By Mel Marquis.
Bundling practices involving patents are fairly common, but in the EU, it can potentially raise competition law concerns. This short comment provides an overview of how EU law may apply in this context.
Bundling by a dominant enterprise might be more likely to restrict competition than the same practice adopted by a non-dominant firm because the degree of market power differs; however, the analytical framework should be the same. There should be an integrated assessment of (i) the nature and degree of actual or likely foreclosure and (ii) actual or likely efficiencies arising from the practice. Yet in Europe it is not clear that this currently is the case. The rigor of the assessment may differ depending on whether an enterprise is dominant. This suggests that the competition rules of the Treaty on the Functioning of the European Union (TFEU) may be applied in a divergent manner. However, a trend toward greater coherence may be afoot.
The Gist of an Analysis Under Article 101 TFEU
Beginning with the legal treatment of exclusionary conduct under Article 101, the general framework (unless a block exemption applies) involves an analysis of foreclosure effects and of any counter-strategies available to competitors, as exemplified by Delimitis v. Henninger Bräu.
The assessment of tying/bundling of complementary products (such as where consumables are tied to a patent license) under Article 101 would follow that framework. In a first step, the degree of foreclosure should be assessed. This requires an analysis of the parties’ market power and the proportion of the market for the tied product that is subject to the tied purchases. If the agreement is between non-competitors and the licensor has a market share (on any affected technology or product market) exceeding 30%, or if the combined market share exceeds 20% in the case of an agreement between competitors (i.e., if a block exemption does not provide an automatic safe harbor), and if the tie/bundle causes non-trivial foreclosure, then the conduct would likely have to be justified by efficiencies (e.g., where a tie permits savings of transaction costs, or where it is necessary for a technically satisfactory exploitation of licensed technology).
An agreement between competitors whereby they bundle substitute technologies would normally be regarded as a serious infringement. If the technologies are substitutes, such an agreement is likely to diminish (inter-technology) competition, resulting in relatively higher royalties and diminished choice for licensees. In such a scenario, a defensive claim to the effect that the bundling generates efficiencies and therefore justifies an exemption under Article 101(3) would likely fail.
Tying and Bundling from the Perspective of the EU Courts
The EU courts normally follow established case law, much of which was developed at a time when economic analysis was limited. In recent Article 102 cases, the European Court of Justice (ECJ) has suggested that so-called “loyalty rebates” a.k.a. “exclusive rebates” (conditional on the customer buying exclusively from the dominant firm) should be considered taking account of “all relevant circumstances.” This implies that functional analysis is creeping, inch by inch, to center stage. A new case concerning the tying or bundling of a patent with another product would be helpful to determine with more certainty whether the ECJ intends its recent case law (not just on rebates but on other pricing practices as well) to provide guiding principles for a whole range of commercial practices (including non-price-based conduct).
In light of the old case law, a dominant firm might still want to be prudent if it wishes to tie the sale (or the licensing) of product A to the sale of product B; and it should be prepared, if necessary, to show why the practice is “objectively justified,” or how it produces efficiencies which counterbalance or outweigh its adverse effects while benefiting consumers. In Hilti v. Commission, a dominant enterprise infringed Article 102 when it sold a packaged product called a “powder-actuated fastener system,” which included patented cartridge strips and non-patented nails. This “system” was intended for use in nail-guns, which Hilti also manufactured. In the EU Microsoft v. Commission media player case, a dominant enterprise infringed Article 102 when it pre-installed its media players in its “ubiquitous” Windows software. Although the then-Court of First Instance (CFI) accepted the Commission’s analysis of competitive effects, it did not consider that analysis to be essential for establishing an abuse. It would have been sufficient for the Commission to show only: Microsoft’s dominance on the market for PCOS; separate demand for media players; and inability of customers to purchase Windows from Microsoft without the company’s media player. On the whole, the old case law seems more rigid than necessary to prevent anticompetitive tying.
Article 102 may also apply to a dominant enterprise that obliges a licensee to accept a bundled product consisting of desired but also un-desired patents (including the bundling of standard-essential patents with patents that are not standard-essential). Adopting a prudent approach, a dominant patentee would allow a licensee to choose which patents it wishes to in-license.[1]
In brief, there does not yet appear to be a confirmed legal requirement for a decision-maker to apply an effects-based analysis in an Article 102 case involving the bundling of patents. However, this old approach may evolve when an opportune case emerges. Such a new direction would entail a rough convergence (by its nature a problematic concept) with Section 2 case law in the US, where a rule of reason applies to the tying of patented products with consumables (see Illinois Tool Works).
Tying and Bundling from the Perspective of the European Commission
The Commission prioritizes exclusionary (unilateral) conduct cases on the basis of economic analysis. Its general approach to tying/bundling under Article 102 may be stated succinctly: if an undertaking is dominant on the market for the tying product, it will take action if the tying and the tied products are distinct (i.e., there is separate demand for the tied product) and the tying or bundling is likely to lead to anticompetitive foreclosure. The term “anticompetitive foreclosure” generally means that the foreclosed rival is as efficient as the dominant firm. If the Commission finds apparently abusive conduct, it will nevertheless evaluate any relevant efficiency arguments (savings of production, distribution, or transaction costs). In the case of bundled complementary patents, the Commission would also consider any ‘Cournot effect,’ i.e., the internalization of pricing externalities (which would normally arise if the respective royalties for the complementary technologies were being determined by different entities that do not adequately take account of the other royalties to be charged).
In short, the Commission—a well-resourced agency that has significant economic expertise at its disposal—explicitly embraces economic analysis of tying and bundling for the specific purpose of prioritizing cases and deciding where to focus its efforts.
Conclusion
The legal treatment of patent bundling under Article 102 has traditionally been guided by the early development of the approach to exclusionary conduct by dominant firms—where the law was formalistic and interventionist. The situation is evolving, and in the future a more economically comprehensible doctrine may emerge, which would make Article 102 more coherent with Article 101.
Why have the EU courts clung to the past in some areas of Article 102 jurisprudence? First, EU judges are often conservative, at least outside the context of fundamental rights. Second, the ECJ is a collegiate body with no dissenting opinions. If a majority in a 3- or 5-judge (or in exceptional cases, a 15-judge) chamber favors change but cannot persuade the minority to budge, the status quo wins. Third, in Article 102 jurisprudence, there may be a bias against dominant companies. This is a European version of the old “big is bad” bias. Fourth, some would say that in Europe there is an abiding attachment to the notion that market operators should be protected when their freedom to compete is jeopardized. A negative spin on this is that it is a “protect competitors, not competition” framework. A positive spin is that the freedom-to-compete approach keeps markets open, contestable, and vibrant, while at the same time vindicating individual rights and economic liberty. The view of the present commentator is that the individual views of the ECJ’s judges across the span of generations has been far too heterogeneous to ascribe to them a common philosophical commitment to any particular, coherently constructed paradigm of competition law. As suggested above, the same heterogeneity feeds a status quo bias, since the status quo represents a common denominator for the 28 judges of the ECJ.
*The featured image is from The Federal Circuit Already Follows ABA and IPP Recommendations, The Intellectual Property Blog, https://blog.lrrc.com/ip/2017/04/10/federal-circuit-already-follows-aba-ipo-recommendations/.
Mel is Part-time Professor at the EUI in Florence, Italy and Professore a contratto at LUMSA in Rome. He is also a Senior Fellow and LL.M instructor at the University of Melbourne Law School. He has been Chair Professor at the Central University of Finance and Economics in Beijing, and he has lectured at various universities in Europe, the U.S., Japan and China. He has practiced law in the U.S. and Belgium. He publishes on antitrust issues in various regions of the world and his latest publication is Competition law in the Philippines: economic, legal and institutional context, advance access, Journal of Antitrust Enforcement, pp. 1-44., DOI: doi: 10.1093/jaenfo/jnx014.
[1] Separate issues arise in a scenario where the dominant firm refuses to grant a license for a given technology. There is no space in this short comment to discuss the substantial body of EU case law on refusals to license intellectual property rights.
