Credible Commitments: Petroleum Exchanges

“The task of linking concepts with observations demands a great deal of detailed knowledge of the realities of economic life” (Koopmans, 1957, p. 145). The phenomenon of petroleum exchanges has puzzled economists for a very long time. It routinely comes up in antitrust cases and investigations. The 1973 case brought by the United States Federal Trade Commission against the largest petroleum firms maintained the view that exchanges were instrumental in maintaining a web of interdependencies among those firms, thereby helping to effect an oligopolistic outcome in an industry that was relatively unconcentrated on normal market structure criteria. The more recent study on The State of Competition in the Canadian Petroleum Industry likewise.regards exchanges as objectionable. The Canadian Study, moreover, produces docu-ments—contracts, internal company memoranda, letters, and the like—as well as deposition testimony to support its views that exchanges are devices for extending and perfecting monopoly among the leading petroleum firms. Such evidence on the details and purposes of contracting is usually confidential and hence unavailable. But detailed knowledge is clearly germane—and often essential—to a microanalytic assessment of the transaction cost features of contract.

1. The Evidence from the Canadian Study 

Volume V of the Canadian Study deals with the refining sector. Arguments are advanced and supporting evidence is developed that interfirm supply arrangements permit the principal refiners to perfect oligopolistic restrictions in the following four respects:74 75 (1) valuable knowledge about investment and marketing plans of rivals are disclosed by such agreements (p. 56); (2) leading firms are able to control lesser firms by manipulating the terms of exchange (pp. 49-50); (3) competition is impaired by conditioning supply on the payment of an “entry fee” (pp. 53-54); and (4) exchange agreements impose limits on growth and supplementary supply (pp. 51-52).

The first two fail to pass scrutiny of the most rudimentary comparative institutional kind. Thus assuming that trade between rivals is efficient and that unilateral supply agreements (if not exchange) will be permitted, the objec- tionable information disclosures attributed to exchanges would presumably continue—since investment and marketing plans will be unavoidably disclosed in the process. Accordingly, evaluated in comparative institutional terms, the information disclosure objection is properly regarded as an objection to long- term trade of any kind. Exchanges are not uniquely culpable.

The suggestion that exchanges are anticompetitive because they permit firms to realize unfair bargaining advantages is similarly misplaced. The correct view is that firms should always be expected to realize such bargaining advantages as their positions lawfully permit. Absent a showing that exchanges are different from unilateral trades in bargaining respects, that objection is properly disregarded also.

The entry fee and marketing restraint objections are more substantial, however, and warrant elaboration.


The entry fee objection to exchanges is that this has foreclosure conse- quences. That such fees are required as a precondition for trade, or at least the sale of product at favorable prices, is set out in the Canadian Study as follows (pp. 53-54; emphasis added):

Evidence of an understanding that a fee relating to investment was required for acceptance into the industry can be found in the following quotation from Gulf: “We do believe that the oil industry generally, although grudgingly, will allow a participant who has paid his ante, to play the game; the ante in this game being the capital for refining, distributing and selling products.” (Document #71248, undated, Gulf)

The significance of the quotation lies equally in the notion that an “entry fee” was required and in the notion that the industry set the rules of the “game.” The meaning of the “entry fee” as well as the rules of the “game” as understood by the industry can be found in the actual dealings between companies where the explicit mention of an “entry fee” arises. These cases demonstrate the rules that were being applied—the rules to which Gulf was referring. Corhpanijs which had not paid an “entry fee,” that is, companies which had not made a sufficient investment in refining capacity or in marketing distribution facilities would either not be supplied or would be penalized in the terms of the supply agreement. [Emphasis added]


The Canadian Study notes that exchanges were made conditional on growth and territorial restraints and regards both as objectionable. The Imperial- Shell exchange agreement, under which Imperial supplied product to Shell in the Maritimes and received product in Montreal, is cited in both connections (p. 51):

The agreement between Imperial and Shell, originally signed in 1963, was renegotiated in 1967. In July 1972, Imperial did this because Shell had been growing too rapidly in the Maritimes. In 1971/72, Imperial had expressed its dissatisfaction with the agreement because of Shell’s marketing policies. Shell noted:

“There [sic] [Imperial’s] present attitude is that we have built a market with their facilities, we are aggressive and threatening them all the time, and they are not going to help and in fact get as tough as possible with us.” (Document #23633, undated, Shell)

Imperial renewed the agreement with Shell only after imposing a price penalty if expansion were to exceed “normal growth rates” and furthermore stipulated that “Shell would not generally be allowed to obtain product from third party sources” to service the Maritimes (p. 52).

Gulf Oil likewise took the position that rivals receiving product under exchange agreements should be restrained to normal growth: “Processing agreements (and exchange agreements) should be entered into only after considering the overall economics of the Corporation and should be geared to providing competitors with volumes required for the normal growth only.”76 It furthermore sought and secured assurances that product supplied by Gulf would be used only by the recipient and would not be diverted to other regions or made available to other parties (p. 59).

2. Interpretations 

These practices are subject to several interpretations. One is that the entry fees and marketing restraints are both anticompetitive. A second is that efficiency purposes are arguably served, especially by the former. A third is that there are mixed effects.


The two polar contracting traditions for evaluating nonstandard or un- familiar contracting practices are the common law tradition and the antitrust or inhospitality tradition. Whereas contractual irregularities are presumed to serve affirmative economic purposes under the common law tradition, a deep suspicion of anticompetitive purposes is maintained by the antitrust (or inhospitality) tradition.

The inhospitality tradition belongs to the monopoly branch of contract and is supported by the widespread view that economic organization is tech- nologically determined. Economies of scale and technological non- separabilities explain the organization of economic activity within firms. All other activity is appropriately organized by market exchanges. Legitimate market transactions will be mediated entirely by price; restrictive contractual relations signal anticompetitive intent.

The authors of the Canadian Study are evidently persuaded of the merits of that tradition. Long-term trade among rivals of any kind is suspect. And exchanges, which represent an irregular if not unnatural contracting form, are especially objectionable. Not only do exchanges facilitate information disclosure and permit bargaining muscles to be flexed, but they are used punitively against nonintegrated independents who, because they have not paid an entry fee, are denied product on parity terms. Furthermore, the marketing restraints associated with exchanges are patently offensive.


Unlike the inhospitality traditiorv, the transaction cost approach is in the common law tradition. A comparative institutional orientation (Coase, 1964) is maintained. “Defects” are thus objectionable only where superior feasible alternatives can be described. Inasmuch as the information disclosure and bargaining concerns raised by the authors of the Canadian Study continue under unilateral trading, they are set aside, and attention is focused on entry fees and marketing restraints.

  1. Entry The entry fee issue is the matter of special interest to this chapter. Long-term exchange agreements permit firms to secure product in geographic markets where own production is not feasible because economies of scale are large in relation to their own needs. The amount of product in question may nevertheless be substantial. Firms with whom exchange agreements are reached will thus construct and maintain larger plants than they otherwise would. Specific investments in dedicated assets are made as a consequence of such agreements.

If supply agreements were of a unilateral kind and the buyer was unable or unwilling to offer a hostage, contracts of the kind described in Chapter 7 as type II would presumably be negotiated—whence the trading price would be p =fv2 + kl(i — p). If instead the contract is extended to include bilateral rather than unilateral trade, the contract is converted to one of type III. Although exchange agreements stipulate the physical flows of product, the effective price is p -v2 + k, which is less than p. Moreover, the parties have the incentive to exchange product so long as realized demand price in both regions exceeds v,,10 which is the marginal cost supply criterion. Assuming that demands in the two regions are highly correlated, the parties will nor­mally reach common decisions on the desirability of trade.”

  1. Marketing restraints. The supply and growth restraints discussed by the Canadian Study can be looked at in three ways. First, they can be viewed as a means by which to protect the exchange agreement against unilateral defection. Second, such restraints may serve strategic market division Third, restraints may serve to regularize markets. These are not mutually exclusive.

Only the first purpose is consonant with an efficiency interpretation. The argument here is that marketing restraints help to preserve symmetrical incen- tives. Such symmetry could be upset if one of the firms were to receive product in its deficit region from third parties. Such a firm might then be in a position to play one supplier off against the other. Or symmetry could be placed under strain if one party were to receive product from the other such that it began to grow “in excess of normal”—in which event it might be prepared to construct its own plant and scuttle the exchange agreement. Marketing restraints that help to forestall such outcomes encourage parties to participate in exchanges that might otherwise be unacceptable.


Monopoly explanations are commonly advanced when economists, law- yers, or other interested observers come across contractual practices they do not understand. Inasmuch as “we are very ignorant [in this field], the number of ununderstandable practices tends to be very large, and the reliance on a monopoly explanation frequent” (Coase, 1972, p. 67). A rebuttable presumption that nonstandard contracting practices are serving affirmative economic purposes, rather than monopoly purposes, would arguably serve antitrust law and economics better than the inhospitality presumption, which until recently has prevailed.

The presumption that exchanges have efficiency purposes could be chal-lenged on any or all of three grounds. First, it might be argued that exchanges are merely a clever device by which to deny product to nonintegrated rivals. Refusals to sell to nonintegrated firms on p terms would support that contention. (It is plainly unrealistic, however, for buyers that have not made credible commitments to expect to receive product at p.) Second, the market in question could be shown to have troublesome structural properties. The issue here is whether the requisite preconditions for market power—mainly high concentration coupled with high barriers to entry l3-Ve satisfied. A third would be that the preconditions for efficiency are not satisfied. Factors favorable to the efficiency interpretation are the following: The exchange should be of a long-term kind; the amount of product exchanged should represent a significant fraction of plant capacity; but economies of plant scale should be large in relation to the amount of product traded. Exchanges for a small quantity of product where economies of scale are insubstantial are much more problem- atic.

To be sure, exchanges might simultaneously serve efficiency and anti- competitive or other antisocial purposes. Here as elsewhere, where tradeoffs are posed, they ought to be evaluated.

Source: Williamson Oliver E. (1998), The Economic Institutions of Capitalism, Free Press; Illustrated edition.

Leave a Reply

Your email address will not be published. Required fields are marked *