The Peter Principle: Do People Really Rise to Their Level of Incompetence?

Laurence J. Peter and Raymond Hull meant the Peter principle as a joke. In their 1969 satire The Peter Principle, people in a hierarchy keep being promoted for doing their current job well until they reach a job they do badly, and there they stay. Many readers took it as a serious point about how promotions work.¹

Is it true? Partly. The joke rests on a real tension: the skills that earn a promotion are not always the skills the next job needs. The strongest test so far, a 2019 study of sales workers at US firms by Alan Benson, Danielle Li and Kelly Shue, reported that the best sellers were the likeliest to be promoted and tended to make weaker managers.² Part of the slump after a promotion may be plain statistics, and a famous case for promoting at random comes from a model, not from real firms.

In practice, organizations can reward strong work without forcing a move into management, and anyone offered a step up can try the new job’s work before taking its title. This review looks harder at one finding from our guide to what research says about getting ahead at work.

What is the Peter principle? A 1969 joke with a serious core

Peter was a Canadian educator who began as a teacher in Vancouver in 1941 and later taught education at the University of Southern California in the US.³ Hull wrote most of the text from Peter’s research. The logic is simple: a competent person is promoted into a job that needs new skills. If they manage it, they are promoted again; if not, they stop rising and stay in a job they do poorly, which Peter called their plateau.¹

The book’s own example is a schoolteacher who is good with children, becomes a capable assistant principal who handles parents and staff well, and then struggles as principal, a job that turns on relations with the school board. Each step was earned, and the last one asked for something the earlier jobs never tested.¹

Strip away the jokes and the principle makes a claim you can check, built on two conditions. Promotions must reward performance in the current job, and the next job must need different skills. Where both hold, a mismatch is likely; where either fails, the principle has little to say. So when a new manager struggles, ask which of the two conditions held before blaming the person.

Myth

The Peter principle is a proven law of how organizations work.

Fact

The 1969 book was satire. The best real-world test, in sales, supports a narrower claim: promotions can reward skills the next job does not use.

Is the Peter principle real? What a study of sales workers found

The best evidence supports a narrower version of the Peter principle. In a 2019 Quarterly Journal of Economics paper, Benson, Li and Shue reported that firms favored their top sellers for promotion over colleagues with more promise as managers, and that teams improved less under managers who had sold more. The authors call this the first large-scale empirical evidence of the principle.²

The study

Moderate evidence

Promotions and the Peter Principle (2019), The Quarterly Journal of Economics

Drawing on client firms of a sales-software company, the economists scored each new manager by how much their team’s sales improved. Among workers who only just made the cut for promotion, higher past sales went with weaker managing. Twelve months after a promotion, higher sellers passed over for a weaker-selling teammate were 23 percentage points less likely to still be at the firm. The authors estimated that promoting purely for management potential would have raised average manager quality by about 30%.²

Two limits matter. The study covers one occupation in one country, picked for its clean performance measure, and manager quality means only the team’s sales. It is also an observational studyobservational study: A study in which researchers record what people already do or are exposed to, rather than assigning anyone to anything. It can show that two things go together, not that one causes the other, because the groups being compared may differ in other ways as well.Full entry in the glossary of company records, not an experiment, though the authors tested and rejected rival explanations, such as a lucky streak before promotion or differences in the teams new managers were given.²

Promoting the top seller need not be a mistake, the authors argue. A promotion is also a prize, and passing over the best seller has a cost, as the departures show. Some firms seemed to weigh that trade-off: those whose managers ran larger teams, and those with stronger pay for performance, put less weight on sales when promoting.²

The study also found a better signal that firms largely ignored. The more colleagues a worker typically shared sales credit with, the better a manager they tended to make, yet promotion did not reliably follow. The authors suggest firms may hold back from rewarding it because shared credit is easy to game.² Picture a team where one rep closes the biggest deals alone while another keeps getting pulled into colleagues’ accounts to help them close. The first gets the title; the second has been practicing the job.

If you decide promotions, ask who already does parts of the next job, not only who is best at this one. And if you are passed over, the urge to leave is a pattern the study recorded, which is worth recognizing before you act on it.

Regression to the mean: why results often dip after a promotion

Some of the slump after a promotion would happen even if every promotion were right, the economist Edward Lazear argued in a 2004 paper in the Journal of Political Economy. His case is a theoretical model, not new data: performance mixes lasting ability with luck, and people tend to be promoted after a lucky run.⁴

1. Usual level: the lasting skill under each result
2. A good run: skill plus some luck, which is often when promotion comes
3. Back toward usual: the luck does not travel, so results drift down even when the choice was sound

The logic is regression to the mean. To be promoted you must clear a standard, and people who clear it are, on average, having a better-than-usual spell. Luck does not follow anyone into the new job, so their results there drift back toward their usual level. Lazear argued that firms that understand this set the bar higher, that the drop is biggest in jobs where luck matters most, and that the same logic explains why film sequels tend to disappoint.⁴

A laboratory experimentlaboratory experiment: An experiment run under conditions the researchers set up and control, often with students doing short tasks. Control makes cause and effect easier to pin down, but the tasks are simpler than real life, so results do not always carry over to work or home.Full entry in the glossary backs the logic under narrow conditions. In a 2007 study by David Dickinson and Marie Claire Villeval with 190 undergraduates in Lyon, France, performance fell after promotion only when the random part of the task’s score was large. When participants chose their own task instead, a set promotion rule still usually sorted people into jobs better, probably because people misjudged how much of their score was luck.⁵

Lazear’s model cannot explain the sales study, though. If luck were the whole story, the best sellers would still make the best managers, just less spectacularly than their records promised. Benson, Li and Shue found the opposite, so in sales the mismatch of skills looks real.²

An account manager promoted after the quarter a huge client signed will likely have a worse next year, and would have even had she stayed put. So judge promotion candidates on several periods, not their best one, and don’t read your own dip after a move as proof you were promoted too far.

Should companies promote at random? The Ig Nobel simulation

A 2010 computer simulation suggested that promoting people at random could beat promoting the best, but only under one strong assumption. Alessandro Pluchino, Andrea Rapisarda and Cesare Garofalo of the University of Catania, Italy, modeled an organization of six levels and tested promotion rules in it; their paper appeared in Physica A.⁶

Everything turns on an assumption about skill. If competence at one level says nothing about competence at the next, the paper’s “Peter hypothesis”, promoting the best member made the simulated organization about 10 percent less efficient over time; if competence carries over, promoting the best helped. Random promotion gave small gains under either assumption, so the authors judged it one of the two best rules when you cannot tell which world you are in; the other alternates between promoting the best and the worst.⁶ The work won the 2010 Ig Nobel Prize for management, awarded for showing mathematically that organizations would become more efficient if they promoted people at random.⁷

Treat it as a thought experiment: it contains no real employees, and its conclusion follows from its assumptions. Real jobs probably sit between the two extremes. A skilled lab scientist who becomes head of the lab keeps her scientific judgment but must now handle budgets and hiring. The model forces a useful question: how much of the skill being rewarded today does the next job use?

The main sources, side by side:

How companies avoid the Peter principle

Organizations limit the Peter principle by rewarding great work without forcing a move into management, and by choosing managers on evidence of managing, as some firms in the sales study already did for jobs leading larger teams.

A long-standing fix is a dual ladder: one track leads into management, the other offers advancement for technical work. Industry has used it for years to motivate technical staff, though its effectiveness has often been questioned. In a survey of nearly 1,500 engineers and scientists at nine US organizations, published in 1986 by Thomas Allen and Ralph Katz, preferences differed markedly by age, with a strong increase in the share who wanted more interesting project work over either kind of advancement.⁹ So before building ladders, ask people what they want: in that survey, not everyone wanted a higher rung.

Three other practices follow from the research above, though none has yet been studied as a remedy:

• Judge candidates on signs of the next job’s skills, such as helping colleagues succeed, the signal the sales study found firms underused.
• Look at several periods of results rather than a best quarter, so luck counts for less.
• Make moves testable and reversible: a time-limited acting role, a project to lead or cover for a manager on leave. The writers Brian Christian and Tom Griffiths have proposed a hierarchy in which people regularly move up or back down, so anyone promoted past their level is soon moved to where they work well.¹

Before you promote someone

• Does the new job need skills the current one never tested?
• Have they already done some of that work, and how did it go?
• Are you judging several periods of results, not their best one?
• Is there a way to step back without disgrace if it does not fit?
• Could you reward them without changing their job?

Before you accept a promotion, ask whether you want the job

The Peter principle assumes everyone wants to climb, and you don’t have to. Peter and Hull’s tongue-in-cheek advice was to stay below your level of incompetence, even by appearing less able than you are, a tactic they called creative incompetence.¹ The serious point needs no faking: a promotion is a new job.

So find out what the days in the new role are made of before you say yes. If it means managing, ask for a slice of it first: lead a project, run a meeting series or cover while your manager is away. Don’t judge by feel alone; participants in the lab study appeared to misjudge how much of their own success was luck. If you test the work and still want it, our guide to making the case for a promotion covers the next step.

What is the Dilbert principle, and what is the opposite?

The Dilbert principle is a different joke: that companies promote their least competent people into management to keep them away from the real work. Scott Adams, the creator of the Dilbert comic strip, set it out in a 1995 Wall Street Journal article and a 1996 book. Unlike the Peter principle, it assumes the promoted person was never good at their earlier job.¹⁰

The Peter principle has no agreed opposite. One candidate is the Paula principle, a term some authors use for the pattern of women being kept in jobs below their abilities.¹ Another is the mirror image in Lazear’s model, in which people turned down for promotion tend to do better afterwards.⁴ Both point the same way: a hierarchy can misplace people below their level as well as above it. A struggling manager parked in a vague “special projects” post and a sharp analyst never put forward are two sides of one problem, so look for both when you judge how your organization places people.

The bottom line

Where promotions reward the current job and the job above calls for other skills, good workers can land in roles that fit them poorly, and part of any slump after a promotion is luck running out. Organizations can reward excellent work without forcing a change of job and choose managers on evidence of managing. If you are offered a promotion, try the next job’s work before you take its title.