Return-to-office mandates, cost cutting, and generative AI have all become proxies for a deeper argument: what exactly is a manager for. The data show the middle is under stress and, in many places, beginning to thin. Globally, manager engagement fell nine percentage points since 2022, with the sharpest annual drop from 27 percent to 22 percent between 2024 and 2025 (Gallup, 2025). In the United States, employers were advertising 42 percent fewer middle-management roles at the end of 2024 than in spring 2022 (Deloitte, 2025). Analysts expect the structure itself to change. Gartner forecasts that by 2026 one in five organizations it tracks will use AI to flatten their hierarchies, and a Fortune analysis reports Gartner’s expectation that those firms will eliminate over half their current middle-management positions (Fortune, 2025).
Yet eliminating a layer does not eliminate the work that layer performed. The strategic question is which of those jobs should vanish, which should be automated, and which should be reimagined close to the frontline.
What is actually ending
Most managers spend their weeks on the wrong work. In a 2025 global survey of nearly 10,000 leaders, Deloitte found managers devote almost 40 percent of their time to firefighting and administration, and only 13 percent to developing their people. Over a third feel unprepared for the people side of the role (Deloitte, 2025). That is not a staffing problem alone. It is a design choice that crowds out coaching and judgment with forms and escalations.
The costs show up in morale and performance. Global employee engagement sank to 20 percent in 2025, the lowest since 2020, with Gallup estimating roughly 10 trillion dollars in lost productivity, about 9 percent of global GDP (Gallup, 2025). The drop among managers is the most telling signal for our theme. If the people responsible for multiplication of effort are themselves depleted, the middle stops adding value and starts blocking it.
Leaders know their structures are not helping. In its State of Organizations research, McKinsey reported that only half of more than 2,500 leaders say their companies are prepared to anticipate and react to shocks, and two thirds see their organizations as overly complex and inefficient (McKinsey, 2023). Complexity creates demand for coordinators. AI and better process design remove that demand. The temptation is to infer that the manager is obsolete. That would repeat an old mistake.
The debate over “bossless” organizations has long been polarized. Celebrants talk as if hierarchy were a disease. Skeptics dismiss self-management as a naive experiment. A more useful question is where self-management works, and why. That is the balanced frame argued by Bernstein, Bunch, Canner, and Lee in their assessment of holacracy and related models, which in 2016 highlighted Zappos as the largest adopter of holacracy at the time and urged separating promise from hype (Bernstein, Bunch, Canner, and Lee, 2016).
Where the middle really disappears
There are places where the layer has been designed out, and at scale. Buurtzorg, the Dutch home-care organization, operates more than 900 self-managing teams with a maximum of 12 nurses each and roughly 14,700 employees as of January 2021. It has two directors, around 50 people in back office, and no middle managers or supervisors (Martela and Nandram, 2025). The scaffolding is light. About 21 internal coaches support roughly 950 teams, roughly one coach for more than 45 teams, and those coaches have no formal authority and no responsibility for team productivity (Martela and Nandram, 2025).
The financial picture is nuanced. A KPMG analysis around 2015 reported Buurtzorg’s overhead at 8 percent compared with a sector average of 25 percent in Dutch home care, yet the organization’s cost per hour was higher than peers, 54.47 euros versus 48.74 euros (Martela and Nandram, 2025). The lesson is not “self-management is cheaper.” It is “self-management shifts where the money goes.” More cost sits with professionals who do the work, and less with layers coordinating it.
Boundary conditions matter. Martela and Nandram identify that the model works best where interdependence between teams is low and frontline professionals are both highly motivated and competent (Martela and Nandram, 2025). Home care fits this profile because units can serve local neighborhoods with clear mandates and limited cross-team coupling. If your business depends on tight cross-functional synchronization, you will still need strong mechanisms for coordination. The question is whether that coordination is a job title in the middle or a set of transparent practices distributed across teams.
What replaces the layer in most firms
Flattening is not only a budget exercise. It is an information exercise. AI lowers the cost of status updates, project tracking, and answers to routine “who does what by when,” which historically anchored many manager roles. Gartner expects that one fifth of organizations it follows will use AI to flatten by 2026 (Deloitte, 2025). Fortune, summarizing Gartner’s forecast, argues the bigger shift is from knowledge possession to explainable execution, and that is where middle-management roles thin out fastest (Fortune, 2025). Treat that as a provocation, not gospel. The point, grounded in the surveys, is that much of what managers spend time on can be automated, and much of what they are not doing enough of, coaching and building capability, cannot.
If engagement is design, not fate, then the direction is clear. In best-practice organizations, 79 percent of managers are engaged versus a global average of 22 percent, which Gallup uses to argue that the gap reflects organizational choices rather than the business cycle (Gallup, 2025). The firms that win this transition will do three things: strip administration from the role, make performance visible at the team level, and invest in managers as multipliers who teach, not traffic cops who route.
Here is the practical read. Start by clarifying the non-negotiable work only managers can do in your context. If your current role description devotes most space to approvals, status reporting, and cross-unit escalations, assume AI will eat half of it within two years. Use that to shrink the layer where teams can truly self-manage, as Buurtzorg shows for low-interdependence work (Martela and Nandram, 2025). Where the layer remains, redesign it around outcomes people feel, not paperwork. Set expectations and measurement for coaching time and capability lift, because surveys suggest managers are not spending time there today (Deloitte, 2025) and the cost of getting this wrong is visible in the global engagement data (Gallup, 2025).
The end of the middle manager is not an inevitability. It is a consequence of structural choices. Some organizations will remove the layer where the work allows. Many more will keep a version of it that looks different: fewer people, better tools, clearer mandates, and measurable impact on teams. The test is simple. If your middle manager vanished tomorrow, would your customers or your frontline notice for the wrong reasons. If not, you have a layer to redesign or retire.
Sources
- Bernstein, Bunch, Canner, and Lee (2016). Beyond the Holacracy Hype. Argues for a balanced, evidence-based view of self-management and notes Zappos as the largest holacracy adopter at the time, grounding the debate about removing managerial layers. https://hbr.org/2016/07/beyond-the-holacracy-hype
- Martela and Nandram (2025). Buurtzorg: scaling up an organization with hundreds of self-managing teams but no middle managers. Documents a large-scale self-managed model with no middle managers, its light coaching scaffolding, boundary conditions, and nuanced cost picture. https://link.springer.com/article/10.1007/s41469-024-00184-y
- Gallup (2025). State of the Global Workplace 2025. Reports global engagement at 20% and a sharp drop in manager engagement, quantifying the cost and signaling dysfunction in the middle. https://www.gallup.com/workplace/659610/state-global-workplace.aspx
- Deloitte (2025). 2025 Global Human Capital Trends. Shows managers’ time is skewed toward admin and firefighting, that many feel unprepared, and that US postings for middle managers fell 42% since 2022; also cites Gartner’s forecast on AI-enabled flattening by 2026. https://www.deloitte.com/us/en/insights/topics/talent/human-capital-trends/2025.html
- Fortune (2025). Surviving the great flattening: the coming extinction of the middle manager. Reports Gartner’s forecast that by end-2026, 20% of surveyed organizations will use AI to flatten and eliminate over half of middle-management roles in those firms. https://fortune.com/2025/09/19/surviving-great-flattening-coming-extinction-of-middle-manager-layoffs/
- McKinsey (2023). The State of Organizations 2023. Finds only half of leaders feel prepared for shocks and two thirds see their organizations as overly complex and inefficient, supporting the case for simplification and role redesign. https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/the-state-of-organizations-2023