Getting the EU SEP Regulation right — Part 2: EU-wide law because of German rulings?

Imagine the following situation: you have half a dozen glassses of different sizes and colors on a table. The hypothetical objective is that each should be approximately 90% full. For the first five of those glasses, that is more or less the case. But the sixth and largest one of those glasses is almost empty: it’s at 20%. Now, the logical thing to do would be to pour some more water into that sixth glass until it also reaches 90%. What the EU is contemplating with the proposed SEP Regulation, however, is to determine how much water that sixth and largest glass needs (a lot!), and to pour the same quantity of water into each of the other five glasses, too. Indiscriminately. Not because it’s the right thing to do, but because that’s the way it operates, and if all you have is a hammer, everything looks like a nail.

Nowhere in the European Commission’s own impact assessment, all of the related material and even in the position papers of those lobbying for the SEP Regulation is there a reference to a single problematic court ruling in a SEP case in an EU member state other than Germany. The Commission’s legislative proposal appears to make a case for harmonization, but also says (incorrectly) that courts in different EU member states reached largely the same results in SEP cases, albeit on different grounds.

This legislative initiative would never have been taken if not for some unbalanced and partly outrageous decisions made by different German courts over the course of the last three to four years. But if the problem of SEP overleveraging is geographically confined, there is a clear and present danger of breaking the balance that other courts (such as in the Netherlands) have struck.

If there was a concern over the influence of German judges at the Unified Patent Court, given that most UPC cases (and possibly all UPC SEP cases) have so far been filed with German Local Divisions, it would at least make sense to await how the UPC’s jurisprudence evolves. The judges of the UPC Court of Appeal are from several different countries. There is a problem with the way German judges approach SEP cases (as will be explained in more detail toward the end of this article).

Subsidiarity means that a problem that is best solved at the local (or, in this case, national) level should be addressed there. In this case, the ones supporting the proposed regulation may believe it’s a “collateral benefit” if the rules apply EU-wide, but if the objective is to strike the right balance, then there will be collateral damage.

In 2019, even Volkswagen was not concerned about SEP injunctions in Germany and told its government so

There was a period from the mid 2010s to approximately the start of the pandemic when implementers of standards were not—or at least not profoundly—worried about German SEP injunctions. And as recently in 2019, the company that is arguably threatened by such injunctions more than any other (because of its manufacturing and sales in Germany), Volkswagen, even told the participants in an “expert talk” (a roundtable by any other name) hosted by the Federal Ministry of Justice of Germany, which was about a German patent law reform and particularly the idea of recalibrating the criteria for patent injunctions: “We are not here to talk about SEPs.” VW wanted to focus on the unsolved problem of non-SEP injunctions.

The reason VW’s then-chief IP officer Uwe Wiesner said so was because those opposing a meaningful reform of Germany’s patent injunction statute argued that non-SEP injunctions weren’t much of a problem, and if companies like VW feared injunctions, it was about SEPs, but—they said—that problem had been solved by the European Court of Justice in Huawei v. ZTE (2015). VW disagreed that non-SEPs were not an issue (they had just paid half a billion dollars to Broadcom over a non-SEP that was later deemed invalid), but essentially agreed that there was no need for help from the German legislature with respect to SEP injunctions.

By the time the German patent reform bill was passed into law (summer of 2021), the pendulum had already swung back. That is a problem, but not an EU-wide one that would warrant EU-level intervention.

Let’s now take a look at how German SEP case law evolved. There was a Stone Age, an Age of Enlightenment, and unfortunately there is currently a New Dark Age.

Evolution of German SEP jurisprudence (1/3): the Orange-Book-Standard era (“the Stone Age”)

The first part of this artile series already mentioned—and implicitly criticized—the 2009 Orange-Book-Standard decision by the Federal Court of Justice of Germany (German-language court website). In that case, there was no FRAND pledge (and if there had been one, a third-party benefiary couldn’t have enforced its rights under German law anyway), but certainly market power because of a DVD standard.

In Germany, patent injunctions are practically automatic. Until the 2021 reform, the statute didn’t even mention the possibility of an injunction being denied on equitable grounds, and even now that it does, there hasn’t been a single case (in more than 2 1/2 years) in which the denial or tailoring of an injunction was even seriously considered.

Orange-Book-Standard sought to be minimally invasive: while affirming the principle that any patent infringement should result in an injunction, its logic was based on the concept that an injunction must be denied if the circumstances are so egregious that enforcement would immediately result in an antitrust violation. The underlying Latin rule (dolo agit qui petit quod statim redditurus est) means that he who seeks something from a court that must be immediately returned acts in bad faith.

Under Orange-Book-Standard, the only chance for an implementer to avoid an injunction was to make a binding offer to the patentee on terms so very generous that the patentee would commit an antitrust violation by rejecting them. In other words, the measure was not where the patentee’s demand was FRAND, but the implementer’s proactive offer effectively had to be supra-FRAND in order to eliminate any doubt that rejection would constitute an antitrust violation.

In fact, some German judges explained in open court they would assume that the patentee is entitled to an injunction unless there was not the slightest doubt that the license terms were too good to be refused. In two parallel Microsoft v. Motorola cases, a royalty demand of more than 2% of Microsoft’s Xbox and Windows sales by the patent holder (over video codec patents) did not dissuade the Mannheim Regional Court from entering those injunctions (the enforcement of which was prevented only by a U.S. antisuit injunction) as the focus was solely on Microsoft’s counteroffer.

In that same pair of cases, the Mannheim court even forced Microsoft to withdraw its invalidity challenges to the patents-in-suit, arguing that according to its interpretation of Orange-Book-Standard, a continuing invalidation attempt would constitute behavior that cannot be reconciled with the notion that an implementer would already have to behave in accordance with the terms of a hypothetical future license agreement before it is even concluded. That was an extremist outlier position that no other court adopted, and even in Mannheim it was not known to reoccur. Microsoft even moved a warehouse out of Germany (to the Netherlands) because of German SEP case law representing too much of a business risk.

There were also other issues, such as with respect to the amounts to be deposited (not a problem for Microsoft, but for cash-stripped defendants).

Evolution of German SEP jurisprudence (2/3): the Huawei v. ZTE era (“the Age of Enlightenment”)

The unsustainability of Orange-Book-Standard (and what the lower German courts made of it) got the EU involved for the first time when formal antitrust complaints were lodged over the behavior of Samsung and Motorola. Those companies, which are now firmly in the implementers’ camp, weren’t monetizers, but their actual objective was “mutually assured destruction”: they wanted to create a situation in which Apple and Microsoft, who were asserting non-SEPs though some were de facto commercially essential, would agree to zero-zero cross-licenses.

The fact that the EU helped to redress the balance in German SEP enforcement at the time does not mean that the proposal that is currently on the table is workable and proportionate. But it is true that the German judiciary had gone off the deep end and its self-radicalization ended only when the European Commission’s Directorate-General for Competition (DG COMP) announced an interim decision in a Samsung case. The announcement appeared to say that an implementer merely had to be willing to negotiate a license deal in order prevail on a FRAND defense.

That gave some German judges pause. At the time, SEP cases were rarely brought in Munich. For SEPs, the key venue was Mannheim, followed by Dusseldorf. In early 2013, not long after the EC’s late 2012 Samsung announcement, the Dusseldorf Regional Court referred Huawei v. ZTE to the ECJ. In Mannheim, a Nokia v. ViewSonic case would have raised similar issues at the same time, and the court surprisingly (contrary to the preleminary views indicated at trial) rejected the complaint (no infringement), thereby ducking the FRAND issue.

After the ECJ decided Huawei v. ZTE in 2015, there was a period of several years during which SEP holders didn’t obtain injunctions in the EU (not even in Germany) that made headline news. It was a quiet period. SEP holders focused on non-EU jurisdictions, particularly the U.S. International Trade Commission. And in the late 2010s, the UK became an attractive venue because of Unwired Planet v. Huawei (injunction enforceable unless implementer agrees to take license on FRAND terms to be determined by UK court).

Evolution of German SEP jurisprudence (3/3): the Sisvel v. Haier era (“the New Dark Age”)

After a few years of faithfully applying and correctly interpreting Huawei v. ZTE, the German patent judiciary started to look for ways to dial back. The Munich I Regional Court and one of the two patent litigation divisions of the Mannheim Regional Court (the Second Civil Chamber under Presiding Judge Dr. Holger Kircher, who had also presided over the above-mentioned Motorola v. Microsoft cases) started to doubt that under the FRAND negotiation framework of Huawei v. ZTE a SEP holder’s royalty demand would have to be analyzed before the implementer’s counteroffer, which would only be reached if the SEP holder’s initial offer was deemed FRAND. The Munich court was working on SEP Local Rules, and the Mannheim court held a Nokia v. Daimler trial in the spring of 2020.

The Mannheim court’s Second Civil Chamber was also calling into question the safe harbor that an implementer can avoid an injunction by agreeing to a license on FRAND terms that, failing an agreement between the parties, would have to be set by a German court, a safe harbor that actually goes back to Orange-Book-Standard. The regional appeals court (Karlsruhe Higher Regional Court) disagreed in an order on a motion to stay enforcement, but the Mannheim court didn’t give up.

The floodgates were opened in late 2020 with the Federal Court of Justice ruling in Sisvel v. Haier. In that case, the defendant was an utterly unwilling licensee who made a counteroffer only after about four years, shortly before an appellate hearing and at a time when the patent-in-suit had already expired. At that stage, the question of injunctive relief was purely academic except for the purposes of fee-shifting (“loser pays”). But the legal question of whether or not to sequence the Huawei v. ZTE analysis was obviously of transcendental importance.

The Huawei v. ZTE sequence would be the following:

• The SEP holder has to give an infringement notice.
• The implementer has to express an interest in taking a license.
• The SEP holder then has to make an offer, which has to be FRAND (otherwise no injunction).
• Only if the SEP holder’s offer is FRAND does the implementer have to make a counteroffer. If the defendant fails to make that offer, or it is not FRAND, then there will be an injunction.

The Federal Court of Justice made the disgraceful decision (though Haier was definitely a company that deserved to be enjoined in that case) to skip the analysis of the SEP holder’s offer. Like the Munich court and one division of the Mannheim court were contemplating at the time, the Federal Court of Justice said that Huawei v. ZTE wasn’t a strictly sequential test: all sorts of behavioral reasons could give rise to a finding that the implementer’s expression of an interest in a license on FRAND terms was not sincere in the first place.

The presiding judge who was the “mastermind” (or “master copyist” of the Munich and Mannheim schools of thought) of this decision even made it clear at subsequent conferences that he wouldn’t be prepared to submit that interpretation of Huawei v. ZTE to the ECJ in the form of another preliminary reference. He argued that the ECJ would always just point to what it originally said, as opposed to providing new clarifications, but they could have tried at least once.

Sisvel, to be clear, cannot be blamed. Litigants seek the best outcome, and Sisvel’s counsel had every reason to portray Haier as an unwilling licensee.

As a result of Sisvel v. Haier not making it clear that only the most egregious of circumstances would justify a decision on purely behavioral grounds, German SEP injunction law has turned into a kindergarten exercise of trying to identify the slightest failure of an implementer to act like a perfectly willing licensee in order to hand down injunctions no matter how far outside the FRAND range a SEP holder’s royalty demand may be.

Even the term kindergarten is a euphemism for the “logic” of the Mannheim Regional Court’s Second Civil Chamber’s Nokia v. OPPO SEP decision: the court held it against OPPO that it offered a lump-sum license as opposed to a per-unit royalty (with periodic accounting of the number of units sold), and not much later a publicly accessible UK court ruling revealed that Nokia actually told the High Court of Justice in London the very opposite: Nokia argued in the UK that only a lump-sum license structure would be FRAND.

The decision that has mobilized the automotive industry more than any other, though, and which appears to have been viewed most unfavorably by EU-level policy makers is IP Bridge v. Ford. In that case, the Munich I Regional Court’s Seventh Civil Chamber (then under Presiding Judge Dr. Matthias Zigann) did not consider it a prerequisite to an injunction that the SEP holder (Japanese licensing firm IP Bridge) would have to make a FRAND licensing offer relating to its own (relatively small) portfolio: it was deemed sufficient to offer a pool license that involved the vast majority of all 4G SEPs.

Unlike Sisvel v. Haier and what some made of it, IP Bridge v. Ford was neither intellectually dishonest nor disgraceful or idiotic. But it was exceedingly disproportionate: it effectively means that if a patentee proves that one SEP is valid and essential, the implementer has to license 10,000 patents worldwide. The number 10,000 is not a wild exaggeration: a leverage of 1:10,000 is indeed the order of magnitude we’re talking about here if we count assets as opposed to patent families.

Also, there were other decisions, particularly in Dusseldorf, where implementers were obligated to take pool licenses instead of just bilateral licenses. But the pools involved were much smaller.

Ford would have eventually taken the pool license anyway. If they had settled only with IP Bridge at the time, that wouldn’t have solved Ford’s problem. They were being sued by half a dozen other licensors of the same pool. Just like Daimler (now Mercedes) initially opted for piecemeal resolution (first Sharp, then Nokia, then Conversant) before taking a pool license, Ford would have preferred a one-stop solution over a far more costly set of 50+ bilateral licenses as well.

Why does German jurisprudence sooner or later gravitate toward SEP overleveraging?

There are cultural reasons for that, meaning the legal culture, the background of the judges and the country’s overall culture.

The fact that third-party beneficiaries’ rights to enforce contracts are not recognized (unlike in common law jurisdictions or the French stipulation pour autrui) doesn’t make the FRAND part easier, but that’s just a coincidence. It just happens to be like that in Germany, but is not a cultural issue.

The readiness to grant injunctions all the time is indeed very German. It’s ingrained in the German legal system, which absolutely prioritizes the performance of obligations. Something that is prohibited must be stopped, period. No way that monetary compensation could be adequate if a right is violated.

Counterintuitively, that doesn’t mean that the rule of law is truly strong in Germany, or that German judges do a particularly good job at defending it. A former president of Germany’s Federal Constitutional Court, Hans-Juergen Papier, published a book in 2019 in which he warned against citizens losing faith in the country’s waning rule of law.

It is indeed true that German laws tend to be enforced very vigorously and even rigidly against minor offenders while murderers and other perpetrators are treated with exceeding leniency. That results in absurd situations. A grandma who stubbornly refuses to pay the mandatory fees of the country’s state-run channels (which use that money to subsidize professional soccer in the country, to pay their executives and stars sky-high salaries, and to support the government through propaganda) may go to jail. But someone who raped a 10-year-old may be released on probation. It has happened more than once that children under the age of 14 read on the internet that they cannot be punished for any crime they commit, and then committed premediated murder, in cold blood. German judges or the association representing them wouldn’t speak out, and the jurists in the government would continue to defend the status quo.

Put differently, you don’t want to be a patent infringer in Germany, you don’t want to be a grandma who doesn’t pay the mandatory fees for state-run TV and radio, but it’s a great country to be a criminal 13-year-old as you can literally kill with impunity. The German government even had to change the law because its country’s criminal judges were so lenient that burglars were generally released on probation.

Unlike the concept of “equity” in common law, the philosophy of most German judges is that “Recht” (law) does not equal “Gerechtigkeit” (equity, justice). You can hear judges say so in court. You can hear German lawyers say so.

In the SEP context, but also when it comes to the question of non-SEP injunctions, a major part of the problem is that German judges are specialists, not generalists. By contrast, U.S. federal judges have to deal with cases across all fields all the time, so they tend to be fairly familiar with economic concepts. German patent judges do nothing but hear patent cases. They become very good at that, to the extent that some of them deserve an honorary degree in telecommunications network engineering. But their focus is narrow.

In the U.S. and the UK, judges typically work in private practice before they are appointed to the bench. In private practice, they learn what it’s like when you run your own “business” (a law firm) and have to keep it going to pay your bills, as opposed to going from school into public service and getting your fixed and 100% secure monthly pay. In private practice, they experience what it’s like when they have a client discussing with them what a devastating effect a court ruling would have on their business. German judges don’t have that experience. They live in their own world until they retire.

There are some German patent judges who definitely could have been great lawyers (as litigators and as legal advisers). For instance, Judge Andreas Voss (“Voß” in German) of the Karlsruhe Higher Regional Court and formerly Mannheim Regional Court is very charismatic and knows how to explain why one position makes sense and another is absurd. He will retire soon, and if someone persuaded him to become a partner (even if just “of counsel”) in a law firm, he could have a great second career, also with a view to the UPC. While his decisions were sometimes difficult to understand, Judge Dr. Kircher is probably the rhetorically strongest German patent judge. And let there be no doubt that Judge Dr. Zigann could be a first-rate litigator.

But they all became judges without a prior “gig” in private practice, in the real economy. And that makes it harder for German judges to figure out what a licensing negotiation means and how the threat of injunctive relief impacts it. They’re averse to economics: they don’t want to deal with numbers and economic logic. They know it’s outside their comfort zone that is due to a “high school, law school, bench” CV.

It’s disproportionate to change EU law over issues specific to one country

The situation in Germany definitely requires another recalibration comparable to what was described above as the Age of Enlightenment that was unfortunately too short (and where, to be fair, SEP holders might have faced too much hold-out).

The question is how to achieve such improvement. The EU Commission could have tried soft law in the form of a guidance on SEP enforcement, such as on how to apply Huawei v. ZTE. It could have made court filings. It could have talked to the German government about it.

Instead, the proposal that is on the table would throw out the baby with the bathwater, complicating SEP enforcement across the EU. And it wouldn’t even address the issues described above. The various German patent infringement courts will attach different levels of importance to any EUIPO-led FRAND determinations. In the end, SEP holders will still find venues where the post-Sisvel v. Haier behavioral focus will prevail. The question of whether a SEP holder must offer a bilateral license or can point to only a pool license (no matter how large the pool is) is not addressed by the regulation at all.