A Dutch nurse in a Buurtzorg team does not call a supervisor to schedule a visit, order supplies, hire a colleague, or decide how a patient's care should change. There is no supervisor to call. According to a peer-reviewed case study by Martela and Nandram (Journal of Organization Design, 2025), Buurtzorg ran more than 900 self-managing teams, capped at roughly twelve nurses each, with about 14,700 staff (as of January 2021) and only two directors, some fifty back-office people, and no middle managers or supervisors at all. Roughly twenty-one internal coaches supported around 950 teams, and even they held no formal authority and carried no responsibility for team productivity.
Those numbers travel well as a headline and badly as a blueprint. When organizations copy Buurtzorg, they usually copy the org chart: dissolve a management layer, announce that teams are now "self-managing," and wait for the savings. The savings are real but easy to misread. The same study reports overhead of about 8 percent against a Dutch home-care sector average nearer 25 percent, a figure drawn from a KPMG evaluation from the mid-2010s rather than a current audit. And the cost is not lower on every axis: Buurtzorg's cost per hour of care was actually higher than a comparison provider's. The gains show up in fewer hours of care needed per client, not in a cheaper hour. Strip out the model's logic and you keep the higher hourly cost without the mechanism that offsets it.
What actually travels
The exportable part of Buurtzorg is a belief about where competence lives. A community nurse, standing in a patient's kitchen, usually knows more about what that patient needs than anyone in a head office. The model's design decision is to put authority where the information already is, and then to protect that decision from the normal organizational instinct to claw control back. That instinct is the enemy. It is the same instinct Ricardo Semler describes hitting at twenty-five, when he collapsed in a factory and a doctor told him he was the healthiest sick person the doctor had seen, nothing wrong except the way he was living. As Ricardo Semler has told it, he began that afternoon dismantling the machinery of control: the org chart, the time clock, the six-day week. Later Semco stopped counting hours and simply measured whether the pump got delivered, and absenteeism fell. Buurtzorg is that same wager, made permanent and made about care instead of pumps.
That wager is portable across borders and even across sectors, because it is a stance, not a structure. You can hold it in a hospital in São Paulo or a bank branch in Stockholm. What you cannot do is hold the structure without the stance and expect the numbers.
What stays home
The Buurtzorg researchers are unusually blunt about boundary conditions, and this is the part copycats skip. The model works best where three things hold: low interdependence between teams, so a team can act without waiting on five others; frontline workers who are highly motivated and genuinely competent, so autonomy produces good decisions rather than confident bad ones; and highly customized services, where local judgment beats a central script. Community nursing sits squarely in that sweet spot. A tightly coupled assembly line, a call center run to the second, or a business where teams must constantly synchronize does not.
The authors also note why Buurtzorg held its shape while flatter tech firms such as GitHub and Valve added hierarchy as they grew. Buurtzorg was designed around autonomous teams from the start, with a hard cap on team size, rather than bolting self-management onto an existing pyramid. Structure followed philosophy, not the reverse. Import the philosophy first and the structure has something to grow from. Import the structure first and it grows back into a pyramid, because the underlying beliefs never changed.
There is also a quieter caution. The most cited external figures, the overhead comparison above all, come from a specific time and a specific evaluator, and one of the model's most sympathetic researchers has research ties to the organization. None of that makes the case false. It makes it a case, not a law. The honest reading of Buurtzorg is not "abolish managers and watch overhead fall." It is "find the work where frontline judgment is decisive, put authority there, cap team size so coordination stays human, and accept that this will not fit every kind of work."
The Dutch export, then, is a question, not a template. Where does the real knowledge sit, and are you willing to move power to meet it? Answer honestly and Buurtzorg travels. Answer by redrawing the boxes and it stays home.
Sources
- Martela, F. & Nandram, S. (2025). Buurtzorg: scaling up an organization with hundreds of self-managing teams but no middle managers. Journal of Organization Design 14. Documents >900 self-managing teams, ~14,700 staff, no middle managers, and the boundary conditions for the model. https://link.springer.com/article/10.1007/s41469-024-00184-y
- Corporate Rebels (2025). Frederic Laloux, Jos de Blok, Lisa Gill at Corporate Rebels Summit 2025. Describes Buurtzorg (founder Jos de Blok) as roughly 15,000 nurses in autonomous self-managed teams. https://www.corporate-rebels.com/blog/big-news-frederic-laloux-jos-de-blok-lisa-gill-at-corporate-rebels-summit-2025