Ben was CEO of Nene CCG for 8 yrs. He was a Director of an acute trust and has run national improvement programmes. He is a founding director and Principal Consultant of Ockham Healthcare, which he describes as ‘a platform for change’.

The Prisoner’s Dilemma, Game Theory and the NHS

The reader is probably familiar with the prisoner’s dilemma. It uses game theory to show why two rational individuals would choose not to cooperate, even if it appears in their best interest to do so. Here it is applied to whole system working in the NHS.

Two organisations within a healthcare system are facing financial difficulties. One is a CCG and one is an acute trust. Both are in serious trouble with their regulators because of this. When the financial allocation is announced for the year there is insufficient funds available for both the CCG and acute trust to achieve financial balance.

 The CCG and acute trust negotiators develop the chance of a bargain. Each organisation is given the chance either to: ‘betray’ the other by seeking maximum money from the other (for one maximising activity and re-basing MRET, and for the other stringent fines and penalties), or to cooperate and agree how the money in the system can be apportioned fairly according to what is available. The offer is:

  • If the CCG and acute trust both negotiate their own position only (‘betray each other’), both will suffer significant overspends of over £10M.
  • If the CCG achieves its contractual goals with (‘betrays’) the acute trust, the CCG will break even and the acute trust will overspend by £15M
  • If the acute trust achieves its contractual goals with (‘betrays’) the CCG, the acute trust will break even but the CCG will overspend by £15M
  • If the CCG and acute trust agree to share the pain by cooperating and working together they will both only overspend by £5M

The result of the prisoner’s dilemma is that because negotiating their own position offers a greater reward than cooperation, all purely rational organisations will betray the other, making this (1) the only possible outcome. If an NHS organisation has even the chance to meet its ‘statutory duty’ to break even, or alleviate the extent of its overspend, it will take it. The interesting part of this result is that pursuing individual reward logically leads both of the organisations to betray, when they would get a better reward if they both cooperated.

The impact of regulators is that the normal systematic bias towards cooperative behaviour that organisations may be expected to display (i.e. why the prisoner’s dilemma might not apply) is countermanded by the behind closed door pressure that organisations are put under by their regulators to meet their financial duties. This is exacerbated by the overall financial pressure the NHS is under.

When the prisoner’s dilemma is played repeatedly between the same prisoners, and consequently both prisoners continuously have an opportunity to penalize the other for previous decisions, then two classically rational players will betray each other repeatedly, for the same reasons as the single shot variant. As a result it is no surprise that the contract round between the CCG and the acute trust never really yields a different result!

So what does the prisoner’s dilemma really teach us? Primarily, that the default position for healthcare organisations is away from true whole system working, and that seemingly selfish organisational behaviour is purely rational. Whilst whole system working might seem straightforward, it actually requires irrational behaviour. Regulators that are responsible for the interests of individual organisations within a system only make this irrational behaviour even more unlikely. As a result, it will require more than well-meaning appeals to the greater good and platitudes within the five year forward view for whole system working to become a reality.

Together we've got the public sector covered