The case against policy targets
At one level, the UK Secretary of State for Health should be congratulated. He set his department a target of 100,000 viral tests by the end of April, and they achieved the target. At another, he has provided us with a case study in public choice incentives and dodgy targets.
As a baseline it is very hard to know whether this target has made any difference at all. There is no counterfactual to test what would have happened without the target. Although claimed, there is little evidence to suggest this target drove delivery faster. The UK’s testing capacity would have increased regardless. There is a global demand for tests, from multiple bodies, and rising supply as a result. The tests are not made by the state and do not require the state to ensure their efficient deployment.
At most, we can say that the state’s involvement has prioritised their use for frontline services during a time of extreme scarcity. Not a situation that will persist. There is also evidence of statistical chicanery in meeting it, with distributed tests being counted, not just those completed. This is similar to claiming invitations to job interviews should count as employment. A useful reminder just how subjective and political seemingly factual national statistics can be.
The target has distracted from what is it was about the target that actually mattered. Although we do not know how COVID-19 will be beaten, we have a fair idea that the countries best prepared, with the lowest death rates executed an early test, trace, and contain policy. South Korea and Taiwan are the most notable examples. They had the capability and capacity to identify new cases, who were then isolated, without a generalised lock-down. Both learned lessons from the first SARS outbreak in 2002-03 and were ready when this crisis hit. Why, then, was the target not simply to build such a system, ready for when the lockdown could be lifted?
Most of the rest of the world, to varying degrees, were not ready, including the United Kingdom. There has been a lively media debate about the politics of small differences between them, but they are all examples of a generalised policy failure amplified in no small part by insufficient urgency from the World Health Organization and insufficient honesty from Chinese authorities. The UK then has been playing catch-up ever since the crisis started. However, it is not a game that can be won. The lack of preparation conditioned what is happening now and cannot be redressed retroactively, only mitigated. The UK death toll will be higher and our lockdown likely longer as the result of past decisions that cannot be changed now.
The testing target then is both simultaneously a measure of something and nothing. It is something given the UK needs to increase testing capacity. It is nothing given that the number itself is meaningless. The target is not the target.
Targets have always been a feature of policy, but became more popular in the UK in the 1990s, championed by a growing management consultancy industry, offering tools for professional delivery by bringing corporate management techniques to the coasting gentleman amateur culture of the British civil service. McKinsey coined the slogan ‘if it can be measured it can be managed’, and by implication that unmeasured activity was either suspect or misdirected. New Labour had consultants train their future Ministers and put five targets into a pledge card for the 1997 election. This, before driving an explosion of policy targets when in Government. Housing targets, climate targets, child poverty targets, NHS waiting time targets, diversity targets – the list is long, and, met or not, many have survived and multiplied since.
The point of doing this was to drive action and accountability. If there is a clear target, and everyone understands how they can help deliver it, activity is focused, enabling people to deliver the possible to what can seem like impossible deadlines. That is the claim for UK testing growth in April. By stating these targets in public there is accountability. Only, if the target is political, as many public policy targets must be, then the action being driven may be useless or mendacious, and the accountability for it a distraction. It has never been clear, for example who goes to jail if low carbon targets are missed. Good measurement of bad targets is not management, just wasteful data collection, a job creation scheme for consultants.
The Minister’s testing target, set at the beginning of April, relates to a problem that is real: the UK lacks the tools to test resistance. But instead of a national media debate about why the UK lacks that capacity and still does, we got lazy ‘gotcha journalism’. The question of who was responsible matters to the media, as scalping a politician or senior official is news. It matters not so much to the question of how you solve the problem. The Minister then picked a target that addressed a communication concern, the daily media bashing on a problem he cannot instantly fix. But, it has not addressed the actual concern, how you get capacity for quality tests that enable containment.
The South Korean system works because after a positive test, all possible contacts are contacted, contained, and tested until there are no further positive tests. Delivery of controls is decentralised, and resources follow outbreaks. Test errors means it is not perfect, but it is good enough to get the P-factor of transmission well below 1. The UK cannot do this. The suspect quality of available tests aside, the systems, experience and training are not in place. In that regard a capacity for 100,000 a day is more useful than 50,000 or 25,000 tests, but it not a measure of success. A better target would have been an open-ended commitment to get to the systems needed, not the PR target.
In taking that approach the UK has unintentionally borrowed from the playbook of the Soviet Union. Specifically, their habit of setting five-year plans, including detailed output target-setting for every product, most famously tractors. The targets were meaningless. Tractors are tools, not outcomes, and not ones that can be used very effectively without the skills, parts and infrastructure of modern farming. Demonstrating you hit the target meant glory (or rather, in Stalin’s days, avoiding the Gulag) for those involved. It did not necessarily translate into better farming. The central committee also had to perfectly second guess what parts would be needed for the supply chain, repairs, training, and impacts on inter-related sectors like pesticides and perishable warehousing. They didn’t; no one can, central planning doesn’t work. The target was not the target.
In the West conversely the decision to buy tractors was mostly left with individual farmers. Decentralised, subject to market mechanisms, it proceeded at a pace consummate with improving farming and market demand. Supply chains are self-organising and adapt in real time to issues of supply and shortages. There was no target.
To be clear then, duff central targets for a political purpose are a bad habit, best avoided. They convey an entirely false prospectus that a Minister is in charge and can control events. They distort priorities, people seek to meet the target rather than solve the underlying problem. Any political benefits further are fleeting, containing the occasional hostage to fortune, such as if the tests don’t work.
Herd immunity was a sensible target, common to all pandemic survival plans. It can be met either by containment or by deployment of a working vaccine. But in the UK it was made to sound like a callous plan for culling the elderly, due to the failure to lock down quickly, which could only mean achieving it by letting the infection run its course.
One policy and communication error then appears to have led to another, substituting a sensible mission, with a gimmick. The challenge still remains that mission, but the media will now be expecting a new gimmick with which to play Gotcha. It would be better not to play.