Standard Essential Patents (SEPs): Costs And Benefits Of Broad Application

Editor: Please tell us about your background.

Lynde: I am an economist with a doctorate from UC Berkeley with over 30 years’ experience working for the federal government, in academia, and in consulting. I am now a Vice President at Cornerstone Research, where we specialize in economic and financial consulting, including the support of expert testimony. I consult on intellectual property portfolio valuation and have served as an expert witness around the world, including in intellectual property infringement matters and in high-profile cases involving standard essential patents (SEPs). Prior to this, I was a partner at PricewaterhouseCoopers. Before that, I was a professor of economics at the City University of New York. My background in economics is related to policy; my undergraduate background is in electrical engineering at UC Berkeley.

Editor: Please define SEPs. What is the standard-setting organization (SSO) that identifies them?

Lynde: There are actually many standard-setting organizations. Many of these SSOs have subcommittees that set particular standards. These subcommittees require contributing engineers to disclose whether they have patents or patent applications that may be essential to the standard, i.e., you would be infringing the patent if you implemented the standard. Only the courts can determine if a patent is actually essential, and since that is an extensive process, few make it all the way to a determination of legal essentiality.

Editor: Why is it important for in-house counsel and patent managers to understand that this state of the law is in flux and should be followed closely?

Lynde: There has been a rapid evolution of the case law relative to damages in the federal courts, the ITC, and internationally. Supreme Court decisions will have dramatic impacts on patent values and business strategies focused on protecting intellectual property.

Editor: What industries are most commonly associated with SEPs? What industries may adopt them in the future?

Lynde: These are industries where technical interoperability is key for commercial success – telecommunications, computing-related industries, and others. Such products are useful to the public because they must agree with each other, which is made possible by compliance with the same protocol. Economists recognized long ago that industries where so-called network effects were important can benefit from some cooperation between suppliers to agree on a common standard. This cooperation is pro-competitive and benefits consumers, such as having one wall socket voltage, or all telephones able to interpret communications codes. The public hugely benefits from this standardization. The Department of Justice (DOJ)/Federal Trade Commission (FTC) IP licensing guidelines set forth such pro-competitive conditions, allowing competitors to contribute complementary assets to a standard.

Editor: Would you explain the rules by which SEPs are licensed?

Lynde: There is no one set of rules, but one policy common to all SSO agreements is that you promise to license your standard-related technology to all comers on FRAND (fair, reasonable, and nondiscriminatory) or RAND (reasonable and nondiscriminatory) terms. The purpose is to prevent “hold-up” – that is, ex post opportunistic behavior extracting more value than the intrinsic value of the patented technology – thereby inappropriately taking advantage of the fact of standardization per se. These terms are, however, never really defined in the typical SSO agreement. It’s easier to agree if the terms are vague. So after the fact we are left trying to determine what “reasonable” means. Parties may differ on this.

One approach economists have taken is to consider comparable patent pool rates. As these, ideally, can indicate approximate multilateral ex ante negotiated rates, they can give objective measures of value without hold-up (ex ante value takes care of that, as only intrinsic value is bid before a technology is adopted into a standard), and accounting for stacking (everyone contributing needs to agree that the total royalty stack makes sense commercially). Another objective approach the courts have taken is apportioning value based on the smallest saleable unit that embodies the invention (Cornell v. HP); while this is susceptible to arguments about downstream complementary value of the technology, it can be useful in appropriate circumstances. Both these approaches have been used in recent landmark RAND cases.

Editor: Has the meaning of the RAND or FRAND commitment been clearly defined?

Lynde: I do not think so, and it may never be, as SSOs can’t be very explicit. Any ex post determinations are just that – after the fact. Antitrust authorities in the U.S. have been talking to many of the world’s SSOs, asking them to be clearer about what they mean by it. As economists, we are clear we do not want these organizations to facilitate any form of anti-competitive collusion or downstream opportunism that would not benefit the public. But as a matter of economics, we only have a theoretical framework of pro-competitive, RAND-compliant rates; in practice, what can be used as a benchmark? Any actual bilateral agreement, whether reached in settlement of litigation or not, may be tainted with hold-up. As an empirical matter, we can’t measure that. Theoretically, bargaining theory suggests that patent pools might provide an appropriate benchmark. But the literature points out that practical pool arrangements have a number of imperfections that make their adoption as a benchmark difficult in certain circumstances, even if those arrangements might apply well in others. Various court decisions are beginning to sketch out what RAND means, e.g., Microsoft v. Motorola, now on appeal; Ericsson v. D-Link; and Innovatio v. Cisco. My testimony on Microsoft was the basis of a RAND ruling based on pool benchmarks; the court in Ericsson adjusted post-jury verdict rates to a similar range; and the court in Innovatio adjusted “smallest saleable unit” metrics to a similar range for the same standard (802.11). Courts are just beginning to rule on this, and appeals are in process.

Editor: Do SEPs have economic value materially different from the economic value of ordinary patents?

Lynde: It's hard to generalize because to have a patent in a standard is itself highly valuable. Many ideas are of little value unless incorporated into a widespread interoperability standard. For patents of average value, the effect on economic value could go either way. On the one hand, the standard might involve a large number of units, so that the aggregate value could be very large even if the per unit value is small. On the other hand, the RAND promise, which attaches to the SEP, is an encumbrance in the sense that a claim for value should not reach the level of hold-up – i.e., claiming part of the value of the standard itself as opposed to the intrinsic value of the patent – at least as far as economic policy is concerned. The extraordinary intrinsic value of some rare patents for which there is truly no substitute at the standard formation stage would remain – but hard to distinguish from hold-up value.

Editor: From an economist’s perspective, would it be useful to impose a fee scale on the amount owing in royalties by the appropriate SSO?

Lynde: Economists are leery of that. We don’t want to risk any collusion in downstream excessive price coordination. Even if the IP license guidelines carve out a safe harbor, it's a delicate issue — except for royalty-free, which all standards allow as de facto RAND compliant. Even though some SSOs allow parties to disclose what non-zero royalties they plan to charge, most parties do not wish to disclose what they consider reasonable, which is not conducive to promoting the proper level of coordination and cooperation required for a RAND-compliant royalty ex post facto. Hence the practical problems we face in RAND licensing.

Editor: Why would a company want to join an SSO if it could ask for large royalties on its own?

Lynde: If a company is not participating in the SSO, it is unable to discuss the technical merits of the standard and whether its technology should be part of it. Most engineers definitely prefer to be part of the process. If you are waiting and hoping that the standard goes in a particular direction, in all likelihood you will be out of luck for receiving royalties. If you do participate and your technology is adopted, even if the rate is modest on a per unit basis, the numbers of units sold may be so huge as to be extremely beneficial. In the case of the 802.11 Wi-Fi standard, for example, the count next year should be in excess of 350 billion units.

Editor: Please discuss recent case law on this issue.

Lynde: For many years, in the area of SEPs we have not had any objective benchmarks to define the meaning of “RAND-compliant reasonable,” and in my opinion a lot of mischief has followed from that ambiguity. But recently, we have had several cases that address the issue forthrightly. In Microsoft v. Motorola, a landmark case, Judge Robart wrote a bench opinion on what a RAND-compliant range of royalties would be for two sets of SEPs. Critical pieces of evidence he considered were comparable patent pool arrangements that I had set out in my testimony. A jury found that Motorola’s initial demands of over $4 billion per year were so much greater than the RAND range that they constituted breaches of contract with regard to the relevant SSOs. A more equitable solution would be a couple of million dollars per year.

In another important recent case, Innovatio v. Cisco et al. (where I also testified), Judge Holderman determined a RAND-compliant rate based on RAND-specific evidence, as opposed to ignoring the RAND obligation as in a regular patent damages determination of “reasonable royalties.” Plaintiff’s counsel in this case sent demand letters to Starbucks and mom-and-pop coffee shops, alike, for infringing their Wi-Fi patents. This, along with other similar cases, has almost every state attorney general, and Congress, up in arms about the issue of NPEs. Major OEMs intervened, and the plaintiff claimed a “reasonable” royalty of over $1 billion for their Wi-Fi SEPs. The judge ruled instead in favor of pennies per unit, RAND royalties consistent with testimony of the defendant’s economists.

Editor: From an economic expert’s point of view, what are the challenges in litigating internationally, since there are many different standards being set in Europe and Asia?

Lynde: Most standards of interest are international. The ITU in Geneva sets many worldwide telecommunications standards. Actually, most standards have international implications. The challenge is twofold. The venue may be geographically in the U.S., Europe, or Asia, but the damages scope is limited to the country of the patents in suit; and all parties always want a universal resolution. Litigation is usually in the U.S., occasionally in Germany and Tokyo District Court. International arbitration is more frequent if the venue is not a U.S. court, but the same question of international impact is relevant for ultimate settlement. I have testified by translated written submission in Tokyo District Court. My French has come in handy in international arbitrations in Paris and Geneva.

Editor: Please tell our readers about your recent conference with the Stanford Institute for Economic Policy Research (SIEPR).

The views expressed in this article are solely those of the interviewee, who is responsible for the content, and do not necessarily represent the views of Cornerstone Research.