Ideas to Action:

Independent research for global prosperity


Global Health Policy Blog


On June 4 Gavi and partners launched a COVID-19 Vaccine AMC, one element of a broader facility for securing access to the vaccine(s) for all those in need, when these become available. As more information on their design and scope becomes available we will be delving into the details of these arrangements.

Not business as usual

A conventional R&D process takes 10 to 15 years to develop a new vaccine, progressing linearly from high-risk early research/pre-clinical development to clinical development and finally to manufacturing and commercialisation. Attrition rates are high and the probability of success relatively low—especially for newly emerging pathogens. As a result, the commercial returns to vaccine development are typically orders of magnitude lower than those for other fields of pharmaceutical development such as cancer drugs.

Figure 1. Vaccines research development cycle

A chart showing the vaccines research development cycle<br />

Source: ABPI via GSK site

But the global race to develop a COVID-19 vaccine is far from conventional. An effective vaccine would offer one of the few credible exit paths from the world’s worst public health and economic crisis in recent history. Scientists have called for a “new pandemic paradigm [for vaccine development], with a fast start and many steps executed in parallel before confirming a successful outcome of another step, hence resulting in elevated financial risk.”

Figure 2. Difference between traditional vaccine development and development using a pandemic paradigm

A chart showing the difference between traditional vaccine development and development using a pandemic paradigm<br />

Source: New England Journal of Medicine

The missing “development” in the R&D process

Several policy initiatives thus aim to accelerate the timeline for vaccine development and manufacturing; these include the US Operation Warp Speed, ACTA, and the Gavi AMC. Generally speaking, these initiatives support or rely on early-stage research through intensive “push” financing; and create incentives or offer direct financial support to build downstream manufacturing capacity, which would allow rapid vaccine production if and when a successful vaccine makes its way through phase two trials.

In addition, the authors of this blog have proposed a potentially complementary mechanism—the Benefit-Based AMC (BBAMC). The BBAMC proposal is distinct from other policy initiatives in its focus on “pull” incentives for product development. We do not assume that current push funding mechanisms will necessarily be sufficient to bring the safe and effective vaccines we need to market.

It’s important to question and critically examine whether our assumption is correct. Is pull funding for R&D feasible to design, or even needed during this unique global crisis—given the unprecedented mobilization of financial resources and political and corporate commitment to support product development and roll-out? Would the BBAMC solve for a major hurdle to vaccine development—or do we come offering a solution in search of a problem?

In the political and academic discourse, the challenges inherent in vaccine development have largely been overshadowed by an emphasis on the need to scale up manufacturing rapidly, to meet what is expected to be unprecedented volumes for a global roll out. The “speed premium” of scaling and distributing a vaccine as quickly as possible appears to serve as the underlying rationale for the aptly-named “Operation Warp Speed.” The pull and push incentives proposed for manufacturing capacity by the AcceleratingHT team—and planned for the Operation Warp Speed execution phase—differ substantially from the conventional pull mechanisms (e.g. pre-COVID-19 discussions on AMCs, market entry prizes, or impact funds) in that they focus on manufacturing scale-up and not on product development.

A flowchart showing the research and development process<br />

Do we even need pull funding for covid-19 vaccine development?

Ultimately, the question of whether pull funding is necessary to drive product development ought to be empirical. We could decide based on answers to questions such as: are venture capitalists investing in biotech start-ups with potential candidates? Are private sector developers pursuing vaccine candidates that have not received direct support from CEPI or bilateral funders in the US, UK or China? Are the multinational companies investing in multiple candidates within and outside their agreements with public-sector funding agencies? Are large pharmaceutical companies buying out smaller start-ups with promising candidates? Lots of “yes”-es would imply sufficient incentive (either from expected market or push funding or both).

In the absence of strong empirical evidence either way, we have heard three common arguments about why pull funding for development is not necessary, and we also lay out why we remain unconvinced with these arguments and think that pull funding is indeed necessary to get to a highly efficacious COVID-19 vaccine faster.

Push funding for early stage research can indeed be cheap and efficient; but the costs and risks grow. Pull funding matters more as market risk increases. Without significant private investment, there is a real risk that HIC taxpayers overpay through a non-transparent, bureaucrat-led process with limited accountability and at risk of the sunk costs fallacy.

The quantities of push funding for early candidates may completely remove industry early development risk; as a result, leaving no incentive to cut projects that no longer seem optimal, increasing the risk to payers of another Tamiflu-like episode.

Argument 1: The pharmaceutical industry expects a large (enough) and durable market for a successful vaccine.

There is no real commercial risk: High Income Country (HIC) markets, with the motivation to address the health and economic losses are significant, and ought to offer a sufficiently large incentive to mobilize the development investment.

Why We’re Unconvinced:

  • In our view, significant commercial uncertainty remains. Epidemiologically, the disease could either become endemic or fizzle away; here the SARS and MERS precedents are not encouraging. The economic recession could lead to drastic cuts in aid and domestic healthcare budgets that might lead governments to shirk their earlier commitments, while other technological breakthroughs (e.g. a cost-effective cure) could undermine the anticipated size and nature of a market. These uncertainties are more pronounced the longer it takes to develop a vaccine.
  • Although there is a high-income market, it is definitely not a “normally functioning one.” The deal frenzy between individual governments and single firms combined with push funding commitments, represent a form of R&D central planning with little historical precedent. For those “not picked,” the HIC market now looks rather uncertain, and for those interested in solving the problem and getting a vaccine, a policy of relying on governments to pick winners should be regarded with caution at best. Current speed-focused models imply a winner-takes-all approach, especially to those who are not in the lead and there is little discussion of what would happen to second (and perhaps better performing or safer) entrants. The US announced its five “picks” on June 3, triggering concern by scientists and regulatory experts. In terms of target performance, companies are working towards a fairly widely defined WHO Target Product Profile (TPP), without explicit incentives to address market or population specific needs, including differential safety, effectiveness and cost-effectiveness of different vaccine profiles in different populations and geographies. But such performance attributes may well make all the difference in how quickly and effectively we can control the pandemic.

Argument 2: Let’s use push funding all the way to a winner; it’s cheaper and more straightforward.

Given the significant amounts of public funding already committed to R&D (e.g. see here and here), it will be easier and cheaper to push through promising candidates all the way to market, including pushing and bearing the risk of manufacturing scale-up before a product is shown to work.

Why We’re Unconvinced:

  • Reliance on push alone means we need to be confident in the government’s ability to pick winners. Past experiences with picking winners this way have been underwhelming. Current portfolio optimisation models (e.g. here and here and several unpublished), not (yet) subjected to peer review, cannot be seen as a panacea given the paucity of generalisable experience in COVID-19 vaccine development.
  • Push funding for early stage research can indeed be cheap and efficient; but the costs and risks grow in Phase II B and beyond. Pull funding matters more as market risk increases. Without significant private investment, there is a real risk that HIC taxpayers overpay through a non-transparent, bureaucrat-led process with limited accountability and at risk of the sunk costs fallacy-i.e. “we paid for it to be developed so now we ought to buy and use it.”
  • The quantities of push funding for early candidates may completely remove industry early development risk; as a result, leaving no incentive to cut projects that no longer seem optimal, increasing the risk to payers of another Tamiflu-like episode.

Argument 3: We can rely on pharmaceutical companies to do the right thing.

Companies are vocal in their commitment to bring forward a vaccine to market and make it available at cost; these are extraordinary times when the profit motive does not (and/or should not) dominate decision-making.

Why We’re Unconvinced:

In our mind, credible pull incentives for development, alongside research push, are needed to further de-risk the development process for both payers (governments and donors) and suppliers (industry and investors), to crowd in private investments reducing the cost to HICs taxpayers, to share the risk of development between the government and the companies, and to inject payer accountability to the process.

Beyond COVID-19: solving for future challenges

The type of development pull incentives we are proposing—based on market commitments funded through countries’ health budgets with donor funding supporting procurement for the poorest countries—will allow governments to specify what they are looking for. It will also help to avoid them being locked in to vaccine candidates that look promising now but turn out to either be unsuccessful or of limited use (potency, durability, relevance to population sub-groups, or safety). At the same time, pull incentives could drive a significantly greater pool of private capital to invest in late stage development. Such a pull-driven model could more effectively address enduring global challenges of this nature (the next emerging pathogen, AMR, or TB) without large upfront allocations of taxpayer funds. This type of pull allows a larger group of countries to contribute to the costs and risks of vaccine development, and forms an able and arguably critical complement to any push funding that may be available, helping to broaden the pool of candidates and accelerate their development and manufacture.

At the same time, such pull arrangements acknowledge the value of innovation, committing governments and donors to pay for performance and moving the discussion away from development cost transparency, cost-plus pricing or IP licencing, strongly opposed to by industry leaders. In our proposal the risk of development is shared between payers and innovators: a successful company is paid, and paid more if it produces a better vaccine—even in the presence of say an effective treatment and second entrants, especially if better performers, can compete for a share of the market.  

We assert pull funding for the development part of the process matters offers a “belt and braces” approach for both payers and innovators, to the risky endeavour of rapid COVID-19 vaccine development. The recent decision by the US Government to narrow the shortlist of “winners” down to five raised the stakes significantly. We need a Plan B if those candidates fall short of expectations and aspirations. A credible pull mechanism would offer companies and governments precisely that.


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.

Photo credit for social media: