The technologies and contractual options discussed above are displayed sche- matically in Figure 7-2. If pˆ < p1 < p‾, then the relevant nodes are A and C: e uyer will either ask the supplier to employ the general purpose techno ogy an will pay p, upon delivery of product for which orders are’con- lrme , or e will ask that the supplier make specific investments for which die buyer offers safeguards. If instead pˆ < p‾< p1, then the relevant nodes are B and C: Only the special purpose technology will be employed, with respect to which some buyers will offer safeguards while others will not.
The argument that buyers can affect the terms and manner of supply by offering (or refusing to offer) hostages has ramifications for Robinson-Patman price discrimination and to an understanding of franchising and two-part pricing.
The Robinson-Patman Act has been interpreted as an effort “to deprive a large buyer of [discounts] except to the extent that a lower price could be justified by reason of a seller’s diminished costs due to quantity manufacture, delivery, or sale, or by reason of the seller’s good faith effort to meet a competitor’s equally low price.” Plainly, thatp is less thanp in the hostage model has neither quantity nor meeting competition origins. Neither is it contrary to the public interest. Indeed, it would be inefficient and unwarranted for a producer to charge the same price to two customers who order an identical amount of product, but only one of which offers a hostage, if (1) investments in specialized assets aie required to support the transactions in question, or (2) if, because of a refusal to make a credible commitment, transactions of the second kind are produced with a general purpose (but high- cost) technology.
The missing ingredients, plainly, are the differential commitment to buy (as reflected by the willingness to offer hostages) and the differential incentives to breach once hostages have been posted. The confusion is explained by the propensity to employ conventional (steady state) microtheory to the neglect of transaction cost aspects. Rectifying that involves examination of the microanalytics of transactions, with special reference to asset specificity and the hazards thereby posed, and evaluation of alternative contracts with respect to a common reference condition, prospective breakeven being a useful standard. Once that is done, a different understanding of many nonstandard or unfamiliar contracting practices, many of which are held to be presumptively unlawful, frequently emerges.
Klein and Leffler (1981) argue that franchisees may be required to make investments in transaction specific capital as a way by which to safeguard the franchise system against quality shading. As Klein (1980) puts it, franchisers can better
. . . assure quality by requiring franchisee investments in specific . . . assets that upon termination imply a capital loss penalty larger than can be obtained by the franchisee if he cheats. For example, the franchiser may require franchisees to rent from them short term (rather than own) the land upon which their outlet is located. This lease arrangement creates a situation where termination can require the franchisee to move and thereby impose a capital loss on him/up to the amount of his initial nonsalvageable investment. Hence a form of collateral to deter franchisee cheating is created, [p. 359].
The arrangement is tantamount to the creation of hostages to restore integrity to an exchange.
That logic notwithstanding, the use of hostages to deter franchisees from exploiting demand externalities is often regarded as an imposed (top down) solution. Franchisees are “powerless”; they accept hostage terms because no others are available. Such power arguments are often based on ex post reasoning. That the use of hostages to support exchange can be and often is an efficient systems solution, hence is independent of who originates the proposal, can be seen from the following revised sequence.
Suppose that an entrepreneur develops a distinctive, patentable idea that he sells outright to a variety of independent, geographically dispersed suppliers, each of which is assigned an exclusive territory. Each supplier expects to sell only to the population located within its territory, but all find to their surprise (and initially to their delight) that sales are also made to a mobile population. Purchases by the mobile population are based not on the reputation of individual franchisees but on customers’ perceptions of the reputation of the system. A demand externality arises in this way.
Thus, were sales made only to the local population, each supplier would fully appropriate the benefits of its promotional and quality enhancement efforts. Population mobility upsets this: Because the cost savings that result from local quality debasement accrue to the local operator while the adverse demand effects are diffused throughout the system, suppliers now have an incentive to free ride off of the reputation of the system. Having sold the exclusive territory rights outright, the entrepreneur who originated the program is indifferent to those unanticipated demand developments. It thus remains for the collection of independent franchisees to devise a correction themselves, lest the value of the system deteriorate to their individual and collective disadvantage.
The franchisees, under the revised scenario, thus create an agent to police quality or otherwise devise penalties that deter quality deterioration. One possibility is to return to the entrepreneur and hire him to provide such services. Serving now as the agent of the franchisees, the entrepreneur may undertake a program of quality checks (certain purchasing restraints are introduced, whereby franchisees are required to buy only trom qualified suppliers; periodic inspections are performed). The incentive to exploit demand externalities may further be discouraged by requiring each franchisee to post a hostage and by making franchises terminable.
This indirect scenario serves to demonstrate that it is the system that benefits from the control of externalities. But this merely confirms that the normal scenario in which the franchisor controls the contractual terms is not an arbitrary exercise of power. Indeed, if franchisees recognize that the demand externality exists from the outset, if the franchisor refuses to make provision for the externality in the original contract, and if it is very costly to reform the franchise system once initial contracts are set, franchisees will bid less for the right to a territory than they otherwise would. It should not therefore be concluded that perceptive franchisors, who recognize the demand externality in advance and make provision for it, are imposing objectionable ex ante terms on unwilling franchisees. They are merely taking steps to realize the full value of the franchise. Here as elsewhere, contracts must be examined in their entirety.
3. Two-Part Pricing
Victor Goldberg and John Erickson describe an interesting two-part pricing scheme that they observed in the sale of coke. The producer both sold coke to the calcincr and owned and leased the land upon which the plant of the calciner was built. Inasmuch as the coke was sold for “about one-quarter the current market price of equivalent quality coke (1982, p. 25), Goldberg and Erickson conjecture that “the rental rate was above the fair market rate and that the contract was designed to ensure that the calciner would continue to perform” (p. 25). Assuming that marginal costs are much less than average, such an arrangement can be interpreted as one by which the parties are attempting to strike efficient pricing terms that approximate those of the hostage model.
The pricing of utility services, whereby ex ante installation fees are paid by subscribers, also has interesting two-part pricing attributes.16 The risk that sellers will expropriate buyers upon receipt of advance payment can be mitigated by creating a specialized third party, which for convenience may be referred to as a regulatory commission (Goldberg, 1976a). Utilization of utility services can be priced so as to more nearly approximate marginal cost.
More generally, Goldberg and Erickson conjecture that nonlinear pricing schemes are much more widespread than is commonly believed ./They further point out that such arrangements are often very subtle and wilTrfequire detailed knowledge of contracts to investigate (1982, pp. 56-57).
Source: Williamson Oliver E. (1998), The Economic Institutions of Capitalism, Free Press; Illustrated edition.