Part I - Reconsidering Key Theoretical and Policy Issues

1 New and Heightened Public–Private Quid Pro Quos Leveraging Public Support to Enhance Private Technical Disclosure

Peter Lee

Biopharmaceutical companies developed safe and effective COVID-19 vaccines in record time, thus providing hope in a devastating pandemic. While these vaccines have saved countless lives, global inequality in access to vaccines, particularly between developed and developing countries, has been highly controversial. While numerous factors contribute to such inequality, intellectual property (IP) rights have attracted significant attention. Biopharmaceutical companies hold patents on COVID-19 vaccines, and critics have argued that exclusive rights have constrained access to these lifesaving resources. Accordingly, developing countries, public health advocates, and even the US government argued for a recently enacted waiver of international IP rules with the aim of enhancing global manufacturing and distribution of patented COVID-19 vaccines.¹

Not surprisingly, biopharmaceutical patentees opposed this IP waiver. Among their objections, they argued that weakening patents would do little to promote widespread production of COVID-19 vaccines. They asserted that even if third parties were not constrained by patents, they would still lack critical technical knowledge for manufacturing vaccines in industrial quantities. In particular, third parties would lack two overlapping categories of technical knowledge held by vaccine developers: tacit knowledge, which constitutes personal, experiential knowledge that is not amenable to codification, and trade secrets, which constitute codified and uncodified technical knowledge that firms deliberately keep secret.

This state of affairs not only jeopardizes global access to COVID-19 vaccines, but also reveals a troubling paradox at the heart of the patent system. The patent system represents a quid pro quo in which inventors receive exclusive rights in exchange for disclosing a novel invention. Biopharmaceutical patentees, which enjoy exclusive rights over COVID-19 vaccines, have ostensibly disclosed their technologies. Yet these same patentees argue that third parties cannot manufacture these patented vaccines in the absence of privately held tacit knowledge and trade secrets. This chapter examines the causes and implications of that paradox and proposes several ways to resolve it.

The chapter explores several mechanisms to compel greater technical disclosure by patentees and other beneficiaries of public innovation support. It first focuses on modifying the patent quid pro quo to increase technical disclosure by patent applicants and patentees. It proposes rehabilitating the “best mode” requirement of patentability, and it considers the possibility of extending disclosure requirements for a finite period of time after patent filing. Beyond the requirements of patentability, this chapter argues that public funding provides a valuable lever for compelling greater technical disclosure by private innovators, including many patentees. Such measures would promote greater codification of tacit knowledge and public disclosure of trade secrets related to practicing publicly funded innovations. The chapter then focuses on the unique challenges of transferring purely tacit knowledge, which is not amenable to codification. Such knowledge is best transferred through direct interactions between technology generators and adopters. Imposing an obligation of direct tacit knowledge transfer through the patent system would be overly burdensome and fall outside the patent quid pro quo. However, the chapter suggests that additional policy levers can help motivate such tacit knowledge transfer and establish infrastructure to facilitate it.

Section 1 introduces the problem of unequal access to COVID-19 vaccines and the concern that patents contribute to such inequality. It also describes the movement to temporarily waive global IP rules to enhance access to patented vaccines. It further explores the argument that weakening patents would not appreciably increase generic production of COVID-19 vaccines because third parties lack the tacit knowledge and trade secrets to manufacture them. Section 2 discusses the paradox wherein biopharmaceutical patentees have ostensibly disclosed their COVID-19 vaccine technologies, yet third parties cannot practically manufacture vaccines without private knowledge from those patentees. It explores the importance of tacit knowledge and trade secrets to the effective manufacturing of patented vaccines, particularly in industrial quantities. Section 3 explores mechanisms to increase the disclosure of private technical knowledge. It suggests reforming patent law and utilizing the lever of public funding to compel greater technical disclosure by patentees and private innovators benefitting from government support. Section 4 explores the challenges of transferring purely tacit knowledge and proposes policy measures to promote such transfer.

1 Patents and the Challenge of Global Access to COVID-19 Vaccines

The introduction of safe and effective COVID-19 vaccines in late 2020 was a crucial turning point in the pandemic. Based in large part on massive government funding, biopharmaceutical firms introduced several vaccines, including the newest generation of so-called mRNA vaccines from Moderna and Pfizer–BioNTech. While these vaccines provided enormous relief, their unequal distribution quickly generated significant concern. Disparities in access have been especially stark on the global landscape, particularly between developed and developing nations. For instance, individuals in wealthy and middle-income countries received approximately 90 percent of the first 400 million doses of COVID-19 vaccines.² As of September 2022, 72.5 percent of individuals in high-income countries had received at least one dose of a COVID-19 vaccine, but only 22.8 percent of people in low-income countries had received at least one dose.³

While numerous factors contribute to such grossly unequal access, IP rights have attracted significant attention. Although biopharmaceutical companies introduced COVID-19 vaccines in record time, they had been developing and patenting the technologies underlying those vaccines for years. Empirical research shows that private companies have filed about 70 percent (80 out of 113) of the patent families covering the newest generation of mRNA vaccines.⁴ A handful of companies – Moderna, CureVac, BioNTech, and GSK – own about half of the mRNA vaccine patent applications.⁵ Proponents of strong IP rights argue that patents were necessary to induce biopharmaceutical companies to develop COVID-19 vaccines and that they will be necessary to encourage similar innovations to combat future pandemics.⁶ However, critics contend that the exclusivity conferred by patents has constrained access to COVID-19 vaccines around the world, particularly for low-income countries.⁷ Access constraints were particularly pronounced in the first year after the introduction of vaccines, before biopharmaceutical firms ramped up supply.⁸ Even today, developing countries have limited access to the newest and most effective COVID-19 vaccines, mRNA vaccines, which are produced by Moderna and Pfizer–BioNTech.⁹ As of July 2022, 93 percent of all mRNA vaccine doses had gone to wealthy countries.¹⁰

The concern that IP rights can constrain access to vaccines (and other technologies needed to fight the pandemic) motivated calls to weaken those rights. Attempts to do so, however, faced obstacles from the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement).¹¹ The TRIPS Agreement, which is part of the framework that created the World Trade Organization (WTO), establishes minimum standards for IP protection for all WTO member states. It represents the result of “upward harmonization” that requires relatively stringent protection for IP rights – including patents – for almost all countries in the world.¹² Among other provisions, TRIPS establishes an expansive conception of patentable subject matter – which includes health technologies, such as vaccines – and imposes regulations on the granting of compulsory licenses.¹³ As such, any member state that weakens patents in derogation of TRIPS minimum requirements would violate its WTO obligations.

To mitigate this barrier, in October 2020 India and South Africa proposed a temporary waiver of various TRIPS provisions in light of the exigencies of the coronavirus pandemic.¹⁴ This so-called TRIPS waiver would temporarily suspend TRIPS requirements for IP protection for innovations related to the “prevention, containment or treatment of COVID-19.”¹⁵ To the surprise of many, in May 2021 the Biden Administration announced its support for a limited version of a waiver that would temporarily lift TRIPS obligations for most developing countries with respect to patents on COVID-19 vaccines.¹⁶ In June 2022, the WTO adopted such a limited waiver for patented COVID-19 vaccines.¹⁷

Not surprisingly, biopharmaceutical patentees opposed the TRIPS waiver. Their principal argument was that a TRIPS waiver, and a concomitant weakening of patent rights, would undermine incentives to invent, both in the present and going forward. Additionally, opponents of a TRIPS waiver argued that weakening patents would do little to achieve the waiver’s goal of increasing global manufacturing and distribution of COVID-19 vaccines. Biopharmaceutical patentees asserted that even if governments did not enforce patents, unauthorized third parties would not be able to manufacture COVID-19 vaccines without tacit knowledge and trade secrets from vaccine developers themselves.¹⁸ This argument had, for a while, particular traction coming from Moderna, which publicly pledged in October 2020 that it would not assert its vaccine patents against entities manufacturing COVID-19 vaccines during the pandemic.¹⁹ Moderna has subsequently reneged on its pledge in several ways, thus calling into question whether generic manufacturers can reasonably rely on it.²⁰ Moderna, however, continues to maintain that it will not assert its patents against manufacturers in developing countries and that its patents are not preventing generic manufacturing of its COVID-19 vaccine. However, the company opposed the TRIPS waiver, and it has refused to publicly disclose its tacit knowledge and trade secrets for manufacturing its vaccine.²¹

2 The Patent Quid Pro Quo, Tacit Knowledge, and Trade Secrets

Biopharmaceutical patentees cite the inability of third parties to manufacture patented vaccines without proprietary tacit knowledge and trade secrets as a reason to oppose the TRIPS waiver. This chapter, however, argues that this phenomenon reveals a more fundamental divergence between existing patent practice and the overarching principles of the patent system. Specifically, biopharmaceutical patentees’ nondisclosure of private knowledge necessary for manufacturing patented vaccines offends the essential bargain at the heart of the patent system.

The patent system represents a “quid pro quo” in which inventors receive twenty years of exclusive rights in exchange for disclosing a novel invention.²² In the United States, various disclosure obligations are codified in statute and elaborated in case law. In particular, a patent must: teach a person of ordinary skill in the art how to make and use an invention, provide an adequate written description of the invention, and (at least technically) disclose any best mode the inventor knows as the most effective way of practicing it.²³ Robust disclosure plays a central role in the patent system. Patent disclosures comprise an “invisible college of technology” that enriches the public storehouse of knowledge and represents one of the primary benefits of the patent system.²⁴

Among other functions, robust patent disclosure ensures that competitors are on an equal footing with patentees upon patent expiration. As the Supreme Court observed:

[U]pon the expiration of that [patent] period, the knowledge of the invention inures to the people, who are thus enabled without restriction to practice it and profit by its use. To this end the law requires such disclosure to be made in the application for patent that others skilled in the art may understand the invention and how to put it to use.²⁵

While patent expiration means that the public can practice an invention without restraint, the public gains the information to practice that invention immediately – at the time of patent grant – rather than at the end of the patent term.²⁶ Robust technical disclosure represents the consideration that inventors provide in exchange for exclusive rights. However, the controversy over access to patented COVID-19 vaccines gives rise to an unsettling paradox: if biopharmaceutical patentees have disclosed their COVID-19 vaccines in patents, why do third parties need so much private (undisclosed) knowledge to practice them?²⁷

While disclosure plays a central role in the patent system, patent disclosure is limited in several ways.²⁸ At a foundational level, patent law requires disclosure through codification, but not all technical knowledge is capable of codification. As further explored later in this section, purely tacit technical knowledge may be highly valuable for practicing a patented invention, yet it is not amenable to codification. Additionally, the patent disclosure requirements focus on enabling a basic version of an invention, which may be a far cry from a fully developed commercial product.²⁹ This emphasis on enabling a basic version of an invention serves to limit patent disclosure, particularly given that inventors tend to file patent applications as soon as possible on early-stage, embryonic inventions.³⁰

Relatedly, priority rules discourage patent applicants from adding “new matter” to their disclosures after filing.³¹ As a result, the disclosure obligation is largely “fixed” at the time of filing a patent application, which further limits patent disclosure. While inventors continue to gain important knowledge about their creations throughout patent prosecution and commercialization,³² the patent system actually disincentivizes patent applicants from disclosing such information. More broadly, patentees have strong commercial incentives to superficially comply with the requirements of patent disclosure while disclosing as little information as possible.³³ It is very difficult, moreover, for the Patent and Trademark Office (PTO) or courts to know if a patentee is retaining private knowledge about an invention that the patentee should disclose. All of these factors combine to significantly limit patent disclosure.

Of particular note is that patents do not disclose significant amounts of tacit knowledge about inventions. Because tacit knowledge plays a central role in the controversy over access to patented COVID-19 vaccines, some further elaboration is warranted. Tacit knowledge encompasses personal, experiential knowledge that is not amenable to codification.³⁴ For example, a professional tennis player could write instructions on how to serve a tennis ball, but such instructions would necessarily fail to convey tacit knowledge derived from years of training, inherent athletic skill, and even muscle memory.³⁵ In the realm of novel technologies, tacit knowledge entails “non-codified, disembodied know-how” possessed by an inventor.³⁶ It consists of “intangible knowledge, such as rules of thumb, heuristics, and other ‘tricks of the trade.’”³⁷ In the context of COVID-19 vaccines, biopharmaceutical patentees have developed tacit knowledge in the course of developing and commercializing their vaccines, and they argue that third parties cannot manufacture these vaccines in industrial quantities without it.

In describing tacit knowledge, it is useful to draw several distinctions. First, tacitness is not a binary on–off designation but a question of degree. At one end of the spectrum lies purely tacit knowledge, which is incapable of codification. At the other end of the tacitness spectrum is latent knowledge, which is technically codifiable yet not presently codified.³⁸ Second, tacit knowledge has an intrinsically dynamic character. Novel technologies often arise with a significant tacit dimension, as perhaps only the inventors themselves can truly understand them. However, as novel principles become part of the generally accepted knowledge in a field, tacitness decreases.³⁹ Third, tacit knowledge may be useful for understanding a basic invention, but it can be particularly useful for extending, modifying, and commercializing that invention.⁴⁰ The process of translating a new invention into a commercial product presents a host of technical challenges, and the tacit knowledge of the original inventor can be very helpful in overcoming them. Almost by definition, however, tacit knowledge related to a patented invention is not disclosed in the patent.

In addition to tacit knowledge, patents may fail to disclose proprietary trade secrets relevant to practicing a patented invention. A trade secret consists of technical or business information that derives economic value from secrecy and is the subject of reasonable efforts to maintain that secrecy.⁴¹ There is some overlap between tacit knowledge and trade secrets, though the two categories of information are far from coextensive. Because of its difficult-to-convey nature, tacit knowledge may satisfy the secrecy requirement to qualify for trade secret protection; indeed, firms often protect tacit knowledge as trade secrets. However, trade secrets encompass a much wider range of undisclosed information, including codified knowledge, such as instructional manuals, research and testing data, and manufacturing specifications. For example, a written vaccine “recipe” with detailed instructions to make a COVID-19 vaccine does not represent tacit knowledge, and firms are likely to protect such information as a trade secret.

Patentees routinely do not disclose tacit knowledge and trade secrets related to practicing their inventions. Of course, one of the functions of the patent system is to incentivize the codification and public disclosure of otherwise tacit knowledge.⁴² Technically speaking, however, the patent system can only stimulate the codification of latent knowledge; purely tacit knowledge is not capable of codification. Furthermore, as mentioned, patentees have significant incentives not to disclose invention-related trade secrets as long as they can appear to satisfy the disclosure requirements of patentability.

Undisclosed tacit knowledge and trade secrets, moreover, can be critical to practicing and commercializing a patented invention. In the life sciences, for example, when biotech scientists disclose a novel biologic compound in a patent, they often retain substantial tacit knowledge regarding their creation.⁴³ Patent disclosures simply cannot convey all the nuances and details of how inventors create and use complex biological macromolecules. Furthermore, while the tacit knowledge of inventors is helpful to producing a biologic compound in a laboratory setting, it is especially helpful to manufacturing such compounds in industrial quantities. According to legal scholars Nicholson Price and Arti Rai, “slight variations in the manufacturing process can change the quality, safety, or efficacy of the final product.”⁴⁴ In some cases, such knowledge ultimately becomes codified for internal purposes, in which case a biotech firm may protect it as a trade secret. Such private information may be highly valuable to practicing a patented invention, yet patents often do not disclose it.

Tacit knowledge and trade secrets play an important role in enabling the manufacture of patented COVID-19 vaccines. As noted, Moderna and Pfizer contend that even in the absence of patents, unauthorized manufacturers would be unable to produce their vaccines because the process is too complex and requires specialized facilities.⁴⁵ Academic commentators confirm this view, arguing that for “some complex COVID-19 vaccines and biological therapeutics, fast manufacturing, particularly of products originally developed by other firms, will require not only physical capacity but also access to knowledge not contained in patents or in other public disclosures.”⁴⁶ In similar fashion, vaccine expert Alain Alsalhani from Doctors Without Borders noted: “You need someone to share all the process, because it’s a new technology … One of the problems we have is that the scientific literature about industrial-scale manufacturing of mRNA vaccines is so slim. This is why it’s not just about a recipe, it’s about an active and full tech transfer.”⁴⁷ Transfer of private information – including tacit knowledge and trade secrets – is critical for the manufacture of patented COVID-19 vaccines.

3 Leveraging the Patent System and Government Funding to Increase Disclosure of Private Technical Knowledge

A Modifying the Patent Quid Pro Quo

The current state of affairs reveals an unsettling paradox: biopharmaceutical patentees have ostensibly disclosed their vaccines, yet third parties cannot practically manufacture them without private information held by patentees. This paradox, moreover, reveals a conflict between the overarching aims of patent disclosure and the current state of the doctrine. Patent disclosure seeks to put other technical artisans on cognitive footing comparable to the patentee. However, this objective is not met by biopharmaceutical patentees who have ostensibly disclosed their technologies. This in turn suggests the need to modify the existing patent quid pro quo. This chapter provides several suggestions for increasing the disclosure requirements of patentability, which would compel greater disclosure of invention-related tacit knowledge and trade secrets.

The chapter first suggests strengthening the “best mode” requirement of patentability. Under US patent law, the enablement requirement mandates that a patent must teach a technical artisan in the field how to make and use an invention.⁴⁸ As noted, this requirement is aimed at enabling a basic version of a patented invention. Technically, US patent law also requires patent applicants to disclose the best mode for practicing their inventions, which encompasses any “specific instrumentalities or techniques which are recognized by the applicant at the time of filing as the best way of carrying out the invention.”⁴⁹ The best mode requirement aims “to restrain inventors from applying for patents while at the same time concealing from the public preferred embodiments of the inventions they have in fact conceived.”⁵⁰ The requirement has both subjective and objective elements. If a patent applicant has subjective knowledge of a best mode at the time of filing a patent application, the applicant must disclose it in an objectively adequate manner.⁵¹ Historically, the best mode requirement has provided an incentive for patent applicants to disclose invention-related trade secrets.⁵² In theory, it can also promote disclosure of certain kinds of tacit knowledge (which may or may not be formally recognized as trade secrets).

While the best mode requirement plays a valuable role in compelling disclosure of private information, it has attracted criticism for unduly increasing the expense and complexity of litigation.⁵³ Accordingly, in 2011, Congress reformed the best mode requirement in a manner that renders it essentially toothless. Disclosing any known best mode is still technically a requirement of patentability, but failure to do so is no longer a permissible ground for canceling, invalidating, or rendering unenforceable a patent claim.⁵⁴ This chapter argues for restoring the best mode requirement as a fully enforceable patentability requirement. Doing so would compel patentees to disclose private knowledge (including tacit knowledge and trade secrets) concerning the best way to practice their inventions. More broadly, this change would help achieve the overarching objective of placing competitors on equal cognitive footing with patentees. Rehabilitating the best mode requirement would help mitigate the anomaly where, for instance, vaccine developers obtained patents but retained private information critical to practicing their inventions.⁵⁵

A more aggressive, and more controversial, variant of this proposal would increase the time period over which patent applicants and patentees must comply with the disclosure obligations of patentability – including a rehabilitated best mode requirement. Current patent doctrine assesses compliance with the disclosure requirements of patentability as of the date of filing a patent application.⁵⁶ However, inventors continue to gain valuable information about their inventions throughout the processes of prosecuting and ultimately commercializing their patents.⁵⁷ Indeed, it is possible that vaccine developers satisfied the enablement and best mode requirements for their vaccines as of the date of filing and developed knowledge about the best way of manufacturing COVID-19 vaccines at a later date, which would fall outside the statutory disclosure requirements. To address that condition, this chapter suggests extending the disclosure requirements for some reasonable period of time (for example, five years) after patent filing. In essence, patentees would have an ongoing requirement to update disclosure of a best mode for a period of five years after filing a patent application. Failure to do so could lead to denial of patent claims (if the patent is still in prosecution) or invalidation of patent claims (if the patent has already issued). Such an ongoing obligation of technical disclosure would provide significant incentive for patentees to disclose and update private information relevant to practicing an invention.

A more specific variant of this latter proposal would target technical disclosure for a class of regulated technologies that includes COVID-19 vaccines. Any proposal to compel a patentee’s disclosure of tacit knowledge and trade secrets will encounter difficulties of monitoring and enforcement. The PTO and courts cannot easily know what is in a patentee’s mind, and they may be unaware that the patentee has knowledge (or updated knowledge) of a best mode within five years after filing a patent application. In some cases, however, it is evident that the patentee possesses such private information because the patentee discloses it to another government entity. Vaccines, along with diagnostics and therapeutics, are somewhat unique among patented technologies because they are heavily regulated by government agencies such as the US Food and Drug Administration (FDA), the European Medicines Agency, and comparable agencies in other jurisdictions. As a condition of obtaining regulatory approval, developers of these technologies must often submit detailed manufacturing information to regulators. Such submissions can compel the codification of tacit knowledge and the disclosure of codified trade secrets.⁵⁸ Regulators may also engage in hands-on investigation of manufacturing processes; as part of granting authorization for Moderna’s COVID-19 vaccine, the FDA sent inspectors to Moderna’s production facilities and clinical trial sites.⁵⁹ The FDA ordinarily treats such submissions as confidential, thus allowing them to remain the private knowledge of submitters. However, the patent quid pro quo provides a lever for compelling patentees to disclose such information publicly. If such manufacturing knowledge exists at the time of filing a patent application (or, under the more aggressive proposal, within five years of filing), then it would fall within the mandate of the best mode requirement to disclose it.

Of course, these proposals to enhance the disclosure requirements of patentability raise several complications. Rehabilitating the best mode requirement would give rise to several objections that led Congress to weaken it in the first place. Critics contend that the best mode requirement increased the expense and complexity of litigation, especially the need to inquire into an inventor’s subjective knowledge at the time of filing a patent application.⁶⁰ Furthermore, the best mode requirement was unique to US patent law, and rehabilitating it would undermine international patent harmonization.⁶¹

There is reason to believe, however, that these criticisms are overblown. As Professors Brian Love and Chris Seaman argue, the requirement to disclose a best mode (rather than maintain it as a trade secret) could actually decrease litigation expense and complexity by reducing instances when a patentee asserted both patent infringement and misappropriation of trade secrets against a defendant.⁶² Against objections that a rehabilitated best mode requirement would undermine international harmonization and burden foreign inventors, it is important to note that US law already has a best mode requirement. Rehabilitating the best mode requirement would simply give more teeth to an existing obligation of US patent law with which all inventors (domestic and foreign) should comply.⁶³ While it would be ideal for other countries to adopt the best mode requirement, such widespread adoption would not be necessary to meaningfully increase patent disclosure. Given the lucrative nature of the US market, inventors from around the world routinely seek to patent their technologies in the United States, where they are legally obligated to disclose a best mode if they know of one.

Extending the time period for disclosure requirements would also raise several technical complications. Such extension would require changing prevailing rules and practice whereby a disclosure is largely “fixed” at the time of filing a patent application.⁶⁴ Furthermore, safeguards would have to establish that a patentee could amend a specification for the purposes of updating a best mode, but such amendments could not be the basis for expanding claims. Additionally, given that patents (and patent applications) often change hands, policymakers would have to consider how assigning a patent would affect ongoing disclosure obligations. A logical option would be for disclosure obligations to follow assignment of the patent; that is, the assignee (who is likely to take the lead in commercializing a patent) would bear the obligation of updating the best mode for a prescribed period of time.

Finally, a requirement for patentees to publicly disclose information submitted to other regulatory agencies also raises certain challenges. Existing doctrine holds that forced public disclosure of legally protected trade secrets by government agencies may constitute a taking that falls within the protections of the Fifth Amendment.⁶⁵ This suggests that public agencies can only take such information for public use, and they must provide just compensation to the trade secret holder. However, unlike an ex post taking, the proposal here envisions an ex ante agreement by a patent applicant to publicly disclose invention-related knowledge in exchange for exclusive rights. As such, these obligations would not fall within the ambit of the Fifth Amendment any more than the general disclosure requirements of patent law, which compel the disclosure of private information in exchange for a government benefit. Additionally, there is some concern that forcing public disclosure of regulatory information would discourage patentees from seeking regulatory approval for their vaccines, diagnostics, and therapeutics. However, given that regulatory approval is a necessary gateway to marketing and thus profiting from these innovations, it is unlikely that heightened disclosure requirements would significantly chill such submissions.

B Leveraging Government Funding

While modifying the patent quid pro quo can lead to greater disclosure of tacit knowledge and trade secrets related to practicing inventions, the patent system is not the only policy lever for increasing access to private technical knowledge. The federal government provides massive funding to private technology firms (many of which are also patentees), and this public funding provides leverage to insist upon greater dissemination of private information by funding recipients. In the context of patented COVID-19 vaccines, the federal government could condition massive funding for vaccine developers on commitments to disclose or share tacit knowledge and trade secrets for manufacturing them.⁶⁶

The federal government has been investing for decades in the technologies underlying COVID-19 vaccines. It has long supported research on coronaviruses,⁶⁷ and publicly funded research on vaccines for other conditions, such as HIV and MERS, contributed to developing today’s COVID-19 vaccines.⁶⁸ The federal government’s support for COVID-19 vaccines is most evident for the newest generation of mRNA vaccines.⁶⁹ Federally funded research was critical to developing several innovations at the heart of these vaccines, such as genetically engineered spike proteins⁷⁰ and techniques for modifying mRNA to allow it to evade the body’s immune system.⁷¹ Quite simply, federal funds were crucial to developing COVID-19 vaccines.

While the federal government has supported research leading to COVID-19 vaccines for decades, its most visible contributions have occurred since the outbreak of the pandemic. In late April 2020, the Trump Administration launched Operation Warp Speed, an ambitious initiative aimed at producing 300 million doses of safe and effective COVID-19 vaccine.⁷² This initiative provided about $18 billion to six vaccine developers, including Moderna, Pfizer, and Johnson & Johnson. Operation Warp Speed provided vaccine developers with several kinds of financial support, including grants to cover vaccine development, advance-purchase commitments for final doses, and, in some cases, both. In addition to financial support, Operation Warp Speed also provided logistical and operational support to expand manufacturing capacity for some grantees.⁷³ Federal support helped rapidly accelerate vaccine development. While it ordinarily takes three to nine years to move from sequencing a virus to Phase 1 clinical trials,⁷⁴ in the case of COVID-19 vaccines that time period was significantly condensed to about ten weeks.⁷⁵

These enormous public contributions to private vaccine development provide the federal government with certain claims on resulting vaccines. The prospect of leveraging public funding to enhance access to vaccines – particularly in developing countries – has attracted significant attention.⁷⁶ This chapter, however, argues that governments can leverage public funding to enhance access not only to finished vaccine doses but also to the tacit knowledge and trade secrets necessary to manufacture them. In this manner, government agencies can exploit a different kind of public quid pro quo other than the patent system to increase technical disclosure by private innovators.

Government agencies can leverage public innovation funding to essentially bargain for greater codification and disclosure of private technical knowledge. At root, this would simply be an application of traditional contract principles: the federal government provides enormous funds to contractors, and it can condition such funds on those contractors codifying tacit knowledge and conveying trade secrets for practicing subject technologies. For instance, in Operation Warp Speed, the federal government could have included a provision in multibillion-dollar contracts that required grantees to codify and publicly disclose (or privately share) best manufacturing techniques for any successful COVID-19 vaccines. These agreements would have comprised ex ante bargains in which the federal government negotiated with contractors to deliver not just some good but also the knowledge for making it, and they would be less intrinsically coercive than ex post takings. If the contractor did not want to commit to codifying and disseminating private knowledge, including tacit knowledge and trade secrets, it could decline to take federal funds.

Notably, existing federal procurement law already contemplates this kind of quid pro quo of public funds for access to private information. In general, civilian federal procurement contracts are governed by the Federal Acquisition Regulations (FAR).⁷⁷ Subject to some exceptions, the FAR provides the federal government with “unlimited rights” in data first produced under subject contracts and data delivered under subject contracts.⁷⁸ While these provisions do not compel the codification of tacit knowledge, they provide valuable access to codified trade secrets related to government contracts. The FAR defines “data” expansively to include all “recorded information.”⁷⁹ This includes “technical data,” which comprises “recorded information (regardless of the form or method of the recording) of a scientific or technical nature (including computer databases and computer software documentation).”⁸⁰ Furthermore, the FAR defines “unlimited rights” as enabling the government “to use, disclose, reproduce, prepare derivative works, distribute copies to the public, and perform publicly and display [data], in any manner and for any purpose, and to have or permit others to do so.”⁸¹ Government data rights represent a bargain in which contractors must provide access to data in exchange for federal funding.

These government data rights may provide an avenue for the federal government to widely disseminate the COVID-19 vaccine recipes arising from billions of dollars of public funds.⁸² Consistent with the FAR, the authorizing statute for the Biomedical Advanced Research and Development Authority (BARDA) requires it to condition grants to contactors on obtaining access to “all data related to or resulting from countermeasure and product advanced research and development.”⁸³ BARDA’s Operation Warp Speed contract with Moderna specifically requires Moderna to share all of its submissions to the FDA with BARDA.⁸⁴ It also allows BARDA not only to obtain all raw data produced under the contract but also to share it with outside parties, consistent with the FAR.⁸⁵ Public Citizen’s analysis of BARDA’s 2020 contract with Moderna for $483 million yielded two important insights concerning data:

First, BARDA gained access to the entire “vaccine recipe.” This includes Moderna’s dossiers containing chemistry, manufacturing, and controls information, which provide manufacturing instructions in step-by-step detail. Second, BARDA obtained “unlimited rights” to data first produced by Moderna using contract funding (“Unlimited Rights Data”).⁸⁶

Under prevailing regulations, a federal agency only obtains “limited rights in data generated prior to entering or outside the scope of the contract, and data developed at private expense.”⁸⁷ While this provision provides a slim avenue for Moderna (and other contractors) to claim that certain information can remain proprietary, the redacted nature of publicly disclosed contracts prevents knowledge of the exact scope of data over which the government only has “limited rights.” Analysis based on available sources, however, suggests that the scope of “limited rights” data is quite narrow and that the government enjoys “unlimited rights” in data concerning scaling up of the manufacture of Moderna’s vaccine and transferring production to other manufacturing sites.⁸⁸ Government officials, including Senator Elizabeth Warren, have pressed the Biden Administration to clarify what kinds of data are subject to unlimited rights by the government under the BARDA–Moderna contract.⁸⁹ From all available evidence, however, it appears that public funding and government procurement regulations provide federal agencies with broad rights to private data from vaccine developers, including information on how to manufacture COVID-19 vaccines.

4 Relational and Organizational Mechanisms to Promote Tacit Knowledge Transfer

While leveraging patent rights and public funding can compel innovators to disclose private information – including tacit knowledge and trade secrets – such measures are in some ways limited. As noted, some tacit knowledge is not amenable to codification at all. While enhanced disclosure requirements can help capture the “low hanging fruit” of latent knowledge, they cannot compel the codification of purely tacit knowledge. Furthermore, while leveraging public quid pro quos can encourage the disclosure of codified trade secrets, technology transfer through codified texts can be rather inefficient. Relatedly, heightened disclosure requirements may lead to unhelpful “data dumps” of technical information that is expensive to codify yet may not be particularly valuable to technology adopters. Accordingly, this chapter now turns to relational and organizational mechanisms for transferring technical knowledge and policy levers that can promote such transfer.

Given the difficulty of transferring purely tacit knowledge, oftentimes the only (or most efficient) way to effectuate such transfer is through direct interpersonal interaction with the inventor. As economist Joanne Oxley observes, tacit knowledge “is extremely difficult to transfer without intimate personal contact, involving teaching, demonstration, and participation.”⁹⁰ Similarly, economist David Teece has likened the transfer of tacit knowledge to an apprenticeship model in which an apprentice works directly alongside a master craftsperson.⁹¹ The value of interpersonal interaction with an inventor persists even when that inventor has ostensibly “disclosed” an invention in a patent. While reading a text is valuable, sometimes there is no substitute for directly talking with an inventor about a novel technology.

The active participation of inventive entities in technology transfer serves several valuable functions. First, direct transmission of tacit knowledge can aid a technology adopter in assimilating a new invention. Second, as noted, an inventor’s tacit knowledge may be particularly useful for extending, modifying, and commercializing an invention.⁹² In the life sciences, for example, direct interactions with inventors can greatly accelerate industrial-scale manufacturing of biologic products. As economist Gary Pisano observes, “In the absence of well-defined and well-understood scale-up recipes, ensuring product integrity requires extensive interactions between the scientists who designed a cell in the laboratory and bioprocessing engineers charged with developing the production process.”⁹³ An inventor’s participation in active technology transfer can allow a technology adopter to benefit from the inventor’s tacit knowledge to navigate unpredictable challenges on the path toward commercialization.

Given the highly personal nature of tacit knowledge transfer, relational and organizational mechanisms are critical to effectuating such transfer. Inventive entities and technology adopters use a variety of interpersonal and organizational linkages to transfer technical knowledge.⁹⁴ In some cases, firms licensing a patent will hire the inventor as a consultant, thus obtaining direct access to the inventor’s tacit knowledge. For example, firms licensing university patents routinely hire the faculty inventors of those inventions to aid in technology transfer and commercialization.⁹⁵ Organizational linkages between inventive entities and technology adopters can also promote tacit knowledge transfer. This can be accomplished through joint ventures between technology firms or research consortia, such as SEMATECH, a consortium of US semiconductor firms that facilitates “cooperative research, development, and testing projects.”⁹⁶ At the far end of the spectrum, the goal of transferring tacit knowledge between two entities can even motivate them to integrate, becoming a single organization. This is evident, for example, in the vertical integration of small biotech firms that produce novel biologic compounds and large pharmaceutical companies that develop those biologics into marketable drugs.⁹⁷ Such vertical integration accelerates tacit knowledge transfer by bringing technology generators and adopters under the same organizational roof.

Tellingly, relational and organizational mechanisms play an important role in transferring tacit knowledge and trade secrets for manufacturing COVID-19 vaccines. Moderna and Pfizer emphasize the difficulty of transferring their technology to cast doubt on the efficacy of the TRIPS waiver. However, Moderna and Pfizer have actively transferred their technology to foreign entities, thus demonstrating the feasibility of such transfer. They have employed relational and organizational mechanisms that facilitate a high degree of interaction between technology generators and adopters. Put differently, they have transferred technology within what I have called “bounded entities” – organizational constructs such as fully integrated firms, joint ventures, and even “thick” contractual relationships between long-term partners.⁹⁸ Such bounded entities facilitate the intensive communications and interpersonal interactions that are necessary to transfer purely tacit knowledge and that accelerate the transmission of virtually all technical knowledge.

At one end of the spectrum, Moderna is “transferring” its vaccine technology internationally within its own corporate boundaries by establishing manufacturing sites in Kenya, Australia, and Canada.⁹⁹ Such “in-house” transfer illustrates the principle that it is easier to transfer tacit knowledge within an organization rather than between two separate ones. Additionally, both Moderna and Pfizer have transferred vaccine technology internationally by utilizing a different kind of “bounded entity”: “thick” contractual engagements with long-term partners that facilitate repeated interactions. For instance, Moderna entered into a ten-year “strategic collaboration agreement” with Swiss chemicals and biotechnology company Lonza to manufacture Moderna’s COVID-19 vaccine.¹⁰⁰ Far from a one-off engagement, this long-term agreement provides for active technology transfer from Moderna to Lonza.¹⁰¹

Pfizer has agreements with over twenty contract manufacturing organizations around the world that provide for intensive technology transfer.¹⁰² Again, these are not one-off interactions in spot markets. While Pfizer’s technology transfer engagements usually last up to three years, in the case of its COVID-19 vaccines, it accelerated that time frame to between five and eighteen months. As Pfizer describes it:

For the COVID-19 vaccine, the team at the external facility would need to be trained on many aspects of this complex manufacturing process – from learning the intricacies of formulating lipid nanoparticles that encapsulate the mRNA and sterilizing the product to make it safe for injection to filling it into vials, labeling the vials, packaging them, and distributing them around the world.¹⁰³

Such “thick,” intensive interactions over long periods of time accelerate technical knowledge transfer, especially the transfer of tacit knowledge.¹⁰⁴

What are the implications of this relational model of tacit knowledge transfer for the patent quid pro quo? One possibility would be to mandate that patentees directly work with technology adopters to transfer tacit knowledge and other technical information to satisfy the enablement and best mode requirements. For a variety of reasons, however, this chapter argues against such a proposal. Patent law currently contemplates disclosure of technical information through codification, and requiring other forms of information sharing, such as interpersonal interactions, would constitute a major paradigm shift. It would, of course, be impossible for a patentee to individually transmit tacit knowledge to the same universe of entities that can read a patent; while codified text is nonrival and nonexcludable, interpersonal interactions are rivalrous and excludable. The personnel demands of transferring tacit knowledge to all technology adopters would be onerous and would require redirection of technical staff from other responsibilities. Furthermore, the PTO and courts are ill-equipped to assess the sufficiency of interpersonal tacit knowledge transfer by patentees. Ultimately, this chapter does not argue for requiring patentees to engage in interpersonal tacit knowledge transfer. The patent system, however, is only one of several policy levers available to encourage such transfer.

This chapter argues that governments can play an important role in facilitating this relational model of technology transfer. First, governments can condition public funding on commitments by grantees to pursue relational modes of tacit knowledge transfer. This represents another example of a public–private quid pro quo. Government agencies can leverage public funds to encourage grantees not only to codify tacit knowledge (and disclose codified trade secrets) but also to actively transfer tacit knowledge to technology adopters. To maintain feasibility, this would not entail a broad obligation to work with all parties that wished to adopt some government-funded technology – such an obligation would place enormous burdens on government grantees. Rather, this approach would entail individually negotiated agreements by which government grantees would commit to transferring tacit knowledge to select downstream manufacturers (which they could approve) as a condition of receiving public funds.¹⁰⁵

Such agreements could have significantly enhanced tacit knowledge transfer from Operation Warp Speed. For instance, federal agencies could have conditioned funds on grantees agreeing to actively transfer resulting technical knowledge to a predetermined and mutually agreeable list of vaccine manufacturers. Such commitments to actively transfer technology could have been a condition of receiving research and development funds and/or advance purchase commitments, which were worth billions of dollars to vaccine developers. Had federal agencies adopted this approach, technology transfer agreements with less than a dozen sites around the world would have greatly accelerated global production of vaccines (particularly mRNA vaccines) at a critical time.¹⁰⁶

Public funds may be necessary not only to incentivize relational tacit knowledge transfer, but also to enable it. Building relational and organizational links to transfer tacit knowledge is costly. As such, if a federal agency negotiates for grantees to transfer technology through consulting engagements, demonstrations, and on-site problem solving, that agency may have to fund such activities.¹⁰⁷ Additionally, government entities can support relational technology transfer in other ways as well. For instance, State Department officials can assist with visas allowing for the travel of key technical personnel, as they did in Operation Warp Speed.¹⁰⁸ Furthermore, public funds may be valuable not only to encourage innovators to “push” tacit knowledge to technology adopters, but also to increase the “absorptive capacity” of those technology adopters.¹⁰⁹ Investments in equipment, training, and even hiring personnel can greatly assist foreign entities seeking to absorb technology from the United States and other countries. For instance, Operation Warp Speed could have devoted funds to enhance the absorptive capacity of vaccine manufacturing facilities around the world to help ramp up vaccine production.

Second, beyond providing funding, governments and international organizations can actively build knowledge-sharing infrastructure to accelerate tacit knowledge transfer. For example, in May 2020 the World Health Organization (WHO) and its partners established the COVID-19 Technology Access Pool (C-TAP). This initiative created a resource for sharing “intellectual property, knowledge and data” concerning innovations for fighting the pandemic.¹¹⁰ Importantly, C-TAP is more than just a passive repository of information. One of the implementing institutions within C-TAP, the Tech Access Partnership, “facilitates connections between experienced manufacturers and local manufacturers in developing countries to share key data, knowledge and other relevant support though a coordinated network.”¹¹¹ While illustrating the power of public institutions to facilitate tacit knowledge transfer, unfortunately no major biopharmaceutical firms have yet to participate in C-TAP.¹¹²

Public institutions, however, have had more success with establishing a technology transfer hub for mRNA vaccines in South Africa.¹¹³ The WHO, a South African consortium, and partners from COVAX have established a hub in which “[f]oreign manufacturers will share techniques with local institutions and WHO and partners will bring in production know-how, quality control and will assist with the necessary licenses.”¹¹⁴ South African researchers at this technology transfer hub recently recreated a prototype of Moderna’s mRNA vaccine (without any assistance from Moderna).¹¹⁵ The tech transfer hub is now transferring mRNA vaccine technology to six African nations.¹¹⁶ While technology transfer and development would have proceeded much faster with the participation of Moderna or Pfizer, this success illustrates the potential for public infrastructure to catalyze tacit knowledge sharing.

5 Conclusion

Controversy over access to patented COVID-19 vaccines has revealed a significant technical challenge to ramping up global production of these essential resources. While patient populations lack access to vaccine doses themselves, third-party manufactures lack access to the information and knowledge to manufacture them, particularly for the newest generation of mRNA vaccines. This state of affairs is paradoxical given that biopharmaceutical patentees have ostensibly disclosed their vaccines as part of the patent quid pro quo. This chapter has explored various causes and implications of this paradox. In a variety of ways, patent disclosure is often limited. In particular, patentees routinely do not disclose tacit knowledge – personal, experiential knowledge that is not amenable to codification – and trade secrets related to their inventions. As a result, patentees often retain valuable private information about their inventions, such as COVID-19 vaccines, while also enjoying patent exclusivity.

To ameliorate this situation, the chapter suggests new and heightened public–private quid pro quos to increase technical disclosure by private innovators. It first argues that patentees’ retention of private knowledge necessary to practice patented inventions offends the social bargain at the heart of the patent system. Accordingly, it suggests increasing patent law’s disclosure obligations by rehabilitating the best mode requirement. This requirement – which already exists but is rarely enforced – compels patentees to disclose private knowledge about the best way to practice a patented invention. This chapter has also raised the possibility of extending the time period for disclosure obligations beyond the date of filing to capture additional technical information gained by patentees. In the case of patented vaccines and other health products, this approach may also compel patentees to disclose manufacturing information submitted to regulatory agencies. More broadly, federal funding represents a powerful lever for enhancing access to private technical information. In the context of Operation Warp Speed, government agencies could have conditioned billions of dollars on commitments by grantees to disclose latent knowledge and codified trade secrets.

Such obligations, however, are limited to the extent that purely tacit knowledge is not amenable to codification. Rather, such knowledge is best transferred through relational and organizational linkages between technology inventors and adopters. This chapter argues against requiring patentees to engage in interpersonal tacit knowledge transfer as part of the patent quid pro quo. However, it argues that public institutions can encourage such activity through leveraging public funds and establishing infrastructure to catalyze technical knowledge transfer.

2 Global Medical War Chest

Lawrence O. Gostin

China’s release of genomic sequencing data for SARS-CoV-2 on January 11, 2020 was the start of a race to develop vaccines and other medical countermeasures for COVID-19. The Coalition for Epidemic Preparedness Innovations (CEPI) quickly contacted its partners that were developing novel platforms that could be adapted to new pathogens and offered financial support to direct their efforts to SARS-CoV-2.¹ Less than ten weeks later, on March 16, Moderna’s vaccine candidate entered a phase 1 clinical trial, and on July 27 it entered phase 3.² The vaccine candidate – a novel messenger RNA (mRNA) vaccine – was based on a similar vaccine Moderna had been developing for MERS. Biotech companies from around the globe also launched clinical trials of SARS-CoV-2 vaccine candidates. China and Russia began deploying their vaccines for selected populations even before Phase 3 trials were completed. By December 2, the United Kingdom granted approval for another mRNA vaccine manufactured by Pfizer/BioNTech, with US approval for both mRNA vaccines following closely behind.³ By December 29, the UK approved yet another vaccine developed by AstraZeneca/Oxford. In just one year, scientists from around the world developed safe and effective vaccines against COVID-19, a triumph that is unprecedented in human history.

How was it possible to develop COVID-19 vaccines at “pandemic” speed? Scientific understanding and technologies for genomics and structural biology have exploded. High-income countries and biotech companies also had clear incentives to invest substantially in COVID-19 vaccine development, given the global spread of the virus and its massive health, economic, and social impacts.

The COVID-19 pandemic overwhelmed the world’s capacity to respond effectively. Non-therapeutic interventions (such as masks, distancing, and stay-at-home orders) failed to keep the virus under control in most countries. Vaccines became the only way to return to a semblance of normalcy, with schools and businesses fully open, people socializing, and travel resuming.

Previous Ebola outbreaks demonstrated the difference that a timely vaccine can make in confronting a deadly threat – and the factors that contribute to costly vaccine development delays.

The Ebola epidemic in the Democratic Republic of the Congo (DRC) began on August 1, 2018, ending two years later on June 25, 2020. The epidemic was particularly challenging because it took place in an active conflict zone.⁴ Were it not for the deployment of a promising vaccine, rVSV-ZEBOV, the death toll would have been far greater.⁵ The vaccine was administered in rings to high-risk individuals who were geographically or socially connected to patients: contacts, contacts of contacts, and frontline responders. The vaccine proved effective, building on evidence gathered from Guinea in 2015. In both outbreaks, researchers found that vaccinated individuals did not contract the Ebola virus.⁶

The rVSV-ZEBOV vaccine was not new. The Public Health Agency of Canada applied for a patent on the vaccine in 2003.⁷ But the first human clinical trial didn’t begin until over a decade later when the West African Ebola crisis finally spurred action. NewLink Genetics commenced a phase 1 clinical trial in 2014, supported by the National Institutes of Health (NIH) and US Department of Defense.⁸ NewLink then licensed vaccine rights to Merck Pharmaceuticals, which pushed the vaccine into additional clinical trials. In late 2019, the vaccine would ultimately become the first Ebola vaccine to be approved by regulators in Europe and by the US Food and Drug Administration (FDA).⁹

Why did such a promising vaccine with the potential to save thousands of lives languish for years? The intellectual property (IP) system does not generally incentivize companies to produce vaccines or medicines intended for small or uncertain markets. With vaccines costing millions of dollars to bring to the market, pharmaceutical companies, largely located in high-income countries, hesitated to invest in vaccine candidates intended primarily for low-income countries or sporadic outbreaks of diseases.¹⁰

The devastating West African Ebola epidemic galvanized political will and funding for Ebola research and development (R&D), which more than tripled from 2014 to 2015.¹¹ While the US government provided the majority of resources for Ebola and other viral hemorrhagic fevers, funding from private industry increased seven-fold.¹² Merck donated vaccine doses, while the Vaccine Alliance (Gavi) provided $1 million for operational costs.¹³ The DRC, supported by the World Health Organization (WHO), Médecins Sans Frontières (MSF), and the United Nations Children’s Fund (UNICEF), implemented ring vaccinations.¹⁴ Governments, international organizations, public–private partnerships, and nongovernmental organizations (NGOs) offered additional support.¹⁵

Vaccines and medicines are essential components of our medical war chest. The Ebola vaccine development highlights how promising technologies can languish due to lack of funding and political attention. But it also demonstrates how we can overcome market disincentives through targeted financing and partnerships to harness varied expertise. The COVID-19 pandemic was a game changer, demonstrating scientific prowess beyond what would have seemed possible.

This chapter explores the gap between technology’s promise and our ability to realize the global public goods of vaccines and medicines. This gap stems from significant market disincentives in the R&D process, along with clinical trial challenges and regulatory hurdles. Yet a range of innovative financing strategies to delink R&D costs from vaccine and drug prices, along with well-designed and ethically run clinical trials, can fill this gap, facilitating development of urgently needed medical countermeasures.

1 Why Outbreak Disease Research Can Falter and Fail

In 2018, the WHO identified eight priority diseases (including Ebola virus disease, SARS, and Zika) that warranted accelerated R&D given their epidemic potential, and the absence of efficacious medical countermeasures.¹⁶ In many cases, development of countermeasures for these diseases has stalled at the clinical testing phase.¹⁷

Given that billions of dollars are spent on pharmaceutical research each year, what explains the lack of medical countermeasures for outbreak diseases? At its core, it’s a question of market incentives: R&D is expensive and the market for technologies to fight infectious diseases with epidemic potential is often small and uncertain. Even with a predictable market of consumers in high-income countries, the process of bringing a new health product to market is expensive and time-consuming. Including out-of-pocket expenses and opportunity costs, bringing a new drug to market in the United States costs roughly $2.6 billion by one calculation, and takes over a decade.¹⁸ And successful regulatory approval is not guaranteed. Most new drugs that enter clinical trials fail; the rate of successful FDA approval is estimated at only 11–14 percent.¹⁹

For outbreak disease countermeasures, the lack of a clearly defined market exacerbates the costs and risks. Outbreaks, by definition, are sporadic and unpredictable. It is nearly impossible to know precisely which pathogens will cause health emergencies, or when. And even identifying high-risk pathogens is only a first step as pathogens mutate over time. Pharmaceutical companies also cannot predict the number of doses needed to contain an outbreak, which often is quite small. And low- and middle-income countries are often hit hardest by novel pathogens. Consequently, industry lacks financial incentives to develop products for many novel diseases.

Each stage in the medical countermeasure development process highlights the deficiency of existing market incentives to stimulate innovations. Most drug and vaccine development originates with basic scientific research, usually conducted by government-funded academic researchers.²⁰ Traditionally, governments fund research for diseases that cause the most illness and death in their populations, such as cancer, diabetes, or cardiovascular disease. This is beginning to change. The United States, for example, has classified Ebola as a bioterrorism threat and “category A priority pathogen.” But the NIH has not prioritized other novel pathogens, and thus extant pipelines reflect fewer drug candidates.²¹

The second step in the process is product development, where researchers translate basic scientific findings into a drug or vaccine candidate. This resource- and time-intensive process involves identifying a candidate, optimizing it to lessen unintended interactions, and testing it for toxicity. If successful in animal models, the drug or vaccine undergoes three phases of clinical trials, using increasing numbers of patients, to gather evidence about the candidate’s safety and efficacy. Clinical trials are challenging even under the best conditions, but when the study timeline must be compressed to fit within an outbreak, then lengthy regulatory review, dosage shortages, low participation, and poor trial design can cripple a trial’s chance of success.²²

In some cases, a medical countermeasure can gain emergency use authorization (EUA) before completing the entire clinical trial process. For example, in May 2020, the FDA authorized the emergency use of the drug remdesivir to treat COVID-19 patients hospitalized with severe respiratory symptoms.²³

The FDA’s decision followed two clinical trials demonstrating efficacy in COVID-19 recovery. The pharmaceutical company Gilead originally developed remdesivir to treat Ebola, but it had been found ineffective for that purpose.

Emergency use authorization can be controversial precisely because phase 3 clinical trials are not always completed. For example, on August 23, the FDA issued an EUA for investigational convalescent plasma for the treatment of COVID-19 in hospitalized patients.²⁴ The EUA coincided with warnings from senior National Institute of Health officials that the data were not sufficient for such authorization.²⁵

If a product successfully completes clinical trials, its sponsor must obtain regulatory approval from each country in which the product will be marketed and distributed. Countries usually have their own national regulatory agencies, and the rigor of agency reviews is highly variable. For countermeasures primarily used in lower-income countries, drug sponsors must navigate regulatory regimes that often lack capacity to conduct robust and independent reviews. Thus, sponsors often also seek approval from a more stringent regulatory agency, such as those in the United States, the European Union, or Japan, and prequalification from the WHO, which is frequently viewed by low-income countries as a prerequisite for national approval.

Finally, after receiving regulatory approval, the product must be manufactured at scale and introduced into the country, raising a new set of logistical and operational challenges. Manufacturing capabilities in lower-income countries are often weak. Successful product introduction requires effective distribution channels, ensuring sufficient supply and procurement. Health providers must be made aware of the new product, and, if necessary, trained in its use.

2 Overcoming R&D Obstacles

Scientific understanding and technologies for developing medical countermeasures have exploded in recent years. Advancements in genetics allow for early pathogen sequencing, helping identify the proteins needed for vaccine design. Two technologies have especially propelled vaccine development: synthetic vaccinology and platform technologies. In synthetic vaccinology, sequencing data are digitally communicated – from scientist to scientist and country to country – without the need for transferring biological samples.

With platform technologies, instead of “one bug, one drug,” the idea is to support development of multiple vaccines or drugs with one or more components.²⁶ As demonstrated by COVID-19 vaccine development, platform technologies can speed up and simplify vaccine production, enabling the rapid development of “plug and play” vaccines where multiple different vaccines can be developed using the same system. Platform technologies can also reduce the need for cold storage, easing the burdens of stockpiling and distribution.

While the science behind developing countermeasures has progressed rapidly, the incentives and regulatory process for translating scientific research into safe and effective products with market approval have lagged behind. The COVID-19 pandemic has compelled investment and regulatory systems for governing R&D to “catch up” to the science, with reforms to overcome many of the barriers in the R&D process – yet challenges remain.

A Coordinating Research and Development

The 2014–16 West African Ebola epidemic highlighted major gaps in R&D coordination between stakeholders: confusion and poor organization led to delays in affected countries receiving funds, equipment, medical countermeasures, and personnel.²⁷ In the wake of the outbreak, global commissions examined the international response and identified the importance of cooperation and coordination.

Coordination of stakeholders’ R&D activities helps ensure efficient use of scarce resources, prioritize work on the most worrying diseases and most promising technologies, reduce duplicative activities, and marshal technical expertise. With COVID-19 vaccine development proceeding at record pace, coordination has proven critical on other grounds: both ensuring sufficient global supply of the safest and most efficacious COVID-19 vaccines and equitably distributing those vaccines to the world’s population.

In April 2020, WHO launched the Access to COVID-19 Tools (ACT) Accelerator, supported by public and private actors including the European Commission, the Bill and Melinda Gates Foundation, CEPI, Gavi, Global Fund, UNITAID, and the Wellcome Trust.²⁸ The ACT Accelerator aims to align global efforts for equitable access to new COVID-19 diagnostics, therapeutics, and vaccines. The WHO, given its technical expertise and unique legitimacy for global leadership, has developed global policy recommendations on the use of COVID-19 vaccines though its Strategic Advisory Group of Experts (SAGE) on Immunization.

Yet equitable and timely access to COVID-19 vaccines simply could not be achieved without reaching the scale of vaccine production necessary to meet vast global need. For manufacturers, expanding production for vaccines that are still being developed, and may never get approved, comes with enormous financial risk. Many countries and groups of countries, including Canada, China, the United States, and the European Union, addressed this challenge through bilateral agreements with vaccine manufacturers to meet their own COVID-19 needs – a troubling development which coined the term “vaccine nationalism” – the hording of vaccines by governments to meet their populations’ needs.²⁹

On May 15, 2020, the Trump Administration launched Operation Warp Speed, which had the goal of delivering 300 million COVID-19 vaccine doses across the United States, with initial supplies by January 2021.³⁰ A key element was to manufacture vaccines at industrial scale while they were simultaneously undergoing clinical trials. Normally, vaccines are not manufactured at scale until after regulatory approval. As only 11–14 percent of countermeasures typically are approved, waiting for approval before large-scale manufacture reduces the possibility of wasted investments.

However, in Operation Warp Speed, the US government took on the financial risk, allowing vaccine developers to expedite production of the most promising vaccine candidates. In July 2020, the US government announced a $2.1 billion deal with vaccine makers Sanofi and GlaxoSmithKline to develop their COVID-19 vaccine candidate and produce 100 million doses by 2021 – bringing the list of US-supported COVID-19 vaccine candidates to six.³¹ Operation Warp Speed’s overall budget was $10 billion, with $6.5 billion directed toward countermeasure development, and the remainder for NIH research.

Other high-income countries, along with some middle-income ones, have undertaken similar efforts to rapidly develop and produce COVID-19 vaccines. France, Italy, Germany, and the Netherlands formed the Inclusive Vaccine Alliance, with a deal to purchase 400 million doses of AstraZeneca’s vaccine for EU member states.³² The United Kingdom signed its own deals with AstraZeneca and other companies as well. All told, high-income countries had agreements covering 4 billion doses as 2020 drew to a close, while middle-income countries had agreements for close to 3 billion doses, which threatened access of low- and many middle-income countries.³³ Meanwhile, China and Russia were developing their own COVID-19 vaccines, with promises to share them widely, including widespread interest in Latin America, Asia, and the Middle East,³⁴ The Serum Institute of India produced 50 million doses of the AstraZeneca vaccine even before the United Kingdom and India granted regulatory approval.

Beyond the enormous financial risks governments take by investing in the large-scale production of still unapproved vaccines, this siloed, country-by-country approach is concerning for two major reasons. First, competition among vaccine candidates could lead countries to “panic buy” – creating a bidding war and driving up prices on the global market. Second, given that many vaccine candidates will fail, access to successful candidates may become limited to the few privileged countries that selected them. Lower-income countries, lacking the ability to take financial risks and enter advance supply agreements, would be left especially defenseless.

To avoid this dire outcome, the COVID-19 Vaccine Global Access (COVAX) Facility was assembled by CEPI and Gavi, working under the WHO’s leadership.³⁵ Founded in 2017, CEPI’s mission is to engage governments and the commercial sector to improve public health preparedness – which proved critical during COVID-19. Gavi was founded in 2000 to expand childhood vaccine campaigns in the world’s poorest countries but has since expanded its charge to procure COVID-19 vaccines globally. The COVAX Facility is a financing and procurement mechanism that pools countries’ demand and resources for COVID-19 vaccines. By inviting all countries to join, COVAX takes a collective approach, allowing for a much larger portfolio of vaccine candidates than countries can achieve on their own, thus reducing the risk that countries will fail to secure access to successful vaccines. For manufacturers, it reduces the risks that major production investments could result in no or low demand.

Costing an estimated $18.1 billion, COVAX aimed to deliver 2 billion doses of safe and effective COVID-19 vaccines, ones that have passed regulatory approval or WHO prequalification, by the end of 2021.³⁶ Vaccines were to be distributed to all participating countries equally, proportional to population size, initially prioritizing health-care workers and then expanding to cover the most vulnerable 20 percent of every participating country. As of December 2020, 190 countries had agreed to participate in COVAX – though not the United States or Russia. Wealthier nations would self-finance their vaccines from their own public budgets, and partner with ninety-two low- and lower-middle income countries supported through voluntary donations.

Global public–private partnerships, such as CEPI and Gavi, have been highly effective at bringing rapid funding and expertise to thorny health problems such as COVID-19. They have demonstrated the ability to bring governments together to increase the purchasing power of lower-income countries, while also benefitting higher-income countries that can expand their portfolio of vaccine candidates during an outbreak. But still, uncertainty remains whether COVAX will be able to entice higher-income countries to make the political and financial commitment for COVAX to succeed. In rich countries, national leaders poured resources into their own vaccine development and purchasing, taking care of their own citizens first – while also taking up large portions of the global manufacturing capacity, jeopardizing the possibility of low-income nations having access to COVID-19 vaccines.

But the reality must not be ignored: without investments in equitable vaccine distribution, diseases will continue to circulate among poor, unvaccinated populations. Viruses could mutate, gaining virulence and transmissibility, and spread throughout the world, infecting even vaccinated populations. Strategies for financing the development and distribution of outbreak countermeasures are discussed below.

B Financing Innovative R&D

Whether R&D is coordinated by the WHO, governments, partnerships, or private entities, its acceleration requires sustainable funding. Traditionally, governments have offered most R&D funding, supplemented by philanthropy and private industry. In 2015, governments funded 63 percent of neglected disease R&D. Philanthropies, mainly the Gates Foundation and the Wellcome Trust, contributed 21 percent of global funding, and industry contributed the remaining 17 percent.³⁷ In 2015, industry increased Ebola R&D sevenfold over the previous year, becoming the second largest funder of Ebola R&D and other African viral hemorrhagic fevers, behind only the NIH.³⁸

In 2016, the Commission on a Global Health Risk Framework for the Future recommended a global commitment of an incremental increase of $1 billion per year to accelerate R&D of drugs, vaccines, personal protective equipment, and medical devices.³⁹ Yet much more is needed. Developing COVID-19 countermeasures has demanded multibillion-dollar investments in R&D, and has stimulated the use of financing mechanisms to secure funding.

Funding and Financing Mechanisms

Funds can be used to stimulate R&D and overcome market failures, but the question remains how best to channel resources to maximize results. “Delinking” R&D costs from the price of health products is an important concept. The current IP system encourages companies to recoup development costs through high prices, which can make drugs and vaccines unaffordable. The IP system also protects companies from competition by providing patents, exclusive licenses, and regulatory exclusivities – driving up the costs of drugs and vaccines even further. This system can cause two major problems: first, companies may simply avoid developing products needed in uncertain or unlucrative markets; and second, high prices can put essential medicines out of reach of the world’s poorest.

By delinking R&D costs from product prices, financing mechanisms aim to encourage innovation without prohibitively expensive prices. Delinkage includes mechanisms to offset development costs through upfront payments or back-end rewards, or through the strategic softening of intellectual property protections. Governments and partnerships relied on these mechanisms to stimulate the development of COVID-19 vaccines, and ultimately ensured their affordability.

Offsetting R&D Costs: Push and Pull Mechanisms

Financing to offset development costs can either provide funding upfront (called a “push”) or offer a financial reward once a product has been developed (called a “pull”).⁴⁰

Push mechanisms include up-front payments such as grants and innovation funds and are highly effective in spurring R&D for neglected diseases. In 2016, for example, the United States Agency for International Development (USAID) issued a challenge called, “Combating Zika and Future Threats.” Parties from academia and industry submitted proposals to combat Zika, and USAID awarded $15 million in grants for projects that ranged from insecticide-treated sandals to mosquitos infected with bacteria that prevent the spread of disease to humans.⁴¹

Although push financing mechanisms are popular, funding is not contingent upon the recipient successfully furthering R&D. “Pull” mechanisms, alternatively, reward successful innovation with financing. Pull mechanisms come in a variety of forms, including prizes, transferable vouchers, and advance market commitments. For example, the FDA Priority Review Voucher program awards a transferable voucher to the sponsor of a new drug to treat delineated tropical diseases, such as Ebola, Zika, and Lassa Fever. The holder of the voucher can redeem it for expedited review of a new drug or can sell it to another company. For example, a voucher awarded to Knight Therapeutics in 2014 for its leishmaniasis drug was sold to Gilead Sciences for $125 million. Gilead used the voucher for accelerated review of its HIV drug Odefsey.⁴² The race for the COVID-19 vaccine was financed through both push and pull mechanisms. As part of Operation Warp Speed in the United States, the Biomedical Advanced Research and Medical Authority (BARDA) had a budget of $6.5 billion to “push” the development, manufacture, and distribution of promising COVID-19 vaccine candidates.⁴³ BARDA formed agreements to provide millions and even billions to vaccine companies (including Pfizer, AstraZeneca, Moderna, Johnson & Johnson, Novavax, Sanofi, and GlaxoSmithKline), to cover the costs of testing, commercialization, and manufacture. BARDA also partnered with manufacturers and producers of supplies such as vials, syringes, and plastic containers. By securing the financing up-front, companies can focus on rapid vaccine development without incurring much financial risk.

Similar to BARDA, CEPI’s role in the COVAX Facility was to push the development of COVID-19 vaccines by signing contracts with developers of promising candidates to help fund research, clinical trials, and manufacturing capacity.⁴⁴ CEPI initiated partnerships with nine vaccine companies to finance the development of their COVID-19 vaccine candidates.

Concurrent to CEPI’s push mechanisms within COVAX, Gavi employed a pull mechanism: advance market commitments (AMC).⁴⁵ The goal of AMCs is to counter market forces that push down drug prices by ensuring a stable market for a product once it is developed. Many factors can drive down drug prices, including the absence of a lucrative market in low-income countries, and the practice of “bulk purchasing” vaccines. Under bulk purchases, health ministries purchase most vaccines, using their buying power to negotiate lower prices. From the country’s perspective, this is a wise use of limited resources, but by driving vaccine company profits even lower, bulk purchasing can disincentivize vaccine development targeted at low-income countries.⁴⁶ With AMCs, a purchaser promises vaccine developers that it will buy a certain number of doses, at a predetermined price, if the developers can clear the necessary regulatory hurdles.

In 2010, Gavi successfully piloted an AMC for pneumococcal vaccines.⁴⁷ Donors committed $1.5 billion to the World Bank to guarantee the price of pneumococcal vaccines once developed, and suppliers agreed to provide vaccine doses at a predetermined price. In return, each manufacturer received its proportional share of the committed funds.⁴⁸ This financing program has been credited with an estimated 225 million children vaccinated against the virus across sixty low- and low–middle-income countries within the decade after the program was launched.⁴⁹

Within COVAX, Gavi worked with participating countries and donors to assemble the funding to enter into AMCs with CEPI’s COVID-19 vaccine development partners.⁵⁰ As with the pneumococcal vaccine program, the AMCs sought to help ensure that the funding of vaccines for low-income countries remains stable, thus incentivizing their development and manufacture. Gavi estimated that $2 billion was needed to provide 1 billion doses to ninety-two low-income countries through AMCs. By the end of October 2020, Gavi had successfully raised these funds from participating countries (including Italy, the United Kingdom, Canada, and Norway), as well as the private sector, yet as 2020 drew to a close, the COVAX AMC still needed nearly $5 billion, while more than $2 billion was still required for country readiness and late-stage clinical trials.⁵¹

As demonstrated by COVAX, push and pull mechanisms can work in concert: initial financing agreements reduce the risks of product development, while purchase agreements contingent on product approval help assure a stable market in low-income countries.

Decreasing R&D Costs: More Flexible IP Protections

The push and pull mechanisms discussed above all involved funding to offset R&D costs, lessening the need for companies to set high drug or vaccine prices to recoup costs, while leaving IP protections intact. But problems remain. Successful COVID-19 vaccines developed in rich countries, outside of COVAX’s AMCs or other agreement, could be unaffordable for poorer populations in all other countries. Reducing IP protections in defined scenarios, such as patent pools, licensing agreements, and open-source approaches to R&D, is thus essential to global access.

Patents increase drug prices by providing companies an exclusive right to use, make, or sell their inventions for a defined period of time, blocking competition for the length of the patent period. And without competition, the patent holder can charge high prices even for essential drugs.

Patent pools aim to lower IP barriers to competition and affordable pricing. The Medicines Patent Pool is an existing United Nations-backed entity that pools patents for HIV, hepatitis C, and tuberculosis medicines. During the COVID-19 pandemic, the Pool’s scope was expanded to cover COVID-related medical products. Under the Pool’s model, patent holders voluntarily license patented products to the Pool, which then sublicenses them to generic manufacturers. The voluntary licenses are generally restricted geographically to low- and middle-income countries so that the patent holder retains exclusive rights in lucrative markets. Royalties (albeit low) are paid to patent holders upon sale of the resulting products.

In May 2020, the WHO and Costa Rica launched the Solidarity Call to Action as a complement to the WHO’s ACT Accelerator.⁵² The Solidarity Call asked countries to ensure that all COVID-19 publicly funded and donor-funded research outcomes are made affordable, accessible, and available on a global scale through provisions in funding agreements (for example, non-exclusive voluntary licensing), as well as national legal and policy measures to lower barriers such as intellectual property rights. In an “open source” approach, the Solidarity Call encourages research outcomes to be published with no restrictions, and collaborative efforts are taken in pre-competitive drug discovery. By December 2020, about forty countries had endorsed the Solidarity Call, but notably the United States, China, the United Kingdom, and other countries with high vaccine development capacity had not.⁵³

Not surprisingly, the pharmaceutical industry strongly opposed the Solidarity Call.⁵⁴ Companies that are investing billions in developing COVID-19 countermeasures could lose incentive to innovate if they perceive that their intellectual property rights could be jeopardized. Industry has similarly opposed countries’ efforts to lay the legal groundwork for issuing compulsory licensing agreements for COVID-19 countermeasures if necessary. Under compulsory licensing, which the 2001 Doha Declaration allows under extraordinary circumstances to protect public health, governments can grant a license to a public agency or generic drug maker to copy a patented medicine without the patent owners’ consent.⁵⁵

To avoid this, companies have voluntarily entered into licensing agreements to supply COVID-19 countermeasures to low-income countries. For example, AstraZeneca reached an agreement with the Serum Institute of India to supply 1 billion doses of its COVID-19 vaccine candidate, once approved, to low- and middle-income countries including India.⁵⁶

The coming together of individual countries and industry stakeholders to overcome IP barriers is important. And yet, as long as these efforts remain disjointed, equitable access to outbreak countermeasures among the world’s poorest remains at stake. Global solidarity in efforts like the Medicines Patent Pool and Solidarity Call to Action may be the only true solution to ensuring affordable countermeasures universally.

C Facilitating Product Approval

Even with good coordination and adequate funding, promising medical countermeasures often fail during clinical trials or are delayed due to the product approval process. Under an accelerated timeline, the Ebola vaccine did not even receive full regulatory approval until five years after it entered clinical testing. Clinical trials and regulatory approval are both vital to ensure safety and effectiveness, and yet research and regulatory processes could be far more efficient.

Resolving Clinical Trial Design Conflicts

The challenges of designing and conducting clinical trials for outbreak diseases slow or stop promising products from advancing. Clinical research can often only be conducted during a major outbreak, when there are sufficient numbers of patients to support well-designed trials. In fact, the WHO Ethics Working Group has stated that only by conducting clinical trials in outbreak settings could clinicians be assured that scarce resources were being put to their best use.⁵⁷ Developing a clinical trial for implementation during an epidemic is especially complex. Authorities must determine which products are sufficiently promising to be tested during the compressed timeline of an outbreak.⁵⁸ Researchers must design the trials considering factors such as relevant clinical end points, the study size, the effect of herd immunity, and whether to include vulnerable populations like pregnant women and children. Prior to approval, products typically undergo three phases of clinical testing, a process that can take a decade.

Development of the Ebola vaccine from 2014 to 2019 yielded several lessons on expediting vaccine development.⁵⁹ Researchers were given more flexibility in conducting clinical trials. Regulatory agencies from the United States, Canada, and Europe collaborated closely with each other and with National Regulatory Authorities of the impacted West African countries, sharing information on vaccine candidates and testing protocols.⁶⁰ Regulatory agencies also consulted with researchers on the safety and efficacy thresholds required for vaccine approval.

These lessons facilitated COVID-19 vaccine development: researchers were authorized to conduct combined phase 1/2 and phase 2/3 trials, simultaneously testing safety and efficacy to cut months off the clinical trial process.⁶¹ Companies such as Pfizer were authorized to design trials where multiple vaccine candidates were tested in parallel.⁶² Both the WHO and FDA advised on clinical trial study design, outlining that vaccines must prevent infections or reduce the severity of COVID-19 cases by at least 50 percent to be approved. The FDA’s Fast Track designation allowed drug sponsors to interact with the FDA review team about clinical trial concerns such as study design, safety data, dosing, and biomarker use. In the FDA’s guidance for COVID-19 Fast Track review, the agency emphasized the need to include diverse populations in clinical testing, using sufficiently large population sizes to detect safety or efficacy issues, and conducting post-market studies to continue evaluating safety and efficacy even after approval.⁶³

The 2014–2016 Ebola epidemic also revealed an R&D pitfall: stakeholders had no clear agreement on what and how to share epidemiological and research data. This slowed experts’ understanding of the outbreak and hindered the response.⁶⁴ The lack of a sharing platform led to post-outbreak calls to develop better incentives and mechanisms for sharing data. It also incentivized the WHO’s ACT Accelerator for sharing COVID-19 data, as discussed previously. But sharing data remains difficult. Intellectual property and ownership claims arise, such as who owns submitted data, and who has the right to access and benefit from their eventual commercialization. Questions arise on patients’ privacy and consent, whose personal information should be protected when their data are shared. And researchers bear uncertainty about their rights to publish previously submitted data, and potential reputational damage if early research findings are later undermined. Despite these concerns, it remains essential that information on drug and vaccine efficacy, as well as adverse events, are disseminated quickly and openly to ensure rapid development of safe and effective outbreak countermeasures.

Steps taken before an outbreak occurs can facilitate the rapid development of countermeasures. It is important, for example, to identify promising products during inter-epidemic periods. Identification by the WHO of technical specifications for drugs and vaccines for priority pathogens can reduce the time spent evaluating products. And work by CEPI to advance vaccine candidates and platform technologies should make identifying products for clinical trials simpler.

In addition, using inter-epidemic periods to manage regulatory and administrative tasks can facilitate a more rapid response. For example, developing generic clinical trial designs for likely outbreak scenarios and getting buy-in from stakeholders such as ethics boards and communities will provide affected parties with an advanced starting point for discussions when an outbreak occurs. Similarly, protocols and platforms for sharing outbreak and countermeasure data should be established far in advance of an outbreak, with regulatory guidelines on protecting intellectual property and patients’ privacy.

Reforming Product Registration

Once a product is developed, regulatory approvals are needed prior to marketing a drug or vaccine in a country.⁶⁵ Products must be registered with a country’s national regulatory agency before they may be sold in the country, ensuring that the products are safe, effective, and meet quality manufacturing standards. Regulatory hurdles can delay the time for products, which are typically developed in higher-income countries, to be distributed in the lower-income countries where they are often needed most.

Products registered in low- and middle-income countries tend to follow a three-step registration process. First, products are registered in the country where they are manufactured. This initial registration often occurs under “stringent regulatory authority,” adhering to the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). The ICH is a collaboration since 1990 between the United States, the European Union, and Japan to harmonize the scientific and technical aspects of drug registration. An increasing number of generic products, however, are first registered by other national regulatory authorities, such as those of India and China.⁶⁶

After initial product registration, the manufacturer can apply to the WHO for prequalification, typically a prerequisite for international aid agencies, such as Gavi, UNICEF, or the Global Fund, to purchase a product for distribution. For example, WHO prequalification was a prerequisite for COVID-19 vaccine developers participating in COVAX’s AMCs.⁶⁷ The WHO assesses the product’s performance, its risk-to-benefit ratio for the intended population, and whether the product is suitable for the proposed use.⁶⁸ Lastly, the product is registered with the national regulatory agencies in the low- to middle-income countries where the product will be sold. These authorities work to ensure the safety, efficacy, and quality of the products as applied to the health needs of their citizens.

A 2016 study on product registration in Sub-Saharan Africa found that the time between a product’s first regulatory submission and its approval ranged from four to seven years.⁶⁹ This lag – where the drug has been approved by a stringent regulatory authority but is not accessible to vulnerable populations in low-income countries – has led to calls for reform.

Proposed reforms to improve efficiency can enhance both the clinical trial and product approval processes. Agencies can avoid duplicative review by leveraging stringent regulatory assessments. Instead of repeating steps, such as inspections of manufacturing facilities, subsequent assessments could focus on activities that fill key gaps. Eliminating duplication could shorten registration time. Countries can also standardize registration requirements among regional or global partners. For example, technical registration requirements differ among African countries, but efforts to develop uniform standards are occurring. In 2012, regulatory agencies in Burundi, Kenya, Rwanda, Tanzania, Uganda, and Zanzibar launched the African Medicine Regulatory Harmonization program, which strives to encourage regional collaboration and harmonization of regulatory standards.⁷⁰

Just as with clinical testing of outbreak countermeasures, costly delays could be avoided if efforts to reform the product registration process are initiated well ahead of an outbreak.

Safety and Ethical Considerations

Reforms to facilitate the testing and approval of outbreak countermeasures could provide drugs and vaccines to at-risk populations sooner – saving countless lives during future epidemics. In some cases, however, expediting countermeasure development implicates safety and ethical considerations that, if ignored, could result in harm to individuals and societies, and destroy trust between authorities, researchers, and communities. Concerns about cutting ethical or scientific corners are especially acute when there is political pressure to bring drugs and vaccines to the market before completion of clinical trials.

The Declaration of Helsinki requires independent ethics review of human participant research. Even in a health emergency, ethical values are vital.⁷¹ At its core, a clinical trial must have sufficient scientific and social value to justify its risks and burdens, and be designed to create quality data to guide regulatory agencies and clinicians. Respect for the participants and community is another core requirement, including participants’ rights to informed consent and privacy. Communities should be meaningfully engaged, respecting values, cultures, and traditions, with host countries and local researchers treated as equal partners. Trials should be conducted to ensure that benefits and burdens are distributed equitably. Vulnerable populations should be identified and protected. Once the trial is complete, the community and participants should be informed of the trial results, and have access to successful medical countermeasures.

Even with seemingly straightforward ethics principles, ethically appropriate decisions can be complicated, especially during health emergencies. In some settings, trials might exclude pregnant women and children to avoid health risks. But in other settings, such as when pregnant women are especially vulnerable to a disease like Zika, it may be preferable to include them in clinical trials. Early phase 3 clinical trials of COVID-19 vaccines, for example, excluded children, thus creating uncertainty as to the vaccines’ safety and effectiveness in this population. The policy implications are huge, as it took longer to determine whether children could be vaccinated to enable schools to safely open.

During the 2014 West African Ebola outbreak, an ethical debate raged on the use of randomized controlled trials (RCTs), typically the “gold standard” for clinical trials, for Ebola vaccine candidates.⁷² In most RCTs, one group receives an experimental intervention while the other receives a placebo or the conventional care. While RCTs may be the fastest way to generate high-quality data about efficacy and safety, RCTs are not ethical when conventional care means a high probability of death, as with Ebola.⁷³ Withholding promising interventions under these circumstances also creates distrust and animosity among communities and researchers. For these reasons, a number of leading voices argued against RCTs in the context of the deadly Ebola epidemic.⁷⁴ A US National Academies of Sciences committee tasked with reviewing the Ebola vaccine clinical trials acknowledged that uncontrolled trials may be warranted under certain circumstances, such as the unavailability of another treatment as a control, and certainty that patients who don’t receive an intervention will have a poor prognosis.⁷⁵

As recognized in the National Academies of Sciences report, the safety and ethics considerations behind clinical trials are highly contextual to the condition being studied. Consider vaccine development for COVID-19, a disease with a far lower death rate than Ebola. As the “race for the vaccine” progressed, many scientists voiced concerns that skipping steps to vaccinate more persons sooner could do more harm than good. In 2017, a dengue vaccine was pulled from the market in the Philippines after the understudied vaccine was attributed with causing severe cases of dengue.⁷⁶

Similar outcomes could have resulted from COVID-19 vaccines being hurried to the markets. In June 2020, China approved the use of an experimental vaccine, manufactured by the company CanSino, for the country’s military. Phase 1 and 2 trials of CanSino’s vaccine had demonstrated largely mild adverse reactions in some patients, though 9 percent of overall patients had severe side effects that “prevented activity.”⁷⁷ In August 2020, Russia’s Ministry of Health approved a COVID-19 vaccine that had been tested on just seventy-six people by the Gamaleya Research Institute.⁷⁸ The vaccine had undergone phase 1 testing on volunteers from Russia’s military – whose ability to render informed consent is highly questionable. Dubbed “Spuknik V,” scientists around the world denounced the approval as premature and inappropriate, as the vaccine had yet to be proven safe and effective for a large group of people.⁷⁹

In the United States, the FDA by law can only approve vaccine candidates that have been proven safe and effective in phase 3 trials. Vaccine candidates in phase 3 enroll tens of thousands of people with diverse health circumstances from across the country, which is critical to determining safety and efficacy in a real-world setting.⁸⁰ If the FDA determines that a vaccine is safe and effective, the agency can approve the vaccine through an emergency-use authorization prior to the trial’s completion. Yet the US regulatory system was put under pressure by the Trump Administration and Operation Warp Speed. Many experts worried that the FDA would succumb to the pressure to approve a COVID-19 vaccine prior to the presidential election in November 2020.⁸¹ Unlike in Russia, the FDA has an independent advisory committee that reviews approval applications. Still, concerns arose when the FDA issued and later revoked emergency approval for hydroxychloroquine to treat COVID-19 patients. The drug, which had been praised by President Trump, was found ineffective at treating COVID-19, and associated with severely adverse cardiac events.⁸² Fortunately, the FDA performed admirably in granting emergency use authorization for COVID-19 vaccines. The agency used its scientific advisory committee, disclosed all data transparently, and granted authorization only after all the processes were completed.

Unproven countermeasures come with enormous safety risks, underscoring the need for fully informed consent for participation in clinical trials. Candidates for COVID-19 countermeasures could cause serious adverse reactions, or even make COVID-19 infections more lethal – leading to thousands of needless hospitalizations and deaths. Aside from these immediate harms, approving unproven countermeasures contributes to distrust of science when they are found unsafe or ineffective at preventing disease in a population. Vaccine hesitancy has been a major challenge globally, resulting in a resurgence of measles and other childhood diseases.⁸³ In a poll from August 2020, one-third of Americans, and over 40 percent of non-white Americans, said they would not get a COVID-19 vaccine.⁸⁴ Such high refusals jeopardize immunity, particularly among populations that have suffered a history of medical inequities. A rigorous scientific process for putting new countermeasures on the market is absolutely critical to countering public distrust of science and political leaders.

Rigorous scientific process must go hand in hand with transparency from decision-makers at every stage of the R&D process. Transparency helps garner public trust and achieve cooperation with public health recommendations. Operation Warp Speed was criticized when scientists involved with the program disclosed they were excluded from decisions to select the vaccine candidates to receive funding for rapid testing and manufacture.⁸⁵ In a letter signed by over 400 experts in infectious diseases, vaccines, and other medical specialties, the group implored FDA Commissioner Stephen Hahn to disclose the agency’s deliberations on whether to approve a COVID-19 vaccine.⁸⁶ Access to this information would have enabled scientists and health professionals to independently assess and, ideally, promote a safe and effective COVID-19 vaccine to the American people. On a global scale, the WHO must be provided full access to robust information on countermeasure approval decisions. With trusted and informed “gatekeepers” in place from multiple governance realms, the world can maximize the use of safe and effective countermeasures, with a defense against potentially harmful ones.

D Enabling Scientific Innovation

Modern medical tools are essential to combat ongoing threats to global health. Technology offers a way to stock our medical war chest before the next outbreak, epidemic, or pandemic. Vaccines are among the greatest public health achievements in the modern era. Discovering and deploying vaccines against outbreak diseases would be a sound investment in national and global security. Therapeutic agents such as drugs and biologics would lessen suffering and death worldwide. Improvements in diagnostic devices, surveillance, and data-sharing platforms could enable faster and more efficient detection and response to outbreak pathogens and reduce mortality and morbidity.

Science has the potential for major innovation, often our last defense against catastrophic consequences of pandemic disease. But the financing, law, and ethics must be in place – not just when an outbreak strikes, but more importantly during periods of calm. Lurching from complacency to crisis, and back, will never reduce global vulnerabilities. Collectively, policies and processes that support all the building blocks of research and development can save millions of lives. It is wise to remember that there is an ongoing struggle between pathogens with vast power to mutate and to kill, and science with its capacity to prevent and treat disease. For science to prevail over Mother Nature, we need to invest and prepare, building scientific and manufacturing capacity well before the next pandemic strikes.

3 COVID-19 and Boundary-Crossing Collaboration

Laura G. Pedraza-Fariña

COVID-19 unleashed a perfect storm that exposed deep cracks in the foundations of our public health and scientific research infrastructures.¹ As public health budgets faced cut after cut in the past several years,² states found themselves ill-equipped to mount a coordinated response: scrambling to secure enough ventilators and personal protective gear, and failing to consistently test, trace, and quarantine those traveling to and from high-infection areas.³ We now know that these shortcomings, despite spurring heroic efforts to jerry-build solutions with limited time and tools, cost countless lives.⁴

For its part, research infrastructure in the United States and the world over, notwithstanding its many formidable successes, is increasingly fragmented by area of expertise. COVID-19 has laid bare the perils of this fragmentation. As presentations of COVID-19 continue to baffle researchers, the virus is playing a game of cat and mouse with scientific specialties: first thought to be a garden-variety respiratory virus calling for traditional interventions such as oxygenation and ventilation, new findings about its effects on blood cells and blood circulation dynamics implicated a second set of experts such as hematologists and cardiologists.⁵ The virus’ wide-ranging dermatological symptoms suggest that dermatologists may also have an important role to play in our understanding of COVID-19.⁶ Many patients recover from COVID-19 infection only to find themselves besieged by sequelae that span medical specialties and defy scientific understanding: psychotic episodes, failing memories, and chronic fatigue point to a neurological component to the virus’ march through the human body.⁷ And this is only the medical treatment side of the COVID-19 puzzle: disagreements have cropped up on questions about mechanisms of viral spread, pitting aerobiologists, physicists, and computational scientists against infectious disease clinicians on whether COVID-19 is airborne.⁸ The World Health Organization (WHO) itself has framed debates around the virus’ mechanism of transmission as a fight between scientific specialties, with Benedetta Allegranzi, the WHO technical lead for the task force on infection control, questioning “why … these theories [of aerosolized viruses are] coming mainly from engineers, aerobiologists, and so on, whereas the majority of the clinical, infectious-disease, epidemiology, public-health, and infection-prevention and control people do not think exactly the same.”⁹

In short, COVID-19 is one of those boundary-crossing problems whose comprehensive understanding and solution requires the assembly of teams that cut across specialties. And yet our innovation ecosystem – and our funding structures – remain stubbornly organized around disciplinary lines. This tug-of-war between specialization and boundary-crossing gives rise to one of the thorniest innovation policy challenges of our times: how do we build cohesive innovation communities that nonetheless are willing and ready to cross boundaries and collaborate with outsiders? Sociologists of science who have studied the problem of interdisciplinary collaboration understand it as somewhat of a Goldilocks dilemma: getting innovation policy right requires just enough trust and cohesion and just enough cognitive diversity to generate boundary-crossing teams that can work well together.¹⁰

As much as COVID-19 illustrates the shortcomings of our siloed medical and scientific professions, it also represents an opportunity to rethink and reorganize scientific research infrastructure. COVID-19 has become a scientific “nucleating event” of sorts: forcing researchers from many specialties into fragile but promising forms of collaboration around a shared – and pressing – problem, and giving rise to multiple infrastructures to facilitate such collaboration. This chapter argues that forging sustainable cross-cutting collaborations will require ongoing policy action along three axes: (1) building information-sharing infrastructure; (2) creating cross-disciplinary teams; and (3) countering anti-innovation norms. The chapter proceeds as follows: Section 1 summarizes current research from sociology and history of science on the interplay between specialization and intellectual migration in scientific and technological innovation. Section 2 compares two different initiatives, Accelerating COVID-19 Therapeutic Interventions and Vaccine (ACTIV) and Operation Warp Speed (OWS), as case studies to help develop and illustrate my three policy recommendations for boundary-crossing innovation in pandemic preparedness. Section 3 summarizes likely hurdles to assembling and funding cross-disciplinary teams, together with examples of prior successful initiatives that can serve as blueprints for future funding efforts. Section 4 concludes.

1 The Perils of Our Specialized Innovation Ecosystem

Specialization is an important, even indispensable, component of scientific and technological innovation. At its most basic level, specialization allows for the efficient management of an ever-expanding reservoir of scientific and technical knowledge, knowledge that can be filtered through the specialization “sieve” to create more easily categorizable units of knowledge. Scientific disciplines and subdisciplines function as a sieve by developing both research priorities that guide their members in choosing what problems to focus on out in the real world, and research tools and methodologies that govern how community members study those real-world problems.¹¹ Sociologists and historians who study the evolution of science, technology, and medicine often liken scientific specialization to the process of community-building. Forget the image of the genius scientist working alone in a laboratory – at its core, scientific work is a communal enterprise. Scientists work in communities that are held together by a set of tacitly agreed-upon research questions, methodologies, and mechanisms for evaluating what constitutes “good work” within that community.¹² It turns out that analyzing what scientists do from the perspective of what scientists who are members of a particular scientific community do is quite helpful to understand the evolution of innovations in science, technology, and medicine. Community in science – as in any other area in life – has incredible upsides: it generates a shared set of background assumptions (what sociologists call background “social norms”) that, in turn, engender trust among its members. Such community norms, however, also have powerful downsides for innovation.

A crucial downside of allowing innovation to proceed in relatively isolated scientific and medical communities is that real-world problems do not come so neatly packaged. To the contrary, society’s most pressing problems often require solutions that combine insights from diverse scientific specialties. For example, many of the twentieth century’s groundbreaking scientific discoveries, such as the physical structure of our genetic material¹³ or the existence of Big Bang radiation,¹⁴ emerged out of the combination of insights from multiple scientific specialties.

Those community norms that are so helpful in segmenting and organizing knowledge, however, fall short when innovation requires teams with diverse expertise. In their most pernicious forms, the same community norms that build trust and cohesion among community members in fact discourage collaboration across specialties by sowing distrust toward the research questions or methods of other disciplines. In prior work with coauthor Stephanie Bair, we have termed these counterproductive social forces “anti-innovation” norms and argued that a crucial goal of any innovation policy should be to identify and mitigate their negative impact.¹⁵

Anti-innovation norms can have a devastating effect on breakthrough innovation. If there is one insight that has emerged from years of joint studies in history, sociology, organizational economics, psychology, and law it is that breakthrough innovation requires not only trust but also cognitive diversity.¹⁶ Trust provides the social glue and shared language that allow scientists to get to work on a set of common goals. But diversity, in both thought styles and methodologies, is a second indispensable ingredient. Put differently, many breakthrough discoveries emerge through the recombination of ideas, methodologies, and ways of framing problems from two or more distant disciplines. Trust and diversity, however, are forces that often pull in opposite directions. This is because trust tends to emerge from sameness – scientists tend to “trust” research and researchers who adhere to the background research priorities and methodologies of their chosen scientific community. Getting innovation policy right, therefore, is somewhat of a Goldilocks dilemma: too much specialization, too much cohesion, deprives communities of fresh ways of looking at problems and novel methodologies; on the other hand, simply throwing together teams of scientists from widely divergent backgrounds can create cacophony rather than innovation. We need just enough trust and cohesion and just enough cognitive diversity to generate teams that work well together.

For example, the discovery of the physical structure of DNA benefitted from an unusual environment in which both trust and diversity coexisted in a fragile equilibrium. In the context of Nazi Germany, when many scientists were not permitted to attend official seminars, Max Dellbrück organized a “little private club” at his mother’s house where theoretical physicists and biologists (two groups that did not routinely interact with each other) came together. As Dëllbruck puts it, “discussions we had at that time have had a remarkable long-range effect, an effect which astonished us all.”¹⁷ Scholars have hypothesized that the cohesion among these “Dellbrück club” scientists, catalyzed by their Nazi opposition, allowed for fruitful communication to take place across disciplinary boundaries.¹⁸ Indeed, recent empirical network analyses have identified one particular network configuration – the “structural fold” – as essential for generating creative, high-impact ideas.¹⁹ Structural folds are characterized by “cognitive distance” and “intercohesion.” Cognitive distance stands as a network measure for diversity of ideas: the more cognitive distance a relevant group enjoys, the more diverse the knowledge sources that are available to the group.²⁰ Intercohesion refers to the area of overlap between two or more cognitively distant groups. Group members in this overlap area belong to more than one cohesive group; as a result, they can facilitate the emergence of social trust between cognitively distant group members.²¹

2 Building Information-Sharing Infrastructure

Sometimes the hurdles to creative information recombination are surprisingly simple. Information can be “sticky” and remain trapped within particular scientific communities despite it being ostensibly available to the public. It may appear surprising that lack of access to relevant information is still an important hurdle in our modern internet-dominated era. In reality, while the internet offers access to a vast amount of information, this vastness creates a second-order problem: an overabundance of information and a filtering–sorting–prioritizing problem.²² All things being equal, communities will prioritize reading information featured in their own specialty-specific journals and presented at their society’s meetings or trade shows. Having meaningful access to information requires investment in physical and electronic infrastructures that cut across disciplines, for example by creating repositories and protocols for data sharing around a common problem. It also requires mechanisms to transfer “know-how” – ways of performing experiments or delivering treatments that are often tacit (or hard to codify into written language) and, instead, require hands-on training in a particular discipline.

In the area of biomedical innovation, the US response had a promising start through the ACTIV program.²³ The ACTIV program recognized an important coordination gap in the US vaccine and therapeutics research infrastructure. In the words of its founders, “there was no true overarching national process in either the public or private sector to prioritize candidate therapeutic agents or vaccines, and no efforts were underway to develop a clear inventory of clinical trial capacity that could be brought to bear on this public health emergency.”²⁴ This gap is unsurprising, given that more traditional market-based mechanisms for funding downstream research, such as patents and trade secrets, are woefully inadequate to incentivize investment in products like vaccines, for which the unpredictability of epidemics and the unavailability of a steady market create a large gap between potential private returns – which are low and uncertain – and public benefit – which is very high.²⁵ This context was ripe for the creation of alternative mechanisms for research funding. Although several alternative mechanisms for vaccine development already existed prior to the COVID-19 epidemic, the scale of the pandemic catalyzed an unprecedented level of funding, collaboration, and coordination at both the pre-clinical and clinical trial stages in the United States.²⁶

The ACTIV program served as a fulcrum that connected participants across the public/private divide and across institutional boundaries, that created infrastructure to share resources, and that codified tacit knowledge by summarizing best practices and facilitating training across institutions. Each one of its core four working groups (pre-clinical therapeutics, clinical therapeutics, clinical trial capacity, and vaccines) created publicly available databases with curated information that included clinical research protocols, candidate therapeutic compounds, and emerging COVID variants.²⁷ The vaccines working group sought to provide crucial infrastructure to harmonize clinical trial data, allowing information about outcomes and other patient variables to be compared across trials.²⁸ Because COVID-19 cares little about political borders, an effective pandemic response will ultimately require global coordination and collaboration. ACTIV members were well aware of this need, seeking to coordinate their response with international public and private entities.²⁹ It is hard to gauge the full impact of ACTIV’s global collaboration network, in large part because the role of ACTIV in the COVID-19 response has been supplanted by OWS – a much better-funded White House initiative with a different design from the ACTIV coordination network, and under the direction of the Biomedical Advanced Research and Medical Authority (BARDA), not the National Institutes of Health (NIH).³⁰

This notwithstanding, ACTIV’s creation of publicly available data repositories and infrastructure for data sharing, as well as its codification of tacit knowledge, are likely to have social spillover effects (in the form of best practices for future therapeutic candidate selection processes and clinical trials) that extend well into the future and greatly outweigh the NIH’s initial investment (Figure 3.1).

Figure 3.1 Accelerating COVID-19 Therapeutic Interventions and Vaccines (ACTIV)

Each working group contains informal information exchange channels, with work products shared with the public in a codified manner through open access portals. ACTIV, www.nih.gov/research-training/medical-research-initiatives/activ (last visited Jan. 7, 2023).

Operation Warp Speed, an $18B collaboration between the Department of Health and Human Services (HHS), Department of Defense (DoD), and private companies, is justifiably credited with accelerating safe and effective vaccine candidates to approval for use by the US population.³¹ Operation Warp Speed is, in many ways, a success story that underwrites the power of a targeted, hierarchical, command-and-control structure for vaccine development. By acting as the centralized command-and-control authority that also incorporated earlier collaborative efforts like ACTIV,³² OWS effectively became the principal US vaccine effort. Although also a public–private partnership that connected many of the same players previously involved with ACTIV, OWS operated under a dramatically different model. While ACTIV fostered networked connections across the private–private and private–public divides and sought to develop infrastructure to make core information widely accessible, OWS is a centralized operation with BARDA at its core, managing bilateral, and largely secret, contracts with pharmaceutical companies.

Two features differentiate OWS’s overarching structure from that of ACTIV: first, OWS’s insistence on secrecy³³ and, second, its disengagement from global stakeholders (Figure 3.2).³⁴ It is far from clear whether OWS’s centralized design and its secret bilateral contracts were key to the speed of the US vaccine response. A review of the literature suggests that two more important OWS levers were (1) its vast funding (which allowed for simultaneous, as opposed to sequential, clinical trials, process development, and manufacturing scale-up)³⁵ and (2) logistical expertise in global procurement.³⁶ What is clearer, however, is that at the leading edge of technology, innovative ideas often originate in the informal networks of learning and collaboration that cut across firms.³⁷ This, in turn, suggests that rapid information sharing across collaboration networks, rather than secrecy, holds the key to an effective and fast response, especially in the context of a novel virus whose defeat requires rapid technological innovation and global technological diffusion in the face of uncertainty. If OWS shifted the focus from ACTIV’s largely open network of public–private and private–private collaboration to a collection of secret bilateral agreements, it also focused exclusively on a domestic response. Absent from OWS’s mission is any mention of working with global partners to standardize and coordinate development, testing, manufacture, and distribution. Instead, efforts such as COVID-19 Vaccine Global Access (COVAX) have tried to fill this last gap, but without US support until recently, only an anticipated 20 percent of vaccine needs are expected to be met in approximately 190 countries.³⁸ Mistrust from this vaccine nationalism and lack of multilateral global collaboration has profound impacts – complicating the fight against more virulent variants,³⁹ potentially limiting post-vaccination travel to regions that are not politically allied,⁴⁰ and increasing vaccine hesitancy domestically.⁴¹

Figure 3.2 Operation Warp Speed

Bilateral contracts define the extent of information exchange between the government and private companies. Contracts are not transparent or available to the public. Information is not shared publicly or across participants. Nicholas Florko, New Document Reveals Scope and Structure of Operation Warp Speed and Underscores Vast Military Involvement, Statnews (Sep. 28, 2020), www.statnews.com/2020/09/28/operation-warp-speed-vast-military-involvement/ (last visited Jan. 7, 2023); Coronavirus: DOD Response, U.S. Department of Justice, www.defense.gov/Explore/Spotlight/Coronavirus/Operation-Warp-Speed/ (last visited Jan. 7, 2023).

To address these challenges, and to lay the groundwork for a robust global response to future pandemics, the United States and the rest of the developed world need to do more than donate vaccine doses, facilitate bilateral transfers of intellectual property (IP), or even waive IP protection on vaccines – as crucial and as welcome as these interventions are.⁴² Although several of the technologies implicated in COVID-19 vaccines are subject to patent protection, patents alone are not the most significant hurdle to worldwide vaccine availability. Even if all patent rights to COVID-related technology were to be instantly waived, the challenges to manufacturing large numbers of doses of COVID-19 vaccine would remain daunting: other drugmakers would still lack sufficient available inputs, trained personnel, and detailed knowledge about vaccine production technology, in particular with respect to the novel mRNA technology employed in the Pfizer and Moderna vaccines.⁴³ In other words, a key hurdle to scaling up global vaccine production is the lack of manufacturing know-how, which is often kept as a trade secret by originator drugmakers.⁴⁴ Simply waiving trade secret protection, however, would not close this gap: know-how transfer, in particular when new technologies are involved, is notoriously tricky, often requiring the type of “learning-by-doing” that can only happen through immersive training. This type of sticky know-how, also known as “tacit knowledge,” is precisely the reason why licensing deals between academia and industry often include clauses that require academic scientists to remain employed as industry consultants long after the licensing deal is signed.⁴⁵

While OWS lacked a global coordination component, HHS and DoD need look no further than their own initiatives to seize on strategies that can be adapted for know-how transfer on a global scale. For example, HHS’s ACTIV promotes a model of data sharing and know-how transfer mediated by a government agency (NIH) that facilitates knowledge flows across participants and creates standardized knowledge platforms. ACTIV, however, lacked a mechanism for the type of experiential workforce training that is so crucial to the transfer of tacit knowledge, relying instead on the ability of their individual network members to train their own. When local technological capacity is limited – as is the case in global manufacturing capacity – workforce training becomes an indispensable component. Research on innovation clusters, however, suggests a solution: the creation of regional know-how transfer hubs that geographically concentrate training, manufacturing, and innovation.⁴⁶ Although specific individual countries may not yet have the necessary technological capacity to absorb such know-how transfer bilaterally, that capacity is more easily developed at the regional level. Regional know-how transfer hubs could serve as a centralized training center for several recipient countries, both facilitating the transmission of tacit knowledge behind new vaccine technologies and helping develop local capacity by creating informal local networks of information exchange. In fact, DoD-funded manufacturing institutes, such as Advanced Functional Fabrics of America and others within the Manufacturing USA umbrella,⁴⁷ are examples of domestic regional manufacturing hubs that are based on many of these principles of hub design: geographic concentration of technological expertise; a strong focus on building technological capacity through hands-on training; a network design for the exchange of information across the public–private and private–private divide; and a central infrastructure for maintaining shared protocols and best practices. Investing in the development of regional innovation and manufacturing capacity can have important positive spillover effects for future pandemics, leading to increased regional innovation capacity with the ability to quickly adapt technological solutions to local conditions. Countries such as Indonesia, Thailand, and Vietnam provide a powerful illustration of these positive spillovers: having participated in an influenza vaccine technology transfer program spearheaded by the WHO in 2005, they are some of the only lower-income countries that are now producing COVID-19 vaccines.⁴⁸

At the global level, the WHO has previous experience with a hub-and-spokes design for technology transfer in the vaccine space – having helped set up the Netherlands Vaccine Institute (NVI) influenza hub and the University of Lausanne’s hub for the production of vaccine adjuvants.⁴⁹ Capitalizing on this experience, the WHO launched a COVID-19 mRNA vaccine technology transfer hub on June 21, 2021.⁵⁰ At the regional level, the COVID-19 vaccine crisis – and the stark contrast between vaccination rates in developed against developing countries – has prompted African leaders to reconsider their long-standing reliance on foreign vaccine imports.⁵¹ The African Union has announced the creation of the Partnership for African Vaccine Manufacturing – a regional vaccine tech transfer hub with a strong emphasis on research and development through public–private partnerships that include African research universities.⁵² Both of these proposals deserve robust support – both financial and logistical – from the United States and other developed countries. Finally, the WHO’s COVID-19 Technology Access Pool (C-TAP)⁵³ also aims to increase access to technological know-how by seeking licensing agreements with pharmaceutical companies. Thus far, however, voluntary participation in C-TAP has been very limited.

The bulk of private and public investments to curb the pandemic to date have largely focused on developing vaccines and therapeutics and on solving manufacturing hurdles, but much remains to be learned about the biology of COVID-19 itself.⁵⁴ A long-term program of pandemic preparedness will require key actors to expand their focus from vaccines and therapies to understanding the underlying biological mechanisms by which COVID-19 and similar viruses interact with the immune system. Public and private actors should embrace and expand upon the types of boundary-crossing collaborations that were emblematic of the ACTIV initiative, not shelve them as a finished product following the launch of successful vaccines. Cross-cutting knowledge about mechanisms of viral infection is likely to prove crucial to efforts to prevent and treat future outbreaks. The next section summarizes likely hurdles to assembling and funding cross-disciplinary teams, together with examples of prior successful initiatives that can serve as blueprints for future funding efforts.

3 Creating Cross-Disciplinary Teams and Dismantling Anti-innovation Norms

Creating infrastructures and protocols for data sharing is a first and important step in making scientific information meaningfully available across disciplines and institutional boundaries. But the problem of sticky information extends beyond creating platforms and protocols for efficient knowledge flow and know-how transfer. Scientists from a single community may not in fact know how to process information from other disciplines, not because it is inaccessible or tacit but because of what network scholars have termed the “cognitive distance” problem.⁵⁵ Cognitive distance problems are less about information flow and more about the ability to combine seemingly disparate pieces of information and different ways of approaching research questions into a coherent new whole.⁵⁶ In other words, solving or even framing a problem at the intersection of two or more scientific communities can rarely be accomplished by members of a single community. Rather, it requires the assembly of cognitively diverse teams that can come up with new ways to recombine existing knowledge and fresh ways to find and frame new problems, in the process creating novel thought styles and frameworks. In prior research with Stephanie Bair, we described how cognitive distance can arise from three types of social norms that emerge in scientific communities and that act to prevent collaboration across community boundaries: (1) research priorities; (2) methodology; and (3) evaluation norms.⁵⁷ Prioritizing different research questions and different ways to go about answering those questions and evaluating the quality of those answers can lead to two communities having radically-different approaches to a project – approaches that may at times appear incompatible. This conundrum – the simultaneous need for cognitive diversity and the persistent resistance to such cognitive diversity in scientific communities – has preoccupied historians, philosophers, and sociologists of science dating as far back as Ludwick Fleck and including the foundational work of Thomas Kuhn, both of whom described the evolution of science as a clash between different thought styles or scientific paradigms.⁵⁸

What has come to be known as the “airborne vs. droplet controversy” in COVID-19 is a textbook example of the hurdles created by different research priorities, methodologies, and evaluation norms. The controversy pitted two broad communities – and their different methodology and evaluation norms – against each other. To epidemiologists and infectious disease specialists, the long-held default assumption that framed their approach was that influenza viruses (of which coronavirus is a type) spread largely through droplets and fomites, in contrast to other viruses, such as measles, which are aerosolized.⁵⁹ This background understanding of COVID-19 helps explain why this group of researchers more readily accepted the possibility of fomite transmission than of aerosolized transmission, even though the available evidence could support either transmission pathway.⁶⁰ Epidemiologists and public health experts also framed the relevant research question as identifying the “clinically relevant” viral transmission pathways. This framing privileged methodologies that emphasized observational and statistical analysis of real-world transmission events through techniques such as contact tracing, cluster analysis, and R naught measurements. It also implicitly rejected as somewhat suspect any evidence that was not directly tied to clinical outcomes.⁶¹

The members of the second group in this controversy, engineers and aerosol scientists, are in some ways outsiders to the public health policy space, with less influence on public health agendas and recommendations. In a letter to the World Health Organization, members of this group essentially called the WHO and CDC to task for a skewed analysis of the available data: arguing that the WHO had accepted incomplete information when it came to demonstrating large droplet and fomite transmission, but demanded more before endorsing measures to prevent aerosol transmission.⁶² Other critics have lamented this “overly medicalized view of scientific evidence” as an approach that pays insufficient attention to data obtained from controlled laboratory experiments or computer modeling.⁶³

The dynamic between these two communities reflects anti-innovation norms at play – norms that stubbornly police community boundaries and prevent cross-pollination. Much of the research on the potential for aerosol transmission of influenza viruses had been available in the engineering literature for at least the past ten years.⁶⁴ That this controversy erupted amidst the pandemic is a testament to the power of crises to reveal disciplinary clashes in methodologies and research priorities that can spend years quietly simmering beneath the surface, or not even recognized as such – due to the siloed nature of scientific disciplines. This very visible public controversy may have had the salutatory unintended consequence of revealing how the set of different research, methodology, and evaluation norms held by both communities ultimately prevented collaboration and hindered efficient public health measures.

COVID-19 has emerged, hydra-like, as a disease with many faces: the virus can be experienced either as a mild flu or as a deadly pathogen by individuals who appear – by most clinical measures – similarly healthy. Some patients recover only to experience an array of “long-Covid” symptoms, ranging from fatigue to psychiatric disorders and the sometimes lethal multi-inflammatory syndrome.⁶⁵ Scientists do not yet know why there is such an enormous variability in host susceptibility to COVID-19.⁶⁶ This uncertainty befits COVID-19’s status as “one of the biggest evolutionary events in the last hundred years.”⁶⁷ Our science policy strategy should similarly reflect the magnitude of COVID-19’s impact on human biology. By seeking to assemble cognitively distant teams that are likely to lead to breakthrough innovation, we stand the best chance of producing the most useful knowledge about the basic mechanisms of COVID-19 function, and the development of new treatments. Difficulties in assembling cognitively distant teams, however, are bound to emerge not just in analyzing mechanisms of viral transmission but also in research attempting to understand the basic biology of COVID-19, research which to date has received less coverage and has appeared less pressing.

COVID-19 catalyzed the founding of a number of consortia, as well as the repurposing of existing ones to tackle the pandemic. Many of these consortia share several of the hallmark characteristics of successful boundary-crossing collaborations (Figure 3.3). First, COVID-19 sparked the formation of consortia organized around tackling a particular boundary-crossing problem (here, different aspects of COVID-19 infection) rather than around disciplinary lines, which is the typical structure of more traditional funding mechanisms.⁶⁸ This focus on problem-solving helped collapse strong methodological preferences, and brought together experts from a wider range of disciplines.

Figure 3.3 The hallmarks of successful boundary-crossing collaborations

Second, and quite remarkably, most consortia – including those with private industry leaders – have adopted an open science model for sharing both data and reagents. Take, for example, the COVID-19 High Performance Computing Consortium – an initiative spearheaded by IBM, the White House, and the Department of Energy National Laboratories.⁶⁹ The consortium was created to “provide COVID-19 researchers worldwide with access to the world’s most powerful high performance computing resources that can significantly advance the pace of scientific discovery in the fight to stop the virus.”⁷⁰ It relies entirely on voluntary contributions of high-performance computing resources by its members, which span the public–private divide. These resources have been put to use to model mechanisms of viral spread, to design more efficient ventilators, and to identify therapeutic targets, among other projects. Selected research projects are required to produce open results and release their data to the public. Similarly, in Europe, the Excalate4Cov consortium (a public–private partnership) has committed to open-access publishing for all of its data.⁷¹ Two genetics consortia, the COVID-19 Host Genetics Initiative and the COVID Human Genetics Initiative, are also built upon open access to data, at least among consortium members.⁷²

Finally, several consortia rely upon strong informal norms of trust and reciprocity, either by bringing together friends and colleagues who have worked together in the past, or by calling upon norms of shared sacrifice in the face of the pandemic.⁷³ Commentators have celebrated the flourishing of open-sharing consortia as inaugurating a new era of widespread collaboration across the public–private scientific divide. Yet predictions that a new open-science model will enhance innovation while preserving wide access to breakthrough treatments are likely premature. Many of these consortia will need to grapple with the allocation of IP rights and the likely need for further incentives to shepherd potential treatments through clinical trials and commercialization. Preserving openness of results – while crucial for rapid follow-on innovation – is not the same as guaranteeing affordable access to treatment, since the allocation of IP rights is likely to impact pricing decisions. A long-term strategy should consider both how to make data widely available to researchers and how to ensure that resulting treatments are affordable to the public.

Consortia also need to consider structures for the dissemination of tacit knowledge beyond the confines of their core membership – structures that go beyond open-access publishing and include mechanisms for know-how transfer, as outlined in the prior section. Many of these consortia have also focused on narrow types of boundary-crossing, most notably by applying supercomputing resources and artificial intelligence tools to speed up research and discover new patterns or connections in data. We may term this type of boundary-crossing “expertise migration,” which puts the tools of one community in the service of problems found in other. Although this strategy represents an incredibly fruitful tool for accelerating research, it can also fall short – without additional efforts – of creating the type of scaffold across communities that can bring diverse areas of inquiry in communication around a shared problem.

To foster the type of long-term team-building efforts that are necessary for sustained research with the potential for breakthrough innovation, additional key elements are needed. Taken together, these elements constitute the common denominator in a variety of studies of successful boundary-crossing consortia.⁷⁴ First is the involvement of a high-status intellectual actor, or an “anchor tenant” (such as a university or research institute), with a high degree of trustworthiness and a network of preexisting relationships that can increase trust and mitigate the risk of boundary-crossing projects. Second is a wider view of interdisciplinarity not only as intellectual migration but as intellectual co-production of new knowledge frameworks. Finally, some studies suggest that short-term “scaffolding” grants may be sufficient to catalyze long-lasting connections across disciplinary boundaries.⁷⁵ For this reason, governmental incentives for boundary-crossing research need not be large, costly grants.

Funding agencies such as the NIH are well-placed to invest in assembling boundary-crossing teams. The ACTIV initiative highlighted earlier is a good example of how the NIH can create the type of successful scaffolding that brings together multiple players to work on a common problem. But there is another initiative in the NIH’s recent history that more closely resembles the type of funding and infrastructure that would make the creation of COVID-19 interdisciplinary consortia possible. In 2001, under Director Elias Zerhouni’s leadership, the NIH funded a series of interdisciplinary consortia through its “Roadmap” initiative to create the “Research Teams of the Future.”⁷⁶ Zerhouni envisioned these teams as gathering “the expertise of nontraditional teams with divergent perspectives that cut across disciplines” to tackle “the puzzle of complex diseases,” which each individual NIH institute, working alone, was ill-equipped to support on its own.⁷⁷

At the time, this initiative proved controversial and was surprisingly short lived (2005–2012).⁷⁸ But in the context of the COVID-19 crisis, Zerhouni’s admonition that understanding complex diseases requires the assembly of nontraditional teams appears prescient. The reasons for pushback against it – namely that scarce resources to fund principal-investigator-led projects were being siphoned away to create these “teams of the future” – are much less relevant in the face of a clear public health emergency, the specter of future pandemics, and the public willingness to target investment in research to prevent them.

4 Conclusion

COVID-19 has both revealed cracks in the siloed foundations of our research infrastructure and provided us with the impetus and models to correct them. The NIH, and other grant-making agencies, should encourage the creation of consortia that include broader types of boundary-crossing that go beyond using supercomputers to address disciplinary questions and that encompass potential collaborations that cut across medical, basic science, and social science disciplines. Only then will the world truly be ready to face the next pandemic.

4 Legal Paradigms and the Politics of Global COVID-19 Vaccine Access

Matthew M. Kavanagh and Renu Singh

Well before an effective COVID-19 vaccine had been developed, governments and global health institutions were structuring efforts to equitably disseminate them worldwide. Heads of state from many of the world’s most powerful governments, United Nations officials, leaders of global health institutions, powerful philanthropists, and CEOs gathered on private zoom calls and then at public events. They pledged global solidarity and designed a complex web of new institutional arrangements intended to ensure distribution of vaccines would happen on a globally fair basis.

The opposite happened. A year after the first vaccines were registered, 9 billion doses had been administered, but just 1 percent of them were delivered in low-income countries;¹ 72 percent of the population in Western Europe had been fully vaccinated, but just 4 percent in Western Africa.² The highest profile global vaccine equity effort, the COVID-19 Vaccine Global Access Facility (COVAX), achieved less than half of its goal of distributing 2 billion doses in 2021.³ While global governance efforts may yet manage wide vaccination coverage, they did not achieve their stated goal of equitable distribution.

The explanation for this failure, despite backing from powerful individuals and institutions, lies at the intersection of law and politics – in the rise of dueling law and policy paradigms for the achievement of vaccine equity, the success of a paradigm based on voluntary action over legal instruments, and the prevailing political context that made the dominant paradigm predictably ineffectual.

One paradigm assumed that governments should leave in place intellectual property (IP) and other market arrangements that create global monopolies over production of each vaccine developed – limiting supply. Constructed primarily by high-income country governments, philanthropies, and private sector actors, that paradigm focused on coordinating demand and incentivizing countries to voluntarily pool their purchases so that, as effective vaccines came online, limited supply could be fairly distributed. But these assumptions were not shared by all, or even most, governments of the world. An alternative supply-focused paradigm supported largely by low- and middle-income country (LMIC) governments and civil society organizations, instead concentrated on a principle of openness. It proposed greater use of legal authority and sharing of vaccine knowledge to open production worldwide.

These approaches could have been complementary (for example, pooling procurement while compelling the sharing of technology) but, in a remarkable breakdown of international cooperation, there was never serious arbitration among powers. None of the major international venues for negotiation – the UN General Assembly, World Health Assembly, World Trade Organization (WTO), and so on – took up these questions to reach agreement across different interests. As such, two separate paradigms developed, competing for attention, and the interests of powerful global actors ultimately kept the supply/openness paradigm from gaining political traction on the global health policy agenda.

In theory, either approach could have worked. Indeed, some suggest the model behind the dominant approach of voluntary coordinated action amidst monopolies was sound, undermined primarily by the lack of a permanent, rapid financing mechanism and by “unexpected” behaviors by states and companies.⁴ Next time, it is argued, it could work.

We argue that this rationale lacks a firm understanding of politics. Robert Putnam long ago described the “two-level game” in geopolitically important issues in which engagement between states is shaped by the politics inside countries.⁵ In that context, vaccine nationalism and hording by wealthy nations was entirely predictable to observers of the politics of 2020–2021 – characterized by rising populism, growing international rivalries, and a retreat from multilateralism. Yet the paradigm that gained dominance in global health policy required norms of sharing and international cooperation to compel states to limit their own access so other, less-powerful states could get doses, and to ensure pharmaceutical companies filled orders for global health initiatives ahead of those of powerful governments. Missing was a realistic vision of delegated authority as no legal measures bound either states or companies to allocate limited doses ethically.⁶ Failure to achieve vaccine equity, we argue, is explained not by unforeseen technical challenges, but by the fundamental misalignment between the dominant policy paradigm and the international and domestic politics of the moment.

Authors in this volume make a wide range of important proposals on IP, innovation, and access. The question we ask is: which of these might work in an actual pandemic? By tracing the first year of COVID-19 vaccine distribution, we show the critical importance of aligning choice of policy mechanisms with political forces. Indeed, we argue that an openness paradigm may have been more effective not only for reasons of justice, but because it could accommodate populist politics and vaccine nationalism. Important nonstate actors from international organizations and foundations appear to have believed they could work with monopolies and motivate states to prioritize working toward vaccine equity without a robust use of law. They were mistaken. The alternative was a strategy based on legal agreements between states to share knowledge and technology and the use of legal authority by states to compel companies to share so that each country or regional bloc could set up production of effective vaccines for their own population. This strategy did not require countering broad state self-interest and might well have achieved a more equitable outcome.

If global governance mechanisms are to succeed in stopping future pandemics, far greater emphasis will be needed on sharing technology – not just for normative reasons of justice but for the practical crafting of approaches capable of achieving equitable outcomes in the real-world geopolitical context.

1 Creating Vaccines, Creating Vaccine Inequity

That a safe and effective vaccine against the SARS-CoV-2 virus could be developed within a year was far from guaranteed – the previous record was four years for mumps in the 1960s.⁷ Yet a mix of previous investment, global coordinated effort, and a bit of luck rapidly produced multiple COVID-19 vaccines. In December 2020, the United States, United Kingdom, and European Union all approved key vaccines and they began deploying them in large numbers. China and India also quickly approved domestically developed vaccines, following Russia, which had been the first country to do so.

By the end of June 2021, six months into vaccine roll-out, the United States had enough vaccines to cover all its priority populations of health workers and people over sixty-five. High-income countries (HICs) had 90 percent of what they needed.⁸ Low-income countries, on the other hand, had received only enough vaccines to cover 12 percent of their highest priority populations.

While official mortality figures imply that the majority of COVID-19 deaths occurred in HICs – which might make vaccine inequality more justifiable or less harmful – mortality data is highly underreported from LMICs.⁹ Indeed, the majority of cases and deaths in LMICs have likely gone unreported. An analysis of “excess deaths,” accounting for this underreporting, shows that, once vaccines began rolling out, the share of excess deaths in HICs fell and the vast majority of COVID-19 deaths were occurring in LMICs by early 2021.¹⁰ As vaccine coverage rose and cases fell, HICs lifted restrictions and moved to resume normal life. On July 4, 2021, US President Joe Biden declared that “we’re closer than ever to declaring our independence from a deadly virus.”¹¹

As many had predicted, however, leaving large portions of the world unvaccinated led to several new variants as the virus mutated. In mid-March the Delta variant arose in India, which at the time had 2 percent vaccine coverage. Later the Omicron variant arose – likely in Southern Africa, where vaccine coverage rates remained below 25 percent and high levels of immunocompromised individuals are suffering from HIV, cancer, and other diseases.¹² These variants led to a push for boosters throughout HICs – re-exerting pressure on vaccine supply in LMICs.¹³ Throughout this period, HICs focused first and foremost on covering their entire populations.

By the end of the year, vaccine inequity had continued unabated (Figure 4.1) and more booster shots had been administered in HICs than first shots in LMICs. The World Health Organization (WHO) reported that just one in four African health workers received a full course of vaccine.¹⁴

Figure 4.1 Global distribution of vaccines v. population, January 2022.

Sources: Our World in Data, Schellekens, Pandem-IC, World Health Organization

2 Competing Law and Policy Paradigms

Which ideas become policy solutions and which of those make it onto the political agenda of international policymaking has long been studied.¹⁵ In global health, this is particularly complex because of the number of different levels and fora in which international deliberations happen over health – from the World Health Assembly of the WHO to the UN General Assembly to the boards of various health financing agencies.¹⁶ In this case, none of these emerged as a single legitimate space for authoritative policymaking on vaccine access. Instead, groupings of governments and private actors came together in a more ad hoc way. Politically important gaps emerged, like the absence of both the United States and China – the world’s largest economies – which for different domestic, political reasons absented themselves from global coordination efforts.¹⁷ With neither a hegemonic country nor an authoritative international organization forcing all actors into negotiation, policy was made by self-selected groups and little political negotiation occurred directly between higher- and lower-income countries over the equity approach. Vaccine equity efforts emerged into what we characterize as two competing policy paradigms.¹⁸ While there is much that is synergistic about the approaches, the actors, ideas, and context of global public health in 2020 resulted in framing these as different and opposing paradigms. A handful of actors, notably the WHO, unsuccessfully sought to advance both approaches. This division is at the heart of the limited equity achieved to date.

A Leaving IP and Monopolies in Place: A Voluntary Paradigm Focused on Demand

At the March 2020 meeting of the G20, policy leaders from some of the world’s biggest economies began to coalesce around a plan for vaccine access to be built not through global agreement but instead through voluntary action by a group of “countries, international organizations, the private sector, [and] philanthropies.”¹⁹ The Access to COVID-19 Tools Accelerator (ACT-A) was launched at an event a month later, co-hosted by the leaders of France, the European Commission, the WHO, and the Gates Foundation. ACT-A set up a time-limited collaboration focused on cooperation between existing global public health actors – Gavi (formerly the Global Alliance for Vaccines and Immunization), the Coalition for Epidemic Preparedness Innovations (CEPI), the Global Fund, Unitaid, and the WHO.²⁰ Its initial governance centered around ten HIC governments along with key private foundations and the WHO (see Figure 4.2). Representatives of the pharmaceutical industry were key players involved from the start, with LMIC governments appearing in its governance only at a later stage.²¹

Figure 4.2 ACT Accelerator governance structure, June 20, 2020

Source: European Union, Coronavirus Global Response, June 2020, https://global-response.europa.eu/system/files/2020-06/CGRS_United_final.pdf

COVAX, housed at Gavi, became the vaccine pillar of ACT-A. Its goal was to bring the acute phase of the pandemic to a swift end by guaranteeing “rapid, fair and equitable access” to vaccines – aiming to “ensure that people in all corners of the world will get access to COVID-19 vaccines once they are available, regardless of their wealth.”²²

The law and policy agenda behind COVAX was based on the preferences of its main political sponsors – governments, companies, and foundations located in HICs. It positioned the private sector as the main driver of innovation and had little to say about IP – accepting, without debate, that the same system of global monopolies that governed other pharmaceuticals would be maintained. This predominant agenda grounded its strategy in voluntary interventions by companies and donor governments meant to organize the demand side of vaccine production. It focused on the creation of advanced purchase agreements to incentivize development, pooling demand through centralized procurement to increase purchasing power, negotiations with companies making vaccines, and clear demand-signaling that would act as a market-based incentive for producers to expand their capacity. “Self-financing” upper- and upper-middle-income countries were to pay in advance for the option to buy vaccines for their own populations while also financing the purchase of vaccines for LMICs. The primary incentive for HICs to procure their vaccines through COVAX was that it would serve as a de-risking mechanism and “insurance policy” – limiting the need to invest in multiple vaccine candidates (some of which would fail) and ensuring that they would have access to whichever vaccines proved successful without having to gamble their investments on the right vaccines.²³ However, these countries still had the option to negotiate bilateral deals with vaccine makers. LMICs, meanwhile, would have access to doses through the advanced market commitment, financed by donations from philanthropy and governments, as well as the contributions of self-financing countries. By pooling procurement, all countries would benefit from economies of scale and improved buying power.

Equity was to be achieved through two phases – first by procuring and allocating at least 2 billion doses by the end of 2021 – enough to equally cover 20 percent of all participating countries’ populations, protecting the individuals at highest risk everywhere.²⁴ Afterwards, additional doses would be allocated in response to epidemiological conditions, according to a threat and vulnerability formula developed by a joint taskforce of the WHO and Gavi.²⁵

COVAX’s focus was on procuring and delivering the vaccine doses, and on assisting LMICs to ensure that they had the logistical frameworks needed to deliver vaccines to their people. By November 2020, COVAX had raised $2 billion, meeting its 2020 goal.²⁶ That was augmented by a US pledge shortly after President Biden’s inauguration, along with other funders, such that by April 2021 $6.3 billion had been pledged and by June COVAX exceeded its goal with $9.6 billion pledged.²⁷ Funding was, however, slow to arrive as HICs focused more on financing their own purchases first.

This approach did not seek to reach enforceable agreements among states or to place legal obligations on either states or vaccine manufacturing companies. States did not require companies that received public research funding to share technology or agree to COVAX allocations in advance. Companies maintained monopoly control over the production of each vaccine, including IP rights, and it was up to each company to decide whether to sell doses to COVAX (or to LMICs directly), in what quantity, and on what timeline. Neither states nor companies were compelled to prioritize COVAX orders, though companies were urged to voluntarily sell to COVAX and countries to share “surplus” doses from their bilateral negotiations.²⁸

From the start, many leaders in the Global South expressed concern about this approach. African leaders, for example, said their goals were to vaccinate far more than 20 percent of their populations and complained they were scarcely consulted in mid-2020 when the program set that target.²⁹ They questioned why COVAX was based on a model that included no obligations for companies to fulfill African orders or share technology so African companies could make vaccines for their own populations.³⁰

These measures could be complementary. But the agenda of the initiative was narrowed to fit the policy preferences of key members of the coalition backing it, including HIC governments and companies. Pooled demand, for example, could be complementary to an open approach that compelled sharing of knowledge and IP. Ironically, HICs pursued at least limited use of legal mechanisms domestically. US President Joe Biden, for example, used the Defense Production Act to compel companies to collaborate on expanding vaccine production. The WHO and many LMIC leaders also advocated for an integrated strategy.³¹ In addition, there was a global precedent set earlier in the HIV/AIDS pandemic, when the Doha Declaration on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement) and Public Health clarified the urgency and legality of sharing of information and compulsory licenses so that antiretroviral drugs became more accessible and affordable in LMICs.³² But the ACT-A paradigm explicitly excluded calls for more compulsory legal efforts at a national or international level or for a focus on sharing technology.

While political leaders such as EU President Ursula von der Leyen spoke about the “global public good”³³ – such an approach to shared know-how and public production, aligned with economic understandings of a “public good,”³⁴ was not on the agenda.

B Few Doses, Little Equity: Failure of a Paradigm

Ultimately, during the first year of vaccine delivery, the demand-focused/voluntary mechanisms were unable to secure anywhere near the doses needed to achieve equity – even after defining equity and setting goals that some criticized as insufficient. In April, COVAX forecast that it would have 835 million doses to distribute by August, 1.4 billion by October, and 2.2 billion by the end of 2021.³⁵ But major producers refused to commit to selling doses to it. Pfizer, for example, agreed to sell less than 2 percent of its supplies to COVAX; by November, Moderna had promised just 34 million doses and delivered none.³⁶ Instead, these companies prioritized delivery to HICs. Initially, COVAX depended on major deliveries of the vaccine developed by Oxford–AstraZeneca and produced by the Serum Institute of India (SII). However, when there was a major surge of the virus in March, the Indian government put a halt to vaccine exports, much as the EU had done previously.³⁷ COVAX ultimately reached half its 2021 goal of 2 million doses in January 2022.

Governments in Africa, Asia, and Latin America that tried to obtain access to vaccines directly had the same problem. South Africa bilaterally and the African Union as a bloc both deployed emissaries to try to secure supplies from major producers, and only after many months did they finally begin receiving supplies toward the end of 2021.³⁸ Drug companies dragged out negotiations, and they demanded that governments absolve them of all liability and promise sovereign assets as collateral.³⁹ It was even revealed that millions of COVID-19 vaccines being produced at a Johnson & Johnson-contracted factory in South Africa were being shipped to Europe and North America instead of filling African orders.⁴⁰

Meanwhile, HICs used their economic and political power to secure first access to doses in excess of what was needed for their priority populations – in many cases enough to vaccinate their entire populations many times over. The EU, for example, ordered 1.75 billion doses from Pfizer/BioNTech, 300 million from AstraZeneca, 310 million from Moderna, and 240 million from Johnson & Johnson to cover a population of 447 million people.⁴¹ The United Kingdom, United States, Canada, and Israel ordered enough doses to cover their entire populations between 2.5 and 5 times. In total, HICs, home to 1.2 billion people, placed orders for over 7 billion vaccine doses. Leaders applied a range of tactics to ensure they were at the front of the line – from export controls to personal contact from presidents asking CEOs to put their orders at the top of the list.⁴² While wealthy governments ordered based on uncertainty of which vaccines would prove effective early on, laying bets on all products to cover their risk, by mid-2021 multiple effective vaccines were approved in Europe and North America. Yet there were few moves to release ordered doses so that high-risk populations in LMICs could get access before young, healthy populations in the Global North.

Amidst scarce vaccine supply, doses became a diplomatic tool. The United States and “Team Europe” distributed hundreds of millions of vaccines bilaterally and through COVAX. China and Russia moved even earlier to promise their vaccines to dozens of Latin American, Asian, and African countries.⁴³ Many of these promises came with subtle or not-so-subtle strings attached. Danish journalists, for example, reported that Rwanda rejected 250,000 doses when it became clear they were meant to help persuade Rwanda to host asylum seekers deported from Denmark.⁴⁴

C Avoiding Monopolies and Waiving IP: An Alternative Paradigm

While it did not win the day, an alternative paradigm did emerge at almost the same time as the dominant paradigm. It took aim directly at the assumption that monopoly production could deliver during a pandemic and proposed instead a new set of agreements to share technology, waive WTO rules on patents and IP, and focus on maximizing global production. The key idea of this paradigm was to focus more on supply than on demand – achieving equity not by sharing of doses or by signaling demand to originator companies, but by removing monopolies over knowledge and using state power to spur production of effective vaccines by multiple manufacturers throughout the world. In this way, the subject of the policy paradigm was not limited doses but knowledge. The transfer of technology from a handful of originator companies to public and private sector producers, particularly in the Global South, was the goal to maximize supply.

These ideas draw in part from experience with the global AIDS response.⁴⁵ The international community had been incredibly slow to build mechanisms to get HIV drugs to LMICs. Even after action began, success was found only after a shift from distributing a limited supply of high-priced, brand-name medicines to licensing of technologies, production in LMICs, and a supply focus that reduced the price of AIDS drugs by 99 percent.⁴⁶ Coming after millions had died and via pressure from global social movements, the focus on open, affordable supply was key alongside increased aid and pooled procurement.⁴⁷ Many of the same transnational HIV advocacy networks of physicians, lawyers, activists, and Global South governments proposed this alternative paradigm during COVID-19.

Political leaders from the Global South advanced this alternative paradigm at the same time the voluntary/demand paradigm was being put forward by leaders based largely in the Global North. On March 23, 2020 the President of Costa Rica, Carlos Alvarado Quesada, proposed a memorandum of understanding among states to share rights to technologies funded by the public sector among all member countries of the WHO. This included pooling patent rights and designs as well as “regulatory test data, know-how, cell lines, copyrights and blueprints for manufacturing diagnostic tests, devices, drugs, or vaccines.”⁴⁸ The Presidents of South Africa and Senegal and the Prime Minister of Pakistan expanded on this idea in May 2020 in an open letter, joined by dozens of former heads of state and international leaders.⁴⁹ They called for a global agreement implemented under the authority of the WHO that ensured mandatory sharing of COVID-19-related knowledge, data, and technologies; the pooling of intellectual property; coordinated expansion of manufacturing capacity; and a commitment to make COVID-19 vaccines free at the point of service.

In many ways, the vaccines developed by US, EU, and UK sources are good candidates for a public goods approach that focuses on the sharing of technologies. The Moderna vaccine was developed by the US National Institutes of Health (NIH) and supported by $2.5 billion in public funding from the United States for development, clinical trials, and production.⁵⁰ The European Union was a major contributor to BioNTech’s work developing its vaccine through the European Investment Bank and multiple EU research and development (R&D) programs.⁵¹ And the Oxford vaccine was made possible by major public support from both EU and UK governments.

Under the open paradigm, it was proposed that the know-how behind the vaccines resulting from these public investments would be shared widely. Several models were proposed, including licensing by originator companies to multiple other manufacturers, pooling of knowledge and IP, open-source sharing of vaccine know-how, creation of technology transfer hubs.⁵² In addition, a major focus was to be placed on expanding manufacturing capacity, particularly in LMICs, to make the vaccines.⁵³

Key to this would be the effective use of legal and policy tools and of state power to incentivize action by companies, create structures for cross-national sharing, overcome IP barriers, and, where necessary, compel sharing.⁵⁴ Various enforceable global legal frameworks have been proposed to ensure these rights and tackle vaccine nationalism.⁵⁵

In May 2020, a month after the launch of ACT-A, the WHO and several national leaders launched the COVID-19 Technology Access Pool (C-TAP). This followed a resolution by states at the World Health Assembly calling for the pooling of technology and the recognition of COVID-19 vaccinations as a global public good.⁵⁶ Thirty countries and several international organizations supported the launch of the pool, but there was very little overlap between the coalition of HICs, foundations, and industry groups backing ACT-A and the primarily Global South countries backing C-TAP.⁵⁷ Under C-TAP, partners including Unitaid, the UN Technology Bank, Medicines Patent Pool, United Nations Development Programme (UNDP), and Joint United Nations Programme on HIV/AIDS (UNAIDS) would support technology transfer and voluntary licensing of COVID-19 vaccines, along with capacity-building efforts, so that companies primarily in Africa, Asia, and Latin America could make the vaccines.

Apart from the WHO, few of the ACT-A political backers and no G7 countries joined the C-TAP effort. By the end of 2021, no major company had agreed to license its technology through the voluntary C-TAP mechanism, and no country had tied its R&D funding to the sharing of technologies globally. There was also no move toward a global agreement on the sharing of COVID-19 vaccine doses or technologies between HICs and LMICs.

In October 2020, South Africa and India proposed a third element to the openness paradigm – waiving states’ obligations under the WTO to recognize IP protections on COVID-19-related technologies.⁵⁸ This proposal would return national legal prerogative to governments to decide the level of IP protection for COVID-19 vaccines and technologies without facing sanction under WTO TRIPS rules.⁵⁹ This would allow governments to provide legal certainty to those considering investment in new and retrofitted factories to produce vaccines in LMICs, similar or identical to those approved globally, even without the full permission of originator companies.⁶⁰ It would also remove legal barriers to coordinated multi-country production and approaches, since TRIPS provisions for countries without manufacturing capacity are cumbersome and have only been used once – by Rwanda and Canada in a complex process that took years.⁶¹ Producers would still have to secure the know-how – from existing producers, from others who know how these vaccines are produced, or from their own research – but they would not face IP lawsuits or prosecution, which would be important for spurring global production.

The proposal was, in many ways, a very limited one – it did nothing to change patent status in any country that did not wish to act, and it was only temporary. Nonetheless it came up against fierce opposition from industry, governments with significant originator pharmaceutical industries, and IP maximalists who said it would undermine innovation, among other claims.⁶² The proposal ultimately gained the support of over 100 countries, but the WTO’s norm of operating by consensus allowed a handful of countries including the United States, several in Europe, and Japan to block full negotiations on the text of any waiver.

The Biden Administration reversed the US position shortly after taking office – announcing on May 5 that it would back a waiver and support moving to text-based negotiation.⁶³ This shifted the international politics of the question significantly, pushing other holdouts to agree to serious negotiations. However, this shift had little immediate effect, as the focus of opposition simply changed to within-negotiation stalling. The EU, for example, put out its own alternative proposal which many saw as a tactic to distract.⁶⁴ By the end of 2021 – a year after vaccine approvals – a waiver had still not been authorized by the TRIPS council.

Industry and some HIC governments claimed that manufacturing in LMICs, particularly for the most effective mRNA vaccines, was not feasible and could not be started soon enough to matter.⁶⁵ They claimed LMIC producers lacked capacity, financing, and technical acumen, and that originator producers like Pfizer, Moderna, and Johnson & Johnson were the only feasible solution to expand production.

Supply-focused proponents showed that each of these barriers could be overcome. Funding to expand manufacturing became available even before vaccines were approved – with $4 billion announced by the World Bank in October 2020.⁶⁶ The African Union launched the Partnership for African Vaccine Manufacturing in April and secured a major commitment from the Africa Export–Import Bank and African Finance Corporation to fund expansion in multiple countries. Technical know-how was also procured. Thailand, for example, built a partnership between University of Pennsylvania researchers – who had done much of the original research behind the mRNA vaccines – and the Ministry of Health’s pharmaceutical production company to set up mRNA production, even designing their own.⁶⁷ Untapped production capacity was identified in a wide range of countries, including Bangladesh, South Africa, Senegal, Egypt, India, Brazil, and Thailand.⁶⁸

Perhaps the clearest example came when the South African government and the WHO announced an mRNA vaccine production hub that put all the pieces together – the South African company Biovac would act as manufacturer, Afrigen Biologics and Vaccines as developer, a consortium of universities would provide the mRNA know-how, and the Africa Centres for Disease Control and Prevention (Africa CDC) would provide technical support.⁶⁹ What was missing, however, was the “recipe” for an approved vaccine – which neither Moderna nor BioNTech/Pfizer was willing to share.

AstraZeneca made some partial moves, striking a deal with the Global South’s biggest producer of vaccines, the SII, to make hundreds of millions of doses on its behalf for sale to COVAX and directly to countries in the Global South. This deal, however, did not approach the kind of open sharing advocated by the supply/open paradigm’s proponents – using an exclusive licensing agreement for certain territories to simply expand the SII’s monopoly over production. As a result, in March 2021, when India was hit by a second wave, the government’s ban on exports shut down supplies for much of the world. COVAX at this point was largely dependent on SII – which was to produce a majority of its planned supplies for the first half of 2021 – and had no alternative in a context of constrained supplies and monopoly production.

A set of vaccines from China, Russia, and Cuba were shared with greater openness. The Russian Sputnik V vaccine, for example, was offered on an open license basis to manufacturers in many LMICs.⁷⁰ However, in the context of vaccine diplomacy, supplies were negotiated country by country and their quality and efficacy was questioned compared to the more sought-after mRNA vaccines.⁷¹ Technology transfers and vaccine sharing of Russia’s Sputnik V and its further iterations were already taking place before it had undergone all clinical trials, and by the end of 2021 there was still skepticism among international scholars about the validity of the results provided.⁷² Nevertheless, producers in India, Serbia, Argentina, and Iran set up production lines.⁷³ And yet a continuing lack of data, broken promises, corruption, and the start of the Ukraine conflict complicated Russia’s efforts.⁷⁴ Meanwhile, China became one of the major COVID-19 vaccine providers to LMICs in 2021 through donations and mostly bilateral commercial mechanisms, initially focusing on its neighboring countries and those in Africa. Even amidst concerns about how China would vaccinate its own large population and still be able to provide doses abroad, by August 2022 it had delivered at least 140 million doses as donations to 100 countries.⁷⁵

HIC governments have the legal authority to compel sharing of vaccine know-how.⁷⁶ In the United States, for example, the Defense Production Act gives the government wide authority to compel actions from companies during crises. Title 1 gives the government explicit power to allocate “technical information” needed to secure “national public health” – which clearly covers know-how to produce vaccines.⁷⁷ The government could, for example, compel sharing of vaccine-production know-how through the Biomedical Advanced Research and Development Authority (BARDA), which could then train producers around the world to make vaccines. BARDA, established in 2006, is part of the US Department of Health and Human Services and provides a system to address public health emergencies with medical countermeasures including the development of necessary vaccines. With governments having invested heavily in the development of these vaccines, statutes such as the US Bayh–Dole Act also provide authority to compel sharing of government-funded know-how for the public good. The NIH even holds a patent on key mRNA technologies and could demand broader access in exchange for licensing the patented technology.⁷⁸

While HICs had the legal authority to act, they chose not to – likely because of the power of the pharmaceutical industry and the high political level at which a decision to use a mechanism such as the Defense Production Act would have to be taken.⁷⁹ This suggests that further reforms, clarifications, and legal avenues for compelling the sharing of government funding during a crisis would be helpful. For example, proposals to embed clear clauses in government funding contracts for research on pandemic-related technologies like vaccines that explicitly authorize government use or compel the sharing of intellectual property and technology transfer with lower-income countries would make this legal authority clear and perhaps more easily used.⁸⁰

By the end of 2021, however, despite multiple opportunities and backing from NGOs, LMIC governments, and international public health authorities, the supply-focused/openness paradigm had failed to garner sufficient political support to advance significantly. No agreement was ever struck at the WHO on sharing technologies, and while a significantly altered version of the WTO proposal was eventually passed in June 2022, its late timing and provisions significantly narrowed which member states could use it and called into question whether it could still have an impact.⁸¹ Ultimately, a strong norm of maintaining IP rights and monopoly power and the prioritization of domestic concerns won out over a worldwide public health emergency.

3 Politics and Power: Explaining the Failure of the Dominant Paradigm

Both policy approaches could theoretically deliver vaccine equity. Real-world success, however, depended on the global and domestic political contexts in 2020 and 2021. In international politics, states make a wide variety of international commitments – whether, and under what conditions, they are likely to keep them has been widely studied.⁸² Even in the absence of formal treaties, international norms play a key role in motivating state behavior, including the area of health, but compliance is based in part on the strength and socialization of a given international norm.⁸³ Compliance with international commitments also depends greatly on domestic politics and the political attributes of “competing interests.”⁸⁴

In this case, failure of the demand-focused/voluntary paradigm to secure equity was foreseeable and foreseen. Achieving equity under this paradigm, which preserved production monopolies and placed allocation in the hands of vaccine manufacturers, required that pooled procurement mechanisms such as COVAX would be able to get equal access to vaccine doses, that companies would fill orders based on a framework of equity, and that powerful states would refrain from monopolizing doses so that vulnerable groups in all countries could be vaccinated before turning to young, healthy people.

Yet the norms supporting equitable shared access between countries to a limited pool of vaccine doses were remarkably weak. Meanwhile, dominant political forces were lined up in the most powerful states to drive vaccine nationalism. Indeed, leaders’ own statements and actions revealed, early on, that their “two level game”⁸⁵ involved ambiguous commitments to equity alongside simultaneous actions to secure enough doses to cover their entire populations as quickly as possible (often several times over). It also comes as no surprise that countries will look inward and focus on their own security when faced with an external threat, regardless of whether it is a public health emergency. Thus, a global health approach dependent on avoiding vaccine nationalism was, from the start, set against political forces it was unlikely to overcome.

Indeed, HIC governments responded by putting coverage of their entire adult populations as their top priority, and they secured preferential access to the vast majority of supplies available through HIC-based producers, leaving little supply for the rest of the world. Even as inequity prolonged the pandemic and gave rise to variants that disrupted life worldwide, throughout the first year of the global distribution of COVID-19 vaccines, access for LMICs was primarily dictated not by globally coordinated efforts but by the relative scarcity of doses and the location of the manufacturers.

In prioritizing sharing of vaccine know-how so that production could take place in Africa, Asia, and Latin America, the supply/openness paradigm explicitly recognized and sought to accommodate the effects of vaccine nationalism and weak international norms by shifting the actors involved.⁸⁶ Even if this was theoretically not the fastest route to deliver doses, expanding the number and geographic location of producers would have shifted the incentives – allowing HIC-based companies to serve “their” markets first while Asian, Latin American, and African producers served theirs. This aligned with political forces of the time, but remained low on the global health agenda, allowing inequity to thrive.

A Weak Norm Building and Soft International Commitment

The primary mechanism to secure state compliance under the demand-focused/voluntary paradigm was the building of international norms of shared allocation by HICs, appeals to enlightened self-interest, and a project designed to “de-risk” investment. In this sense, global health actors worked as norm entrepreneurs – a familiar role for global health institutions⁸⁷ – trying to disseminate and encourage internalization of the idea that equitable sharing of limited supplies was in the enlightened self-interest of all countries.

A series of global public events, largely virtual due to the pandemic, were created to give governments and global health leaders a platform for norm-building. The launch of ACT-A and COVAX in April 2020 was co-hosted by the French and EU Presidents, Bill Gates, and WHO Director-General Tedros Adhanom Ghebreyesus. President von der Leyen promised the EU’s commitment to develop a vaccine, “produce it and to deploy it to every single corner of the world.”⁸⁸ This was followed in September 2020 by a high-level event that featured heads of state claiming “to build stronger political consensus for a coordinated global response to COVID-19, and champion the importance and urgency of equitable access to new tools, especially effective vaccines.”⁸⁹ Speakers included heads of state from Germany, the United Kingdom, Canada, Norway, South Africa, and Sweden as well as executives from Johnson & Johnson, AstraZeneca, and various UN agencies and nongovernmental organizations (NGOs).

Pledging sessions and political events aimed to raise funding for COVAX, secure donated doses from HICs, and build norms that appealed to the enlightened self-interest of HICs. In one official’s words, “no nation can act alone in a global pandemic. Vaccinating as many people as possible, as quickly as possible, is the only way to reduce the tragic loss of life, end the pandemic, and move us toward economic and social recovery.”⁹⁰ Special envoys were appointed to lead this norm-building work – Ngozi Okonjo-Iweala, former Nigerian Finance Minister (before her election to lead the WTO); Andrew Witty, former CEO of GlaxoSmithKline; and later Carl Bildt, former Prime Minister of Sweden. These efforts, however, built only very weak normative infrastructure, with commitments to funding but little that would constrain powerful states from acting in their own self-interest.

Meanwhile, the international context of rising populism and nationalism was hardly conducive to norm-building. Governments from the world’s two largest economies, the United States and China, did not meaningfully participate in ACT-A. The Trump Administration’s “America First” foreign policy was driving withdrawal from the WHO and disengagement from international efforts, while the United States’ and Europe’s increasingly aggressive stance toward China on COVID-19 undermined trust. Even in Europe, much of the political energy was taken up negotiating Brexit, pushing vaccine equity low on the agenda.

There was no use of formal mechanisms, legal or political, to achieve compliance with actions to promote equity. International instruments for ensuring state compliance range from “hard” binding international law with precise commitments, obligations to act, sanctions for noncompliance, and a third party delegated to implement (for example, WTO rules), to “soft” commitments between states that lack these characteristics.⁹¹ In this case, commitments were even softer than past political declarations on global health from the UN General Assembly. The United Kingdom, for example, promoted an “unprecedented global agreement” called the COV-Access Agreement “to give everyone equal access to new coronavirus vaccines and treatments around the world.”⁹² However, the document bore none of the hallmarks of a significant international agreement. It was signed by twenty countries, almost all HICs, and included only vague promises, such as “commit to the shared aim of equitable global access to innovative tools for COVID-19 for all.” It did not give any international institution (such as the WHO) power to control global allocation, and it established no firm commitments or definition of equity. For example, it did not commit HICs to prioritize the vaccination of vulnerable people in LMICs before young, healthy people in their own countries or even to share excess vaccine doses.

With little firm commitment and no significant stick to ensure compliance, the carrot offered under this paradigm to induce participation also proved quite weak. COVAX sought to incentivize HICs to participate in the pool, which would enable it to allocate ethically among all countries. COVAX was framed as “a critical insurance policy that will significantly increase their chances of securing vaccines, even if their own bilateral deals fail.”⁹³ The risk of making advanced financial commitments to vaccines with unknown efficacy would be spread across countries. COVAX would guarantee the ability to cover up to 50 percent of the population, though without a specific timeline.⁹⁴ But most powerful countries did not actually see these issues as a major risk or excessive investment. They made deals for all or most viable candidates and, with a desire to cover 100 percent of their populations, had every incentive to defect even if they participated in COVAX.

B Domestic Political Incentives Make Demand-Side Paradigm Untenable

Political leaders in most countries have relatively short time horizons, particularly those facing an election in the near term.⁹⁵ In a context of weak international norms and political agendas dominated by COVID-19, leaders prioritized the threat of their own citizens having to wait for their vaccines over the injustice of highly unequal vaccine distribution or even over the threat of a long-lasting pandemic continuing to cause disruption. Even as global health plans focused on vaccinating vulnerable people and health workers worldwide first and HIC leaders were promising to share, they were signaling a very different intention domestically.⁹⁶ None made real plans to slow vaccine access for their populations in order to make supplies accessible to those most in need in LMICs. Efforts were on full display to use political, economic, and strategic power to secure doses for their entire populations as rapidly as possible to the exclusion of others. This was clear long before the first vaccines were available.⁹⁷ Key leaders in LICs voiced their concern that this meant voluntary mechanisms would not work, yet gained little traction.

In the United Kingdom, for example, Prime Minister Boris Johnson came under significant pressure domestically to address the failed British response and remove unpopular lockdown orders such as the much-criticized 10pm pub curfew. Promising everyone in the United Kingdom would get rapid COVID-19 vaccine access became a clear political priority for a threatened government. Trying to stave off a revolt within the Tory party, a government source was quoted promising: “There is a possibility that one day soon we will wake up and Brexit will be done and we’ll have the Oxford vaccine.”⁹⁸ In May 2020, the United Kingdom signed an £84 million deal with AstraZeneca, giving it priority access to 100 million doses. Business Secretary Alok Sharma said, “[t]his deal with AstraZeneca means that if the Oxford University vaccine works, people in the UK will get the first access to it.”⁹⁹ By August, the government has secured preferential access to 340 million doses from Pfizer, Johnson & Johnson, and Novavax – enough for five doses per person in the United Kingdom.¹⁰⁰

In the United States, the Trump Administration failed to respond effectively to the start of the pandemic and was already facing a political crisis in a presidential election year. This dramatically increased the stakes for providing a safe and effective vaccine as soon as possible – and ideally before the November election, as Trump himself said. Indeed, a major point of contention in the campaign became whether Trump was putting undue pressure on regulators to approve a vaccine in time to help him politically.¹⁰¹ Operation Warp Speed (OWS), a public–private partnership initiated in May 2020, aimed to have “substantial quantities of a safe and effective vaccine available for all Americans by January 2021.”¹⁰² By October 2020, OWS had spent at least $12 billion on COVID-19 vaccine contracts to ensure priority access for the United States.¹⁰³ Facing pressure from Congress at the time, Dr. Anthony Fauci predicted that enough doses could be secured for all Americans by April 2021.¹⁰⁴ Senator Tom Tillis also introduced the America First Vaccine Act, which would have required that any vaccine developed with US funding go first to Americans “before it goes to other countries.”¹⁰⁵ Trump agreed, saying, “Day 1 that it’s approved, it’ll be available to the American people immediately,”¹⁰⁶ and issuing an executive order stating that sharing could only happen after all Americans had access. Even after the Biden Administration took charge, powerful domestic political actors pushed for a faster roll-out to all Americans. Congressional committees investigated what more companies and the government could do to procure more supplies “as quickly as possible so we can get them into the arms of more Americans.”¹⁰⁷

In the European Union, President von der Leyen faced pressure from member states frustrated that there was no unified plan to purchase enough COVID-19 vaccines to rapidly vaccinate all of Europe. A letter from six member states warned, “[t]he present situation has raised questions about Europe’s preparedness for pandemics.”¹⁰⁸ This came after a “traumatic event” in which the Trump Administration was rumored to have tried to buy up preferential access to the German company CureVac’s vaccine – resulting in an emergency meeting and announcement of an €80 million plan to help CureVac test and manufacture its vaccine in the EU.¹⁰⁹ France, Germany, Italy, and the Netherlands joined together to create the “Inclusive Vaccine Alliance,” which aimed to ensure vaccines would be produced “on European soil” to secure preferential access for European populations – threating EU cohesion. Von der Leyen, a leading voice for COVAX, responded to this pressure by working to secure any available vaccines, not for COVAX, but for the EU – texting and calling company CEOs herself to secure doses.¹¹⁰ The eventual European plan that emerged focused on getting 70 percent of Europeans vaccinated as rapidly as possible, with no provision to delay roll-out to young, healthy people in favor of the most vulnerable in LMICs.¹¹¹

In addition, facing an upcoming election, Israel’s then-Prime Minister Netanyahu also made securing COVID-19 vaccines for the entire population a center of his campaign – even negotiating directly with Pfizer’s CEO and paying top dollar to receive enough mRNA vaccines to vaccinate the entire population in a matter of months.¹¹² Further, Canada’s Minister of Public Services and Procurement, announcing a major vaccine deal in August 2020, said, “[g]iven intense global competition, we are taking an aggressive approach to secure access to the most promising candidates so that we will be ready to vaccinate all Canadians as quickly as possible.”¹¹³

In this context, political analysis shows that an approach based on pooled procurement and voluntary action by HIC governments and pharmaceutical companies was always unlikely to secure vaccine equity.

4 Conclusion

The global law and policy approach to securing shared, equitable access to COVID-19 vaccines failed. It did so despite remarkable science and despite commitments that came from powerful states well before a vaccine was even available. And it failed despite a great deal of work by impressive institutions to develop an innovative and complex approach to coordinated demand.

Fundamentally, the law and policy paradigm that came to dominate the response – based on a consensus of mostly HIC actors – was misaligned with the political realities of 2020–2021. Vaccine nationalism was predictable in a global context of rising populism and a serious public health emergency. The world’s biggest economies were led by the Trump and Xi administrations, and even those states promising cooperation and shared access signaled their intention to prioritize vaccines for their own populations. Further, key actors decided not to pursue significant legal agreement among states to bind governments or companies to provide vaccines for priority populations in LMICs before shipping enough to HICs to vaccinate, and even boost, their entire populations. Domestic political pressures trumped weak international norms in ways predicted by international relations literature.¹¹⁴ Alternative proposals might have made a difference – providing an option that did not require countries to abandon their immediate self-interest in securing doses for their whole populations. Focusing on sharing vaccine knowledge and technology through waiving intellectual property and compelling technology transfer might have allowed rapid expansion of production to Africa, Asia, and Latin America to expand supply. It would have required overcoming opposition from the pharmaceutical industry, but that at least represents a far narrower interest to counter than nationalism and populism, and one with some precedent. The structure of global health policymaking, however, kept this off the table.

COVID-19 will not be the last pandemic. Looking ahead, far more attention is needed to deploying law in ways designed to succeed in the real-world political context. Commitments to share knowledge and technologies are not easy to secure – but they are far more likely to succeed in moments of crisis than the sharing of limited supplies. Rethinking the policy paradigm for access to medical technologies in a pandemic as well as reorganizing power in global health will both be needed to prevent pandemic inequalities of the future.