I just read a short review by David Lorimer of a book The Tyranny of Metrics, author Jerry Z. Muller. This problem of over-reliance on metrics was apparent during my career in industry, and is increasingly apparent in the world today, particularly in management and government.
Metrics are defined and targeted, and actions are put in place to achieve the targets. All very focusing, you might think, but what is lost is attention to the real world and its appropriate management – all the other factors not involved in the metric. We begin to manage by the numbers, and not by the need. For example, the quarterly results, the bottom line, become all-important, never mind we have to ‘lose’ hundreds of workers in transferring jobs to places where labour is cheaper and standards are less, destroying social cohesion, never mind we have to make the social care system unnecessarily harsh to achieve savings, and on and on.
Unfortunately, this obsession with metrics is undermining professionals across many fields. As the review points out, “judgement based on experience is the hallmark of an accomplished professional in every field”. The over-emphasis on metrics is slowly undermining the accumulated wisdom of centuries of professionalism. It also overrides any consideration of human values.
Worse, computers and so-called artificial intelligence are increasingly being used in situations that can supposedly be managed by algorithms. Indeed, this can be very helpful to the professional – enabling better-informed decisions. The problem, however, comes when the algorithm is given control, and the professional is cut out of the process – what a great cost saving that would be – but at what cost?
It’s all about what sort of world we want to live in. Metrics and algorithms are good servants, but incompetent masters. The world is not algorithmic, and we cannot allow our societies to be ruled by rigid targets and faceless algorithms. There lies an ultimate tyranny.
The driverless car presents an interesting challenge to my rant. Can we ever trust its algorithms and let them loose willy nilly on the roads? They may actually prove safer in the long run than the average ‘human’ driver. So why not? At the end of the day, people must remain in charge at some level (maybe not in the car) and take the key decisions.