This contribution is aimed at exploring what is the right role the “ND” portion within the FRAND commitment should play when determining if a certain license is indeed not FRAND. The literature on the subject is indeed broad, but this aspect has largely been downplayed, mostly relegated to the role of a nuance of the overall test aimed at determining when the royalty level can be said to be fair and reasonable. A valuable help to this purpose will be found in the recent EU jurisprudence on discriminatory conduct pursuant to art. 102, lett. c) TFEU where the Court has recently shed some lights on the concept of competitive disadvantage as direct consequence of the discrimination.

Sommario/Summary:

1. The treatment of discriminatory practices in EU competition law provisions. - 1.1. The requirement of comparable transactions in EU competition law. - 1.2. The element of competitive disadvantage. - 1.3. Clarification of the nature and proof of the “competitive disadvantage” element. - 2. The peculiar case of SEPs. - 3. The “ND” part of the FRAND commitment. - 4. LTA vs. ATA and the implications for 5G standards and IoT technologies. - 4.1. General vs. Hard-edged discrimination. - 5. Conclusion. - NOTE

1. The treatment of discriminatory practices in EU competition law provisions.

The principle of non-discrimination is a core one within the law of the European Union. Articles 2 and 3 of the TEU make clear that the EU is funded on values of democracy and equality, where every form of discrimination will be fought against [1]. The right of an individual not to be discriminated against has been further recognized as a fundamental one by Article 21(1) of the Charter of Fundamental Rights. According to the EU Courts, especially within the framework of the case law developed around the Charter of Fundamental Rights, the non discrimination principle would be a particular expression of the more general principle of equality (contained in art. 20) [2], the latter demanding that that comparable situations must not be treated differently and that different situations not be treated in the same way [3], unless such treatment is objectively justified [4]. As far as competition law is concerned, it is well known that both Articles 101, 2°, lett. d) and 102, 2°, lett. c) of the TFUE prohibit conduct whereby two or more undertakings jointly or a single dominant firm apply «[…] dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage» [5]. The provisions have been interpreted as prohibiting discriminatory practices, on any grounds [6], provided that the discrimination would exert a (negative) commercial impact toward one or more firms acting at the same or at different levels of the production chain. Clearly, the conduct is prohibited, provided that there is not an objective justification for the discriminatory conduct [7]. Abusive conduct prohibited by the EU competition law provisions are punished when they affect trade between member States, the theory of harm being construed as either direct prejudice to consumers or, more broadly, as the disruption of the competitive structure of the market [8]. It is interesting to notice, however, that in the specific case of discriminatory practices, the very same provision (i.e. 102, 2°, lett. c) of the TFUE) requires as an explicit consequence of the conduct that trading partners are placed at a competitive disadvantage. This specification has important implications because it suggests not only that this type of conduct must target trading partners of the dominant firm and not consumers [9], but it is specifically demanded that [...]

1.1. The requirement of comparable transactions in EU competition law.

The analysis of the EU jurisprudence with regard to discriminatory conduct has developed around the concept of comparable transactions and the concept of competitive disadvantage. The comparability of transactions is crucial to assess the anti-competitive nature of the practice, as indeed a difference in the terms and conditions offered could find justification in the fact that the different counterparts are indeed not comparable because, for example, the undertakings are different in structure and size, they operate at different levels of the production chain, or they are located in different regions or Member States where there could be different market conditions, different sectoral regulations, etc. In a quite old decision on practices prohibited by Art. 60, 1°, second indent, of the Treaty, in the common market for coal and steel, the Commission clarified that for transactions to be deemed comparable a) they ought to be concluded with purchasers who compete with one another, or who produce the same or similar goods, or who carry out similar functions in distribution, b) they ought to involve the same or similar products, and in addition, it demanded that c) there ought not to be other relevant commercial features that essentially differ [13]. This line of reasoning has been then followed in later and more recent cases where the analysis of whether the commercial conditions were indeed discriminatory has been pursued by taking into account whether the commercial partners receiving the different offers were indeed situated in a comparable position on the market [14].

1.2. The element of competitive disadvantage.

The milestone case dealing with discriminatory practices in the EU is the very well known United Brands case [15]. As known, UBC used to charge different selling prices for the very same good (i.e. bananas with identical features and branded with the same trademark) to distributors/ripeners located in different Member States. According to the Court, not only was UBC gaining extra profits by charging discriminatory prices to some distributors, but such prices were also excessively high, as they had been wrongly calculated by leaving out one stage of the chain (i.e. the very same distributors) and by just taking into account the interaction between demand and supply between the vendor and the ultimate consumer [16]. In this case, however, the distributors were not competing among themselves, as the markets where the bananas were sold were indeed national. Hence, strictly speaking, there was not a competitive disadvantage of distributors intended as the diminished capability to compete among themselves, but rather an economic prejudice for the ones charged more, as compared to the others who would buy the same identical goods at a far more convenient price. Surely, the overall conclusion of the Court, in this case, was heavily influenced by the summing up of all conduct contested to UBC and, in particular, by the so called “green banana clause” having the effect of impeding exportation of the product from one Member States to another, pre-empting arbitrage, with the ultimate consequence of fortifying national barriers, against the goal of creating an EU internal market. However, in this case, the element of the competitive disadvantage was largely ignored [17], and the finding of abuse merely rested on the discriminatory treatment in terms of prices for equivalent transactions (i.e. the selling of the very same product) to similarly situated customers (distributors facing the same unloading and transportation costs). In other subsequent decisions, the EU Courts have seemingly ignored the element of the competitive disadvantage and have been contented with a finding that the discriminatory conduct would – or be likely to – cause a distortion of competition [18]. In 2007, the European Court of Justice rendered its sentence on the British Airways v. Commission case [19]. At that time, the debate on the “more economic” approach to Art. 82 (then) of the Treaty was a front [...]

1.3. Clarification of the nature and proof of the “competitive disadvantage” element.

The issue of how to properly construe the element of the competitive disadvantage within a unilateral discriminatory conduct has been recently discussed in great length by the Court of Justice thanks to a preliminary ruling lodged by the Tribunal da Concorrência, Regulação e Supervisão (Competition, Regulation and Supervision Court, Portugal). The case concerned the alleged discriminatory conduct pursued by the Portuguese Collecting Society for neighboring rights (Cooperativa de Gestão dos Direitos dos Artistas Intérpretes ou Executantes) “GDA” who had apparently set different terms and conditions – and in particular a higher royalty rate – to MEO, an entity which provided a paid television signal transmission service and television contents, as compared to the one practices to its competitor NOS Comunicações SA (“NOS”) [26]. The case was set aside by the Portuguese CA who, after an in-depth analysis of the market, and relying on costs, income and profitability structures of the retail offerings of the television signal transmission service and television content, concluded that the tariff differentiation did not have any restrictive effect on MEO’s competitive position, and lacking proof that the discriminatory conduct would actually put one or more undertaking at a competitive disadvantage no distortion of competition could be assumed (hence there could be no abuse of dominance). MEO appealed, arguing that the approach of the CA was flawed, insofar as it examined whether there had been a significant and quantifiable distortion of competition, in terms of actual harm suffered by the discriminated firm, whereas the EU case law had just asked that the discriminatory conduct be capable of distorting competition  on the relevant market[27]. Despite finding the NCA discourse sound and reasonable, the referring Court was not entirely convinced by such straightforward application of the effects-based approach to unilateral discriminatory conduct and its compatibility with the EU jurisprudence. Therefore it lodged a preliminary ruling. In its judgement, the Court of Justice referred to a great extent to its previous ruling in British Airways; it clarified, however, some significant features. First of all, the Court explained that it is not necessary for a finding of abuse pursuant to Art. 102, 2°, lett. c) that the abusive conduct [...]

2. The peculiar case of SEPs.

A standard-essential-patent is a patent insisting on a teaching, often a sliver of technology, that forms a standard [34]. While sometimes standards emerge de facto, as a result of a natural process within the market whereby consumers gradually gather all around the same product [35], often market participants themselves recognize the importance to elicit together a standard to be commonly adopted within the industry [36]. In this latter case, specific entities are created – so called standard setting organizations (i.e. SSO) – which will deal with the standardization process and make sure that the best technology is selected and that it is then accessible to all market participants [37]. This precisely because, although not legally binding, conformity to such standards becomes extremely valuable to manufacturers willing to produce standard-compatible devices, as their products would be far less appealing to consumers when not standard-compliant (think of a new generation smartphone only capable of connecting to GSM technology) [38]. When access to the standard becomes then essential to compete in the market, control – thanks to IPRs – on such technologies becomes the key to open the door to a wide stream of profits, as all manufacturers will need a license from the IPRs insisting on the standardized product or technology [39]. In order to avoid opportunistic behaviours by SEP holders, both in the phase before the standard has been selected and after [40], many SSOs have adopted detailed IPRs policies which compel their members to i) timely disclose the holding of patent(s) that might insist on a technology considered to be chosen as standard or to be further developed to become a standard; ii) endorse a duty to license such rights, to whomever interested, on fair, reasonable and non-discriminatory conditions (so called FRAND terms) [41]. The duty to disclose should eliminate the risks of so called patent ambush, as it happened in the Rambus case [42], while the commitment to license should eliminate the risk that the SEP holder incurs in unfair and abusive conduct. In fact, the requirement of fairness and reasonableness clearly aims at preventing exploitative unilateral abuses in the form of excessively high royalty rates and unfair trading conditions [43], which the SEP holder might try to impose on third [...]

3. The “ND” part of the FRAND commitment.

As well known, while the literature has been mostly keen to the appropriate determination of the FRAND royalty rate on the perspective of how to measure a fair and reasonable price for the license [50], the “ND” aspect of the FRAND commitment has remained largely in shadow. This could be mostly attributed to the fact that many commentators have interpreted the “ND” prong as a feature to be read together with – in a way to reinforce – the element of fairness. In this way, non-discrimination would be read as a nuance of the general obligation to act in good faith and grant access to all interested parties, without any discrimination among them both with regard to access itself and to the contractual conditions of such access [51]. In the second part of this study, we are going to explore how we can better construe the concept of ND in the FRAND obligation by factoring in the recent achievements of the EU jurisprudence on art. 102, 2° lett. c). To this purpose, two questions are relevant. The first question concerns whether the SEP holder can or cannot discriminate among different categories of implementers situated at different levels of the production chain by choosing at which level to license or whether, on the contrary, respect of the FRAND commitment demands that he has a general obligation to license to whomever asks for a license, regardless of which level he is situated at. The second question regards whether the ND prong should be interpreted as an implicit element, automatically satisfied when the FRAND royalty is set, or rather whether it should play a more active role within likely antitrust controversies on the matter. These questions, of course, are only relevant assuming that antitrust in general has a role to play in such cases [52].

4. LTA vs. ATA and the implications for 5G standards and IoT technologies.

Differently from what generally happens with general IP licensing, access to the essential technology forming the standard is likely to be claimed by a quite heterogeneous type of undertakings along the production chain: from the manufacturer of the component that physically implements the standard (for example, a manufacturer of a chipset embedding the wireless communication technology) to the producer of the complex end product sold to consumers (i.e. the smartphone vendor) [53]. Furthermore, the producers of end products incorporating the standard can be very different ones: from smart-phones companies to smart-fridge vendors to car manufacturers. Since patent holders are supposed to extract profits only once from the patented product, because of the so called patent exhaustion doctrine, the question of whether the ND portion of the FRAND commitment gives rise to an obligation to license to whoever asks for a license carries significant monetary implications. Indeed, should SEP holders be compelled to license at the component level, they would exhaust their rights there; however, they would gain way more by licensing far downstream to end product manufacturer, where the profits are higher [54]. Scholars have taken different stances on such a controversial issue. Proponents of the so called “access to all” approach argue that, in order to be adequately compensated for their investment in technology, SEP holder should be allowed to choose the most profitable category of market operators to whom to license (hence they should be entitled to choose to license exclusively at the end product level) [55]. Within this framework, implementers situated in other (most likely upstream or intermediate) levels would simply access the standard with no need for a license [56] and the ND prong of the FRAND commitment should only demand that SEP holders do not discriminate between implementers within the same level of the market [57]. A corollary (i.e. that SEP holder does not discriminate between licensees who happen to be situated at the same level of the production chain and are direct competitors) which seems in line with the later findings of the Court of Justice in Meo. Proponents of this approach further stress that when 5G will be completely operational, and the Internet of Thing will enshrine a new set of interoperable services, the full functionality of the standard will be captured only at the level of the end [...]

4.1. General vs. Hard-edged discrimination.

At a general level, it is believed that the ND portion of the FRAND commitment is intended to promote widespread adoption of a standard within a certain industry, as it ensures market participants that if and when they will implement a standard technology into their product, they will not be disadvantaged vis-à-vis competing market participants [68]. By the same token, it has been underlined that non-discrimination commitments can also work as strong embankments against potential hold-up by SEPs holders [69]. An in-depth analysis of the ND limb of FRAND commitments can surely be found in the British case Unwired Planet v. Huawei, where two possible interpretations of the ND portion were compared: a first “general” non-discrimination approach vis-à-vis a hard-edged one, eventually adopting the former. Justice Birss embraced the general non-discrimination approach favoured by UP, which claims that the ND portion is «[…] part of the overall assessment of the inter-related concepts making up FRAND by which one derives a royalty rate applicable as a benchmark. This rate is non-discriminatory because it is a measure of the intrinsic value of the portfolio being licensed but it does not depend on the licensee» (italics added) [70]. Therefore, the general approach would imply that a royalty rate appropriately determined is, per se, also non-discriminatory, with no need for further comparison [71]. By contrast, the hard-edged approach would imply a distinct element to be factored in the analysis, «capable of applying to reduce a royalty rate […] which would otherwise have been regarded as FRAND» and this by taking «[…] into account the nature of the particular licensee seeking to rely on it» [72]. The approach of Justice Birss has been later confirmed by the Court of Appeal [73], whose judges considered the hard-edged alternative as being excessively strict and incapable of achieving a «[…] proper balance between a fair return to the SEP owner and universal access to the technology, without threat of an injunction» [74]. The Judge of the CA further stated that that «an effects-based approach to non-discrimination is appropriate» and that a non-discrimination rule had the potential to harm technological development of standards if it has the effect of compelling the SEP [...]

5. Conclusion.

Is it really true that a properly tailored ND rule would affect technological developments? The FRAND debate could probably earn a lot by looking more closely at the recent jurisprudence of the CJEU in non-discrimination cases. The Court of Justice, indeed, has gone from a first per se approach in UBC where the mere act of discriminating between market operators was enough to rule in favour of abuse, to a more thoughtful approach which depicts the theory of harm around the concept of distortion of competition and competitive disadvantage. The most recent cases have shown that a discriminatory conduct can only be punished if it causes a distortion of competition in the form of an alteration of the competitive relationship of downstream operators. It specifically requires that the undertakings in the downstream market be direct competitors, and it seems to require evidence that the conduct has somewhat a substantial impact on the overall market position of the discriminated party. Such insights could be preciously imported into the FRAND framework to construe a third intermediate approach between the general one, which has the practical result of depriving the ND limb of any significance, and the hard-edged one, which is blamed to threaten innovation in standards. Sidak has recently put forward a framework whereby to properly assess the ND limb of the FRAND obligation, it would be necessary to pursue a three pronged analysis: 1) analyse whether the supposed discriminated implementer is similarly situated to others by assessing whether they use the standard for comparable products, etc. 2) analyse whether the SEP holder is treating the implementers differently; 3) whether there is any objective justification for the different treatment [75]. Our intermediate approach would suggest to allow the allegedly discriminated willing licensee to claim that the proposed offer is not FRAND every time he is able to prove that the offer places him at a competitive disadvantage vis-à-vis other firms that are his direct competitors [76]. Not only, therefore, similarly situated customers, but similarly situated customers who must be in direct competition between themselves [77]. In this case, it would be upon the alleged discriminated firm to prove that the conduct of the dominant firm has had an impact on his costs, profits or any other relevant interest.

NOTE