Three product directors left the same technology company over an 18-month period. Their exit interviews cited familiar phrases: limited growth opportunities and unclear career paths. Standard turnover narrative.

When HR dug deeper, the pattern was different. All three had reported to managers who were running 47+ meetings per week, approving expense reports at 10 PM, and managing 16-plus direct reports. The managers weren't disengaged by disposition. They had no structural capacity left to develop the people beneath them.

That company isn't unusual. It's representative.

In 2024, global manager engagement fell from 30% to 27% — a three-point decline that, according to Gallup's chief workplace scientist Jim Harter, represents a direct threat not just to organizational performance but to GDP growth (Gallup, State of the Global Workplace: 2025). Individual contributor engagement, by contrast, held flat at 18%. The divergence between those two numbers is the signal. The system isn't failing workers uniformly. It's failing the specific layer responsible for translating strategy into execution, developing talent, and sustaining culture. When that layer erodes, everything it was holding up erodes with it.

The question this article addresses is not whether manager burnout is real — that case is closed. The question is why well-resourced organizations continue to treat it as a people problem when the evidence consistently points to a design problem.

The manager role was never simple, but for most of the twentieth century, it was at least coherent. A manager hired, directed, evaluated, and advanced a defined team. Decision rights were clear. Administrative burden was modest. The span of responsibility, while significant, was relatively bounded.

That coherence began to unravel in the 1990s as organizations delayered in pursuit of efficiency. The logic was straightforward: remove management overhead, push accountability lower, accelerate decisions. The results were mixed from the start. Research on organizational flattening found that firms with fewer middle-management layers often exhibited greater centralization at the top — the opposite of the intended effect — as senior leaders absorbed decisions previously distributed throughout the hierarchy (Boston University Questrom School, Flattening the Firm, February 2026).

The post-pandemic period accelerated the compression. Between spring 2022 and the end of 2024, U.S. employers advertised 42% fewer middle management positions (Deloitte, Global Human Capital Trends 2025). Gartner projects that by 2026, 20% of organizations will use AI to further flatten their structures, eliminating more than half of current middle-management positions. Those who remained absorbed the work of those who were cut, while simultaneously being handed new mandates: manage hybrid teams, lead AI adoption, sustain psychological safety, drive transformation initiatives, and deliver results in tighter economic conditions.

McKinsey's Power to the Middle research, updated in 2024, documented the result: middle managers now spend close to 50% of their time on non-managerial tasks — administrative approvals, coordination, reporting, and status meetings that could be asynchronous (McKinsey). Deloitte's 2025 Global Human Capital Trends survey found that 40% of managers' time is consumed by firefighting and administrative work rather than by coaching and development, which distinguishes the role (Deloitte). Forty percent of managers report that their mental health declined when they took on the role. Harvard Business Impact research found that 85% of mid-level leaders report experiencing burnout weekly (Harvard Business Impact, December 2025).

These are not outliers from particularly dysfunctional organizations. They are population-level findings across industries and geographies. The role, as currently designed, is structurally broken.

The 70% figure from Gallup — that 70% of team engagement variance is attributable to the manager — has appeared in organizational research for years, but its implications become newly urgent when read alongside the 2025 engagement data.

When managers disengage, their teams disengage at a predictable and measurable rate. The 2025 Gallup data show global employee engagement falling to 21%, the second decline in 12 years, with the first being the COVID-19 year of 2020. The magnitude of the 2024 drop was equivalent to the pandemic decline, despite no comparable external economic shock (Gallup, U.S. Employee Engagement Sinks to 10-Year Low, January 2025). What changed was the internal condition of the management layer.

The gender and age dimensions of the data further sharpen the diagnosis. Female manager engagement fell seven points year-over-year. Managers under 35 fell five points. These are not marginal numbers. They indicate that the groups most likely to represent the next generation of senior leadership — and the groups carrying the greatest cultural expectations for relational and developmental management — are disengaging at the fastest rate. Organizations are burning out the people they can least afford to lose.

The attrition logic that follows is straightforward but often underweighted in executive decision-making. When a manager leaves, their direct reports lose a champion, a developer, and a sensemaker. Some portion of those reports then leave. The institutional knowledge embedded in that manager-team relationship — who is ready to promote, what is broken in the process, where the strategic risk actually lives — walks out with them. External backfill at senior levels costs more and carries higher failure rates than internal development. The compounding cost of one burned-out manager's exit is almost never captured in standard attrition metrics.

The original article's most actionable observation — that the most fixable problem is administrative burden — holds up under closer scrutiny, but the numbers are worth examining in more detail.

McKinsey's breakdown identifies the categories: approximately 18% of a manager's week consumed by approvals, expense workflows, and compliance documentation; 14% on status meetings that carry no decision value; 10% on rework and context-switching caused by unclear decision rights. Together, those categories represent roughly 42% of a 40-hour week spent on work that does not require managerial judgment (McKinsey, Power to the Middle).

Deloitte's 2025 data reinforces the scale. Forty-one percent of all employee time — not just managers — is spent on work that doesn't contribute to organizational value creation, according to Deloitte's 2025 Global Human Capital Trends survey (Deloitte, Reclaiming Organizational Capacity). For managers, whose non-value work is disproportionately concentrated in bottleneck roles — the single approver, the institutional memory repository, the escalation endpoint — the burden is not evenly distributed across their week. It is concentrated in moments when it becomes a constraint on everyone they manage.

Cut 10 hours of administrative work from a manager's week, and the math shifts materially. That is enough time for substantive one-on-ones with eight to ten direct reports. Enough time to notice who is struggling before it becomes an HR issue. Enough time to think about whether the team's current priorities align with what the strategy actually requires next quarter. The case for investing in admin support, automating approval workflows, and redesigning span of control is not a wellbeing argument. It is a capacity argument. Organizations are spending significant resources managing the consequences of decisions — attrition, disengagement, strategic drift — that a simpler structural intervention could prevent.

The relationship between span of control and manager effectiveness has been studied for decades, but it is consistently treated as an HR efficiency metric rather than a strategic design choice.

Research on frontline hospital managers found that spans of 15 or more direct reports were directly predictive of work-life balance problems and intention to quit (NCBI/PubMed). Organizations that reduced the span of control to eight to ten direct reports saw measurable declines in manager stress and corresponding improvements in team engagement. The math is not complicated: a 40-hour week divided across 15 people allocates approximately 2.6 hours per person per week — insufficient for any meaningful development. Halve the span, and the manager suddenly has the structural capacity to do the job the organization claims to need them to do.

Span of control decisions are also consequential for strategy execution in a way that organizational design conversations rarely acknowledge. Deloitte's research found that organizations with strong management capability report up to 15% higher financial performance than those with weaker management (Deloitte). The mechanism is not mysterious. Managers with adequate capacity to coach and develop their people produce teams that perform better, adapt faster, and surface problems earlier. The return on investment from right-sizing the span of control is recoverable against standard retention and performance metrics — but it requires treating the management layer as infrastructure, not overhead.

Across all of this, a secondary problem amplifies it: most managers have never been adequately prepared for the role. Less than half of the world's managers — 44% — report having received any management training (Gallup, State of the Global Workplace: 2025). Deloitte's 2025 survey found that 36% of managers felt unprepared for the people-management dimensions of the role at the time they took it.

The pipeline logic matters here. Organizations promote their best individual contributors into management, then give them the title and accountability, with minimal scaffolding. Gallup's data offers a sharp illustration of what basic training does: providing manager training improved manager thriving rates from 28% to 34%, with further improvement to 50% when the training was paired with active encouragement and ongoing development. An untrained manager running 15 direct reports in an organization that fills their calendar with status meetings is set up to fail. The resulting burnout is then misattributed to the manager's personal resilience.

The strongest counterargument to the "manager burnout as design failure" thesis comes from the organizational efficiency school — specifically, from those advocating for flatter structures and reduced middle management as a competitive advantage. The logic is coherent: fewer layers mean faster decisions, less bureaucracy, more direct connection between senior leadership and frontline execution. Meta, FedEx, and others have made high-profile bets on this model. A Boston Consulting Group analysis predicts that eliminating traditional middle-management layers can boost agility and efficiency, particularly as AI assumes supervisory tasks.

This argument deserves to be taken seriously, because it is not wrong about the problem. Middle management, as currently configured, often does create bureaucracy. Managers managing managers create signal degradation. Administrative layers slow down decisions. The efficiency critique of the current model is legitimate.

Where it fails is in conflating the role with its current dysfunctional design. You can cut the title without eliminating the function, but organizations that try this discover that humans will still lead and be led regardless of org chart labels. "Shadow hierarchies" emerge. Unofficial decision-makers appear. The sensemaking and translation work that middle managers do doesn't disappear when the role is eliminated; it migrates to whoever fills the vacuum, often without accountability or clarity (Boston University Questrom). Deloitte's 2025 research found that Gartner itself — which projects significant AI-driven management cuts — acknowledges that the core capabilities managers provide (coaching, development, psychological safety, strategic translation) will always be necessary. The question is not whether those capabilities are needed. It is whether the organization is willing to create the conditions for their exercise.

Flattening a broken role does not fix it. It distributes the breakage more widely.

The management audit has to come before the management initiative. Before investing in a new leadership development program or launching a manager effectiveness survey, executive teams need an honest accounting of how managers currently spend their time. Have a cross-functional sample of 15–20 managers track their time for two weeks across four categories: direct report development, strategic execution, administrative tasks, and meetings. If administrative work and low-value meetings exceed 40% of the week, the variable to fix is not managerial capability. It is a system design. Development investment layered on top of an overloaded role produces marginal gains at best.

Administrative work below the judgment threshold should be systematically removed. Identify which approval workflows require actual managerial judgment and which exist because no one redesigned the process. Budget approvals below a defined threshold can be automatic. Status updates can be dashboard-based rather than meeting-dependent. Compliance documentation can be pre-populated. Each of these changes is small on its own. Cumulatively, they can return 8–12 hours per week to the manager's calendar — enough to materially change what they can do for the people they lead.

Span of control is a strategic investment decision, not a headcount efficiency target. The pressure to increase manager-to-employee ratios is almost always a cost-reduction impulse. The cost-benefit calculus looks different when attrition cost, ramp time for external hires, and performance degradation from understaffed teams are included. Organizations that treat span of control as a lever for manager effectiveness — targeting eight to ten for first-line managers, four to six for senior managers — tend to recover the investment through retention and performance within 12–18 months. The math is not difficult. The organizational will to run it is.

Female and young managers require targeted attention, not generalized programs. The seven-point drop in female manager engagement and five-point drop among managers under 35 are not statistical noise — they indicate specific, compounding pressures that generic manager support programs will not address. Female managers absorb disproportionate emotional labor expectations on top of the same impossible administrative load. Younger managers are navigating role complexity and span of control without institutional experience to buffer them. Both groups need structural relief and explicit development investment, not the same webinar everyone else gets.

Training is necessary but insufficient on its own. Gallup's finding that manager thriving improves from 28% to 50% when training is paired with ongoing encouragement is worth taking seriously — but the sequence matters. Training a manager who is drowning in administrative work is unlikely to move the needle. The structural preconditions — cleared calendar space, a reasonable span, and clear decision rights — must exist before development investment can compound. Organizations that provide training before the structural work is done are buying a palliative, not a solution.

Manager burnout is not a failure of resilience. It is the predictable output of a role that has been systematically overloaded, under-resourced, and treated as a cost to minimize rather than a capacity to invest in. The organizations that have figured this out are not running better wellness programs or conducting more frequent check-ins with their managers. They are redesigning the role itself — clearing administrative burden, right-sizing spans of control, and restoring the structural conditions under which a manager can actually do the job they were hired to do.

The data is unambiguous: companies with strong management outperform those without it by up to 15%. Seventy percent of the variance in team engagement runs through the manager. And right now, global manager engagement is at its lowest point in years and continues to fall. The question for every executive reading this is not whether the management layer matters. The question is whether you are running the organization in a way that allows it to function.