Anticompetitive effects of two types are commonly attributed to inte- gration. price discrimination and barriers to entry. The argument has been compactly expressed in the following terms: “. . . vertical integration loses its innocence if there is an appreciable degree of market control at even one stage of the production process. It becomes a possible weapon for the exclusion of rivals by increasing the capital requirements for entry into the combined integrated production processes, or it becomes a possible vehicle of price discrimination” (Stigler, 1968, p. 303).
1. Price Discrimination
That even perfect price discrimination can lead to allocative efficiency losses, under conventional partial equilibrium assumptions, is evident from the discussion in Section 3.1 of Chapter 1 (though this does not establish that it usually will). My interest here, however, is somewhat different. Attention is restricted to the ways by which vertical integration contributes to the price discrimination result.
Successful price discrimination requires that differential demand elasticities be discovered and sale arranged in such a way as to preclude reselling. Users with highly elastic demands who purchase the item at a low price must not be able to service inelastic demand customers by acting as a middleman, all sales must be final. Although vertical integration may facilitate the discovery of differential elasticities, which is an information advantage, it is mainly with respect to the nonresale condition that it is regarded as especially efficacious.
Thus, let it be assumed that demand elasticity differences are known or easily knowable. Assume also that it is lawful to include a nonresale stipulation in a sales contract. In what respects then can vertical integration help to achieve a price discrimination result? Price discrimination is clearly practiced in some commodities without recourse to vertical integration (witness electricity and telephone service). What are the distinguishing factors? Legality considerations aside, presumably it is the cost of enforcing (policing) terms of the contract that are at issue. Some commodities apparently have self-enforcing properties — which may obtain on account of high storage and resupply costs or because reselling cannot be arranged inconspicuously. The absence of self-enforcing commodity properties is thus what makes vertical integration attractive as a means of accomplishing discrimination. Where promises cannot be believed, the firm may integrate forward into the more elastic market in order to preclude arbitrage between it and another market to which the firm wishes to charge a higher price.
2. The Condition of Entry
Stigler observes: “. . . it is possible that vertical integration increases the difficulty of entry by new firms, by increasing the capital and knowledge necessary to conduct several types of operation rather than depend on rivals for supplies or markets” (1968, p. 138). But why should a simple increase in capital requirements, without more, impede entry? Bork, for example, contends: “In general, if greater than competitive profits are to be made in an industry, entry should occur whether the entrant has to come in at both levels or not. I know of no theory of imperfections in the capital market which would lead suppliers of capital to avoid areas of higher return to seek areas of lower return” (1969, p. 148). Similarly, Bowman observes that “difficulties of access to the capital market that enable X to offer a one dollar inducement (it has a bankroll) and prevent its rivals from responding (they have no bankroll and, though the offering of the inducement is a responsible business tactic, for some reason cannot borrow the money) . . . [have] yet to be demonstrated” (1973, p. 59).
The issue, evidently, is whether an increase of the financial requirements is attended by an adverse alteration of the terms under which capital becomes available. A simple explanation is that borrowing by the firm to finance additional plant and equipment is akin to borrowing by the consumer to mortgage a house, but such an analogy is at best imperfect. The firm is borrowing funds in anticipation of realizing a prospective stream of earnings. These prospective earnings, as well as the resale value of the assets in question, are used to support the loan in question. The home-owner, by contrast, is not ordinarily able to augment his earnings by his purchase of a house. Thus, whereas the householder who successively in- creases the size of his mortgage eventually incurs adverse capital costs because the risks of default are greater, the firm need not likewise be impeded. Wherein, then, if at all, does vertical integration by established firms disadvantage prospective entrants on account of capital market “defects”?
An assessment of the issues will be facilitated by setting out the specific alternatives. Thus, suppose that two distinct stages can be identified in the industry in question, and let these be designated I and II, respectively. Assume, furthermore, that stage I in the industry is essentially monopolized while stage II may or may not be integrated. The question now is whether a potential entrant, who has developed a technologically satisfactory stage I substitute and has an established reputation in stage I related activity, will be unaffected by the integrated condition of stage II. Consider, in particular, the following conditions: (1) the monopolistic stage I producer is not integrated, in which case the prospective new entrant can enter at stage I only and utilize the facilities of stage II producers (suitably expanded if necessary), and (2) the monopolistic stage I producer is integrated into stage II so that either (a) the new entrant himself comes in at both stages or (b) independent new entrants appear simultaneously at both stages. If Bork and Bowman are correct, the cost of capital ought to be independent of these conditions.
To contend that the terms of finance are the same under 2 (a) as they are under 1 implies that the capital market has equal confidence in the new entrant’s qualifications to perform stage II activities as it does in firms that are already experienced in the business. Except in circumstances where experienced firms are plainly inept, this is tantamount to saying that experi- ence counts for nought. This, however, is implausible for transactions that involve large, discrete rather than small, but recurring commitments of funds. Although transactions of the latter type can be monitored reasonably effectively, on the basis of ex post experience, this is much less easy for transactions of the large, discrete variety—which are the kind under consid- eration here. Reputation, which is to say prior experience, is of special importance in establishing the terms of finance for transactions that involve large, discrete commitments of funds.
The reasons for this are to be traced in part to the incompleteness of information regarding the qualifications of applicants for financing. Faced with incomplete information, suppliers of capital are vulnerable to opportunistic representations. Unable to distinguish between those unknown candidates who have the capacity and the will to execute the project successfully from opportunistic types who assert that they are similarly qualified, when objectively (omnisciently) they are not, the terms of finance are adjusted adversely against the entire group. Furthermore, and of special relevance to the issue at hand, if omniscience is lacking, then, as between two candidates for financing who would be judged by an omniscient assessor to have identical capacities and wills to execute the project, but only one of whom has a favourable and widely known performance record, the unknown candidate will find that he is disadvantaged.
Moreover, where both candidates are equally suspect, but one has access to internal sources of financing while the other does not, the candidate requiring outside financing may be unable to proceed. Timing, in this connection, can be of critical significance. If one firm moves to the integrated structure gradually and finances the undertaking out of internal funds, while the second firm perceives the market opportunity later but, to be viable, must move immediately to a comparably integrated structure, the second firm may have to contend with adverse capital market rates.
The learning-by-doing conditions referred to earlier are also germane to an assessment of the earnings opportunities of an integrated versus non- integrated new entrant. By assumption, the prospective entrant is well- qualified in stage I respects. If learning-by-doing yields significant cost advantages and if the prospective entrant has no special qualifications in stage II respects, will his incentive to enter be any the less keen if, by reason of integration, he must now come in at both stages? I submit that if the knowledge gleaned from experience if deeply impacted, which is to say that it is not generally known or easily made knowable to those who lack experience, and if it is very costly to hire away the requisite experienced personnel from the integrated firm (due possibly to the fact that the knowledge advantage which the experienced labor force has acquired is dispersed among a team of individuals, negotiating the transfer of which is prohibitively costly), the prospective entrant is plainly at a disadvantage. Information impactedness and imperfect labor markets thus combine to explain the cost disadvantage of the otherwise qualified new entrant in relation to the experienced firm. Were the monopolistic stage I producer not to have integrated into stage II, so that the prospective entrant need come in at stage I only and could rely on already experienced stage II firms to acquire the necessary capital to expand appropriately and service his stage II needs, capital costs would be lower and the prospect of entry thereby enhanced.
The problems, moreover, do not vanish if the new entrant comes in at stage I only and relies on independent entry into stage II to occur. (The comparison here is condition 2(b) in relation to condition 1.) Not only is the cost of capital adjusted adversely against would-be new processors here, by reason of the lack of experience referred to above, but simultaneous yet independent entry into both stages may be impeded because of nonconvergent expectations” — which is to say, there is a risk that interdependent decisions between stages will fail to be made in a compatible way (see Section 1.5 of Chapter 5). Lack of common information among independent stage I and stage II specialists with respect to the market opportunities which they confront and doubts regarding the true investment intentions and contractual reliability of other parties are the apparent impediments to effective coordination. Ultimately, however, the problems are to be attributed to the human and environmental factors described in Chapter 2.
To be sure, the argument has no special monopoly power significance unless the industry in question is already very concentrated62 or, in less con- centrated markets, conditions of effective collusion, which include collective refusal to deal, obtain. In such circumstances, however, actual competition, by itself, cannot be expected to self-police the market in a way that reliably assures the competitive outcome. Accordingly, potential competition has an important role to play. If potential entrants regard imitation of prevailing vertical structures as contributing importantly to the prospect of successful entry (as they may in highly concentrated industries), vertical restrictions that require funds to be raised by less, rather than more, experienced firms can have entry impeding consequences.
The financing issue, thus, is not that capital markets perversely avoid earnings opportunities, which is the test proposed by Bork, or that financing cannot be arranged under any terms whatsoever, to which Bowman refers. Rather, the cost of capital is at issue. If a prospective new entrant has the self- financing to come in at one stage (or can raise the capital at reasonable terms, perhaps because of a proven capability at this stage of operations) but lacks the self-financing and incurs adverse terms should he attempt to raise the capital to come in at the second stage, the condition of entry can clearly be affected by pre- existing vertical restrictions.
3. Circumventing Regulation
A third type of antisocial consequence is sometimes imputed to regulation: circumventing regulation. This can occur either at a general level (often in the context or wartime restrictions) or in connection with public utility regulation. As Coase (1952, pp. 338-339) and Stigler (1968, pp. 136- 137) have both pointed out, vertical integration is sometimes employed as a device by which to evade sales taxes, quota schemes, and other methods of nonprice rationing. Since such efforts by the government to interfere with the price mechanism typically apply to market-mediated but not to internal transactions, the shift of such transactions from the market to the firm serves to circumvent these regulatory schemes. This is perfectly straightforward and is derived from received microtheory without appeal to transaction-cost considerations.
Conventional microtheory can also be made to address the following issue: Can a regulated firm that is permitted only a “fair” rate of return in supplying final goods or services effectively evade the regulatory restraint by integrating backward into own-equipment supply? As Dayan (1973) has shown, such backward integration will permit the regulated industry to earn monopoly profits if either equipment transfer prices or the rate of return at the equipment supply stage is unregulated. At least some regulatory commissions appreciate this and have effectively extended regulation to include equipment supply; Michigan Bell is an example (Troxel, 1966, p. 168).
Backward vertical integration in regulated firms can also lead to technical progress distortions. As is generally acknowledged, rate of return regulation introduces a static bias in favor of capital intensive techniques. The intertemporal extension of this bias is to favor technical change of a capital intensive kind.
Of course such intertemporal biases can be transmitted across a market, in a derived demand fashion, as well as between stages of a regulated firm that is integrated backward into own equipment supply. I submit, however, that it is apt to be more pronounced in the latter. Since the integrated firm’s equipment business is not really “up for bid,” it is less apt to be confronted with the need to choose publicly between alternative equipment techniques. The nonintegrated firm, by contrast, does need to make and defend such choices—assuming, of course, that there are at least several alternative sources of supply from which to select. In circumstances where relatively more labor intensive options objectively represent least- cost techniques* this is more likely to be publicly exposed, and hence such techniques adopted, in the context of nonintegrated equipment supply.
The above assumes that competition at the equipment supply stage is feasible. Where it is not, backward vertical integration, for the reasons given in Chapter 5, is arguably efficient. If, however, but for such backward vertical integration, competitive equipment supply at some future time would develop, a tradeoff between monopolistic distortions in present supply and regulatory distortions in future supply needs to be faced. Concerns with backward vertical integration are obviously attenuated to the extent that one has confidence that regulation is or will become effective across both stages.
Source: Williamson Oliver E. (1975), Markets and hierarchies: Analysis and antitrust implications, A Study in the Economics of Internal Organization, The Free Press.