Thursday, October 17, 2013
Posted by Simon Steel (Harkins Cunningham LLP)
RAND Obligations Outside the SSO Setting: Some Perspectives from History and Analogy
In the standards-rich, networked economy of the twenty-first century, there is a natural tendency to think of commitments to standard-setting organizations (SSOs) as the main basis, and the model, for obligations to license patents on reasonable and non-discriminatory (RAND or FRAND) terms. But it was not always so. A look back at the history of RAND licensing obligations in twentieth century patent/antitrust cases offers a broader perspective, and suggests responses to some objections to enforcement of RAND obligations outside the SSO context.
However, while the old cases legitimize RAND as a workable antitrust remedy, they are less helpful when it comes to when and how strongly a voluntary commitment should be enforced. Other contributors to these debates, including in this symposium, have tackled these issues at the level of antitrust, IP, patent misuse and contract doctrine. In this piece, I will skip the doctrine and offer an analogy from the world of real property to suggest what an ideal resolution (which would likely require legislation) might look like.
Twentieth Century Patent RAND
RAND patent licensing obligations are not a new innovation by SSOs; they are a remedy that has been used for various purposes in antitrust cases fairly regularly for over 60 years. In 6 separate cases in the 1950s, federal courts upheld antitrust consent orders imposing “reasonable and non-discriminatory” licensing obligations on patentees alleged by DOJ to have divided or otherwise controlled product markets by means of restrictive provisions in patent licensing or pooling agreements. RAND (and other) compulsory licensing obligations were also imposed over objections in a fully litigated case, the GE lamps case, as a means, less restrictive than divestiture, to restore competition to markets harmed by patent-based tying, and in a private action, the glass tempering case, . The substantive antitrust liability theories at play in some of these cases might seem questionable today in light of the modern contraction of per se liability and modern understandings of the procompetitive benefits of patent pools and vertical restraints, but RAND obligations were clearly a common, established remedy for patent licensing practices viewed as anticompetitive. As the Supreme Court confirmed in 1974, RAND licensing requirements are “well-established forms of relief when necessary to an effective remedy, particularly where patents have provided the leverage for or have contributed to the antitrust violation alleged.” None of the early cases concerned the enforcement of a prior RAND licensing commitment, whether to a SSO or to someone else. But the lack of such a commitment did not stop DOJ, the courts and the consenting defendants from adopting RAND as an antitrust remedy. Moreover, vague as the phrase “reasonable and non-discriminatory” may seem in the abstract, it seems in practice to have provided sufficient clarity to guide negotiations, since there is no reported subsequent adjudication on the meaning or application of that phrase in any of the 1950s cases. And DOJ, with the courts’ approval, constructed a more or less standard apparatus to deal with any disputes that might arise: (1) if the “reasonable” level of royalty was disputed, the patentee could apply to the court that imposed the RAND obligation for a “reasonable” royalty determination, as to which the patentee would have the burden of proof in a proceeding in which the potential licensee and DOJ would also be heard; and (2) having requested a license, the potential licensee could proceed pending royalty determination as if licensed, with the court having discretion to impose interim royalties and to require retroactive payment of royalties once determined.
DOJ and the courts got more specific about the meaning of RAND in the following decades. For example, in a 1969 case, DOJ brought a section 1 case against 3M for abusive patent infringement litigation coupled with territorial and other restrictions on licensees, resulting in a RAND licensing consent order that specified that licenses must be non-exclusive and “non-discriminatory as among licensees procuring the same rights under the same patents,” and with a detailed procedure that re-appeared in subsequent DOJ consents: a 30-day deadline for the patentee to respond to license requests with proposed royalties; 120 days for the parties to negotiate royalties; freedom thereafter for either party to apply to the court for royalty determination, with the burden being on the patentee; a right for the applicant to practice the patent in the interim, subject to interim and retroactive royalties set by the court; and any judgment as to reasonable royalty in one case being binding in future cases on the same patent. The 1975 Manufacturers Aircraft consent added a further wrinkle: the RAND licensing obligation imposed there was subject to a condition that any applicant for a RAND license from the defendants which itself had a patent in the same area must make a reciprocal offer of RAND licensing. And on remand in the Glaxo case in 1974, the district court added to the now-standard RAND provisions a clause prohibiting the defendant from transferring the patents at issue other than with a binding promise from the transferee to be bound by the RAND commitment.
Meanwhile, RAND also acquired legitimacy and substance in the courts as a substantive basis for finding patent misuse. In the 1959 ASC glass tempering case, the Third Circuit held that compulsory package licensing – refusing to license one patent on its own, rather than as part of a packaged patent portfolio license – both violated a prior court order imposing RAND obligations and constituted patent misuse. In the 1969 Uniroyal case, the Southern District of New York relied on the RAND concept in holding that a patentee with a patent monopoly on the use of an unpatented type of weedkiller was guilty of patent misuse when it used that patent to monopolize the product market by refusing to license other manufacturers, or users who bought the product from other manufacturers, under the use patent, other than at royalty rates so high relative to Uniroyal’s product price that product competition was foreclosed.
Where does this twentieth century history take us in the end? Individual cases do not take us very far, particularly insofar as they rest on per se liability theories and broad notions of patent misuse that may not survive modern precedents such as Princo. But collectively, twentieth century history does establish that RAND orders have long been a recognized antitrust remedy outside any SSO context; that courts can construct simple and consistent rules and procedures for judicial resolution of any disputes about the meaning and implementation of RAND; that courts can and have specified what their RAND orders require with respect to such issues as preventing evasion by transfer of patents, barring compulsory package licensing, and barring royalty rates that foreclose competition; and that, in practice, judicial guidance on RAND outside the SSO context has typically been sufficient to avert the need for the parties to return to court.
A RAND Commitment Registry – An Idea Derived from Real Property Analogies The twentieth century patent-RAND history provides support for RAND enforcement against several potential objections. No, RAND is not a special SSO thing. Yes, it is an established antitrust remedy. Yes, courts can enforce and specify it. And no, it is not so indefinite as to guarantee endless insoluble litigation about its meaning.
The problems that remain concern RAND commitments: whether, to what extent, for and against whom, and how they should be enforced. Under section 1 of the Sherman Act, bait and switch tactics to exploit lock-in, such as promising RAND licensing to lure market participants into adopting your standard, and then reneging, should qualify as exclusionary conduct that will lead to liability if the requisite monopoly or danger of monopoly is shown; and section 5 of the FTC Act and unfair competition law may offer enforcement prospects when monopolization cannot be shown. And contract and promissory estoppel doctrine can also provide a basis for enforcement of RAND commitments, especially (but not only) when there is an obvious relationship between the parties, as between members of an SSO.
However, while there are strong grounds for antitrust or contractual enforcement in most if not all cases of a dishonored RAND commitment, such enforcement can be complex and does not seem to capture the basic reality. Assuming that it is a real commitment, and not just puffery, a firm that makes a RAND commitment is making a specific promise to the world at large about how it will use a specific item or items of property. And, whether it does so in order to enjoy the benefits of membership of an SSO, or in order to persuade an SSO to adopt its technology as a standard, or in order to persuade the market, outside an SSO, to adopt its technology as a standard or platform, or in order to persuade antitrust enforcers to permit a merger or close an investigation, or in order to win goodwill from developers, open software advocates or politicians, it is generally doing so to gain a substantial commercial advantage based on public trust that it will honor its commitment.
In assessing whether and how such a commitment should be enforced, complex after the fact inquiry into market power, contractual relationships or individual reliance makes little policy sense. When a legal proceeding is about damages, causal and relationship issues are rightly central. But a RAND dispute is generally about specific performance. (An “efficient breach” theory of why damages should substitute for injunctive relief seems implausible where the injunctive relief takes the form of compelling a reasonable exchange.) If a commitment is serious and clear enough to merit enforcement, and it was made with an expectation of significant commercial benefit, it makes sense to enforce it.
This suggests that what would be desirable, if we could escape doctrinal tangles, is a regimen that determines clearly ex ante what the content and scope of a given RAND commitment is, that distinguishes clearly between serious commitments and puffery, without relying on arbitrary distinctions between SSO and non-SSO commitments, that is simple and predictable for both the committing party and the potential licensee as to how the commitment will be interpreted and enforced, that allows parties to frame their RAND commitments, and SSOs to frame their rules as to RAND commitments, as they wish, so long as they are clear, that makes RAND commitments publicly available and clear, and that offers a simple, central enforcement mechanism. What would such a system look like? My suggestion is that it would not be a system that relies on varying state law rules of contract and promissory estoppel, on antitrust enforcement discretion, or on complex, fact-intensive private antitrust trials. It would not create opportunities for forum-shopping, and it would not place the burden on the public, or on excellent public interest efforts such as the American University effort described in Jorge Contreras’s comment, to find, catalog and track RAND commitments. Nor would it be a system that creates opportunities for evasion by separating the data on the commitment from the data on the patent, which the committer may have transferred to a third party.
Rather, my suggestion is that the ideal system for dealing with patent RAND commitments going forward would be one administered centrally by the PTO and modeled loosely on real property title registries. RAND commitments to the general public are akin to public rights-of-way or other easements over real property: they are encumbrances, in favor of the general public, on the property. And, just as the scope and validity of an easement on real property is efficiently and publicly settled by appending it, delineated however the property owner consents it to be delineated, to the record of title for the underlying realty at the local property registry, an official, conclusive (apart from fraudulent registration) documentation of a RAND commitment could, with appropriate legal authorization, be appended to the record of the patents it encumbers at the PTO. Moreover, that federal registration could be made the basis for a federal cause of action to enforce RAND commitments which could be channeled into appeals to the Federal Circuit, thus minimizing forum-shopping and facilitating the development of coherent and predictable doctrine interpreting RAND commitments.
There are many possible objections to this suggestion – not least that it would likely require legislation and extra funding for the PTO, which could be politically difficult. But the expense of centralized title registration has long been proven worthwhile for real property, and it is at least as important to provide clarity with respect to rights to patents – particularly patents important enough to be subject to public RAND commitments. There would, of course, be transitional problems with respect to patents already subject to RAND commitments, and the formality and transaction costs (including costs of determining precisely which in a large portfolio of patents are included in a general RAND commitment) could deter some RAND committers from registering. Ultimately, however, honest RAND committers will commit if the benefit of committing outweighs the cost, and strengthening, publicizing and clarifying the commitment by means of a PTO record should generally enhance any net positive value of a RAND commitment. And the market could learn to devalue – and SSOs and antitrust enforcers could amend their rules and practices to deem insufficient – unregistered RAND commitments.
This suggestion may or may not ultimately feasible, and the devil would be in the details if it is. But in an area strewn with complex overlapping IP, antitrust and contract doctrine, stepping back to look at broader analogies, and trying to imagine what an ideal regime would look like, seems like a worthwhile project.