Hitting the target but missing the point
During the pandemic the UK government tied itself up in knots over metrics. The Health Secretary originally promised 10,000 Covid-19 tests a day by the end of March 2020, a target not reached, and then replaced this with a more ambitious target of 100,000 tests a day by the end of April. Under pressure, perhaps embarrassed by the low rate of testing in comparison with many of our European neighbours, particularly Germany, the government had to be seen to be acting, and to be doing so seriously. Every subsequent press conference saw government spokespeople claiming that they were ‘straining every sinew, working night and day’ to make sure that staff in the NHS had everything they needed. As UK citizens we were offered quality, “we’re working very hard”, and quantity, “we’ll deliver you 100k tests”
At no point did we enter a discussion about why we might not have met the first target before reaching for another abstract ideal, a new target, or whether the new target has any relation to what we actually needed.
When questioned about the complaint that there is not enough Personal Protective Equipment (PPE) for frontline health staff, government spokespeople came up with more numbers. There are X million pieces of equipment in the system; they have spent Y million pounds procuring it. We’ve spent money on the health service like we have never done before.
It provokes me think about the role of metrics in general, and targets in particular, as hedges against the anxiety provoked by uncertainty. In a situation where there is a high degree of anxiety about the future and when people look to authority figures for reassurance, there might be a tendency to pick a number, any number, in order to contain and soothe and to pretend to be in control. We are asked to put our trust in numbers and thus the authority figures offering them. Numbers are grasped at and repeated like a religious incantation. Perhaps the assumption by those repeating the numbers is that there is proportionality between the size of the number and the reassurance experienced. If the government is spending more money than ever before, then this must be good, right? Millions and millions of pieces of PPE requisitioned has got to be helpful for frontline workers, hasn’t it?
At the risk of comparing the banal with the extreme seriousness of what the world faced in terms of the lethal pandemic, in the same week that the UK Health Secretary announced a new set of targets for testing, my own institution announced new metrics for researchers throughout the university in terms of research income, publications and PhD completions. These are universal, irrespective of the intellectual discipline of particular school they are imposed on, and their record to date of achieving anything similar. One can only suppose that the idea of imposing the same targets across all schools in the university has some attachment to the notion of ‘fairness’, which is a cult value for all bureaucracies. The directionality is different though: this is not the management of the university under scrutiny telling us researchers what they will do, but setting a set of metrics for what they expect of us.
I am guessing that they won’t be impressed with a qualitative argument that we are straining every sinew, working night and day, to achieve them. That’s becasue they are used a means of disciplining and punishing.
The literature on metrics is now wide and deep. We know, for example that targets make a difference in organisations not because they work but because they are made to work.[1] Making them work might mean staff massaging the numbers, gaming the system, or doing work-arounds, particularly if there are sanctions attached to not meeting them. There are now a number of egregious examples of how public institutions have met their metrics and have been deemed good organisations by regimes of inspection, only for this judgement to have covered over poor practice. There are many similar examples of institutions hitting their targets but missing the point. We know that because targets are informed by cybernetic systems thinking they oblige workers to pay attention to aspects of service which politicians or authority figures deem important rather than what their local situations demand.[2] Instead of asking how the work is working and dealing with the situation in hand, we are obliged to work to what someone who usually sits remotely considers to be important. We know that experts in their field can experience them as tyranny that undermines their professional judgment and can cause the opposite of what they claim to prevent[3]. We know that despite the deluge of metrics, there is no greater trust in public government as a consequence.[4] And we know that metrics don’t just measure reality, they shape the reality to be measured and what gets noticed.[5] [6] We can no longer imagine a world where there are no metrics for organisational life, and our numbers and audit regimes are methods of trying to organise and systematise our uncertainties.[7]
None of is to denigrate those areas of public life where numbers are important for modelling, for telling us about health and social outcomes, and for reassuring us that a particular medicine is effective and safe to use. Instead I am dealing with organisational performance targets, often built on proxies and sometimes arbitrarily arrived at to convey certitude, to exculpate, to discipline and control.
So one of the central functions of metrics is to reassure, to claim authority and to assert control, either for authority figures to claim they are in charge, or to establish control over others. They serve as a defence against anxiety and unpredictability and are a claim on authority.
One interesting argument from the economist John Kay’s book Obliquity[8] is that goals are best achieved indirectly. The happiest people are not necessarily the people who strive for happiness, and the most successful companies in Kay’s experience are the ones who care about what they are doing, rather than the ones who focus on targets.
So if my university followed Kay’s advice we would spend more time thinking about research, discussing our research, and uncovering what it is that gets in the way of our doing more and better research. We would generate a lively discussion about what we are doing and why we care about it. The numbers would then follow. Equally, the answer to the question, ‘how many tests should we carry out in the UK’ is as many as we need depending on the circumstances in which we find ourselves. This might involve paying attention to our current situation and extrapolating from that rather coming up with an arbitrary figure. This would demand humility and honesty to talk about how the work is working, what we are able to do and what we are not, and why.
[1] Townley, B (2008) Reason’s Neglect, Oxford: Oxford University Press.
[2] Seddon, J. (2008) Systems Thinking in the Public Sector, London: Triarchy Press.
[3] Muller, J. (2018) The Tyranny of Metrics, Princeton: Princeton University Press.
[4] O’Neil, O. (2002) A Question of Trust, The Reith Lectures, BBC. http://www.bbc.co.uk/radio4/reith2002/
[5] Scott, J.C (1998) Seeing Like a State, Yale: Yale University Press.
[6] Beer, D. (2016). Metric Power. London: Palgrave Macmillan.
[7] Power, M. (2007). Organized Uncertainty: Designing a World of Risk Management, Oxford: Oxford University Press.
[8] Kay, J. (2011) Obliquity: why our goals are best achieved indirectly, London: Profile Books.
