We all know very well that - well, we all should know very well that - a pure free market in pharmaceutical drugs simply will not work. It costs $2 billion to do the research and gain authorisation to sell such a drug. It costs perhaps $5 to make a dose of said drug - often enough 50 cents. So, if anyone and everyone could simply make the drug once approved then no one would spend the $2 billion. Therefore patents on drugs.

Yep, quite right, those who - or their health care systems - need the drugs right now are held to ransom to pay that $2 billion. The saving grace of the system being that the patent protections mean we do get new drugs. The patent lasts 20 years, it takes 10 years to get to authorisation, so after 10 years of being in use that patent lapses and everyone can make the drug for that 50 cents.

It’s possible that this is not the very best system possible. But the point to be made is that this clearly and obviously a constructed market. We have invented, out of whole cloth, that idea of ownership of the proof of authorisation in that patent. This isn’t a right that exists in nature, it’s something we’ve invented to try to make the system as a whole work better.

That is, people can be entirely right in claiming that good government can make markets better. Indeed this is true.

Almost a century on from the groundbreaking discovery of penicillin by Alexander Fleming, his scientific successors are racing to save modern medicine.

Infections that were once easy to cure with antibiotics are becoming untreatable, and a novel treatment for bacterial infection is the holy grail for teams of researchers around the world.

However, severe financial challenges have left the pipeline of new antibiotics thin and fragile – and treatments are unavailable in many of the places they are most needed. Big pharmaceutical companies have left the field in search of greater profits elsewhere, and talented researchers have opted for new jobs in more stable sectors.

This is a problem. The really big one here being that the financial incentives of that patent conflict, mightily, with the medical incentives. As and when someone has spent that $2 billion to get a new antibiotic to market we don’t want them selling the heck out of it for 10 years to make that money back. Instead, we want this new drug to be used as sparingly as possible - preferably never in fact - so that we don’t promote the evolution of the bugs to be resistant to it. We want that new and fully effective antibiotic to remain fully effective by not using it, or only using it when all else has failed.

Our constructed market doesn’t work for new antibiotics. OK then:

The issue needs incentives that push innovation, De Felice says, such as grants to support early-stage research from governments and the third sector.

It also needs incentives that pull drugs through to market and guarantee companies a return on their investment, even if the antibiotics are not used but held in reserve as a last resort for particularly severe infections.

Some of those programmes exist already. In the UK, drug companies can receive a fixed annual fee for new antibiotics regardless of how much they are used. The subscription model bases payments on how valuable the drugs are to the health system.

So, a new market is being constructed. This looks like it will cure that specific problem but of course it may not. So, continual innovation in the market itself is something we should at least hope to aim for.

The larger point here being that it is not true that the market fails. Rather, specific markets, given their structure and incentives, can fail. If that’s so then the correct answer is to construct another market with different incentives.

For example, we all know that Nick Stern comment upon climate change - the biggest market failure ever. At which point the Stern Review doesn’t say abandon markets, it recommends constructing a new market, with the correct incentives, by having a carbon tax.

Markets fail so use markets.

Of course, there are even some things that markets cannot handle. Say, keeping the French at bay so government should indeed run the Royal Navy. But these things are few and far between - unlike the beliefs of those who twitter on about strategic plans with strict conditionality.

Tim Worstall