Briefing
Making You Think 2 · Primer and discussion · About 16 minutes in two parts
Download the MP3 · Linda and Leonard, AI voices, read a script written from the sources below.
Download the MP3 · Linda and Leonard, AI voices, read a script written from the sources below.

This episode is about The Peter Principle: Why Things Always Go Wrong, by Laurence J. Peter and Raymond Hull. Find a copy on Amazon Australia, or ask your local bookshop or library.
Three questions to ask yourself
Two things to try, alone, with a pen
Laurence Johnston Peter was born in Vancouver, British Columbia, on 16 September 1919. The New York Times obituary reports that he taught in Vancouver schools, took a bachelor's degree in 1957 and a master's in education in 1958, and worked as a psychologist and assistant professor at the University of British Columbia. In 1966 he moved to California to teach education at the University of Southern California. His first book was a textbook, Prescriptive Teaching (McGraw-Hill, 1965). By the obituary's account the Peter Principle manuscript was first sent out in 1964, was turned down by thirty publishers and was bought by William Morrow for US$2,500; it came out in 1969, sold more than 200,000 copies in its first year, and he left academic life in 1970. The obituary also reports that he refused offers to become a management consultant, saying he did not want to rise above his own level of competence. He died at home in Palos Verdes Estates, California, on 12 January 1990, aged 70. Raymond Hull (1919 to 1985) is described by his British publisher as a playwright, television writer and lecturer. BC BookWorld reports that the two met as strangers at an amateur production at Vancouver's Metro Theatre.
Written from public sources and checked on 12 October 2026. The script was drafted from search-inside snippets of the book, the publishers' descriptions, the New York Times obituary and one chapter summary, and reviewed by a person who has read the book. Read in full: the obituary, the UK publisher's page and the Catania simulation paper. Read in part: the sales study by Benson, Li and Shue (abstract, introduction, data and main results) and Ironies of Automation (abstract and first section). Read in abstract only: Lazear's paper. The link to AI tools is the presenters' own, not the authors'. Check the book itself before quoting.
Before we record, every claim the hosts make is checked against the sources (this table was last updated 12 October 2026). Confirmed means a primary source or two independent reports; Reported means one party says so; Partly right, Disputed, Unverified and Not yet known mean what they say, and the hosts say so on air. This is the table as we left it.
Sign up for our podcast at agile-insights.com.au.
Linda If you reward people for doing a job well by giving them a different job, you will keep moving them until they reach one they cannot do... and then leave them there. I'm Linda, and this is the second episode of Making You Think, a series that takes one book or idea at a time. That idea is the Peter Principle, from a book of the same name, first published in nineteen sixty nine. By the end of this part you will know who wrote it and how the argument works in three moves. You will also know what a few of its best terms mean.
Leonard And I'm Leonard. Start with who wrote it.
Linda First, the authors. Laurence J Peter was born in Vancouver in nineteen nineteen. The New York Times obituary reports that he taught in Vancouver schools. It says he took his first degree in nineteen fifty seven, in his late thirties. He then worked, it says, as a psychologist and assistant professor at the University of British Columbia. In nineteen sixty six he moved to California to teach education at the University of Southern California.
Leonard And his co author?
Linda His co author, Raymond Hull, was also born in nineteen nineteen. Hull's British publisher describes him as a playwright, television writer and lecturer. Hull died in nineteen eighty five, and Peter in nineteen ninety.
Leonard So, that's the authors. Now the book itself. The story of the book matters, and nearly all of it comes from that same obituary. It reports that Peter first sent the manuscript out in nineteen sixty four, and that thirty publishers turned it down.
Linda Thirty.
Leonard Thirty. William Morrow, in New York, then bought it for two thousand five hundred American dollars. By the obituary's account it sold more than two hundred thousand copies in its first year. And it was written as satire. The obituary's headline says so, and the British publisher still calls it satirical. Peter is reported to have said he was only kidding...
Linda So, a joke with an argument inside it. Here is move one. People are promoted for how well they do the job they have now. The book's own sentence reads: in a hierarchy every employee tends to rise to his level of incompetence. Notice the word... tends. Our example, not theirs: the best salesperson on a team is the obvious choice for sales manager. Nobody is being foolish. Today's results are the only evidence the firm can see.
Leonard So that's move one. Let me check I have that. The promotion is a prize for the last job, and it is also a bet on the next one. Those are two separate things, and the firm is treating them as one.
Linda That's it. Go on.
Leonard I can see why. If I sold the most, it would feel unjust to promote someone who sold less. But it only goes wrong if the new job is truly different. Is it?
Linda Good question. That is move two. The next job is a different job. Selling is one skill, and managing people who sell is another. One summary of the book gives its early example as a teacher who is promoted to principal. Teaching a class and running a school are separate kinds of work. Being good at the first tells you little about the second.
Leonard Give me a picture.
Linda Think of a brilliant chef who is made restaurant manager. She now spends the day on rosters, suppliers and complaints. The food was never the problem, and it is no longer her job...
Leonard And if she turns out to be a poor manager, what then? I think most of us know the answer. She does not get the next promotion, because she is not doing well. But nobody sends her back to the kitchen either. That would look like a punishment, and it would embarrass whoever promoted her. So... she stays. Is that the third move?
Linda It is. Move three: nobody goes back down. A person stops being promoted when they stop doing well, and then they stay where they are. The authors call that post your level of incompetence. It means the first job you cannot do. It does not mean you are a fool. You were good at everything one step below.
Leonard That's a relief.
Linda For now. From there the book draws what it calls Peter's Corollary. In time, every post tends to be filled by someone who is not competent to do it. And the real work, the book says, is done by people who have not reached that level yet.
Leonard Present company included, I hope.
Linda Let's hope.
Leonard So, three moves down. Now the terms. They even invented a mock science for it, hierarchiology, the study of hierarchies. Which of their other terms are worth keeping?
Linda Three. The first is percussive sublimation, which is being kicked upstairs. Someone is given a grander post to move them out of the way.
Leonard I have seen that done.
Linda Most of us have. The second is the lateral arabesque. That is a sideways move dressed up as progress, usually summarised as a move to a longer title. The book treats both as false promotions, and so as only apparent exceptions to the rule.
Leonard And the third?
Linda The third is creative incompetence, the title of chapter fourteen. It is about avoiding the last promotion without refusing it. As summaries describe it, you look unsuited in some harmless way... and stay in the job you do well.
Leonard Tempting.
Linda Don't you dare. So, those are the terms. That is the argument, and it needs only those three moves. So here is one question to take with you. Think of the last promotion you saw up close. It might have been your own, or a colleague's. What was it a reward for... and what was it a prediction of? Those may have two separate answers. Try to name both before you go on.
Leonard Take your time with that one. I am not going to answer that for anyone. Give it a minute before you press play again. In Part Two we ask whether the Peter Principle is true. We put the case at its strongest. Then we look at the one large study that tested it, and at two complications that change how you should read it. Join us there.
Linda If you reward people for doing a job well by giving them a different job, you will keep moving them until they reach one they cannot do... and then leave them there. I'm Linda, and this is Part Two of the second Making You Think. That idea is the Peter Principle, from the nineteen sixty nine book by Laurence J Peter and Raymond Hull. Your level of incompetence is the first post you cannot do. Peter's Corollary says that in time every post tends to be filled by someone who cannot do it. By the end you will have the best case for the idea, the evidence, two complications, and some thinking to do.
Leonard And I'm Leonard. First, the case at its strongest. The argument needs only three things to be true. People are promoted on how they do today's job. The next job asks for different skills. And there is no easy way back down.
Linda The three moves from Part One.
Leonard The same three. None of those needs a fool or a villain. Every decision in the chain looks fair when it is made. Everyone in the story was good at their work, one level below. The fault sits in the rule... and a rule can be changed.
Linda That's the case for. Now the case against. The weakness is just as plain. It was written as satire. The New York Times obituary of Peter says so in its headline, and reports him saying he was only kidding. A satire can be right, but it cannot be its own evidence.
Leonard So, is there any?
Linda Then, in twenty nineteen, three researchers put it to a test. Alan Benson, Danielle Li and Kelly Shue were at the University of Minnesota, M I T and Yale. Their study appeared in the Quarterly Journal of Economics. They call it the first large scale evidence of the Peter Principle.
Leonard Here is what they had. Records from a company that sells sales performance software, covering two thousand and five to two thousand and eleven. One hundred and thirty one firms based in the United States. Thirty eight thousand, eight hundred and forty three sales workers, of whom one thousand, five hundred and fifty three became managers.
Linda And what did they find?
Leonard Two findings matter. The top seller on a team was about three times as likely to be promoted in a given month.
Linda No surprise there.
Leonard None. And doubling a person's earlier sales went with a six point one per cent fall in the sales of each person they then managed...
Linda There is a third finding. People with a record of working with others on shared sales turned out to be better managers. Yet that record did not help them win the promotion.
Leonard No. Naturally not.
Linda So that's the evidence. That leads to the first activity. It takes five minutes and a sheet of paper.
Leonard Pens out, then.
Linda Pens out. Write down the five things you spent last week doing. Beside them, write the five things the person one level above you spent last week doing, as best you can tell. Tick anything on both lists, and count the ticks. Then look at the unticked items on their list. Which of those has anyone ever seen you do?
Leonard I will leave you with your two lists... and make a claim. That study settles it. The best sellers were promoted, and the better they had sold, the less their new teams sold. Peter and Hull were joking in nineteen sixty nine, and fifty years later the data say the joke was accurate. The firms were getting it wrong. If I ran one of them, I would stop promoting on sales tomorrow...
Linda Tomorrow? The authors would not go that far, and I think they are right. They say they do not conclude the firms are making mistakes. A promotion is also a prize. The chance of winning it may make the whole team sell harder, and a firm might accept weaker managers as the price.
Leonard Hm. Go on.
Linda And their version of the principle is narrower than the book's. They show firms give too much weight to today's performance. They do not show that every post ends up badly filled.
Leonard I accept that. The pattern is there in those sales teams, and whether it is a mistake depends on what the promotion was buying.
Linda Good. That's settled, then. Two complications to come.
Leonard Now the first complication. It comes from a paper by Lazear in the Journal of Political Economy, in two thousand and four. We are going only on the paper's own summary. A promotion shows that someone met a standard. Part of meeting any standard is luck... and luck does not repeat.
Linda Sadly not.
Leonard No. So on average, performance falls after a promotion. The summary adds that firms allow for this, and the effect still never disappears.
Linda Put plainly, some of the decline is luck running out. Imagine someone promoted after their best year ever. Their next year would probably have been more ordinary in any job, old or new.
Leonard So is it only luck?
Linda I do not think it is only luck, and this is our reading, not Lazear's. In the sales study, a visible habit predicted who would manage well. Luck alone would not produce that. Both things can be true at once. So a question for you. Think of someone you have said is not up to the job. How would you tell a person in the wrong post... from a person in their first hard year?
Leonard Take your time with that one. Hold that, because the second complication is about the famous cure. The cure people quote is... to promote at random. It comes from a paper in the journal Physica A in twenty ten, by Pluchino and two colleagues at the University of Catania. It won that year's Ig Nobel Prize for Management. It is a computer simulation, with no real workers in it.
Linda Nobody was harmed, then.
Leonard Not one. The model organisation has one hundred and sixty posts on six levels. Suppose skill in the new job has nothing to do with skill in the old one. Then promoting the best lowered efficiency by about ten per cent, and promoting the worst raised it by about twelve...
Linda Here is the half that is usually left out. The same model was run with the opposite assumption, that skill carries over from one level to the next. Then the result reverses.
Leonard Reverses?
Linda Reverses. Promoting the best raised efficiency by about nine per cent, and promoting the worst lowered it by about five. Random promotion gave a small gain in both worlds, of one or two per cent. So the finding is not that random is best. It is that random is the choice you will not regret... if you cannot tell which world you are in.
Leonard I had only heard the first half.
Linda Most people have. So, two complications down.
Leonard One short step sideways. The authors wrote in nineteen sixty nine, long before these tools. What follows is our own extension, not theirs. Suppose an AI tool does a small task well, and so it is handed a wider one. That is the promotion rule again: a reward for the last job, with no evidence about the next. And a person moved from doing the work to checking a machine's work has been given a different job too.
Linda Go on.
Leonard A nineteen eighty three paper called Ironies of Automation, about industrial control and not AI, saw that coming. After automation the operator is left to watch the machine and take over when it fails, with skills that have gone unpractised.
Linda I would hold that loosely. We found no study that tests it. But it points at a useful question, for people and for tools. What can you observe now that is about the next job, and not this one? That is the second activity. Think of someone you might one day recommend for a bigger job. Write down what you have seen them do well. Now cross out everything that is only about their present job.
Leonard A shorter list, I expect.
Linda Usually. Then write one thing you could ask them to try in the next month that would tell you something about the next job.
Leonard Worth doing slowly. Before we close, what is not known. The one large body of data covers a single occupation, in one country, over seven years. We found nothing that tests the full claim, that whole organisations fill up with people who cannot do their jobs. So a last question. Have you ever seen someone moved back to a job they were good at... and treated as a success for it? If not, why not?
Linda Sit with that for a moment. So, what to carry away. One, a promotion is a reward and a prediction at once, and it helps to ask which you are making. Two, the evidence supports the pattern in one setting, and its own authors stop short of calling it an error. Three, some of what looks like decline is luck running out, so be slow to judge a first hard year. And four, look for evidence about the next job before you fill it. Read the book if you can, and check our account against it.
Leonard Please do.
Linda If you'd like to know more, sign up for our podcast at agile-insights.com.au.
Leonard Thanks for listening.
Every episode tagged #leadership
All of Making You Think (13) · Follow Making You Think in Apple Podcasts or any podcast app
Everything from Agile Insights: briefings, Making You Think and the reference series