We seem to have something of a break through in the treatment of peanut allergies:

The first medical treatment for children with peanut allergies is likely to be approved next year but there are concerns about its affordability, even though it consists essentially of peanut flour.

That the treatment exists is just lovely. It’s essentially Mithridates all over again - minor and increasing doses to build tolerance. The idea’s not new - indeed one of us knows a child having this treatment right now.

The complaints about cost have surface validity. It’s just peanut flour so why does it cost £17,000 a year? The answer being because someone had to test this old idea. Lots of people in white coats had to spend years of their lives working on whether it worked, how it worked and whether they could prove it worked.

It’s most appealing to look at the cost of sales of some new thing and demand that some reasonable margin above that is enough. But to do so is to ignore those development costs. Yes, obviously, development costs of something that exists are sunk costs but this is our public goods problem. Non-recovery of sunk costs won’t change, one iota, the value of this peanut allergy cure. But they’ll hugely affect any future intention to invest in solving another of man’s ailments.

The argument that all of this should be publicly funded also doesn’t change the basic calculation. It’ll not even change who pays. If the NHS buys privately developed drugs then the taxpayer pays. If government paid scientists develop the drug then the taxpayer pays. The argument necessarily devolves down to efficiency, a state run bureaucracy does rug development better or worse than a profit driven private sector?

Someone, somewhere, has to pay those development costs. If we use the private sector to do the development then the private sector must be paid those costs - or we’ll not have future private development.

There is no way out of this, no cute trick. Costs of development are costs and they’ve got to be paid somehow.