During the 1990’s and early 2000’s the UK government relied heavily on target setting as a way to manage the public sector. We had:
No doubt we had targets for the target setters as well. UK PLC was run by setting targets and expecting the managers to hit them.
Did it work?
Here is a very interesting paper that answers that very question. Unfortunately it is the sort of paper that takes 30 minutes, your full concentration and a strong cup of coffee to read. If you don’t have those let me give you a taste…
There were 3 clear effects from all this target setting:
1. Ratchet effects
The ratchet effect happens as target setters progressively make the target harder and harder to hit, gradually ratcheting up performance year on year.
The problem with ratcheting up the target is that if you make 101% of the target one year you will be asked for 105% the next, so nobody in his right mind would knock the target over and hit 150%, after all what chance would you stand the following year?
You might think this is cynical, so let me ask you… Did you make sure your spent all of your budget last December just so it wasn’t taken away from you this year?
Yes? I did, and that is what causes the ratchet effect.
2. Threshold effects
The threshold effect happens when the target creates a step in performance.
Instead of a spread of results performance clusters around the target. Those who are below the target strive to hit it (by about 101%) where as those who are performing far above the target take their foot off the gas and coast down to it. Why invest your resources in something you won’t get thanked for?
Once agin you can challenge me with cynicism so let me ask another question… If you were a teacher with a target to get your children through an exam where would you focus your efforts?
- The children who could easily pass the test
- The children who will never pass the test
- The children who might just pass the test
Which did you chose? Targets create thresholds in performance.
3. Output distortion
The last effect of target setting is to distort the output, a politically correct way of saying to cheat — to make the numbers by fair means or foul.
Of course you or I would never cheat, we are fine upstanding members of society.
Let me give you a couple of examples of how fine upstanding members of the medical profession behaved when faced with targets:
- Doctors were told that patients shouldn’t wait more than 48 hours to see a General Practitioner. This was an easy target to hit. They simply stopped taking appointments to see anybody more than 2 days in advance — goodbye waiting list.
- Hospital managers were given a target that emergency admissions should be given a bed within 12 hours. Once again this was an easy target to hit. They took the wheels off the gurneys the patients were lying on, converting them into “beds”.
So targets have effects, just not the ones you would expect
Still not convinced? Then let me leave you with a final thought…
The more taxing government targets were nicknamed P45 targets. (P45 is the reference code for the UK tax form entitled “Details of employee leaving work”.)
How would you behave if your children’s welfare depended on hitting a target?
Let me ask again, did the targets work?
Don’t ask me, I am horribly biased, you will have to read the report and draw your own conclusions.
If you enjoyed this post click here for updates delivered straight to your inbox
Read another opinion
Image by zomie84
Hi James
I agree with you 100%. Setting targets often has unintended consequences.
I remember when one particular automobile importer’s dealerships in Germany were set targets for Post Purchase Customer Satisfaction. The satisfaction score was measured through a survey to be completed by the customer within a short-period of buying a new vehicle. Missing the target had financial consequences for a dealership. So what did the dealerships do? That’s right, they offered customers a free post-delivery inspection, a voucher for Euro100 towards their next inspection, or some other other incentive if the dealership could fill in the survey itself on behalf of the customer. Most customers were only too happy to let the dealership fill in the survey whilst they pocketed the incentive. Dealerships consistently scored high on satisfaction. And their margins were not affected.
The curious thing about this story is that the vast majority of dealerships consistently scored high on satisfaction anyway. Introducing the target created over-compensating behaviours that were, for most dealerships, simply unnecessary.
Graham Hill
@grahamhill
James – I enjoyed reading your blog, and felt a close connection to the reward matter you discussed. When I’m working with companies on business development strategy, I find myself often saying (or at least, thinking), “be careful what you reward, because you might get it.”
Your discussion reminds me of the Cobra Effect, in which the British government in India put a bounty on cobras in order to check the growing population of the snakes. Initially it helped, but the effect was to have an epidemic population of the snake. The government’s bounty encouraged people to breed cobras to get the bounty that the government awarded.
The backfire didn’t end there. When the government caught on to the practice, they immediately rescinded the bounty award, and the outcome was – you guessed it – breeders released their now-worthless snakes.