There has been a recent flurry of policy and regulatory activity related to standard-essential patents (SEPs). First, Samsung announced it was dropping its requests for injunctions on SEPs in its ongoing patent battles with Apple. Then, curiously, the European Commission filed a “Statement of Objections” (first round of European antitrust charges) against the Korean company for the very practice it had just ceased. On the other side of the Atlantic, the U.S. FTC announced that it had reached a settlement with Google (and its subsidiary Motorola Mobility) preventing the company from seeking injunctive relief on SEPs if the potential licensee was willing to undergo binding arbitration to determine “reasonable” licensing terms for Google’s patents. Then the U.S. Department of Justice and the U.S. Patent and Trademark Office issued their own policy statement counseling against SEP injunctions is many cases.
The crux of the debate centers on how much flexibility SEP holders have to negotiate licensing terms for their patents that are promised to be licensed under “(fair) reasonable and non-discriminatory” terms (FRAND or RAND). Specifically, the debate focuses on this question: when is it appropriate to enjoin infringing products from the market if licensing negotiations break down? Historically, FRAND commitments have been relatively ambiguous, giving those holding SEPs broad (but not unlimited) flexibility to negotiate “reasonable” bilateral deals. Currently, there is a movement afoot to give SEP holders less negotiating flexibility. This will have both positive and negative consequences.
Much hay has been made over SEPs and injunctions lately, but there has been little discussion of the historical context, which is truly necessary to understand this complex issue and the role SEPs have played in the larger smartphone wars.
(For a more in depth discussion on SEPs, see CCIA’s filing to the FTC on the Commission’s Consent Decree with Motorola Mobility.)
Although some commenters and regulators have implied that SEP injunctions are the main cause of the current smartphone patent wars, this simply isn’t the case. Indeed, over the last few decades, there has been an explosion of both innovators declaring patents to standard setting bodies, and implementers investing billions in products that implement these standards. The fact that both sides of the market flourished during a period of time that there was no legal restraint on the ability to get injunctions on SEPs is dramatic evidence that the standard-setting system was not broken.
That isn’t to say the system is perfect. Far from it. Sometimes the incentives to “play nice” break down and hold-up does occur. Sometimes a company who was living by the spirit of their FRAND commitment goes bankrupt and the SEPs get sold off to patent trolls who seek excessive licensing fees. These problems should not be downplayed, but they are problems that have been at the margins of the system.
Ambiguous FRAND commitments have also had positive benefits. Some standards contributors have decided to keep their SEPs for defensive purposes. Often these firms don’t proactively track down companies for licensing revenue and only assert their SEPs against companies that seek licenses from (or sue) them. And given the realities of the technology marketplace, where millions of patents apply to technology devices, other firms use their SEPs to gain “freedom to operate” (aka the ability to bring their products to market) by seeking large portfolio cross-licenses. If companies can force mandatory arbitration on just the SEP components of these licensing proposals, there is less incentive to negotiate these large deals where companies essentially agree not to sue each other.
These types of defensive SEP strategies have two pro-competitive effects. First, they function to keep the licensing burden of particular standards lower than if everyone sought to charge everyone else for their SEPs. Second, they prevent lawsuits. In a world where ‘everyone infringes everyone elses IP’, the preventing lawsuits function is particularly important.
If competition agencies rein in the flexibility of FRAND terms without other action being taken to control the patent flood, this will likely lead to more litigation in the short run, not less, as the rivals of companies with SEP-heavy portfolios are incentivized to push their litigation advantages while their opponents are aggressively rebuilding their defensive portfolios. Furthermore, given the courtroom success that other litigants have achieved with controversial software and design patents, it is reasonable to assume that these patents will become the core of new defensive strategies.
Although gaining more clarity and providing real teeth to the FRAND licensing pledges that accompany many SEPs is probably a good thing if part of a larger patent reform effort, any competitive gains achieved through discrete action will quickly be lost if the root causes of the patent wars are not adequately dealt with.
Standard-essential patents are just that — patents that are essential to specific technical standards. Standards often result when companies collaborate to build or design essential components, specifications or processes used by many technology companies. (However, proprietary standards can result when one company makes a product or technology that becomes the de facto industry standard, like the VHS cassette which was developed by JVC.) The reason that all sorts of different electronic devices and household tools can plug into the same AC electrical wall sockets is that many industry leaders came together to agree on common specifications (at least on a national or regional basis) during the initial phases of the roll out of the electrical grid. In creating a common standard, industry agreed on voltages, number of pins and pin sizes and certain safety specifications. This standardization allowed for robust competition and innovation in a wide range of electrical device markets. Standards are even more important today in computer and electronics markets. As the American National Standards Institute (ANSI) says, “Today, just about everything associated with a computer, especially those connected to the Internet, is based on an industry standard.”
Although some standards, such as many Internet standards — including TCP/IP, HTML and XML — forego patented technology altogether (thus obviating the need to take a license), many other standards include patented technology. Patents included in an industry standard are referred to as the previously mentioned, and now infamous, standard-essential patents.
When patented technology is included in a standard, a potential problem arises. The holder of an SEP would be conferred significant market power if they decided to use it aggressively after everyone had implemented the standard. That is why standard setting organizations (SSOs) usually require companies wishing to have their technology incorporated in a standard to make up front FRAND pledges (or, in some cases, royalty free pledges) to increase industry certainty and speed adoption. Why build your products to a certain standard if it means that holders of SEPs have you over a barrel when you do? Thus, in order to incentivize firms to invest in implementing standards, they must be able to rely upon a promise that the SEP owner will not “hold them up” for a royalty that appropriates all of the profits to be earned from selling the infringing device. In other words, the SEP holder must promise that it will seek fair, reasonable, and non-discriminatory (FRAND) royalties.
In the smartphone and device space, many SEPs pertain to WiFi, 3G and 4G (such as WiMAX and LTE) technology that allows all wireless broadband devices to utilize the networks of wireless broadband providers.
The potential competitive problems associated with SEPs are what give competition authorities a say about how they are used.
A Brief History of FRAND
The preceding discussion thus begs the following question: What does FRAND mean? The short answer is no one knows.
“Reasonable” means that the licensing rates should not be excessive (both individually and in the aggregate of the total “cost” of implementing the standard) and comport with the relative value of the patent or patents contributed to the standard. “Non-discriminatory” requires that licensors treat licensees in a “similar” manner. However, that’s where the clarity ends.
“Reasonable” is also incredibly subjective. A consensus is building that “reasonable” means the value of the technology or patents contributed to the standard before the standard was agreed on — or phrased another way, the difference in value between patents incorporated in the standard and the next best alternative. This is great from a theoretical perspective, but actually getting parties to agree on the relative “value” of an individual patent to a standard is incredibly contentious and virtually impossible to define in most cases. In the smartphone context, where devices are composed of a quarter million patents or more, determining the relative “value” of any individual patent is a crapshoot of epic proportions. Plus, everyone thinks that their contribution is incredibly important. And “non-discriminatory” is not much clearer either. It is well recognized that “non-discriminatory” does not mean treating everyone the same. Licensing terms can vary, for example, based on the creditworthiness of the licensee or how they incorporate the standard into their products. Furthermore, big companies often forego direct cash payments for cross-licenses of the other party’s IP, so determining the precise value of bilateral terms is very difficult.
This ambiguity has been the defining feature of the history of FRAND. In 2002 FTC hearings on the subject, all of the all-star panelists agreed that the term lacked real meaning and largely agreed that attempting to clearly define it was either quixotic, unnecessary or even harmful. Here are some quotes from the panelists:
Carl Shapiro (former Chief Economist for the DOJ) summed up the consensus nicely when he said: “It just seems there’s a lot of running room between different interpretations of fair, reasonable and non-discriminatory when we’ve got complex terms and conditions that are integral to the whole process.”
Paul Vishny (former General Counsel for the TIA): “Do I think the agency should define [FRAND]? I really think not because I think what is reasonable and what is not reasonable will vary in many cases in many ways.”
Despite the inherent ambiguity, at least prior to the current patent wars, FRAND licensing usually worked fine — at least for companies that actually produce viable products — because companies know that they have to sit across the table from their competitors at the next standard meeting and that the other guys had their own SEPs that they could use as weapons. (In game theory, this is called the “repeated game”, where individual parties don’t “cheat” in the short run because they know it will be costly over time.)
In fact, most major problems with SEPs prior to the current thermonuclear patent warfare were largely when they fell into the hands of patent trolls who were not interested in bringing their own products to market (or held by companies that transformed from manufactures to patent trolls when their business models soured). The major SEP cases that involved the FTC (including the non-patent troll Dell case) also involved clearly deceptive behavior (patent ambush and/or submarine patents) or the explicit repudiation of agreed licensing terms after the patent changed hands.** (This behavior is decidedly different than the current wave of SEP suits, in which the companies using SEPs in lawsuits are, by and large, using their SEPs defensively or suing companies that are already suing them. But more on that later.)
When patent owners don’t have to worry about bringing products to market or engaging in future good faith standard setting negotiations (aka patent trolls), they can abuse the market power conferred after a standard has been adopted without fear of future retaliation. Therefore, the “repeated game” scenario that compels “good” behavior mentioned above does not apply.
For example, a troll who got its hands on a WiFi-essential patent could sue everyone producing wireless broadband products (and their customers too) knowing full well that they are not vulnerable to counterattacks because they themselves don’t have products on the market. In the process, the troll could use the threat of an injunction to extract much more than the “true value” of the patent because companies implementing the standard are (1) by definition infringing the SEP, (2) locked into using the standard (and the patent in question) because designing around the patent is not an option short of dropping the whole standard and (3) highly vulnerable to their products being taken off the market via injunctions.
The current state of affairs begs the question: Why are companies like Motorola and Samsung using SEPs more aggressively than they have in the past?
In short, Apple started a thermonuclear patent war and Google purchased Motorola and its SEP-laden patent portfolio to defend Android from attacks by Android’s competitors. Around the same time, Samsung — also a major hardware company that relies on Google’s open source Android operating system to power its mobile devices — started shipping more phones. A lot more phones. Eventually even more phones than Apple! The battle lines were drawn.
SEPs as a Strategic Policy Weapon
Before the smartphone patent litigation epidemic, most of the major patent holding companies in the tech space recognized the importance of FRAND commitments, but rejected government interference in curtailing how they were used, or in setting ex-ante rates. Even though ambiguous FRAND commitments were not ideal, most companies argued that these were complex transactions and standard processes that — for the most part — worked just fine. Plus, from the standpoint of corporate patent lawyers, they did not want the government to impose conditions that would impede their strategic use of patents in the future. (All things being equal, lawyers usually like more — not less — flexibility.)
A view widely held in IP departments of major tech companies — articulated by Microsoft in a June 2011 filing to the FTC for a workshop it held on patents and standards — was that the government should not mess with SEPs. It said:
Concerns about “patent hold-up” should not extend to any bilateral business disagreement between two companies regarding proposed licensing terms. These discussions typically pertain to a broader set of questions than just the proposed licensing terms for essential patent claims reading on a standard. In addition, if the Government were to attempt to quasi-regulate RAND licensing terms, then they arguably should review the interplay among all of the substantive terms (and not just the monetary component) for all aspects of patent licensing terms. Yet that would likely be unworkable.
(Also of note, Google and Apple were not among the 49 parties to submit comments taking views. My organization, CCIA, also filed comments, noting that the explosion of poor quality patents was complicating the standards process.)
Apple escalated what was a series of border skirmishes on the edges of the smartphone patent ecosystem to full-blown nuclear war a couple of months before Google announced its acquisition of Motorola and its large patent portfolio. The patent cold war (replete with some minor shooting wars on the periphery) between the major players became hot when Apple had so much to lose from the inevitable commoditization of the smartphone and tablet space that the heavy cost of waging war seemed to be worth the price. Given that the Android manufacturers were willing to accept much smaller margins on their products than Apple (whose amazing 40%+ margins had skyrocketed it to the most valuable company on Earth), Apple was faced with two choices: cut profit margins and invest more in R&D to compete aggressively or be eventually consigned to a high-end, niche player in the smartphone and tablet ecosystem. Apple liked neither option and decided costly courtroom warfare was better than actual product competition. And, in doing so, changed course on long-held views on the anticompetitive use of intellectual property.
(As an aside, being a market leader with huge profit margins by definition not only attracts competition in the current product market, but it also invites disruptive competition. Newer competitors, like Samsung, Amazon and new Chinese entrants, are perfectly fine developing products — and more open ecosystems — that live on razor-thin margins and are more willing to quickly evolve new business models and technologies.)
Once the war heated up, incentives changed.
Where once most tech companies with major patent portfolios had either asked that the government stay out of the way on SEPs or chose to remain agnostic, now the battle lines were drawn around who had what types of patents. Where “more flexibility” to use patents how you please was nice when things were peaceful, the realpolitik of a shooting war meant that the patents were now weaponized. Contemporary expedience replaced future flexibility as the loadstar of corporate IP policy.
A line began to form between hardware companies (including technology licensing companies, like Qualcomm, who develop technology and then license their IP) who by necessity had patent portfolios chock full of SEPs (their products need to work on the same airwaves and interoperate with wireless infrastructure providers) and software companies (who had portfolios with comparatively few SEPs because they design their own unique software offerings). That explains why Apple and Microsoft both issued their proclamations against getting injunctions on SEPs the same month that Google purchased Motorola. (As Earle Thompson, then the Intellectual Asset Manager at Texas Instruments, said in the 2002 FTC hearings, the threat of an injunction is the “big difference and a big lever” in IP negotiations.)
It also helps explain why Google, the darling company of the patent reform movement who had been largely silent in the niche realm of SEP policy, was less gracious with its SEP commitments. Without the potential for injunctions, Motorola’s patent portfolio would be much less effective in defending Android from patent attacks. Google and Samsung were seeking injunctions against Apple and Microsoft for leverage in negotiations as Apple and Microsoft held injunctions (or were actively seeking injunctions) against on many Android products.
Ministers of Wartime Propaganda
To extend this war analogy further, as the esoteric world of intellectual property and standards policy became a key battleground of the global smartphone war, corporate PR machines on both sides began intensely pumping out wartime propaganda. The problem was much of the nuance described above was lost.
Apple stressed — and rightly so — the inherent problems around seeking injunctions on SEPs (however, as discussed earlier, the problem had previously centered on patent trolls). What the Cupertino giant failed to mention was that (a) it started this mess and was asserting equally controversial abstract software patents, and that (b) the SEP suits it was worried about were attempts by Samsung and Motorola to defend themselves and keep their products on the market. (Clarification: Although Samsung, HTC and Motorola initiated some suits, they did so after the war had been started. In patent warfare, a good defense necessarily involves a strong offense as you need to choose the products, patents and legal venues that serve you best, as that is what your opponents are doing to you. You don’t want to limit yourself exclusively to the battles and terrain chosen by your opponents.) These countersuits largely exemplify why companies were unwilling to completely rule out injunctions on SEPs: they wanted the flexibility to use their patent portfolios defensively if attacked by competitors.
The hardware companies, by contrast, largely downplayed the problems inherent with seeking injunctions on SEPs, while focusing — rightly so — on the low-quality, abstract software patents and ubiquitous design patents that were being asserted against them that were essentially de facto standards that should have never been issued in the first place. In other words, Apple’s patents were so ambiguous and broad that companies that made smartphones and tablets could not possibly make products that consumers wanted without using rounded rectangles or basic touch screen functionality — much like actual SEPs. Furthermore, hardware companies contended that many of Apple’s patents should never have been issued in the first place as they were obvious and preempted by previous technology (contentions that hold more weight as the PTO has started invalidating many of Apple’s core patents).
Unsurprisingly, each side focused on what the other side was doing wrong and downplayed their own faults.
SEPs Are Just the First Piece of Patent Peace
Immediately after the FTC settled its case with Google, FTC Commissioner Rosch told a reporter, “But I will tell you this — we designed [the Google SEP settlement] to bring patent peace.” Now, I may agree with Rosch’s sentiment, but it is only part of the job. Unilaterally disarming one side of the patent war — particularly the non-aggressor side — is an interesting way to achieve peace. It might work in the short-run (as one side can’t fight back as effectively), but you can be assured that Samsung, Google and other major SEP holders will begin stocking up on non-SEP patents that read on their competitors products. (Notice that Google withdrew its SEPs in its current battle against Microsoft in the ITC, but still continued the suit with the non-SEP patent at play.)
Given that current smartphones incorporate at least 250,000 U.S. patents (and growing), procuring patents to use against your competition is not difficult, especially given the relatively lax patent-issuance standards currently employed by the USPTO. The SEP settlement could even make companies, at least on the margins, less likely to participate in standard-setting bodies. Or design new standard consortia that remove FRAND obligations — so we are back in the old ambiguous pre-smartphone war world all over again.
The FTC solution — which is reasonable in a vacuum — does little to fix the long-term problem. If the massive problems with patent quality and the explosion of software patents are not fixed, the FTC’s victory will be largely pyrrhic. Samsung, Google and other companies with SEP heavy portfolios have seen the writing on the wall and started ramping up their filings and acquisitions of non-SEP patents, many of them being software patents. In fact, in the new shooting war we have now, an effective patent strategy is as much about acquiring patents that your competitors are likely to infringe upon (even if it is not a core technology of the patent filer) as it is about acquiring patents that protect your own products.
SEPs were the weapons of choice for companies like Samsung and Motorola when they were ambushed as they were the only weapons they had in their arsenals that they could effectively use to attack their competitors. A couple years from now, that will not be the case. Fiduciary duty and the strong desire to compete in the market will drive them to build bigger, scarier patent portfolios with patents that can’t be neutered by the competition agencies of the world.
In this scenario, nobody wins — except patent lawyers and patent trolls who will likely see an increased supply of patents on the secondary market. Real patent peace needs a comprehensive solution.
Appendix: FTC SEP Cases Prior to the Smartphone Wars
**The 3 major cases were:
1996 Dell Case: During its participation in the standards-setting process, Dell intentionally withheld patents pertaining to computer bus technology and sought to enforce its patent rights after the standard had been settled upon. The FTC won a settlement in this case that precluded Dell from enforcing its patents.
2006 Rambus Case: The chip manufacturer Rambus (which turned into a patent troll when its business model took a turn for the worse), participated in a JEDEC standard setting process for SDRAM and then withdrew from the group shortly before the standard was complete. JEDEC members pledged to keep patented technology out of the standard, but when Rambus withdrew it augmented its own patent portfolio with applications designed to capture much of the standard. It waited several years until the standard was solidified and on the market and then Rambus “ambushed” several of the standard’s implementers. The FTC challenged the action but ultimately lost because it failed to demonstrate that the standard body would not have implemented Rambus’s patented technology even if Rambus had fully disclosed its patent position.
2008 N-Data Case: The FTC reached settlement with the patent troll Negotiated Data Solutions (N-Data), who acquired crucial Ethernet SEPs that originally belonged to National Semiconductor, reneged on the original commitments on the patents and tried to drastically raise licensing fees after the entire computer industry had already adopted the Ethernet standard.