Public policy issues of three kinds are posed by the conglomerate. First, there is the possibility that conglomerates are especially prone to engage in what are thought to be anticompetitive practices. Charges of reciprocity and of predatory cross-subsidization are commonly lodged against conglomerates. Second, conglomerate mergers may impair competition by weakening the threat of potential competition or by giving rise to a condition of conglomerate interdependence. Finally, very large conglomerates may pose social and political issues. Consider these seriatim.
1. Anticompetitive Conduct
Reciprocal buying involves an informal conditional purchase agreement, such that I buy from you if you buy from me. Though unlawful, it is nevertheless thought to be widely practiced. Since the opportunity to match purchases with sales is more extensive in a multiproduct than in a specialized firm, conglomerates are said to make sales to the possible disadvantage of undiversified but otherwise qualified rivals. Complaints of reciprocity, or reciprocal buying potential, almost always accompany antitrust actions against conglomerate acquisitions.
Three questions arise in connection with the charge of reciprocity. First, are there alternative interpretations of the incentive to engage in reciprocity? Second, what organizational structures are best suited to support reciprocity? And third, what does the evidence show with respect to the proclivity of conglomerates, in relation to other firms, to engage in such practices?
Stigler observes that an affirmative case can be made for reciprocity on the following grounds:
The case for reciprocity arises when prices cannot be freely varied to meet supply and demand conditions. Suppose that a firm is dealing with a colluding industry which is fixing prices. A firm in this collusive industry would be willing to sell at less than the cartel price if it can escape detection. Its price can be reduced in effect by buying from the customer-seller at an inflated price. Here reciprocity restores flexibility of prices.
An assessment of the consequences of reciprocity is thus uncertain: it may be an unfair sales technique, but it may also be an indirect way of evading an oligopolistic price structure. Since there are other ways of shading prices if reciprocity were disallowed, however, and since reciprocity, whatever its origins, may be continued because it suits the bureaucratic preferences of the sales staff, public policy disdain for reciprocity seems to me warranted.
Consider, therefore, whether conglomerate firms are especially given to this practice. Organization form is relevant in this connection. The general proposition is this: M-form enterprises are less given to reciprocity than are U- form firms, ceteris paribus. The reason here is that divisionalization concept is corrupted if cross-divisional reciprocity is attempted. Accordingly, it is not sufficient to rest a charge of reciprocity on product mix considerations; proclivity must also be considered. As between conglomerates, those that are organized as M-form firms are poorly suited to engage in such practices. Although failure to make organization form distinctions easily leads to undiscriminating charges that conglomerates quite generally are prone to engage in reciprocity, even the courts appear to recognize the importance of organization form in assessing proclivity charges.
Consider, finally, the evidence. Blake observes in this connection that “empirical research, if it could be carried out, would show that reciprocity is as inevitable a result of widespread conglomerate structure as price rigidity is a consequence of oligopoly structure” (1973, p. 569) —where, apparently, the latter, and hence the former, is believed to be extensive. A recent study by Markham (which was unavailable to Blake), in which the organizational underpinning for reciprocity (in the form of a trade practices department) is examined in relation to the degree of firm diversification,109 concludes otherwise: “… highly diversified companies are no more, and may be even less, given to reciprocity than large corporations generally” (1973, p. 176).
The argument with respect to cross-subsidization is that profits earned in one sector of a business may be used to offset temporary losses in another sector that is engaged in predatory price-cutting. Again this is unlawful. But again it is important to distinguish between predatory cross-subsidization, undertaken for the purpose of damaging rivals, and the rational reallocation of internal resources to support divisions which, if the divisions in question were to be organized instead as free-standing enterprises, would not be funded by the capital market. This last need not to be regarded as cross-subsidization in any objectionable sense since, because of the information and incentive advantages of internal in relation to external organization, risks can be efficiently accepted by an M-form enterprise that the unassisted capital market would reject. M-form firms, which have deeper knowledge of the underlying data, may therefore be prepared to sustain current losses that outside investors would consider unacceptable. (See the reasons given in Chapter 8.)
Whether conglomerate firms are especially given to the practice of predatory cross-subsidization is seriously to be doubted. For one thing, any multimarket organization — including specialized firms operating in geographically dispersed product markets —can engage in such practices. In addition, the organization form distinction is again relevant: the M-form structure is seriously compromised if predatory cross-subsidization is attempted.
While the evidence on predatory price-cutting is incomplete, the examples cited by the U.S. Federal Trade Commission Staff Study of Con- glomerates suggest that it is the relatively specialized multimarket firms that are most given to this behavior.110 Also, Markham’s examination of the proclivity to coordinate pricing decisions within each of 204 large manufacturing concerns revealed that pricing decisions were left “almost entirely to division or operating unit managers” in the more diversified of these firms (1973, p. 91).
One concludes that while antitrust vigilance with respect to reciprocity and cross-subsidization is warranted, conglomerate acquisitions ought not to be regarded with special animus on either account.
2. Actual and Potential Competition
Actual competition arguments of two kinds are leveled against con- glomerates. First, diversification by large corporations is said to result in increasing concentration in individual markets ( Shepherd, 1970, pp. 140- 141). Berry’s subsequent examination of the evidence, however, concludes otherwise: Although the market position of the leading firms of an industry is protected from erosion by the entry of small firms, the “market share of entering large firms is acquired at the expense of the leading four firms” ( 1974, p. 202).
The second objection to conglomerates in actual competition respects is that a condition of mutual interdependence recognized develops among them, as a result of their multimarket contacts, and that this results in an attenuation of price competition. Blake, moreover, contends that this is not merely a hypothetical possibility but that there is now “hard evidence to support the no longer novel theory — and widely held belief in the business community — that large conglomerates facing each other in several markets tend to be less competitive in price than regional or smaller firms” ( 1973, p. 570).
There are two problems with Blake’s argument. First, I would scarcely characterize the evidence on which Blake relies upon as “hard.” Part of the evidence cited by Blake is Scherer’s discussion of the “spheres of influence hypothesis” (1970, pp. 278-280). But Scherer is very careful to characterize the evidence quite differently — noting that, even with respect to the prewar international chemical industry, which, aside from marine cartels, is his only Western example, the evidence is fragmentary. With respect to other industries he concludes that “there is a dearth of evidence on spheres of influence accords” (1970, p. 279).
Second, the definition of a conglomerate requires attention. Are all specialized firms (such as National Tea, to which Blake earlier refers) that operate similar plants or stores in geographically dispersed locations really to be regarded as conglomerates? By stretching the definition of a conglomerate to include geographically dispersed but otherwise specialized enterprises, is the number of nonconglomerate large firms shrunk to insignificance?
I submit that if the conglomerate is to be defined in product diversifica- tion terms, Blake (and the U.S. Federal Trade Commission) ought to be expected to generate examples of conglomerate abuse from the universe of product diversified firms. If all large multimarket firms, whatever their product specialization ratios, are the objectionable subset, it is these, rather than product diversified firms (which is the narrower definition of the conglomerate), that warrant attention. As things stand now, the facts, with respect to conglomerates, have yet to be assembled.
The potential competition argument is that prospective entrants into an industry should be barred from acquiring a large firm already in the industry if, but for such acquisitions, smaller firms (toeholds) in the industry would be acquired, and subsequently expanded, or de novo entry would occur instead. The removal of a likely potential entrant from the edge of the market, by permitting it to make a large acquisition, relieves the threat of entry and, arguably, impairs competition.
The anticompetitive potential of conglomerate mergers, in potential competition and other respects, is characterized by Blake as “so widespread that it might appropriately be described as having an effect upon the economic system as a whole — in every line of commerce in every section of the country” (1973, p. 567). He accordingly proposes that conglomerate acquisitions by firms above a specified size [the subset of firms that are to be restricted is not explicitly identified, but Blake makes several references to the top 200 firms (1973, pp. 559-569) ] be accompanied by a spinoff of comparable assets (1973. p. 590) and further stipulates that no toehold exception should be permitted (1973, pp. 590-591).
The basis for his refusal to admit a toehold exception is, however, somewhat unclear. For one thing, the acquisition of a very small firm scarcely, by itself, contributes much to the growth of the large firm. Correspondingly, requiring the large firm to release assets in an equivalent amount whenever a toehold acquisition is made is scarcely more than a nuisance.20 Furthermore, toehold acquisitions made for the purpose of securing a position that will subsequently be expanded is internal growth of the sort Blake favors. Either there is little point to Blake’s toehold argument,21 or he regards expansion by small firms as socially to be preferred to similar investments by large firms.
Assuming, arguendo, that the same investments will be made whether the small firm is acquired or not, it is easy to agree with Blake. But it is doubtful that the same investments will actually occur. This raises transfer process issues.
The examination of these matters in Chapter 10 suggests that small firms apparently enjoy a comparative advantage at early and developmental stages of the technical innovation process. Large, established firms, by contrast, display comparative advantages at large-scale commercial production and distribution stages.22 Not only may the management of the small firm lack the financial resources to move to the commercial stage in any but gradualist manner because its credit standing does not permit it to raise significant blocks of capital except at adverse rates,23 but the management of the small firm may be poorly suited to make the transition. Different management skills and knowledge are required to bring a project successfully to large-scale commercial development than may have been needed at earlier stages. If, because of management experience and team considerations, similar to those described in Section 2.2 of Chapter 6, the talents needed to facilitate internal expansion cannot be costlessly identified and assembled, transferring the project to an established firm that already possesses the requisite talents may be more economical instead. (Again, it is transactions, not technology, which dictate this result.) Put in these terms, it is unclear that the no toehold position survives.
The law with respect to the effects of conglomerate acquisitions on potential competition also appears to be moving away from the broad position ( “in every line of commerce, in every section of the country”) set out by Blake. Commissioner Dennison, speaking for a unanimous commission in the recent decision by the FTC in Beatrice Foods, described the required factual proof that potential competition has been or probably will be reduced as follws:
Complaint Counsel in essence attempt to rest their case on the existence ol concentration ratios alone. The test for finding injury due to elimination of a potential competitor is not simple. Additional factors enter into any analysis of the loss of a potential competitor. Among these are: trends toward concentration in the market; extensive entry barriers: high probability that the lost potential competitor would have actually entered the market; whether the lost potential competitor was one of only a few such potential competitors and whether, if he had entered the market, his new competition would have had a significant impact on price and quality. Although the number of competing firms or trends toward concentration may be enough without more to condemn many horizontal mergers between existing rivals in a market, the condition of entry by new firms as well as these other factors mentioned above must be considered when dealing with elimination of a potential competitor.
The reference to the condition of entry in this statement warrants additional development.
As Turner (1965) has forcefully argued, potential competition is apt to be impaired if one of a few most likely potential entrants acquires a firm that exceeds toehold proportions. If the industry in question is highly concentrated, so that, but for the threat of potential competition, competitive results will not reliably obtain, the quality of competition is degraded by the loss of one of a few “most likely potential entrants.” I would like to urge that the appellation “most likely potential entrant” has genuine economic significance, as contrasted with transitory business significance, only to the extent that nontrivial barriers to entry into the industry in question can be said to exist.
The antitrust distinction to be made is between firms which, for transitory reasons, may have demonstrated an acquisition interest in the industry and firms which, in entry barrier respects ( which introduces non- transitory considerations), are strategically situated to enter. Because the interest of firms of the first kind is unlikely to persist, being explained by such factors as current interest of the chief executive, temporary cash balances, immediate income statement considerations, and the like, prohibiting entry by acquisition to such firms is of little affirmative economic purpose. No long-term benefit to potential competition is thereby secured. Rather, the principal effect is to shrink the acquisition market, which impairs both the market for corporate control and the incentives for entrepreneurs to invest in new enterprises.
The situation is quite different, however, if the industry in question has nontrivial barriers to entry and the firm evidencing an acquisition interest is one of only a few firms for which de novo or toehold entry would be very easy. Consider in this connection the entry barrier conditions identified by Bain ( 1956), namely, economies of scale that are large in relation to the size of the market, absolute cost advantages, and product differentiation. Although Bain describes these barriers without reference to specific firms, plainly the height of the barrier varies among possible entrants. Those for which the barriers are least are the firms that are usefully designated, in potential competition terms, as most likely potential entrants, ceteris paribus.
Thus, though economies of scale may be large in relation to the size of the market, this impediment to entry is apt to be less severe for those few firms which have closely complementary production processes and sales organizations. Similarly, a few firms may be well-situated with respect to absolute cost advantages. Although patents may constitute a severe impedi- ment to entry, high-grade ore deposits may be in limited supply, or special- ized labor skills are required, a few firms are apt to stand out from all the rest by reason of a complementary technology, which facilitates inventing around the established patents, because they possess medium grade ore deposits, or because their labor force has acquired, in a learning-by-doing fashion, the requisite specialized skills. Product differentiation advantages are likewise attenuated for those firms that market related types of consumer goods and themselves enjoy brand recognition.
In circumstances, however, where all such barriers to entry are negligible (economies of scale are not great; patents, specialized, or otherwise scarce resources are unimportant; product differentiation is insubstantial), no small subset of firms can be said to enjoy a strategic advantage. Correspondingly, it is fatuous to attempt to identify a group of most likely potential entrants the loss, by acquisition, of any of which would significantly impair the quality of potential competition.
3. Giant Size
The above assessment suggests that the conglomerate form poses much less severe public policy issues than Solo, Blake, and others have indicated. Public policy will be served by identifying specific instances where conglomerates pose problems rather than by mounting a broadscale attack. Specific abuses (for example, reciprocity) ought to be challenged, but conglomerate acquisitions ought not to be arrested on this account. Similarly, potential competition cases ought to be brought only where nontrivial entry barriers exist and the acquiring firm can be characterized as a most likely potential entrant.
An exception to this case-by-case approach might, however, be war- ranted where the acquisition of already large firms by other very large firms is contemplated. Such acquisitions might routinely be accompanied by a divestiture of equivalent assets. For one thing, as Hofstadter has observed, the support for antitrust rests less on a consensus among economists as to its efficiency enhancing properties than it does on a political and moral judgment that power in the American economy should be diffused (1964, p. 113). The wisdom of such populist social and political attitudes is illustrated by the misadventures of the ITT Corporation in domestic and foreign affairs (Sampson, 1973). Since much of Blake’s disenchantment with conglomerates (1973, pp. 574, 576, 578, 579, 591), appears to be attributable to a concern that giant size and political abuse are positively correlated,113 I would urge that the case be made expressly in these terms. If giant firms rather than all conglomerates are the objectionable subset, attention ought properly to be restricted to these.
A requirement that very large firms divest themselves of equivalent assets when larger than toehold acquisitions are made is also favored by the prospect that this will serve to curb bureaucratic abuses associated with very large size. Although such divestitures sometimes occur voluntarily (Coase, 1972, p. 67), such efforts predictably encounter bureaucratic resistance. If, however, such divestiture commonly has beneficial effects of an organizational self-renewal sort, making divestiture mandatory is scarcely objectionable. It merely strengthens the hand of those in the firm who are anxious to forestall bureaucratization. Absent such a rule, internal agreement on divestiture may be difficult to secure; parties with vested interests will make partisan (opportunistic) representations that will be difficult to reject. Given such a rule, however, the general office can simply plead that it has no choice but to divest (assuming, that is, that a large acquisition is to be made). The preferences of the general office are thus made more fully to prevail.
Source: Williamson Oliver E. (1975), Markets and hierarchies: Analysis and antitrust implications, A Study in the Economics of Internal Organization, The Free Press.