Last week, while reading a letter to the Independent newspaper, I came across Goodhart’s Law for the first time. (To read this letter, click here and search for ‘Goodhart’.)
Expressed in simple terms, Goodhart’s law states that:
So, for example, if you set schools a target for the proportion of pupils gaining A* to C passes at GCSE, this will cease to be a good measure of how well the school is functioning because it changes the way the school behaves: as we all know, it puts pressure on teachers to concentrate their efforts on improving the performance of borderline pupils (who might just get a grade C with additional teaching) at the expense both of brighter pupils and those who stand no chance of getting a grade C no matter how much extra tuition they are given.
The logic of Goodhart’s law is that centrally set targets cannot work. If you have a small number of crude targets (like the one I have just mentioned) this will distort the way teachers and organisations work, yet if you try to overcome this by making the targets more complex and more numerous then you cripple organisations with an impossible bureaucratic burden. And of course, all targets (whether crude or complex) take the attention of organisations and teachers away from what should be their main focus – doing their best to promote real learning for each and every pupil/student/learner.
Below are links to a couple of good articles about the negative effects of Goodhart’s Law on learning.
Peter Mortimore writing in the Guardian in May 2008
An article by Michael E McIntyre, Emeritus Professor of Applied Mathematics and Theoretical Physics at the University of Cambridge
Let me finish this post with a quotation from the McIntyre article. I feel sure this is an idea I will return to again and again in this blog. If we want to promote real learning…
Comments