By Garrard R Beeney and Renata B Hesse
- Properly organised patent pools benefit both licensors and assignees and may be the purest market example of the well-established efficiencies of combining complements.
- A deep dive into their development charts various improvements and areas ripe for further change.
- It also highlights why pools are a practical and often cost-saving solution for manufacturers dealing with hundreds or even thousands of SEPs.
Over the past 25 years multiple patent owners have formed dozens of patent pools – where groups of rights holders offer the marketplace a single standard license to their patents – and joint licensing programmes. Thus, pools have facilitated thousands of consensual licences, unburdened by the inefficiency of litigation or other disputes, and have enabled the widespread adoption of beneficial, innovative technologies. Yet despite this, certain critics continue to voice their distrust of the model and the potential for abuse.
Failure to recognise how both innovators and implementers can benefit from efficient pools creates a distorted licensing narrative. Imagine the impact on new product innovation, sales and technological development without multiple successful pools based on licensing successive and more powerful generations of video compression and connectivity technologies? The smartphone as we know it today likely would not exist, either in terms of innovative products or product sales. When it comes to properly run pools, the interests of innovators and implementers are largely aligned.
This is not to say that pools cannot be improved. Pools can be mismanaged and can fail to achieve the benefits offered by the model. Below we explore several efficiency-enhancing structural changes that have developed over the years. No doubt those advances will continue.
Patent pools – the arguments for
The world’s leading patent pool administrators – including Access Advance, MPEG LA, Sisvel and Via Licensing – currently list 39 licensing programmes and over 5,224 licensees.
Pools not only provide access to cutting-edge technologies on standard terms but also drive further innovation in at least two ways. First, pool revenue received by innovators incentivises and funds further R&D. Second, the availability of a single licence to a wide assortment of patents held by multiple entries encourages implementers to adopt new technologies, safe in the knowledge that they will have access to necessary intellectual property at a known and transparent cost.
A successful pool allows innovators of all sizes to sidestep the need to build expensive licensing structures, with revenue instead fueling innovation. Without this incentive, there will be less innovation. And with less innovation, implementers will sell fewer products.
This principle has been recognised by both economists and government agencies. In their 2015 paper, “The Effect of Patent Pools on Patenting and Innovation”, Justus Baron and Tim Pohlmann found that patent pools had significant positive effects on further patenting, while Section 5.5 of the US Department of Justice and Federal Trade Commission’s Antitrust Guidelines for the Licensing of Intellectual Property acknowledges that they benefit innovation.
Moreover, because a pool administrator has no means to force an implementer to accept a licence, executed pool licences reflect market terms, which empowers implementation at a cost that is acceptable to the licensee.
Without patent pools, innovators and implementers face the uncertainty and inefficiency of:
- thousands of bilateral licence transactions;
- varied royalty rates that can distort competition;
- duplicated substantial licensing transaction costs; and
- multiple other inefficiencies.
This was illustrated starkly in Microsoft Corp v Motorola Inc (795 F 3d 1024, 1043-44 (9th Cir 2014)), when the Court of Appeals for the Ninth Circuit upheld a lower court finding that a FRAND royalty rate could be as much as three times the pool rate to account for the value that a patent holder receives from a grant-back licences and promotion of the standard.
As a final argument, patent pools reduce uncertainty about innovator hold-up.
A rich history
Patent pools are not a new solution to the problems of multiple patents covering a single product. Over a century ago, on the eve of the United States’ involvement in World War I, airplane engine manufacturers agreed to a pooling arrangement proposed by the National Advisory Committee for Aeronautics.
The pool addressed manufacturer reluctance to build much-needed engines because of the threat of patent infringement suits. Accordingly, it set standard royalty rates, which substantially reduced the royalties that certain patentees had threatened to seek in litigation.
The problems faced by engine manufacturers a century ago – when many necessary patents were held by multiple entities – are all the more real today. How does a manufacturer navigate a thicket created by multiple patents covering the product that it wants to market? Such issues include determining:
- the total royalties payable for patented technologies adopted in a product;
- whether a licence under all necessary patents will be available at a fair royalty; and
- whether differing licence terms and costs offered to different implementers will distort competition among them.
While pools and their reception have evolved substantially since the airplane engine pool, the fast pace of technological advancements makes patent pools an increasingly important solution today.
Modern technological developments and enhanced innovation have increased the prevalence of products covered by multiple patents. The prime example is smartphones, which not only enable telecommunications but also provide audio and video entertainment, business computations, email, web surfing, data storage and photography. The fundamental question becomes: what value do different patented technologies provide to such a device and how can a patentee and implementer agree on this? Patent pools answer this by relying on market-accepted standard terms, thereby reducing friction and costs that might otherwise dominate licensing discussions.
In addition, the pool model addresses the proliferation of standards set by standard development organisations (SDOs). These bring together innovators from multiple companies, universities and research institutes to develop the best and most innovative technological solutions, which often cannot be devised without collaboration.
The success and proliferation of SDOs recognises that standards provide important interoperability, which increases competition, and that technological development is accelerated by collaboration between innovators, to the ultimate benefit of consumers. It is because of the standardisation of phone connectivity that customers of a particular cellular service provider can choose from a wide array of branded handsets without worrying about compatibility.
In a typical SDO process, participants submit proposed solutions to identified technological problems that must be solved in order for the new technology to meet the established goals. The SDO working groups then agree on the best solution to achieve their objectives. Such inventions are often protected as patents, which, as a result of their adoption, then become SEPs (ie, the standard cannot be practised without infringing that patent).
However, not all patents essential to a standard are of equal value. Some may be foundational, others less significant, and some subject to dispute about whether they are essential at all. As standardisation bodies and others address proposals to revise the procedures by which a patentee declares its patent to be essential, there can be little doubt that increasing technological complexity means that a growing number of solutions will be required and more SEPs are likely to cover products that incorporate a standard. For example, it took the addition of a staggering number of inventions to advance from 2G to 5G connectivity.
Without a pool licence granting rights to use the thousands of patents necessary to practise a standard, implementers would face multiple inefficiencies, uncertainty and higher costs. What is more, they would have to negotiate numerous licences with multiple entities. This raises transaction costs and likely also raises aggregate costs for royalties. Implementers would also encounter uncertainty about the total cost of employing technology as they entered into multiple negotiations and would face the risk that not all patentees are willing to provide a licence or at least one under reasonable terms.
Assuming that the necessary licences were available and executed, implementers would likely then have to provide sales or royalty reports to all or most licensors, while facing other inefficiencies. Although an implementer may elect to assume these challenges, a pool licence helps to avoid them.
These benefits are not mere theories. They have been demonstrated by the actual impact of patent pools in the market. The history of the MPEG-2 video pool (mirrored by the success of MPEG audio pools) demonstrates the multiple advantages that pools confer on economic development, scientific innovation and sales and use of new and beneficial products.
When the gains of digital video over analogue video were envisaged some 30 years ago, concerns were raised about the amount of digital information or bits necessary to represent an analogue video in digital format. The solution was to compress the necessary digital information to represent the video into as few bits as possible. Innovative companies and universities from around the world met under the auspices of the MPEG to craft a compression technology that would enable digital video and audio.
The multi-year collaborative process of developing MPEG-2 largely concluded in 1994. However, once the standard was defined, concerns were raised about the need to license multiple patents owned by multiple entities that had contributed their innovations, with questions over whether uncertainty about the availability of these would delay or prevent wide-spread adoption. Multiple other unknowns threatened the use of this innovative technology, including the following concerns:
- Would all necessary patent owners agree to license their patents essential to MPEG-2?
- What would be the total requested royalty by all patent owners?
- Would similarly situated implementers be charged different royalties, risking distortion to competition?
- What would be the transactional costs of negotiating multiple licences and providing multiple royalty reports?
- Would litigation and/or hold-up create additional costs and inefficiencies?
These questions were answered when the MPEG-2 developer-patentees offered all their essential MPEG-2 patents in a single standard licence through a licensing administrator called MPEG LA. The rest, as they say, is history. After the launch of the MPEG-2 pool, the MPEG-2 technology became one of the most successful standards ever developed. It was an IP game changer for implementers as well. It appears that none of the thousands of implementers that executed the MPEG-2 pool licence have ever faced litigation from any MPEG-2 SEP owner participating in the pool.
Billions of devices have now been sold with MPEG-2 technology, the vast majority of which are licensed through the standard pool licence. Innovators and implementers alike have enjoyed minimal transaction costs, while similarly situated innovators have accepted licence terms available to all and paid the same royalty rate.
The MPEG-2 video pool, like the MPEG audio pools, has also had a positive impact on innovation by helping innovators to achieve a fair return on their R&D investment, MPEG continues to develop more standards, leading to more and more powerful compression technologies (eg, AVC, HEVC and now VVC – for more on this see “Qualcomm takes pole position for next-generation video standard patents“) – each time roughly halving the amount of digital information needed to produce the same quality video.
The collective licensing of patents owned by multiple entities does, theoretically, present opportunities for abuse. Areas of hypothetical concern include:
- combining patents that may be substitutes rather than complements;
- using licensee sales or other confidential information to adversely affect competition;
- using the pool to reduce competition in downstream licensed products;
- delaying innovation by requiring grant-backs from licensees, which fail to reward licensee innovations related to the licensed technology; and
- mandating package licences that may include patents that a licensee does not need in order to practise the licensed technology.
However, all of these issues have been recognised and addressed in structures adopted by many pools.
Acknowledging both the novelty of their proposed pool solution to the MPEG-2 patent thicket and the fact that structures could be implemented to address theoretical risk without reducing efficiency, in 1997 the MPEG-2 SEP owners sought guidance about their proposed pool structure from the US Department of Justice (DoJ), the European Commission and the Japanese Fair Trade Commission.
As a result, the DoJ issued a business review letter while the European Commission issued a comfort letter. In oral communications, the Japanese Fair Trade Commission agreed that the proposed pool structure likely enhanced efficiencies and competition. In the succeeding 25 years, no competition authority in the world seems to have challenged a pool that employs the approved structure and safeguards of the MPEG-2 pool.
These pro-competitive safeguards have since been implemented in a number of successive pools. They address multiple theoretical risks of pools and joint licensing programmes. For example, to address whether pools combine complementary patents rather than substitutes, many pools independently evaluate whether licensed patents are in fact essential before including them.
Many pools also provide licensors with the means to challenge the essentiality of any licensed patent, along with economic incentives for successful challenges. Such measures further ensure that pools package complements, not substitutes.
To avoid anti-competitive effects, pools generally keep competitively sensitive information about individual licensees confidential from licensors. Pool meetings are regularly attended by competition law counsel to further ensure that anti-competitive conduct is avoided.
Pool licence terms are generally transparent and frequently published so that implementers are aware of the IP costs of including a licensed technology in their products. Pools also generally make clear that any grant-back obligation, if required, is on terms that provide a fair return to licensees so that incentives to innovate are not adversely affected.
Significantly, most pools allow or even require that all licensors bilaterally license their pool patents outside the pool licence. Thus, by definition, the pool becomes an alternative to the world that would exist without it (ie, individual licensing negotiations). Any implementer is free to consider (or reject) this alternative without affecting the options otherwise available to it.
Arguments have been raised that once certain pools with a large number of licensors and licensed patents have been formed, bilateral licensing becomes impractical. In effect, they claim that pool terms as compared with bilateral licensing are so good that the latter becomes an unattractive choice. However, this misses the point. Regardless of how the terms of a pool compare with bilateral options, the fact remains that a typical pool adds a choice and does not foreclose possibilities that existed pre-pool. If a bilateral licence becomes unattractive, this is merely because the pool is far more efficient. Moreover, when a licensor seeks a bilateral consensual licence rate that is in part based on its share of the pool royalties, the pool can have a positive impact on negotiations.
While these structures are well known and regularly cites and employed by pools, one critical safeguard all but eliminates the possibility of abuse: pools, by their nature, depend on consensual licensing. If terms are rejected by the market of licensees, the pool will fail.
This fact significantly disciplines the terms that pools set for licences. Indeed, if pool licence terms – frequently crafted in consultation with licensees – are rejected by the market, the pool licensors will be either forever fighting an ocean of expensive litigation or watching while implementers use their licensed technology without paying for it. In either case, the pool fails.
Another frequently overlooked structural safeguard is the fact that patented inventions can be implemented without the owner’s permission. Thus, implementers can – and often do – add patented technology to their products before seeking permission (ie, before entering into a licence). Unlike an electric company that can turn off its services when bills are not paid or a vendor that can refuse to make further deliveries if sale terms are not met, rights holders cannot themselves prevent the unauthorised use of their intellectual property.
When faced with unauthorised use, a patentee has three choices:
- allow its valuable property to be used without consent or remuneration;
- engage in costly and uncertain litigation; or
- offer licence terms to which the implementer will agree.
Because the goal of successful pools is to enter into hundreds or thousands of licence transactions, they must offer terms that the market will accept.
Additional safeguards post-MPEG-2
Based in part on the lessons learned from MPEG-2 and the enormous success of the MPEG video and audio pools for innovators, implementers and consumers, successful post-MPEG pools have not only adopted these safeguards but have developed additional structures to guard against potential abuse.
Many successful pools now commit to licensing on FRAND terms following the examples of licence terms advocated by several SDOs. More recent successful pools include those for HEVC, 3G, 4G and 5G connectivity standards (including multi-standard programmes), Wi-Fi and W-Lan standards, TS-UNB, AAC, MPEG-4 audio and video, DRM, AGORA-C, OCAP-thru2way, AVC, QI, EV charging, EVS, AV1, VP9, ATSC, VC-1, MVC, 1394, DASH, DVB-T, DVB-T2, DVB-S2, DVB-2SX, DVB-SIS and DVB-CSA.
Whether considered a contract between the patentee and the SDO to which implementers may be third-party beneficiaries (as in the United States) or obligations that may arise from market power (as in many EU member states), SDO FRAND obligations are imposed on innovators and implementers in many SEP license negotiations.
Pools have also evolved to address the economic issues that arise when a manufacturer already has a licence to some, but not all, patents licensed by the pool. Rather than seek the full licence rate from partially licensed licensees, they provide so-called ‘netting’ to ensure that a licensee receives credit for existing licences and does not pay twice. This incentivises licensors to enter into bilateral licences outside the pool; licensees and licensors know that they can agree to a bilateral licence without the former being penalised if it also wants a pool licence.
In addition, pools have adopted royalty distribution formulas that incentivise greater participation by a larger number of licensors to increase their value and efficiency. The DoJ acknowledged these efficiencies in its 2020 business review letter regarding the Avanci 5G patent pool, citiing a study by Anne Layne-Farrar and Josh Lerner, “To Join or Not to Join, Examining Patent Pool Participation and Rent Sharing Rules”(29 Int’l J Indus Org, 294, 295 (March 2011)). Of course, the more licensors and licensed patents there are, the more efficient a single pool licence becomes. These distribution formulae may also incentivise smaller licensors to join pools and enable smaller inventors to monetise their valuable intellectual property.
Many pools have gone to even greater lengths to make their terms attractive to licensees. Initiatives include:
- fixed annual royalties to avoid the transaction cost of reporting sales;
- annual rate caps when justified in certain markets;
- sliding scales of rates;
- lower rates for basic versions of licensed technology that use fewer tools and licensed patents than other versions;
- incentives for early licensee adoption and ending hold-outs; and
- other licensee-friendly means to achieve the ultimate pool objective: a large volume of licence transations on consensual terms.
While successful patent pools are not the sole cause of increased innovation and consumer welfare, efficient access to patented technology provided by pools and a fair return to innovators have played a material part.
Widespread recognition of the benefits of patent pools
The fact that pools provide efficiency-enhancing benefits to the IP world has been recognised for decades. WIPO has said that: “Patent pools represent a common instrument to deal with…inefficiencies[,] allowing complementary IP assets to be organized under a single contract to be not only cross-licensed among patent pool members but also licensed to interested third parties.”
More importantly, government agencies and courts have repeatedly recognised that pools – employing some or all of the safeguards noted above – enhance consumer welfare.
After a review of multiple pools, the DoJ concluded in 2007 that “each of these patent-pooling proposals were likely to create substantial integration efficiencies by reducing the time and expense of disseminating the patents to interested licensees, clearing blocking positions, and integrating complementary technologies”.
The DoJ continues to acknowledge these benefits today. In July 2020 it approved a pool for 2G, 3G and 4G SEPs used in vehicles and smart metres. And just this year, it praised a pool licensing non-SEPs formed by 15 leading universities in the United States. According to the DoJ, the pool efficiently addressed “institutional challenges unique to universities in the physical science context”.
The European Commission has been similarly supportive. In 2014 the European Competitiveness and Sustainable Industrial Policy Consortium wrote: “Specialized pool administrators including MPEG LA, Via [Licensing], and Sisvel have attempted to set up pools for a wide range of standards… proven to be a successful mechanism to facilitate licensing.”
In 2015 the commission’s Expert Group on Patent Aggregation outlined the many advantages of patent pools, including improving transparency, reducing asymmetry and even fostering “socially desirable technologies”. One example cited was the London Declaration on Neglected Tropical Diseases, a collaborative disease eradication programme from 2012 that demonstrated how pools can be used to service wider societal goals.
In 2017 the commission encouraged patent pools that comply with EU competition law, stating: “They can address many of the SEP licensing challenges by offering better scrutiny on essentiality, more clarity on aggregate licensing fees and one-stop-shop solutions.” The overall goal was to “incentivize the development and inclusion of top technologies in standards, by preserving fair and adequate return for these contributions”.
And as recently as 25 November 2020, in addressing the covid-19 pandemic, the commission stated its support for “voluntary pooling and licensing of IP”, noting that it was “looking into ways to incentivize the rapid pooling of critical IP in times of crisis”.
Clearly, multiple markets have benefitted from pools. Next week, we will examine the recent criticisms of them and evaluate certain suggested changes to the model’s structure.
The views expressed in this article are those of the authors and do not necessarily represent the views of Sullivan & Cromwell or any of its clients. The authors wish to acknowledge the assistance of Paul von Autenried in preparing this article.
Garrard R Beeney
Sullivan & Cromwell LLP
Renata B Hesse
Sullivan & Cromwell LLP