The consequences of measurement

Goodhart a British economist is widely held to have formulated the equivalent of Heisenberg’s Uncertainty Principal for economics. It states as follows

Any statistical relationship will break down when used for policy purposes

A simpler formulation from a US academic translates this as:

If a measure becomes a target, then it ceases to be a measure

Now if I had my way I would have this engraved on tablets of stone and placed in the entrance hall of every Government department in the world. If I just look at the press this week we have had (i) stories about the Ambulance service in Britain misusing the classification system to achieve targets and worst still (ii) a UK Government Think Tank saying that although all key targets for the Health Service have been met, the overall service has deteriorated. The problem is compounded by a view of measurement that is stuck in the hypotheses-empirical tradition of evidence. An approach that has value in some cases, but little credibility in conditions of uncertainty and complexity. My next Frontiers article will address this, looking at ways to achieve un-gameable measurement systems that represent complex environments without assuming linear causality.

In a very real sense you get what you measure and you live to regret it. In the case of government we are the ones to suffer. Of course it could get worse. When Six-Stigma starts to loose its shine then Governments will adopt it as industrial best practice and after that the bottom 10% of all patients will be culled for the sake of the health system as a whole.