10 December 2009
For the past several years, US courts have been struggling to articulate the limitations imposed by Section 2 of the Sherman Act on dominant firms seeking to engage in “bundling.” Bundling typically involves providing customers a discount when they purchase products spanning several different product lines. The bundled discounts most frequently challenged under Section 2 are those where the bundle includes both “monopoly products,” over which the dominant firm arguably wields monopoly power, and “competitive products,” relative to which the dominant firm faces competition.
The thrust of the antitrust challenge tends to focus on an allegation that the defendant is using its power in the monopoly market to attempt to monopolize the competitive market. Courts have reached drastically different conclusions on the legal standards applicable to this kind of conduct, which makes counseling in this area particularly difficult.
The Ninth Circuit recently further complicated the analysis of bundled discounts by suggesting that bundled discount cases may be resolved by the legal standards applicable to price squeezing claims. Its decision in Doe v. Abbott Labs., 571 F.3d 930 (9th Cir. 2009), highlights the continued ambiguity surrounding the legal standards applicable to bundled discounts by a dominant firm.
The Market for Boosted Protease Inhibitors
The plaintiffs in Doe alleged that Abbott Labs was attempting to monopolize the market for “boosted” protease inhibitors. Protease inhibitors are considered the most potent class of drugs available to fight the HIV virus. Norvir is the brand name for a patented compound called ritonavir. Abbott originally introduced Norvir in 1996 as a stand-alone protease inhibitor, with a recommended daily dose of 1,200 mg/day, priced at approximately $18. Later, it was discovered that small doses of Norvir used in conjunction with another protease inhibitor would boost the effectiveness of the other inhibitor. Norvir’s use as a booster also reduced the side effects associated with high doses of protease inhibitors and slowed the rate at which HIV developed resistance to the other protease inhibitors. When used as a “booster,” however, the recommended daily dose of Norvir was only about 100-400 mg/day. By 2003, the average price for a daily dose of Norvir had fallen to $1.71.
In 2000, Abbott introduced Kaletra, a single pill containing the protease inhibitor lopinavir as well as ritonavir, which is used to boost lopinavir’s effectiveness. Kaletra was effective and widely used, but it caused some patients to experience significant side effects.
Three years later, two new protease inhibitors were about to be released to the market: Reyataz, by Bristol-Myers Squibb, and Lexiva, by GlaxoSmithKline. Studies showed that when boosted with Norvir, these new drugs were as effective as Kaletra, and were more convenient to some patients. After Reyataz was successfully introduced into the market, Kaletra’s market share fell more than Abbott had anticipated. Moreover, the average daily dose of Norvir also fell. While prior protease inhibitors required 200-400 mg of Norvir per day to benefit from its boosting effect, Reyataz was effectively boosted by only 100 mg/day of Norvir.
On December 3, 2003, Abbott raised the price of Norvir from $1.71 to $8.57 per 100 mg, but kept the price of Kaletra constant. Abbott claimed that this price increase brought the price of Norvir closer to its considerable clinical value. A group of plaintiffs filed suit, claiming that the price increase violated Section 2 of the Sherman Act. The Section 2 claims centered on the allegation that Abbott was using its patent-conferred monopoly over Norvir to attempt to monopolize the boosted protease inhibitor market, in which Kaletra competed with Reyataz and Lexiva. A central component of the plaintiffs’ claims was the understanding that Kaletra, which effectively bundled Norvir with Abbott’s inhibitor lopinavir, was “discounted” to the point that competing firms could not offer customers a competitive price for boosted inhibitors.
The Legal Standards Applicable To Bundled Discounts
Much of the modern controversy over the legal standards governing bundled discounts by dominant firms results from the Third Circuit’s decision in LePage’s Inc. v. 3M, 324 F.3d 141 (3d Cir. 2003) (en banc). The plaintiff in LePage’s was a manufacturer of private label transparent tape. 3M manufactured Scotch tape and had a dominant share of the US transparent tape market. LePage’s argued that 3M engaged in a series of related actions, including offering certain types of bundled rebates, that were designed to restrict the availability of lower-priced transparent tape to consumers.
In the challenged bundled rebates, 3M offered discounts to major customers conditioned on purchases spanning six of 3M’s diverse product lines. For each product line, 3M fashioned growth targets specific to each customer. The size of the rebate was linked to the number of product lines in which the growth targets were met. Failure to meet the growth target in a single product line would cause the customer to lose the rebate across all of the products it had purchased from 3M.
3M argued that, pursuant to the US Supreme Court’s decision in Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 (1993), LePage’s was required to prove that 3M was selling tape below some measure of its costs. The Third Circuit, however, declined to adopt the per se rule of legality 3M sought. Pointing to its prior precedent in SmithKline Corp. v. Eli Lilly & Co., 575 F.2d 1056 (3d Cir. 1978), the Third Circuit noted that a monopolist’s decision to link a product on which it faced competition with products on which it faced no competition had long been subject to searching inquiry under Section 2 of the Sherman Act. The court therefore affirmed the jury’s Section 2 monopolization verdict despite the absence of evidence that 3M had priced tape below some measure of its costs.
The Ninth Circuit tackled the issue of bundled discounts in Cascade Health Solutions v. PeaceHealth, 515 F.3d 883 (9th Cir. 2008). Cascade involved allegations of anticompetitive bundled discounts in the markets for primary, secondary and tertiary hospital services in Lane County, Oregon. A jury found in favor of the plaintiff on claims — including Section 2 claims — attacking bundled discounts. On appeal, the Ninth Circuit issued an order inviting amicus briefing on the bundled discount issues, then conducted a comprehensive analysis of the applicable antitrust framework.
The court began with the premise that bundled discounts are pervasive and generally procompetitive. It noted, however, that it was possible for a monopolist to use a bundled discount to exclude a competitor of equal or superior efficiency from the market and thus to reduce consumer welfare in the long run. The court determined, moreover, that by relying on a broader range of products, a bundled discounter could achieve competitor exclusion without sacrificing any short-run profits.
Because the jury instructions in Cascade had been based upon the Third Circuit’s decision in LePage’s, the Ninth Circuit first had to determine if a Section 2 plaintiff could attack bundled discounts qualitatively, by demonstrating market foreclosure, or if a cost-based standard should apply instead. The court surveyed the views of commentators that antitrust claims attacking bundled discounts are similar in certain respects to predatory pricing claims and similar in other respects to tying claims. The Ninth Circuit ultimately concluded that “[a] bundled discount, however else it might be viewed, is a price discount on a collection of goods.” It therefore found particularly relevant a long line of Supreme Court rulings showing that courts evaluating price-cutting behavior must tread with particular care because “mistaken findings of liability would chill the very conduct the antitrust laws are designed to protect.” The Ninth Circuit found “the course safer for consumers and our competitive economy” would be to hold that Section 2 claims “cannot be satisfied by reference to bundled discounts unless the discounts result in prices that are below an appropriate measure of the defendant’s costs.”
Its decision in favor of a cost-based test for liability obliged the court to resolve several thorny questions about how to formulate such a test. Unlike single-product cases, in which courts simply ask whether the defendant priced its product below its incremental cost of producing that product, bundled discount cases present far more challenging scenarios. As noted above, a monopolist engaging in bundled rebates across several products can exclude equally efficient rivals without sacrificing short-term profits. Simply asking whether the defendant is pricing above its incremental costs therefore fails to capture potentially anticompetitive bundled rebates. With these considerations in mind, the court sought to evaluate the various rules of law proposed by the parties and the amici.
The defendant in Cascade and some of the amici advocated the adoption of an “aggregated discount” rule, under which bundled discounts could violate Section 2 only if the discounted price of the entire bundle was below the bundling firm’s incremental cost to produce the entire bundle. Such a test essentially ignores the multiple-product aspects of bundling and applies the single-product predatory pricing framework of Brooke Group. The court rejected that test, finding that multiple-product cases present different concerns that require a different mode of analysis.
The court then proceeded to consider the legal standard proposed in Ortho Diagnostic Sys., Inc. v. Abbott Labs., Inc., 920 F. Supp. 455 (S.D.N.Y. 1996). The Ortho test, designed specifically for multi-product cases, “deems a bundled discount exclusionary if the plaintiff can show that it was an equally efficient producer of the competitive product, but the defendant’s bundled discount made it impossible for the plaintiff to continue to produce profitably the competitive product.” The standard appreciates that a cleverly designed bundled discount can have the effect of excluding an equally efficient competitor from the marketplace. It therefore permits a plaintiff to prove that it was equally efficient in producing the competitive product – and that the challenged bundled discount must therefore necessarily rely on factors other than superior efficiency to foreclose competition.
The Ninth Circuit recognized that the Ortho test was better than the aggregated discount rule in identifying bundled discounts that threatened competition. Nevertheless, the court perceived two significant problems in administering the test. Both problems resulted from the fact that the Ortho test is subjective because the legality of a challenged bundled rebate depends upon the cost structure of the plaintiff challenging it.
First, the court was sensitive to the fact that the Ortho test would not provide particularly helpful guidance to sellers interested in offering procompetitive bundled discounts. Because dominant firms are unlikely to have access to information about their competitors’ costs, they could never be assured that a proposed bundled discount would not violate Section 2.
Second, the court was concerned that the Ortho test would necessitate multiple lawsuits challenging the same bundled discount. The court envisioned a scenario in which a monopolist offering a bundled discount faced competition from two firms, one of which was more efficient than the monopolist at producing the competitive product, while the other was less efficient. Under the Ortho test, the discount would violate Section 2 relative to the more efficient competitor but would be perfectly legal relative to the less efficient competitor. Thus, if the less efficient competitor brought an antitrust challenge and failed, the more efficient competitor would be forced to bring a second suit to obtain relief. The Ninth Circuit was skeptical of any rule “that might encourage more antitrust litigation than is reasonably necessary to ferret out anticompetitive practices.”
The court therefore decided to adopt what it termed a “discount attribution” standard amounting to an objective version of the Ortho test. Under that standard, which was advocated by several antitrust commentators and amici, the full amount of the bundled discounts given by the defendant is allocated to the competitive product. The bundled discount can violate Section 2 only if the resulting price of the competitive product is below the defendant’s incremental cost to produce it. By using the defendant’s own cost structure — rather than the plaintiff’s cost structure — to evaluate the legality of the bundled discount, the “discount attribution” standard will condemn only those bundled discounts that have the potential to exclude an equally efficient producer of the competitive product. The standard also avoids the practical counseling difficulties and multiple lawsuits that troubled the court relative to the Ortho test.
Finally, the Cascade court turned to the technical question of how to measure the defendant’s incremental cost of producing the competitive product. Relying on its prior precedent in the single-product context, the court determined that the appropriate measure of costs in bundled-discount cases was average variable cost.
Evaluating the Conduct in Doe as a Bundled Discount
It was against this background that the district court in Doe attempted to evaluate Abbott’s conduct. Abbott filed a motion to dismiss, arguing that plaintiffs’ claims were foreclosed by the Ninth Circuit’s analysis in Cascade. The district court denied the motion (see In re Abbott Labs. Norvir Antitrust Litigation, 562 F. Supp.2d 1080, 1091 (N.D. Cal. 2008)), relying on its analysis in the parallel case of Meijer, Inc. v. Abbott Labs., 544 F. Supp.2d 995 (N.D. Cal. 2008).
The court denied Abbott’s motion for two reasons. First, the court was skeptical that Abbott’s conduct could properly be analyzed as a bundled discount. The court deemed it “not readily apparent that Kaletra consists of two products at all — ritonavir and lopinavir are combined in a single pill.” Indeed, Abbott did not offer lopinavir as a stand-alone protease inhibitor, and the FDA licensed lopinavir for use only as a component of Kaletra. Because there was no independent price for lopinavir, the court deemed it impossible for Abbott to offer a “discount” on lopinavir when sold as part of Kaletra.
Second, the court held that even if Abbott’s conduct were to be analyzed under the rubric of bundled discounts, the Cascade test would not apply. The court found the test inapplicable because the stated goal of the Cascade test (making unlawful only those pricing schemes that would exclude equally efficient competitors from the market) would not be met by applying the rule to Abbott’s conduct. The court illustrated its reasoning with numbers from the record. Abbott charged $17.14 for 200 mg of Norvir. A dose of Kaletra containing the same amount of Norvir was priced at $18.78. Because the imputed price of lopinavir — the second component of the Kaletra compound — was $1.64, Abbott’s pricing could fail the Cascade test only if its average variable cost of producing lopinavir was greater than $1.64. But the cost of manufacturing Kaletra was no more than pennies per pill. Thus, Abbott’s pricing of Kaletra could virtually never be subject to liability under the Cascade test.
As the court observed, no newly developed protease inhibitor could ever be profitably sold at a price reflecting Abbott’s average variable cost of production, “because the manufacturer would never be able to recoup its huge research and development costs.” In other words, the court found that “unique structural characteristics of the pharmaceutical industry, where fixed costs in the form of investment in research and development dwarf variable costs,” rendered the Cascade test inapplicable.
The court briefly considered modifying the Cascade test to utilize a measure of cost that better addressed the realities of the pharmaceutical industry, but it found such an approach “difficult to implement in practice.” It therefore found the Cascade test an inappropriate measure of Abbott’s conduct because application of the test would immunize conduct that could drive an equally efficient competitor from the protease inhibitor market.
Following the district court’s decision, the parties entered into a high/low settlement. They agreed to base the ultimate settlement amount on how the Ninth Circuit resolved several questions on interlocutory appeal, including whether the Cascade test applied to Abbott’s conduct.
Evaluating the Conduct in Doe as a Price Squeeze
The Ninth Circuit defied the parties’ expectations by resolving the case on a different ground. According to the Ninth Circuit, the Supreme Court’s recent decision in Pacific Bell Telephone Co. v. Linkline Communications, Inc., 129 S. Ct. 1109 (2009), immunized Abbott’s conduct from antitrust liability.
Linkline involved allegations of a “price squeeze” in the wholesale and retail markets for the provision of DSL service. Internet service providers sued AT&T, alleging that the company’s high wholesale prices for DSL transport services, combined with its low prices for retail DSL service, had unlawfully squeezed their profit margins. The Court analyzed each aspect of the alleged squeeze separately.
First, relying on its recent decision in Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004), the Supreme Court found that AT&T’s wholesale prices were immune from antitrust scrutiny. FCC regulations required AT&T to provide DSL transmission services to the plaintiffs. But the Court determined that AT&T had no obligation under the antitrust laws to provide services to its rivals. Given that AT&T could, consistent with Section 2, refuse to provide DSL transmission services to the plaintiffs under any terms, it necessarily followed that Section 2 could not condemn providing those services at prices higher than the plaintiffs would have preferred.
Second, the Court evaluated plaintiffs’ claims that AT&T’s retail prices were too low. Noting the strong antitrust policies in favor of low prices to consumers, the Court found that low prices could violate Section 2 only if they were predatory under the Brooke Group analysis — a theory the plaintiffs had not adequately alleged. Finding that plaintiffs’ price squeezing claim was “nothing more than an amalgamation of a meritless claim at the retail level and a meritless claim at the wholesale level,” the Court found AT&T immune from antitrust liability.
The Doe court found its case to be indistinguishable from Linkline. Conceiving of plaintiffs’ claims as amounting to an allegation that Abbott’s prices for Kaletra were too low and its prices for Norvir were too high, the Ninth Circuit evaluated each claim separately. Because plaintiffs had alleged no refusal to deal at the booster level and no below-cost pricing at the boosted protease inhibitor level, the Ninth Circuit found Abbott’s conduct to be beyond the range of Section 2.
Evaluating the Conduct in Doe as Monopoly Leveraging
The plaintiffs in Doe resisted application of Linkline by arguing that they had alleged neither price squeezing nor bundled rebates, but rather had alleged a free-standing claim for monopoly leveraging. The Ninth Circuit had previously approved such a theory of Section 2 liability in Image Tech. Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195 (9th Cir. 1997). That case involved an allegation that Kodak had refused to sell its parts to independent service organizations and that the company “used its monopoly power over Kodak photocopier and micrographic parts to attempt to create and actually create a second monopoly over the service markets.”
The Doe plaintiffs maintained that their theory was indistinguishable; that is, they alleged that Abbott was using its monopoly power over the booster market to attempt to create a second monopoly over the boosted protease inhibitor market. The Ninth Circuit was unconvinced, pointing out that “Image Technical involved a refusal to deal.” Under those circumstances, the court found that Linkline’s logic compelled the conclusion that the Doe plaintiffs had not stated a claim under Section 2.
Unfortunately, bundled discounts continue to defy categorization into easily-applicable antitrust standards. Even within the Ninth Circuit, the extent to which the Cascade test was modified by Linkline and Doe remains unclear. The facts in Doe closely resemble neither the traditional multiple product offerings of bundled rebate analysis nor the wholesale/retail market distinction in price squeeze analysis. The Doe court’s logic suggests that Section 2 defendants within the Ninth Circuit could now enjoy even broader protection than was available under Cascade. Nevertheless, particularly given the Third Circuit’s decision in LePage’s, offering bundles that include both competitive products and products over which the bundler may have monopoly power remains subject to unclear antitrust standards.