People with fashionable opinions have a tendency to become overconfident in those opinions. They rarely encounter serious objections to them, so they start believing that there aren’t any.

This explains, for example, the phenomenon of the clueless-yet-ultra-confident Twitter Marxist: they post some lame anti-capitalist cliché, get thousands of likes and retweets, and start thinking “Wow, I’m a genius!”

Perhaps paradoxically, something similar can happen at the opposite end of the popularity spectrum. If you have a moderately unpopular opinion, your opponents will at least feel the need to spell out what they think is wrong with it. But if your opinion is really unpopular, your opponents will no longer do that: they will just hysterically scream at you.

Thus, the person with the popular opinion and the person with the extremely unpopular opinion have one thing in common: they rarely encounter well-argued, reasonable objections to their position. This can lead both of them to believe that no such objections exist, making them prone to overconfidence. “I must be right – everyone agrees with me!” and “I must be right – that’s why my opponents are all so hysterical” ultimately lead to the same place.

I sometimes wonder whether there is a danger that my work on healthcare falls into the second category. I have been making the case for replacing the NHS with a market-based social health insurance (SHI) system for quite some time now, and I certainly can’t complain about a lack of interest – but I almost never come across serious attempts at a rebuttal. Virtually all the objections are along the lines of: “How dare you?!? The NHS literally saved my life!!” and “Bet you’ve been bribed by US healthcare corporations to write this garbage! You want to destroy our NHS, and leave people to die in a ditch for a quick buck!!” (see e.g. pp. 8-13 for a selection).

With that in mind – I am glad that the Nuffield Trust, a think tank which specialises in health policy, has now published a brief critique of SHI systems, as part of a “mythbuster” blog series written by their CEO Nigel Edwards.

It is a fair-minded and evidence-based good-faith critique, which raises valid objections. If we want to debate the pros and cons of different healthcare systems – this is exactly the way we should do it.

But while I approve of the format, I am nonetheless not convinced by the substance of Edwards’s argument.

Bismarck vs Pinochet?

Edwards is not an opponent of SHI systems per se. He accepts:

“These schemes […] generally provide a high level of cover to all […]

[H]igh-income countries with social health insurance (SHI) models […] are achieving better results than the UK on measures such as heart attack and cancer survival.”

But what he does not believe is that introducing such a system today, when you don’t already have one, would replicate those outcomes. He is particularly sceptical about the UK government’s ability to do this in the right way:

“[A] shift to social insurance in the UK […] could lead to a number of adverse changes […]

The risk is that in aiming for Otto von Bismarck’s venerable German system, we get General Pinochet’s”.

Some of those adverse changes he fears are “the exit of higher earners to private insurers” and “the departure of many NHS staff in areas with a high demand for private care”, leaving a “rump of NHS services providing poor services to the chronically ill and less well off”.

Is that a real risk?

Edwards appears to assume that in a British SHI system, statutory insurance would compete directly with conventional commercial insurance. That is imaginable, but it would be a very unusual arrangement. There are, as far as I know, only two OECD countries where that is the case, namely, Chile and Germany. (Otto von Bismarck’s system has that in common with General Pinochet’s.)

The main difference between SHI and conventional insurance is that under SHI, insurance premiums are not risk-rated, which means that people in good health (“good risks”) cross-subsidise people in poor health (“bad risks”). In a commercial health insurance system, there is no such systematic redistribution: good risks pay low premiums, and bad risks pay high premiums. SHI systems also involve varying degrees of redistribution from the rich to the poor, which commercial health insurance does not, or at least not on its own.

Both systems are internally consistent – but we can see why they do not sit easily side by side. If the rich and healthy can choose between an insurance option where they have to cross-subsidise the poor and sick, and an option where they don’t have to do that, the latter option will obviously be more lucrative for them. 

This is, in practice, not a huge problem in the German system – but it might very well be under an identical system in the UK.

However, I cannot see any reason why a British version of SHI would replicate this particular oddity of the German system, which has its roots in the politics of the Wilhelmine era. Otherwise similar systems in the Netherlands and Switzerland do not allow such opt-outs from SHI: there, the SHI system covers the entire population, and conventional commercial health insurance acts as a complement to it, not as a substitute. This is a more coherent arrangement, which is surely how you would do it if you introduced an SHI system today.

System-level reform vs reform within the system

Edwards seems to believe that the variation between different types of healthcare systems is less important than the variation within a given type. All systems come in good and bad versions, so system-level reform is a distraction: you are better off reforming the kind of system you already have. Specifically, he mentions some systems that are not too dissimilar from the NHS, and that achieve substantially better outcomes.

It is, of course, true that if you only look at health systems through a “Bismarck vs Beveridge”, a “single-payer vs multi-payer” or a “public vs private” lens, you are not going to get very far. I also agree that we should not discount reform options within the system.

But there are nonetheless some regularities. While Edwards manages to find some better examples of NHS-type systems, I notice that he does not find an example of a terrible SHI system. That is reassuring, given that this type of system has been around for well over a century, and covers well over 100m people today. Conversely, it is not hard to find other NHS-type systems that suffer from similar problems as the UK’s. Long waiting times, for example, are as much of an issue in the Mediterranean state-run health services and in New Zealand as they are here.

Either way – the pro-SHI argument was never that switching to a different funding model would mechanistically make everything better. Rather, the argument is that such a switch could be a catalyst for many other changes – including some of the ones that Edwards would like to see within the current system.

Too much disruption?

If we could wave a magic wand to replace the NHS with an SHI system, I suspect Edwards would not have a problem with that. But he fears that “the path to getting there would be very difficult. It […] would be hugely disruptive, likely unpopular among staff and perhaps the public, and would take many years.”

It certainly would take many years – which is why I have always resisted the temptation of claiming that switching to an SHI system could somehow act as a quick fix for whatever the NHS crisis of the day is. But when Edwards describes system change as “disruptive”, he apparently assumes that it would mean shutting down the current system, and building a new one from scratch. 

Admittedly, proponents of SHI systems usually fail to spell out a clear transition path: their idea of how to get from here to there is more like the “How to draw an owl” meme. But the general idea is to carve the new system out of the old one, or rather, to create conditions under which one system can morph into the other over time.

When the NHS’s “internal market” reforms were introduced in the early 1990s, critics complained that this would set the NHS on a slippery slope towards the creation of a marketised system. Those critics were wrong. We know that, because three decades have since passed, and it has still not happened. But the internal market structure could, in principle, be used in such a way. It is not a slippery slope, because there is no automatic slipping going on: you would still have to pull the sleigh. But the basic infrastructure for a smooth transition is there.

Conclusion

Nigel Edwards raises the sort of critical questions that proponents of SHI systems need to be able to answer. How is the transition supposed to work, in practice? Which complementary reforms are required to make such a change work? Which version of SHI are we talking about, given the differences between them? What is this supposed to achieve, and why can this not be achieved via reforms within the system we have?

But a “myth-buster”, it is not. There is nothing mythological about the failures of the NHS, and the superiority of SHI systems. They are very real indeed, whether we have the courage to admit it or not.