If there is one piece of information I wish more decision makers knew about, it is this one. It’s called Goodheart’s Law, and it states:
“The moment a measure becomes a target, it ceases to be a measure”
So simple, but yet so profound. I can think of many failed initiatives that could’ve been avoided if whoever decided on measurement criteria knew this.
I’ll give you a few examples: a South African television news program recently did an expose on the apparent misshandling of rape cases by the SA Police Service. One of the main complaints was that many rape cases that are reported to the police are never logged i.e. no case files are opened. During the interviews it was found that local police stations are measured on their ability to reduce the number of rapes in their area. This in turn was driven by Governments need to reflect a decrease in violent crime incidents when they report on the statistics.
We all know that lowering the incidence of rape is not an easy thing, but with their bonusses (or even their jobs) dependent on showing a smaller number of rape cases, police officers just decided not to log all the reported instances. Problem solved …
Today I was chatting to a doctor who needs to measure the performance of medical personnel below him. Upon receiving the performance review sheets, he was told by HR to remember that the results HAD to reflect a ‘bell curve’. They ignored his protest that in a hospital environment, poor performance (as indicated on the left of the bell curve) could not be tolerated, and all doctors had to perform above average, otherwise they’d be killing people. So he was left, walking around the hospital looking at his staff and wondering who the poor unlucky person would be that he would have to assign a 1 or a 2 performance rating to in order to satisfy the Bell curve requirement set by HR.
When will people stop incentivising people to be dishonest, stupid or both in order to satisfy inappropriate measures?