#862 2007 · Gavi, the Vaccine Alliance (with six donor governments and the Gates Foundation) · Global health / vaccine finance
Six donors pledged $1.5bn for a vaccine — payable only if it was built for the poorest
the problem
Pneumonia killed half a million children a year; a vaccine existed but was priced and designed for rich countries
background
Pneumococcal disease killed around half a million children under five every year, the leading vaccine-preventable cause of death in young children. A conjugate vaccine had existed since 2000, but it was tailored to developed-world strains and priced far beyond what poor-country immunisation programs could pay — so the countries with the highest death toll were the last to get it, typically a decade or more after rich ones.
The usual levers were push and patience: fund the research, or wait for prices to fall and negotiate bulk purchases afterwards. But the research was already done; what kept the vaccine out of poor countries was price and supply. No manufacturer would invest in the capacity and reformulation a developing-country vaccine required for buyers who could promise nothing — and post-hoc haggling had proved that point for years.
what everyone would do
The standard levers are push and patience: fund the research, or wait for prices to fall and then negotiate bulk purchases. But the research was done — a pneumococcal vaccine had been licensed since 2000 — and post-hoc haggling cannot persuade any manufacturer to build production capacity for markets that can promise nothing.
what they saw
The blocker wasn't science — a vaccine existed — but that no firm would build capacity for buyers who couldn't promise payment. The donors became that buyer in advance — for the right product at the right price.
the move
In 2007 six donors — Canada, Italy, Norway, Russia, the UK and the Bill & Melinda Gates Foundation — pledged $1.5 billion to an Advance Market Commitment run by Gavi: a binding promise to buy pneumococcal vaccine at a guaranteed price from any manufacturer that met WHO's target product profile and committed to supplying at a low, sustainable price once the AMC money ran out. The price was capped at $3.50 a dose, with a further $3.50 per-dose subsidy on the first 21% of doses under each contract. Pfizer and GSK signed supply agreements from 2010; India's Serum Institute joined in 2020.
why it works
Manufacturers invest in capacity against demand they can bank, not against need. By signing a binding purchase promise before a qualifying vaccine existed — open to any firm that met WHO's target product profile — the donors converted the poorest countries' invisible demand into a contract, and the price cap plus time-limited per-dose subsidy made early entry profitable while locking in a low sustainable price after the subsidy expired. Because the promise was a market rather than a grant, it rewarded exactly the behaviour aid money usually can't buy: building production and reformulating for poor-country strains ahead of any order.
the payoff
By end-2019, 215 million children were reached — coverage above the global average — and the pilot averted an estimated 700,000 deaths.
where it breaks
It fails when the product already exists — then the AMC is a subsidy to incumbents, which is largely what happened here: no research was accelerated. It fails when the buyer lacks cost data to negotiate the guarantee, as the $3.50 ceiling could likely have been lower. It fails when supply contracts rest on low forecasts: shortages of up to 29 million doses delayed 23 country introductions. And with no technology-transfer requirement, only two firms supplied the AMC for a decade — a developing-country manufacturer entered only in 2020.
what came after
The pilot became the template for COVAX, the pandemic vaccine facility — and the object lesson in the mechanism's limits. MSF's audit found it accelerated no research at all (the vaccine already existed), that 82% of the money went to the two incumbent manufacturers, and that under-forecast supply contracts left shortages of up to 29 million doses in 2012–14, delaying introductions in 23 countries. Gavi's own commissioned evaluations confirmed the weaknesses.
references
- [1]About the pneumococcal AMCGavi, the Vaccine Alliance, 2025gavi.org
- [2]Analysis and Critique of the Advance Market Commitment (AMC) for Pneumococcal Conjugate Vaccines — Executive SummaryMédecins Sans Frontières Access Campaign, 2020msfaccess.org