What’s happening with proposals for a WTO waiver of COVID-related IP?

By Nicholson Price, Rachel Sachs, Jacob S. Sherkow, and Lisa Larrimore Ouellette

If COVID-19 were a pandemic movie, we’d be very close to the end since we’ve identified several excellent vaccines; the conventional biomedical innovation narrative often ends with the product being fully developed. But we’ve still got a long way to go with COVID-19, and the biggest challenge is getting the vaccines to billions more people (and getting them to take the vaccines, but that’s a separate topic). Only 0.3% of global doses have been administered in low-income countries, many of which are confronting severe outbreaks. A staggering 1 million infections were reported in Africa in just one month, with few vaccines in sight. By some estimates, much of the world’s population won’t be vaccinated until well into 2023. Amid pledges of sharing vaccines, perhaps the most prominent policy debate today is about waiving intellectual property rights to COVID-19 technologies, including vaccines. In this post we explain what’s being proposed, what’s happening with the waiver negotiations, and what impact these negotiations might have.

What does TRIPS require at baseline? 

Negotiations over the intellectual property rights supporting COVID-19 technologies are taking place against the backdrop of the World Trade Organization (WTO) Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS). The TRIPS Agreement creates obligations for WTO member states (with exceptions for least-developed countries). For example, member states must provide inventors with the opportunity to obtain patents that last for 20 years from the date of filing, and must provide for patent protection for pharmaceutical products, rather than processes alone.

Only member states can enforce TRIPS obligations. Members who wish to initiate dispute resolution proceedings against other countries they believe are violating TRIPS may do so through the WTO’s dispute resolution mechanism, with possible trade sanctions as a remedy. But since December 2019, the WTO’s Appellate Body—charged with hearing appeals and resolving inconsistent settlement panel rulings—has lacked a quorum. As a result, the WTO’s ability to resolve disputes and engage in enforcement is now limited, as dispute resolutions are only binding if neither party appeals or the appeal is resolved. The Trump Administration spent years blocking any potential nominees to the appeals panel, and thus far, the Biden Administration has done the same.

TRIPS Article 31 and 31bis do permit member states to engage in compulsory licensing, permitting either “use by the government or other third parties authorized by the government,” as long as certain conditions are met, including “adequate remuneration” paid to the rightholder. Originally, TRIPS only permitted compulsory licensing for domestic production—and although many countries did have the ability to produce their own pharmaceuticals, not all did. In the COVID-19 context, for instance, most countries currently lack facilities with capacity to produce COVID-19 vaccines. Africa, for example, currently only has three existing facilities with vaccine substance manufacturing capacity, one each in Egypt, Tunisia, and Senegal, though others are planned. None currently make mRNA-LNP vaccines.

In 2005, an amendment to the TRIPS Agreement was approved (though it was first agreed to in 2003 and not formally ratified until 2017) that would enable member states to produce a pharmaceutical product subject to a compulsory license solely for purposes of export to another eligible member state. In theory, this provision would allow a country with current capacity to manufacture COVID-19 vaccines to issue a compulsory license for the production and export of the vaccines to developing countries. But the procedures established by the amendment were complicated, and the amendment has only been used once—in 2007, when Rwanda used it to import HIV medications manufactured in Canada.

To be sure, formal WTO sanctions are not the only deterrent to compulsory licenses. Countries who engage in compulsory licensing also may worry about unilateral action outside WTO procedures—such as pressure exerted by the United States through its Special 301 Report—as well as retaliation from companies who are the subject of those licenses. In 2007, after Thailand issued compulsory licenses on a number of HIV-AIDS medications, Abbott Laboratories—the manufacturer of one of those drugs—retaliated by declaring that it had “elected not to introduce new medicines” into the country. 

How would the IP waiver proposal work and how are negotiations progressing?

In October 2020, two WTO members, India and South Africa, proposed a waiver of certain provisions of the TRIPS Agreement, namely, Sections 1 (copyrights), 4 (industrial designs), 5 (patents), and 7 (trade secrets) of Part II of TRIPS. The stated aim of the waiver was to “meet the growing supply-demand gap” of COVID-19 related medical products through the “unhindered global sharing of technology and know-how in order that rapid responses for the handling of COVID-19 can be put in place on a real time basis.” That is, by removing TRIPS requirements for domestic IP enforcement, affected countries could—according to the proposal—more easily manufacture COVID-19 related medical products. The basis for such a proposal was stated as stemming in part from “several reports about intellectual property rights hindering or potentially hindering timely provisioning of affordable medical products to the patients,” namely, two newspaper articles (one from Bloomberg, the other from the Louisville Courier Journal) about N95 mask shortages early in the pandemic. (Both articles focused on the N95 mask shortage early in the pandemic and whether the Defense Production Act could have been invoked to increase their supply. We previously wrote an explainer about the Defense Production Act and the mask shortage.) 

The biggest surprise in subsequent IP waiver discussions, perhaps, is the support from the U.S. Trade Representative, the agency responsible for negotiating trade agreements. On May 5, the USTR announced the Biden Administration’s support for waiver of WTO rules for vaccine-related IP and the U.S.’s participation in “text-based negotiations” at the WTO. India and South Africa’s broader IP waiver proposal, as of May 18, had 62 co-sponsors, and an amended version of the proposal was submitted to the WTO on May 25. 

India and South Africa’s waiver, as currently proposed, is expansive. It is not limited to vaccines, but includes any “health product technologies” used for the “prevention, treatment or containment of COVID-19.” The waiver would last for 3 years, although termination would not be automatic but instead require an agreement of the General Council, i.e., WTO consensus to end the waiver period. (This would make ending the waiver just as difficult as enacting it.) Mechanically, the waiver would not automatically end covered IP rights in any given country; rather, it would allow member states to engage in certain IP practices—diminishing of trade secrets, issuing compulsory licenses—or change their domestic IP laws without the threat of being haled into a WTO dispute. The extent to which countries would change their domestic laws remains unclear. But countries could not use WTO enforcement mechanisms to curtail such practices.

Implementing a WTO waiver is easier said than done. WTO decision-making normally involves the consensus of all 164 members. Currently, consensus seems unlikely given the opposition by Germany (home to CureVac and BioNTech), South Korea, Switzerland, Japan, the UK, and Australia, among others. Theoretically, if consensus cannot be reached, WTO rules permit waiver proposals based on a vote by three-fourths of all members (i.e., 123 countries) if floated by the WTO’s Ministerial Conference, but this is exceptionally rare: “[I]n practice, almost no matter has been decided by voting.” And even if the voting option were pursued, as of this writing, the number of WTO members supporting IP waiver is only about half of the votes needed. 

Negotiations also take time. TRIPS Council members have been negotiating the waiver proposal in June and July, with plans to report to the WTO General Council July 27-28—but the informal negotiations paused without obvious progress as the negotiators recessed until September. It seems unlikely that negotiations will be concluded before the next Ministerial Conference (November 30 to December 3 in Geneva). And even if some version of IP waiver makes it through the WTO in December, national-level implementation would take additional time.

What impact would a WTO IP waiver have?

The importance of the waiver proposal has been contested by academics. Over 180 IP faculty signed a letter in support of the waiver, stating that “existing provisions within the TRIPS Agreement are not sufficient in a pandemic context, whereby global access to vaccines produced at speed and scale is in all our interests.” If the costs of negotiating licenses with separate IP rightsholders or seeking compulsory licenses is deterring firms with manufacturing capacity, an IP waiver could reduce these problems (assuming national action to subsequently alter IP rights). On the other hand, a different group of IP faculty have argued that “IP rights might so far have played an enabling and facilitating rather than hindering role in overcoming Covid-19, and that the global community might not be better off by waiving IP rights, neither during nor after the pandemic.” 

As we have explained, for COVID-19 vaccines, a waiver of patent rights alone would likely have little impact. Jorge Contreras has argued that “the main benefit of the waiver could be in the area of non-patent IP,” such as a country attempting to “mandate that foreign companies operating in the country disclose their proprietary manufacturing, storage, and testing information to local producers under a compulsory license.” To be clear, waiving trade secret laws related to COVID-19 vaccines would not automatically lead to sharing of manufacturing information, training of personnel, or access to scarce raw materials. These are significant barriers, as detailed by Christopher Rowland in his Washington Post account of Pfizer’s manufacturing troubles. Nor is it clear how a mandatory disclosure obligation would work against a recalcitrant manufacturer; there are few, if any, guides. As Contreras notes, mandatory trade secret transfer “would be unprecedented in the international arena.”

An IP waiver is not the only international proposal for expanding access to COVID-related technologies. The G20 has floated a voluntary IP licensing scheme. The EU has issued its own WTO proposal that focuses on limiting export restrictions, expanding domestic manufacturing production, and operationalizing the sharing of manufacturing information and licensing. Wealthy countries like the United States can also donate excess supply—and help purchase additional doses for the rest of the world (though some barriers exist). As we stated in August of last year: “[T]he economic and public health impact of COVID-19 counsels that whatever risk exists from payers overpaying for a patent-protected COVID-19 vaccine is a mere rounding error compared to the enormous harm—both economic and human—from the pandemic. If Pfizer stands to reap $2 billion to end a scourge that has already claimed 700,000 lives and stands to cost $82 trillion globally, it’ll be worth it.” Since then, the number of lives claimed has increased nearly six-fold, and vaccine distribution has become an international tragedy. Expanding global access to effective COVID-19 vaccines should be an urgent priority for policymakers in the United States and abroad—no matter which policies end up helping us get there.

This post is part of a series on COVID-19 innovation law and policy. Author order is rotated with each post.

New Free Patent Casebook by Masur & Ouellette

As previously announced, I have been writing a new free patent law casebook with Prof. Jonathan Masur (Chicago Law), and we're very grateful to everyone who has provided helpful comments on the beta edition over the past year.

We are excited to be releasing the 1st edition. You can download a free PDF or purchase an at-cost color printed copy through Amazon here: https://www.patentcasebook.org/

We've tried to achieve a number of goals with this project beyond simply lowering the cost of course materials:

• The casebook is heavily problem-focused, including problems that can be used for out-of-class assignments, in-class multiple-choice polls, and small-group activities. Some problems are drawn from real cases, while others are hypotheticals constructed to isolate and explain difficult concepts.
• A key goal was bringing conceptual clarity to the details of modern patent practice, which means that cases are very heavily edited, some topics are presented through means other than case excerpts, and we have many diagrams and notes to help explain the points that are most likely to trip students up.
• Finally, we have worked to place patent law in its social context, including by surfacing issues of race and gender and exploring the connections between patent law and inequality among innovators and inequality in access to innovations.

For adopting instructors, we have an accompanying slide deck (including many multiple-choice questions and figures from the patents at issue in excerpted cases) and a teacher's manual with answers to practice problems and suggestions for Q&A with students about cases. We are also happy to share a Word version of the casebook.

The casebook is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. We are likely willing to authorize many derivative uses; please contact us to discuss.

If you have questions, suggestions, or interest in adopting the casebook, please let us know at ouellette@law.stanford.edu and jmasur@uchicago.edu.

We're delighted that patent law instructors will have multiple free course materials to choose from for the coming academic year. For those looking for free or low-cost options across different areas of IP, James Grimmelmann maintains a helpful compilation here.

Rob Merges Guest Post: Who Gives a Hoot About Minerva? The Patent Act and the Common Law of Patents

Guest Post by Rob Merges, UC Berkeley

In the immediate, practical sense, the Minerva opinion registers like the mildest tremor on the landscape of patent law. With a few tweaks of the standard patent assignment agreement, and putting aside the potential that the Federal Circuit will bollix up the follow-through, the opinion changed very little.

But, sometimes, a ripple on the surface denotes more dramatic movement in the deep crust. So it may be with this prosaic little case of assignor estoppel. Justice Barrett’s dissent signals a potentially radical reappraisal of the many common law rules that supplement, permeate and modify the body of operational U.S. patent law. If the signals are portents, then many settled doctrines of patent law – and other fields of IP law as well – have been quietly but surely put into play.

Holding

The holding in Minerva Surgical, Inc. v. Hologic, Inc., No. 20-440, 2021 WL 2653265 (U.S. June 29, 2021) is that assignor estoppel survives. It has, one might argue, a slightly slimmer profile than in the strongest form of the doctrine, which, borrowing from “warranty of title” principles in real property, is that an assignor is estopped from attacking any patent on an assigned invention. In Minerva, the Court returned the doctrine to its equitable roots. From now, only some assignments give rise to estoppel: only those that include an implicit or explicit representation that the inventor believes the assigned (and claimed) invention to be valid. As put by Justice Kagan: “The doctrine applies when, but only when, the assignor's claim of invalidity contradicts explicit or implicit representations he made in assigning the patent.” Minerva Surgical, Inc. v. Hologic, Inc., No. 20-440, 2021 WL 2653265, at *3 (U.S. June 29, 2021). The idea is that, to be estopped from making a statement (“this patent is invalid”), one must be on record as having made a prior, contradictory, statement (“I believe this patent to be valid”). A simple assignment – “I assign my rights to you” – is presumably not enough to show such a representation. Something more is needed now.

It remains to be seen what kind of representations are good enough. The lower courts will need to work out what constitutes an “implicit” representation of validity. Perhaps signing a conventional inventor’s oath in additional to a standard, simple assignment will be enough. The oath is required by the PTO at the time a patent application is filed, and it usually includes a statement that the signing party believes they are an inventor or co-inventor of a claimed invention. Perhaps some supplemental language in the standard assignment form might become conventional instead: something along the lines of “my invention may include embodiments, and be covered by claim language, not present in any draft specification made as of this date [of signing], and I acknowledge that these embodiments and claim variants are included in what I mean by ‘my invention’.”

However practitioners absorb and apply the majority holding, little will probably change. Because of its specific facts, the biggest ramification is that Minerva could call attention to the common practice of third party acquisition – and later amending – of “open applications”, pending patents used as the basis of later-filed, broader claims, often in an attempt to capture later-developed embodiments. (These broadening amendments are likely to run into validity challenges under the written description requirement. See Merges, Software and Patent Scope: A Report from the Middle Innings, 85 Tex. L. Rev. Vol. 1627, 1652-54 (2007) (describing a “misappropriation [of third party inventions] by amendment” rationale for written description in broadening-amendment cases). Putting this aside, Minerva is an unremarkable thumb-screw adjustment to patent doctrine, in keeping with many recent Supreme Court cases. See, e.g., Kimble v. Marvel Entertainment, LLC, 576 U.S. 446 (2015) (re-affirming prohibition on post-patent-expiration royalties, with some “safe harbor” practices advocated to soften the impact of the doctrine); Stanford University v. Roche Molecular Systems, Inc., 563 U.S. 776 (2011) (re-affirming assignment priority rules in cases of conflicting patent assignments).

Dissent

On the other hand, if the approach announced by the dissent takes hold, or is extended to other doctrinal areas, the resulting legal changes could be felt on a massive scale. It might invite open season on a host of never-codified common law rules long understood to apply in the patent context.

The evidence of this potential movement is in the three-Justice dissent from the still-new pen of Justice Barrett. This dissent applies the ascendant principles of interpretive textualism in a way that challenges reams of Supreme Court precedent on how statutory patent law fits into its common law backdrop. The dissent emphasizes that the Patent Act never mentions assignor estoppel. “Not one word in the patent statutes supports assignor estoppel, and the majority does not claim otherwise. ‘[T]his Court [doesn't] usually read into statutes words that aren't there,” Romag Fasteners, Inc. v. Fossil Group, Inc., 140 S.Ct. 1492, 1495 (2020), but that is just what the majority has done in this case.” Minerva Surgical, Inc. v. Hologic, Inc., No. 20-440, 2021 WL 2653265, at *11 (U.S. June 29, 2021) (Barrett, J., dissenting). Justice Alito agrees with this aspect of the Barrett dissent. “[T]he majority . . .  adopts a text-blind method of statutory interpretation with which I cannot possibly agree.” Minerva Surgical, Inc. v. Hologic, Inc., No. 20-440, 2021 WL 2653265, at *11 (U.S. June 29, 2021) (Alito, J., dissenting).

And again: “The Patent Act of 1952 sets forth a comprehensive scheme for the creation and protection of patent rights. But it nowhere mentions the equitable doctrine of assignor estoppel, which precludes inventors who file patent applications from later saying that the patent is invalid. To the contrary, where the Act does address invalidity defenses, it states that invalidity “shall” be a defense “in any action involving the validity or infringement of a patent.” 35 U.S.C. § 282(b). The text includes no exception for actions in which the inventor is the defendant.” Minerva Surgical, Inc. v. Hologic, Inc., No. 20-440, 2021 WL 2653265, at *13 (U.S. June 29, 2021).

Justice Barrett’s opinion also narrowly cabins traditional rules about Congressional ratification of prior cases, and traditional understandings of the perpetuation of well-recognized common law rules. The dissent hints that there are broad, preclusive effects that flow from the text of the Patent Act. If aggressively pursued, this tack could result in attacks on doctrines such as patent exhaustion, prohibitions on post-expiration patent royalties, and a host of other common law rules that routinely and conventionally augment the text of the Patent Act. 

This is not only ahistorical. It is counterproductive. Because many common law principles apply to patent transactions – assignments, licenses, mortgages, and the like – this restrictive approach undermines the private law function of patents. The uses of patents in business arrangements and organizations are often overlooked in discussions of patents and patent policy. (I review the history of the “private law of patents” at length in my [plug alert] forthcoming book, American Patent Law: A Business and Economic History, 1790-Present (Cambridge Univ. Press, 2021).) But for many inventors and companies, patents are important assets in the raising of money and the execution of business strategies. It is ironic, perhaps, that Supreme Court Justices labeled “conservative” would threaten to undermine private law arrangements having antecedents back to the 1790s. But here we are.

The Barrett dissent also shows an indifference to practical consequences that is – regrettable. And quite inconsistent with the very long history of patent jurisprudence at the Supreme Court. By tradition we would expect the big C Court to at least engage in the big policy questions at stake here: Would the abrogation of assignor estoppel lead to more, and more effective, patent challenges? Would the unsettling of patent assignments – the opening of a new level of invalidity risk faced by patent assignees – be good or bad overall? Is it fair for an inventor to challenge a patent on which she is listed as a contributor?

To her credit, Justice Kagan does engage with this tradition. She says, for example, that: “Lear [v. Adkins] counseled careful attention to the equities at stake in discrete patent contexts—and expressly distinguished assignor from licensee estoppel.” Minerva Surgical, Inc. v. Hologic, Inc., No. 20-440, 2021 WL 2653265, at *8 (U.S. June 29, 2021). And in footnote four of the opinion Justice Kagan addresses the crux of the policy conflict that informs the Minerva decision:

“Even beyond promoting fairness, assignor estoppel furthers some patent policy goals. Assignors are especially likely infringers because of their knowledge of the relevant technology. By preventing them from raising an invalidity defense in an infringement suit, the doctrine gives assignees confidence in the value of what they have purchased. That raises the price of patent assignments, and in turn may encourage invention.”

Id., at *9, fn 4.

Justice Barrett would ignore these issues. She would have it that the interweaving of common law principles can occur only under narrow circumstances. The starting point, the main focus, is on whether the text of the Act codifies a legal rule. Where the text does not codify a rule, that same text must be closely parsed to see if it is at all inconsistent with the proposed rule. Often, Justice Barrett implies, it will be.

Traditional practice is stood on its head here. Justice Taft in Westinghouse Elec. & Mfg. Co. v. Formica Insulation Co., 266 U.S. 342 (1924), asked whether the statute explicitly precluded assignor estoppel. (This just ahead of Taft’s terrific discussion of the application of common law bona fide purchaser rules in the patent context – another classic case of patent law supplementation.) Justice Barrett asks instead whether the 1952 Patent Act implicitly “incorporates” the estoppel doctrine. Minerva, supra, at *18 (U.S. June 29, 2021) (Barrett, J., dissenting). Barrett is quick to conclude that a rule not spelled out in the text of the Patent Act will often be precluded by general language in the Act. In Minerva, the general statement in 35 U.S.C. § 282 that one accused of infringement can argue invalidity becomes in the hands of Justice Barrett a rule precluding common law principles that might limit validity challenges. Under this logic, rules not mentioned in the Act are treated as a “gloss” on the text of the Act: a bad thing, unless ratified by Congress or very deeply grounded in the past. (The 1000 year history of estoppel by deed meets the test; the “shaky”, merely hundred year history, of assignor estoppel does not.)

Conclusion

The merit of the majority opinion is that for the most part it left well enough alone. It did not question the underlying fairness of assignor estoppel. And it fell back on, rather than sought to minimize, the long history of the doctrine as an adjunct to statutory patent law. Justice Kagan summarized the doctrine this way:

“By saying one thing and then saying another, the assignor wants to profit doubly—by gaining both the price of assigning the patent and the continued right to use the invention it covers. That course of conduct by the assignor strikes us, as it has struck courts for many a year, as unfair dealing—enough to outweigh any loss to the public from leaving an invalidity defense to someone other than the assignor.”

Minerva, at *9 (U.S. June 29, 2021) (footnote omitted). 

As Justice Kagan put it: “Assignor estoppel was by 1952 . . . a background principle of patent adjudication, and Congress gave no indication of wanting to terminate it or disturb its development. Nor has Congress done so since that time.” Minerva Surgical, Inc. v. Hologic, Inc., No. 20-440, 2021 WL 2653265, at *7 (U.S. June 29, 2021) (Kagan, J.).

Whatever one thinks of Supreme Court stewardship of patents over the years, the Court has always at least tried to steer the system toward its founding goal of encouraging invention while limiting the economic drag of patents. The radical dimension of the dissent is in its disregard for all this: its insistence that the key questions, and the only answers to be legitimately sought, are ones of textual meaning. Whatever its appeal or justification in other fields, this approach is simply inconsistent with the very long history of U.S. patent law. The text of the Patent Act is, and is rightfully, the essential focal point of many patent decisions. But common law rules of all shapes and sorts are an essential part of patent law. The common law backdrop helps facilitate patent-based transactions, and private ordering in general. The best readings of the Patent Act never lose sight of this.