The flat organization has become one of management's most seductive myths. Promising agility, empowerment, and freedom from bureaucratic drag, it captures the imagination of leaders weary of organizational sclerosis. Companies from Valve to Zappos to Medium have publicly embraced radical flatness, and management literature has elevated the rejection of hierarchy to something approaching moral virtue.

Yet the empirical record tells a more sobering story. Many celebrated flat organizations have quietly reintroduced layers, pivoted to hybrid structures, or watched their flatness curdle into political dysfunction. The pattern is so consistent it demands explanation: why do organizations that abolish hierarchy so often find themselves rebuilding it, sometimes in worse forms than they dismantled?

The answer requires us to reconsider hierarchy not as a regrettable artifact of industrial-era thinking, but as a coordination technology that solves specific organizational problems. When designed well, authority structures clarify decision rights, accelerate execution, and create developmental pathways. When abolished without replacement, the functions hierarchy served don't disappear—they migrate underground, where they operate without accountability or design intent. The serious question for organizational designers is not whether to have hierarchy, but how to architect authority structures that capture coordination benefits while minimizing the pathologies that give hierarchy its bad name.

Hierarchy persists across every form of complex human organization—military, religious, corporate, governmental—because it solves coordination problems that scale-free networks cannot. Understanding why requires examining what hierarchy actually does, beyond the caricature of command-and-control. At its core, hierarchical structure performs three irreducible functions: it allocates decision rights, it routes information, and it scaffolds human development.

The decision-rights function addresses what economists call the problem of residual control. When circumstances arise that contracts and procedures cannot anticipate—which is constantly, in any dynamic environment—someone must have the authority to choose. Hierarchy specifies in advance who decides what, eliminating the costly negotiations and political contests that would otherwise consume organizational energy. A flat structure does not eliminate the need for these decisions; it merely obscures who is empowered to make them.

The information-routing function is equally consequential. In any organization beyond Dunbar's number, no individual can maintain direct relationships with all colleagues, much less synthesize information across the entire enterprise. Hierarchy creates legitimate channels through which information aggregates, gets filtered, and reaches decision-makers in usable form. Without designed channels, information either drowns leaders in undifferentiated noise or fails to reach them at all.

The developmental function is perhaps the most underappreciated. Hierarchy creates a ladder of progressive responsibility through which people develop judgment, build capability, and demonstrate readiness for greater scope. The supervisor-subordinate relationship, properly executed, is one of the most powerful developmental contexts ever devised. Flat structures often eliminate these relationships without replacing the developmental architecture they provided.

Recognizing these functions reframes the design challenge. The question shifts from whether to have hierarchy to how to fulfill these functions effectively. Some organizations may distribute decision rights more widely, route information through different mechanisms, or develop people through alternative structures—but they cannot abandon the functions themselves without consequence.

Takeaway

Hierarchy is not primarily about power; it is a coordination technology that allocates decision rights, routes information, and develops people. Eliminating the structure does not eliminate the need for these functions.

When formal hierarchy is dismantled, informal hierarchy fills the vacuum—and the informal version is typically less accountable, less legitimate, and less responsive to organizational needs than the structure it replaced. This dynamic was first systematically documented by Jo Freeman in her 1972 essay The Tyranny of Structurelessness, written about feminist collectives but applicable to any organization that mistakes the absence of formal structure for the absence of structure itself.

In ostensibly flat organizations, power accrues to those with proximity to founders, charisma, longevity, technical reputation, or social capital. These informal hierarchies operate without job descriptions, performance reviews, or appeal mechanisms. A formal manager can be evaluated, coached, and replaced; an informal power broker operates in a zone of organizational ambiguity that is remarkably difficult to challenge. Employees often report that flat organizations feel more political, not less, precisely because the rules of advancement and influence become opaque.

Holacracy and similar self-management frameworks attempt to solve this by codifying decision rights through elaborate governance protocols. The honest acknowledgment embedded in these systems is revealing: even committed flatness advocates concede that some formal structure for authority is necessary. What they offer is not the elimination of hierarchy but its reorganization—often into systems of considerable complexity that participants must learn and continuously navigate.

The pathology of hidden hierarchy extends to accountability. In a flat structure, when something goes wrong, blame diffuses. When something goes right, credit concentrates around those with the most informal capital. Both outcomes corrode the relationship between performance and recognition that healthy organizations depend upon. The visibility that formal hierarchy provides—however imperfect—at least creates targets for legitimate critique and reform.

The deeper issue is that humans are status-seeking creatures, and any group of humans will rapidly develop status hierarchies whether designed or not. The choice facing organizational designers is not between hierarchy and its absence, but between hierarchies that are explicit, accountable, and purposeful, and hierarchies that are implicit, unaccountable, and emergent.

Takeaway

Structurelessness is a myth. The real choice is between designed authority that can be examined and reformed, and emergent authority that operates in the shadows without accountability.

The serious work for organizational designers is constructing authority structures that capture hierarchy's coordination benefits while minimizing its well-documented pathologies: rigidity, information distortion, deference cultures, and the suppression of dissent. This requires moving beyond the binary of flat versus hierarchical to a more nuanced design vocabulary that distinguishes between different dimensions of authority.

Elliott Jaques' work on requisite organization offers one productive starting point. His core insight is that organizational layers should correspond to differences in cognitive complexity and time horizon of decisions, not to status or compensation. When layers are calibrated to genuine differences in the work performed, hierarchy feels legitimate; when layers exist merely to create career steps or pad compensation structures, hierarchy generates the cynicism that fuels flatness movements.

A second design principle involves separating the dimensions of authority that traditional hierarchy bundles together. Authority over what work gets done, how it gets executed, who performs it, and how performance is evaluated need not concentrate in a single role. Modern organizational designs increasingly distribute these dimensions—product managers may own the what, technical leads the how, capability managers the who, and dedicated coaches the development. This decomposition retains formal accountability while preventing the concentration of power that breeds dysfunction.

A third principle is designing for legitimate dissent. Healthy hierarchies include explicit mechanisms—skip-level meetings, ombuds functions, dissent protocols, anonymous channels—that prevent authority from suppressing the information leaders most need to hear. Bad news travels poorly up steep hierarchies; design must counteract this gravitational pull. The Toyota production system's famous andon cord, empowering any worker to halt production, exemplifies how hierarchy can preserve front-line authority over critical decisions even within a layered structure.

Finally, optimal hierarchy design treats the structure as a living system requiring continuous calibration. Spans of control, layer counts, and decision rights should be revisited as the organization's strategy, scale, and environment evolve. The goal is not the perfect static design but the capacity to redesign as conditions change—what Edgar Schein called the adaptive culture embedded in the organization's deeper architecture.

Takeaway

Good hierarchy is calibrated to the actual complexity of work, distributes authority across dimensions rather than concentrating it, and includes designed mechanisms for legitimate dissent.

The romance of flatness reflects a legitimate frustration with hierarchy's pathologies—the petty tyrannies, the bureaucratic friction, the suppression of talent below the executive layer. But the response to bad hierarchy is not no hierarchy; it is better hierarchy. Organizations that pretend to abolish authority structures rarely succeed in doing so, and the structures that emerge in the absence of design are typically worse than those they replaced.

The mature position recognizes that authority is a design variable, not a moral failing. The functions hierarchy performs—decision rights, information routing, human development—must be performed by some structure. The question is whether that structure will be intentional or accidental, accountable or hidden, calibrated to the work or arbitrary.

For senior leaders, this reframes a generation of organizational debate. Stop asking whether your organization should be flat. Start asking whether your authority structures are legitimate, whether your layers correspond to genuine complexity differences, and whether dissent can travel upward without distortion. Those questions, taken seriously, lead to organizations that are neither rigidly hierarchical nor naively flat—but something more difficult and more durable: deliberately designed.