Forced Ranking - Goodhart's Law redux?

In my last post, I talked about Goodhart's Law (and a related concept, known as Campbell's Law).  In essence, both Goodhart's Law and Campbell's Law say the same thing (Goodhart's Law was first described in economics, specifically in monetary policy, while Campbell's Law comes from the social sciences) - once an organization starts to use a quality measure as a target, the measure ceases to be useful as an indicator of quality.  As I stated in reply to a comment to my original post, I think part of the explanation for Goodhart's Law is that when we tie individual metrics to some kind of incentive, whether it is monetary (e.g., pay-for-performance) or something else (e.g., college ranking in the annual U.S. News and World Report), we have essentially substituted extrinsic motivation for intrinsic motivation.  Extrinsic motivation tends to "crowd out" intrinsic motivation - in other words, by linking achievement of a certain level on a quality measure with some kind of reward, such as a pay bonus or recognition by an external agency, we have discouraged employees from doing something because it is the right thing to do (doing good for its own sake). 

Several years ago, the legendary management guru, Jack Welch started to use a technique known as forced ranking at General Electric, Co.  Forced ranking is a management technique in which managers rank individual employees every year based on their performance.  Fair enough so far, correct?  The top 20% of employees are the "A group", the middle 70% of employees are the "B group", and the bottom 10% of employees are the "C group."  Importantly, the C group employees may actually be good employees who are achieving performance targets every quarter; however, they still rank at the bottom of the organization in terms of their performance.  Jack Welch was a strong advocate for providing financial rewards, praise, and recognition of the "A group" employees, while "B group" employees were provided with coaching and mentorship.  Welch often said of the "C group" employees, "They have to go."  By ranking employees, Welch and others believed that employees would compete with each other and the organization as a whole would benefit. 

Even if forced ranking is not followed exactly as Jack Welch originally described, we see certain elements of it in almost every performance management system used today.  For example, many organizations use the "Fails to meet expectations," "Meets expectations," and "Exceeds expectations" in their performance management system, which is essentially ranking employees by whether or not they achieve certain performance metrics.  Moreover, in some cases, forcing managers to rank employees is a way to address "grade inflation" with performance ratings.  For example, when I was in the Navy, a simple five-point scale was used to rate every sailor in approximately five to six categories.  Even though a "3" was equal to "average" or "meets expectations", getting a "3" was essentially a career killer.  Hence, most commanding officers gave a lot of 4's and 5's, and if a sailor was up for promotion, a "5" was absolutely essential (and usually provided, as long as the individual was a hard worker and motivated to excel). 

Forced ranking certainly has its supporters - there are definite advantages, but in many cases, these are outweighed by the significant disadvantages!  Think of how difficult it would be every year to have to terminate your bottom performers!  What forced ranking is saying is that the "C group" employees aren't even worth the effort to try to rehabilitate or "coach up" to the "B group."  While some level of competition is okay, the kind of cut-throat competition that many companies that used forced ranking was counterproductive in the end.  Think of how competitive pre-med students are and you will have a good idea of what I am talking about.  Teamwork, at least the kind of teamwork that is necessary in today's environment, just doesn't happen when employees are competing this much with each other.  Finally, by linking promotion (and termination) with performance, forced ranking has essentially created the conditions described by both Goodhart's Law and Campbell's Law.  By making a performance evaluation a target to achieve for each individual employee, the performance evaluation has ceased to be useful as a measure of the quality of an individual employee's work production. 

It is no wonder that a number of organizations (including, by the way, General Electric, Co.) that previously used forced ranking have stopped using this management technique.  For all intents and purposes, forced ranking, at least in the form that Welch described, has fallen out of favor.  However, as I mentioned above, we still have some vestiges of forced ranking with most of our current performance management systems, especially when they use classifications such as "Meets expectations" or "Exceeds expectations."  Goodhart's Law would suggest that even this simple classification should not be used.  Maybe Goodhart was right...