Search the archive.

The UPSTO’s 2019 Patent Eligibility Guidelines Demonstrate Reform is Possible

Admittedly, the USPTO’s 2019 patent eligibility guidelines have brought back some stability and certainty back to the examination of the patent applications under the Alice-Mayo inquiry. Andrei Iancu, the USPTO Director at the time, stated that the purpose of the new guidelines was to “improve certainty and reliability” in the agency’s review of patent applications under § 101.61 Among others, one key reform in the UPSTO’s 2019 guidelines, which is reflected in PERA, is to clearly demarcate the patent eligibility inquiry under § 101 from the separate but equally important patentability requirements of novelty (§ 102), nonobviousness (§ 103), disclosure (§ 112), and utility (§§ 101 and 112). One recent empirical study concludes that the 2019 guidelines have achieved this goal.62 (This may account for the slight 0.5% increase in patent applications in 2022, after mostly declining annual patent application numbers in the U.S. since 2018.63)

Narrowing eligibility standards for inventions undermines the patent system's ability to foster collaboration and discourages public disclosure of inventions. Congress should override the judicially created requirements and take steps to ensure the test for patent eligibility is more clear and predictable. Without the changes made by the draft bill, the following problems created by the current confusion in the law of eligibility will continue.

1. The ability to protect inventions is reduced. As recognized by President Lincoln, who was an inventor, the patent system adds "the fuel of interest to the fire of genius." n9 When the government grants patents to inventors, it gives them the right to exclude others from practicing the patented inventions. By enabling inventors to exclude others from practicing their inventions, and allowing inventors to generate a return on their investment of sweat and treasure, the patent system incentivizes creativity and the generation of new inventions.

Do you agree with Judges Lourie and Newman? Does § 101 require a Congressional fix or should we let the courts continue to work things out?

I agree that the law of patent eligibility needs clarification by Congress. The current uncertainty in patent eligibility is the result of exceptions solely created by the Supreme Court. The uncertainty has existed for decades, with the scope of what is and is not patent eligible ebbing and flowing and tests for eligibility appearing and disappearing. The courts have on many occasions sought to devise a test for patent eligibility of computer implemented inventions. But even after all that time and effort, as the Federal Circuit noted, there is no “single, universal definition of ‘abstract idea.’” See Amdocs Ltd. v. Openet Telecom, 120 U.S.P.Q.2D 1527, 1531 (Fed. Cir. 2016).

IV. EVOLVING AND INCONSISTENT COURT DECISIONS CREATE UNTENABLE UNPREDICTABILTY AND UNCERTAINTY

The “Findings” section of the current draft of PERA notes “extensive confusion and a lack of consistency” in how courts are applying the judicial exceptions to § 101. I agree. This confusion and inconsistency makes it difficult to counsel clients on whether or how a given innovation can be patented in the U.S. It also makes it difficult to assess the value of patent portfolios for corporate valuations or transactions, since that depends on the validity of the constituent patents.

A related concern, framed most prominently by Andy Coan, is that courts have difficulty crafting rules that can be applied cleanly and reliably.312 Any lack of clarity, or the choice to employ standards rather than rules, risks overwhelming the courts’ dockets with litigation by parties who are confused, motivated by strategic considerations, or both.313 The Supreme Court’s jurisprudence on patentable subject matter may be an example of this failing. While the Court has set out three concise categories of exceptions to patentable inventions—laws of nature, natural phenomena, and abstract ideas314—those rubrics have defied easy explication. The Court has had to return repeatedly to each category in an effort to provide definitional rigor to them and has thus far failed to craft clear tests for what is in or out of bounds in terms of subject matter. Thus, even if the federal courts possessed the expertise for detailed decision making about patent and copyright policy —and perhaps the Federal Circuit might—they lack the capacity to do so at the necessary scale.315 Only the more political branches possess the personnel to process these problems.

Second, why are there judicially-created exceptions in the first place? If the United States Government decides not to issue patents to certain types of creations that would otherwise fall within the four statutory categories, such as diagnostic techniques, it is up to Congress to consider, debate, and decide what those exclusions are. Courts are not well-suited to consider what are fundamentally major policy issues. The Constitution binds the Judiciary to hearing cases and controversies.30 As such, courts must take cases as they come, and decisions must be limited to the facts and issues at-hand. This is not conducive to crafting broad patent policy and providing certainty to industry. In addition, decisions in one case as to whether an invention falls in or out of a judicial exception leaves future applicants, litigants, and courts to guess how the exception might apply to their inventions, which are often very different from the precedent.

This leads to the third part of the problem: The judicially-created exemptions are confusing. The guessing game of how to apply these decisions to other inventions leads to distorted legal outcomes and significant uncertainty. For example, in Athena Diagnostics v. Mayo Collaborative Services, the panel majority held that medical diagnostic methods for “diagnosing neurological disorders” are not patent eligible because they “are directed to a natural law and lack an inventive concept.”31 The panel majority reached this holding despite acknowledging that “Congress intended statutory subject matter to include anything under the sun that is made by man.”32 This is because the panel majority felt bound by the Supreme Court’s exception analysis of an entirely different invention in Mayo Collaborative Services v. Prometheus Laboratories33 (referred to as the Mayo/Alice framework by practitioners), stating: “[W]hether or not we as individual judges might agree or not that these claims only recite a natural law, the Supreme Court has effectively told us in Mayo that correlations between the presence of a biological material and a disease are laws of nature.”34 Various judges in four dissents to a denial of rehearing en banc voiced dismay over the confusing nature of the judge-made exceptions. The common theme amongst the dissents was the idea that one Supreme court decision—in this case, Mayo—could not be read so “broadly” to preclude any and all types of diagnostic methods that the Supreme Court inherently could not have considered in merely one case and controversy involving a separate patent.35 As Chief Judge Moore observed, however, the fact is that, “[s]ince Mayo, we have held every single diagnostic claim in every case before us ineligible.”36 Indeed, the Chief Judge said, “[w]e have turned Mayo into a per se rule that diagnostic kits and techniques are ineligible.”37

The Electronic Frontier Foundation (EFF) is a 501(c)(3) nonprofit corporation. 501(c)(3)s are largely prohibited from lobbying for candidates or legislation, even though a small portion of a 501(c)(3)’s resources can be used to lobby for legislation. The EFF lobbies a lot on many issues.

Among those issues, for many years, the EFF has lobbied heavily on legislation gutting patent law. This is not surprising because the EFF receives funding from Google, Facebook, and other Big Tech multinational monopolies that have no allegiance to the United States. The EFF is really just a Big Tech shill used to destroy free market competition that could damage their paymasters’ monopolies.

CHRISTOPHER COONS: Thank you, Director Kappos. Thank you to all the witnesses on this second panel. I might, at the risk of reopening a now well-traveled ground, Mr. Kappos, if you could just illustrate with an example, a claim currently ineligible that would be eligible under PERA, compared to a claim that would be ineligible under both current law and PERA. And Mr. Jones has argued that football plays and marriage proposals would be eligible under PERA. Do you agree?

DAVID KAPPOS: Yeah, starting with -- with football plays and marriage proposals, of course not. The essence of PERA is to provide in this area that an ineligible process, marriage proposal as an example, does not become eligible because you add -- put a ring on someone's finger. A football dance in the end zone does not become eligible because you had spike of football at the end of it. That's the language that that was included in the legislation that very clearly says that claim does not become eligible, If there's -- unless there's a no practical way to perform it without the use of a article of manufacture.

Another common criticism of patents based on the “rewards” view is that patents restrict innovation by discouraging the use of patented inventions. According to this argument, patent owners can “hold up” infringers by threatening injunctions and damage remedies for infringement.157 Infringers are said to be unaware of the patents and surprised by legal claims. The “hold-up” argument further assumes that infringers are locked in to the technology because of high costs of designing around the patented technology and the costs of switching to a new technology. The “hold-up” argument concludes that patent owners will take advantage of these costs to increase royalty demands in license negotiations.

The theoretical “hold-up” argument tends to ignore incentives for coordination through the market for inventions.158 Patent owners and producers have incentives to contract before any infringement occurs and before producers make complementary investments. Evidence for this comes from extensive investment and industry growth that occurs after licensing and cross-licensing of inventions.

The conflation of holdup and market power leads to three fallacies that underpin patent-holdup theory. Once the mechanics of holdup are loosened from their moorings in economic theory, it becomes possible to simultaneously claim that patent holdup is a variant of holdup as it is understood in mainstream economics and define it in ways that are inconsistent with the meaning of holdup as it is understood in mainstream economics. Patent holdup elides key assumptions of the standard theory and transforms necessary conditions for holdup into sufficient conditions for holdup. The implications are fundamental. In the established theory, firms—working together—will make structural, contractual, and behavioral adaptations in order to prevent holdup, thereby sustaining trade and investment in equilibrium. In patent-holdup theory, by contrast, firms cannot adapt and solve the problem wrought by opportunistic renegotiation of a contract, because the game begins after the R&D is completed and manufacturers invest. Adaptations to prevent holdup are ruled out by construction, and market failure is inevitable.

The conflation of holdup and market power leads to a second fallacy. Patent-holdup theory claims that the same manufacturing firms can be held up many times over, resulting in a phenomenon called royalty stacking. In point of fact, however, holdup cannot occur many times over to the same firm. A firm's quasi rents (the difference between its revenues and its short-run costs) can be extracted only once. Any attempt to extract more revenues would cause the firm to shut down. Royalty stacking, by contrast, is about the exercise of market power by multiple input suppliers to downstream firms. Although this multiplicity of input suppliers might be an inefficient organization of a market, it nonetheless can be a long-run equilibrium, unlike holdup.

At the same time that there are self-evident stelae contradicting patent-holdup theory, there is no positive evidence in support of its core predictions. Damien Geradin and Miguel Rato,24 Damien Geradin, Anne Layne-Farrar and Jorge Padilla,25 Vincenzo Denicolò, Damien Geradin, Anne Layne-Farrar, and Jorge Padilla,26 Richard Epstein, F. Scott Kieff, and Daniel Spulber,27 Kirti Gupta,28 Anne Layne-Farrar,29 J. Gregory Sidak,30 and Edward Egan and David Teece31 review the literature on patent holdup, patent thickets, and royalty stacking.32 All of these studies reach the same general conclusion, which is perhaps best summed up by Layne-Farrar: “Certainly the theories have been developed, but the empirical support is still lacking. Despite the fifteen years that proponents of the theories have had to amass evidence, the empirical studies conducted thus far have not shown that holdup or royalty stacking is a common problem in practice.”33

According to patent-holdup theory, excessive patent royalties cause slow rates of innovation. As an empirical matter, the aggregate royalties paid by licensees in any industry can be estimated on the basis of the SEC 10-K and 40-F filings of the patent-licensing firms.21 The data on the canonical case of mobile phones shows that the cumulative royalty yield from the twenty-one largest patent licensors in the mobile phone value chain was only 3.3 percent of a mobile phone's average sales price in 2015. That ratio has been fairly stable since at least far back as 2007.22 Researchers have parameterized royalty-stacking models from the patent-holdup literature using actual price and quantity data, and have discovered that the royalty yield predicted by the models is more than twenty times higher than the actual royalty yield and about four-fifths of the price of a smartphone. They have also found that no individual patent licensor earns an individual royalty consistent with the hypothesis that it operated as a monopolist.23

The facts of fast and continuous innovation in the mobile phone industry—one of the stelae of patent-holdup theory—are evident to anyone with a smartphone in their pocket. Three decades ago, a mobile phone cost the current equivalent of $10,000, was the size of a brick, weighed a kilo, and enabled its user to make a half-hour call before going dead. Today, a smartphone has more computational power than the supercomputers that guided the Apollo missions to the moon, allows a user to produce and share data, video, and audio files with anyone on the planet, costs an average of $300—and also happens to make a phone call.

Like the theory of the peaceful Maya, patent-holdup theory had its own set of facts—stelae, as it were—that contradicted the theory. Patent-holdup theorists asserted that innovation in SEP-intensive IT products was under threat: excessive royalties were discouraging new firm entry and reinvestment by existing firms. They called particular attention to the threat to innovation in mobile telephones and personal computers, as well as in extensions of those products in the IoT.15

Economists measure rates of innovation by examining relative rates of change of quality-adjusted prices,16 and one can download the publicly available, product-by-product, quality-adjusted price data compiled by the Bureau of Labor Statistics in order to carry out an analysis of innovation rates across products and within products over time.17 An analysis of that data shows that from 1997 to 2013, rates of innovation in phone equipment (which includes low-tech items such as fax machines and landline phones, as well as wireless phones) was 10 percent per annum faster than the economy-wide average. The data show that the rate of innovation in portable and laptop computers was faster still—31 percent per annum faster than the economy-wide average. Similar rates of innovation are observed in other SEP-intensive IT products such as video equipment, audio equipment, desktop computers, and televisions. Furthermore, rates of innovation in SEP-intensive IT products have not slowed over time relative to the rates of innovation in similar, non-SEP-intensive IT products.18 For example, the rate of innovation in SEP-intensive laptop computers compared with non-SEP-intensive mainframe computers shows that SEP intensity was associated with faster innovation.19

Another common critique of patents that is based largely on the “rewards” view is that IP rights block incremental inventions.141 This concern is applied to both individual patents and to sets of complementary patents, referred to as patent hold-ups, patent thickets, patent logjams, or the anticommons. Little if any evidence is advanced for these policy concerns.

However, the history of patents demonstrates that the patent system has accommodated a continual stream of both major inventions and incremental inventions. As Adam Mossoff observes, “From the sewing machine to automobiles to airplanes to radios, incremental innovation is omnipresent in the historical evolution of science and technology.”142 George Selgin and John Turner explain how James Watt’s steam engine patent did not prevent subsequent innovations, but “may actually have hastened the development of the high-pressure steam engine by inspiring Richard Trevithick to revive a supposedly obsolete technology so as to invent around them.”143

SEP = Standard Essential Patents –they govern a wide range of technology (like USB ports, for example). FRAND = Fair, Reasonable And Non Discriminatory. It’s a licensing principle applied to SEPs – if a company commits to developing a standard for particular tech, they can also commit to licensing the patent under FRAND

Both the biotechnology and mobile phone industries are commonly held up as paradigm–examples of anticommons that produce underuse. Heller and Eisenberg’s original article, for example, took biotechnology as its case study.83 Similarly, “[a]n influential literature claims that standard setting in the smartphone industry creates monopoly power” that produces underuse.84 This Section addresses both examples directly and shows that they do not support Heller and Eisenberg’s anticommons thesis.

Experience does not bear out Heller and Eisenberg’s anticommons thesis in the patent context. In many situations where multiple patents covering many products in many contexts, there are extraordinary rates of innovation despite the fact that barriers supposedly exist.13 This should give immediate pause to anyone trying to understand whether or not there is a policy issue of the kind Heller and Eisenberg suggest with respect to the socalled “anticommons.” As discussed below, one can seriously question whether the “anticommons tragedy” arises all that frequently. Every day, many firms face the challenge of assembling thousands of inputs to make complex systems. The existence of products that require licenses from multiple patent owners supposedly leading to what Heller and Eisenberg characterize as a “tragedy of the anticommons” does not, in practice, seem to lead to serious problems.

Empirical work appears to confirm this commonsense intuition. For example, one study by Walsh, Arora, and Cohen did not show support the anticommons tragedy thesis for biotech.14 Instead, they found that patents only posed a relatively small number of obstacles.15 Further, the “solutions” to these obstacles included licensing, doing the research beyond the reach of patents, and outright infringement.16 In addition to the Walsh et al. study, Fiona Murray and Scott Stern (2007) see little evidence of harmful effects created by patent thickets.17 Their test examines knowledge difference rates associated with the publication of a patent.18 Anticommons theory predicts a drop–off in citations to the research once patents are granted; they find at most only a modest drop off.19

Although conventional wisdom and classical economics define patents as monopolies by which the incentive to invent is balanced against restraints on access and higher short-term prices, this is a fundamental misunderstanding of the nature and function of patents. Patents and other intellectual property rights, such as copyright and trademark, are not merely incentives to create, but also incentives to commercialize innovation. They are property rights. Thus, they represent an equal opportunity for any person who creates a new invention to secure the fruits of their labors, just like any person who works as a farmer or worker should have secured to them the fruits of their productive labors. Thus, patents, like all property rights, are the basis for commercialization activities, such as obtaining venture capital financing, entering into license deals, and creating new commercial structures for efficiently placing new products and services into the hands of consumers, such as the franchise business model invented by U.S. patent owners in the nineteenth century.

In the healthcare market, for example, this has meant an ever-increasing supply of cuttingedge medical treatments and increasing availability of older medical treatments that are now “off patent.” Patents not only function for companies to recoup billions in investments and thousands of labor hours in creating new drugs and other healthcare innovations, they facilitate extensive licensing and information-sharing agreements that efficiently distribute these healthcare innovations to patients. These extensive manufacturing, commercial distribution, and information-sharing agreements were the launch pad for the unprecedented response by the biopharmaceutical sector in inventing, producing, and distributing billions of doses of the COVID-19 vaccines during the pandemic—an achievement never before accomplished by the biopharmaceutical sector since the invention and patent for Aspirin in 1900 and the invention of vaccines in the 18th century.1 Although drug prices are a subject of policy debate, it is important to recognize that 95% of the essential medicines identified by the World Health Organization are in the public domain; thus, these drugs are available for production by any generic company wishing to sell them in the healthcare market in any country in the world, subject to regulatory approval by health officials.2

F. Patent Quality is Best Addressed by Improvements to the Disclosure Requirements

Some have supported the current judicially imposed limits on patent eligibility as a remedy for concerns they have about attempts to enforce poor quality patents. We understand and share those concerns. The risk from poor quality patents has been greatly reduced by judicial decisions on obviousness, patent remedies, and fee shifting. Even more helpful have been the USPTO post-issuance validity review processes provided for in the America Invents Act, which IBM strongly supported.

On the other hand, the pro-Alice narrative that it is enabling quick and cheap invalidations of abusive patent troll cases is not fully borne out either. Although PAE and non-individual NPE patents are somewhat more likely to fail under Alice than practicing entity patents, the difference is not statistically significant. A large fraction of practicing entity patents in the software/IT industry are invalidated too. The biggest effect is on a group that has not received as much attention in the Alice debates: individual inventors and inventor-started companies, both of which fare quite poorly under Alice.94 That could be a good thing (if we think those suits are on weak patents or are overclaiming) or a bad thing (if it reflects something other than the merits). Certainly, some individual inventors have felt that they are not faring well at the patent office and some in Congress seem to be listening. In December 2019, a new House bill (the Inventor Rights Act) was introduced with the intent of strengthening individual inventor patent rights.95

1. Legislative Proposals Prevent Amalgamating Patentability Requirements

Many of the proposals' discretionary limitations are focused on isolating patentable subject matter considerations from other patentability issues. For example, PERA, the AIPLA-IPO Proposal, and the ABA Proposal all explicitly prevent a court from considering any requirement which stems from another section of the Patent Act.207 Thus, proposals for a reformed §101 all focus on reestablishing §101 as a distinct requirement apart from novelty, nonobviousness, or other patentability requirements. This would allow inventions embodying practical applications of "abstract ideas" under the Alice/Mayo framework to proceed to the merits of other patentability considerations, rather than amalgamate all patentability requirements under §101 at the motion to dismiss stage.208