Here's how to get billions of COVID-19 vaccine doses to the world
This essay is part of a PIIE series on Economic Policy for a Pandemic Age: How the World Must Prepare.
The unprecedented development of several effective COVID-19 vaccines in less than a year is an historic achievement in the annals of scientific research. No less impressive, however, is the work of the US government in organizing and subsidizing a complex supply chain for manufacturing and distributing the COVID-19 vaccine. American policymakers made large advance purchases of potential vaccines and supported some sponsors—such as Moderna and Johnson & Johnson—carrying out clinical trials, while simultaneously working with myriad far-flung and lesser-known contract manufacturers and suppliers of equipment and ingredients (from cellular material to glass tubing to syringes), to actually produce the vaccines and related supplies. The crash effort that went from science labs to vaccine distribution began under the Trump administration’s Operation Warp Speed and has continued and expanded under the Biden administration.
As Americans prepare to emerge from the pandemic, however, much of the world is still living a nightmare. Global demand for vaccine doses dwarfs the supply. Only a handful of countries are able to manufacture these vaccines at the scale needed to overcome infections and deaths that are still growing—and shifting due to the emergence of viral variants. It is certainly problematic that the United States and some other countries are hoarding doses and critical vaccine supplies for their own domestic populations. But a focus on hoarding alone threatens to miss one crucial point. There is much that the United States can do to help replicate the success of Operation Warp Speed on a worldwide scale to get billions of doses to a global population.
The United States needs to do something new: Devote resources and engage both domestic and international sources in coordinated, targeted investments to build on the existing global infrastructure of labs and manufacturers that produce specialized inputs for the vaccines, whether the drugs or syringes or vials that contain them. The vaccine manufacturing supply chain is already global. But it needs to be better organized and subsidized at multiple levels.
Governments and philanthropists have worked together to scale up HIV medicines and pediatric vaccines, but the level of policy cooperation needed for COVID-19 vaccine production is, admittedly, unprecedented. That’s why a global coordinating body will also be needed to help some countries scale up the production of early stage ingredients and critical equipment, knowing they can incentivize other countries to subsidize expansion of later steps in the manufacturing process that will also need to be scaled up. Skillfully combining those inputs into vaccine output and then guaranteeing access to the final doses and related supplies through international trade will require trust among governments that is currently in short supply. But without guarantees and coordination, individual governments will revert to self-defeating vaccine nationalism.
One way to accomplish this goal would be to set up a new and enforceable COVID-19 Vaccine Investment and Trade Agreement (CVITA). Such a CVITA can draw some lessons from the US subsidization and coordination of its domestic vaccine manufacturing supply chain under Operation Warp Speed. It was not easy to get so many competing entities to work together in the United States. But its success is already clear. The even more complex policy challenge now is to get the rest of the world vaccinated.
How Operation Warp Speed expanded output and input capacity along the vaccine manufacturing supply chain
The US government announced the framework behind Operation Warp Speed (OWS) on May 15, 2020. In contrast to the Trump administration’s mismanagement of the public health response on procurement and distribution of personal protective equipment, social distancing, masking, as well as business and school closures, government officials succeeded in accelerating the development and manufacturing of vaccines for the American public in record time.
The federal government—through the Departments of Defense and Health and Human Services, as well as the Biomedical Advanced Research and Development Authority (BARDA) and other agencies —created OWS to coordinate clinical trials and scale up manufacturing in advance of regulatory approval of potential vaccines. This “at risk” approach—spending money that might be lost if a vaccine was not ultimately approved—was essential to making rapid progress. OWS also helped expedite the development of viable vaccines able to obtain authorization from the Food and Drug Administration for emergency public use.
OWS was not flawless, but there is still much the rest of the world can learn from the American experience. Through OWS and successor efforts in the Biden administration, the United States made large advance purchases and coordinated and matched suppliers with vaccine sponsors to ensure those purchase orders would be fulfilled. OWS subsidized input production capacity, including capital equipment, raw materials, and the syringes, vials, and other ancillary supplies needed for packaging and administering doses into arms. The United States worked with manufacturers and suppliers, invoking the Defense Production Act if necessary, to untangle potential input bottlenecks. The scope of OWS engagement varied depending on the pharmaceutical sponsor and the vaccine.
In July 2020, the United States contracted with Pfizer for 100 million future doses of the vaccine it was developing in partnership with the German biotech firm BioNTech. That sizable advance purchase was enough to help expedite the clinical trials and justify reserving and investing at risk in production capacity at three US manufacturing facilities. Pfizer produced raw materials at its Missouri plant, which were then shipped to its Massachusetts facility and turned into drug substance before being transported to its Michigan site for formulation into the vaccine and packaged for distribution.
With other vaccine candidates, OWS was more extensively involved in scaling up production along a fragmented supply chain.
Consider the further challenges involved when vaccine sponsors include smaller firms, with no prior experience in manufacturing vaccines on a global scale (see figure). In this increasingly typical business model, a biotech firm or university researchers are engaged in the early stage research and development (R&D) of a vaccine candidate before licensing it to a second company to conduct later stage clinical trials, produce the vaccine to commercial scale, seek regulatory authorization for those clinical trials, manufacture, and, ultimately, expand public use.
The process requires adequate supplies of a wide variety of specialized inputs—everything from expensive pieces of capital equipment like bioreactors and filtration pumps to single-use bioreactor bags, adjuvants and lipids—from a range of suppliers. Bulk drug production often requires recruiting partners further along the chain to complete the final “fill and finish” step of adding other ingredients and putting the correct dosage into tiny containers suitable for shipping to health care workers. And, of course, the health care workers require syringes, needles, and personal protective equipment to administer the doses.
One missing input or piece of equipment could grind the entire supply chain to a halt.
The United States spent heavily and relatively early in 2020 to subsidize and invest in many steps of that supply chain (see table). Early funding for the Moderna candidate allowed Lonza to establish a manufacturing facility in New Hampshire at risk. The United States reserved other contract manufacturing facilities, including Emergent Biosolutions in Maryland in May and Fujifilm Diosynth Biotechnologies in Texas in July, ensuring they would be available once additional subsidized vaccine candidates had cleared regulatory hurdles. Grand River Aseptic Manufacturing in Michigan was contracted to fill and finish. And in March 2021, the United States further expanded capacity by contracting with Merck to use some of its facilities to both manufacture and fill and finish the Johnson & Johnson vaccine.
Finally, the US government did not just buy up inputs these companies would ultimately need for manufacturing, it also spent hundreds of millions of dollars in 2020 contracting with other specialized firms to expand those companies’ input production capacity. Cytiva received funding to expand capacity of cellular material, bioreactors, and bioreactor bags. SiO2 Materials Science and Corning were contracted to enlarge production of glass tubing and vials, and Smiths Medical, BD (Becton, Dickinson and Company), Retractable Technologies, and ApiJect Systems scaled up to make more syringes, needles, and single-dose injectors.
|US federal subsidies or contracts to COVID-19 vaccine supply chain, 2020 through March 12, 2021|
|Johnson & Johnson (Janssen)||$21 million||February 11, 2020||Support nonclinical studies and a Phase 1 clinical trial|
|$436 million||March 27, 2020||Amendment|
|$1 billion||August 5, 2020||Demonstrate large-scale manufacturing, 100 million doses|
|$85 million||August 21, 2020||(Unknown)|
|$454 million||November 13, 2020||Amendment, funding for Phase 3 clinical trial|
|Sanofi and GSK||$31 million||April 10, 2020||Accelerate nonclinical studies and a Phase 1 clinical trial|
|$2.04 billion||July 30, 2020||Phase 3 clinical trial, manufacturing demonstration project|
|Merck and IAVI||$38 million||April 15, 2020||Accelerate development of vaccine candidate|
|Moderna||$430 million||April 16, 2020||Accelerate development of vaccine candidate|
|$53 million||May 24, 2020||Expand manufacturing capacity|
|$472 million||July 25, 2020||Support Phase 3 clinical trial|
|$1.53 billion||August 11, 2020||Support Lonza’s manufacturing, 100 million doses|
|$1.67 billion||December 11, 2020||100 million doses|
|$1.75 billion||February 11, 2021||100 million doses|
|Novavax||$60 million||June 4, 2020||Manufacturing components for use in Phase 2/3 clinical trial|
|$1.60 billion||July 6, 2020||Demonstrate commercial-scale manufacturing|
|Pfizer (BioNTech)||$1.95 billion||July 21, 2020||100 million doses|
|$2.01 billion||December 22, 2020||100 million doses, option for 400 million more|
|$2.01 billion||February 11, 2021||100 million doses|
|AstraZeneca (Oxford)||$1.6 billion||October 28, 2020||Accelerate development and manufacturing to begin Phase 3 clinical trial|
|Emergent Biosolutions||$628 million||May 30, 2020||Contract manufacture, fill and finish|
|$20 million||August 6, 2020|
|Fujifilm Diosynth Biotechnologies (Texas A&M University)||$265 million||July 24, 2020||Contract manufacture|
|Grand River Aseptic Manufacturing (GRAM)||$161 million||August 6, 2020||Fill and finish, including for Johnson & Johnson's vaccine|
|Ology Bio||$106 million||August 17, 2020||Fill and finish|
|Merck||$269 million||March 2, 2021||Produce drug substance, formulate and fill vials of Johnson & Johnson’s vaccine|
|Equipment and other input suppliers|
|SiO2 Materials Science||$143 million||June 5, 2020||Glass tubing and vials|
|Corning||$204 million||June 5, 2020||Glass tubing and vials|
|Becton, Dickinson and Co.||$42 million||July 1, 2020||Syringes and needles|
|Retractable Technologies||$54 million||July 1, 2020||Syringes and needles|
|Smiths Medical||$21 million||July 11, 2020||Syringes and needles|
|Cytiva||$31 million||October 13, 2020||Cellular material, mixer bags, and bioreactors|
|ApiJect Systems||$590 milliona||November 19, 2020||Prefilled, single-dose injectors|
|a. Loan to finance 75 percent of the project’s capital costs.|
|Sources: Compiled by the authors from Biomedical Advanced Research and Development Authority, 2021, BARDA’s Rapidly Expanding COVID-19 Medical Countermeasure Portfolio and BARDA’s COVID-19 Domestic Manufacturing & Infrastructure Investments; Novavax; Merck; GRAM; and US International Development Finance Corporation.|
Defense Production Act: Reallocating inputs along the US supply chain
As American manufacturing of COVID-19 vaccines ramped up in 2020, the US government also reportedly invoked the Defense Production Act (DPA) to reallocate scarce inputs from one firm to another. (The DPA can be used to compel manufacturers to accept and prioritize government contracts ahead of private sector orders.) Because some firms may have stockpiled inputs in anticipation of producing vaccine candidates that failed during trials (see again table), the DPA could be useful in freeing them up in the absence of well-functioning secondary markets.
Some vaccine manufacturers requested use of the DPA. In December, the New York Times reported that Pfizer asked the US government to use the DPA “to give the company better access to roughly nine specialized products it needs to make the vaccine,” including lipids. When queried about requests that the US administration invoke DPA, Pfizer’s CEO, Albert Bourla, told CNBC, “we are asking them and I hope that they will do it very soon because, particularly in some components, we are running at critical supply limitations.” In February 2021, DPA was reportedly again invoked to get Pfizer “more manufacturing supplies, including pumps and filtration units.”
As manufacturing starts in Europe and India, input shortages have arisen
As manufacturing began to scale up elsewhere, including Europe and India, concerns arose that its slowness was due to a scarcity of inputs. A shortage emerged for lipids, a key input in the messenger RNA (mRNA) vaccine, only to be aided by Merck agreeing to accelerate supplies for BioNTech. Another involved bioreactor bags, with Stanley Erck, the Novavax CEO, stating: “We just had a minor breakthrough where we thought we were going to be out of [bioreactor] bags in one particular facility. And we just got enough to get through February, March and April. Otherwise, [the factory] would have shut down.” The chief executive of the Serum Institute of India, Adar Poonawalla, went further, linking the global shortage of bioreactor bags to the US policy of prioritizing domestic production, thereby limiting exports of inputs to other countries in need.
European Commission President Ursula von der Leyen has indirectly acknowledged the American approach playing a key role in securing the supply chain for its vaccines, “The US has a strong advantage by having BARDA,” she said. “This is an infrastructure Europe did not have.”
Admittedly, there has been insufficient time and data required to do a full, empirical cost-benefit assessment of the OWS or use of the DPA by the Trump and Biden administrations. OWS surely could have spent resources more effectively, and it will likely emerge that invoking the DPA had important unintended consequences, including on input availability for foreign manufacturers.
Nonetheless, there is little question that US intervention helped scale up inputs and unlock input bottlenecks to expand production overall. Furthermore, it is also hard to argue that the tens of billions of dollars of US federal expenditures were too large, given the pandemic’s ongoing human costs (hundreds of thousands of lives lost) and economic costs (trillions of dollars of lost gross domestic product). The same would be true with regard to resources needed to expand manufacturing to meet global demand for vaccines and the critical inputs needed to make them.
The demand side: Emergence of viral variants increases an ongoing need for vaccines
There is growing optimism that the United States has now secured sufficient doses to vaccinate every willing American adult by the end of May. Yet, this optimism must be tempered by the emergence of variants of SARS-CoV-2. The most worrisome are currently P.1, which emerged in Brazil, and B.1.351, which has spread rapidly through South Africa, despite the country’s earlier relative success in virus containment. The speed with which these variants are replacing previously circulating viruses demonstrates their potential to increase caseloads, hospitalizations, and deaths globally, including in the United States.
The emergence of variants of SARS-CoV-2 will greatly increase the demand for vaccine manufacturing for three reasons.
First, early data suggest that current SARS-CoV-2 vaccines may be less effective against the new variants and more manufacturing capacity is needed to produce the next generation vaccine candidates targeting the variants. The demand for that vaccine manufacturing will be greater if the active ingredients for addressing original and variant strains must be made separately and then combined in a bivalent vaccine, protecting against both the original and variant strains of the virus.
Second, countries will need to achieve higher vaccination rates to reach herd immunity—the point at which a sufficient share of the population has developed immunity so that, even absent mask wearing and social distancing, extensive transmission of the coronavirus can no longer be sustained.
Third, decreasing new infections is the only effective strategy for reducing the risk of variants. That will require increasing global vaccination rates. The longer the coronavirus spreads, the more likely it is to mutate and produce variants that are more contagious, deadly, or resistant to proven vaccines.
More vaccination requires manufacturing more doses of vaccines.
Obstacles to scaling out vaccine manufacturing globally
Quickly scaling up vaccine manufacturing during a pandemic requires policy support, as the American experience has shown. But the expansion cannot and should not be pursued by America alone. Even spreading production to other wealthy countries is unlikely to result in swift enough action to meet global vaccine needs. Some pharmaceutical companies, such as AstraZeneca and Novavax, have already engaged contract manufacturers in emerging-market economies. But there is even more production capacity to tap, and scaling up vaccine manufacturing in this pandemic will require its use. Finally, the most resilient supply chain for future pandemics will be a distributed one that can survive regional or single-country disruptions.
The Coalition for Epidemic Preparedness Innovations (CEPI), Gavi, and the COVAX facility have provided seed funding and led match-making efforts to convince manufacturers to start scaling out global production. The early agreement between Oxford/AstraZeneca and the Serum Institute of India—a company with the capacity to produce billions of doses annually—was the largest and most well-known. Others include Novavax, also a CEPI funding recipient, partnering with the Serum Institute. Novavax also contracted with Takeda in Japan, SK Bioscience in South Korea, Baxter in Germany, and Biofabri in Spain, conditional on its vaccine being approved. CSL in Australia has also agreed to manufacture the Oxford/AstraZeneca vaccine, with subsidies from the Australian government, after the home-grown University of Queensland candidate did not pan out.
Despite the efforts of CEPI and its partners, however, a shortage of inputs and the ongoing threat of export restrictions is impeding the necessary expansion of global vaccine production. Establishing new COVID-19 vaccine manufacturing sites also requires new supply chains to provide them with sufficient inputs—capital equipment, raw materials, and ancillary supplies (see again figure)—to make and deliver those vaccines. Bioreactor bags, lipids, and other inputs are already facing shortages, being used up in the United States and Europe. Specialty syringes are scarce in Japan.
Further scaling out global manufacturing requires the cooperation of multiple countries to subsidize the production capacity of outputs and inputs. One problem is that many of the countries with reliable contract vaccine manufacturers do not have all the necessary local companies to subsidize capacity expansion of needed inputs. Governments elsewhere may have the input-making companies, but in the absence of policy coordination, they don’t have the public health incentive to provide subsidies to reach the scale needed to satisfy global demand. Those input manufacturing countries would only enjoy the “externality“ benefits—i.e., solving their local public health crisis—if they were guaranteed access to other countries’ vaccine output through trade.
For that reason, the ongoing threat of “vaccine nationalism”—in the form of imposing export restrictions on vaccines—is thus another important factor discouraging the subsidization of input capacity. By limiting exports of locally produced vaccines, the United States, European Union, Italy, and India have established a worrisome precedent. The threat that countries with new manufacturing coming online might themselves deploy vaccine export restrictions creates an additional disincentive for other governments to subsidize critical input-providers.
Proposing a COVID-19 vaccine investment and trade agreement
A COVID-19 Vaccine Investment and Trade Agreement (CVITA) is needed to create the incentives to ensure the timely and sizable scaling up of output and input investments to respond to this pandemic and future pandemic threats. Baby steps toward such an agreement are found in the Trade and Health Initiative that a small, but influential, group of World Trade Organization (WTO) members proposed in late 2020. But much more is required.
First, CVITA should be aligned to leverage COVAX, the umbrella for the public and private international organizations that already have joined together for the purchase and distribution of vaccines. Linking the agreement to existing networks of regulators, such as the International Coalition of Medicines Regulatory Authorities, would also help ease concerns and create a more transparent pathway to the licensing of vaccines, instilling global confidence, reducing development costs, and expediting access in poorer markets.
Second, the investment component of the agreement must create a framework to subsidize the full vaccine manufacturing supply chain and especially coordinate expansion of input production capacity, including for bioreactors, bags, cellular materials, vials, stoppers, syringes, and other ancillary supplies. Governments would pay into the investment fund on a subscription basis. Participation of the poorest countries should be heavily subsidized or free.
Third, the agreement should include an enforceable commitment on the part of participating countries to not place export restrictions on supplies of vaccines and related materials destined for other countries participating in the agreement. In effect, subsidized imported inputs would be exchanged for future doses of an exported vaccine. Countries should agree that imposing export restrictions on vaccine output will be swiftly met with trading partners jointly restricting 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—will not work, because everyone will lose. Protections against export restrictions would also provide an incentive for nations to join the CVITA.
Fourth, this type of international policy cooperation demands unprecedented levels of transparency. Trust can only be maintained—decreasing the likelihood of hoarding—if access to information on COVID-19 vaccines and inputs reduces uncertainty. In response to dozens of countries imposing export restrictions on staples during a perceived food crisis in 2008-2011, the G20 created 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. A similar informative monitoring system for vaccines and inputs is needed under CVITA.
Fifth, CVITA needs an effective and transparent administrator who is one part general contractor and one part ombudsperson. When building a house, the general contractor is there to ensure the right inputs are available in enough supply at the right time. The electrician cannot install the wiring before the floors, beams, and rough construction are in place. On the other hand, if the sheet rock has already gone up, the plumber cannot install the pipes. Sometimes, the general contractor will move an extra plumber or electrician off one job so that a different job does not fall behind. At its best, Operation Warp Speed and the DPA were the general contractor the Americans used to help scale up investments in its entire domestic vaccine manufacturing supply chain.
CVITA needs some of that facilitation at the global level. With access to information, it can help coordinate capacity investment subsidies. It can also help address the reallocation of scarce inputs when inevitable bottlenecks materialize, potentially by creating secondary markets. This is critical to ensuring the production process stays on track. This facilitation mechanism can also recognize and prepare for inevitable frictions in scaling out global manufacturing. Shortages will occur. Tensions will rise. Because of scarcity problems, difficult choices will need to be made, and some may be asked to wait. Those challenges have to be resolved quickly, fairly, and transparently.
If global policymakers fail to launch and implement an enforceable COVID-19 Vaccine Investment and Trade Agreement, it will be a lost opportunity. If they succeed, lives will be saved.
1. Recent estimates are that governments thus far have paid $6-$40 for installed capacity yielding a course of vaccines that has a societal value of $5,800 (Castillo et al., Market design to accelerate COVID-19 vaccine supply, Science: Policy Forum, February 25, 2021.)
2. Developing enough vaccines and focusing them on the American population, as opposed to global public health needs, is a positive, not normative, characterization of the Trump administration’s “America First” policies. Elsewhere, we have made the normative arguments against such an approach. See Thomas J. Bollyky and Chad P. Bown, The Tragedy of Vaccine Nationalism: Only Cooperation Can End the Pandemic, Foreign Affairs 99, no. 5 (September/October 2020): 96-109; and Thomas J. Bollyky and Chad P. Bown, Vaccine Nationalism Will Prolong the Pandemic: A Global Problem Calls for Collective Action, Foreign Affairs, December 29, 2020.
3. Pfizer’s vaccine production in Europe took place in Puurs, Belgium.
4. These are sometimes referred to as a contract manufacturing organization (CMO) or even a contract development and manufacturing organization (CDMO).
5. BARDA was able to tap into prior relationships under its preparedness programs through its Centers for Innovation in Advanced Development and Manufacturing, which had been in place since 2012.
6. In addition to those listed in the table, there were separate contracts awarded on August 4, 2020 to procure needles and syringes for the Strategic National Stockpile from Duopross Meditech Corporation ($48 million), Cardinal Health Inc. ($15 million), Gold Coast Medical Supply, LP ($14 million), HTL STREFA Inc. ($12 million), Quality Impact, Inc. ($9 million), and Medline Industries, Inc. ($6 million) and on August 31, 2020 to procure syringes from Goldbelt Security, LLC ($125 million).
7. DPA actions that reallocated inputs may have also delayed production of vaccines purchased by COVAX or other countries, and it could have slowed the manufacture of other essential pharmaceutical products that were deprioritized. In January 2021, for example, Catalent, a contract manufacturer, was forced by the DPA to shift its facilities toward supporting the Moderna and Johnson & Johnson vaccine production and away from fulfilling a prior contract for a thyroid eye disease drug for the firm Horizon.
8. Given the rest of the world is still reeling from the “America First” approach of the Trump administration, it is unlikely to accept being held hostage to such an outcome anyway. It is only necessary to recall the 2020 shortages of personal protective gear and recognition that much of the world’s excess manufacturing capacity for personal protective equipment was located in China, the epicenter of the pandemic.
9. A factor impeding additional arrangements involves transferring know-how from vaccine sponsors to manufacturers. That necessitates the commitment of skilled personnel, time, and investment—all in short supply for firms, many of which were quite small and already stretched thin during the pandemic. Sponsors, especially of the newest mRNA vaccines, will require financial support and adequate assurances of intellectual property rights protection before they voluntarily transfer technology and tap unused contract vaccine manufacturing, even in well-regulated markets. Yet, fears that sponsors would not transfer their IP to manufacturers contributed to calls by countries like India and South Africa to waive IP protection during the pandemic. This approach is unlikely to address the many other bottlenecks preventing the scaling up of global manufacturing. (See Chad P. Bown and Soumaya Keynes, Is the WTO making it harder to end the pandemic? Trade Talks, Episode 150, February 25, 2021.)
10. This may have been less of a challenge in the United States in 2020 for two reasons. First, it had a large internal domestic manufacturing base. Second, since it was the first mover, it could tap into imports if additional inputs were needed, potentially reducing supplies available to second movers.
11. The strategy of vaccine manufacturing governments to subsidize imports of inputs is insufficient to tackle the problem, as this could only divert scarce supplies. Subsidies are needed to incentivize firms to invest in expanding input production capacity.
12. Or imported bulk vaccine, as in the case of Italy. Italy seized potential exports of 250,000 doses of AstraZeneca’s vaccine to Australia for which the contract manufacturer based near Rome (Catalent) was providing fill and finish services based on imports from Belgium or the Netherlands.
13. The United States, and to a lesser extent Europe and India, may have been able to get away with limiting vaccine exports because they had sufficiently large internal markets to provide the inputs needed for production at scale. Smaller countries dependent on cross border supply chains are unlikely to get away with it without consequences.
14. Ideally, this role would be informed when countries deploy their DPA equivalent—to prioritize input reallocation for vaccine production—so that those suffering from secondary effects can be quickly identified and inputs reallocated.