Despite the intensive attention that has been given to the conglomerate phenomenon over the past twenty years —and especially the past ten years— there has been surprisingly little effort made to evaluate the performance consequences of the conglomerate corporation. Markham’s recent study, referred to above, is an exception. Weston and Mansinghka’s review of the literature in 1971 turned up only three studies prior to their own, and two of these were concerned with special rather than general performance characteristics. Attention in this section is mainly confined to an examination of the Reid study and Weston and Mansinghka’s updated review of the types of performance measures considered by Reid. I conclude with some brief remarks on the second part of the FTC Staff Study.
First, however, it may be useful to consider some “indirect” evidence that is relevant to an assessment of the conglomerate phenomenon. The argument here relies on a combination of a priori theorizing and related natural selection considerations. To begin with, there is Chandler’s impressive historical survey of the invention and subsequent diffusion of the multidivisional concept. As noted earlier, this organizing approach was originated in the 1920’s at Du Pont, under the leadership of Pierre S. du Pont, and General Motors, under Alfred P. Sloan, Jr. Shortly thereafter, but apparently independently, Chandler finds similar organizational changes being introduced by Walter C. Teagle at Standard Oil of New Jersey and by Robert E. Wood at Sears and Roebuck. That the new concept was viable is testified to by the persistence of this organizing structure and the resulting success of each of these firms. Even more impressive is the adoption of the multidivisionalization concept among large corporations quite generally,115 including recent adoptions by large European firms.
Although the timing and nature of structural changes in internal organ- ization are attributable to a variety of causes—including accident, oppor- tunism, faddishness ( unthinking imitation )- those structures that prove durable will rarely lack for economic rationality. Chandler’s survey suggests that firms which were early to adopt the M-form structure prospered and, in classic natural selection fashion, advanced relatively. Large rivals were eventually induced to adapt, although often this was delayed until a change in the top executive position occurred. For some it became absolutely essential, as a survival measure, to imitate.
That the M-form structure possesses attractive rationality properties (in a comparative-institutional sense) is revealed by the discussion in Chapter 8, which interprets the M-form structure as an internal organizational response to the problem that the U-form firm experienced in the face of large size and complexity. The conglomerate form of organization that adopts the multidivisionalization concept ( including a strong general office, supported by an elite staff, which is disposed to give strategic direction to the enterprise and vigorously exercise the compliance machinery in implementing its programs) is less a new form of organization than it is a diversified variant on the basic M-form innovation. This is not, however, to suggest that “mere diversification” is easily accomplished or that its consequences are trivial.
Consider now the evidence reported by Reid (1968) and disputed by Weston and Mansinghka ( 1971). Reid was interested in determining whether the interests of managers or those of owners could be said to be promoted by mergers, be they conglomerate or otherwise. As indicators of management interests, Reid chose growth rates of sales, assets, and employment, while ownership interest was measured by growth rates of market valuation of shares, net income to total asset ratio, and net income to sales ratio. As Weston and Mansinghka point out, however, the net income to sales ratio lacks significance in periods when the asset mix is changing — as it often is when diversification is occurring. Moreover, Reid’s study terminated in 1961, which was still early in the conglomerate movement. Reid nevertheless interpreted his results, judged in terms of the above indicated measures, as mainly revealing that management, rather than ownership interests, were being promoted by conglomerate organization.
Weston and Mansinghka dispute these findings, partly on grounds of internal consistency and partly because more recent data (through 1968) contradict them. Thus, they observe that Reid’s “measures of performance were either not significant or significant in the wrong direction for ten or more of the 14 industries for each of each of his six measures” (1971, p. 929), and that while conglomerates did not compare well with nonmerging firms they compared favorably with those that did merge (1971, p. 930). Also, they note that Reid’s principal measure of profitability (net income available to stockholders divided by total assets) is defective. Unless interest charges are added to the numerator or debt financing subtracted from the denominator, firms employing financial leverage (as conglomerates have been prone to) are unfairly penalized.116
Be that as it may, Weston and Mansinghka studied the comparative performance of large conglomerates in relation to other large firms over the interval of from 1958 to 1968 for each of the following five growth rate dimensions, which roughly parallel those used by Reid: total assets, sales, net income, earnings per share, and market price. They found that the mean growth rates for conglomerates were significantly higher on all counts. They then turned to an examination of earnings performance — measured mainly as before or after tax earnings before interest to total assets — and found that while conglomerates in 1958 had inferior earnings records, they had achieved parity or better by 1968. Altogether they concluded that “an important economic function of conglomerate firms has been raising the profitability of firms with depressed earnings to the average for industry generally’’ (1971, p. 934).
Although additional studies of conglomerate performance need to be made, including an effort to make the organization form distinctions proposed in Chapter 8, the provisional judgments that one reaches are that, in principle, the M-form conglomerate is not without real (as opposed to merely pecuniary) economic purpose and that, in fact, even when none of the organization form distinctions set out in the previous chapter are made, conglomerates as a group appear to have some commendable properties — which properties, presumably, would be all the more apparent among conglomerate firms of the M-form kind.
Consider now Part Two of the FTC Staff Study on conglomerates.
The study reviews alleged anticompetitive consequences of large conglo- merate firms (mainly predatory pricing supported by “deep-pocket” resources, mutual forebearance, and reciprocal buying) as well as possible procompetitive effects of conglomerate mergers ( such as expanding toehold positions, economizing on certain administrative functions, and facilitating efficient resource allocation). The FTC Staff Study found these various arguments to be conjectural and, in an effort to resolve the issues, undertook a study of nine large conglomerates.
The Staff Study produced an essentially neutral verdict: neither anticompetitive nor procompetitive results are systematically associated with conglomerate merger activity. The main objection which the FTC Staff Study raises concerns financial reporting by conglomerates. Public assessment of conglomerate performance is made difficult by the failure to report on a divisional basis.
While greater disclosure of divisional performance seems to me feasible and has merit, I question whether the multidivisional model of conglomerate organization on which the Staff Study relies (1972. pp. 44-46) is correctly interpreted. Thus, the study notes that while changes in auditing were usually made after acquisition ( 1972. p. 50) and that new financing was usually arranged through conglomerate headquarters (1972, p. 53). both of which have a bearing on corporate control and overall resource allocation processes, few efforts were made to centralize the R & D, advertising, purchasing, and so forth, of the operating divisions ( 1972. p. 57). The FTC Staff Study concludes, correctly I think, that refusal to centralize this last group of functions is consistent with the multidivisional model of organization. But it then contends that, absent intervention in the operating affairs of the divisions, the possibility of efficiency improvements is seriously to be doubted ( 1972, p. 85). Indeed, they appear to be of the opinion that conglomerate synergism requires active involvement by the general management in operating affairs; accordingly, the multidivisional structure is held to be in “conflict with the synergism argument’1 ( 1972. p. 192).
Their discussion of new investment likewise warrants scrutiny. The FTC Staff Study takes the position that “really substantial improvements in productive efficiency would probably have to be accompanied by expendi- tures for new plant and equipment11 ( 1972, p. 68). An attempt is therefore made to establish whether investments in acquired firms increased or decreased following the acquisition. While no significant pattern is detected, investment declines outnumbered increases (1972, pp. 69-72).
I submit that both the treatment of new investment and the discussion of the relations between operating involvement, synergy, and organization form are mistaken. With respect to investment, the question is not whether additional resources are invested in the acquired divisions, but rather whether resources are allocated among the divisions in favor of high yield activities. If some of the acquired firms are generating excessive cash flows, in relation to their investment opportunities, while others are not generating enough, a mixture of increases and decreases is to be expected — which is what they report.
The discussion of operating involvement, synergy, and organization form seems to miss entirely the proposition that the conglomerate is to be regarded as a miniature capital market. Synergy in a conglomerate is not mainly attributable to the headquarters supply of consulting services to the divisions. More important are the internal incentive and resource allocation changes that the M-form conglomerate is able to effectuate. Operating involvement is emphatically not the device by which synergy is mainly realized.
Source: Williamson Oliver E. (1975), Markets and hierarchies: Analysis and antitrust implications, A Study in the Economics of Internal Organization, The Free Press.