Forced Rankings: Truly Meritocratic or Just a Lottery?

Companies often use forced ranking systems: in a team, the bottom 15 % are fired and the top 15 % promoted, regardless of actual performance. The idea sounds simple—create a "competition" to push everyone to excel. But what really happens? A recent study, using a simulation with almost 1 000 engineers split into 142 teams of seven, showed that even with random team assignments, 32 % of termination and promotion decisions are wrong. In other words, a third of employees are judged solely by team composition, not by their own abilities. When real differences in manager skill are added, the error rate climbs to 53 %, with more false positives (unearned promotions) than true recognitions. Some suggest cross‑team calibration, but this turns evaluation into an influence contest: the most persuasive manager gets the promotion, even if undeserving. Over time, employees notice these random outcomes, become risk‑averse, and top talent may leave. The most efficient solution would be to let accountable managers decide, but legal constraints and the demand for a "transparent" process keep forced rankings alive—not because they truly motivate, but because they appear systematic. The result? A system that looks fair but actually distributes rewards and penalties almost at random.

The simulated model uses an agent‑based approach where each engineer is represented by an agent with a competence score drawn from a normal distribution. Teams are formed randomly, mirroring project assignment practices in many tech firms. The forced ranking mechanism imposes a fixed percentage threshold for promotions and terminations, regardless of the overall score distribution. Classification error is calculated by comparing an agent's true position in the global distribution with the ranking decision. Introducing variability in managerial decision‑making ability (modeled as a bias term drawn from a normal distribution) adds a "evaluation bias" that amplifies errors. Sensitivity analysis shows that increasing team size slightly reduces error, but does not eliminate the randomness introduced by the fixed‑percentage constraint. From an economic‑theoretic perspective, forced ranking violates incentive compatibility: employees can no longer distinguish effort from outcome, leading to "gaming" behavior and reduced productive efficiency. Implementing evaluation mechanisms based on individual objectives, continuous feedback, and managerial accountability proves more robust for optimal human‑resource allocation, but requires a more flexible regulatory context than that imposed by forced ranking policies.