"Is the NHS in a Death Spiral?"- Responses

One of the joys of doing this substack is the response to pieces, in the comments and by email, particularly those on complex policy issues. We must surely have the most thoughtful comment section on the internet. This was more true than ever for my NHS post last week, given the large number of you that work in or adjacent to the health service. It felt worthy of a follow-up piece.

I haven’t tried to discuss every interesting comment – or this post would be 20,000 words long, but I have picked up on challenges and suggestions that I thought raised questions worthy of further consideration. I’ve sorted these into five categories:

1. Privatisation and Treasury Brain (or whose fault is this?)

2. Social Insurance (did I gloss over it too quickly?)

3. Efficiency vs Resilience (how do we draw broader lessons from this?)

4. Baumol’s Cost Disease (a bit more of an explainer)

5. Sorting the little things….

1.       Privatisation and Treasury Brain

Some of you felt I was being too lenient on the Government when I dismissed the idea that this crisis was part of a plan to privatise the NHS. Nobody – well none of you anyway – thinks there’s some shadowy “deep state” conspiracy waiting to be uncovered. That would require a previously hidden level of competence from the Conservative party. But as an emailer pointed out:

“I wonder whether those advancing the ‘Tories are conspiring to run down the NHS to privatise it’ argument you dismiss up top are really arguing the scenario you posit toward the end:

‘The alternative to fixing it is not total collapse but what’s happened with dentistry – a messy shift to an unplanned two tier system whereby anyone who can afford to do so buys health insurance (with private emergency care emerging over time too) and the NHS is left as a provider of last resort, with politicians finding it ever harder to persuade voters who don’t use it to fund it to an acceptable level.’”

Or as Simon Frankau put it:

“I believe in not ‘a conspiracy to run down the NHS in order to privatise it’, but a plan to drive a substantial fraction of healthcare private by gently running down the NHS. As you say, they did this for dentistry, so why not the rest of the healthcare system? Hubris meant that they assumed a much larger, much more complicated and much more important set of systems would behave in the same way. And here we are.”

In a way it doesn’t really matter why we’re in such a mess, it’s the government’s responsibility whether they wanted us to be in this position or not. But I think distinguishing between an intentional desire to shift to a different system and incompetence does matter in fixing the problem. There are certainly plenty of people in the Tory party who think we should explore moving to an insurance-based and/or privately run system, but these people tend to think the government has been too generous to the NHS. Indeed there is widespread frustration in the meeting rooms of right-wing think-tanks that Ministers have consistently refused to engage with the politics of what they would see as genuine reform.

The government’s position, throughout the last 13 years, has been to desperately try and avoid the NHS becoming a political issue at all. They know full well voters don’t trust them on the topic and they want to talk about almost anything else. They have, when crises emerge, put more money into the system – especially since 2019. Department of Health spending this year will be £26 billion higher than 2019 in real terms. And as I set out in the original post staffing has increased a lot. The problem is that this funding has always come in response to things deteriorating and so been distributed in an unstrategic way. It would have been far better to increase spending more during the previous decade and invest it in infrastructure, social care and so on and thus avoid such a big crisis now. What we’ve seen in the last few years though is not the behaviour of an administration actively trying to force the system towards unplanned chaotic privatisation.

This is why I’ve found the government’s handling of the pay dispute with nurses and paramedics so baffling. If you wanted to downplay the issue, and reduce the sense of crisis, you’d do a one-off payment asap and focus your “union paymasters” row on Mick Lynch and the RMT (even though the RMT isn’t actually affliated to Labour). From reports it seems that the Treasury are insisting this can only happen if the money comes from elsewhere in the Department of Health budget. If true it is a classic example of Treasury Brain – with an overfocus on cost control distorting cross-government prioritisation.

The role of Treasury Brain, rather than deliberate political strategy, in causing our current troubles was underlined by a comment from John Gieve, formerly an extremely senior civil servant at the Treasury and deputy governor of the Bank of England, and now the Chair of an NHS Trust:

“I agree that the system lacks the spare capacity to handle even mild variations in demand (or supply like the change in EU workforce). That reflects a long period in which high utilisation was seen as a measure of efficiency and perhaps also a belief at the centre that activity would rise to meet any increase in capacity. But it has also reflected a prolonged squeeze on capital spending (alongside health education, social care and public health) in order to maximise the current budget for the NHS which is seen as the top political priority. In successive Spending Reviews and Budgets the NHSE and DH have pressed for the maximum headline increase they can and the Treasury has sought to offset the increases by squeezing the separate more vulnerable budgets. While everyone can see that is a mistake, they can't find a way out and, of course, finding funds for a bigger pay increase will exacerbate the problem.” (My bolded italics)

As I’ve written before I think the split we have in government between a No. 10 responsible for political prioritisation and a Treasury responsible for cost control makes no sense and will continually lead to this kind of situation. I’ve seen the same thing happen with education where the Secretary of State and the Prime Minister always want to protect the headline schools budget and the Treasury insist this is paid for by cutting, amongst other things, post-16 provision. So we now have the ridiculous situation of a child getting roughly the same amount of money spent on them from 5 to 16 as they did in 2010, and then 25% less from 16 to 19. To be clear I don’t think this is the fault of the people working in the Treasury – this is what they are asked to do. No one in government ever decided to push the system in this direction, it’s a function of the way spending reviews are done, the structure of government, and an absence of strategic oversight from the centre.

2.       Social Insurance

A number of those commenting felt I was too quick to dismiss the case for shifting to a social insurance model. The point I was keen to make in the post is that those who tend to argue for such a shift are usually starting from the wrong premise – that the NHS is expensive and inefficient and that is a function of it being fully “nationalised”. The opposite is true – social insurance systems have an additional layer of bureaucracy and less rationing so they are more expensive and spend more on administration. However, I noted one potential and substantial benefit – it would probably be easier to get more money for our health system if people felt additional payment went directly to their care rather than general taxation. John Gieve agrees. After noting that the political nature of health spending in the UK means fluctuations are not related to actual demand he says:

“Of course introducing a social insurance structure would add a layer of admin but, in principle, it might produce a more stable pattern of funding which was more related to underlying demand and thus would provide a better basis for planning capacity, workforce, and investment over the long term.”

Daniel Alex made a similar point:

“The problem with ‘publicly accountable services’ [is that] public accountability doesn't seem to make for very good decisions a lot of the time. You say that social insurance would at least have the advantage of being able to charge rich people more; surely the other advantage is that it might take the system away from us electors and our skewed priorities (e.g. our fixation on hospital medicine). A bit of a counsel of despair but perhaps realistic.”

I don’t disagree with this. If I could get in a time machine and go back to the 1940s I’d recommend starting from a different place. But given where we are the costs of shifting to a social insurance system seem insurmountable – both in terms of the time and resources to manage such a huge upheaval in the midst of crisis and the politics, which would be impossible for Labour and horrible for the Tories. If that is the case then it is not something upon which it is worth wasting time. Moreover levels of patient satisfaction under New Labour show that it is possible for the NHS to work, it just requires a very high level of prioritisation.

This may change if my worst case scenario happens and we drift towards a two tier system where increasingly large numbers of people have private insurance. Mark Segal noted that:

“You mention Canada - I don't know how nuanced the ‘debate’ is in the UK, but in Canada it is not; it's largely about whether there should be more privatization, with little being said about whether the existing system can be reconfigured to work better.”

And Jack Smith replied:

“I think another reason why the debate tilts in the binary private v. public direction in Canada is because approx. 2/3 Canadians have complementary health insurance, usually funded by their employer. That, I think, introduces a private component to the system which makes people more comfortable with it.

In my opinion, having lived in both countries, I actually think the UK’s relative lack of private insurance explains why it’s such a taboo - the implicit comparison is the US rather than a more private-tilted version of your own system. I’d expect the same debate as in Canada to happen more in the UK, years or decades down the line if the NHS continues to deteriorate and employers offer complementary insurance as a benefit.”

That makes sense to me.

3.       Efficiency vs Resilience

The most popular comment on the post came from Simon Frankau – who works at Google doing clever things that I don’t understand:

“As someone who works on the reliability of complex technical systems, I found the approach of this analysis very interesting, partly for what it doesn't cover. [Queuing theory] seems key to the situation: The idea that a small increase in utilisation of a system near capacity can lead to an unbounded increase in waiting times. When your system's at the edge, tiny changes can have a huge effect.

People tend to not only think that systems behave linearly on the edge, but not realise that efficiency costs responsiveness. We can see this in the government's focus on NHS "efficiency", while failing to understand that necessarily increases waiting times. "Slack" is not inefficiency, it's not even emergency capacity, it's a required component of any effective, responsive system. To compare, it's not unusual for large computer systems to run at 50% capacity on their time-sensitive workloads, to avoid queue build-up.

There's a lot of other useful research and ideas around these systemic failures, including resilience, cascading failure and metastability. The NHS, built on people who care, is the kind of system that can survive an impressive amount of damage without showing much of it. The government can cut too much for too long (even if ‘cut’ is chronic underinvestment), and people try to make up the gap, so the feedback loop shows ‘these cuts are fine’. One area can attempt to cover for another, so damage spreads - we end up with damage to social care eventually breaking urgent care. There's a theme of ‘normalisation of deviance’, where the system is pushed outside of its safe operating envelope, no severe problems are seen, and this becomes the new (unsafe) normal, and risks pile up. In short, by the time the damage really shows, it's far too late and reversing it becomes extremely hard.”

I can’t pretend to be an expert in queuing theory but Simon is absolutely right that systems can suddenly go over a tipping point if warning signs are ignored for too long. The issue of efficiency vs resilience in the public sector is one that only really started to get widely discussed (in wonk circles) after covid highlighted the problem of “running systems hot”. And I don’t think we have even begun to develop a language around this that would resonate with voters.

Waste on the other hand is a very intuitive and well established concept which no one could claim to support. Of course there is actual waste in any large bureaucracy that does not contribute to capacity. But how do we redefine excess capacity as necessary and not wasteful? I’ve been reading a lot about Thatcherism recently (for a future post) and in the late 70s one of the complaints about the NHS was there were too many empty beds! I don’t know the answer but I guess it’s a positive that at least it’s starting to be discussed. In the meantime, though, I worry that other bits of the public sector are not far away from their equivalent tipping points, even though they seem to be coping now.

4.       Baumol’s Cost Disease

An interesting question via email from a former senior civil servant:

“I wonder: do you think there’s a limit (probably sadly a long way away) to the Baumol effect? Presumably at some stage, some sectors are *so* productive that they cap out the number of people they need. Then, even if those people are highly paid and the jobs are in demand, there aren’t many of them, so there’s abundant labour, so… I presume? the low productivity / high labour sectors get to pay less? Or, put another way, AI making the world super productive could = take loads of jobs = lots of labour fighting for the remaining roles?”

A quick reminder that Baumol’s cost disease is the idea that the sectors that aren’t getting more productive are often those involving a lot of human contact. They see costs going up because they have to compete for staff against sectors that are increasingly productive. Baumol came up with his theory in the 1960s and so far it has been extremely predictive. Healthcare, education, childcare etc have all become more expensive while manufactured goods have become far cheaper. In rich countries vastly more people are now employed in services than in manufacturing.

Not being an economist I don’t know if there is a good answer to the question above (suggestions in the comments welcome). But I do think that the main issue with Baumol is bringing ourselves to accept it’s not really a problem. The costs of caring professions will keep increasing for the foreseeable future but this is because others things are getting cheaper. As Baumol himself wrote in 2012 – at the age of 90(!):

“Despite their ever increasing costs, stagnant-sector services [he gives healthcare and education as examples] will never become unaffordable to society. This is because the economy’s constantly growing productivity simultaneously increases the community’s overall purchasing power and makes for ever improving overall living standards….

…This is not merely naïve optimism but something we have already experienced. The exploding cost of hospital care and galloping college tuition increases since the Second World War have not prevented us from consuming these and other services and goods. Indeed, we now live longer than ever, and a continually rising share of the population attends college. Instead, the true threat to this desirable future is foolish public policy.”

Well quite. We have to stop pretending that it’s possible for the costs of these services not to grow relative to others. The more we try to suppress costs to current levels or below the worse services will get. And that risks growth (look at the increase in numbers on benefits who are too ill to work).

Ignoring Baumol leads to one type of foolish public policy discussion in which reform is seen as being an alternative to additional money when, in fact you cannot do one without the other. As one hospital director put it in an email:

“Within local areas, hospitals are often the bit of the system with both the most capacity to enable change but also the most flashing red lights. It is easy for them to dominate by default by merit of being the greatest pool of clinical staff to work on reform. However, decent meaningful reform takes a lot of clinical engagement and currently staff don’t have the time to do this. Executives are understandably focused on keeping emergency pathways safe. Hospitals and systems that are being long sighted invest in longer term change but it is hard to do so when central govt is focusing more than ever on performance management of the existing model in the run up to the next election.

We are fundamentally asking the system to do too much at once. Running high volumes through the existing model whilst making it less efficient (due to the discharge exit block). Degrading the intrinsic capacity though underfunding of estates and workforce. Imposing structural change and requiring root and branch reform when there is no protected capacity to do so. What is required are choices about priorities - but none of them are easy ones. Should we deliver even fewer hip operations so that we can invest our clinical time in a reformed model of care that won’t deliver benefits for a couple of years? Might be right for the patient but what if it means we get less money next year and more scrutiny for not hitting targets? Etc”

5.       Sorting the little things….

A few people mentioned a couple of quick fixes to help with staffing. Sorting the pensions caps that are pushing higher paid doctors to reduce hours or retire early seems an obvious one. Labour have said they will do this but as ever with tax changes it’s not at all straightforward, even if there’s little doubt it’s a problem.

On the other hand – fixing GP visas seems eminently doable and not that complicated. As the FT explained last week:

“Newly qualified GPs from overseas face a unique bureaucratic hurdle because they are required to obtain a two-year ‘Tier 2’ skilled worker visa after their three-year training visa expires, putting additional burdens on GP practices. The issue does not arise for other medical specialities, which take at least five years, after which all newly qualified doctors can apply for permanent status.”

Coming full circle back to questions of government responsibility. This is the real problem – not a hidden desire to foist right-wing ideas on us all – but a fundamental lack of competence and ability to manage a crisis properly. Even relatively small measures that would make a difference, are left undone. That’s what we should be angry about.