The reason relative ranking beats absolute targets

A team paid for hitting a fixed number hits it and stops. A team paid on relative position keeps working, because there is no point at which position is secured. The second scheme produces more output and a different set of problems, several of which do not appear in the output figures.

What everyone sees

Ranking is read as healthy competition and targets as clear expectations. Managers usually justify the choice by fairness or clarity rather than by ceiling effects, and the ceiling is the mechanically important difference: a target defines a point at which further effort earns nothing.

What is actually happening

Lazear and Rosen’s tournament model shows that rank-order compensation can elicit effort efficiently when individual output is noisy, because relative position filters out shocks common to everyone. It also removes the ceiling. The cost is documented in the same literature: tournaments discourage cooperation, invite sabotage where it is possible, and become demotivating when ability differences are large and known, since the outcome stops being contestable.

Why it stays hidden

The failure modes hide because they are not in the same units as the output. Sabotage, hoarding of information and the withdrawal of the visibly outmatched all reduce total value while leaving individual scores intact or improved. The scheme reports its own measure faithfully, and nothing in that measure records what was done to competitors rather than to the work.