By Timothy Muris
Timothy J. Muris is a Foundation Professor of Law at George Mason University’s Antonin Scalia Law School and Senior Counsel at Sidley Austin LLP. He was the Director of the Federal Trade Commission’s Bureau of Competition, 1983-85, and Chairman of the FTC, 2001-04. Since leaving the FTC, he remains active on these antitrust issues both as a scholar and expert for a variety of clients, including ACT. This article reflects his own views.
On January 17, 2017, the Federal Trade Commission (FTC) sued Qualcomm, alleging it had illegally monopolized markets for chipsets used in smartphones.[1] Three days later, Apple made many of the same allegations against Qualcomm.[2] Apple’s contract manufacturers also have made similar antitrust allegations in their counterclaims against Qualcomm.[3] And other interested parties—such as Intel, Samsung, the American Antitrust Institute, and Act | The App Association– have filed amicus briefs in support of the FTC action against Qualcomm.
While the cases focus on alleged harm to competition within baseband (or modem) chipset markets, they also address important issues relating to standard setting, FRAND obligations, and licensing of intellectual property. Obviously, Qualcomm will have its chance to rebut the allegations, but the factual charges are compelling. And if the facts pled are proven, these cases fit comfortably into conventional antitrust precedent, as Judge Koh held in rejecting Qualcomm’s motion to dismiss.[4] To support this conclusion, I first discuss the background of the charges, then turn to the heart of the cases involving the chipset market, and finally explain why the cases against Qualcomm follow traditional antitrust principles without weakening appropriate intellectual property protections.
BACKGROUND
Some history is helpful. The original digital cellphones were based on standardized protocols, for which Qualcomm developed key intellectual property for one of those protocols (2G CDMA) and became the dominant supplier of chipsets using that protocol. Qualcomm’s original exploitation of its 2G CDMA intellectual property (in terms of profits and dominance in chipsets) is not at issue. Instead, the question is whether Qualcomm has since illegally maintained its chipset monopoly through exclusionary conduct, including licensing practices that may have broken its promise to license on fair, reasonable, and non-discriminatory (FRAND) terms.
Several facts point to illegality. Today, smartphones use the far more advanced 4G LTE protocol for which Qualcomm is just one of many companies that have contributed intellectual property. Even though its earlier intellectual property development should give it no distinct advantage in today’s marketplace, Qualcomm nonetheless still dominates the market for advanced LTE chipsets and nearly all of its former competitors have exited the business.
Beyond that anomaly, Qualcomm’s licensing practices are peculiar. Qualcomm’s intellectual property contributes to a chipset that is only one component of modern smartphones. A company in Qualcomm’s position would typically license its FRAND-encumbered intellectual property to other chipset manufacturers, and Qualcomm acknowledges it has a FRAND obligation not to keep its standard essential patents (SEPs) to itself. But that is not Qualcomm’s practice. Qualcomm refuses to license its SEPs to competitors (e.g, Intel). Instead, it licenses its patents to the handset manufacturers (i.e., Apple, Samsung, etc.) and earns a royalty rate based on the sales price of the handset.
That leads to a further anomaly. The modern smartphone includes features and functionality far beyond the basic functions of the initial cellphones. Today’s smartphones are still used for phone calls, but are also computing and multimedia platforms. A typical smartphone includes an advanced microprocessor, a sophisticated graphics processor, memory, location awareness, touch technology, voice recognition, high-definition cameras, video and music replay, power management, and an advanced operating system. Each of these technologies provide numerous benefits to end users independent of the phone connections. Yet, as new technologies and functionalities are added to a smartphone and built into the sales price, Qualcomm continues to claim a percentage of that increasing value through its licensing demands.[5]
According to the pleadings, Qualcomm obtains royalties far larger than if it licensed its patents to component manufacturers based on the price of the components in which its patents are used. Indeed, Apple’s complaint alleges that it pays more in royalties to Qualcomm than it pays all other cellular patent holders combined. At the same time, Qualcomm profits heavily from its chipset business and maintains a dominant share in that market.
THE CASES AGAINST QUALCOMM FOCUS ON CHIPSETS
Qualcomm’s licensing practices seem to breach Qualcomm’s FRAND obligations and both the FTC and Apple complaints allege such a breach. Most antitrust cases involving improper conduct related to SEPs focus on technology markets, i.e., relevant markets for the licensing of intellectual property. The allegation typically is that the SEP owner has acted deceptively and/or breached its FRAND obligation, leading to higher prices or other anti-competitive effects within the technology markets in which the SEPs are licensed and compete.
In the two recent Qualcomm cases, however, the alleged harm to competition is not within technology markets for licensing patents. Rather, it is alleged that Qualcomm has illegally maintained a monopoly (and harmed competition) within product markets for selling chipsets used in handsets (such as smartphones). Thus, the cases differ from recent cases involving SEPs and, in fact, are better viewed as traditional monopolization cases. In other words, although the alleged improper conduct involves potential exclusionary practices related to the licensing of SEPs, the focus is on harm to competition within traditional product markets, here broadband (or modem) chipsets.
Within those markets, several allegations of improper conduct are made. Both pleadings allege exclusive dealing through Qualcomm agreements with Apple that effectively precluded Apple from buying chipsets from other manufacturers. While it will have to be proven that the agreements harmed competition, alleging exclusive dealing is standard antitrust fare.
The pleadings also contain two interwoven narratives involving how Qualcomm’s licensing practices improperly maintain dominance in chipsets. First, Qualcomm has a FRAND obligation not to discriminate in its licensing, including licensing chipset manufacturers. Licensing chipset manufacturers, however, would extinguish its intellectual property rights under patent law and it would have no basis then to seek royalties from the downstream handset makers such as Apple and Samsung. Because Qualcomm wants to license handset makers directly, Qualcomm refuses to license other chipset manufacturers (in violation of FRAND commitments), obviously believing it more beneficial to instead earn royalties from downstream sales of handsets. In so doing, the antitrust pleadings argue that Qualcomm illegally structures its patent licenses to handset makers to impair the ability of other chipset manufacturers to compete.
Second, while Apple and the FTC make their arguments slightly differently, both challenge what the FTC calls Qualcomm’s “no license-no chips” policy, under which Qualcomm will not sell chipsets to any handset manufacturer unless the manufacturer agrees to a patent license under Qualcomm’s terms. This appears to be the crux of the alleged illegal behavior. Both Apple and the FTC note that Qualcomm is such a dominant supplier of chipsets that if a handset maker loses access to Qualcomm chipsets, it risks crippling damage to its business. As a result, Qualcomm can impose severe licensing terms for access to its intellectual property as a condition of access to Qualcomm chipsets. Indeed, the FTC complaint describes internal Qualcomm documents that discuss how conditioning the sale of product on acceptance of a separate license can allow Qualcomm to obtain non-FRAND licensing terms.[6] Among other effects, this strategy discourages Qualcomm licensees from testing in court whether Qualcomm’s licensing demands are consistent with its FRAND obligations because, until now, no licensee dared challenge whether Qualcomm’s demands were excessive. Until very recently, handset makers have not challenged Qualcomm because of the risk of losing needed access to Qualcomm chipsets and rival chipset makers are not licensed in the first place.
Thus, the FTC and Apple complaints allege that, through its “no license-no chips” policy, Qualcomm requires handset makers to accept licenses that effectively “tax” rival chipset manufacturers, impairing their ability to compete and illegally maintaining Qualcomm’s chipset monopoly. This “tax” allows Qualcomm to manipulate the effective price for chipsets. When a handset manufacturer uses a Qualcomm chipset, the price has two parts: (1) the price for the chipset and (2) the price for the intellectual property license that Qualcomm requires separately. The effective price paid combines these two parts. If a handset maker tries to avoid buying chipsets from Qualcomm, it still must pay Qualcomm the second part of this combined price under its intellectual property license. Rival chipmakers use Qualcomm’s SEPs and Qualcomm, despite its FRAND obligations, refuses to license rival chipset makers (which again would extinguish any intellectual property claims Qualcomm has on downstream handset manufacturers). The allegation is that Qualcomm manipulates the price by loading more of the combined price a handset maker pays into the intellectual property license, a price it will still have to pay if it buys from Qualcomm’s rivals. This works like a “tax” on rival chipset maker sales, as buyers have to pay this extra amount if they want to avoid buying chipsets from Qualcomm. Conditioning the ability to license intellectual property on the licensee’s purchase of a good or a service also is standard antitrust fare to be evaluated under the rule of reason.[7]
This licensing structure likely allows Qualcomm to obtain royalties above its FRAND commitments. Although, per above, the two cases do not allege harm to competition in licensing technology markets resulting from Qualcomm supra-FRAND demands, they do claim that Qualcomm’s licensing and chipset sales practices make their licensing demands much more immune to court resolution of FRAND compliance. Specifically, a handset manufacturer that decides to challenge whether Qualcomm’s licensing practices violate FRAND under contract or patent law would risk foregoing chipsets that are critically needed for its business. Thus, Qualcomm’s licensing practices may interfere with proper FRAND determinations under patent and contract law and cause patent hold-up. Once one accounts for the combined price Qualcomm charges, the result is higher effective prices from all suppliers across the chipset marketplace.
THE CASE AGAINST QUALCOMM FOLLOWS TRADITIONAL ANTITRUST PRINCIPLES
There is little doubt that, if Qualcomm monopolizes one or more chipset markets, and maintains this monopoly through restrictive conduct that impairs rival chipset manufacturers resulting in higher prices, then the conduct is illegal under Sherman Act § 2. That is a standard antitrust principle independent of the context of patent licensing, FRAND commitments, and standard setting.[8] Accordingly, the Qualcomm case breaks no new legal ground. Reflecting that reality, in a thorough, 50-page opinion in the FTC case that addresses many more claims and facts than discussed in this blog, district court Judge Koh concluded that the FTC’s complaint and alleged facts would sustain a Sherman Act case.
While the case breaks no new ground in terms of antitrust principle, what is noteworthy about the case, connecting it to the recent history of antitrust enforcement related to SEPs, is how the restrictive conduct at issue, discussed above, also involves Qualcomm’s licensing practices and its FRAND obligations. Again, standard antitrust law holds that product monopolies can be maintained illegally through intellectual property licensing.[9] Thus, if the facts as alleged are true, they state a claim under the Sherman Act.
One question that has arisen is the extent to which the FTC and Apple must prove that Qualcomm’s royalty rates exceed those consistent with its FRAND obligation. Such excessive royalties are implicit in the logic of the cases, and a plaintiff must establish harm to competition to prove monopolization. The question is what proof is sufficient to establish an anticompetitive effect and, more specifically, whether harm to competition can be established without a detailed expert analysis showing actual prices exceed but-for prices. While in a monopolization case the FTC and Apple must establish harm to competition and elevated pricing (as part of establishing what lawyers call a prima facie case), it is possible to prove the alleged Qualcomm conduct and use economic expert testimony opining that this conduct would logically lead to higher pricing. Among other allegations, if the “no license-no chips” policy effectively prevents customers from contesting infringement, validity, or FRAND compliance, it can follow as a matter of economics that customers then pay higher royalty rates. In other words, a prima facie case of harm to competition can be established through a combination of evidence and economic inferences from that evidence.
This issue of proof is important because it can be difficult to assess whether actual royalty rates are inconsistent with FRAND, as the task requires calculating detailed FRAND rates.[10] In its pleadings, neither the FTC nor Apple calculates a FRAND rate, but they do note that Qualcomm’s 5% royalty rate on phones has not changed over time, which does seem inconsistent with a proper setting of a FRAND rate given that the base of this calculation includes all the new smartphone features. Similarly, one amicus brief notes that Qualcomm earns more than four times the amount of cellular royalties that Ericsson does even though the two companies have a comparable number of novel SEPs for the 4G LTE protocol. Such evidence also suggests Qualcomm’s royalty rates are not consistent with FRAND. Again, such evidence can help support the basic case that anticompetitive effects are continuing.
Assuming a prima facie case is established, Qualcomm would then have the ability to rebut that case, including evidence that pricing in fact is not elevated. Obviously, if Qualcomm establishes that its royalties are consistent with FRAND, e.g., using expert witnesses, such proof would be a strong factor in its defense. If the plaintiffs can establish a prima facie case without such proof, as seems likely if the facts pled are proven, the burden will be on Qualcomm, but Apple and the FTC may also choose to undertake such detailed expert calculations to bolster their cases.
Finally, when assessing these cases under modern antitrust law, it is worth emphasizing that because the cases are grounded in established antitrust law – the monopoly maintenance paradigm – they do not represent a weakening of appropriate intellectual property protection as suggested by some. It has long been recognized that IP rights are not absolute, and that their misuse can be a proper subject of antitrust law, if the stringent requirements of the law are satisfied. Moreover, the conduct here involves a web of interrelated practices, including various allegations of exclusionary conduct, adding a layer of complexity to the case that likely magnifies the harm to competition from a breach of FRAND commitments. In any event, the focus on the complex conduct in this case that includes IP obligations is a proper basis for antitrust enforcement.[11]
[1] FTC v. Qualcomm, Case No. 5:17-cv-00220 (N.D. Cal. January 17, 2017); The FTC’s case alleged both violations of antitrust law (Complaint, ¶ 147(a) and ¶ 147(b)) and then alleged in the alternative a stand-alone FTC § 5 violation (¶ 147(c)). My discussion here focuses on the FTC’s complaint as alleging an antitrust violation. On June 26, 2017, the Court rejected Qualcomm’s motion to dismiss the case.
[2] Apple Inc. v. Qualcomm Inc., Case No. CV0108 (S.D. Cal. Jan. 20, 2017). Even though the allegations in the Complaint significantly parallel the FTC’s, Qualcomm chose simply to answer the complaint without filing a motion to dismiss. Apple subsequently amended its complaint on June 20, 2017.
[3] See Qualcomm Inc. v. Compal Electronics, Inc., FIH Mobile LTD., Hon Hai Precision Industry Co., LTD., Pegatron Corp., and Wistron Corporation, Case No. 3:17-CV-01010-GPC-MDD, Defendants’ Answer and Defenses; Counterclaims; Demand for Jury Trial, pp. 80-88 (U.S. Dist. Ct. S. Cal, filed July 18, 2017).
[4] See note 2.
[5] These issues relate to the problems that arise in the FRAND context when royalties are not based on sales of the smallest saleable unit. See Joseph Kattan, Janusz Ordover & Allen Shampine, FRAND and the Smallest Saleable Unit, Competition Policy International (September 2016).
[6] This discussion in the FTC complaint, at ¶¶ 97-98, was originally redacted, but then unredacted and discussed in Judge Koh’s opinion rejecting dismissal of the case. See Order Denying Motion to Dismiss, slip op. at 11.
[7] See “Antitrust Guidelines for the Licensing of Intellectual Property,” Sec. 5.3, issued by the U.S. Department of Justice and the Federal Trade Commission (January 12, 2017) (citing United States v. Paramount Pictures, Inc., 334 U.S. 131, 156-58 (1948) (copyrights); Int’l Salt Co. v. United States, 332 U.S. 392 (1947) (patent and related product), abrogated in part by Ill. Tool Works, Inc. v. Indep. Ink, Inc., 547 U.S. 28 (2006)).
[8] See, e.g., United States v. Microsoft, 253 F.3d 34, 57-59 (D.C. Cir. 2001); III Philip Areeda and Herbert Hovenkamp, Antitrust Law § 651 (defining exclusionary conduct for purposes of Sherman Act § 2).
[9] See note 8, at Sec. 2.1.
[10] For examples of such calculations in a court setting, see In re Innovatio IP Ventures, 2013 WL 427167 (N.D.Ill. 2013) and Microsoft Corp. v. Motorola, Inc., 2013 WL 2111217 (W.D.Wash. 2013).
[11] For a discussion of recent patent law developments and antitrust enforcement related to patents, see Timothy J. Muris, Bipartisan Patent Reform and Competition Policy, American Enterprise Institute publication (May 2017) (http://www.aei.org/publication/bipartisan-patent-reform-and-competition-policy).