While the coronavirus has hit all countries, some of the world’s other serious diseases disproportionately afflict the poorest nations. And here lies the big challenge with finding a cure. The vast majority of ideas for new drugs fail, and even if a drug succeeds, it costs an average of nearly $3 billion to develop and gain approval (see Chapter 1 of the book). A pharmaceuticals company is unlikely to sink such expenditure if those helped by the drug are unable to pay for it.
The book cites the Mectizan Donation Program, where Merck took the extraordinary decision to give away a river blindness drug for free to some of the poorest countries in the world. This was an amazing humanitarian act, but companies clearly can’t do this with every drug. How can society encourage companies to find cures for diseases suffered by the world’s poor?
The easy answer is for governments and charities to give subsidies. And that’s the right answer. But, the much harder question is how to structure the subsidy. One common approach is financing the input of research – providing funds before the research has been actually carried out, to enable it to take place. This is indeed how grants to university professors are typically awarded. The main problem is that you’re providing funding based on the research proposal, and it’s very difficult to predict which proposals might actually succeed. Some of the most innovative ideas may seem crazy and infeasible at the proposal stage. Indeed, there are many famous examples of ground-breaking papers that, even when they were completed, got rejected by multiple academic journals because they were so novel. Imagine how negative the response would have been to the ideas at the conception stage, before the researchers even had any results.
Despite all the flaws with input-based funding, this is the predominant model for university research because there’s often no other option. If funds weren’t provided at the outset, the professor simply would have any money to conduct the research. But with large pharmaceuticals companies, that’s not the case. They have substantial funds generated by the profits of existing drugs, and can raise more through debt or equity markets. Indeed, healthcare firms are very willing to finance new medicines if the likely revenues are high (e.g. into a cholesterol-lowering drug, which addresses a developed-country problem). Unlike universities, the problem with companies isn’t that they can’t finance research to address developing-country diseases, but they don’t want to.
So a different form of subsidy might be particularly effective – funding the output of research. Here, once a drug has been successfully developed, you subside the price. This enables companies to sell to developing countries at a low price that they can afford, but – thanks to the subsidy – receive a high price that eventually recoups their substantial development costs. The key benefit is that it only rewards successful research. It doesn’t end up wasting funds on ideas that sound good on paper but are unimplementable in practice, or on professors that come up with great ideas but, after receiving funding, get distracted and spend more time consulting.
And output-based funding also addresses another problem – drug access. Even after a drug has been successfully developed, the challenge is to ensure that it is widely available. But companies have the incentive to underinvest in production – not only because building production capacity is expensive, but also keeping output low also allows them to set prices high.
For all these reasons, output-based funding sounds a good idea in theory. But does it work in practice? A new paper by Michael Kremer (co-winner of the 2019 Nobel Prize in Economics), Jonathan Levin (Dean of Stanford’s Graduate School of Business) and Christopher Snyder (Professor at Dartmouth) studies the effect of such a scheme. This involved the Gates Foundation and five countries pledging $1.5 billion in 2007 to develop a vaccine for pneumococcus – a respiratory disease that killed 700,000 children under five in developing countries each year.
Even after deciding to undertake output-, rather than input-based funding, there remain important design issues. A simple design would be to have firms compete for the subsidy, award it to one and subsidise each dose that it provides. But then the company has little incentive to expand its production capacity. It will get the entire $1.5 billion at some stage. Investing in more capacity to produce more doses only means that it will get the funds faster, but it won’t get any more funds since there’s a fixed $1.5 billion.
A companion model by Michael, Jonathan, and Christopher show that a better way is to tie the funds to production capacity commitments. Firms make bids that commit to supplying a certain amount of the drug each year. The annual need for pneumococcal vaccine is 200 million doses. So, a firm committing to supply X million doses per year would receive a X/200 share of the $1.5 billion fund. In return, they have to sell each dose for $3.50, which is close to cost. (The US Centers for Disease Control and Prevention pays $137). Overall, the scheme ensures that countries could buy the vaccines at close to cost, while the suppliers received sufficient revenue to encourage them to expand production.
How did the scheme work out? By 2016, 160 million doses of the vaccine were being distributed each year across 60 of the 73 countries covered by the programme. With three doses needed for a vaccination, over 50 million children are immunised each year. The researchers estimated that 700,000 lives have been saved. Note that the research doesn’t mean that AMCs are the only way to subsidise drugs. There remains a role for both input and output-based funding, the latter being particularly important where researchers need the funds for the research to be conducted. But, given the massive scale of health crises that the world faces, it’s useful to have many tools in our arsenal. And this particular tool can be used to encourage development of innovations outside medicine, such as new varieties of crops.