Standardised uncertainty

Standardised technology is everywhere, from Blu-Ray to cellphones, from light fixtures to wireless internet access. Because of industry standards, a consumer does not need to know whether his Blu-Ray player was manufactured by Sony, Samsung or LG in order to watch a film. Industry standards likewise allow consumers to use laptops manufactured by Lenovo to wirelessly access the Internet through routers manufactured by Cisco. This convenience has been a major boon to consumers.

Widespread industry standards provide several additional benefits. Standards lower costs by increasing product manufacturing volume, which benefits consumers and industry participants alike. They increase price competition by eliminating switching costs for consumers who wish to switch from products manufactured by one firm to those manufactured by another (eg, switching from a Sony Blu-Ray player to a Samsung one), and by reducing barriers to entry for new competitors. Finally, standards reduce the risk to industry participants and consumers who adopt a technology at its outset of investing scarce resources in a proprietary technology that may not gain widespread adoption. Standardised technologies are more likely to gain broad acceptance, which is typically perpetuated over time because yesterday’s standardised products usually interoperate with today’s.

However, one complication with standards is that it may be necessary to use patented technology in order to practise them. This has become an increasingly common issue in recent years, but one on which courts and licensing professionals have only recently started to focus. Both the licensors of patents necessary to practise a standard – called standards-essential patents – and those that need to license those patents worry that standards-essential patents create opportunities for abuse.

According to at least some licensees, the same interoperability benefits that make standards beneficial to consumers also create opportunities for companies that hold standards-essential patents to exercise disproportionate market power. The theory goes that once a standard becomes widely adopted, companies cannot effectively compete in the industry without implementing it; consumers do not want to purchase the one product that is incompatible with the others on the market. The significant investment required to develop products and technologies that conform to the standard make it prohibitively expensive for an entire industry to switch to a different standard (not to mention the lack of backwards compatibility that such a move would entail). Licensees accordingly worry that the holder of a standards-essential patent could potentially hold the entire industry hostage by extracting supra-competitive royalties based on the fact that technology was standardised and on the prohibitive cost of switching rather than the actual value of the patent at issue. The actual value of that patent may be extremely small despite its inclusion in a standard. Indeed, much of the value comes from the fact that the patent is standards-essential, rather than from its technological contribution. As many licensees like to point out, patented technology is not necessarily included in the standard because it is the best technological solution. Often, several equally viable alternatives exist and the particular technology is incorporated into the standard simply because a choice has to be made.

Standards-setting organisations, such as the Institute of Electrical and Electronics Engineers (IEEE), have issued rules designed to address this hold-up issue. Among other measures, standards-setting organisations ask potential holders of standards-essential patents to commit contractually to license their patents on fair, reasonable and non-discriminatory (FRAND) terms to others interested in implementing the standard. The intuition is that this commitment minimises hold-ups by establishing ground rules for licensing standards-essential patents before the industry is locked into a standard.

However, some licensors of standard-essential patents worry that the FRAND commitment and attempts so far to enforce that commitment in litigation have given licensees disproportionate power. Courts have increasingly accepted the argument that injunctive relief for patent infringement is incompatible with the FRAND commitment, for a variety of reasons. In Apple Inc v Motorola, Inc, for example, the Federal Circuit recently held that Motorola’s commitment to license its standards-essential patents to anyone intending to implement the standard foreclosed injunctive relief. Although the decision refused to adopt a per se rule that injunctions are never appropriate in FRAND cases, it strongly suggests that injunctions will be rare. That is problematic, according to licensors, because the absence of injunctive relief in conjunction with the expense of patent litigation can lead to reverse hold-up, whereby a licensor is supposedly forced to license at less than fair market value. Under this view, a licensee supposedly risks little by delaying good-faith negotiations and insisting on a below-market royalty because (setting aside litigation costs and reputational harms) the worst outcome in litigation will be an order to pay a market-rate royalty similar to that paid by all others practising the standard.

Whatever the prevalence of hold-up and so-called reverse hold-up generally, the recent emergence of non-practising entities (NPEs) which acquire and seek to assert standards-essential patents points to one area where hold-up is a real concern. Those entities have no interest in practising the standard and consequently no concern that their unreasonable licensing demands may lead others to license standards-essential patents on similarly unreasonable terms. NPEs simply want to maximise their royalty payments. How and even whether to deal with the unique challenges that NPEs create is an issue that standards-setting organisations and courts have only recently started to confront.

Regardless of whether hold-up or reverse hold-up is a significant problem, parties attempting to negotiate a licence to a standards-essential patent, even in good faith, often do not see eye to eye. ‘Fair’ and ‘reasonable’ are subject to interpretation. Even the ‘non-discriminatory’ requirement has been loosely interpreted by some. Is a standards-essential patent owner violating that contractual commitment by licensing its patents to any end-product manufacturers that wants to include standards-practising components in its products, but refusing to license to the component manufacturers themselves? The answer to that may seem obvious, but some licensors have different opinions.

The increasing number of potentially standards-essential patents for recent standards further complicates the negotiating process. What constitutes a fair and reasonable royalty to practise a standards-essential patent depends, among other things, on how many other standards-essential patents exist for the same standard. Some standards-setting organisations do not catalogue how many patents each of their standards incorporate, instead accepting blanket assertions from standard-setting participants that whatever standards-essential patents they may hold will be licensed on FRAND terms. Moreover, some of those standards-essential patents may not issue until years after the standard is adopted; patents do not issue instantly. All of this can make it difficult for the parties to a licensing negotiation to know whether the patent under discussion is one of 100 standards-essential patents for a given standard or one of 1,000. FRAND negotiations, in short, can sometimes involve significant uncertainty.

Recent court decisions have highlighted the risks that this uncertainty creates. A licensor of a standards-essential patent whose initial licensing offer is not made in good faith has breached its contractual FRAND commitment, according to a 2012 decision out of the Western District of Washington in Microsoft Corp v Motorola, Inc. So too the licensor that makes an initial good-faith offer, but whose final offer is not FRAND. The jury in Microsoft v Motorola awarded Microsoft over $14 million in damages for Motorola’s breach of its FRAND commitment, even though the district court needed a bench trial to determine a reasonable FRAND royalty range. A recent jury verdict out of the Northern District of Illinois in Fujitsu Ltd v Tellabs, Inc found that Fujitsu had breached its FRAND commitment in similar circumstances. That court is currently considering whether the breach renders Fujitsu’s standards-essential patent unenforceable against Tellabs. There may also be significant risk for a licensee that rejects a FRAND offer. As previously noted, an injunction is not normally available if a party is found to have infringed a standards-essential patent. However, the Federal Circuit decision in Apple Inc v Motorola, Inc implies that an injunction may be available if the licensee rejected a FRAND-compliant licensing offer.

Despite highlighting the stakes, recent decisions have only started to offer licensors and licensees guidance on how to evaluate whether an offer is in good faith or FRAND compliant. In re Innovatio IP Ventures, LLC (out of the Northern District of Illinois) and Microsoft v Motorola modified the Georgia-Pacific factors which guide the hypothetical licensing negotiation that is frequently used to calculate patent damages to account for hold-up, royalty stacking and other unique FRAND considerations. Presumably, that modified framework will provide some guidance to licensing professionals engaged in actual licensing negotiations. Some worry that Innovatio and Microsoft v Motorola may tilt the negotiation too heavily in favour of licensees and undervalue standards-essential patents. Regardless, not all courts agree with the approach in those cases. An Eastern District of Texas decision in Ericsson Inc v D-Link Systems, Inc, for example, held that the FRAND commitment did not require modification of the Georgia-Pacific factors at all. Future decisions – or future action by standards-setting organisations – will have to further flesh out the proper analysis.

Because FRAND licensing is such an important and rapidly developing area, this issue’s roundtable offers the perspectives of four thoughtful experts on the difficult issues that FRAND licensing raises. Their backgrounds are diverse – a former Federal Circuit judge, a senior in-house licensing policy maker, a leading economics expert and a former chair of a standards-setting committee – and their experience is extensive. Their thought-provoking perspectives are sure to provide all of us with valuable insights into how to approach FRAND licensing and litigation issues.

What effect does the FRAND commitment have on the willingness of patent holders to participate in the standards-setting process?

Greg Leonard, economist and partner, Edgeworth Economics

“It remains to be seen whether companies that recently divested their device businesses while retaining their licensing/R&D businesses drop out of standards-setting organisations”

Greg Leonard (GL): FRAND commitments have been required by a number of standards-setting organisations for a long time and participation rates by major industry players has always been high. I do not foresee that changing for any industry player which is both an implementer and a patent owner, because the benefits of standardisation to such players are so large. However, it remains to be seen whether companies that recently divested their device businesses while retaining their licensing/R&D businesses drop out of standards-setting organisations. Over time, the adverse effect of such a trend on standards setting may be limited because, without a device business as a driver, these companies may no longer generate as many innovations and patents as they did previously and thus their continued participation in standards-setting may not be as important going forward.

Gil Ohana, senior director, antitrust and competition, Cisco Systems

“Any implementer of a standard that is willing to take a licence on terms that are independently determined to be FRAND must therefore be free of the threat of an injunction”

Gil Ohana (GO): Many companies that drive innovation and new product development choose to participate in standards-setting organisations in order to promote interoperability through industry standards to create a broad market opportunity for the entire industry. Standards support interoperability and in turn help to drive growth opportunities for entire industries, benefiting the companies that supply the standardised products and the customers who use them. For standardisation to provide these benefits, participants often agree to license patents on FRAND terms in exchange for being able to participate in the development of standards that will create growth opportunities for their companies.

Despite FRAND licensing requirements, numerous industry participants – including companies that monetise their patent portfolios by licensing patents to implementers of standards – continue to see the benefit of participation. The existence of FRAND licensing requirements has not discouraged participation in any noticeable way. I suspect that this is because owning patents that are essential to standards can create valuable licensing opportunities which exceed what would be available to the same patentees if their patents were not essential to implement standards. In return for what may be far broader opportunities to license, patentees give up some rights that they would otherwise enjoy, including the right to enjoin or exclude accused infringers.

Randall Rader, former chief judge of the Court of Appeals for the Federal Circuit

“While I have seen no data indicating a reluctance to participate in standards setting, I would expect important technology developers to withdraw if the process frustrates the market for the recovery of technology development costs”

Randall Rader (RR): While I have seen no data indicating a reluctance to participate in standards setting, I would expect important technology developers to withdraw if the process frustrates the market for the recovery of technology development costs. For instance, I was aware a few years ago of a country that wished to use standards setting to acquire technology below its market value. This policy would simply frustrate technology transfer and availability.

Do you believe that patent holders are actually extracting hold-up value when negotiating licences with companies that practise the standard?

GO: Patentees that own patents that are essential to implement widely adopted industry standards can, if permitted to enjoin or exclude accused infringers or license on non-FRAND terms, obtain royalties that are wholly unrelated to the benefit that their patented invention contributes to the standard. Instead, they can demand royalties based on the expense that the accused infringer would incur in withdrawing its product from the market if the sale of that product were enjoined after the product had been introduced to the market. By appropriating what economists refer to as ‘switching costs’, patentees can derive a windfall that far exceeds what their invention would be worth without the injunction threat. In that situation, a patentee is seeking hold-up value.

There is growing awareness among judges, competition enforcement agencies and standards-development organisations that seeking hold-up value is anti-competitive and discourages innovation by shifting royalty income towards owners of patents that are essential to implement standards, but which do not describe important innovations. Because of that awareness, the ability of patentees to extract hold-up value is diminishing. This concerns some owners of standards-essential patents that benefited from the availability of injunctions and exclusion orders.

Assuming that patent hold-up is a real problem, how should standards participants and the courts use the FRAND commitment to address it?

GO: First, they should understand the FRAND commitment as a commitment to license on reasonable terms. A commitment to license is a commitment not to enjoin. So any implementer of a standard that is willing to take a licence on terms that are independently determined to be FRAND must therefore be free of the threat of an injunction.

Second, they should continue their efforts to provide additional guidance as to what FRAND means. Unfortunately, some patentees – in particular, patent assertion entities that acquire patents which are subject to FRAND licensing commitments – take extreme positions regarding what licensing terms are consistent with FRAND, while nevertheless claiming that the terms they are offering are fair and reasonable. Courts and standards-development organisations (as well as competition enforcement agencies) have an important role to play in identifying additional principles to help define what the FRAND commitment means. It has been gratifying to see courts beginning to play that role, notably in Judge Robart’s decision in Microsoft v Motorola and Judge Holderman’s decision in Innovatio. Standards-development organisations also have a critical role to play in identifying the principles that their members wish to be relevant when it comes to determining FRAND licensing terms. Policies that clarify what terms are FRAND would allow courts to apply the views of participants in standards development rather than identifying their own methodology for resolving licensing disputes as they try to translate FRAND to the facts of a particular dispute.

Matthew Shoemake, serial entrepreneur

“Royalty stacking is a material problem due to the complex nature of modern standards, which are based on multiple patents”

Matthew Shoemake (MS): Standards participants can make licensing terms clear before the standard is ratified. This gives members of the standards body the ability to exclude technology that has an excessive, non-FRAND royalty.

However, in general, standards bodies are not eager to be pulled into litigation. Thus, it falls to the courts to interpret FRAND policies. In doing so, they must recognise that there is a third-party beneficiary situation (ie, FRAND commitments are submitted to the standards-setting organisation) for the benefit of all. Given that a standard could adopt multiple technologies, the situation is highly competetive. There are entities willing to make their technology available to standards at no cost, which means that for standards such as WiFi, the patent costs should be zero or near zero. Relatedly, courts should realise that there is a winner-takes-all element in standards development. In addition, there is an incentive to contribute intellectual property to standards-setting organisations even if it does not result in royalties (eg, prestige or product sales). Courts must recognise the value of the standards-setting process itself – for which patent holders should not be compensated – and the patent landscape that pertains to a standard, as there may be hundreds if not thousands of patents that apply to a standard. Finally, courts should try to understand the nature of the marketplace that the standard is targeting (eg, is pricing in the market priced by manufacturing cost or IP cost?).

What is your view on whether royalty stacking is a problem for standards participants?

GO: Hundreds or thousands of patents may be required to implement common interoperability standards that we all use every day. It would be great if the owners of those patents could pool all essential patents for a particular standard and make them available for a single price. However, the most aggressive licensors prefer not to join pools. In the absence of comprehensive pools, implementers are left with the likelihood of serial licensing negotiations with multiple licensors, each of which will argue that the patent it owns is uniquely valuable.

Beyond the significant transaction costs that this can result in for licensees, this creates what I call the ‘Lake Wobegon’ problem. As listeners to Prairie Home Companion know, Lake Wobegon is the mythical Minnesota village where the women are strong, the men are good looking and all the children are above average. Even a lawyer can understand that, even in Lake Wobegon, all the children cannot be above average. Likewise, every standards-essential patent cannot be a uniquely valuable standards-essential patent.

Asking courts to take stacking into account is, in effect, asking them to presume as a starting point that each standards-essential patent is of average value to the standard. The patentee should be free to argue that its patent is uniquely important and the accused infringer should be free to argue the opposite. However, if neither sustains its burden, then all that the patentee gets is a proportional share of the entire value that the standard contributes to the product in which it is implemented. Following the Federal Circuit’s lead in LaserDynamics and subsequent cases, this means the smallest saleable unit that implements the patented invention, with further apportionment to address the fact that the standard itself may contribute only a portion of the overall value of the relevant component. Indeed, a complete analysis of apportionment must also consider the value attributable to the manufacturer’s contributions (eg, a reputation for quality and reliability, brand name, product design, packaging, advertising, marketing, sales and customer support). The patentee has no claim to damages based on the value contributed by any of these factors.

MS: Royalty stacking is a material problem due to the complex nature of modern standards, which are based on multiple patents. Thus, the number of patents required to practise the product should be taken into account when determining a royalty rate. Further, successful standards tend to result in increased patent filings in the area, which exacerbates patent stacking.

Given that the extent of the royalty stack for a given standard is often unknown, how can parties to a FRAND negotiation and the courts effectively deal with that uncertainty?

GL: One approach is to determine the overall royalty burden that was anticipated by the members of the standards-setting organisation or a reasonable expectation as to what the overall royalty burden would have been when the standard was set. Then, divide this overall royalty burden among the standards-essential patents based on their relative contributions. This ensures that each standards-essential patent receives its appropriate royalty while at the same time avoiding an excessive royalty burden that might have threatened the standard’s adoption. This approach also avoids the problem of one standards-essential patent owner attempting to claim for itself value that was actually created by the standards-essential patents owned by others.

RR: The parties are in the best position to address this problem because they generally know the source and value of all the technology contributing to the standard. The courts are at a distinct disadvantage because they have only one patent – somewhat mislabelled as ‘the’ patent essential for a standard – without full awareness of the extent of the entire standard process or the rest of the market. Thus, the courts are at a disadvantage in their ability to set the accurate value of this single, perhaps small aspect of the entire product market.

Significant liability and remedies issues in FRAND cases turn on whether a FRAND licensing offer was made and whether that offer was rejected. How can parties determine whether a licensing offer is FRAND?

GO: Hopefully, the analysis in Microsoft v Motorola and Innovatio will be adopted by other courts. That will help the parties to a FRAND negotiation assess the reasonableness of the licensing terms that they are offering or being asked to accept in a more granular way than the simple term ‘fair and reasonable’ currently permits. I agree with the recent observation of Federal Trade Commission Chair Edith Ramirez that: “Greater clarity on the terms of a FRAND license is likely to facilitate private negotiations and limit the need to seek a third-party determination of a FRAND rate.” However, if private negotiations fail, courts must be free to adjudicate a FRAND rate, either with or without consideration of related patent-specific issues such as infringement and validity.

Do you think courts should view damages differently in cases involving FRAND commitments and standards-essential patents?

GL: I have no problem with courts’ attempts to modify the Georgia-Pacific factors to tailor them to the context of a FRAND royalty analysis. However, at the same time, the Federal Circuit has moved reasonable royalty damages analysis in non-standards-essential patent cases toward the idea that the royalty should be limited to the invention’s footprint in the marketplace, which means that the royalty should reflect only the value that the invention creates and should not include value created by other technologies, patented or otherwise, or value from contributions from other sources. This is similar in spirit to the FRAND concept that a FRAND royalty should not reflect the value created by the standard itself. In a sense, I think we are seeing a convergence conceptually between what is being done in FRAND royalty cases and what is being done in non-standards-essential patent cases.

RR: The ultimate challenge is to determine the proper value of the technology in the marketplace. If the manufacturer can show that the standard will create distribution that the technology would not have enjoyed otherwise, then it might be able to argue for a slight discount in the market rate because the standard has contributed to an enhanced market value. However, this would come only after setting the market value, independent of the standards-setting itself.

Should standards-setting organisations provide more guidance on what the FRAND commitment entails?

MS: In my experience with 802.11, significant guidance and warning are already provided. There are operating procedures in place that regularly inform attendees of their obligations and duties with respect to FRAND policies.

That said, if rights holders abuse the policies and seek hold-up value, the standards-setting organisation can take further action and require a so-called ‘FRAND-Z’ assurance. Under this, the rights holder asserts that it will not only be fair and reasonable, but will also license at a zero rate (eg, the Bluetooth SIG). Such standards-setting organisations can be highly successful because there are always multiple technologies available and there seems to be a willingness to contribute intellectual property at zero royalty. Companies that sell products may be and have been willing to license intellectual property for free to enable the standard and the marketplace it helps to create.

If standards-setting organisations and participants cannot agree on a meaningful definition of what the FRAND commitment entails, what are the implications for future standards-setting efforts?

GL: Courts have started clarifying the meaning of these commitments. It is possible that some entities, particularly patent owners without product businesses, may stop participating in standards-setting organisations in response to these events. However, in general, standardisation is too valuable to patent owners with product businesses to be abandoned by them.

Non-participant patent owners may seek hold-up value in licensing negotiations with implementers of a standard and, in principle, this could lead to excessive royalty burdens that have an adverse impact on adoption of the standard. To address this potential problem, standards-setting organisations may try to identify and avoid technologies covered by patents owned by non-participants. Otherwise, if such a technology is incorporated into the standard, the non-participant patent holder may be able to extract hold-up value. However, non-participants’ ability to extract hold-up value may be constrained by recent US court decisions, which limit reasonable royalties in patent infringement litigation to the invention’s footprint in the marketplace (ie, the value it creates). In my view, this limitation should be interpreted as excluding hold-up value. Then, non-participants may be able to recover only royalties similar to what they would have got had they participated in the standards-setting organisation and made a FRAND commitment.

GO: We are already seeing some of the implications. One is the increasing popularity of standards-development efforts with a limited technical scope which are organised around default or mandatory royalty-free IP rights policies (eg, Bluetooth, Universal Serial Bus, HTML and DOCSIS and Packet Cable – the standards that cable television operators use to provide broadband). I think that this trend will continue, particularly if some of the formal standards-development organisations that have been most resistant to reform persist in letting a small but vocal minority of their membership delay meaningful and necessary clarifications to IP rights policies.

Do you think standards setting against the backdrop of FRAND on balance promotes innovation, or does the FRAND commitment discourage innovation by allowing others in the industry to free-ride off the R&D of others?

GO: Standards setting certainly promotes innovation. It is an open question in my mind whether FRAND licensing promotes innovation. Many widely adopted standards have been created under royalty-free licensing models – I mentioned some examples earlier. The success of the Ethernet local area networking standard, which we at Cisco know well, was in part the result of an early agreement between Intel, Digital Equipment and Xerox to license essential patents for an all-you-can-eat price of $1,000.

The success of standards created under royalty-free licensing models or for which patent monetisation was a trivial consideration should remind us that many companies without outbound licensing models regularly participate in standards development and contribute valuable technologies for use in standards. They participate and contribute because they benefit from the commercial opportunities that standards create to sell more products and services, not because they want to license standards-essential patents. This should suggest that the connection between innovation through standards development and FRAND licensing is uncertain. What has become all too certain is the connection between FRAND licensing for standards such as LTE and WiFi and expensive, time-consuming and disruptive patent disputes.

MS: Standards setting in a FRAND context certainly promotes innovation. It is intended to create an ecosystem that allows manufacturers and consumers of products to rest assured that they can obtain access to all intellectual property needed to practise the standard. Without FRAND policies, manufacturers would be reluctant to burden their patent with a FRAND commitment, because they would not be assured that they could gain access to the other patents needed to build the product that they intend to sell. Such a situation would be stifling to the market for standards-based products, because manufacturers would likely make the business decision to build proprietary products rather than interoperable, standards-based products.

Further, in my experience, industry participants do not free-ride off the R&D of others. The truth is that the industry comes together in standards-setting organisations to choose from many different approaches, which may be used to solve a problem and work together towards a world with interoperable, fairly priced and technically sound products for consumers.

That said, standards-based markets such as WiFi, Bluetooth and LTE enable cost-based rather than IP-based pricing. Contributors of intellectual property to standards-setting organisation must realise that they are likely contributing to a standard that enables a cost-based market. If those contributors (or parties to which they transfer their intellectual property) try to shift the economics of the technology segment to an IP-based pricing model, they will break the standard that they are contributing to and another standard will be needed to take its place. This is not good for consumers or the industry as a whole.

Watch closely

Although standards and standards-essential patents have existed for years, the courts have only recently started to define the rules for licensing and enforcing those patents. The Federal Circuit – with the exception of a few paragraphs regarding injunctions in Apple v Motorola – has yet to weigh in. Many important and fundamental areas remain where parties have only a handful of district court decisions and regulatory statements to guide them. What is a standards-essential patent, for example? What does a good-faith opening offer look like? How does a party evaluate whether a licensing offer is FRAND-compliant? Given the rapidly evolving legal landscape surrounding FRAND, it is an area that we should all watch closely.