The inhospitality tradition to which 1 referred earlier9 held that nonstandard modes of contracting were presumptively anticompetitive. The argument, moreover, was very sweeping. No effort was made to delimit applications to a subset of activity where the anticompetitive concerns were thought to be especially severe. Rather, customer, territorial, and related contract restraints were held to be presumptively unlawful, without qualification.
That policy position was based on two lines of argument. The affirmative argument was that rivals, distributors, customers, and so on are somehow “disadvantaged” when nonstandard contracting is employed. That was but-tressed by the view that true economies take a technological form, hence are fully realized within firms. Since there is nothing to be gained by introducing nonstandard terms into market-mediated exchange, the use of contract restraints was presumed to have anticompetitive purpose and affect.
Both lines of argument are related and mistaken. The notion that all relevant economies have technological origins is the more obviously mistaken of the two. At best it is a convenient fiction, as is surely evident from the contracting schema in Chapter 1 (to which reference has been made repeatedly throughout the book).
To prohibit contract restraints for trades that are supported by specific investments is to insist, in effect, that all k > 0 contracts be of a node B kind. That is patently inefficient in circumstances where effective contractual safeguards of a node C kind can be fashioned. It bears repeating, moreover, that price and governance structure are determined simultaneously, in an internally consistent relation to each other.
That last introduces contracting in its entirety considerations. It is easy to conclude, upon examining a contract at a point in time, that one of the parties to the exchange is disadvantaged by the restraint—in the sense that the restrained party would behave differently if the restraint were removed. Thus franchisees would frequently exercise the option to buy supplies (product; replacement parts) from unauthorized suppliers if it were permitted. That supposedly demonstrates that manufacturer insistence that purchases be made only from authorized suppliers is one-sided and anticompetitive.
Such a myopic conception fails to recognize that the tertns under which the original franchise was struck reflect the associated restraints. It is understandably attractive to have your cake (low price) and eat it too (no restrictions). But both the theoty and the practice of contract preclude that.
The Schwinn case, which was argued in 1966 and decided in 1967, reflects those confusions. The main arguments and their premises áre examined in section 6 of Chapter 4. With one exception, they will not be repeated here. Consider, however, the government’s views on vertical integration wrsus vertical restraints—“a rule that treats manufacturers who assume the distribution function themselves more leniently than those who impose re-straints on independent distributorn merely reflects the fact that, although integration in distribution sometimes benefits the economy by leading to cost savings, agreements to maintain resale prices or to impose territorial re-strictions of unlimited duration or outlet limitations of the type involved here have never been shown to produce comparable economies.” The clear or internal over market modes of organization is consonant with the then prevailing preoccupation with technological features and the corresponding disregard for transaction costs.
That orientation did not withstand subsequent criticism. The mistaken reasoning of Schwinn was corrected only a decade later when the Supreme Court decided the GTE-Sylvania case. The Court held that
[vertical] restrictions, in varying forms, are widely used in our free market economy. [Moreover, while] there is substantial scholarly opinion and judicial authority supporting their economic utility, [t]here is relatively little authority to the contrary. Certainly there has been no showing in this case, cither generally or with respect to Sylvania’s agreement, that vertical restrictions have or are likely to have a “pernicious effect on competition” or that they “lack … any redeeming virtue. ’ . . . Accordingly, we conclude that the per se rule in Schwinn must be overruled.
The intellectual basis for assessing the merits of alternative modes of organization evidently experienced substantial changes in the ten-year interval between those two opinions. Public policy was transformed as a consequence.12 Subsequent revisions in public policy toward price discrimination, franchise restrictions, reciprocity, basing point systems, block booking, and the like are also to be anticipated if recent scholarship on those matters is equally persuasive.
Lest the affirmative case for vertical restrictions become the new orthodoxy, however, it should not be concluded that such restrictions are unproblematic. For one thing, there is the usual caveat that vertical restrictions can be and sometimes are used to support horizontal cartels. Resale price maintenance, for example, can serve dealer cartel purposes; and vertical restrictions can also serve to regularize a manufacturers’ cartel (the facilitating practices doctrine). But the issues go deeper than that.
Thus consider a vertical restriction that has two effects: It helps to mitigate free rider effects, and thus restores incentives to engage in valued promotional and related sales and service activity, and it serves as a device by which to price discriminate. The first effect is generally in the public interest. The second may be, but it need not. As I have discussed elsewhere, efforts to monetize consumers’ surplus can yield net private gains and net social losses if the transaction costs that attend those efforts are substantial (Williamson, 1975. pp. 11-13). Specifically, three effects of price discrimination have to be distinguished: (1) What had been consumers’ surplus in a uniform pricing regime is monetized (let this be (7,); (2) net revenue is further augmented by the sale of added product made possible by price discrimination (let this be V-,, and assume, for convenience, that price discrimination is perfect); and (3) transaction costs are incurred in introducing and policing the practice by which perfect price discrimination is achieved (let this be T). Net private gains will then obtain if Δπ= V1 + V2 – T > 0, but a social gain will be realized only if ΔW = V2 — T > 0. The possibility that Δπ > 0 and ΔW < 0 must thus be admitted. The monetization of consumers’ surplus on intramarginal product is the troublesome factor that yields this mixed result.
Source: Williamson Oliver E. (1998), The Economic Institutions of Capitalism, Free Press; Illustrated edition.