Netflix

standard historian's explanation of this trend is that once science became a profession rather than the province of rich amateurs, prizes were no longer a suitable way of funding innovation. Hanson is not convinced. "Most academics who study the issue of prizes have focused on what a prize does to the behaviour of researchers, versus a grant,'' he says. "But there's another aspect: what does the person giving the prize or the grant get out of it?''

He argues that grants are more appealing than prizes to bureaucracies for many reasons, not all admirable: "With grants, there's all sorts of possible patronage and corruption.'' Even leaving aside outright graft, there is plenty of opportunity for cosiness and cliques. Then there is the mundane fact that grants are easier to account for in an annual budget than a multi-million prize that could be paid tomorrow, in a year, or never. For Hanson, it was for these reasons, rather than any intrinsic merits, that grants elbowed aside prizes in the 19th century.

Prizes may be making a comeback because of all the money now available from private foundations - which demand results. Not only the X Prizes and the Millennium problems prize, but even the pneumococcal vaccine prize is part-funded by private money. Yet governments are getting in on the act. The US space and defence research agencies Nasa and Darpa both use innovation prizes, and other government agencies look likely to follow with, for example, an "H prize'' for advances in hydrogen fuel technology.

If Hanson is right, this new trend is a welcome swing of the pendulum towards a modest use of prizes. But not everyone is convinced that prizes will live up to the hype.

"The literature has pushed them as a silver bullet; more recently there's been a bit more sobriety in the debate,'' warns Andrew Farlow, an expert in the economics of vaccines at Oxford University. "How much genuine risk-taking can it pull along?''

The problem is not the principle, he argues, but the details. A vaccine for HIV is a distant and costly prospect, and might require a $10bn or $20bn prize. Inevitably, companies and their shareholders will question whether the prize would be honoured in full. The triggers for releasing some of the prize money are difficult to define: early vaccines would probably be expensive, fallible and risky, but better than nothing. Donors would not want all the money to go to those efforts and leave none to encourage superior successors. Try framing "good enough'' in legalese, when billions are at stake.

Donors might pay a lot more than they needed to for a substandard product, or the prize might be too restrictive and too small to generate any interest at all. That would drain attention, enthusiasm and political will. "It all sounds like good economics, but whether you could ever set a prize big enough or correct enough to work in those cases is doubtful,'' Farlow concludes.

But