IPR Licensing and Antitrust – The Transatlantic Divide

posted by A Layne on Wednesday, September 23rd, 2009 at 5:08 am

Competition agencies on both sides of the Atlantic have been keen to pursue cases on IPR licensing in standard setting. In the United States, it is the Federal Trade Commission leading the charge, with the Dell Corporation decision in 1996, the Union Oil Company of California (Unocal) consent order in 2005, the Negotiated Data (N-Data) settlement of 2008, and the long but ultimately unsuccessful Rambus case that came to a close just this year. In Europe, the European Commission also investigated Rambus, picking up the strands laid out by the FTC and reaching a tentative settlement in June of this year. The EC also opened a formal investigation of Qualcomm in 2007 (after two years of “informal” investigation). And at least one European member state has entered the fray, albeit at the court level rather than at the level of the national competition authority, with the IP-Com v. Nokia litigation in Germany.

Ex ante expectations

A common thread in all of these matters – either in Europe or the US – is the divergence between ex ante expectations/obligations and ex post behaviors. Consider Rambus. The concern here was that Rambus chose to participate in discussions at the standards body JEDEC but failed to fully disclose all of its relevant patents during the meetings. When JEDEC included Rambus technologies in its memory standards (SDRAM and DDR-SDRAM), its members did not expect to pay any royalties for them. Rambus, however, did expect to be paid for its intellectual property, and thus it pursued licensing fees and sued several JEDEC members for patent infringement once the JEDEC standards hit the marketplace.

At the heart of the Rambus dispute in the US was JEDEC’s IPR policy, which requested members to disclose potentially relevant IPR to other JEDEC members during working group meetings. The FTC argued that this policy created a reasonable expectation on the part of JEDEC members that Rambus’ silence meant it had no relevant IPR. This was the line the EC took in its investigation as well. By voluntarily joining but not disclosing its IPR, then, Rambus acted deceptively. In the US, however, the Court of Appeals for the Federal Circuit disagreed. It found JEDEC’s policy so vague as to not imply any obligation of disclosure.

But disclosure is not really the fundamental issue; ex ante expectations are. Joining a standard setting body is voluntary – firms do not have to participate. Once a firm has joined, however, there are certain expectations regarding its behavior. Members of standards organizations want ex ante IPR disclosure because they want to be able to anticipate and plan for the cost of implementing a standard once it’s ready for commercialization. Seen in this light, a failure to disclose relevant IPR before a standard is promulgated can distort competition, perhaps even if the IPR holder does not attempt patent holdup ex post (as Rambus was accused of doing). Clearly ex post holdup, enabled by a surprise release of IPR that were heretofore hidden is a problem because it exploits investments that firms have made to implement the standard. It is the combination of those investments plus the surprise of IPR unknown ex ante that can enable a patent holder to charge licensing fees in excess of the economic value of the patented technology.

But suppose a firm fails to disclose its relevant IPR and then promises ex post to license those IPR on a reasonable and nondiscriminatory (RAND) basis. Is there harm here too? Possibly. For firms making regular disclosures and who are known to hold patents reading on a standard, missing a patent here or there isn’t likely to change anything. Fellow standards body members are already on notice that this firm has valuable patented technologies that it plans to license. In most cases, licensing negotiations cover entire portfolios, so the odd patent left off the official disclosure forms generally won’t materially affect the terms. If, however, the non-disclosure is on the part of a firm that is not otherwise known to hold relevant IPR, the omission could be significant. This follows because adding in this firm to the list of known patent holders may alter the aggregate royalties that must be paid to implement a standard by enough to affect an implementer’s decisions. Since the standards body members were not anticipating paying any royalties to this firm, if the cumulative total is appreciably different, non-disclosure may be seen as “deceptive” even if the terms are reasonable in isolation.

Ex ante licensing announcements

The theme of ex ante expectations versus ex post conduct is also the crucial one in the other standards cases listed above. For example, the N-Data case involved ex ante licensing commitments that a later patent holder attempted to renege on. In that case, a firm participating in IEEE for the development of Ethernet standards announced in 1994 that it would license its technology to all comers for the flat rate of $1000. IEEE then incorporated the technologies into its Ethernet standard the following year. A few years later, in 1998, the patents changed hands and the new holder attempted to announce different, although supposedly RAND, licensing terms.

The facts in N-Data are clearly different from those in Rambus: the patents were fully disclosed to IEEE, there was no surprise, and the ultimate patent holder was ostensibly not attempting holdup. And yet, the underlying issue is the same: the members of IEEE expected to pay $1000 ex ante but N-Data attempted to change that deal ex post.

Digging a little deeper into the case, however, one might question whose ex ante expectations were not met. In particular, N-Data attempted its alteration of the licensing terms in the late 1990s, several years after the standard was published and had begun to be commercially implemented. Arguably, the chain of firms holding the patents could have reasonably expected that any firm implementing the IEEE Ethernet standard would pay the announced $1000 flat fee. Certainly those firms that had paid the $1000 fee before N-Data acquired the patents would not have been affected by its newly announced terms. But as many as 64 implementers had not paid that fee when N-Data acquired the patents. One can question whether firms failing to pay the ex ante announced terms had any reasonable expectation for those terms to hold in perpetuity, even though they were technically infringing the patents by implementing the standard without a license.

The transatlantic (competition) divide

As this brief review of Rambus and N-Data illustrates, IPR licensing in standard setting can involve competition issues. Withholding IPR during ex ante standard formulation can lead to the anticompetitive disclosure of other firms (by excluding their IPR from the standard) and it can affect consumer prices on the marketplace ex post. Likewise, failing to honor an ex ante licensing commitment can distort competition and may also affect downstream prices. And failing to license IPR that is known to support a standard may affect the IPR holder’s competitive position in the marketplace, especially if that IPR holder is not vertically integrated and relies on licensing revenues as its source of profits.

It is not surprising, then, that both European and American competition agencies have seen a role for antitrust law in IPR licensing in standards. There are, however, some differences in approach. Most fundamentally, competition laws differ across the two jurisdictions. So, while the FTC could pursue Rambus for illegal acquisition of a monopoly through deceptive conduct at JEDEC, the EC could not. Instead, had the matter not settled, the EC’s charge would have to have been exploitive or excessive pricing.

On the other hand, agencies on both continents have at their disposal laws regarding “unfair” competition and “deceptive” behavior. That means that an N-Data type case could be pursued in Europe as well. And indeed, the Qualcomm matter currently in limbo before the EC does involve allegations that the firm holding so many pivotal patents for 3G mobile telecom standards has reneged on its promise to license its IPR on FRAND terms.

Finally, the debate over the relevance of other legal tools aside from antitrust policy appears to be largely centered in the US for the time being. Thus some American scholars have suggested that an alleged reneging of licensing terms (à la N-Data or Qualcomm) is not really an antitrust issue but is instead better dealt with using contract law. And of course, as always in the US, there is the tension over which cases the agencies should bring and which matters should be left to private litigation among the parties directly affected.

Questions for debate

The discussion above raises a number of interesting issues.

1. Should IPR policies at standards bodies be revised or refined to offer better guidance to their members and, most importantly, to set appropriate expectations for both IPR holders and implementers? Is that even possible given the diversity of interests at play in the typical standards body and the fact that firm participation is entirely voluntary?

2. Should standards bodies’ IPR policies be enforced in some fashion, aside from the current ad hoc approach taken by the FTC and EC? If so, how and by whom (again bearing in mind that most of these bodies are voluntary associations of private firms and universities)?

3. Given the fundamental differences in antitrust law in the US as compared to Europe, can we realistically expect convergence on these matters?

4. Is antitrust/competition policy really the best vehicle for addressing the kinds of issues that arise in IPR licensing in standard setting? Do we even need to pick just one legal mechanism, or is there value in using a multipronged approach?

recent DISCUSSION

Marcus Glader says:

Re (1) and (4): A multipronged approach appears to be way forward. The SSOs should be encouraged to make refinements to their IPR policies. Ex ante negotiation models and other potentially promising solutions to hold-up problems may work in some contexts but not in others. Certain issues can be dealt with as a matter of contract law, in particular where the IPR policies are clear. But it is unlikely that a district court judge, in contractual dispute between two parties, could impose an effective global remedy to a reneging IP holder. As in many other business contexts, it is important that abuses ultimately can be remedied through antitrust intervention if necessary. It is in the shadow this possibility that industry participants in most cases will find a reasonable solution, either on a bilateral or multilateral basis, without involving the antitrust authorities.

Re 3: I’m not sure the transatlantic divide is that significant. The US agencies and the EC Commission seem to agree on the important policy issues. While there are significant legal differences between Article 82 and Section 2, the end result may not differ that much. From a legal perspective, it seems to me that a patent ambush may have an exclusionary element as well as exploitative under Article 82. And the remedies currently considered by the EU Commission in Rambus are in any event similar to those that the FTC had in mind. Similarly, in Broadcom v Qualcomm, the US Third Circuit Court of Appeals held that a patent holder’s intentionally false promise to license essential proprietary technology on FRAND terms, coupled with an SSO’s reliance on that promise and the patent holder’s subsequent breach of that promise, is actionable anticompetitive conduct under US law.
The antitrust authorities also tend to analyze agreements on standards in conformity under Article 81 and Section 1, respectively.
Marcus Glader says:
Niklas Lindblom says:

Re (1): When it comes to standard bodies’ IPR policies, the clearer they get the better. That said, I would think that the problem rather lies in the fact that IPR holders sometimes, more or less deliberately, fail to meet the requirements of ex-ante disclosure. Therefore I would argue that the issue to tackle is how one should deal with non-disclosure. Forcing all IPR to be licensed on FRAND terms is perhaps one way, but that comes with a set of further questions…
Re (2): I do believe that IPR policies should be enforced. The question of how this can be done is harder.
Re (3): Since standardization is increasingly global, so should the behavior of authorities. The US and the EU are fundamentally different in several ways which unfortunately makes it hard to see any convergence in the near future. However, one would hope that the benefits of a mutual approach would become apparent to American and European legislators and agencies, sparking joint efforts to deal with IPR matters. Do you think that the problem with convergence could run deeper than just differences in the Competition/Antitrust laws (particularly related to IP law and litigation tradition, including settlement and damages)?
Re (4): I believe that the key is to set regulatory requirements of patent disclosure and agree with arebentisch’s view on an “integrated approach”. If a harmonized approach on patent and licensing rules can be reached, I think that would prove a more effective tool that regular antitrust law.
Timothy Simcoe says:

Some thoughts on the discussion questions:

(Re 1) It’s hard to argue that more clarity from SSOs would hurt. And the goal would certainly be to align expectations. But given the inherent conflict of interest between implementers and IPR holders, we should not expect protests that “the other side didn’t live up to expectations” to disappear. After all, it’s just cheap talk.

(Re 2) I’d love to hear a lawyer’s opinion on the prospects for using contract rather than competition law….

(Re 4) Patent reform is obvious place to look for alternative policy tools (see arebentisch’s reply to my post on ex ante). Two possibilities are speeding up patent review to resolve uncertainty, and creating a “registry” to keep track of non-assertion covenants, licensing terms or other restrictions that arise from the standards process
arebentisch says:

ad 4: An “Integrated approach” would be perfect. Competition law is for market conduct. But for the ESOs and SSOs the EU may set simple supreme rules and disclosure requirements. The inherited problems emerge from the fact that even the official SSOs are kind of NGOs which cannot exercise statutory regulatory powers themselves. The EU legislature enjoys these powers.I would argue the golden path in EU-Europe would be to get the EPC implemented as an EU directive, so substantive patent law becomes part of the acquis communautaire, otherwise the EU could also harmonise national patent law to achieve the same. Then a fully integrated EU order policy is possible, which sets ESOs rules, competition and patent rules.

IPR Licensing and Antitrust – The Transatlantic Divide

recent DISCUSSION

Comment on this post

About the Author

A Layne

Talkstandards.com

An active online community where developers, researchers, policymakers and other interested parties can share ideas and collaborate on the global standards system.