Deciphering Non-Discriminatory Licensing Terms Under a FRAND Commitment

by Arielle Koppell*

To what extent can a SEP holder can discriminate in how it licenses to suppliers without violating its FRAND commitment? In this Contribution, Arielle Koppell (’19) considers whether and how SEP holders can discriminate in licensing. Ultimately, this Contribution argues that a SEP holder should be able to arrange differential licensing terms for vertically integrated and non-vertically integrated licensee counterparts require its licensees to purchase tied non-SEP components when those non-SEP components are functionally related.

As technology becomes increasingly sophisticated, the need for them to be interoperable around a single standard becomes increasingly important to consumers. At the same time, interoperability presents important antitrust and contractual issues. Numerous intellectual property holders around the globe have achieved the pinnacle of profitability on their patents: being incorporated into a standard by a standard-setting organization (SSO) due to the innovations presented by their patented technology. In order to be incorporated, patent holders will frequently make voluntary commitments to the standard-setting organization, promising to abide by a “FRAND” commitment in how they subsequently license their patent, thereafter known as a standard-essential patent (SEP). A FRAND commitment requires the patent holder to engage in fair, reasonable, and non-discriminatory conduct in the price and conduct by which it licenses its SEP. In addition to securing these commitments, the majority of standard-setting organizations issue policy statements that support firms’ rights to gain access to the patented technologies that are required for them to achieve standard compliance.²

While what constitutes fair and reasonable conduct under a FRAND commitment in setting licensing fees has been expounded upon in American and European competition law, the extent to which a SEP holder can discriminate in how it licenses to suppliers without violating its commitment is less apparent.³ For instance, it is unclear whether a SEP holder is obligated to license the patent at issue on a standalone basis, should the licensee request it, or whether the patent holder can bundle its SEP with non-SEPs, licensing at the technology portfolio level.⁴

This Contribution will argue that a SEP holder should be able to arrange differential licensing terms for vertically integrated and non-vertically integrated licensee counterparts, regardless of whether that makes them similarly situated. Further, this Contribution will argue that it is legal for a SEP holder to require its licensees to purchase tied non-SEP components when those non-SEP components are collateral input or functionally related to the SEP.

* * * * *

There is widespread disagreement about what the “non-discriminatory” principle of FRAND commitments requires of SEP holders in their licensing practices. Several commentators believe that it requires all implementers of the standard to be offered licenses to the SEP and all similarly situated firms to pay the same royalty rate.⁵ Justice Birss of the High Court of Justice of England and Wales supported this view in Unwired Planet v. Huawei, reasoning that a FRAND royalty rate should be set based on the value of the SEP rather than by the licensee’s characteristics, such as its size.⁶ Still, he limited the ability for future licensees to challenge a license allegedly granted under FRAND terms upon learning that a similarly situated licensee was charged less, absent a competitive distortion between the two licensees.⁷ Parallel analysis drove the judgment of the Court of Justice of the European Union in MEO v. Autoridade da Concorrência, which found that price discrimination by a dominant firm is prohibited as an Article 102(c) TFEU violation when it tends to distort competition in a downstream market.⁸

Others disagree, writing that non-discrimination should not be read as an inflexible commitment for SEP holders to license at identical terms to all potential licensees.⁹ These courts and commentators have interpreted non-discrimination as to allow a SEP holder to charge different royalty rates to licensees based on their size or market share, requiring those who sell more SEP-incorporating products to pay higher levels of royalties.¹⁰ Disagreement over how to interpret the non-discrimination prong of FRAND commitments is only further complicated by a wide range of perspectives on which factors are relevant to find licensees similarly situated.¹¹

Further, this area of debate does not speak directly to how the non-discrimination prong of FRAND applies to SEP holders engaging in differential licensing practices and offering different non-price licensing terms across their licensees.¹² For instance, a SEP holder may choose to selectively license its SEPs to non-vertically integrated firms that it does not directly compete with (firms that arguably incorporate the standard by using the SEP to create a standard-compliant secondary product that the patent-holder does not produce). Licensing only at the end-user product level is not uncommon amongst technology firms, but there is a dearth of judicial authority justifying whether this practice conforms to or departs from the firm’s FRAND commitment.¹³ Setting licensing practices that allow the SEP holder to exclusively license at the end level in a supply chain affords them the opportunity to maximize the profits they can earn from royalties while avoiding triggering the patent exhaustion doctrine. They justify their ability to constrain access to their SEP by claiming that their direct competitors who are not vertically integrated downstream are not implementers of the standard.¹⁴

Component makers (direct competitors of the SEP holders) may argue that under the standard-setting organization’s IPR policies, they should receive a license that provides them with the full make, sell, and use rights over the patent and that, due to patent exhaustion doctrine, they should be able to re-sell the patent to downstream customers. In response, the SEP firm may attempt to set differential royalties based on the types of end-products sold by the licensee firms into which the patent will be incorporated. The licensee firms would argue that such conduct is inconsistent with the FRAND commitment under the spirit of the entire market value and smallest saleable unit rules, which serve to prevent the royalty amount awarded to the SEP firm from being based on a function of the end-product price (which can be distorted by unrelated factors) rather than the true value of the patent itself.¹⁵

The entire market value rule bars a plaintiff from seeking damages based on the entire end-product price in a patent infringement case and requires their proof to instead focus on the value of the patent to the “smallest saleable patent-practicing unit” (or component) within the product, unless the patented technology drives demand for the whole product. Component-maker (vertically-integrated) licensees would argue based on the entire market rule that they are entitled to license at the level of the smallest saleable unit (here, exclusively the SEP). To the extent this rule sheds light on non-discriminatory conduct in the FRAND context, it would weigh against allowing the SEP firm to license only to non-vertically integrated licensees.¹⁶ Licensing to a component manufacturer rather than an end-user manufacturer would also lower the likelihood that a SEP holder may try to exercise patent hold up and extract greater royalties from licensees, as it is more difficult to determine the value of a patent to an end-user product manufacturer (due to the heterogeneity among end products) than it would be for upstream component manufacturers.¹⁷

This analysis may be correct under patent law, but many antitrust scholars and practitioners aptly believe that FRAND commitments should not be governed by patent law restrictions and instead interpreted through contract law. It may seem to prioritize form over substance to require SEP firms to license on non-discriminatory terms, but to then allow them to engage in price discrimination by imposing contractual restrictions on their offered licenses. Still, the FRAND commitment is generally viewed as a contract between an SEP holder and SSO, with technology licensees and users playing the role of third-party beneficiaries and governed by conflict of law rules. For instance, though not a FRAND case, Justice Thomas, writing for the majority in Quanta Computer v. LG Electronics, noted that the Court’s exhaustion ruling (finding a patent holder’s attempt to restrict the scope of authorized re-sale ineffectual) did not preclude arguments based on a strictly contractual theory of liability.¹⁸

In fact, every U.S. court that has addressed a claim for injunctive relief on a FRAND-committed SEP has done so under contract law rather than antitrust law principles. Antitrust sanctions in fact may be harmful by over-deterring procompetitive participation in SSOs as liability would turn upon whether a patent infringer was a truly willing licensee – an unclear factual determination that would make SEP firms more reluctant to incorporate their patent into a standard, knowing they could not recoup their losses against patent infringers who use their technology without a valid license. The prospect of antitrust liability would also enable an infringing user to negotiate with the SEP firm in bad faith, understanding that its exposure is capped at the FRAND royalty rate and encourage those users to engage in reverse-holdout to defer payment of FRAND royalty rates.¹⁹

Under antitrust principles, a SEP firm’s refusal to license on a portfolio basis to vertically-integrated licensees does not violate its FRAND commitment, and if it did, still does not constitute an antitrust violation under U.S. law. Some allege that refusal to license at the component level results in the vertically integrated SEP holder bundling its SEP component with its portfolio. Even if this licensing offer could indeed be characterized as a bundle, the firm’s conduct would not lead to foreclosure of the component market if the SEP firm did not assert its patents at the component level and licensed its portfolio to end-user suppliers on FRAND terms irrespective of where they sourced their components. The bundle offered by the SEP firm could be competitively replicated by end-user suppliers by mixing and matching the components sold by non-integrated component suppliers and the patent portfolio of the integrated SEP firm. Therefore, the otherwise legally cognizable theory of raising rivals’ costs could not be employed by a non-vertically integrated licensee, and no antitrust problem would arise.²⁰ Moreover, to find a violation of a FRAND commitment based on a SEP firm’s refusal to license rivals, a licensee plaintiff would need to show that such refusal was anticompetitive on its own terms and is thereby unlikely to succeed (since unilateral refusals to license are virtually always regarded as lawful).²¹

The latest iteration of this issue is presented by the Federal Trade Commission’s complaint, filed in the U.S. District Court of the Northern District of California against Qualcomm in January 2017. The Court denied Qualcomm’s motion to dismiss in June 2017, writing that Qualcomm licenses its SEPs to original equipment manufacturers (OEMs), but “refuses” to license its SEPs to competing modem chip manufacturers.²² It held that the FTC adequately alleged that Qualcomm has an antitrust duty to license its SEPs to modem chips competitors under Section 2 of the Sherman Act and, by extension, Section 5 of the FTC Act.²³  Despite the immature state of the law on this issue, the Court observed, notably, that “recognizing a[n antitrust] duty to deal in this case would not require courts to play a larger role in setting the terms of dealing than the role that courts already play in determining appropriate royalties in patent cases.”²⁴

Following this acknowledgement of a duty for a FRAND-committed SEP holder to deal with all competitors based on Verizon Comm’s, Inc. v. Law Offices of Curtis v. Trinko²⁵ as a viable theory of liability, the Court granted the FTC’s motion for partial summary judgment on that theory in November 2018.²⁶ In its opinion, the Court wrote that Qualcomm’s FRAND commitments include an obligation to license to all comers, including [directly] competing [OEMs].²⁷ It also held that the non-discriminatory prong of FRAND requires that SEP holders not be enabled to selectively license across its supply chain to non-integrated suppliers.²⁸ Though it found Qualcomm’s conduct violated Section 5 of the FTC Act, the Court’s analysis still mostly drew upon contractual interpretation of the SDOs’ IPR policies.²⁹ Moreover, it is unclear how this judgment will impact the Department of Justice’s policies on the issue, given AAG Makan  Delrahim’s recent public comments that breaches of a licensing commitment should not give rise to antitrust liability,³⁰ that a refusal to license a valid patent should be per se legal, even if the patent is part of a standard,³¹ and that agencies should not “transform commitments to license on FRAND terms into a compulsory licensing scheme.”³²

While SEP firms are required to license to similarly-situated licensees on non-discriminatory terms, there is widespread disagreement over what constitutes similarly-situated licensees and what is meant by non-discriminatory terms and conditions. Under the terms of ETSI, an important SSO, licensing terms to independent licensees need not be identical in order to be non-discriminatory. Moreover, similarly-situated licensee analyses have been made in the past based on differences in output levels (through production and/or sales), differences due to the outcome of a hypothetical pre-standard setting bilateral negotiation, and information about (prior to the selection of a standardized technology) the incremental value from the incorporated patented technology accrued by each respective firm compared to the next best alternative becoming incorporated. Another approach interprets the non-discrimination obligation as applicable only to vertically-integrated licensors (the final product price less the incremental cost of all non-SEP inputs).³³ Under each of these interpretations, offering different licensing terms to a non-integrated versus integrated licensee would not necessarily engender a breach of non-discriminatory conduct required by a SEP firm’s FRAND commitment.

There are several legitimate business reasons that explain why technology firms commonly license at the end-user device level, which may be acceptable to courts. For instance, in the telecommunications industry (encompassing 3G and 4G cellular standards), the nature of the technology does not relate to components, but to systems or networks that they are designed to optimize. Firms also prefer to extract higher royalties by licensing to end-product manufacturers, helping them to avoid patent exhaustion. Moreover, when a firm’s SEP component is functionally related or connected to its non-SEP components, bundling them together in a license helps the firm reduce administrative costs for itself and the licensees and increases its ease of verifying the number of units of its SEP-incorporated components sold.³⁴

* * * * *

When firms offer SEPs to licensees on a stand-alone basis on FRAND terms irrespective of whether they source other components from the SEP holder, and when those firms do not assert their SEPs at the component level, the firm’s pricing of its component is effectively constrained by the prices offered by competing non-integrated component manufacturers. Therefore, no competitive distortion occurs.³⁵ Firms with SEPs can impose different licensing terms on similarly situated integrated and non-integrated licensees without breaching their FRAND and subsequently engendering antitrust or contractual liability. In fact, openly acknowledging the legality of these already commonly practiced, although legally ambiguous, licensing activities will likely spur individual firm and overall industry innovation, and therefore should be endorsed, rather than prohibited, by courts.

* Arielle Koppell is a 3L at New York University of Law. This piece is a commentary on issues raised by the 2018 Global Antitrust Institute (GAI) Invitational Moot Court Competition held by Antonin Scalia Law School in Washington D.C. The problem presented a non-jurisdictional fact pattern raising a number of issues, including whether it is discriminatory for a SEP holder to refuse to license a standard essential patent to non-vertically integrated suppliers, resulting in an antitrust law violation or contractual breach of a FRAND commitment. It also raised the issue of whether licensing to suppliers at the technology portfolio level amounts to a breach of antitrust law. The views expressed in this article do not necessarily represent the views of the author. Rather, this article is a distillation of the analysis the author represented at the GAI Invitational, updated to reflect several doctrinal and policy shifts that have since occurred.

2. See Anne Layne-Farrar & Koren W. Wong-Ervin, Standard-Essential Patents and Market Power, at 37 (Geo. Mason L. & Econ. Res. Paper No. 16-47, 2016).

3. Whether licensing terms and fees are non-discriminatory is a separate inquiry from whether they are fair and reasonable. The non-discriminatory FRAND analysis focuses on differences in terms and fee arrangements offered by the SEP holder across licensees.

4. Id. at 38.

5. See Dennis W. Carlton & Allan L. Shampine, An Economic Interpretation of FRAND, 9 J. Competition L. & Econ. 531, 546 (2013).

6. See [2017] EWHC 711 (Pat) (Apr. 5, 2017) (hereinafter Unwired Planet) at ¶175, 806(8).

7. See Unwired Planet at ¶501.

8. See Case C-525/16, MEO v. Autoridade da Corroncência, ECLI:EU:C:2018:270 ¶24 (Apr. 19, 2018).

9. See Daniel A. Crane, Patent Pools, RAND Commitments, and the Problematics of Price Discrimination, in Working Within the Boundaries of Intellectual Property: Innovation Policy for the Knowledge Society 371, 373 (Rochelle C. Dreyfuss et al. eds. 2010); see also U.S. Int’l Trade Comm’n, Initial Determination of Administrative Law Judge, In re Certain Wireless Devices with 3G Capabilities and Components Thereof, No. 337-TA-800 at 432 (A.L.J. Shaw July 29, 2013) (“The FRAND nondiscrimination requirement . . . does not require uniform treatment across licensees, nor does it require the same terms for every manufacturer or competitor.”).

10. See Jorge L. Contreras, A New Perspective on FRAND Royalties: Unwired Planet v. Huawei, at 5 (U. of Utah C. of L. Res. Paper No. 206, 2017); see also Carlton & Shampine, supra note 5, at 546 n.41 (“There is no consensus in the literature as to whether FRAND should imply a uniform royalty level for all licensees [or whether different royalties can be charged] depending on each player’sbargaining power and business features.”); St. Lawrence Communs. LLC v. Motorola Mobility LLC, 2018 U.S. Dist. LEXIS 25229, 18 (E.D. Tex. 2018) (“[T]he Court is not persuaded that . . . negotiating different rates and terms for different licensees when presented with different circumstances . . . for a FRAND encumbered patent [is patent misuse].”); Richard J. Gilbert, Deal or No Deal? Licensing Negotiations in Standard-Setting Organizations, 77 Antitrust L. J. 855, 875-76 (2011) (“Non-discrimination . . . requires uniform treatment for similarly situated licensees, rather than identical treatment.”) (recommending that licensees be able to choose from the same schedule of royalties so they can select the combination of licensing terms and non-FRAND consideration that maximizes their value).

11. See TCL Commun. Tech. Holdings, Lwtd. v. Telefonaktiebolaget LM Ericsson, No. SACV 14-341 JVS(DFMx), 2017 U.S. Dist. LEXIS 214003, at *100-04 (C.D. Cal. Nov. 8, 2017) (acknowledging that experts define similarly situated firms as those at a similar level in the value chain and equally reasonably well-established in the global market, considering (primarily) their geographic scope, required licenses, and sales volume; defining them accordingly due to use of same technology and factors such as overall financial success or risk, brand recognition, operating system of their device, or existence of retail stores does not bear on the FRAND analysis of whether royalty rates for SEPs are discriminatory); see also Unwired Planet ¶ 487 (finding that both parties’ analyses presumed that “similarly situated parties, equivalent/ comparable transactions, and objective justification were the same under the non-discrimination limb of FRAND as they are in competition law.”); see also Jorge L. Contreras & Anne Layne-Farrar, Non-Discrimination and FRAND Commitments, in The Cambridge Handbook of Technical Standardization Law 186, 202-03 (Jorge L. Contreras ed. 2018, forthcoming) (discussing similarly situated licensees at Section 4).

12. Although ALJ Shaw commented that “nondiscrimination analysis . . . requires an examination of the whole of each license agreement, and not just the effective royalty rate.” U.S. Int’l Trade Comm’n, supra note 9, at 432.

13. Contreras & Layne-Farrar, supra note 11, at 2.

14. This justification was recently deconstructed and dismissed in Federal Trade Commission v. Qualcomm, Inc., No. 17-CV-00220-LHK (N.D. Cal. Nov. 6, 2018).

15. Id. at 7.

16. Id. at 6-7.

17. See Janusz Ordover & Allan Shampine, Implementing the FRAND Commitment, The Antitrust Source 1, 4 (Oct. 2014).

18. See Karl D. Belgum, The Next Battle Over FRAND: The Definition of FRAND Terms and Multi-Level Licensing, 39 New Matter 2, 7-8 (Summer 2014); see Quanta Computer, Inc. v. LG Electronics, Inc., 553 U.S. 617, 637, n.7 (2008).

19. Douglas H. Ginsburg, Koren W. Wong-Ervin, & Joshua D. Wright, The Troubling Use of Antitrust to Regulate FRAND Licensing, 10 Competition Pol’y Int’l Antitrust Chronicle 2, 6-7 (2015).

20. See generally Jorge Padilla & Koren W. Wong-Ervin, Portfolio Licensing, at the End-User Device Level: Analyzing Refusals to License FRAND-Assured Standard-Essential Patents at the Component Level (Oct. 19, 2016), https://ssrn.com/abstract=2806688.

21. See Erik Hovenkamp, Tying, Exclusivity, and Standard-Essential Patents, 14 Colum. Sci. & Tech. L. Rev. 79, 122 (2017).

22. Order Denying Motion to Dismiss, Federal Trade Commission v. Qualcomm, Inc., 17-CB-00220-LHK, at 12-13 (N. D. Cal. Jun. 26, 2017).

23. Id. at 41-42, 46.

24. Id. at 45.

25. 540 U.S. 398 (2004).

26. Order Granting FTC’s Motion for Partial Summary Judgment, Federal Trade Commission v. Qualcomm, Inc., 17-CB-00220-LHK, at 1 (N. D. Cal. Nov. 6, 2018).

27. Id. at 17-18.

28. Id. at 19.

29. Id. at 13-17, 23-24.

30. Assistant Attorney General, Antitrust Law and Patent Licensing in the New Wild West,” Remarks at IAM’s Patent Licensing Conference in San Francisco (Sept. 18, 2018) (available at https://www.justice.gov/opa/speech/
assistant-attorney-general-makan-delrahim-delivers-remarks-iam-s-patent-licensing).

31. Assistant Attorney General, “The ‘New Madison’ Approach to Antitrust and Intellectual Property Law” Keynote Address at University of Pennsylvania Law School (Mar. 16, 2018) (available at https://www.justice.gov/opa/speech/assistant-attorney-general-makan-delrahim-delivers-keynote-address-university).

32. Assistant Attorney General, Remarks at USC Gould School of Law’s Center for Transnational Law and Business Conference (Nov. 10, 2017) (available at https://www.justice.gov/opa/speech/assistant-attorney-general-makan-delrahim-delivers-remarks-usc-gould-school-laws-center).

33. See Edward A. Gold & Scott Weingust, Exploring the Nondiscriminatory Aspect of RAND Licensing Terms, Stout, Resius, Ross 1, 2-3 (Sept. 1, 2014).

34. Padilla & Wong-Ervin, supra note 20 at 18.

35. Id. at 19.