For fifty years, Oncken and Wass’s Who’s Got the Monkey? has been the best explanation of why work boomerangs back to the leader: the leader keeps accepting it. That describes a real failure. It misses the more common one — and the one that costs six weeks instead of six minutes.
Key Takeaways
- The classic: Oncken and Wass’s monkey-on-the-back model says work returns to the manager because the manager keeps accepting the next move.
- The assumption: Every scene in the article turns on possession. The assignment itself is treated as fixed and unambiguous.
- The gap: The more expensive failure is a handoff that never completed, because the person walked away holding different work than the one you assigned.
- Why the discipline cannot catch it: There was no moment of acceptance to refuse, so a manager following the rules perfectly still gets the work back weeks later in the wrong shape.
- The practice: Before the conversation ends, have the new owner say back the job, the finish line, and their latitude. Ninety seconds.
The Article Everyone Can Quote
Some management writing gets read. Very little of it gets into the language. In 1974, William Oncken, Jr. and Donald L. Wass published an article in Harvard Business Review that did — and if you have ever said you had too much on your back, you were quoting it without knowing. “Management Time: Who’s Got the Monkey?” became one of the publication’s two best-selling reprints of all time, was reissued as a Classic in 1999 with a commentary by Stephen Covey, and now sits in Harvard Business Review’s 10 Must Reads for New Managers. Fifty years on, it is still the first thing most executives are handed when they say they are drowning.
The setup is elegant. Oncken splits a manager’s time into what the boss imposes, what the system imposes, and what the manager imposes on themselves — and then observes that a large portion of that last category is not self-imposed at all. It is imposed by the people who report to you, one conversation at a time.
The monkey is the next move. A manager walks down a hallway. Someone stops them with a problem framed as a shared one, opening with the word we. The manager knows enough to be interested and not enough to decide on the spot, so they say the sentence every leader has said: let me think about it and I will get back to you. Twelve words, no meeting required. The monkey has changed backs.
The manager has just accepted a responsibility from their own subordinate and promised them a progress report, which are the two things subordinates normally do for bosses. And now the person who reports to you will stop by cheerfully to ask how it is coming, which Oncken drily calls supervision.
Multiply that by four people and three monkeys a day and by Friday the manager has sixty of them, a closed office door, and a growing reputation for being unable to decide anything. The article ends with a discipline of refusal, five rules, and five degrees of initiative. His first rule — “Monkeys should be fed or shot” — is the line everyone remembers.
It is a wonderful piece of writing and most of it still works. What follows is not a takedown. It is an argument that the model can only see one of the two ways delegation fails, and that the one it cannot see is the more expensive.
The Assumption Underneath
Read the four scenes again and notice what varies and what does not.
In every one, the question is where the monkey is. On the subordinate’s back, on the manager’s back, straddling both. The manager’s error is always an error of acceptance — agreeing to think about it, asking for a memo that will land back in their own inbox, offering to help in a way that makes their approval the gate, volunteering to draft the objectives.
What never varies is the monkey itself. Across all four scenes, Oncken treats the work as a fixed object whose identity is settled and whose only interesting property is location. Both people know what the animal is. The only question is whose back it sits on.
That assumption is doing enormous quiet work, and it is the reason the model has a blind spot.
The monkey model solves for who is holding the work. It cannot tell you whether the same work exists in both heads.The Possession Assumption
The Fourth Scene, Which Almost Says It
Oncken’s last example is the interesting one, and he walks right up to the edge of a different diagnosis before turning away.
A manager has a person newly transferred in to launch a business venture. The manager says they should get together soon to work out a set of objectives for the new role, and offers to draw up an initial draft for discussion. Oncken’s own analysis is precise and worth quoting in substance: the person has the job by formal assignment and the full responsibility by formal delegation, and yet the manager holds the next move, so the person is immobilized.
Stop there. Formal assignment happened. Formal delegation happened. Full responsibility was conferred. And the work is not moving.
Oncken reads that as a possession failure — the manager grabbed the draft. There is another reading, and it explains far more of what happens in real organizations. Nobody in that conversation knows what the job is yet. That is precisely why they need to work out objectives. The manager’s instinct is to write the draft, which is a transmission move, and the transmission move is what stalls it.
The version I watch play out most often does not stall at all, which is what makes it costly. The manager writes the draft. The person reads it, nods, and leaves. Both believe the handoff is complete. And six weeks later the work arrives shaped like something neither of them would have described. No monkey ever jumped. The refusal discipline had nothing to refuse.
What Actually Crosses Between Two People
Here is why that happens, and it is not a character flaw in anyone involved.
Your senses take in information at something like eleven million bits per second. Your deliberate, thinking attention runs at a few dozen bits per second. Researchers argue about those figures — one newer line of work puts intake nearer a billion and conscious thought closer to ten — and every version of the argument makes the gap larger than any manager plans for.
So when you assign something, you compress weeks of context into ninety seconds of speech. The reasoning, the three options you already discarded, the board conversation that changed the timeline, the customer you have in mind, what you would consider finished, how much rope this person has — nearly all of it stays in your head. It has to. There is nowhere else for it to go.
And what does arrive does not get received. It gets rebuilt, inside a structure you did not build and cannot inspect: their last performance review, their read of what happened to the person who tried this in 2023, their assumption about how much latitude your tone implied. Two people leave that hallway holding two different animals, and both of them are certain the handoff went fine.
Chip Heath and Dan Heath named the mechanism in these same pages — the curse of knowledge. Once you know something you cannot recreate not knowing it, which makes you a poor judge of what your own words carry. The version I find most useful comes from a study by a Stanford graduate student named Elizabeth Newton. Tap the rhythm of Happy Birthday on a table for someone. While you tap, you hear the entire song. They hear knocking. Tappers in her study predicted half their listeners would name the tune. Two and a half percent did.
Now put that on the assignment. You are not handing over a monkey. You are tapping one.
Two Failures, One Discipline
There are two ways a piece of work returns to a leader, and conflating them is what keeps competent people stuck.
The first is Oncken’s. You accepted the next move. You said you would think about it, review the draft, make the call, talk to the other department. This failure is visible in the moment, it happens in the conversation, and his discipline of refusal corrects it cleanly. Anyone can learn to catch it.
The second failure produces no moment at all. You accepted nothing. You delegated fully, named the owner, agreed the next step, and put it in the calendar exactly as instructed. Six weeks later the work arrives requiring your judgment to repair, and when you trace it back there is no point where a monkey jumped. The transfer never completed. What returned was not your monkey coming home. It was a different animal you have to deal with, and it is now expensive, and someone did six weeks of sincere work in the wrong direction.
A manager applying Oncken’s rules perfectly is defended against the first failure and completely exposed to the second. Worse, the discipline can make the second one harder to see, because refusing the next move buys weeks of silence and silence reads as ownership.
Refusing the monkey is cheap and visible. A transfer that never completed is expensive and silent, and no discipline of refusal detects it.The Two Failures
Rule Four, With the Right Instinct and the Wrong Reason
The strongest evidence that Oncken sensed the second failure is buried in his fourth rule.
He instructs managers to handle monkeys face-to-face or by telephone and never by mail. His stated justification is possession: with mail, the next move ends up back with the manager. Fair enough as far as it goes.
But that is not the main thing a live conversation buys you, and I suspect Oncken knew it from practice without having a theory for it. Face-to-face is the only format in which you find out what the other person built. You hear the pause before the agreement. You hear them restate the deadline as a different date. You catch the moment their version of finished turns out to be smaller than yours. A memo cannot do that, not because it puts the next move on your desk, but because it gives you no view of the reconstruction.
Oncken had the right rule and filed it under the wrong cause. And this is the pattern worth watching across the management classics: the memorable, quantified half travels for fifty years, and the harder half gets left in the original.
It happened again to this very article. When Harvard Business Review reissued it in 1999, Stephen Covey wrote a commentary observing something Oncken never addressed — that many managers are not victims of monkey-jumping at all but are eager to take the monkeys, and that even confident executives find it painful to hand over control. Covey traced that to a belief that rewards are scarce and fragile. The five rules went on to sell for another quarter century. The commentary about why managers want the monkeys mostly did not travel at all.
The Ninety Seconds That Complete a Handoff
The correction is small, and it goes at the end of the conversation rather than in a new process.
Before anyone leaves, ask the person who now owns the work to tell you three things in their own words: what they are going to do, what finished looks like, and how much latitude they think they have. Then stay quiet and listen for divergence rather than for agreement. Where is their deadline different from yours? Their customer? Their definition of done? Their sense of whether they can spend money without asking?
Notice what this is not. It is not asking whether they have questions, which tests willingness to speak in front of you and returns silence in exactly the rooms where silence means least. It is not sending a written summary afterward, because a summary you wrote is one more tap on the table.
Oncken’s five degrees of initiative deserve the same treatment, incidentally, because they are useful and because they are quietly ambiguous. Recommend and then act, and act and advise at once, sound like clean categories from the manager’s chair. In practice, one person’s level three is another person’s level four, and the gap between them shows up as either an unwelcome surprise or an unnecessary delay. Naming the level is good. Hearing them name it back is what makes it real.
Underneath the practice is a belief, and the belief is the actual work. Most capable leaders operate on an assumption they have never inspected: if I said it clearly, it was received. Hold that assumption and you will read every returning piece of work as a monkey somebody let jump, which sends you looking for firmer refusals. Replace it — you have no evidence of reception until you have seen the decoding — and the last ninety seconds of every assignment change without anyone needing a new operating cadence. No one can empower people; people already have power. Your job is to unleash it, and a handoff that never completed is one of the most common ways that power stays parked.
What you are after is a person who is on the hook for work they can actually describe. Ownership of a job nobody defined is not ownership. It is exposure.
Where This Argument Could Be Wrong
Four places, named before you find them.
Sometimes the manager really does just grab it. Plenty of leaders take the monkey because taking it feels good, feels fast, and confirms that they are still the one who can. Covey saw this in 1999 and he was right. For that failure, Oncken’s discipline is the correct medicine and this article adds nothing.
The calendar problem is real and Oncken solved it. The distinction between boss-imposed, system-imposed, and discretionary time remains one of the more useful pieces of management thinking anyone has produced. Verification does not replace it. Verification is what you spend some of the recovered discretionary time on.
Some of this is not a communication problem at all. Work that keeps returning is often a decision-rights problem wearing a delegation costume — nobody ever allocated the choice, so it comes back to the only person authorized to make it, dressed as a misunderstanding. No amount of verified reception repairs a structure that never assigned the decision. Verification is useful here mostly because it reveals which problem you have.
And the least comfortable one: reading this will not change your practice. Awareness is not the mechanism. Either the playback becomes a standing close on assignment conversations, visible in how meetings end, or it stays a thing you found interesting on a Tuesday. Individual insight against an organized system loses every time.
The Crossing
This argument is one application of a longer framework paper. The Crossing maps the four gates a message passes through between two minds — Intake, Compression, Transit, and Reconstruction — shows where the loss actually occurs, and sets out the practice that verifies reception at each one. It includes the full research base, the figures, and a section on where the model could be wrong.
The Half That Got Kept is a CO2 Coaching series on management classics whose memorable half travelled and whose harder half stayed behind. The next piece examines John Kotter’s fourth error and the reception test his readers left in the original.
Frequently Asked Questions
What is the monkey on the back theory of management?
In their 1974 Harvard Business Review article, William Oncken, Jr. and Donald L. Wass used the monkey as a metaphor for the next move on a piece of work. When someone raises a problem and the manager says they will think about it and get back to them, the monkey jumps from one back to the other. The article argues that managers run out of time while their people run out of work because managers keep accepting these next moves, and it prescribes a discipline of refusing them.
Why does delegated work keep coming back to me?
Two different failures need two different fixes. In the first, you accepted the next move during the conversation, and Oncken’s discipline of refusal corrects it. In the second, you refused nothing and the work still returned weeks later in the wrong shape, because the person left holding a different understanding of the job. No amount of refusing corrects that one, because there was no moment of acceptance to refuse.
Is Oncken’s monkey model still useful?
Yes, for the failure it was built to see. The five rules, the five degrees of initiative, and the insistence that every assignment carries a named next step and a named owner remain good practice. The limitation is that the model treats an assignment as an object with a location, so it solves for who holds the work and cannot see whether the same work was ever transferred.
How do you make sure delegation actually sticks?
Verify the transfer before the conversation ends. Ask the new owner to say back, in their own words, what they are going to do, what finished looks like, and how much latitude they believe they have. Then listen for where their version diverges from yours. It takes about ninety seconds and it is the only evidence you will get that the same piece of work exists in both heads.
What did Stephen Covey add to Who’s Got the Monkey?
Covey’s 1999 commentary named something the original never addressed: many managers are eager to take on their people’s monkeys, and even confident executives find it hard to relinquish control, which he traced to a belief that rewards are scarce. The five rules travelled for another twenty-five years. The commentary largely did not.
