Ten years ago, a CEO curious about self-management asked a believer's question or a skeptic's one: is this the future, or is it nonsense? Both questions have aged out. The Zappos experiment, once the movement's flagship, ended in an instructive retreat: after Tony Hsieh's 2015 ultimatum to embrace Holacracy or leave, 18% of the workforce took the buyout, and by 2017 the company had replaced the system with an internal market of teams running their own P&Ls. Meanwhile Buurtzorg, the Dutch home-care organization, kept scaling the same idea to over 900 self-managing nursing teams, roughly 14,700 people as of January 2021, with two directors and no middle managers at all. Same decade, same philosophy, opposite outcomes. The interesting question was never "does it work?" It was always "when?"
That is the question leaders actually ask now, usually in private, and the research has caught up with it.
What the literature settled
The academic anchor is Lee and Edmondson's 2017 synthesis, which defined self-managing organizations precisely, those that radically and systematically decentralize authority, eliminating the manager-subordinate reporting relationship, and summarized what hierarchy is actually good for: reliable execution of known tasks in stable environments. Its documented costs appear elsewhere: rigidity, suppressed problem-solving across functional boundaries, status differences that stunt people's development. This already dissolves the ideological version of the debate. Hierarchy and self-management are not worldviews; they are tools with different performance envelopes.
The boundary conditions came into focus with the Buurtzorg research. Martela and Nandram's longitudinal study identifies why that organization scaled flat where others could not: teams capped at twelve, low interdependence between teams, highly motivated frontline professionals, highly customized service work, and a support structure of just 21 coaches for roughly 950 teams, coaches who hold no formal authority and no responsibility for team productivity. The authors add the warning every imitator ignores: copying the structures without the underlying philosophy tends to fail. GitHub and Valve, they note, added hierarchy as they grew; Buurtzorg's stability came from being designed around autonomous teams from day one, not converted later.
And the skeptics earned their seat at the table. Foss and Klein's work documents that many nominally flat firms simply grow informal hierarchies, unaccountable ones, in place of the formal hierarchy they abolished, and concludes there is no universally superior structure, only fits and misfits between task, environment, and people.
The mechanism view
Look at what actually functioned inside the successful cases and a pattern emerges: self-management is not the absence of management but its redistribution into explicit mechanisms. Semco's version, as Ricardo Semler has told it, is a clean illustration. Teams could veto a manager, and everyone rated their boss anonymously twice a year, with two consecutive low scores costing the role. The first year was uncomfortable; then something specific happened: managers who used to shout began to ask. The ones who could not adapt returned to technical work at the same salary, a legitimate exit, not a humiliation. Notice what this is: not the elimination of accountability but its inversion. Authority still existed; it simply had to be continuously earned from below rather than granted once from above.
That mechanism lens converts the vague question "should we be self-managing?" into auditable ones. Who allocates work when no one assigns it? Who resolves conflict when no one adjudicates? How does a bad performer, or a bad leader, actually lose their position? Organizations that can answer with named mechanisms (Buurtzorg's team-size caps and coaches, Semco's upward evaluations, Morning Star's peer contracts) are practicing something real. Organizations that answer with values language are running an informal hierarchy with extra steps.
The question behind the question
There is one more reason the leader's question has changed: the alternative is deteriorating. Gallup's 2025 global data shows manager engagement falling faster than anyone else's, from 27% to 22% in a year. The traditional middle layer is being squeezed by AI on one side and cost programs on the other, which means every organization is about to redistribute managerial work whether it wants to or not. Self-management research has become, in effect, the instruction manual for a transition many firms will make involuntarily.
So the honest answer to the CEO's real question, "would this work here?", is a counter-question: which conditions do you already have, and which mechanisms are you willing to build? If the work is customizable, the teams boundable, the information openable, and leadership genuinely willing to let authority be earned from below, the evidence says yes, at scale, for decades. If not, the evidence is equally clear about what you will get: Zappos, not Buurtzorg, and a well-documented retreat dressed up as evolution.
Sources
- Lee & Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing. Research in Organizational Behavior. Canonical SMO definition; hierarchy suits stable known tasks but inhibits cross-boundary problem-solving. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059
- Martela & Nandram (2025). Buurtzorg: scaling up with hundreds of self-managing teams but no middle managers. Journal of Organization Design. 900+ teams, ~14,700 staff (Jan 2021), 2 directors; boundary conditions: low interdependence, capped team size, coaches without authority. https://link.springer.com/article/10.1007/s41469-024-00184-y
- Foss & Klein (2022). Who's the boss? The persistence of entrepreneurial hierarchy in flat organizations. Journal of Organization Design. Flat firms often grow informal hierarchies; no universally superior structure. https://link.springer.com/article/10.1007/s41469-022-00132-8
- Groth, A. (2020). Zappos has quietly backed away from holacracy. Quartz. 2015 ultimatum led 18% to take buyouts; by 2017 holacracy was replaced by an internal market of teams with own P&Ls. https://qz.com/work/1776841/zappos-has-quietly-backed-away-from-holacracy
- Gallup (2025). State of the Global Workplace. Manager engagement fell from 27% to 22% in one year, the steepest decline of any group tracked. https://www.gallup.com/workplace/659610/state-global-workplace.aspx