Wednesday, October 19, 2022

Frustrated over Avanci patent pool's market penetration, critics ask DOJ to revisit Business Review Letter--but their disagreement with German patent law is no U.S. antitrust problem

I am flattered by the fact that both sides of the automotive standard-essential patent (SEP) licensing debate--those with a preference for licensing at the end-product level as well as those blaming Avanci and some of its licensors for not extending component-level licenses--point decision-makers to my writings. About ten months ago, Avanci licensor Acer sued Volkswagen and pointed to a couple of FOSS Patents on so-called licensing negotiation groups. And on Monday, "28 former government enforcement officials, professors, and public interest advocates" wrote a letter to U.S. Antitrust Assistant Attorney General Jonathan Kanter, asking him to revoke or downgrade his predecessor's July 2020 Avanci Business Review Letter" (PDF) that contains references to four FOSS Patents posts (footnotes 9, 16, and 17).

This is appreciated, but it doesn't make me look at the document through rose-tinted glasses. There are clear errors in that letter, and even if one agreed with some or all of the authors' policy positions, their approach is wrong in two fundamental ways:

  1. That they see a problem is no reason to blame a part of the solution for not solving all of the world's problems.

  2. If one thinks it through, the root cause of everything they criticize is unrelated to U.S. antitrust law and actually just a problem with Germany's patent law, including its automatic injunction regime and the "injunction gap." Therefore, instead of writing to AAG Kanter, they should have directed their grievances to Dr. Marco Buschmann, Germany's Federal Minister of Justice.

My policy concern is that this misguided pool-bashing can do damage to the entire patent pool universe. Indeed, MPEG LA, which is actually a competitor of Avanci and especially its sister company (Marconi) filed an amicus curiae brief (PDF) with the Fifth Circuit in a Continental v. Avanci case in support of Avanci because it sought to protect the ability of patent pools to operate. What we're talking about here is not just Avanci and MPEG LA, but so many others, such as Sisvel and Via. What has me profoundly worried is that the signatories of the letter to the DOJ's Antitrust Division seek to mischaracterize as antitrust issues two principles without which patent pools simply cannot function:

  1. No pool can offer licenses beyond the scope for which there is a consensus among its licensors. Yes, that is the narrowest common denominator--like many things in life, like it or not. No, it's not an antitrust issue so long as those licensors who are prepared to grant licenses beyond the scope of the pool remain unrestricted in their ability to accomodate such other licensing requests.

  2. No pool that is set up--like all the pools I know--as a licensing agent (as opposed to an entity that would acquire patents, i.e., be assigned all rights to those patents, which some pool administrators do in addition to acting as a licensing agent) has standing in an infringement action. Only its licensors can bring suit. Now, if you have dozens of licensors, you would inevitably get free-riding (that some wait for others to take enforcement action against unwilling licensees)--possibly to the extent that no one would do anything--unless there is some reimbursement mechanism for scenarios in which a defendant settles not just through a bilateral license but takes the pool license. It's a matter of fairness--also vis-à-vis honest licensees who would otherwise be at a disadvantage as they compete with infringing competitors.

The letter does not argue that patent pools should never be allowed to have a limited scope (field-of-use restrictions; here, vehicle-level licenses to car makers) or shouldn't reimburse contributors for taking the risks and shouldering the overall burden of suing infringers. Furthermore, it does not (as it cannot) allege that Avanci's contracts with licensors preclude them from granting bilateral component-level licenses to automotive suppliers like Continental.

So what are they really complaining about?

What's already clear from the headline of the Public Interest Patent Law Institute's related press release (The DOJ Should Review Avanci’s Patent Pool and Revoke its License to Troll) is that they don't like non-practicing entities (NPEs). However, the fact that Avanci's contributors include a mix of operating companies (such as LG, which is in fact a major automotive supplier) and NPEs is not an antitrust issue. The business of NPEs is legal. As the Business Review Letter and Avanci's related request show, Avanci as a licensing agent does not discriminate against any category of licensors. Whether they could discriminate is a question we don't have to think about: they've chosen not to do it, which is legal--and I don't know any patent pool that doesn't allow NPEs to join.

The letter says Avanci's reimbursement policy--which obviously makes reimbursement subject to someone taking a pool license, as it wouldn't work any other way, and without reimbursement pools would be defenseless against patent hold-out by infringers--leads to the same effect as "pool licensors collectively agree[ing] not to license outside the pool." But that makes no sense because those are two different categories of licensees. Take IP Bridge v. Ford: if Avanci's 4G pool works largely like the 5G pool for which the DOJ issued the BRL, IP Bridge got reimbursed when Ford settled. But Ford couldn't have taken a component-level license because it doesn't make such components. And if IP Bridge had wanted to collect royalties from Ford's suppliers, it would have sued a company like Continental (which was merely an intervenor supporting Ford).

Some Avanci licensors have indeed--after joining Avanci--extended component-level licenses, even Conversant Wireless, which is one of the NPEs the letter particularly criticizes for their business model. Some others are known for their policy of licensing end products: Qualcomm, Nokia, Ericsson, to name but the most famous ones. But they've had that position for a long time, even long before Avanci was founded. It is one of the glaring weaknesses of all that criticism of Avanci that no one can point to a single patent holder who would have extended component-level licenses before, but--because of Avanci--changed that policy. Not even one.

The letter labels as "collusive litigation" certain past situations in which car makers were sued by multiple Avanci licensors in parallel (it is mostly in that context that FOSS Patents gets referenced, as there is no other online resources that tracks automotive SEP litigation so closely). But that's just the inevitable consequence of unlicensed use of patents. For example, Apple is currently being sued by Ericsson (in multiple countries), Optis (after a UK trial, a judgment is in the works), numerous other NPEs (one of which is Clear Crystal Codec), and was on the verge of being sued by InterDigital.

I, frankly, doubt that it would even be illegal for multiple patent holders to coordinate litigation. Whenever MPEG LA announces infringement actions (which they usually bring in Dusseldorf), there are patents from different MPEG LA licensors that get asserted against a single defendant. And it's not unheard of that companies facing infringement lawsuits over the same patent(s) engage in some coordination: they form joint defense groups or compare notes. A group boycott would raise competition concerns; discussing prior art would not.

The "collusive litigation" part culminates in a nonsensical but-for scenario:

"Absent the Avanci pool, members would be competing for licensees [...]"

That is an insult to human intelligence because, as the BRL stated and the letter to AAG Kanter doesn't dispute, SEPs reading on the same standard are by definition complementary: they do not "compete." For competition to take place, you need a substitutive potential. Where you do have competition between patent licensors is when there are alternative technologies: for example, if there are alternative ways of cooling down a seat and different patent holders (or pools) offer patents that solve the problem in different ways. But Nokia isn't "competing" with Ericsson, or InterDigital with Qualcomm. They may compete for engineers and for investors, but never for licensees. Nokia can't go to an implementer and say: "Our patent license costs less than Qualcomm's, so take it!" Undercutting wouldn't work because in the end, the implementer will need to license both Nokia's and Qualcomm's patents--and those of dozens of other SEP holders.

The first accusation in the letter recalls that the BRL did not rule out the risk that "non-infringing manufacturers" could be forced to "settle and take a [...] license." The letter claims that "this is exactly what has happened." But the facts that the letter point to do not support that claim at all. There were no "non-infringing manufacturers" that took a license, just numerous car makers who took a license without litigation--and where litigation was brought and reached the point of a judgment, the automakers lost--most recently Ford. Someone who loses an infringement case cannot be honestly described as a "non-infringing manufacturer[]."

Closely related is the point that infringement litigation fails to "promote the removal of invalid patents from the pool." According to the letter, which then points to IP Bridge v. Ford, "litigation does not lead to validity determinations when licensors seek injunctions, as Avanci’s members have done." The BRL, they say, "does not consider that injunctions before validity determinations force companies to take their products off the market—and risk going out of business—long before any invalidity (or infringement) determinations are made."

By the way, it is simply wrong that they claim Avanci's licensors sought "preliminary" injunctions: those injunctions were permanent ones, after a full trial. Also, they claim litigation was brought in "Frankfurt"--they might have meant Mannheim, but that is a different city and in a different state, about one hour away.

They may not have researched those cases thoroughly enough, but it is clear that they have a problem with how patent enforcement works in Germany. That's why I said further above that they should have written to Germany's Minister of Justice.

It all boils down to four characteristics of German patent law that those Avanci critics don't savor because of the commercial realities they create on a daily basis, with or without Avanci:

  1. German courts routinely enjoin patent infringers as opposed to applying an eBay-like standard. The new proportionality defense is meaningless.

  2. At least the courts in Dusseldorf (where most codec cases used to be filed in the past) and Munich (as in IP Bridge v. Ford) typically tell a SEP implementer that they must take a global pool license as opposed to just a bilateral license. Ford couldn't have solved the problem by just taking a license from IP Bridge, but that's because the Munich court held that Ford violated its FRAND obligations by not making a counteroffer to the Avanci license that IP Bridge pointed to. FOSS Patents was first to publish, translate, and explain that judgment.

  3. At least the courts in Mannheim and Munich consistently rejected defenses by car makers based on the theory that their suppliers were allegedly willing licensees. The courts essentially say that patent holders have the choice against which level of the supply chain to bring their cases.

  4. Finally, there's the "injunction gap": there is no full invalidity defense in a German patent infringement action--just an assessment of the likelihood of success on the merits of a parallel (and slower) nullity proceeding in the Federal Patent Court, based on which assessment the infringement case may or may not be stayed. That indeed leads to invalid patents often surviving just because the infringer comes under settlement pressure before a final validity determination. The German legislature was lobbied about this, and earlier this year a legislative amendment relating to the injunction gap took effect, but as I explained, it's not going to be truly useful and may even be counterproductive.

There is nothing that even the powerful United States Department of Justice can do about how the courts in Germany interpret and apply the laws. This leads us to a very important question: Is a pool like Avanci (or MPEG LA, for that matter) part of the problem or part of the solution?

It is perfectly reasonable to disagree with the four above-mentioned features (or bugs, if you dread them) of German patent law. But the DOJ can't change them, and Avanci isn't responsible for them. So if we talk about a realistic But-For World, we have to do so against the background of what the courts in Germany--with Munich being the primary hotspot--would do with or without Avanci:

  • Those patent holders--including the NPEs whose business model some people don't like--would still exist.

  • Some patent holders (including major operating companies like Qualcomm, Nokia, and Ericsson) would not grant component-level licenses anyway.

  • Car makers would still be implementing those standards (particularly 4G and now 5G).

  • SEP holders wouldn't condone infringement forever. They would sue. There would be more litigation, not less, but regardless of how many cases would be brought, there would be a very significant number. And car makers would often settle after (or in anticipation of) injunctions.

  • As I explained further above, it's preposterous to suggest that SEP holders would be "competing" with each other. Each and every SEP holder would want each and every car to be licensed, and a license to, say, Nokia's patents doesn't mean you don't need to take a license to Acer's patents. A license to patent A is never a defense to an infringement allegation concerning patent B.

In the end, Avanci gets blamed for not convincing companies with a longstanding policy of licensing at the end-product level to change their positions, and for providing a one-stop shop that is optional because licensees could always seek (and sometimes indeed secure) bilateral licenses from any or all of Avanci's licensors. If a court of law tells them they're not entitled to a bilateral license because the pool license is acceptable, that's not an antitrust problem, and when it happens in Germany (as it does), it's nothing the U.S. government is responsible for.

The only realistic But-For World is one in which transaction costs are higher, and license fees would certainly not be lower.

The shortcomings of that letter don't end here. Let me show you a really disingenuous passage. They argue that Avanci's rate increase from $15 to $20 per car (for a 4G license), which was announced well in advance of taking effect and only applied to new licensees, was unjustified for the following reasons:

"There has been no change to the decades-old standards to which Avanci’s license pertains that could justify any fee increase, let alone such a significant one. The only change has been to the number of vehicle manufacturers that have taken Avanci licenses, many of whom did so to avoid preliminary injunctions, both potential and actual."

Nowhere do they acknowledge that the old rate had been in effect for about five years (which was when BMW accepted it); that it won't apply to existing licensees; that there has been and continues to be inflation; that there's nothing wrong with incentivizing compliance and early sign-ups; and above all, that Avanci's pool has grown massively. For instance, the addition of LG alone would have been sufficient to justify the rate increase (which, again, didn't even impact existing licensees).

What's the agenda here? What's likely to happen next?

Let's start with the signatories. Some of them--such as Professor Herbert Hovenkamp--are definitely independent; they may have been disinformed by some others. But some have an agenda, and I believe that agenda extends beyond the automotive sector. The Public Interest Patent Law Institute has previously collaborated with Big Tech astroturfers: they filed an amicus curiae brief together with the so-called Computer & Communications Industry Association (CCIA), which is actually a Cash & Carry Industry Association, and another signatory organization was ACT | The App(le Association. All three of those organizations--PIPLI, CCIA, and ACT--are among the backers of the "Save Our Standards" campaign, which engages in deceptive lobbying and campaigning (they sponsored an "interview" with a small app developer who claimed to have problems with SEP licensing though it never had to license a SEP, much less in the project they described in that interview, which was an app they developed for a state government that is immune to patent law).

The signatories include at least one professor who routinely submits pro-Big Tech amicus briefs on patent law issues, and co-authored a paper with an ACT astroturfer.

That letter to AAG Kanter was probably just the opening salvo, and who knows how much lobbying has taken place behind the scenes already. Also, the timing is interesting because we're approaching the deadline for Continental to file a cert petition after losing its Avanci lawsuit in district court and in the Fifth Circuit.

The letter fails to propose how to address that criticism and still enable patent pools to operate successfully and provide the transactional efficiencies that are their raison d'être. If they want to change German patent law, they should go to Berlin, not Washington.