Some economies have now vaccinated more than half of their populations against COVID-19 and are reopening, while low- and middle-income economies still have limited access in the face of devastating outbreaks.
Supply bottlenecks have been blamed. Though vaccine manufacturers report substantial capacity, essential vaccine manufacturing supplies like giant plastic bags and glass vials are hard to come by, understandably, as countries ordered more vaccines at one time than ever before.
However, these supply-side challenges are overemphasized. The reason why low- and middle-income countries are not further along in their vaccination campaigns comes down to insufficient demand. As Ruchir Agarwal of the IMF and I show in a recent research paper, even though governments have substantial experience implementing vaccination campaigns and most individuals are not hesitant to take vaccines, governments did not commit to buy Covid-19 vaccines from manufacturers early enough (Figure 1).
Figure 1: As of April 2021, despite available capacity for 10 vaccines showing effectiveness in Phase 3 trials, there were not enough advance purchases to cover the world’s population
Source: Agarwal and Reed (2021)
From a commercial perspective, advance purchases help vaccine manufacturers overcome three challenges.
The first is inventory risk. If vaccines are produced but not sold, increased inventory lowers the return on capital. For this reason, manufacturers are often unwilling to reserve capacity for customers in advance, since their ability to juggle capacity across customers is a source of profitability. When demand is stable, this works, and they can maintain just enough inventory to meet it. However, when demand spikes, advance purchases are used to activate additional capacity. For example, in the semiconductor industry, where there is currently a shortage of chips brought on by a boom in demand for durable goods that use them, the car maker Tesla recently made advance purchases of chips in a bid to secure supply.
The second challenge is access to credit. Even if a firm has a legally-binding commitment to buy, it may not have the working capital to buy inputs. Advance purchases can solve this problem by giving the cash up front. Even if cash is not given up front, banks will be more willing to lend when firms have commitments to buy. Kazaz, Webster and Yadav (2021) note the converse is also true--the offer of cheap credit won’t incentivize manufacturers to invest unless they have a committed buyer.
The third challenge is the discovery and resolution of supply bottlenecks. Once vaccine manufactures have commitments to buy, they can commit to buy from suppliers, and so on. For example, firms won’t find out how to get more bioreactor bags until they go looking for them.
Some governments have signed options contracts or memoranda of understanding without a legally-binding commitment to buy. These contracts do not resolve the challenges above. If the option price is less than the advance purchase price (or nothing is paid at all) and there is no guaranteed demand, manufactures delay investment and will not commit to making capacity available on a given date.
Castillo et al. (2021) showed that early investment in COVID-19 vaccine production capacity could be the highest-return public investment ever: “Capacity already in place, some of which was installed ‘at risk’ before clinical trials were completed, is more valuable than capacity that comes online later because it can produce vaccine courses without delay.” Advance purchase contracts are the way to invest ‘at risk’ in capacity.
Operation Warp Speed in the United States accelerated vaccinations by purchasing hundreds of millions of doses as early as July 2020---before emergency use authorization. At the same time, the COVAX vaccine cooperative established its Advance Market Commitment facility to subsidize advanced purchases---before emergency use authorization---and organize procurement for at least 91 of the lowest-income countries..
In our April 2021 paper, we showed that the COVAX facility still needed at least $4 billion in grant funding to buy vaccines for these 91 countries, and that more advance purchases would be required from middle-income countries, including China and India. After a June 2nd summit, COVAX had secured $3.3 billion of the $4 billion we advocated for, bringing its total fundraising to $9.6 billion. Assuming slightly lower prices (or greater beneficiary co-pays) than in our original model, this could be enough to vaccinate 30% of the population in the 91 eligible countries. An additional 30% could be covered by direct procurement efforts and gifts (donations) of advance purchases already made by high-income countries more than their populations. For instance, the African Union has secured an option to vaccinate 30% of its population with the Johnson and Johnson vaccine; recently announced donations by the G7 (at least 800 million doses) would easily be enough to cover 390 million people, at two doses per person, or 30% of the population in the countries not inside the African Union eligible for subsidies under COVAX.
Together, existing contracts and donations appear sufficient to vaccinating 60% of every population, which is the recommendation of the Africa CDC and a mid-point of the theoretical threshold for `herd immunity,’ the point at which the spread of the disease from person to person becomes unlikely and, as a result, the whole population becomes protected — not just those who are immune (our paper describes the formula, and what it depends on). Estimates of available capacity suggest enough vaccines to meet this goal could be delivered by mid-2022 if advance purchases are placed immediately to activate capacity.
This achievement is cause for optimism and not complacency, which there has already been enough of. If $10 billion had all been given one year ago, instead of today, advance purchases could have enabled production to scale earlier, accelerating vaccination campaigns.
Direct procurement efforts also face challenges. In early May, only 5 of 55 African Union members had paid a 15% deposit to secure vaccines through the African Vaccine Acquisition Trust, even though a lending facility has been established to cover the remaining cost. Further, as we describe in our paper, vaccine loan facilities established by multilateral development banks require vaccines to be authorized for emergency use before purchase. This requirement may reduce risk that ineffective vaccines are procured, but it also means these facilities could not fund early ‘at risk’ investment in capacity, in the style of Operation Warp Speed or COVAX.
Greater transparency over each country’s place in line under existing contracts would facilitate gifts and allow an assessment of whether too many orders are concentrated on a single manufacturer, a risk we highlighted in our paper. For example, COVAX has planned to rely heavily on production of the AstraZeneca vaccine in India, which has now banned exports and may not permit them until 2022. COVAX could accelerate deliveries by making advance purchases of the Sinopharm and Sinovac vaccines, which the WHO declared safe and effective when authorizing them for emergency use. These vaccines are produced in China, which has a different input supply network than India, and has consistently exported half of production as it controlled domestic outbreaks. A global emergency is the time to exploit rather than avoid a diversified international supply chain.
In hindsight, COVAX could have diversified the risk of delay by buying vaccines from China as well as India in advance of WHO authorization. It is not clear that Sinopharm and Sinovac were worse bets either, since they both use an older and well-understood inactivated virus technology. Today, Sinopharm has proven more effective than the AstraZeneca, and neither vaccine has been linked to blood clots as is the case of vaccines using the alternative adenovirus technology like AstraZeneca and Johnson & Johnson. Just as COVAX has signed advanced purchase agreements for Novavax---which has not yet received emergency use authorization---COVAX could also purchase CanSino, which requires only one shot and appears just as effective as Johnson & Johnson based on interim data from Phase 3 trials. The three Chinese vaccines have already been authorized by numerous jurisdictions in Africa, Asia, and Latin America.
Inclusive of the $4 billion proposed we proposed in April, the IMF estimated a total $50 billion is needed to end the pandemic, in order to also cover testing and treatment in the interim period when vaccine supply is limited, and to buy additional vaccines as insurance against the event a more transmissible variant pushes the herd immunity threshold higher.
To prepare for future pandemics, an envelope of at least $10 billion should always be available to a facility like COVAX empowered to make advance purchases of vaccines before regulatory approval, enabling Operation Warp Speed for the poorest countries. In this pandemic, donors realized this one year too late.
Some have argued that a waiver of vaccine patents could end the pandemic by allowing new manufacturers to enter the market, further diversifying supply. This might be true, though just as with existing supply governments would need to commit to buy from those manufacturers.
In the long-run, you get what you pay for. Supply bottlenecks may delay the path to global vaccine equity, but without enough demand, we will never reach the goal.
Join the Conversation