Netherlands’  The Hague District Court Orange-Book decision brings up again the discussion about the compulsory licensing defense and the role of antitrust in patent infringement proceedings. Opposing’s Germany highest court the Dutch Court denies the possibility of responding with a compulsory licensing defense in a patent infringement proceeding. How damaging is it that two major patent courts in Europe differ on such important issues? After all, we are supposed to have a harmonized European legal order in patent cases. Or maybe not? At the time of the German Orange-Book-Standard decision the German jurisprudence was engaged in thorough discussions in which major voices called for the ECJ (European Court of Justice) for a preliminary ruling to clarify whether a compulsory licensing defense base on antitrust antitrust law is admissible in patent infringement proceedings . Especially the danger of conflicting court decisions and a threat of a further blow to unified European patent law, as well as a possible revival of forum shopping was a major theme in discussions in Germany.

In general, regarding the German Orange-Book-Standard decision there was nothing new by admitting to object antitrust considerations against the alleged infringement of intellectual property rights. The ECJ has already done so in the striking decisions Magill TV-Guide, IMS Health and lastly Microsoft, although with regard to the latter not in a really direct speech. What is different between the mentioned cases and the German as well as the Dutch Orange-Book-Standard cases is that the former ECJ decisions deal with a total refusal to license IP that is essential to entry a secondary market. The German as well as the Dutch Orange-Book cases differ from this initial position this way that there was not the total refusal of a license by the patentee in question. The sticking point between the parties was the royalty rate for which the technology should be licensed, where there is a general willingness to license the IP on the licensor’s side.

Against this background it becomes clear why it is not easy to reach a concurrent view among courts in Europe on the issues dealt with in the Orange-Book-Standard decisions. The predominantly accepted guidelines of the ECJ regarding compulsory licensing by antitrust law are not readily applicable to these Standard cases, but are open to interpretation. We fail to see how -in the light of ECJ rulings- the recent Dutch judgement more adequately solves the problem than the German judgment. We have to judge the Orange-Book-Standard decisions by their practicability and impacts on prospective licensing behavior.

Applied to the particular Dutch Orange-Book case, the German as well as the Dutch solution would not differ in their results. In both cases the plaintiff would prevail and in fact, that is really noteworthy, for the same reason. Both solutions presuppose that the alleged infringer must have seriously submitted a license proposal to the patentee. SK Kassetten GmbH & Co. KG, the defendant in the recent Dutch proceedings, has never done so. Therefore in the particular case the antitrust law defense would not be applicable, not even under the German rule. Thus, we have to run through the next conditions of defending against an alleged patent infringement just theoretically.

While the Dutch court completely denies the existence of an antitrust law defense the German BGH stipulates further conditions to apply it, namely that the alleged infringer has behaved like a real licensee and paid or at least deposit in escrow a FRAND[1]-like hypothetical royalty. Nonetheless, even the Dutch court acknowledges the problems resulting from determining and agreeing on a FRAND-royalty term, and offers its solution of an interim injunction against the potential licensor, which could be initiated to have the patentee to conclude a FRAND licensing agreement or even replace the patentee’s signature under such agreement. The striking point remains the assessment of a reasonable royalty rate and in both solutions the responsibility of ascertaining what is “reasonable” is left to the court to decide. But in contrast to the Dutch solution, the German judgement sees the court’s responsibility only as the second best approach. The BGH gives more leeway to the parties and would consider the requirement of a presupposed offer to license by the potential licensee also fulfilled, if the potential licensee makes an offer to conclude a license agreement in which the potential licensor determines the amount of royalties according to its own “reasonable” discretion. What is done by this provision is nothing else than giving the power to determine the royalty rate to the licensor. Under this thinking, the licensor still can behave anti-competitive (or against antitrust rules). This is rather unconvincing. It is also not convincing, that the antitrust law defense can result in a compulsory license that is not limited in time. If the defendant in a patent infringement proceeding can actually prove all presupposes of the antitrust law defense the patentee would be completely precluded from asserting its patent. Therefore, the result has rightly lead to vehement criticism as it could potentially result in a completely unwanted and de facto permanent license.

In contrast, the solution of the District Court of The Hague appears more preferable since it permits an interim license agreement which could, at least temporarily, satisfy both the alleged infringer of the patent and the potential licensor until they agree to a mutually negotiated agreement. Faced with the impeding decision of the court the parties would be highly motivated to come to a mutual agreement. Thus, under the Dutch approach there remains a chance to solve the negotiation problem between the parties affected by it. And in this way we have the chance to avoid the introducing of a general price control regarding the licensing of IP, which everybody considers to be undesirable.

On the other hand, the conditions to get a license against the wishes of the patentee appear to be reached easier under the Dutch approach. The potential licensee only has to request the patentee for a FRAND license before he can apply for an interim injunction to get a temporary license granted by the court, provided that the parties can not agree on a license mutually. Compared with this, the presupposed FRAND offer by the potential licensee under the German antitrust law defense appears to be more difficult to fit. The potential licensee risks by any offer that the court would not follow its notion of a FRAND royalty term and thus he would jeopardize the antitrust law defense on the whole by underestimating what figure really fits the FRAND royalty rate. Otherwise, by exceeding a FRAND royalty rate the potential licensee could improve its chance to get a license but he risks paying too much for the license as the patentee has only to agree to the presupposed unconditional offer by the potential licensee. In fact, the Dutch Court is right in criticizing the legal uncertainty which is produced by this approach.

Hence both approaches finally emerge as not really practicable. What we really need are proper and reliable concepts in valuing IP to which the stakeholders as well as the courts could align. We admit, that in regular a fair royalty rates comes from mutual negotiations. Thus, it might be a solution to give up the general secretiveness of royalty rates by establishing a system of an open accessible royalty rate database. With it, royalty rates in question would be more comprehensible for both parties and would give less incentives to raise a controversy.

Stephan Dorn

[1] FRAND refers to “fair, reasonable and non-discriminatory”. Especially in the context of Standard Setting, IP proprietors the technology of which would be essentially used to implement a given standard are asked for a commitment to license their IP on fair reasonable and non-discriminatory basis. The concrete meaning of FRAND is heavily discussed and therefore a contentious issue again and again in proceedings regarding licensing and infringement of IP.