NHS underspends: whose money is it anyway?


140 characters is not sufficient to communicate complicated arguments. I realised this yesterday, when Crispin Dowler (of the Health Service Journal) responded to my tweet – in which I said that the NHS does not have a policy to generate a surplus – with a link to national NHS guidance entitled ‘surplus policy’.

That exchange was connected to a wider issue I have with the HSJ’s characterisation of the NHS’s – and the Department of Health’s – surplus.  

Simply put, the HSJ argues that the Department of Health’s surplus (or underspend) every year is handed back to the Treasury and lost to the NHS for good.  I, on the other hand, side with the Department of Health’s view that the vast majority of the surplus remains in NHS bank accounts, and remains available to the NHS to spend in future years.

Superficially, it would seem that we cannot both be right. But, in fact, we both are.  And so here, in a rather more generous 13,000 characters, is my attempted explanation – an explanation which I apologise in advance for its fiendish complexity.

My view and the HSJ’s view differ because they derive from looking at the same NHS finances through different ends of the telescope: my view is derived from the way in which cash flows around the NHS; and their view is derived from the way the Government as a whole accounts for its spending.

Let me explain the HSJ’s argument first.

Under the system of government accounting, Government departments are set a ‘control total’ of how much they can spend (the ‘Departmental Expenditure Limit’ or ‘DEL’) (OK – there is also a control total called ‘Annually Managed Expenditure’, but it isn’t particularly relevant to the NHS or the Department).  A Government department is given a DEL every year – let us say (for ease of explanation), that for the Department of Health it is a fixed £100 billion, each year and every year. An overspend against this DEL would be a very serious issue – indeed, the Accountable Officer would expect a severe reprimand, or even worse, the sack. So the Department of Health never overspends. But it cannot land expenditure every year exactly on £100 billion, and so it underspends instead. Let us say that the Department of Health underspends by £5 billion in year 1 – it spends £95 billion. However, in year 2 the Department of Health’s DEL is fixed at £100 billion: it cannot make use of the £5 billion it has spare from year 1 – and so this £5 billion, the HSJ would argue, is simply handed back to the Treasury and lost to the NHS for good. And looking at it from their end of the telescope, they are right.

I am hoping you will be with me so far. Now let’s look through the other end of the telescope – from the perspective of an NHS provider, rather than the Government.

For simplicity, imagine that the NHS is comprised only of 10 providers – or ‘NHS trusts’ (ignore commissioners for now – we’ll play them back in later). And now imagine that the £100 billion in year 1 is gifted to each of them in equal shares every year – so at the beginning of year 1, they are gifted £10 billion each in their bank accounts. Now during year 1, each of these NHS trusts is remarkably prudent – each saving 5%, or £500 million, on their initial £10 billion. And so altogether, these 10 NHS trusts have used £95 billion and have £5 billion sitting in their bank accounts as a surplus at the end of year 1.

Because Government accounting works on the basis of when cash is used, rather than when cash is handed out, this looks like £95 billion of spending in the public accounts in year 1 – £95 billion of spending against DEL. However, it is still £100 billion that has been handed out – £5 billion of which has yet to be used. At the beginning of year 2, therefore, each NHS trust has £10.5 billion. Together, the 10 NHS trusts look in their bank accounts and see that they have £105 billion to spend. This is why I argue – in contrast to the HSJ – that the NHS surplus is not lost to the Treasury, but is available to the NHS to spend in the future.

Now let’s marry these two systems up.

In year 2, if the 10 NHS trusts spend all of the money in their bank accounts – which they have every legal right to do – they spend £105 billion. For the Department of Health, this creates a headache because their £100 billion DEL is busted. But, fundamentally, the Department of Health has no legal powers to stop NHS trusts spending their money. It has the responsibility, but it has no power.

What does the Department of Health do in this situation? All it seems it can do is hope – and hope for two things in particular. 

The first is for NHS trusts not to spend all of their surpluses at once – because this would bust the DEL by £5 billion. And the second hope is that if some NHS trusts do spend their year 1 surpluses in year 2, these are offset by other NHS trusts underspending in year 2.

And, remarkably, hope has been good enough for a number of years. Because with hundreds of organisations in the NHS rather than ten, the number or organisations running small surpluses and the number of organisations running small deficits roughly help to balance each other out.

Only, they don’t quite. Because instead, the NHS has run relatively significant (£1 billion+) surpluses for a number of years. And these historical surpluses are now all stacked on top of each other, sitting in NHS bank accounts.

In the event that NHS organisations decided to spend all of these historical surpluses at once, the Department of Health’s DEL would be blown apart. But thankfully, the Department of Health’s finance team has three routes (other than hope) to offset the risk to their DEL.

The first is something called ‘budget exchange’. This is when a predicted underspend on DEL in year 1 is removed and added to year 2. So, in our example, the £5 billion underspend in year 1 could be deleted from the year 1 DEL (which becomes £95 billion) and added to the year 2 DEL (which becomes £105 billion) – which looks like it should remove the problem entirely. However, it doesn’t quite work like this because, to apply the budget exchange, the Department of Health would have to have 100% certainty that its prediction is accurate (if it predicted a £5 billion underspend half-way through year 1, which actually turned out to be only a £2.5 billion underspend at the end of year 1, then its revised DEL would be busted). And so in practice, the budget exchange is applied to a conservative estimate of what the Department of Health believes the underspend to be.

Let us say then that the Department of Health applied a budget exchange of £2.5 billion – meaning that its DEL in year 1 becomes £97.5 billion and its DEL in year 2 becomes £102.5 billion – but that the NHS did indeed underspend in year 1 by the full £5 billion. This means that NHS trusts still have £105 billion in their bank accounts at the beginning of year 2, but the Department of Health has a DEL of only £102.5 billion. And so the Department of Health is exposed to less risk – but still £2.5 billion of risk nonetheless.

The second thing the Department of Health can do is try and understand whether the risk will be realised. And so it looks at the spending plans of NHS organisations and estimates how many of them will access their historical surpluses (or indeed borrow from third parties, which also scores against DEL – but for simplicity let’s leave borrowing out of this). And so let us say that 8 of the 10 NHS trusts are planning to access all of their £500 million year 1 surplus in year 2, but 2 of them are planning to underspend by another £500 million in year 2. This means that the 8 surplus-accessing NHS trusts will spend a total of £84 billion (remember they are each gifted another £10 billion in year 2), but the 2 surplus-building NHS trusts will spend only £19 billion – giving a total spend in year 2 of £103 billion.

Unfortunately for the Department of Health, this still leaves a gap between the amount the NHS wants to spend in year 2 (£103 billion) and its year 2 DEL (£102.5 billion) of £500 million. And so it uses the third tool at its disposal – it can cut the budgets it controls directly by £500 million to offset the risk to its DEL.

Now, the Department of Health does not actually control a huge amount of money directly – but it does have levers at its disposal over one category of budget holders:  NHS commissioners. And this is where we can play them back in. Because the Department of Health (or, strictly, the NHS Commissioning Board) can order commissioners to generate a surplus of £500 million to cover off the £500 million risk to its DEL from providers.

And I am guessing – because 2013-14 is the first year I can recall when a year-end surplus has been demanded of NHS commissioners at the beginning of the year  – that this is the origin of the ‘surplus policy’ tweeted by Crispin Dowler, the relevant section of which is copied below:

“Each commissioning organisation should plan to make a cumulative surplus at the end of 2013/14 of at least 1 per cent of revenue, including any historic surplus not drawn down.”

Back in the real world of hundreds of NHS organisations, applying this 1 per cent to the £80 billion-odd of revenue controlled by both the NHS Commissioning Board and the clinical commissioning groups will generate an underspend amongst commissioners of around £800 million. And although this is called a ‘surplus policy’, that is actually a red herring. It is actually an ‘underspend policy’ for commissioners, designed to offset the risk of unmanaged overspends amongst providers.

The Health Select Committee has now said that it is to look into this policy of ‘underspending’, but I would argue that doing so is another red herring. The NHS doesn’t want to have underspends, any more than the Department of Health does (or the Government, which has a commitment to increase NHS spending in real terms) – but the only reason why these underspends are happening is because the Department is covering off too much risk. It is, simply put, generating more underspends amongst commissioners than providers are generating overspends – and because the Department of Health looks at providers’ spending plans this must be because providers are not spending as much as they thought they would need to. And an investigation into why that is happening – in a world where providers are meant to be squeezed between an unprecedented £20 billion efficiency requirement and the tightest spending settlement the NHS has ever seen – would be very interesting indeed.

It would also help us understand the risk to the whole system of NHS finances, because it doesn’t take a genius to realise that it is unsustainable if providers continue to stack more surpluses on their historical surpluses; a few more years would put a small handful of providers – those with the largest surpluses – in the position of being able to blow the Department of Health’s DEL almost by accident. 

The ultimate solution, which the Health Select Committee should recommend, is for NHS trusts to be taken off the Government’s balance sheet. This would allow them to spend their surpluses without risking the Department of Health’s DEL, it would negate the need for a ‘surplus policy’ to be asked of commissioners – and it would also be the point when it becomes obvious that the NHS’s surpluses were never ‘handed back to the Treasury’ but always remained in NHS bank accounts and available to them to spend (hence the ultimate source of my difference of opinion with the HSJ). But such a recommendation would need to be made by the Health Select Committee to the Office for National Statistics, whose decision a balance sheet reclassification would be (rather than the Government’s).

And for those of you still reading, thank you for bearing with me. And if you found that as complicated as it is, then pity the man who needs to make it work. His evidence to the Health Select Committee says it all:


Q255 Chair: Are we not missing a trick? Is the health service not missing a trick here? If the Commissioning Board was constituted as a trust, then the cash would go into the Commissioning Board at the agreed rate and it would simply carry it forward as a reserve. The health service has 240 trusts, all of whom do that. Why can’t the Commissioning Board do it?

Richard Douglas: This is where, I’m afraid, the public spending framework does get complicated. It does still score against our spending. If you look at the underspend last year, part of that £800-odd million on the revenue was the sum of the foundation trust and the trust underspends in that year. I cannot access that cash. It is there, as you say, in their bank accounts, but it scores as an underspend against our figures.

Q256 Chair: You can’t but they can.

Richard Douglas: But the minute they do, it scores against my spending.

Dr Wollaston: It doesn’t make sense.

Q257 Chair: In other words, these are 240 independent organisations, all of whom have their cash-treasury function carried out by HM Treasury.

Richard Douglas: No; they don’t have it because they have the cash. They have the cash and they spend it, but when we look forward we need to predict at what point they are going to spend it.

Q258 Dr Wollaston: Why not just give them the money and let them manage it?

Richard Douglas: But they do manage it. That is exactly my point. We are not managing it. The trusts themselves are managing it. When you look at the underspend, part of it is the surplus that has been earned by the trusts. If you look at my £800-odd million underspend, roughly £500 million or £600 million of that would be the surpluses in foundation trusts. They have that cash. It is there in their bank accounts and they can spend it in future years.

Q259 Chair: Hang on a second. Can they or can’t they?

Richard Douglas: Yes, they can. I have no control over them spending their cash-absolutely.

Q260 Chair: That is what I thought the situation was, but you appeared to say a moment ago that if they did, that counted against your future year budget.

Richard Douglas: It does count against our spending. It counts against spending but it is under their control.

Andrew George: It is a Treasury rule.

Q261 Chair: It would be useful to have a clarification of exactly how these rules work.

Richard Douglas: I might say that too.