Clearance: Proskauer's Quarterly Antitrust Update - Fall 2013

Heartburn in Massachusetts: The District of Massachusetts Activates FTC v Actavis

In Federal Trade Commission v. Actavis, Inc., the Supreme Court, in a 5-3 decision written by Justice Breyer, reversed the Eleventh Circuit's dismissal of an FTC complaint under Section 5 of the Federal Trade Commission Act[1] challenging a pharmaceutical reverse payment settlement even though the exclusionary effect of the settlement was within the scope of the patent grant. At the same time, the Court rejected the "rule of presumptive illegality" advocated by the FTC and held that the FTC "must prove its case as in other rule-of-reason cases."[2] The decision upended the prevailing Circuit Court view that the competitive effects of settlement pursuant to Hatch-Waxman litigation are assessed by reference to the scope of the challenged patent.[3] While the ruling resolves the circuit split created by last year's Third Circuit decision in In re K-Dur Antitrust Litigation,[4] it dramatically alters the certainty implicit in, and the incentives for settlement of, patent infringement litigation between brand-name and generic pharmaceutical companies.

Supreme Rules of the Road

The Actavis Court provided limited guidance regarding the factors that should be utilized in applying the rule of reason. The goal of the analysis is to determine whether the "objective [of the settlement] is to maintain supracompetitive prices to be shared among the patentee and the challenger rather than face what might have been a competitive market." The Court noted that "traditional" considerations "such as likely anticompetitive effects, redeeming virtues [i.e., procompetitive effects], market power, and potentially offsetting legal considerations present in the circumstances, such as here those related to patents"[5] were among the relevant factors to be used. While "circumstances… related to patents" should be part of the analysis, the Court denied that this would be an administrative burden[6] because antitrust review of a reverse payment settlement "normally" does not require an assessment of patent validity.  Rather, the lower courts may presume that a large reverse payment serves as a surrogate for a patent's weakness and, therefore, its objective is to maintain supracompetitive prices.[7]

To what extent this is a rebuttable presumption remains to be seen; as Commissioner Wright recently noted, "what constitutes a 'large and unjustified' payment" is the central question the lower courts must answer.[8] What is clear is that the patentee may introduce justifications for the payment because the rule of reason implies that "offsetting or redeeming virtues are sometimes present"[9] in reverse payment settlements. Such justifications include "traditional settlement considerations, such as avoided litigation costs or fair value for services."[10] The Court suggested that the lower courts should distinguish between "settlement on terms permitting the patent challenger to enter the market before the patent expires," which would be procompetitive and "to the consumer's benefit," and those containing a "payment in return for staying out of the market."[11] Given that these two propositions are the flip side of the same coin, it is not clear how one would ultimately balance quicker entry against a monetary payment.

In short, the Actavis Court directs the lower courts to formulate the structure of rule-of-reason reverse payment antitrust analysis on a case-by-case basis going forward and to determine the factors that bear on a finding of anticompetitive conduct.[12] In so doing, however, the lower courts should "structure antitrust litigation so as to avoid, on the one hand, the use of antitrust theories too abbreviated to permit proper analysis, and, on the other, consideration of every possible fact or theory irrespective of the minimal light it may shed on the basic question — that of the presence of significant unjustified anticompetitive consequences."[13]

Heartburn in Massachusetts

The District of Massachusetts' recent decision, In re Nexium (Esomeprazole) Antitrust Litigation,[14] denying defendant pharmaceutical manufacturers' motion to dismiss, is one of the first lower court opinions to attempt to define the specific contours of rule-of-reason reverse payment analysis.[15] In Nexium, a putative class of direct and indirect purchasers alleged that AstraZeneca entered into reverse payment agreements with various generic competitors who had filed Abbreviated New Drug Applications ("ANDA") under the Hatch-Waxman Act. Following Ranbaxy's paragraph IV certification as the first-filer and AstraZeneca's commencement of patent infringement litigation, AstraZeneca allegedly "paid" Ranbaxy $1 billion to settle in exchange for a purported six-year delay in generic market entry. The $1 billion compensation was not cash, however, but took the form of AstraZeneca's agreement to refrain from marketing an authorized generic version of Nexium during Ranbaxy's 180-day generic exclusivity period. Such agreements are known as a "No-AG agreement."[16] The authorized generic product would have otherwise stood in direct competition with Ranbaxy's generic Nexium, thereby lowering the price of the drug to consumers as a result of competition. The complaint also alleges that Teva and Dr. Reddy's, both subsequent generic ANDA filers, agreed to settle and delay generic competition in exchange for AstraZeneca's forgiveness of patent infringement damages related to "at-risk" generic launches for different AstraZeneca pharmaceutical products "entirely disconnected from" the Nexium infringement suits.[17]  

Defendants originally moved to dismiss because the competitive effects were within the scope of the AstraZeneca's Nexium-related patents. The district court "hastily" denied the motions to dismiss, but after Actavis was handed down, the court decided to "revisit some of its earlier conclusions," and "adjust its rationale in light of" Actavis.[18] In so doing, the court did not change its ruling, but dismissed the motions to dismiss after applying a structured rule of reason analysis.  

Market Power in a Relevant Market

The court first asked whether the complaint sufficiently alleged market power in a relevant market. Defendants argued that the plaintiffs' proposed market of branded and generic Nexium was too narrow because Nexium was interchangeable with and competed with other products, such as Prilosec. However, the court declined to sustain the motion based on that argument because the complaint alleged that Nexium does not exhibit significant cross-elasticity of demand with other antacids.[19] While plaintiffs' allegation may have been sufficient on a motion to dismiss, it may have a more difficult road in a motion for summary judgment or before a jury. Given that Nexium is a single product market, the complaint's allegation that AstraZeneca was a monopolist able to charge supracompetitive prices for branded Nexium also was sufficient to withstand dismissal.[20]

Harm to Competition

Next the court considered whether the complaint sufficiently alleged that competition was harmed. But for the settlements, plaintiffs claim that the generic manufacturers would have entered earlier, and therefore competition was harmed by the delay in generic competition. AstraZeneca attacked these allegations as rank speculation and Dr. Reddy's argued that it was precluded from entering before the expiration of Ranbaxy's 180-day exclusivity period. The court rejected both arguments. While the court conceded that "waxing poetic on the probability of … generic market entry" may not suffice to raise a triable antitrust issue in some cases, here the fact that the generic companies had launched "at risk" in the past suggested that they would have done so in this case.[21] Moreover, as to the subsequent filers, the court suggested that the settlements facilitated an illegal bottleneck in generic entry among Ranbaxy and the subsequent filers.[22] Finally, relying on the Supreme Court's holding that a large reverse payment may be indicative of anticompetitive effect, the court noted that the AstraZeneca's $1 billion payment "seems like outsize accommodation from a company to whom Ranbaxy was purportedly liable for patent infringement."[23] These factors "sufficiently implicate[d] adverse anticompetitive consequences" to allow the claims to proceed.[24]  

Monetary verses Nonmonetary Compensation

AstraZeneca characterized the No-AG agreement as a license granting Ranbaxy an exclusive license to market generic Nexium during the 180-day exclusivity period. As such, there was no "monetary payment," to stay out of the market. Therefore, argued AstraZeneca, the plaintiffs failed to state a claim under the rules laid down in Actavis. The court rejected this argument out of hand, explaining that "[n]owhere in Actavis did the Supreme Court explicitly require some sort of monetary transaction to take place for an agreement between a brand and generic manufacturer to constitute a reverse payment."[25] Rather, the district court found that interpreting reverse payments to include nonmonetary forms of compensation, such as the No-AG agreement at issue, served to "align the law with modern-day realities."[26] Quoting Actavis, the district court explained that "the likelihood of a reverse payment bringing about anticompetitive effects depends upon its size, its scale in relation to the payor's anticipated future litigation costs, its independence from other services for which it might represent payment, and the lack of any other convincing justification."[27] In other words, a reverse payment that exceeds litigation costs and the value of any services rendered, such as the No-AG agreement with Ranbaxy and the forgiveness of unrelated infringement claims against Teva and Dr. Reddy's, serves as a proxy to demonstrate anticompetitive effects regardless of whether it takes the form of direct monetary compensation.

Justifications

The rule of reason analysis requires the court to balance the anticompetitive effects against the economic benefits of a restraint. In Nexium, the only benefit proffered was the general policy favoring the settlement of patent disputes. As noted above, the Supreme Court rejected the argument that the desirability of settlement was sufficiently procompetitive to outweigh a large reverse payment. As the defendants had not "put forward a shred of affirmative evidence" demonstrating any countervailing benefits of the settlements, the court denied the motion to dismiss.[28]

Conclusion: Looking to the Future

While the Nexium Court is the first to rule on whether nonmonetary compensation constitutes a reverse payment, others are poised to do so in the near future.[29] However, the more challenging question posed by Actavis is the extent, if any, that an assessment of patent validity plays in the analysis. As recently as September 20, 2013, in FTC v. Cephalon, Inc., the FTC moved to preclude Cephalon from introducing evidence of patent validity to defend against allegations that it paid four generic challengers to delay competition with its blockbuster prescription drug Provigil.[30] Citing Actavis' admonition that "it is normally not necessary to litigate patent validity to answer the antitrust question,"[31] the FTC takes the position that "the Supreme Court has held that such evidence does not provide a defense to the charge that a reverse payment was used to prevent the risk of competition."[32] The FTC's position in Cephalon is debatable, particularly where a defendant tenders evidence demonstrating a low likelihood of patent invalidity combined with a substantial economic loss should the patent be declared invalid. Contrary to Actavis' assertion that a large reverse payment "likely seeks to prevent the risk of competition,"[33] even where the risk of invalidity is small, a branded company with a strong, highly valuable patent may find it economically rational to settle with a large reverse payment to foreclose even a miniscule chance of invalidity. Therefore, such cases may represent the instances where evidence bearing on patent validity is necessary because the size of the reverse payment does not properly serve as a proxy for likely anticompetitive effects.  

The Supreme Court in Broadcast Music, Inc. v. Columbia Broadcast System, Inc.[34] famously warned against a "literal approach" to the application of the Sherman Act. Literalism, according to the Court, is "overly simplistic and often overbroad."[35] So we learned, for example, in Broadcast Music, that not all "price-fixing" is a per se violation of the Sherman Act. The concept, of course, is that the antitrust laws are not meant to forbid commercial activity that "promotes competition."[36] However, as Justice Brandeis noted in 1918 and the Court reiterated in Broadcast Music, determining when activities "promote competition" rather than "destroy competition" is not "a simple matter"[37] subject to a "simple test."[38]

In Gulf States Reorganization Group, Inc. v. Nucor Corporation,[39] the trial court was confronted with two issues that perhaps could have been decided literally: (i) when is a contract a "contract, combination or conspiracy" within the purview of the Sherman Act; and (ii) when is a product equivalent to a product market. Given the complexity of the matter, what is a trial court to do when confronted with warring motions for summary judgment? With consent of the parties,[40] the motions for summary judgment were referred to James F. Rill, as special master. Rill "is without question one of the leading antitrust lawyers in the Unites States."[41] Rill served as Assistant Attorney General in charge of the DOJ's Antitrust Division from 1989-92, and under his auspices the FTC and DOJ issued the first joint Horizontal Merger Guidelines in 1992. Special Master Rill recommended granting summary judgment in favor of Nucor. The district court adopted his recommendations, and the Court of Appeals affirmed based on the Special Master's four reports.

Fabricating Antitrust Claims

Nucor Corporation and Gulf States Steel competed in the manufacture and sale of black hot rolled coil steel, which is a form of plain black sheet steel that is rolled into a coil for ease of storage and transportation. When black hot rolled coil steel is bathed in acid and coated with oil the resulting steel is called pickled and oiled steel. In 2000, Gulf States filed for bankruptcy and its assets were put up for sale in a bankruptcy auction. Gulf States Reorganization Group, Inc. ("GSRG") was a newly formed company that wanted to enter the black hot rolled coil steel business by acquiring Gulf States' assets. When Nucor learned that GSRG intended to bid for the steel assets of Gulf States, Nucor entered into a contract with Casey Equipment Corp. ("Casey"), an established company that buys used steel-related equipment for resale to steel manufacturers and also develops industrial parks. Nucor and Casey formed Gadsden Industrial Park ("Park"), the purpose of which was to buy the assets of Gulf States through the bankruptcy auction and resell them.[42] Casey/Park won the auction[43] and then sold and exported the steel making assets to an Asian buyer and developed the land into an industrial park.

GSRG sued Nucor, Park and Casey under Sections 1 and 2 of the Sherman Act, alleging that they conspired to purchase the steel producing assets of Gulf States in order to block competition in the black hot rolled coil steel market[44] and conspired to monopolize the market for black hot rolled coil steel.[45] GSRG also alleged that Nucor attempted to monopolize the market for black hot rolled coil steel in the Southeast U.S.A.[46] in violation of Section 2 of the Sherman Act.[47]

Conspiracy Theory

Section 1 of the Sherman Act proscribes coordinated conduct ("a contract, combination or a conspiracy") that unreasonably restrains trade; it does not proscribe unilateral or independent conduct.[48] Similarly, Section 2 of the Sherman Act proscribes conspiracies to monopolize, which also requires coordinated conduct. On first blush, it would seem that the facts illustrate a classic example of concerted conduct. Here, there was a contract among Nucor, Casey, and Park, each of which were independent entities. Further, Nucor "had an ulterior objective in entering the contract—to exclude GSRG from the market."[49] Quod Erat Demonstrandum.

But not so fast—not all concerted conduct falls within the purview of the Sherman Act. For example, a corporation and its subsidiaries cannot conspire.[50] The Supreme Court has explained that to satisfy the plurality requirement of section 1 independent entities must have "a conscious commitment to a common scheme designed to achieve an unlawful objective."[51]

Special Master Rill noted that this case presented the "infrequent but not unprecedented" situation where a company's agent (Casey), described as an ancillary service provider, performing a discrete task, should not be liable as a co-conspirator because the agent lacked the requisite conscience commitment required by the Supreme Court.[52] First, Casey, Park and Nucor were not competitors in any market.[53] Indeed, Casey/Park lacked "any economic interest in the state of competition in the relevant market."[54] Accordingly, the contract did not eliminate "independent centers of decision making," the core concern of Section 1 of the Sherman Act.[55] Moreover, GSRG presented no evidence that Casey/Park shared or even knew of Nucor's ulterior motive to exclude GSRG from the market. The Nucor-Casey/Park contract was "facially neutral,"[56] an "ordinary commercial brokerage contract."[57] Casey/Park entered into the contract "to perform its usual business" of buying and selling used steel manufacturing equipment for a profit.[58] Unlike negligence law, antitrust does not impose a duty to inquire or have a "should have known" standard. Rather, antitrust requires that "Casey knew of and 'consciously committed' to Nucor's allegedly anticompetitive objective."[59] In its motions for summary judgment, plaintiff submitted no such evidence. Finally, Casey/Park had a legitimate business reason for wanting to sell off Gulf State's assets, namely to provide the financing to purchase the property and develop it into an industrial park.[60] Thus, although the Nucor-Park-Casey agreement was literally a contract, it was not a conspiracy subject to condemnation under the Sherman Act.

"Where there is no Market, there is no Monopoly"

In addition to proscribing conspiracies to monopolize, Section 2 also proscribes unilateral conduct that monopolizes or attempts to monopolize "any part of the trade or commerce among the several States."[61] As noted, in addition to its conspiracy claims, GSRG alleged that Nucor attempted to monopolize the market for black hot rolled coil in the Southeast U.S.A in violation of section 2 of the Sherman Act. The offense of attempted monopolization requires the specific intent by the defendant to bring about a monopoly in a defined relevant product and geographic market and a dangerous probability of success.[62] GSRG had the burden to define and prove the relevant market. As the Special Master noted, "Where there is no Market, there is no Monopoly."[63]

GSRG was formed to get into the business of selling black hot rolled steel coil in the Southeast U.S.A. Literalism would therefore suggest that black hot rolled steel and the Southeast are product and geographic markets. To be sure, black hot rolled steel coil is a product and the Southeast is a "part … of the several states," but is each a product and geographic market for antitrust purposes? GSRG offered the testimony of an expert, who so opined, but did so in a "purely conclusory" manner, that is, without reference to "actual data.[64] Defining a market demands more than conclusions; it requires an examination of a number of factors, including demand substitution, supply substitution, price sensitivity, specialized distribution channels, and industry recognition. [65]

Consistent with Eleventh Circuit case law,[66] the Special Master explained that plaintiff's expert failed to take into account supply substitution when it defined the product market as black hot rolled coil.[67] The district court adopted the Master's recommendation, declaring that "Mr. Rill's analysis is right on target."  

Supply substitution "looks at competition from the production end instead of the consumer end."[68] When the response of other manufacturers is taken into account, the product market is expanded to include pickled and oiled hot rolled coil. Summarizing the findings of the Special Master, the district court wrote that the:

"evidence is undisputed that producers can readily increase their black HRC output by simply not pickling and oiling or performing other standard finishing processes. Pickled and oiled hot rolled coil is essentially hot rolled coil that is subjected to one additional process. When pickled and oiled HRC is sold, the price increase is small and based upon a fixed cost-based price differential (as compared to black HRC). Accordingly, if there were a black HRC price increase in the market, producers could (and would) immediately increase their black HRC output.… Producers of pickled and oiled hot rolled coil already have the appropriate substitute product by simply foregoing the one additional process required to produce the pickled and oiled product."[69]

The Court of Appeals affirmed the decision of the lower court, but wrote only "to explain why cross-elasticity of supply is critical to defining the relevant market in this case."[70] The Eleventh Circuit explained: 

Assume, for example, that Nucor obtains a monopoly of the black hot rolled coil steel market. Through its monopoly, Nucor inflates prices (by, say, lowering the supply of black hot rolled coil steel, which, given a constant demand, increases the price). Such a move would present pickled and oiled steel manufacturers with two options. They could continue to produce pickled and oiled steel at the same cost and continue to sell that product at the same price. Or they could cut the "pickling" processing short (thereby saving the costs of converting black hot rolled coil steel into pickled and oiled steel) and sell the black hot rolled coil steel at the higher price to earn significant profits. In a world of rational economic actors … one would expect that many, if not all, of these manufacturers would choose the latter course. As the district court explained, "[p]roducers of pickled and oiled hot rolled coil [steel] already have the appropriate substitute product by simply foregoing the one additional process required to produce the pickled and oiled product."[71]

Finally, GSRG's definition of the relevant geographic market was too narrow. The Southeast could only be the appropriate market if consumers in the Southeast could not turn to sellers outside that region.[72] In fact, both domestic and foreign manufacturers could expand production capacity and ship "large quantities" of hot rolled coil into the Southeast region.[73] Conversely, there was no evidence that consumers could not turn to sellers outside the region. Accordingly, the Southeast market is "unrealistic."

Conclusion

The final lesson, perhaps unstated, is a rival's "lousy" intent standing alone does not violate the Sherman Act. GSRG undoubtedly had good intent; it wanted to increase competition and also make money. But in the competition for Gulf States' assets, it lost fair and square, notwithstanding Nucor's motivations. Nucor undoubtedly was pleased that a competitor had gone out of business and was under no antitrust obligation to help GSRG start a business in competition with it. Nucor played by the rules set out by the bankruptcy court and did not "conspire" with a horizontal competitor.  

Suppose a manufacturer sells the same products to two retailers who compete against each other in the same geographic area. The manufacturer gives one of the retailers more favorable terms and conditions than the other retailer. Unless the price differential can be justified by savings to the seller or some other defense, the disfavored buyer has a claim against the manufacturer/seller under Section 2(a) of the Robinson-Patman Act ("RPA") [74] for treble damages.[75] But apart from a possible claim against the manufacturer (seller), does the disfavored retailer (buyer) have a claim against his rival under the RPA? In passing the RPA, "Congress sought to target the perceived harm to competition occasioned by powerful buyers, rather than sellers; specifically, Congress responded to the advent of large chainstores, enterprises with the clout to obtain lower prices for goods than smaller buyers could demand."[76] Yet only two sections of the RPA actually concern buyer behavior and it would seem anomalous that one retailer should be allowed to sue another retailer for, in essence, having greater bargaining or haggling skill than his rival. Anomalous or not, our hypothetical disfavored buyer (retailer) would seem to be in luck, as Section 2(f) provides that "[i]t shall be unlawful for any person engaged in commerce … knowingly to induce or receive a discrimination in price which is prohibited by this section." Finally, GSRG's definition of the relevant geographic market was too narrow. The Southeast could only be the appropriate market if consumers in the Southeast could not turn to sellers outside that region.[77] A disfavored buyer, therefore, has a claim against his rival under the RPA if the favored buyer: (i) knows it received a discriminatory price, or (ii) knows it induced a discriminatory price, and the seller is liable under "this section"; that is, section 2(a) of the RPA.[78]

Although recent cases under Section 2(f) are rare,[79] the Ninth Circuit, in Gorlick Distribution Centers, LLC v. Car Sound Exhaust System, Inc.,[80] recently had an opportunity to review the kinds of "knowledge" a favored buyer must have to violate Section 2(f). In Gorlick, Car Sound Exhaust System, Inc. ("Car Sound") manufactured mufflers and catalytic converters and distributed these parts in the Pacific Northwest through Gorlick Distribution Centers, LLC ("Gorlick") and Allied Exhaust Systems, Inc. ("Allied"). Allied, the favored buyer, exclusively sold parts made by Car Sound, whereas Gorlick sold parts manufactured by a number of different companies.[81] Gorlick alleged, and it was generally undisputed, that Car Sound gave Allied lower prices, greater volume discounts and higher year-end rebates than it provided Gorlick. In addition, Car Sound did not charge Allied for shipping, handling or storage, but did charge Gorlick for these services. Gorlick alleged that, under these facts, Allied received preferential treatment in violation of Section 2(f).[82]  

After discovery Allied moved for summary judgment, arguing that it lacked the requisite knowledge required under the statute. The district court granted the motion on that basis and Gorlick appealed, arguing that the district court overlooked the fact that Allied had actual knowledge, trade knowledge, and a duty to inquire whether it was receiving prohibited prices. The Ninth Circuit, relying on the Supreme Court's 1953 Automatic Canteen Co. case,[83] affirmed the lower court.[84]

Knowledge of Better Terms

The statute seems straightforward enough: the favored buyer must know that it is favored. "Buyers are not liable if they are innocent beneficiaries of discriminatory prices."[85] In Gorlick, Allied believed that it received preferential treatment. Allied's salesmen bragged about the better terms they received and internal company memos highlighted its superior position.[86] Moreover, Allied knew that the prices it received were below those on Car Sound's published price list. According to Gorlick, this fact was decisive: "where a seller publishes its prices, any departure from the schedule places the buyer on notice that he is receiving discriminatory prices."[87] The Ninth Circuit interpreted this argument as requiring a ban on bargaining. Accordingly, relying on Automatic Canteen Co., the Ninth Circuit held that the RPA does not "prohibit buyers from haggling for a better deal."[88] Gorlick's rule would put a buyer at risk for liability every time it asked for a lower than listed price. Moreover, the manufacturer may make the discount generally available, and therefore the price may not be discriminatory at all. In fact, the court noted, all of Car Sound's dealers, including Gorlick, were offered bulk pricing even when they purchased less than the target quantities.[89] Discounts that are "generally available" do not run afoul of the Act.[90]

Knowledge of Seller's Defenses

The statute, therefore, requires more than knowledge of the discrimination in price. The favored buyer also must know that the seller has violated Section 2(a). In Automatic Canteen Company of America Co., the Supreme Court held that "there is no substantive violation if the buyer did not know that the prices it induced or received were not" justified by a defense available to the seller.[91]

In other words, the favored buyer must know that the seller violated Section 2(a) and the seller did not have an affirmative defense. RPA section 2(a) provides for two defenses: cost-justified differentials and differentials resulting from changed market conditions. In addition, Section 2(b) provides the defense of meeting competition.

Too Much Knowledge Can Be a Bad Thing

But how is the favored buyer to know whether the seller has a legitimate defense? Indeed, how is the disfavored buyer to know what the favored buyer knows? Too much knowledge or the wrong kind of knowledge could be construed to violate Section 1 of the Sherman Act. In the Great Atl. & Pac. Tea Co. case, which concerned the meeting competition defense, the Supreme Court explained that competitors needed to exercise caution in obtaining actual knowledge:

In a competitive market, uncertainty among sellers will cause them to compete for business by offering buyers lower prices. Because of the evils of collusive action, the Court has held that the exchange of price information by competitors violates the Sherman Act. Under the view advanced by the respondent [FTC], however, a buyer, to avoid liability, must either refuse a seller's bid or at least inform him that his bid has beaten competition. Such a duty of affirmative disclosure would almost inevitably frustrate competitive bidding and, by reducing uncertainty, lead to price matching and anticompetitive cooperation among sellers.[92]

In Automatic Canteen Co., the Supreme Court suggested a work-around to the issue of actual knowledge: the disfavored buyer can demonstrate that the favored buyer had notice of the seller's unjustified price discrimination based on "trade experience."[93] Applying this concept, the Ninth Circuit in Gorlick wrote that knowledge based on trade experience requires proof that Allied and Gorlick purchased from Car Sound the same quantities in the same manner with the same level of effort at substantial price differentials that are not justified.[94] The evidence, however, showed that "Gorlick and Allied were very different Car Sound customers."[95] Allied was Car Sound's number one account, selling fifteen times the dollar amount of product that Gorlick did. In addition, Allied had developed an electronic ordering system that reduced errors and streamlined its dealings with Car Sound. Accordingly, the effort required by Car Sound to serve Allied was considerably less than the exertion required to serve Gorlick.[96] The preferential treatment that Allied received, therefore, was simply "an incentive for its continued loyalty."[97]  

Gorlick also argued that since Allied knew it was getting preferential treatment from Car Sound it had a duty to inquire into the prices offered to its competitors by Car Sound. The court conceded, without deciding, that where a buyer induces discrimination in price, it may have a duty to inquire. But Allied had not induced the price differential and the court saw no reason to "drastically expand the scope of that duty."[98] Had the Ninth Circuit done otherwise, it "would have given give rise to a price uniformity and rigidity in open conflict with the purposes of other antitrust legislation."[99]

Conclusion

The Supreme Court has instructed that the RPA must be construed "consistently with broader policies of the antitrust laws."[100] Mindful of that instruction, the Ninth Circuit in Gorlick carefully construed the various kinds of "knowledge" required by Section 2(f). In so doing, it generally preserved the ability of disfavored buyers to maintain a price discrimination claim against a favored buyer, while at the same time not enabling them to "extend beyond the provisions of the Act." [101]   

Some battles never end, or so it seems. Novell v. Microsoft is the Yankees-Red Sox of the tech world – albeit not as exciting. Microsoft first released its Windows95© operating system in August 1995. Nine years later, on November 12, 2004, Novell sued Microsoft, alleging that Microsoft's refusal to license certain application programming interfaces ("APIs") for its "Office" suite application was an illegal refusal to deal that had the purpose and effect of maintaining its Intel-compatible operating system monopoly in violation of Section 2 of the Sherman Act.[102]

That even a monopolist has no duty to deal with its competitors is one of the bedrocks of antitrust law.[103] Yet, like all rules – even the rule against price-fixing – the law recognizes exceptions. [104] Novell claimed that Microsoft's conduct fell within the well-recognized exception that a monopolist that terminates a prior voluntary course of dealing with a rival without a legitimate business rationale violates Section 2 of the Sherman Act. After 11 years of litigation, a two-month trial, and a hung jury,[105] the district court entered judgment for Microsoft, finding that Microsoft's conduct did not violate the Sherman Act as a matter of law. On appeal, in Novell, Inc. v. Microsoft Corp.,[106] the Tenth Circuit affirmed the lower court, holding that "Microsoft did nothing unlawful," and in fact "engaged in legally permissible competition."  

Ancient History

A computer operating system is a software program that mediates the computer hardware and application software, such as word processing, spreadsheet, presentation, and graphics software. Microsoft's first operating system was MS-DOS, which it had acquired from Seattle Computer Products in 1981, and then licensed to IBM. MS-DOS was not the only version of DOS, but by the mid-1990s, Microsoft was the dominant supplier of DOS, the operating system used in Intel-compatible personal computers.

Users communicated with DOS by typing commands on a keyboard. Apple developed an operating system which allowed users to communicate with the computer by pointing and clicking An input device (a "mouse") on pictures ("icons") on the computer screen, a so-called graphical user interface. Building off of the experience it learned in developing applications software for Apple's graphical operating system, Microsoft, in the mid-1990s, began to develop its own graphical user interface operating system, which would become Windows95©.[107]

In addition to operating systems, Microsoft sold application software, including Word, Excel, and Power Point, which it bundled in an offering called "Microsoft Office." Microsoft, however, was not the only vendor of application software. Independent software vendors ("ISVs"), including Novell, also offered application software in competition to Microsoft Office. Novell offered PerfectOffice©, which included WordPerfect and Quattro Pro (a spreadsheet program). Prior to the introduction of Windows95©, WordPerfect was the leading word processing program on computers using DOS.

The success of an operating system in the marketplace depends in part on the number of applications written for it. In order to write applications for an operating system, the developer needs access to the operating system's programming interfaces ("APIs"). In developing Windows 95©, Microsoft needed to decide to what extent should it provide access to Windows95©'s APIs before its official release. Providing prerelease access to APIs to ISVs, like Novell, would mean that upon Windows95©'s official release, consumers would have the choice of software not just from Microsoft, but from other developers. Microsoft, in fact, provided a prerelease of Windows95© with APIs to ISVs, including Novell, in June 1994. However, in October 1994, Microsoft changed course and advised ISVs that they could no longer rely on the previously published APIs and would not guarantee the operability of the previously published APIs in the final version of Windows 95©.[108]  

Good Manners and Antitrust

As noted, as a general rule, no firm, not even a monopoly firm, has a duty to deal with its competitors.[109] "Antitrust evinces a belief that independent, profit-maximizing firms and competition between them are generally good things for consumers... Experience teaches that independent firms competing against one another is almost always good for the consumer and thus warrants a strong presumption of legality."[110] The rationale for this rule is straightforward. Helping a rival would (i) "paradoxically risk encouraging collusion," (ii) "risk reducing the incentives both sides have to innovate, invest and expand," and (iii) would put courts in the role of central planners required to choose the applicable terms and conditions.[111]

Under certain limited circumstances, however, a monopolist's refusal to deal with its competitors may violate Section 2 of the Sherman Act,[112] which proscribes actual and attempted monopolizations, and conspiracies to monopolize. To establish that the refusal to deal constitutes the offense of unlawful actual monopolization, the plaintiff must prove that the seller possesses monopoly power in relevant product and geographic markets.[113] In the Novell case, Microsoft conceded that it had a monopoly in Intel-compatible operating systems in the 1990s.[114]    

Having a monopoly, however, is not enough to make an act illegal under Section 2.[115] In addition, Novell had to prove that Microsoft "willfully" maintained or extended its monopoly.[116] To prove "willful" acquisition of monopoly power, the plaintiff must prove that the seller engaged in exclusionary conduct, which, generally, is conduct that does not have a valid business purpose and harms consumers.[117] In the context of a unilateral refusal to deal, exclusionary conduct may be demonstrated by showing that there was a prior voluntary profitable course of dealing, which the monopolist terminates and in so doing sacrifices short-term profits in order to harm a rival (that is, acts without an efficiency justification).[118]    

Wasn't this exactly what Microsoft did? Initially, Microsoft shared its APIs with ISVs "in an effort to spur them into writing software," and then pulled the rug out from under them. [119] "A voluntary and profitable relationship clearly existed between Microsoft and Novell";[120] and Microsoft understood that the withdrawal would make it "harder for the likes of Notes [and] WordPerfect" to compete and would give "Office a real advantage."[121] In Aspen Skiing Company v. Aspen Highland Corp.,[122] both the Tenth Circuit and the Supreme Court upheld a jury verdict finding liability when Aspen Skiing, a monopolist, agreed to a sales and a marketing joint venture with its smaller rival, Aspen Highlands, and then discontinued the venture even though the arrangement was profitable.[123] Aspen offered no efficiency rationale for its conduct; instead Aspen "discontinued the arrangement simply to reduce the value of Aspen Highlands, force Highlands to sell, and in this way ... control all four ski mountains in Aspen."[124]  

Even though the Aspen Skiing rule was "at or near the outer boundary of § 2 liability,"[125] surely after years of struggle Novell had finally scored the winning run in the ninth inning. Like mighty Casey, however, Novell struck out. True, Microsoft was a monopolist and, yes, it voluntarily entered into a profitable course of dealing with ISVs; it was also true that Microsoft may have had "an uncharitable intent toward rivals, maybe even a wish to 'hurt' or 'destroy' them.'"[126] But the antitrust laws are not "designed to be a guide to good manners"[127] and the difficulty was that Microsoft "did not sacrifice short term profits, let alone in a manner that was irrational but for its tendency to harm competition."[128] Indeed, Microsoft presented evidence that its "decision came about as a result of a desire to maximize the company's immediate and overall profits."[129] Moreover, even if Microsoft made Windows 95© less attractive in the short term, the withdrawal of the APIs "allowed it to win significant profits in the sale of Office suite applications – and to do so immediately."[130] Thus, the withdrawal of the APIs was a double-edge sword. On the one hand, Windows95© would not be as popular upon its initial release because a wide range of software from different vendors would not be available; on the other hand, "withdrawing access would also make Microsoft's own applications, including Microsoft Office©, more immediately attractive."[131]  

Efficiency and Antitrust

Given that Novell claimed that Microsoft was willfully extending its operating system monopoly, why was it not enough that Microsoft arguably made Windows95© less attractive in the short run? Aspen held that a plaintiff had to show that the defendant's conduct "had no economic justification except its tendency to exclude a rival."[132] In contrast, even though Novell claimed that Microsoft's conduct was directed at monopolizing the operating system market, as long as Microsoft's conduct was profit-maximizing overall, its withdrawal of help was not anticompetitive. "Microsoft is an integrated firm with the goal of maximizing overall profits."[133] According to the Tenth Circuit, antitrust requires firms to seek overall profitability and does not require courts to "disaggregate profits from different lines of business."[134] Any other rule would require large firms "to forego immediate overall gains in order to subsidize a less efficient rival that happens to do business only in one particular product line."[135]

Venue in the Lower Court Case

KM Enterprises, Inc. ("KME"), an Illinois corporation, and Global Traffic Technologies, Inc. ("GTT"), a Minnesota company, are competitors in the market for devices that permit emergency vehicles to send a signal that preempts ordinary traffic lights and thereby allows the emergency vehicle to pass through an intersection with, rather than against, the light. KME filed suit in the Southern District of Illinois, claiming that GTT violated antitrust laws by improperly interfering with competitive bidding on public contracts and engaging in illegal tying. GTT did not transact business in the Southern District of Illinois, although a few of its devices had been installed at intersections in the district and it had made six direct sales in the district over a four-year period. The public bids referred to in the complaint occurred in the Northern District of Illinois. Nevertheless, KME sued GTT in the Southern District of Illinois, alleging violations of the Sherman and Clayton Acts.[137] KME accomplished personal jurisdiction over GTT in the Southern District relying on the service provision of Clayton Act § 12,[138] which provides special (but non-exclusive) rules for venue and service of process in antitrust cases against corporations. It did not rely on the Clayton Act § 12's venue provisions, however, it instead relyied on the general venue provision of 28 U.S.C. § 1391.[139] The district court dismissed the action for lack of venue because GTT did not have sufficient contacts with the district as it did not reside in the Southern District of Illinois and none of the events took place there.

On appeal, KME argued that GTT had sufficient contacts to satisfy Section 1391 and that it was entitled to rely on Section 1391 even though it had relied on Clayton Act § 12 for service of process. KME claimed that it was appropriate to use Clayton Act § 12's nationwide service of process provision, but not Clayton Act § 12's venue provision. Instead, KME relied on Section 1391 for venue. Presumably under this interpretation, the reverse would be possible: service under Fed. R. Civ. Pro. 4(k)(1)(A) and Clayton Act § 12's venue provision. In other words, it could mix and match among the service of process and venue provisions of Clayton Act § 12 and Section 1391.

À La Carte or Fixed Menus

An antitrust action, like all civil actions, is commenced by serving a summons and a complaint. Rule 4 of the Federal Rules of Civil Procedure provides that personal jurisdiction is "established" when a summons and a complaint are served on an entity (i) in a territory where that entity would be amendable to suit under the laws of the state in which the federal court sits or (ii) pursuant to the rules of a specialized federal statute. [140] Clayton Act § 12 is one such specialized federal statute that provides rules for service of process and for venue in antitrust cases. In fact, Section 12 provides for nationwide service of process[141] and is therefore more generous than Rule 4(k)(1)(A).[142]  

As noted, Clayton Act § 12 is both a service and venue provision. And therein lies the problem, as Clayton Act § 12's venue provision and the general venue statute, 28 U.S.C. § 1391, are not consistent with each other. The general federal venue statute for civil actions, 28 U.S.C. § 1391, provides (in relevant part) that venue is proper in a judicial district in which (i) any defendant resides if all defendants are residents of the state in which the district is located or (ii) a substantial part of the events or omissions giving rise to the claim occurred.[143] A corporation resides in a judicial district within which its contacts would be sufficient to subject it to personal jurisdiction.[144] In contrast, Clayton Act § 12 is more generous, providing that venue is proper anywhere the corporation is incorporated ("inhabitant"), "found," or "transacts business."

The Seventh Circuit explained why the issue is important. If a plaintiff must use Clayton Act § 12 as a whole (that is for both service and venue),

[t]hen there exist some limits on where a corporate antitrust defendant may be sued. Though personal jurisdiction is appropriate everywhere under the statute, venue is proper only in the district(s) the corporation inhabits, is found or transact business.… Section 12 is not a restrictive venue provision … but it falls short of providing universal venue in every judicial district in the United States.

But the same cannot be said if we decouple Section 12's clauses and enable a plaintiff to combine nationwide service of process with Section 1391.… Section 1391(b)(1) states venue is proper in any district where the defendant resides … while subsection (c)(2) provides that a corporation "resides" in any district in which it is subject to personal jurisdiction. But if the plaintiff relies on the Clayton Act's nationwide service of process to secure personal jurisdiction, then for purposes of Section 1391 the corporate defendant would reside in every judicial district in the count and venue would be proper everywhere.[145]

Table 1: Summary of Service and Venue Provisions

As noted, KME argued for an à la carte menu under which a plaintiff could chose to use nationwide service of process under Clayton Act § 12 without also satisfying Section 12's venue provision, instead using Section 1391's venue rules. KME did not cook this up out of nothing, but relied on opinions from the Third and Ninth Circuits, as well as the scholarly opinion of Wright & Miller.[146] For example, in Go-Video, Inc. v. Akai Elec. Co., 885 F.2d 1406, 1410-11 (9th Cir. 1989), the Ninth Circuit held that decoupling was consistent with the Clayton Act's general aim of expanding venue in antitrust actions.[147]

In contrast, the D.C. and the Second Circuits[148] have held that Clayton Act § 12 must be used as a fixed menu, that is, if a plaintiff relies on nationwide service of process, plaintiff is limited to the venue choices of Clayton Act § 12. According to these circuits, their holding is mandated by the "plain language" of the statute, which provides that "in such cases" where nationwide service is used to establish personal jurisdiction over a corporation, venue is limited to the corporation's place of incorporation, where it is found or where it transacts business.[149] The common meaning of the term "in such cases" is "previously characterized or specified," and what previously is described in Clayton Act § 12 is the venue restrictions of incorporation, inhabitation, and transaction of business. Moreover, the Second Circuit noted that Congress did not intend "to give plaintiffs free rein to haul defendants hither and yon at their caprice."[150]

Breaking the Tie

Weighing in on the circuit and scholarly split, the Seventh Circuit agreed with the D.C. and Second Circuits' holding that Clayton Act § 12 must be read as a package, but rejected their reasoning.  The plain language is ambiguous, according to the Seventh Circuit, but the reasoning of the Ninth Circuit is "too bizarre and contrary to Congress's apparent intent." [151]

…  allowing  antitrust plaintiffs to mix and match Section 12's service of process provision with Section 1391's general venue provision renders the venue inquiry meaningless, since venue is satisfied in every federal judicial district under subsection (c)(2). This runs contrary to Congress's apparent intent in passing Sections 12 and 1391 that there be some limits on venue, in antitrust cases specifically and in general. Both statutes authorize venue only when certain enumerated requirements are met, be it that the defendant "transacts business" in the district, "resides" there, or something else. It would be quite strange to read two statutes that place limits on venue in a manner that eliminates those limits.[152]  

Thus, a plaintiff has a choice: use Clayton Act § 12 as a fixed menu, or use Section 1391 and the general state long-arm statute or some other source of personal jurisdiction and service.[153]  

Was Venue Proper in the Southern District of Illinois

Applying the fixed menu option, the Seventh Circuit upheld the dismissal, concluding that venue was improper in the Southern District of Illinois under both Clayton Act § 12, and Section 1391. As noted under Clayton Act § 12 venue is proper in any district where the defendant is incorporated, is found, or transacts business. The Seventh Circuit found that none of these were satisfied. GTT is a Minnesota corporation and was not "found" in the Southern District of Illinois; nor did it transact business there. It did not promote its goods through product demonstrations or solicit orders through salespersons in the district, did not ship its products to the district, did not maintain offices or provide customer assistance in the district, had no employees in the district, and did not control a subsidiary or a distributor that transacted business in the Southern District.[154] Moreover, the fact that "GTT's technology is being used in a few places within the district does not demonstrate that it transacts business there."[155]  

For similar reasons, venue also fails under Section 1391. In states with multiple districts, like Illinois, Section 1391 provides that venue is proper in any district in which it would be subject to personal jurisdiction were the district a state.[156] An out-of-state corporation is subject to jurisdiction in the State of Illinois if (under Illinois state law) it is "fair, just and reasonable" to require the defendant to adjudicate in Illinois and (under federal constitutional law) the foreign corporation has minimum contacts with the state.[157] Here, KME lacked the minimum contacts. First, the claims do not relate to any contacts with the state of Illinois. KME's claims center on patent litigation that has nothing to do with Illinois, bid rigging in other states, or sales other than the six sales that were made in the Southern District. Second, "minimum contacts" is not the same as de minimus contacts. "Minimum contacts" means "continuous and systematic general business contacts"[158] and six sales over four years does not satisfy that criterion.

Conclusion

The intersection between general principles of federal personal jurisdiction and venue and the Clayton Act's specific provision of course will remain tangled until the Supreme Court decides to untie the knot. Right now, the universal antitrust venue is an option in the Third and Ninth Circuits, whereas the D.C. Circuit, the Second Circuit, and now the Seventh Circuit require plaintiffs to choose either the Clayton Act as a whole or the general venue statute as a whole.

[1] 15 U.S.C. § 45.