The world's trade ministers are struggling to deliver a concrete response to the urgent appeal by the new director-general of the World Trade Organization (WTO), Ngozi Okonjo-Iweala, that vaccinating the world against COVID-19 "is a moral, practical, and economic imperative." The WTO's twelfth ministerial conference (MC12) at the end of November can make progress by getting trade officials to reengage in the pandemic challenge. To vaccinate the world, they should support a new COVID-19 Vaccine Investment and Trade Agreement that focuses on accelerating immediate-term production.
In the story of COVID-19 vaccines, trade is both hero and villain. The unglamorous, day-to-day import and export of raw materials, equipment, and vaccines taking place under the rules of the multilateral trading system have already helped save millions of lives and livelihoods. The WTO should be celebrated for creating an environment to facilitate this progress, however limited.
But WTO members also deserve criticism for not doing more to scale up vaccine production. Despite some progress on international cooperation, their efforts have been piecemeal, often bilateral and inefficient, failing to leverage the collective action framework the WTO provides. In addition, the activities of some of the WTO members actively engaged in Geneva are not necessarily aligned with Okonjo-Iweala's implicit call for measures that would immediately increase vaccine production.
Accordingly, the United States, European Union, and India must convene a small and select group of critical WTO members to facilitate a plurilateral COVID-19 Vaccine Investment and Trade Agreement. The deal should be agreed by MC12 and focus on accelerating the manufacturing needed to get more than 16 billion additional vaccine doses produced and traded as soon as possible.
COVID-19 vaccine doses constitute progress, but billions more are needed
As of October 6, 2021, roughly 6.5 billion COVID-19 vaccine doses have been administered globally. Together the United States and European Union account for nearly one billion of those doses, increasingly the highly effective mRNA-based vaccines from Pfizer-BioNTech and Moderna (figure 1). India has dispensed more than 900 million doses; much of that is the AstraZeneca vaccine manufactured locally by the Serum Institute of India. China has administered more than 2.2 billion doses of its home-grown vaccines, including Sinovac and Sinopharm.
At the other extreme is Africa, which has administered only 160 million doses. Roughly 4 percent of the continent's adult population has been fully vaccinated. Globally, low-income countries are estimated to have vaccinated less than 3 percent of their populations. The World Health Organization target of vaccinating 40 percent of their populations by the end of 2021 is increasingly out of reach.
It is remarkable that the world acted with unprecedented speed to invent multiple life-saving vaccines, get them through clinical trials and rigorous regulatory processes, and manufacture as well as distribute 6.5 billion doses globally so far. Few would have predicted this achievement at the outset of the pandemic. But the crisis is far from over, and obvious shortfalls need to be addressed. COVAX, the consortium organized by Gavi (the Vaccine Alliance), the Coalition for Epidemic Preparedness Innovations (CEPI), and the World Health Organization, was established early for distributing COVID-19 vaccines to poor countries. But it has failed to do so because it could not secure a sufficient number of doses from companies and the countries where the manufacturing is located. Instead, India, the United States, European Union, United Kingdom, and China all prioritized allocating locally produced doses to their entire eligible domestic populations, including those at low risk.
In the continued absence of vaccine-producing countries agreeing and adhering to a more equitable vaccine sharing scheme based on global public health needs, the only solution is that manufacturers in those countries greatly expand their production targets.
The mathematics are simple but stunning. With the exception of Johnson & Johnson, each of the other vaccines is a two-dose regimen. For a global population of more than 7 billion, the starting point is thus roughly 14 billion doses. With waning immunity, some governments are already granting third doses; universal adoption could push demand to 21 billion doses. Finally, because governments are stockpiling excess doses and because there is some inevitable waste in the system (e.g., expired, or opened but unused vaccines), estimated demand must be increased by 10 percent. To be safe, about 23 billion doses of COVID-19 vaccines may thus be needed in the immediate term.
Going from 6.5 billion administered doses to the capacity to manufacture roughly 16 billion more doses is still a long way off.
Trade and COVID-19 vaccine manufacturing supply chains today
Trade will ultimately be the unsung hero if and when COVID-19 vaccines are eventually credited with helping to control the pandemic. For billions globally, shots will not be available through local production, and imports are the only answer. But even Americans and Europeans living in countries where vaccines are manufactured depend on international supply chains in ways that are still poorly understood.
Start with the cross-border movement of people and ideas. In the United States, two of the three authorized vaccines were invented at least partially overseas. The Pfizer jab was created by Turkish immigrants at BioNTech in Germany; Johnson & Johnson was codeveloped at the Janssen R&D lab in the Netherlands. A similar story has emerged for vaccines being administered across the European Union. US-based scientists came up with the Moderna shot, and British researchers at Oxford invented the AstraZeneca vaccine.
Trade has also enabled the cross-border transfer of technology and development of brand new supply chains for COVID-19 vaccine manufacturing. While each vaccine maker created a unique supply chain capable of producing billions of doses annually, one common feature was trade. Take two examples of the basic two-step process of first manufacturing the drug substance and formulating it into drug product in one set of plants before shipping it to a second "fill and finish" assembly-line style plant where the liquid vaccine is put into millions of vials for distribution.
Pfizer and BioNTech mostly relied on their own manufacturing facilities to perform the first step, even by mid-2021 (figure 2). The flow of technology and ingredients between the Pfizer and BioNTech plants in different countries was one form of trade. A second arose within its European supply chain, where the mRNA vaccine might be manufactured at facilities in Germany or Ireland but then shipped over a border to that second type of plant for bottling in Switzerland, France, or Italy.
AstraZeneca set up a different type of production network, but one that also featured trade (figure 3). The jab was invented in Oxford, but instead of using its own plants, AstraZeneca transferred the technology to contractors to manufacture its vaccine, including companies in many developing countries. The largest manufacturer of the AstraZeneca vaccine is the Serum Institute of India. It is also being produced through networks of facilities in the European Union, South America, Japan, Thailand, and Australia, as well as in the United Kingdom and elsewhere.
Highly specialized inputs are a third way in which trade has been essential. Each COVID-19 manufacturer relies on a host of critical equipment and raw materials—such as bioreactors, bioreactor bags, filtration pumps, filters, cellular materials, vials, stoppers, syringes, and other ancillary supplies—that are often produced only in other countries.
Shortages of those imported inputs often grabbed headlines during the pandemic. The heads of the Serum Institute of India, Novavax, Biological E., and CureVac all complained publicly that such import shortfalls impeded their abilities to reach production targets. (Pfizer and Moderna also complained of input shortages but did not necessarily tie them to imports.) On the other hand, such interdependence and the fear of a trading partner retaliating by shutting down a vaccine input pipeline likely also helped keep trade for finished vaccines flowing in the other direction. Take the lipid nanoparticles critical to the mRNA vaccine of Pfizer-BioNTech (see again figure 2). The United Kingdom was a critical source of lipid nanoparticles early in 2021 for the plants in the European supply chain. Keeping those UK exports flowing during the pandemic was essential to getting those vaccines manufactured and ultimately shipped back to the United Kingdom despite rising UK-EU political tensions.
But trade disruptions are also the problem in the COVID-19 vaccine story. Though less an indictment of the WTO, governments of vaccine-manufacturing economies engaged in "vaccine nationalism" by refusing to share enough doses with COVAX to distribute to health care workers and vulnerable populations globally. The failure to prioritize the global public health crisis has led to additional deaths and to the emergence of lethal virus variants that have spread to vaccine-hoarding countries themselves.
These problems demonstrate the critical need for further geographic diversification of manufacturing facilities. For Africa, COVID-19 vaccine makers have taken baby steps to address this during the pandemic. Johnson & Johnson was the first—its vaccine is already being bottled by Aspen Pharmacare in South Africa, but only starting in July 2021. Pfizer-BioNTech signed an agreement with Cape Town-based pharmaceutical company Biovac, also to begin doing fill and finish for their vaccine, but the South African facility is expected to come online only in 2022. In October 2021, Moderna announced plans to build a 500-million-dose mRNA vaccine production facility in Africa, but the site had not yet been selected. Neither has the timeline for when it would become operational. Finally, in another long-run initiative, the European Commission announced in May it would provide €1 billion to help "develop a number of regional manufacturing hubs across the continent."
Getting COVID-19 shots quickly into the arms of people across Africa, as well as low-income countries elsewhere, cannot wait until more African production comes online sometime late in 2022 or 2023. That goal must therefore rely on trade and expanding production in manufacturing countries today.
International cooperation so far
Policymakers at the highest levels in major vaccine-manufacturing economies have now recognized the need for enhanced cooperation and engagement. Significant steps began in March 2021 and have accelerated since, albeit in a disorganized fashion and not in concert with the WTO. The United States has been heavily involved, in part because the inputs in short supply in other countries were primarily sourced from US manufacturers.
In March, Presidents Joseph R. Biden and Ursula von der Leyen appointed Jeffrey Zients and Thierry Breton to facilitate US-EU cooperation over COVID-19 vaccine supply chains. Their relationship helped resolve input bottlenecks—CureVac is one public example—and was formalized into a joint COVID-19 Manufacturing and Supply Chain Taskforce in September.
There are other examples. A US-India dialogue began in earnest in April 2021, triggered by the CEO of the Serum Institute of India accusing the Biden administration of imposing an "embargo" on US exports of vaccine-making inputs. The United States responded by immediately sending emergency supplies of that equipment, later cementing US-India vaccine collaboration through the "Quad" (with Japan and Australia) in September. Furthermore, the United States, France, Germany, and the World Bank announced funding in June to South Africa's Aspen Pharmacare to expand its vaccine manufacturing.
Manufacturing more vaccine doses more quickly, cheaply, and in more locations also requires expanding the capacity to supply critical inputs and facilitating additional investment. The United States provided some subsidies to companies manufacturing those inputs in 2020 and early 2021 under Operation Warp Speed. In the face of continued input shortages, the Biden administration announced an additional $2.7 billion from the American Rescue Plan in September 2021. Aside from CEPI, few others globally have announced subsidies to expand capacity to vaccine input suppliers.
How to make CVITA a reality
The United States, European Union, India, and their partners in vaccine manufacturing supply chains must now consolidate their fragmented initiatives into a COVID-19 Vaccine Investment and Trade Agreement (CVITA). To start, CVITA would be a plurilateral agreement, demanding participation by those WTO members, as well as the United Kingdom, Switzerland, Japan, Australia, and potentially South Africa. (It could include China, but uncertainty over Chinese vaccines, its mostly local supply chain, and onerous transparency demands described below mean China may be unwilling to participate.)
CVITA must have four components to work:
First, CVITA should be aligned to leverage COVAX, the umbrella for the public and private international organizations that have joined together for the purchase and distribution of vaccines. The committed vaccine output of the producing-economy members of CVITA would be allocated between themselves and COVAX. Other countries would subscribe to COVAX for their vaccine disbursements, with the subscriptions having zero cost (or being highly subsidized) for low-income countries.
Second, the investment component of the agreement must create a framework to subsidize the full vaccine manufacturing supply chains for the committed vaccine manufacturers. "Push" contracts must be written to coordinate expansion of input production capacity to supply COVAX; they must not simply be deals to buy supplies that allow firms to use existing capacity to eventually deliver on their own timetables. Furthermore, since the benefits of such investments spill over outside of national borders, local governments (including the United States) lack the incentive to subsidize input capacity enough to meet global manufacturing demand. Those public investments in supply chains should thus also be funded by the subscriptions paid into COVAX. Finally, CVITA should also support COVAX Marketplace, a secondary market established by CEPI to help short-term reallocation of scarce inputs when inevitable bottlenecks materialize. (This exchange would include for plants set up by manufacturers prior to obtaining regulatory approval where the vaccine candidate subsequently failed to pass clinical trials.)
Third, CVITA should include an enforceable commitment not to place export restrictions on supplies of vaccines and related materials destined for other signatory countries. In effect, subsidized imported inputs would be exchanged for future doses of an exported vaccine. Countries should agree that vaccine export restrictions may enable other treaty participants to retaliate by jointly curbing their supply of inputs to the export-restricting country. This potential mechanism for reciprocity, if made explicit, can be used to convince skeptical domestic audiences that hoarding—while politically tempting—is self-defeating. Since CVITA would not apply to non-participants, it would not alter the current status quo under WTO agreements which permit export restrictions on public health grounds to vaccine-consuming countries refusing to participate. Receiving protection against export restrictions would thus provide an incentive for nations to join the CVITA.
Fourth, this type of international policy cooperation demands high levels of transparency. Trust can be maintained—decreasing the likelihood of hoarding—only if access to information on COVID-19 vaccines and inputs reduces uncertainty. When supply disruptions occur, transparency will also help differentiate between genuine input shortages versus those resulting from export bans.
In response to dozens of countries imposing export restrictions on staples during a perceived food crisis in 2008-2011, the Group of Twenty (G20) developed the Agricultural Market Information System (AMIS) to improve transparency and coordinate policy in the event of sudden scarcity. That system generated information and trust that arguably reduced the use and duration of agricultural export bans in the early days of the COVID-19 pandemic.
These four components must all be part of the deal. If not, CVITA will fail. Firms will refuse investments to achieve economies of scale without guarantees that they can "export" from those facilities. Governments will refuse to subsidize the full supply chain needed to create that expansion if they fear insufficient access to finished vaccines that get manufactured.
Now is the time for CVITA
Establishment of a CVITA faces many obstacles. The unprecedented nature of the pandemic demands new types of international collaboration over different policy instruments because of the cross-border nature of vaccine manufacturing supply chains. To its credit, the WTO Secretariat has played an important role in convening industry, civil society, and policymakers to educate the community about the underlying supply chain challenges and to generate potential policy solutions.
Until now, small groups of well-intentioned negotiators have proposed a few initiatives without making progress on turning those ideas into reality. The Trade and Health initiative received a lukewarm reception, in part because it focused on trade facilitation and stopping export bans. Those are important issues, but negotiators must achieve a joint commitment that members subsidize and provide transparency over the full vaccine manufacturing supply chains as well as ensure that poorer countries are not priced out of the market for the vaccines they are someday able to produce domestically. A second proposal, initially made by India and South Africa, was to waive patents for vaccines. Such a waiver by itself is likely to have only a limited immediate impact on increasing production, given that the main technological impediment to vaccine manufacturing is how to affirmatively transfer production knowhow, not the patent. (There are other impediments to scaling up manufacturing, such as insufficient supply of specialized inputs, inadequate regulatory oversight, and an inexperienced workforce, that a patent waiver would also not resolve.)
A final challenge is that many governments have not adequately included trade ministers in the domestic pandemic policy response. For COVID-19 vaccine supply chain policy, US efforts have been shepherded by the White House (Jeffrey Zients) as opposed to the US Trade Representative (Katherine Tai). In the European Union, policy has been driven by the Commissioner for the Internal Market (Thierry Breton) as opposed to the Trade Commissioner (Valdis Dombrovskis).
The time is now for the US Trade Representative, EU Trade Commissioner, and other trade ministers to become more engaged. Over 16 billion additional doses are still needed to save lives globally. The world economy is suffering trillions of dollars of losses due to the ongoing pandemic. Inequality is rising. Supply chains for other products are under attack.
At MC12 in November, the United States, European Union, India, and a select group of other key countries should commit to a small, plurilateral CVITA to enshrine and expand upon the cooperative steps they are already taking outside of the WTO. The threat that new and more devastating virus variants could emerge, against which existing COVID-19 vaccines would be ineffective, means that no one is safe until the pandemic is under control globally. Trade ministers should do their part to ensure that everyone everywhere has access to COVID-19 vaccines.
1. For an explanation, see Chad P. Bown and Thomas J. Bollyky. Forthcoming, "How COVID-19 vaccine supply chains emerged in the midst of a pandemic," The World Economy.
4. Given recent developments, these proposals update those originally formulated in Chad P. Bown and Thomas J. Bollyky. 2021. Here's how to get billions of COVID-19 vaccine doses to the world. PIIE Trade and Investment Policy Watch, March 18.
5. See Scott Duke Kominers. 2021. World Can Have Covid Boosters and Its First Doses, Too. Bloomberg Opinion, September 28.
6. The failure to be transparent in early 2021 led to allegations the United States was using the Defense Production Act (DPA) to restrict exports of vaccine inputs. The alternative and perhaps dominant explanation, given that domestic manufacturers like Pfizer were also complaining about shortages, is that DPA was used to allocate inputs in short supply globally to the manufacturers most in need and that were making vaccines authorized for public use. See Thomas J. Bollyky and Chad P. Bown. 2021. The Real Vaccine Procurement Problem: Why America Should Make Its Supply Chain More Transparent. Foreign Affairs, June 24.
7. See Thomas J. Bollyky and Chad P. Bown. 2020. Vaccine Nationalism Will Prolong the Pandemic. A Global Problem Calls for Collective Action, Foreign Affairs, December 29.
8. Such a framework would have been more likely to address two fundamental problems to emerge with Indian government policy. First, India waited to subsidize vaccine manufacturing capacity until April 2021, perhaps because it miscalculated the seriousness of the pandemic. Second, the Indian government didn't pre-order sufficient doses; it simply seized doses that had been produced by COVAX-funded technology transfer and manufacturing arrangements.
9. This last issue of poor countries becoming concerned about access to vaccines produced locally emerged in late 2021 when Johnson & Johnson vaccine doses bottled at the Aspen facility in South Africa were found to be shipped back to Europe despite the latter having much higher vaccination rates.