Regular readers of this blog will know the importance of numbers in frameworks. Numbers provide finite anchors against which to determine change – has something improved, has it got worse, by how much? As with most rules and frameworks there is a cautionary note to be made about setting and using targets. Setting numerical targets can be dangerous because they can have unintended consequences. We have seen this in the UK government’s strategy on Covid-19.
From the beginning the government in the UK was panicked about the ability of the National Health Service to cope with the coronavirus pandemic. Various models were run and some showed that the NHS could be overwhelmed by the disease. This led to a strategy, a framework if you like, which was very clear – “stay at home, protect the NHS, save lives”.
At first sight you would say that there is nothing wrong with this. None of us wanted to see the NHS collapse. Whatever was necessary to keep it going would be provided – at any cost. However, looking back over the strategy it has been a disaster. The slogan and the numerical targets for making hospital beds available, instead of saving lives, have cost them.
While the government focused on hospitals, care homes were disregarded. Indeed, up until the 12th March the government said the care homes sector was very unlikely to face an outbreak of the disease. Care homes had to scratch around to obtain supplies of personal protection equipment. The NHS sent elderly patients to care homes without testing them for the virus in order to free up precious beds in the hospitals.
It is estimated that between the middle of March and the end of April, 20,000 people died in care homes in England and Wales as a result of the virus. In that same period, bolstered by billions of pounds and additional resources, around 30,000 people died in the NHS in England and Wales. The setting of targets and the obsession with the NHS somehow made the government blind to the dangers lurking in care homes.
We’ve seen a similar thing happened with targets for testing. The UK government was behind in its testing strategy. In other countries there were early strategies for “test, track and trace”. If you know where the virus is, you can hunt it down, surround it, and sort it out. Possibly because it was on the back foot, the UK government set an ambitious target of performing 100,000 coronavirus tests a day by the end of April. The number was “sort of” achieved. It wasn’t really. Almost 40,000 of the claimed tests were simply mailed out and were counted as tests carried out. No one could say how many were still hanging around on shelves in the bathroom waiting to be used. No one could say how many had been sent back and tested. The obsession with a target number led to poor decision-making.
In 1975, Charles Goodhart, an economist made an important observation. He said "any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes". Think about that for a minute. Goodhart is telling us that if we set a target in order to achieve something, the measurement of the target will start to go wrong. This principle has since become known as Goodhart's Law.
It's not that numbers and causal relationships are wrong, it is that making a measure a target makes it cease to be a good measure. We see this all the time in business. A company makes good profits and as a result sets itself a target of maintaining or improving those profits over time. Though it may do so, it might find that it is achieved at the expense of worse productivity or creative accounting or high staff turnover.
The UK government's ambition to "protect the NHS" was done with the best intentions. However, if the government had taken note of Goodhart it would not have made "protect the NHS" the mantra. "Stay at home, save lives" would have been more appropriate and may indeed have saved many more lives.
When a public enquiry finally takes place into the UK’s response to coronavirus, we will see that “protect the NHS” and the arbitrary setting of targets will show that the government didn’t really understand frameworks and the danger of pressure placed upon targets for control purposes. I wonder if anyone will mention Charles Goodhart?