The “New Madison” Approach to
Antitrust and Intellectual Property Law
Thank you Professor Yoo, and the University of Pennsylvania Law School for the opportunity to speak today. I’m honored to be here for today’s conference on “The Future of Standard Essential Patents.” While the conference is forward-looking, it is fitting that we are here in Philadelphia, the original birthplace of U.S. patent rights. The U.S. Constitution recognizes these rights in Article I, where it empowers Congress to “promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.”
Many in the patent community champion Thomas Jefferson as the father of patent law. President Jefferson’s contributions and public influence in this area cannot be understated, as he was the first lead patent examiner in the United States and made a number of influential writings on patent policy in the early years of the Republic. But, lately, it has been vogue among some critics of the U.S. patent system to selectively quote Jefferson to make the case that intellectual property rights ought to be reined in.
I submit that the true father of U.S. patent law was the Founding Father principally responsible for drafting the Constitution, James Madison.
Madison wrote in The Federalist Papers that “[t]he copyright of authors has been solemnly adjudged, in Great Britain, to be a right of common law,” and that “[t]he right to useful inventions seems with equal reason to belong to the inventors.” Madison went on to note a policy rationale for patent rights, stating that “the public good fully coincides . . . with the claims of individuals.” Analogizing patent rights to common law rights was a truly revolutionary position. In Great Britain, patents were conferred on an arbitrary basis by the King or Queen to their political and economic allies, often with little regard for the utility of the invention.
The notion that “rights” should belong to inventors and that this right “coincides” with “the public good” was not widely shared at the time. Indeed, Benjamin Franklin, the famous inventor, intellectual, and founder of the University of Pennsylvania, took a more magnanimous approach. He wrote in his autobiography that he did not oppose the use of his inventions without compensation, as he had “no desire of profiting from patents himself, and hat[ed] disputes.”
The exchanges between Jefferson and Madison on the question of patent rights in 1788 are therefore illuminating of Madison’s intellectual influence. Reflecting the general anti-monopoly sentiment at the time, Jefferson wrote from his post in Paris that “the benefit even of limited monopolies is too doubtful to be opposed to that of their general suppression.”
In response, Madison acknowledged that monopolies “are justly classed among the greatest nuisances in Government.” But he recognized a limited exception for patents. “[I]s it clear,” he asked Jefferson, “that as encouragement to literary works and ingenious discoveries, [monopolies] are not too valuable to be wholly renounced?” Madison answered his own question, demonstrating a nuanced understanding of how to balance concerns about monopolies with creating incentives to innovate: “Monopolies are sacrifices of the many to the few. . . . Where the power . . . is in the many not in the few, the danger can not be very great that the few will be thus favored. It is much more to be dreaded that the few will be unnecessarily sacrificed to the many.”
Madison understood that replacing monarchy with democracy reversed the threat of the misapplication of power, creating a risk that patent holders might suffer from the tyranny of the majority seeking to benefit unfairly from their innovation.
Madison’s view ultimately prevailed in the text of the Constitution, tying the right to a patent to innovation, or “the progress of science and useful arts.”
Remarkably, the word “right” appears only once in the original Constitution—which took effect two years before the Bill of Rights was ratified—in the Copyright and Patent Clause. The reward of a patent for a fixed period aligned the interests of inventors, who need incentives to innovate, with the interests of the public, who want the fruits of innovation. It was an ingenious compromise that unleashed the power of innovation in the young Republic.
This history would not be complete without noting that, in the end, Thomas Jefferson shifted his perspective on patents to embrace a more Madisonian position. He wrote Madison in 1789 that he would support an article in the Bill of Rights specifying that “[m]onopolies may be allowed to persons for their own productions in literature and their own inventions in the arts” for a fixed term. Jefferson went on to become the administrator of the patent system under the 1790 Patent Act, and authored the subsequent 1793 Patent Act. In his writings, Jefferson voiced his support for patent protection for invention on the ground that “ingenuity should receive a liberal encouragement.”
In recent months I have found inspiration in this history and Madison’s dogged perseverance in favor of strong patent protections—a view that stood at odds with much of the received wisdom and practice of the day.
There has been a shift in recent years toward what I would call a “retro-Jefferson” view of patents as conferring too much power that ought to be curbed, either through reinterpreting antitrust law or establishing patent policies of standard setting organization (“SSO”) that favor implementers who practice on a patent when they build new technologies. Many advocates of reducing the power of intellectual property rights cite the so-called “hold-up” problem in the context of SSOs. As many of you know, I believe these concerns are largely misplaced. Instead, I favor what I call the “New Madison” approach to the application of antitrust law to intellectual property rights.
The New Madison approach, if I may, has four basic premises that are aimed at ensuring that patent holders have adequate incentives to innovate and create exciting new technologies, and that licensees have appropriate incentives to implement those technologies.
First, hold-up is fundamentally not an antitrust problem, and therefore antitrust law should not be used as a tool to police FRAND commitments that patent-holders make to standard setting organizations.
Second, standard setting organizations should not become vehicles for concerted actions by market participants to skew conditions for patented technologies’ incorporation into a standard in favor of implementers because this can reduce incentives to innovate and encourage patent hold-out.
Third, because a key feature of patent rights is the right to exclude, standard setting organizations and courts should have a very high burden before they adopt rules that severely restrict that right or—even worse—amount to a de facto compulsory licensing scheme.
Fourth, consistent with the fundamental right to exclude, from the perspective of the antitrust laws, a unilateral and unconditional refusal to license a patent should be considered per se legal.
To understand what I mean when I say that patent hold-up is not an antitrust problem, it is important to step back to consider the purpose of antitrust law—what it does, and what it should not do. At its core, antitrust law aims to protect competition and consumers.
Antitrust law is guided by a consumer welfare standard, which dates back to the origins of the Sherman Act. The ultimate focus on the consumer gained academic prominence in the late 1970s and 1980s through the intellectual leadership of Judge Robert Bork, Judge Frank Easterbrook, and others. This standard sharpens the focus of antitrust scrutiny to anticompetitive practices that are harmful to consumers, rather than competitors, so that the antitrust laws are not misapplied to advance social goals unrelated to consumer welfare and efficiency.
Importantly, however, the consumer welfare standard is not synonymous with a policy always favoring lower prices. For example, high demand for an exciting new product may drive up its price, but that may simply reflect consumer preference for a superior product relative to alternatives. Antitrust law is intended to protect this behavior, not punish it, so that others will have incentives to innovate and compete themselves, all for the benefit of consumers. Such dynamic competition should be encouraged by our enforcement policies.
Rather than focusing on prices in isolation, antitrust law instead protects consumers where practices also harm competition—that is, they harm some “competitive process” in a manner that causes harm to consumers in the form of above-competitive prices, lower output, or reduced efficiency. Indeed, directly showing harm to end-consumers is not always necessary to prove a violation of the antitrust laws. For example, where collusion among buyers pushes input prices down—what economists call a monopsony effect—that may violate the antitrust laws because there is harm to competition even though it results in lower prices.
This is where theories that unilateral patent hold-up is an antitrust problem go wrong. Stating that a patent holder can derive higher licensing fees through hold-up simply reflects basic commercial reality. Condemning this practice, in isolation, as an antitrust violation, while ignoring equal incentives of implementers to “hold out,” risks creating “false positive” errors of over-enforcement that would discourage valuable innovation.
Advocates of using antitrust law to reduce the supposed risk of patent hold-up fail to identify an actual harm to the competitive process that warrants intervention. If an inventor participates in a standard-setting process and wins support for including a patented technology in a standard, that decision does not magically transform a lawful patent right into an unlawful monopoly. To be sure, that decision gives the patent holder some bargaining power in claiming a piece of the surplus created by standardization. And, it would require the patent holder to live up to commitments as they would have bargained for it, enforceable by contract laws. But standard setting decisions are intended to be a recognition that a technology is superior to its alternatives. A favorable SSO decision, like a patent itself, is a reward for an innovator’s meritorious contribution whose wide-ranging benefits can ripple throughout the economy, contributing to dynamic competition. Arguments that inclusion in a standard confers market power that could harm competition typically rest on the unreasonable assumption that the winning technology is no better than its rivals.
It is therefore unsurprising that proponents of using antitrust law to police FRAND commitments principally rely on models devoid of economic or empirical evidence that hold-up is a real phenomenon, much less one that harms competition. Since hold-up theories gained traction in the early 2000s, it is striking that they still remain an empirical enigma in the academic literature. Antitrust law demands evidence-based enforcement, without which there is a real threat of undermining incentives to innovate.
That is why I believe so strongly that antitrust law should play no role in policing unilateral FRAND commitments where contract or common law remedies would be adequate. I worry that courts and enforcers have overly indulged theories of patent hold-up as a supposed competition problem, while losing sight of the basic policies of antitrust law. They lose sight of the fact that antitrust law is not just remedial; it is, importantly, intended to deter through the threat of treble damages. As enforcers, we have a responsibility to ensure that antitrust policy remains sound, so that U.S. consumers continue to enjoy the benefits of dynamic competition and innovation, and so we do not export unsound theories of antitrust liability abroad, where economically dubious enforcement actions can have serious consumer-harming effects on U.S. businesses, consumers, and workers.
The second premise of the New Madison approach I advocate is that standard setting organizations, as collective bodies, themselves should avoid over-indulging theories of patent hold-up, to the detriment of patent rights. SSOs should instead strive to ensure that their patent policies create maximum incentives for innovators to invent (or at a minimum don’t curtail incentives to innovate), and for licensees to implement.
Achieving this goal is not an easy task. At minimum, it requires a recognition of the fact that implementer hold-out poses a more serious threat to innovation than innovator hold-up. To be sure, both practices threaten to undermine innovation through under-investment in new technology. But, there is an asymmetry between the two: innovators must make significant upfront investments in technology before they know whether it will pay off, whereas implementers can delay at least some of their investments in a technology until after royalty rates have been determined.
To the extent antitrust law should play a role, it is to ensure that concerted action among implementers or innovators does not occur at any level of the supply chain. Specifically, as I noted this past Fall, the Antitrust Division will be skeptical of rules that SSOs impose that appear designed specifically to shift bargaining leverage from IP creators to implementers, or vice versa. What do I mean by that? As enforcers, we have only limited insight into the patent policies of various standard-setting organizations, and we do not seek to impose a top-down mandate to skew the playing field clearly in the direction of innovators or implementers. But we expect there to be some symmetry between these competing interests, which manifests itself in two ways.
First, at SSOs, we hope to see a diversity of views represented on patent policy committees to give us confidence that patent policies are based on reasoned and unbiased decision-making. We strongly encourage SSOs to avoid allowing voting blocks of competitors to dominate decisions on patent policy or on which technology to incorporate into a standard. That kind of action would confirm the Supreme Court’s observation that SSOs “can be rife with opportunities for anticompetitive activity.” Ensuring that no voting blocks take hold would help negate the risk that a rule or standard is the product of a buyer’s or seller’s cartel. As long as an SSO’s IP policies are the product of a consensus or a clear majority that includes both standard-essential patent holders and implementers, the Department of Justice should have no reason for concern. On the other hand, if an SSO’s policymaking decisions appear to be dominated by implementers, and the resulting policies or standards appear to be heavily skewed toward implementers and away from innovators, that’s already two strikes.
Second, I believe innovation policy would benefit from a diversity of patent policies across standard setting organizations. Optimally, competition can begin to emerge among SSOs within the same industry, with dueling patent policies that allow for the more efficient regime to prevail. Across industries, we expect that patent policies and the requirements for the inclusion of patented technology in a standard will vary depending on the technology in question. By contrast, I worry that advocacy by government agencies in recent years could lead SSOs to adopt a uniform approach to articulating specific commitments necessary for inclusion in a standard—an approach that may be skewed too far in the direction of implementers. This unfortunate trend should not continue.
The third premise of the New Madison approach to antitrust law and intellectual property is to respect the core of what it means to hold an IP right—namely, the right to exclude. In his letters to Thomas Jefferson, Madison acknowledged that state-conferred monopolies are “among the greatest nuisances in government,” but maintained that these “nuisances” could be harnessed to serve the greater good of social progress and innovation through patent protection. His analogy of patents to the “common law . . . copyright of authors” in The Federalist Papers is telling because, at the time, the copyright of authors was understood as a property right. Equipping patent holders with the property right to exclude therefore goes hand-in-hand with the goals Madison envisioned for the U.S. patent regime.
Understanding patent rights, once conferred, as a form of property right helps frame the current debate over injunctions, and demonstrates how far we’ve strayed off course. Under current Federal Circuit law, a standard-essential patent holder faces significant difficulty in establishing a right to an injunction instead of damages. In a worrisome trend, some commentators have suggested that the mere act of seeking an injunction order to prevent infringement raises competition concerns, and, with a degree of hubris, litigants have advanced such theories as a basis for antitrust liability. Taken together, these trends fundamentally transform the nature of patent rights away from their constitutional underpinnings. They convert a property rule into a liability rule, and amount to a troubling de facto compulsory licensing scheme.
It is not difficult to understand why that is the case, particularly in the context of standard setting. If a patent holder effectively loses its right to an injunction whenever a licensing dispute arises, or is deterred from seeking an injunction due to the prospect of treble damages, an implementer can freely infringe, knowing that the most he or she will eventually have to pay is a reasonable royalty rate. Implementers have a strong incentive to pursue this course while holding out from taking a license due to the high injunction bar for innovators that make FRAND commitments. It is a harmful arbitrage that should be discouraged.
Some may be skeptical of this claim, given that “willful” infringement entitles a patent holder to compensation up to treble damages. But it is extremely difficult to prove willfulness, a demanding standard that the Supreme Court emphasizes should be limited “to egregious cases of misconduct.” Under recent developments of the law, the standard for obtaining an injunction and the standard for proving willfulness both work to the benefit of implementers and significantly limit the downside risk of infringement. This results in a de facto compulsory licensing scheme for FRAND-encumbered patents deemed “standard essential,” and could serve as a disincentive for innovation or for patent holders to contribute technology to the standard-setting process in the first place. Deterring the right to enjoin other parties from infringement—particularly competitors—seriously reduces incentives to innovate, much in the same way that the DOJ’s enforcement policies in the 1970s prevented field of use restrictions in patent licensing. This can cause great harm to consumers, and is particularly problematic as more and more products and services come to depend on standardized technology.
The foregoing analysis leads me to the fourth premise of the “New Madison” approach, which is that a unilateral and unconditional refusal to license a valid patent should be per se legal. A refusal to license should not be a source for a competitor or customer to seek treble damages under the Sherman Act. That is because competition and consumers both benefit when inventors have full incentives to exploit their patent rights. This requires an assurance to inventors that they need not subsidize their competitors’ business models if they prefer not to do so. The Supreme Court clarified as much in Trinko, explaining that a refusal to deal is not an antitrust violation if the parties have never done business with each other, because “there is no duty to aid competitors.” A de facto compulsory licensing scheme turns this policy underlying the Sherman Act on its head.
To that end, I urge scholars and policymakers to give careful consideration to the underlying policies of the Trinko decision. The Supreme Court emphasized that its earlier Aspen Skiing decision was merely a “limited exception” to the rule that there is no duty to deal under the antitrust laws. But some, particularly some of the newer enforcement agencies abroad, may think the “exception” leaves room for a licensee to bring an antitrust suit if a patent holder terminates or refuses to renew the licensing agreement. The licensor thus could be forced to litigate for years the consequences of a business decision stemming from changed competitive dynamics or a new licensing strategy. Antitrust laws should not be used to transform an inventor’s one-time decision to offer a license to a competitor into a forever commitment that the inventor will continue licensing that competitor in perpetuity.
This past Fall, I urged all of us who care about innovation to consider “fresh thinking” about the implications of SSOs and the proper role of antitrust. So far, I have been encouraged and humbled by the positive response. To look forward to the future of standard essential patents, however, we should take a moment to look back to the wisdom of the Founding Fathers, and the vision of James Madison in particular. He understood the value of strong IP protection as a means of fueling innovation and technological progress. I submit that a “New Madison” approach to these issues may help restore the promise of patent and antitrust law, and unleash America’s full potential for innovation. We should, in the words of Madison, continue to recognize that “as encouragement to . . . ingenious discoveries,” patent rights are “too valuable to renounce,” and that we should fear not that the many are sacrificed to the few, but rather that “the few will be unnecessarily sacrificed to the many.”