HBS’s Working Knowledge recently published an article on using Forced Ranking in performance management. I’ll get into the conclusions of the article, but first – what is Forced Ranking? Well, Forced Ranking is a method by which managers evaluate employees relative to the performance of other employees. It’s a little like the curve I had in college.

With the dreaded curve, all test scores for a particular exam were plotted along a pre-determined curve of how many people should get A’s, B’s and so on and then adjusted to fit. This way, no more than say, 20% of people could get As and no more than 20% could get B’s and 20% were, in fact, forced to get C’s, D’s and F’s so as to fit the pre-determined curve (In all fairness to my alma mater, NYU, I’m not sure if this was exactly the curve they used, but as an example, it suffices).

The thinking goes that people will be motivated to work as hard as possible because small differences in scores can lead to large differences in grades. Likewise, in performance management, Forced Ranking ensures that managers will differentiate between talent and eliminates the possibility of review inflation becuase employees cannot all recieve excellent or even good reviews. If, for example, a company used a 1-5 rating scale and we followed the curve cited above, 20% of employees would get a 1, 20% would get a 2 and so on.

What follows is that now that the top 20% (relative to the rest of employees) and the bottom 20% (again, relatively) have been identified, they can be so dealt with. The top performing employees can be amply rewarded so as to make sure they stay with the company and are motivated to continue to perform. The worst performing employees can be… fired. This is commensurate with the whole idea of Forced Ranking. By removing the worst performing employees and replacing them with the best potential replacements from some applicant pool, the overall performance of the organization can be elevated.

The article goes on to cite a study that bears out this theory. Here’s a link to the study at Personnel Today. From the study via the article: “Results showed that a forced ranking system can improve workforce potential, in the sense that, on average, lower-potential workers can be identified and replaced by workers with higher potential.”

The critics of Forced Ranking say that its gains, if any, are cancelled out by the resulting low morale, blow to teamwork and poor perception of the company by potential employees and investors. But again, the study rebuffs them. In the words of the article’s author, ” They found that the potential problems were in every case balanced by equally compelling benefits.”

It is important to note that the study was not a real-life study. This was a model that took into account some assumptions about the nature of companies and workplaces that you may or may not agree with. I encourage you to read both the article and the study to determine your own feelings on the topic.

So what do you do then? Well, both the article’s author, Dick Grote, and the study he cites argue that the best approach is a temporary one:

Finally, for many years I have argued that for most companies, forced ranking systems should be used for only a few years and then, once the obvious and immediate benefits have been achieved, replaced with other talent management initiatives. While some companies have been successful in using their forced ranking system for decades, I find that most organizations are better served by implementing a forced ranking system as a short-term initiative. Scullen and his fellow researchers confirm that advice. Early in their article they lay out clearly the basic problem of using forced ranking on an ongoing basis: “Despite the allure of having a continually improving workforce, we argue that each time a company improves its workforce by replacing an employee with a new hire, it becomes more difficult to do so again. That is, the better the workforce is, the more difficult it must be to hire applicants who are superior to the current employees who would be fired.”10 Their mathematical simulation demonstrated that the greatest benefits came in the first 3.5 to 4.5 years after initiating a forced ranking system.

What are your thoughts?