At the level of performance, however, both rankings and forced distributions have similar effects.
The most powerful mechanism for increasing performance is actually being clear about performance expectations, in terms of what is expected as well as the standard of performance that is required and rewarded.
If all performance is relative, then there is no absolute standard for performance expectation.
This can be a disincentive for performance, and it can result in overall levels that are poorer than if clear standards were used.
At a technical level, using a forced distribution may be even worse than ranking.
A manager ranking a small group whose performance he or she knows well can at least use some knowledge and expertise to identify top, middle and lower performers, even if they all meet the standard (although I still wouldn't want to use it to generate overall ratings or rewards).
But if managers are pushed into using a bell curve or normal distribution, then they would have to satisfy the requirements of that type of curve.
A normal distribution, or bell-shaped curve, is built on a sample of a large size - well over 100 - which would only apply to the largest 100 or so New Zealand organisations.
More importantly, the causes of the distribution of performance scores would have to be random.
It's a bit of an oversimplification, but if you had a large group of staff and you hired them randomly and managed them randomly, then you would expect a bell-shaped curve of performance scores.
I would suggest that if the conditions for the curve were truly met, the organisation would already be in trouble for low levels of performance and that you should get rid of all those random managers as they aren't adding any value.
* * *
Email a question for Dr Marie Wilson
