Peter Principle

What the Peter Principle is

The Peter Principle is a claim about management hierarchies: an organisation that promotes its people on the strength of how well they do their current jobs pushes each of them upward until they reach the level at which they can no longer cope, where their rise stops and they stay. It is named after Laurence Peter, who published it in a book in 1969.

Its famous wording is satirical in origin, but beneath it lies a serious management observation: the criterion for promotion and the demands of the level promoted to are two different things, and an organisation that uses the first as evidence of the second is using a measure for something it was not designed to measure.

Two assumptions, and it falls if either one does

The principle is not a law that applies to every hierarchy. It is a result that follows from two conditions when they occur together. Reading it as a result without its conditions attributes to the structure of organisations what is really the effect of a choice made inside them:

• The promotion criterion is performance in the current level. If a person is promoted on evidence specific to the target level, the cause of the error is cut off at the root.
• The abilities the two levels require differ materially. If the higher level is an extension of the same work on a larger scale, current performance is reasonable evidence for it, and nothing follows.

The second condition is where the real risk lies, and its clearest case is the move from technical work to managing the people who do it. An excellent specialist is promoted for mastering a tool and solving problems, and the new level asks them to allocate work, appraise other people’s performance and absorb a disagreement between two colleagues. Nothing in their first mastery says anything about the second, either for or against.

The ladder calculation, and exactly what it shows

Take a ladder of five levels and an intake of 1,000 employees entering at the first. Assume the probability that a promoted person succeeds at the next level is 70%. The figure is assumed to show the structure, not an established value, and in the sources we reviewed we found none that gives it a general value.

• First to second: 700 move up and 300 stay where they are.
• Second to third: 490 move up and 210 stay.
• Third to fourth: 343 move up and 147 stay.
• Fourth to fifth: about 240 move up and about 103 stay.

So about 240 of the thousand reach the top, or 24%, which is 0.70 multiplied by itself four times. The remaining 760 or so occupy levels at which their rise came to a halt.

The calculation shows something the verbal version hides: the share whose rise stops grows with the length of the ladder, not with any weakness in the people. A ladder of three levels with the same probability leaves 49% at the top; one of seven levels leaves about 12%. Shortening the ladder changes the result on its own, and changes nothing about anyone’s competence.

A third factor changes the result and is absent from the wording: the organisation’s growth. Seats at the upper levels multiply as the volume of work grows, opening up in front of people whose rise had stopped for lack of a seat rather than lack of ability. The reverse happens in an organisation of fixed size: a higher seat opens only when its holder leaves, so advancement depends on other people’s movement. Reading that difference spares the organisation an explanation that blames individuals for what is really an effect of its size and how fast it is growing.

The confusion inside the famous wording

This is the correction that usually gets missed. The popular version says an employee settles at their level of incompetence, and in doing so it treats two different descriptions as one:

• Not being a candidate for the next level, which is a relative description, comparing the person with others competing for a single seat.
• Not being competent at the current level, which is an absolute description, measured against the demands of the role alone.

The first does not imply the second in any way. In the ladder above, seats at a higher level are fewer than the people at the level below by the very shape of the structure: the wider the span of control, the fewer the layers, and the fewer the seats with them. Most of the people whose rise stopped did so for lack of seats, and are fully competent in their roles.

That is why using the principle to describe a particular employee is a misuse, and it happens often as a verdict on someone who has stayed a long time at one level. That is a separate topic, the career plateau, where a halt in movement may be structural to the hierarchy rather than a description of the jobholder’s ability.

Why the error is not corrected once it has happened

What makes this error heavier than others is that it works like a trap: easy to fall into and hard to climb out of. Three things combine to cause that:

• Moving down reads as a penalty. Returning an employee to their previous level is understood inside and outside the organisation as failure, however it is described, so both sides prefer staying put to correcting it.
• The effect of the error is delayed. Weak management does not show in a month or two. It shows in team turnover a year later and in decisions that came late, and both are attributed to other causes before anyone traces them to the promotion decision.
• The promotion changed pay and grade together. Reversing it is a question for the pay structure, not only the job title, because in a salary structure the grade is a property of the job, not of the person holding it.

What removes the cause

The remedy lies entirely in the first condition, the promotion criterion, not in appraising whoever was promoted:

• Measure the candidate against the demands of the target level, not against their performance record at the current one. That is done with the same tools used for hiring decisions, including a structured interview with questions derived from the higher role, and real tasks from that role assigned before the decision, which is what a work sample test formalises.
• Provide a track on which pay and standing rise without managing anyone. When management is the only way up, everyone who advances is forced into it. Career pathing is where that track is designed.
• Base promotions on a list prepared in advance rather than on a vacancy filled within a week. That is what succession planning provides. A list allows the people on it to be prepared before they are needed; a sudden vacancy allows only the choice of whoever is nearest.
• Set an initial, announced period in the new level, at the end of which the decision is reviewed, on the understanding from the start that it is reviewable. That is what makes reversal an expected step rather than a sudden penalty. What follows from changing a grade or pay after a promotion depends, among other things, on the terms of the contract and what the two parties agreed, and we do not settle it here; the English guide to employment contract requirements in Saudi Arabia is where contract terms are covered.
• Separate what excellent performance is rewarded with from what earns a promotion. Many promotions are not decisions about the structure at all, but the only reward the organisation has available. Once tools exist to reward people within a single level, the link between performing well and moving to a different role comes apart, and each decision is taken for its own reason.

Internal mobility runs in three directions, one of which is the sideways move that carries no visible promotion. It is the least requested direction and the one that builds breadth of experience most, and here it offers another way out: movement that satisfies the need to progress without placing anyone at a level they were never measured for.

How to tell whether it is happening in a given organisation

The phenomenon is described in general terms, and the practical question is narrower: is it happening here? That is not answered with an opinion about anyone, but with three readings of figures that already exist:

• How team metrics vary with who manages the team. Comparing teams with each other over the same period shows whether one team consistently departs from the rest. The condition people forget is that the difference may lie in the work itself rather than in who manages it. The caution that applies to rater bias holds here too: assuming a bias where there is none is an error that mirrors the bias itself.
• What actually fills the new manager’s time. A promoted person who is still doing their old technical work and putting off the work of the new level has not really moved; a task has simply been added to them. That shows in calendars and work schedules before it shows in any metric.
• Reversals set against promotions. An organisation that has not recorded a single reversal of a promotion in years is not an organisation that has never made a wrong decision. It is one whose machinery cannot correct a decision. The absence of a case from the record is not evidence of its absence from reality.

The probability of success, and the figures that follow from it, are assumed to show how the structure works. We estimate no probability of succeeding after promotion and no rate at which the error the principle describes occurs in organisations. What can be said is one thing: the evidence used in a promotion decision belongs to a different level from the one being decided, and that alone is enough to explain the phenomenon without assuming weakness in anyone.

What sits on either side of it

The principle assumes a sound appraisal put to the wrong use.

Rater bias sits on the other side of that assumption. Its patterns make a rating describe the rater or the instrument, while the principle assumes the rating was entirely correct and places the error in using it for a different purpose. Promotion by seniority sits outside the first condition altogether, since promoting on length of service does not use current performance as evidence at all. It produces a different fault: evidence with no bearing on either level.

A bloated hierarchy or a count of its layers is a matter of span of control, not of this principle, and the principle passes no verdict on the person promoted, since most of those whose rise stopped are competent where they are. Nor does it carry any legal position on promotion, on changing pay or on changing contract terms; those are read from their own provisions and nothing about them is borrowed from here.

Related terms

• Succession Planning
• Structured Interview
• Work Sample Test
• Competencies
• Hierarchy of Needs