Random Promotions Beat the Peter Principle

The Peter Prin­ciple states that “in a hier­archy every employ­ee tends to rise to his level of incom­pet­ence” (dis­cussed pre­vi­ously). This prin­ciple is typ­ic­ally observed when pro­mo­tions are rewar­ded based on an employee’s abil­ity in their cur­rent pos­i­tion and provided there is suf­fi­cient dif­fer­ence between the two pos­i­tions.

In such cir­cum­stances, is there a simple way to ‘beat’ the Peter Prin­ciple? Accord­ing to the research that won the 2010 Ig Nobel Prize for Man­age­ment, yes: pro­mote at ran­dom to pre­vent the prin­ciple from com­ing true (pdf, also: arX­iv, doi).

We obtained the coun­ter­in­tu­it­ive res­ult that the best strategies for improv­ing, or at least for not diminishing,the effi­ciency of an organ­iz­a­tion […] are those of pro­mot­ing an agent at ran­dom or of ran­domly altern­at­ing the pro­mo­tion of the best and the worst mem­bers.

The authors of the study have cre­ated a sim­u­la­tion so that you can see the ran­dom pro­mo­tion strategy in action, and it’s worth remem­ber­ing that this coun­ter­in­tu­it­ive and (hopefully) tongue-in-cheek approach is just one of the pos­sible solu­tions to the prob­lem described by the Peter Prin­ciple.

Read­ing up on this, I also came across the rather eleg­ant Gen­er­al­ised Peter Prin­ciple, ori­gin­at­ing from obser­va­tions regard­ing hard­ware at nuc­le­ar power plants:

Any­thing that works will be used in pro­gress­ively more chal­len­ging applic­a­tions until it fails. […] There is much tempta­tion to use what has worked before, even when it may exceed its effect­ive scope.