A leading theoretical need is for additional work on the properties of hierarchy. The mathematical theory of hierarchical forms is not, unfortu-nately, well-developed (Varaiya, 1972, p. 497). Simon’s (1962) interdisciplinary review of the evolution of hierarchy and his heuristic discussion of its advantages provides useful background for such an effort.189
The economic properties of hierarchical modes of organization can be fully established only after additional work is done on bounded rationality and on the efficacy of internal control processes. Lest, however, internal controls be unduely emphasized and pressed with excessive zeal, more attention to the economics of atmosphere is simultaneously required.
The importance of organizational innovation to economic efficiency is poorly understood. Arrow contends in this connection that “the prime need in organizational design is increasing capacity to handle a large agenda. . . .
Shortrun efficiency and even flexibility within a narrow framework of alternatives may be less important in the long run than a wide compass of potential activities” (1974, p. 59). This is a provocative and intuitively appealing statement, but plainly needs to be explained. Both historical and theoretical studies can contribute to our understanding of these issues. The diffusion of organizational innovations — within industries, across industries, and across cultures — both in terms of the mechanics of the diffusion process and the economic consequences associated with organizational innovations of various kinds, of which I would nominate the multidivisional structure as the leading example, warrants investigation.
Assuming that measures of organizational structure can be worked up for a large number of corporations over a significant period of years, both cross sectional and time series studies of the influence of organization form on enterprise performance can be performed. Although individual instances of improved performance are easy to cite, the extent to which performance trends in individual firms are affected by organization form changes can be established only by more systematic study of the time series data. Among the types of cross section studies that might usefully be repeated by introducing organization form as an explanatory variable are studies of rates of return to alternative sources of funds (Baumol et al., 1970; Grabow- ski and Mueller, 1974), industry studies of comparative firm efficiency (Gordon, 1965), as well as more general studies of financial and economic performance (Meyer, 1967). The recent work of Weston and his associates (1970; 1972) on conglomerate performance could also be refined by making organization form distinctions.
Historical studies of vertical integration, such as Buttrick’s (1952), would contribute to an understanding of the reasons for and properties of vertical structures. In conformity with the basic thrust of this book, an emphasis on transactional rather than technological conditions is urged. Studies of quasi-integrated modes of organization (for example, franchising) could also usefully be undertaken. A sensitivity to the tradeoff between the incentives of autonomous enterprises and the controls of integrated structures is apt to be important in this connection.
Addressing the dominant firm issue could proceed along either of two main lines. The ambitious approach would involve a reconstruction of the history of the industry, especially during its intermediate development stage, for the purpose of assessing default and/or chance event failure. This would require that major events be reconstructed and critical decisions made by rivals and prospective entrants be evaluated. A less demanding approach would be simply to assume that dominance has resulted from market failures of the types described and check against the possibility that the dominant firm outcome is attributable, instead, to economies of scale or original unexpired patents.
The latter approach is intellectually less satisfying than an attempt to reconstruct the history of the industry during the period when dominance emerged. Inasmuch as the number of dominant firm industries is relatively small, it may be feasible to take the more ambitious approach by having such studies done by academics. The active interest and support of the antitrust enforcement agencies would nevertheless facilitate such an effort. Collaboration between academics and antitrust enforcement officials with respect to the dominant firm industries would seem altogether warranted.
3. Public Policy Toward Regulation
The plausibility of arguments that large-numbers bidding at the outset reasonably guarantees that a good or service will thereafter be supplied on efficient terms (Demsetz, 1968) ought to be reexamined with more self- conscious attention to the details of the contracting process. Can “irrelevant complications,” such as equipment durability and uncertainty, be assumed away without stripping the argument of all policy significance? When these are introduced and the limits of short- and long-term contracts are acknowledged, is the case for the private sector “solution” still compelling? If not, an assessment of the transactional properties of alternative modes of organizing the activity in question is plainly required before an assignment to the most efficient mode can be made.
Source: Williamson Oliver E. (1975), Markets and hierarchies: Analysis and antitrust implications, A Study in the Economics of Internal Organization, The Free Press.