10 December 2009
United States antitrust law condemns very few types of agreements between businesses outright. That per se status is generally confined to agreements to restrain price or output that are made between participants at the same level of a market - i.e., conspiracies among competitors. Most other claims asserting that a business agreement unlawfully restrains trade under Section 1 of the Sherman Act are evaluated under the rule of reason.
Rule of reason analysis requires an assessment of the actual or likely anticompetitive effects of the challenged practice. For practical purposes, a showing of anticompetitive effects requires a showing of market power. That is because few arrangements that do not involve collusion among competitors can affect the competitive process in the absence of market power. Consequently, if a party with only a small market share engages in a contracting practice, rule of reason analysis ends at the threshold.
In practice, rule of reason litigation often founders on the plaintiff's inability to prove market power. A recent decision by the US Court of Appeals for the Ninth Circuit, however, may make such rule of reason claims substantially easier to maintain.
On April 3, 2009, a divided panel of the Ninth Circuit revived a decade-long lawsuit brought on behalf of a putative class of wholesale gasoline purchasers who had sued nine petroleum companies over their hedging practices involving a special blend of gasoline required by the California Air Resources Board (CARB gasoline). See William O. Gilley Enterprises, Inc. v. Atlantic Richfield, Inc., 561 F.3d 1004 (9th Cir. 2009). The defendants had won summary judgment in an almost identical lawsuit in California state court. That action had alleged that the refiners had conspired to limit supply and raise prices by entering into a series of bilateral exchange agreements that effectively kept CARB gasoline off the spot market.
For once, however, federal court - and federal antitrust law - proved more plaintiff-friendly than California state court. The Ninth Circuit held that market power analysis can aggregate the market share of all counterparties to a single party's contracts, even in the absence of a broader conspiracy to restrain prices or output. Thus, the plaintiffs were entitled to pursue their theory that the refiners' bilateral exchange agreements, while made without collusion, had the aggregated effect of raising prices above competitive levels and therefore violated Section 1 of the Sherman Act under the rule of reason.
The Ninth Circuit recognized that a rule-of-reason claim usually requires proof of market power in order to show that the agreement has anticompetitive effects. Breaking new ground, however, the panel majority held that the effects of each defendant's multiple exchange agreements could be aggregated. This ruling, in turn, allowed market power to be shown by aggregating the market share of the other defendants who were counterparties to a single defendant's series of hedging agreements. That is, even though each defendant lacked market power individually, the claim could go forward, even in the absence of collusive action, so long as the defendants' collective market power was sufficient to threaten anticompetitive effects that might violate Section 1 under the rule of reason.
The effect of the Ninth Circuit's decision could be limited to its unusual factual setting, which involved a concentrated market in which each of the participants entered into hedging agreements with some or all of the others. All of the producers had large sales commitments to their respective distribution networks. Because the market was subject to artificial regulatory limitations, there were no independent producers. Consequently, individual producers had no alternative source of supply except one another. Thus, producers' options for hedging their own sales requirements were largely limited to purchases from their competitors.
But the aggregation principle recognized in this decision instead could have much broader consequences for antitrust litigation. The Ninth Circuit's decision may provide a new basis on which to bring common industry contracting practices under closer antitrust scrutiny. This is because its reasoning seems but one step away from allowing the independent parallel action of a variety of noncolluding market participants to be treated together for purposes of market power analysis.
In 1991, the California Air Resources Board (CARB) adopted regulations limiting the sale of gasoline in California to a new, cleaner burning, but more expensive formulation of gasoline - CARB gasoline - beginning in 1996. That same year, a putative class of retail CARB gasoline purchasers sued nine petroleum companies in California state court (the Aguilar action), asserting violations of California's Cartwright Act The suit alleged that the defendants had conspired to restrict output and raise prices of CARB gasoline by entering into a series of bilateral exchange agreements to ensure that each refiner had an outlet for any surplus and a source of supply to make up any shortfall. Plaintiffs argued that the agreement had effectively prevented most CARB gasoline from reaching the spot market. Two years later, in 1998, Gilley Enterprises filed a Sherman Act complaint in federal court, mirroring the allegations in Aguilar against the same defendants. The federal action was stayed pending resolution of the state court lawsuit.
In 2001, the California Supreme Court upheld a grant of summary judgment to the defendants in Aguilar on the ground that there was no evidence of a conspiracy or collusive action. The defendants subsequently moved for summary judgment in Gilley on the ground of issue preclusion (collateral estoppel). The district court agreed, but granted Gilley leave to amend its complaint.
Gilley's amended complaint alleged that 44 CARB gasoline exchange agreements had the effect of unreasonably restraining trade even in the absence of a conspiracy. The district court dismissed the complaint because Gilley could not articulate how any individual exchange agreement could have anticompetitive effects, as each agreement accounted for only a small percentage of the relevant market.
After the Ninth Circuit remanded for a further opportunity to replead, Gilley filed another amended complaint, relying for its theory of anticompetitive effects on the aggregate market power of each individual refiner-defendant and all other refiners who entered into exchange agreements with it: ''[t]hrough the use of [the defendant's] exchange agreements, coupled with its own refining capacity and that of its contracting partners, [defendant] has obtained sufficient market power to limit the supply of CARB gas to unbranded marketers and to raise the price.''
The district court again dismissed the complaint, holding that Gilley's allegation of the existence of a network of exchange agreements that allowed defendants to coordinate their production and output to limit the amount of CARB gasoline on the spot market ''is, at its core, a conspiracy claim.'' In the district court's view, even if the exchange agreements could be aggregated, there still was no causal connection between the exchange agreements and anticompetitive effect in the absence of a conspiracy.
The court provided a hypothetical: producer A enters into separate exchange agreements with producers B, C, D, E and F. If producer B overproduces gasoline, A may be able to take the excess amount and adjust its own production accordingly. But in the absence of coordinated action among the defendants, one producer cannot have enough control over the refining capacity in its geographic area to keep gasoline out of the spot market and away from unbranded sellers. This is because producers C, D, E and F may also produce excess gasoline that cannot be absorbed by A because A has already taken the overproduction from B. Without further agreement among the producers, the alleged market distortion could not occur.
The Ninth Circuit's Decision
The Ninth Circuit reversed. The court of appeals agreed with the district court that Aguilar precluded a claim that depended upon proof of collusion by the defendant oil companies to control supply and prices. But the majority held that Gilley had stated a valid antitrust claim by pleading that, ''without a conspiracy,'' the exchange agreements, ''when aggregated together,'' had “an anti-competitive effect on competition in the relevant market.''
The Ninth Circuit majority held that the district court should have permitted Gilley to allege the cumulative effects of a single defendant's exchange agreements to show that defendant's market power and anti-competitive effect. Drawing on the analogy of a single party's array of tying agreements with different consumers, the Ninth Circuit held that antitrust law generally permits ''the aggregation of multiple contracts when evaluating the legality of an individual contract.'' The majority declined to limit that principle to exclusive dealing and tying contracts that could be aggregated to show the market power of the single party that imposed those contracts upon its customers. Rather, the court recognized ''no general rule [that] requires that only the easiest cases may be aggregated.''
The court also held that it was improper on a motion to dismiss for the district court to ''''probe the soundness of Gilley's economic theory''' to find that the claim was, at its core, a conspiracy claim. According to the panel, ''[t]he district court may be correct in its understanding of how the economy or the oil business works, but that is a factual assessment not left to the court, even a savvy judge, to decide on a Rule 12 motion.''' Moreover, court of appeals disapproved of the district court's reliance on hypothetical reasoning that applied common sense to the allegations in the complaint. In the panel's view, the courts must accept at face value the complaint's conclusory allegations: i.e., that anticompetitive effects have resulted from the exchange agreements in the absence of collusion, even if the conclusion made no sense in light of the facts pleaded.
The panel was not disturbed by the US Supreme Court's recent requirement in Bell Atlantic Corp. v. Twombly that pleaded antitrust theories be plausible, not merely possible. Instead, the panel viewed Twombly as simply reaffirming the principle that ''[e]ven if ... a savvy court view[s] actual proof of the facts pleaded in the [complaint] as improbable and conclude[s] that a recovery is remote and unlikely, the complaint should still proceed.''
Judge Callahan dissented. She believed that Aguilar precluded the Gilley complaint because, to make sense, the new complaint necessarily relied on collusion. More broadly, she believed that Twombly precluded the majority's indulgence of the complaint's counterintuitive theory of noncollusive anticompetitive effects. And most important, she rejected the use of aggregation to rescue the complaint.
Judge Callahan agreed that Gilley could aggregate each defendant's contracts to determine that defendant's market power. But she rejected the notion that Gilley could aggregate all of the bilateral agreements by all of the defendants. Judge Callahan viewed the complaint's theory as necessarily reliant on market-wide aggregation. It is not clear that the majority's reasoning would permit market-wide aggregation in every setting. In the specific circumstances of CARB gasoline production, however, market-wide aggregation results from the panel majority's holding that Gilley could aggregate the market shares of all participants in each defendant's agreements.
Judge Callahan did not believe that aggregation would salvage Gilley's claim, no matter what market share resulted from the aggregation of each defendant's exchange agreements. In her view, individual exchange agreements, unlike exclusive dealing contacts, do not inherently restrain trade. The use of exchange agreements to avoid selling excess gasoline on the spot market is consistent with each defendant's economic self-interest because spot markets produce lower returns than do sales to branded dealers. As a result, aggregation at most would produce a complaint that pleaded market power without pleading any unlawful conduct. In Judge Callahan's view, a ''narrow focus'' on the effect of the various bilateral agreements ''would convert self-interest parallel action, similar to that found to be legal in Twombly, into an antitrust violation.''
Implications of the Decision
The Gilley decision raises many more questions than it answers. The Ninth Circuit revived allegations that had failed in the accommodating California state courts because there was no evidence of any agreements among suppliers to use the hedging agreements to reduce competition.
The decision might have the relatively narrow effect of permitting aggregation only of a single market participant's contracts in the market power analysis. In that interpretation, the Gilley claims survived only because each defendant contracted with most of the other suppliers in these reciprocal hedging agreements. That is, contracts with parties who were not also competing suppliers might not have had the effect of inflating a single defendant's market power.
On the other hand, the breadth of the panel majority's reasoning could sustain broad new antitrust theories that have little apparent economic basis. Such interpretations could support Section 1 violations based on industry-standard contracting practices, regardless of how indistinct or unsubstantiated their threat to competition. It is easy to imagine at least some trial courts giving the Ninth Circuit's decision this broader reading.
The Gilley decision has two significant implications.
First, in rejecting the district court's role in screening a pleaded theory for economic common sense, the Ninth Circuit undermines the requirement of plausible antitrust pleading in Twombly - a requirement of plausibility that the US Supreme Court has since extended to all civil actions. The majority gave a strikingly narrow construction to the Twombly directive to dismiss antitrust claims that are not plausible in light of basic economic principles. Twombly emphasized that a complaint fails if ''it stops short of the line between possibility and plausibility of 'entitle[ment] to relief.''' Yet the panel treated any analysis of plausibility as purely factual and thus inappropriate at the pleading stage.
In contrast, many other courts of appeals have recognized that Twombly requires a determination of plausibility on a motion to dismiss. The Gilley majority reiterated older precedent holding that dismissals are ''disfavored in antitrust actions.'' But Twombly appeared to have put that disfavor to rest. In finding that a complaint alleging merely parallel conduct did not sufficiently plead a conspiracy, the Supreme Court explained, ''it is only by taking care to require allegations that reach the level suggesting conspiracy that we can hope to avoid the potential enormous expense of discovery in cases with no 'reasonably founded hope that the [discovery] process will reveal relevant evidence' to support a § 1 claim.''
The Ninth Circuit's majority also departed from the spirit, if not the text, of Twombly in chiding the district court for considering whether the asserted anticompetitive effects could occur without a conspiracy. Such a common-sense analysis would seem necessary to ensure that a claim has crossed ''the line from conceivable to plausible.'' Indeed, the Supreme Court explained that the sufficiency of a complaint ''turns on the suggestions raised by [the alleged] conduct when viewed in the light of common economic experience." Yet the Gilley majority found that the district court had exceeded the appropriate judicial role when it attempted ''to determine the soundness of Plaintiffs' economic theory.'' Common sense, it seems, has no place at the pleading stage.
While courts must take as true any allegations of historic fact, it is unclear how a court could determine whether a pleaded entitlement to relief is plausible, rather than merely possible, without probing the soundness of the plaintiff's economic theory. More specifically, the theory adopted by the Ninth Circuit appears in substantial part to provide an end-run around Twombly. One of the central holdings of that decision was that parallel conduct “is just as much in line with a wide swath of rational and competitive business strategy unilaterally prompted by common perceptions of the market” as with a conspiracy. And such ''unilaterally prompted'' conduct is usually analyzed on the basis of each individual actor's market power, if any.
In line with this analysis, the Gilley district court considered whether the pleaded agreements actually reflected anticompetitive control of the market. With neither allegations of coordinated action nor allegations that the exchange agreements required a contracting partner to produce or purchase a pre-established amount of CARB gasoline, it concluded that no defendant could control the output of any of its contracting partners. Moreover, as the Ninth Circuit dissent noted, there are sound economic reasons why, even without collusion, an oil refiner is strongly motivated to avoid having to sell gasoline on the spot markets. Spot markets exist only when a refiner produces more gasoline than can be sold by its branded dealers. They result from refiner inefficiencies and they produce lower returns than the refiner can obtain through its branded dealers. Given these economic realities, then, the theory of noncollusive anticompetitive effects, as opposed to individual interest-seeking, does not make economic sense. The Gilley majority's refusal to recognize any meaningful judicial role in assessing the economic plausibility of the allegations in an antitrust complaint will cause further confusion for litigants and trial courts while prolonging the litigation of meritless claims.
Second, the Ninth Circuit's decision is troubling in that it apparently permits the aggregation of contracts among different parties not acting in concert to determine market power. The decision suggests that the effects of a single market participant's agreements may be aggregated not only to determine that defendant's market power, but also to attribute to it the market power of all entities with which it contracts. If horizontal competitors buy and sell from each other, for example, then their market shares may be aggregated for purposes of antitrust analysis, even in the absence of any collusion between them at either the buyer level or the seller level.
As the dissent pointed out, the complaint appears to allege the aggregation of the contracts of all the defendants, and the majority opinion did not clearly distinguish between aggregating the contracts of one defendant and aggregating the contracts of all defendants (perhaps because each defendant contracted with most or all of the others). Aggregating the contracts of all defendants would include the entire market whenever an entire industry follows a standard contracting practice - here, a hedging strategy. Yet no real ''power'' would result from the standard practice, since no individual defendant - and no coordinated group - would control output and pricing.
Aggregation is commonly used to determine the probable competitive effects of a merger by considering the impact on the competitive landscape of two firms acting in a coordinated manner. And it makes sense to aggregate a single defendant's contracts to evaluate that defendant's ability to affect competition through restrictive practices. Thus, a single defendant's series of tying agreements or unreasonably long exclusive dealing arrangements might be treated together. But that aggregation would reflect the market power of a single market participant able to impose anticompetitive terms on its customers.
In Twin City Sportservice, Inc. v. Charles O. Finley & Co., for example, the Ninth Circuit aggregated the market share encompassed in all of a defendant's unreasonably long concession franchise agreements with follow-the-franchise clauses that, taken together, created barriers to entry and precluded competition within the franchise market. That type of aggregation treated a single defendant's pattern and practice of using exclusive deals to lock up the market incrementally, on the ground that a defendant should not be allowed to do piecemeal what it would be prohibited from doing simultaneously.
Similarly, the market effects of a single defendant's tying arrangements are generally treated as a whole, precluding the need to examine each tied sale independently. But even though tying and exclusive dealing must be evaluated under the rule of reason, their potential for anticompetitive effects is well-recognized (if somewhat controversial) where they foreclose sufficient proportions of a market. This is true, in part, because of the seeming coercion exercised to demand agreement to the tie or exclusive deal.
It is difficult, in contrast, even to articulate how the bilateral exchange agreements seen in Gilley might have actual anticompetitive effects. This is especially the case because the agreements did not require a contracting partner to purchase gasoline under the contract, nor did they limit a partner from selling its own gasoline on the spot market.
In the context of Gilley, all aggregation could show is one defendant's potential market power if it purchased all of its contracting partners' gasoline production. But that is neither the probable, nor even the possible, effect of those agreements, much less a plausible outcome. Rather, the hedging agreements provided refiners with definite prices for any excess production, along with definite access to supply any excess needs if their own production was inadequate.
Although rule-of-reason analysis is necessarily fact-specific, aggregation alone - that is, market power alone - shows only that a challenged practice, if anticompetitive, could have sufficiently widespread effects to harm the integrity of the market as a whole. Even at the pleading stage, however, a plaintiff should be required to articulate a plausible basis for believing that the challenged agreements restrained competition. It should not be sufficient merely to demonstrate that agreements of similar structure covered a substantial share of the market.
Other courts have applied this more stringent analysis. For example, in evaluating whether certain resale limitations in Anheuser-Busch's agreements with its distributors violated Section 1, the Eleventh Circuit took a very different approach to the proper role of aggregation in Section 1 allegations. The court of appeals in that case declined to aggregate Anheuser-Busch's individual distributor agreements for the purpose of showing market power and anticompetitive effect. The court cautioned that, if aggregation were permitted in the absence of a conspiracy among the distributors, ''aggregation of market share would always be required when reviewing vertical restraints.'' That practice, in turn, could threaten ''arrangements traditionally reviewed under the rule of reason, by making market power seem to appear where it does not really exist.'' As the Eleventh Circuit recognized, weakening the market power screen in this way could subject a much wider variety of neutral agreements and business practices to a fact-specific analysis of their competitive effects.
The Ninth Circuit's decision, however, does exactly that. In reality, each refiner's agreements give it power only over its own production and whatever additional purchases it may need to make. In the absence of collusion, an individual oil refiner has no control over the production, supply or pricing of its contracting partners. An analysis that aggregates the market shares of its exchange agreement partners, however, makes it appear that the single refiner has market power to affect prices merely because it can purchase gasoline from other refiners.
The Gilley decision may be short-lived. After the defendants petitioned for rehearing, the court swiftly called for a response and has since considered the petition for nearly six months. But if Gilley remains in place - or if other courts adopt its reasoning - companies facing rule of reason claims may find that a complaint's economic implausibility will not necessarily result in early dismissal. And unless the reciprocal nature of the exchange agreements in Gilley provides a strict limiting principle, a more advanced rule of reason inquiry may result whenever many companies within an industry use similar contracts.
As we were going to press, the Ninth Circuit panel withdrew its opinion and issued a new decision affirming the dismissal. The new opinion rests largely on gaps in the pleading, while offering dicta that echoed many of the ideas in the withdrawn opinion. We will explore the new opinion in the next edition of the Antitrust Review.