Context: The debate over standard-essential patent (SEP) enforcement in Europe and the need for legislative intervention would benefit from greater honesty on both sides. There are serious issues as virtually no company that German courts have found in recent years to implement an industry standard has prevailed on a FRAND defense. One can recognize the unsustainability of the situation without supporting the European Commission’s proposal for a SEP Regulation that is certain to harm innovators but may not bring down SEP licensing costs at all.
What’s new: This week, consulting firm BRELA (Baron Research in Economics & Legal Analytics) published a self-described “Empirical Analysis of the German Caselaw on SEP Injunctions after Huawei v ZTE” (BRELA company webpage) that falsely claims to “depict a reality that is more nuanced than the one described by the critics of the German jurisprudence” and refers to a “significant balancing of the outcomes in German SEP litigation cases” that just does not exist. The authors arrive at those conclusions in a reality distortion field in which they (a) selected a nonsensical period for their analysis that muddies the water and (b) even by their own admission didn’t even look at all the relevant decisions they could have analyzed if the objective was (or had been) to find out the truth. For instance, they count only 14 (!) SEP decisions by the Munich I Regional Court over the course of the last nine years. A non-patent equivalent would be a study focusing on U.S. abortion decisions if that study included a century before Roe v. Wade and furthermore narrowed the focus even for that nonsensically long period so that numerous mainstream decisions would not be counted. In addition selection bias, the study also puts up a smokescreen by conflating the technical merits of SEP cases (essentiality/infringement and validity) with the FRAND defense, while the latter is the sole topic of political debate.
Direct impact: That dual selection bias (nonsensical period and incomplete picture even for that period) renders the “study” completely unreliable for policy-making purposes and makes its conclusions nothing but propaganda designed to benefit the authors’ clients. The problem is that some organizations and individuals still appear to believe that the same methods with which they have so far failed to gain traction in the political process will work if only they throw more money at them. They need to realize that their camp now needs to regain credibility with reasonably balanced and dfensible positions in order to influence the legislative process.
Wider ramifications: It’s important to bear in mind that there is no shortage of similar (if not worse) distortions in the European Commission’s impact assessment and the arguments provided in public by the Commission itself (such as a commissioner’s suggestion that only bad-faith actors criticized the proposal) and by SEP devaluation advocates. For example, the theory that the rapid growth of the IoT segment will threaten the viability of many small and medium-sized enterprises (SMEs) is purely conjectural, and no hard evidence has been provided for widespread SEP-related issues facing SMEs, though ip fray does believe certain SME-specific measures may be warranted and helpful.
ip fray primarily focuses on litigation and licensing questions, but does take an interest in policy matters, especially when there appear to be misapprehensions or, even worse, when there is disinformation.
Let’s take a closer look now at the three key defects (which must be features, not bugs) of the BRELA paper authored by Justus Baron (the front page also mentions his research job at Northwestern University), his partner or employee Santiago Bergallo and Clarivate’s Eric Sergheraert (the front page also mentions his research job at the University of Lille).
1. The period since Huawei v. ZTE is irrelevant because the controlling decision in Germany is now Sisvel v. Haier
The authors picked the wrong start point for the period they analyzed and then cut off the period, without objective justification, in early August 2023.
1.1 Wrong start of period
There is no plausible explanation for the selection of a period starting from the Huawei v. ZTE decision by the European Court of Justice (ECJ) in 2015 than a political agenda to distort the picture in order to do clients a favor. But it’s not going to work. “Nice try” would be a euphemism for that.
In summary, the period on which the authors focus is the one that contains as many years as possible during which it was more difficult to obtain a SEP injunction in Germany than it is now. It was easy before Huawei v. ZTE, and it is easy again after Sisvel v. Haier. The entire implementer-friendly period between Huawei v. ZTE and Sisvel v. Haier, which was merely an episode, is considered by the authors, while of the sole relevant period (since Sisvel v. Haier made an impact), approximately a quarter is missing at the end.
At first sight, someone might indeed think that a landmark SEP decision by the ECJ makes this a reasonable period to select. Was there a “before Huawei v. ZTE” and an “after Huawei v. ZTE“? Yes, there was for purely historical purposes, but it doesn’t make sense to focus on that period if the question is whether there is a problem with German SEP case law and when Huawei v. ZTE has effectively been vitiated in Germany (for all practical intents and purposes, it has been overruled, just that the Federal Court of Justice of Germany lacks the authority to do so officially).
The list of decisions (Appendix A1/list A1) at the end of the BRELA document states only one (!) decision since September 2020 in which a SEP holder was denied an injunction on the basis of a FRAND defense. In that case (Dusseldorf Regional Court, case no. 4c o 42/20), the issue was not even the reasonableness of a royalty rate, but the court merely disagreed with a mechanism for how to avoid double-dipping between two pools (MPEG LA and Access Advance). And indeed, ip fray is not aware of a second German SEP case since September 2020 in which a FRAND defense succeeded. According to rumor, a FRAND defense was rather likely to succeed in early 2023, but that dispute was settled.
The first of the two Sisvel v. Haier decisions came down in May 2020. There were some decisions that came down so shortly after Sisvel v. Haier that the new standard, which is exceedingly hostile to implementers, did not yet come to bear. The last two decisions of that kind issued in August 2020.
1.2 Unreasonable cutoff date
The last decision on the list is from early August 2023. For a paper published in mid-May 2024 (9.5 months later), that is a major shortcoming. Quite a number of SEP injunctions came down in Germany during that period, and not a single FRAND defense is known to have succeeded during that period.
2. Unnecessarily narrow focus on published decisions further distorts the picture
The authors limit their analysis to three sources: online databases of the German judiciary, decisions published on the website of the Dusseldorf-based law firm of Kather Augenstein, and darts-ip (which one of the authors runs). What they did not do is what academic researchers with a serious interest in finding out the truth would be expected to do: to conduct a systematic study of all SEP decisions in Germany. If they had partnered with a German university or research institute, they could have accessed (for purely scientific purposes) any case file they wanted to.
They also didn’t include decisions on which there were reliable published reports, in many cases even press releases by litigants.
They only have 14 Munich SEP decisions on their list. That’s a ridiculously low number, and so far no FRAND defense has succeeded in Munich (even regardless of period).
This unscientific (if not anti-scientific) approach by the authors results in an overweighting of outlier cases. When the law of the land is that a SEP holder who prevails on the technical merits is practically going to obtain an injunction, there’s obviously more of a reason to publish a decision in which it doesn’t happen. But the bigger factor is that Dusseldorf publishes decisions rather consistently, and that is the only court in which a FRAND defense has succeeded (albeit only once) in recent years. Munich, by contrast, does not systematically publish its judgments.
The authors deny that they were doing this for clients. But if they had done it for scientific purposes, they’d have had to approach the topic differently. They’d have had to select the only objectively relevant period (the current situation, or simply “the years after Sisvel v. Haier“), but where they’d have had to cast a wider net is to focus not only on published decisions but also on others they could have had access to or that were reported on. Instead, they cut off the period 9.5 months before the thin paper was published: during that period, no FRAND defense succeeded.
3. Conflation of technical merits and FRAND defense
In the 40-page paper, the term “validity” (or “invalidity”) comes up 41 times. But validity is absolutely not SEP-specific. That is off-topic for a paper that purports to focus on German SEP case law.
The net effect is that those who read the document will sometimes hear about SEP assertions that fail, but what really matters is whether a SEP holder who prevails on the technical merits is denied an injunction.
How nuanced is German SEP case law really?
There is nothing nuanced about a jurisdiction in which numerous SEP cases are adjudicated over the course of several years with a FRAND defense succeeding in only one outlier case (raising a very rare issue that in fact related to only one pool, with even the same pool administrator running other pools that keep clear of that problem).
There is an objective need to fix a problem. How to solve the problem is another question.
The UPC is going to hear its first SEP case only next month, a case in which there presumably is an unwilling licensee (based on the wide acceptance of the royalty demand by its peers and its own litigation tactics, such as by making a simple patent licensing dispute a geopolitical issue in the United States). It would also make sense to see how the UPC’s case law evolves, with that first case being just one, thus not representing a statistical sample and not answering the question of whether the UPC will deviate from German case law, no matter the outcome.
Some of the more recent SEP injunction rulings coming out of Munich (after the BRELA paper’s unreasonably early cutoff date) reflected an effort to go beyond what Sisvel v. Haier would have required. There have been decisions in Munich and other German venues that cannot be criticized for being biased. But it’s not plausible that every single implementer of a standard to have been sued in Germany, apart from one TV manufacturer that raised an issue relating to two overlapping pool licenses, has been an unwilling licensee.
The discussion over how to restore sanity in SEP enforcement is an interesting one, or it would be an interesting one if both sides made an effort to provide policy makers with reliable information instead of sometimes rather gross distortions.