention to relieve these conditions ought seriously to be entertained in the second situation, and in the first when the absolute superiority is not both contemporaneous and indivisible.

2. Historic Accident 

Although all of the firms in an industry may have been performing in a fully creditable, but unexceptional, manner, the dominant firm may be thrust ahead of its competitors by an unusual sequence of fortuitous events. Such an outcome will be referred to as chance event failure.

It is relevant in this connection to distinguish between defects in the competitive process and dissatisfaction with a stochastic outcome. In order to focus strictly on accidental outcomes, the stochastic generating mechanism will be assumed to operate without defect. Unwanted outcomes thus are to be attributed strictly to the laws of chance intrinsic to the probabilistic process and not to systematic distortions in the mechanism.13 Even under these conditions, however, probabilistic processes can be expected to generate, with statistical regularity, occasional outcomes that may be deemed to have “undesirable” properties.14 The argument is illustrated by the simulation experiment reported by Scherer.

Scherer began with industry of fifty equally sized firms and, for each firm, obtained a series of annual growth rates by sampling repeatedly from a common distribution of growth rates. The growth path of each firm was determined in this way; and industry concentration ratios, at ten-year intervals, were computed. The resulting firm size distributions often revealed high concentration, since “[s]ome firms will inevitably enjoy a run of luck, experiencing several years of very rapid growth in close succession” (1970, p. 127).

Moreover, what is relevant for our purposes, concentration once achieved was not easily undone. “Once the most fortunate firms climb well ahead of the pack, it is difficult for laggards to rally and rectify the imbalance, for by definition, each firm — large or small — has an equal chance of [con- tinuing to grow] by a given percentage amount” (Scherer, 1970, p. 127). Thus, repeated application of the same stochastic mechanism that, within a relatively brief interval, gave rise in some industries to concentration cannot reliably be expected to undo this result in any short period of time.145 If, during the course of their development, industries typically undergo changes that progressively reduce the frequency and intensity of the random disturbances to which they are exposed, dominant firm outcomes will be all the more secure.

Three stages in an industry’s development are commonly recognized: an early exploratory stage, an intermediate development stage, and a mature stage. The first or early formative stage involves the supply of a new product of relatively primitive design, manufactured on comparatively unspecialized machinery, and marketed through a variety of exploratory techniques. Volume is typically low. A high degree of uncertainty characterizes business experience at this stage. The second stage is the intermediate development stage in which manufacturing techniques are more refined and market definition is sharpened; output grows rapidly in response to newly recognized applications and unsatisfied market demands. A high but somewhat lesser degree of uncertainty characterizes market outcomes at this stage. The third stage is that of a mature industry. Management, manu- facturing, and marketing techniques all reach a relatively advanced degree of refinement. Markets may continue to grow, but do so at a more regular and predictable rate. More accurate and complete information regarding market developments becomes available. Unanticipated system breakdowns are progressively eliminated as experience accumulates and statistical inference techniques are improved. Plant and equipment are in place; employees with firm-specific experience and attachments are on hand; established connections with customers and suppliers (including capital market access) all operate to buffer changes and thereby to limit large shifts in market shares. Significant innovations tend to be fewer and are mainly of an improvement variety (Kuznets, 1953, pp. 258-267). Circumstances which, at an early stage in an industry’s development, would have given rise to considerable dislocation now are attenuated.

Relevant in this connection is Kaysen and Turner’s discussion of barriers to entry as a function of the stage of development of an industry (1959, pp. 73- 75). As they point out, new entry into a mature industry is impeded by the lack of knowhow, by the difficulty of upsetting established customer connections, and by the absence among would-be entrants of a known performance record to be assessed by factor suppliers (including the capital market). What this argument comes down to, essentially, is this: The potential entrant into a mature industry must not only raise capital sufficient to finance plant and equipment at an efficient scale but must, in addition, have resources sufficient to cover the startup costs which, in a mature industry, may be considerable. Unlike entry at an early stage in an industry’s development, where differential experience and reputation effects are, perforce, negligible, cost differences between established firms and new entrants at a mature stage are to be anticipated precisely because these factors now may be considerable. First-mover advantages may be said to favor those firms which, for whatever reason, were there early.146

Among the first-mover advantages that warrant special attention are those that are associated with the market for management talent. As noted earlier, Kaysen and Turner assume the management talent issue away. The argument goes through if the stock of talent is sufficient to staff prospective rivals of the dominant firm, if this talent can be easily identified, and if prob- lems of transferability and interpersonal compatibility are not severe. Of these assumptions, the first two are probably the most plausible. But for the occasional exception of truly gifted entrepreneurs, talent in sufficient numbers and its easy identification will accordingly be taken as given. Problems of interpersonnal compatibility and transferability, however, are more formidable.

That managers are imperfectly mobile resources — partly because of personal (including family) preferences and partly because employment contracts commonly include such deterrents to mobility as nonvesting — 147 is obvious. Less evident, but nonetheless important, is the matter of who works well with whom under what circumstances. This is the compatibility issue. Barnard expresses it succinctly as follows: “. . . the question of personal compatibility or incompatibility is much more far-reaching in limiting cooperative effort than is recognized” (1962, p. 146). Efficient adaptation to changing circumstances is impaired “when there is incompatibility or even mere lack of compatibility, [since] both formal communication and especially communication through informal organization become difficult and sometimes impossible” (1962, p. 147). Although team considerations might be overcome by raiding the dominant firm to accomplish the transfer of a coherent group, executing such a transfer poses much more formidable problems than does the hiring of a few key individuals.

The relevance of this condition to the present argument is a temporal one. Thus, it is not that early entrants somehow avoid the costs of assembling the necessary talent but that, given that their cost experience is symmetrical and contemporaneous, none could be said to experience a disadvantage. As between established firms and would-be entrants, however, a contemporaneous cost difference in this respect does exist.To the extent that the prospective entrant’s initial costs exceed the steady-state costs of established firms in the industry, a higher price to (steady-state) cost margin will be required to induce entry. Ceteris paribus, nontrivial transaction costs in the market for management talent inhibit entry.

Although the industry life cycle stability properties referred to above might be offset if the industry in question is one with a dominant firm, dom- inance seems more likely to reinforce stability. This obtains for two reasons: the dominant firm is able often to manage variety, and rivalry is apt to be attenuated.148 The management of variety can take several forms.

One of the more important is the supplanting of market tests by engineering and administrative processes. The dominant firm employs an internal selection mechanism to choose among alternative proposals and thereby avoids the unsettling effects of competitive market tests such as would obtain naturally among a group of independently acting firms in a competitive market. Thus, whereas in a fully competitive market the independently acting firm is a variety generator in relation to its rivals, the dominant firm absorbs potential variety through internal decision-making processes. In a similar way, the dominant firm may be able to follow a program of contrived depreciation (Barro, 1972).

Rivalry will be further attenuated if extant and potential rivals behave more discreetly in a dominant firm industry than they would otherwise. Potential entrants in particular may recognize that where any significant market share can be secured only at the expense of the dominant firm, there is an especially high risk of economic reprisal. Although extant rivals sometimes enjoy umbrella effects, they may also forego aggressive initiatives, which would be freely undertaken in a less concentrated industry, if these threaten to upset established relationships.

In short, the degree of uncertainty in a mature industry with a dominant firm can be expected to be less than both (1) what the same industry experi- enced in its intermediate development stage, and (2) what the industry would have experienced in its mature stage but for the presence of a dominant firm. The argument leads to the following proposition: In consideration of the indicated attenuation in the shocks that the system experiences, a position of dominance (achieved by whatever means —including a fortuitous combination of chance events during an industry’s intermediate stage of development) is resistant to undoing once the industry has reached an advanced stage of development. To the extent, therefore, that the Court’s exemption from section 2 scrutiny of dominance acquired through business acumen or historic accident relies implicitly on the assumption that such dominance will soon be upset by the operation of the market, the doctrine is of dubious merit. The combination of first-mover advantages together with normal variety attenuation and absorption processes in a dominant firm market suggest instead that a doctrine which relies on this line of reasoning ought to be regarded warily.

Source: Williamson Oliver E. (1975), Markets and hierarchies: Analysis and antitrust implications, A Study in the Economics of Internal Organization, The Free Press.

Leave a Reply

Your email address will not be published. Required fields are marked *