The middle manager has become the villain of the current restructuring story. The numbers behind the story are real. The conclusion most people draw from them is not.
The role really is in trouble
Start with the evidence that something is wrong. Gallup's 2025 workplace study, drawn from 263,810 respondents, found global employee engagement fell to 20 percent, and that manager engagement dropped 9 points since 2022, with the sharpest annual fall coming between 2024 and 2025, from 27 to 22 percent (Gallup, 2025). The people who are supposed to carry the culture are disengaging faster than the people they manage. Deloitte's 2025 Human Capital Trends, based on nearly 10,000 leaders across 93 countries, shows why. Managers spend close to 40 percent of their time firefighting and on administrative work, and only about 13 percent developing their teams, while 36 percent feel unprepared for the people-management part of the role (Deloitte, 2025). That is not a function failing. That is a job stuffed with administrative debris until the part that justifies it, developing people, gets squeezed to a sliver.
The wider organizational picture is just as unflattering. McKinsey's State of Organizations survey of more than 2,500 leaders found two-thirds describing their organizations as overly complex and inefficient, and only half feeling prepared to react to external shocks (McKinsey, 2023). Layers accumulate faster than value.
Why the extinction narrative overreaches
Into that vacuum comes the sharper claim: that artificial intelligence will simply delete the layer. An essay in Fortune, citing a Gartner forecast, argues that by the end of 2026 around 20 percent of surveyed organizations will use AI to flatten their structure and eliminate more than half of current middle-management positions (Fortune, 2025). That is worth taking seriously and worth labeling honestly. It is a prediction and an opinion, not a measured outcome. Forecasts about management have a poor track record, and the same AI that can strip a manager's informational advantage does not automatically supply the coordination, judgment, and development that the good version of the role provided.
The extinction framing also misreads what actually replaced hierarchy in the cases that worked. When Haier dismantled its middle layers, it did not leave a void. Gary Hamel and Michele Zanini (2018) documented its reorganization into roughly 4,000 self-managing microenterprises, about 250 facing users and the rest supplying components and services internally, coordinated by internal contracts and market transactions rather than by managers. Over the prior decade Haier's gross profit and revenue grew about 23 and 18 percent a year. Flattening there was a design, with new coordination mechanisms deliberately built to carry the load the managers used to bear. Academic analysis treats Haier's Rendanheyi as a genuine management innovation rather than a technology story, documenting how internal contracts and market transactions between microenterprises replaced the coordination that hierarchy used to provide (Frynas, Mol and Mellahi, 2018). The layer went away because something specific was engineered to do its work.
Flattening is a design, or it is just a layoff
The contrast that matters is between flattening that engineers a replacement and flattening that removes the layer and hopes. Buurtzorg runs more than 900 self-managing nurse teams and about 14,700 staff with no middle managers, but it substitutes structure for supervision: small teams, a digital platform, and roughly 21 coaches who develop rather than command (Martela and Nandram, 2024). The managers are gone and the coordination is present. Delete the layer without building that substitute and coordination does not vanish gracefully. It reappears as the firefighting and administrative overload that Deloitte already measured, now with fewer people to absorb it.
This is the honest read of the great flattening. The manager's job as currently constituted is broken, drowned in administrative work and starved of the developmental core. That is a strong argument for redesigning the role or distributing it into a real structure. It is a weak argument for assuming the function can be subtracted for free.
The practical read
Before cutting a layer, answer one question: what carries the coordination, judgment, and development that the layer was doing, however badly? If the answer is a designed system, microenterprises with internal contracts, self-managing teams with a platform and coaches, then flattening can raise both performance and engagement (Hamel and Zanini, 2018; Martela and Nandram, 2024). If the answer is a slide claiming AI will handle it, you are not flattening the organization. You are running a layoff and betting that the work those managers absorbed will quietly disappear. Gallup and Deloitte already show where that work goes when no one is designed to catch it.
Sources
- Gallup (2025). State of the Global Workplace 2025. Global engagement fell to 20 percent and manager engagement dropped 9 points since 2022, evidence the manager role is in crisis. https://www.gallup.com/workplace/659610/state-global-workplace.aspx
- Deloitte (2025). 2025 Global Human Capital Trends. Managers spend about 40 percent of time firefighting and only 13 percent developing people, with many feeling unprepared. https://www.deloitte.com/us/en/insights/topics/talent/human-capital-trends/2025.html
- Fortune (2025). Surviving the great flattening. An opinion essay citing a Gartner forecast that AI will eliminate over half of middle-management roles in some firms by end of 2026. https://fortune.com/2025/09/19/surviving-great-flattening-coming-extinction-of-middle-manager-layoffs/
- McKinsey (2023). The State of Organizations 2023. Two-thirds of 2,500-plus leaders call their organizations overly complex and inefficient. https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/the-state-of-organizations-2023
- Hamel and Zanini (2018). The End of Bureaucracy. Haier replaced middle layers with about 4,000 self-managing microenterprises coordinated by internal contracts, alongside strong growth. https://hbr.org/2018/11/the-end-of-bureaucracy
- Martela and Nandram (2024). Buurtzorg: scaling up an organization with hundreds of self-managing teams but no middle managers. Removed managers but substituted small teams, a platform, and developmental coaches. https://link.springer.com/article/10.1007/s41469-024-00184-y
- Frynas, Mol and Mellahi (2018). Management Innovation Made in China: Haier's Rendanheyi. Treats Haier's flattening as a genuine management innovation built on internal contracts between microenterprises. https://journals.sagepub.com/doi/10.1177/0008125618790244