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.