The Curve Debate – A Matter of Context and Application
The practice of ranking or evaluating employees on a bell curve has been hotly debated in the past two years when some organisations started to make a switch towards a new way of rewarding performance.
The recent article “Forced Ranking and the Bell Curve: How Outdated HR Practices Undermine Employee Performance” neatly summarises some of the common gripes – how the performance in an organisation is not normal, limits on the number of high performers in the organisation, artificial demotion of a top performer after several iterations of ranking and yanking, undermining of employees’ actual contributions, failure of the bell curve in driving performance or engagement, and the list goes on.
The proclaimed solution to this problem was the use of the power law distribution. Advocates claim such a distribution was a better reflection of performance and accounts for the high variability of performance.
However, I feel that the solution is academic and hardly solves the issues associated with the bell curve. For instance, if an employee was demoralised because he was rated 'average' on a bell curve, he is not going to feel any better if he was rated 'average' on the power law distribution. Likewise, the power law distribution does not escape from limiting the number of high performers; in a world with finite resources, management will still need to determine how best to distribute rewards. That will still necessitate some form of ranking.
The choice of what to use should therefore be guided by the business and operating context and how the curve is applied in performance evaluation.
The business health of the organisation and the type of work would be factors to consider in the choice of the curve to apply.
Business Health: An organisation may need to turn around its business and has identified weeding out its poor performers as one of the initiatives. In this case, the rank and yank approach might be appropriate.
Type of Work: For the power law to work, there must be a way to accurately identify what is high performance. This might work for a sports team, which often would have one of two elite players who make a difference. However, apply it to a low tech manufacturing environment, where workers are typically similar, and output is measured by number of units completed, and the power law distribution may not work so well. In such an instance, there may not be a significant differential to identify a group of employees that is truly top performing .
Regardless of the choice of the curve, how an organisation actually uses it matters. For instance, rank and yank is not limited only to the bell curve. I could similarly pick out the worst 10% employees on the power law distribution and fire those employees. Similarly, if the purpose of identifying poor performers was to provide them with extra resources so that they can become more effective, then it does not matter if these employees were identified with a bell curve or a power law distribution.
Instead of a debating which curve to apply, energy should be focused on working closely with management to design a system that works best in the organisation's operating context. Being open and transparent in its communication with employees will also go a long way towards building trust in the system. This is how I think HR can add better value to an organisation.