Every society is haunted by institutions that should no longer exist. Boards that meet quarterly to ratify decisions made elsewhere. Ministries with budgets but no measurable mandate. Professional certifications that gatekeep skills the certified rarely use. These arrangements consume resources, shape careers, and structure expectations, yet a careful audit would struggle to identify what social function they actually perform.

The puzzle is not that institutions die slowly. It is that they often do not die at all. Sociologists working in the neo-institutionalist tradition, particularly Paul DiMaggio and Walter Powell, demonstrated decades ago that organizational forms persist not because they are efficient but because they are legitimate—recognized, expected, woven into the cognitive scaffolding of their fields. Function is a thin reed compared to the thick rope of taken-for-grantedness.

What follows is an analysis of three mechanisms that explain institutional persistence past the point of utility: the inertial drift of zombie institutions, the active defense mounted by constituencies whose interests have fused with the form, and the archaeological deposits that defunct arrangements leave behind to shape what comes next. Understanding these dynamics matters because institutional reform almost always underestimates them. Reformers assume that obsolescence invites replacement. It rarely does. Obsolete institutions are not waiting to be dismantled—they are quietly recruiting allies and rewriting the terrain on which their successors must build.

Ulrich Beck coined the term zombie institutions to describe forms that are dead but still walking—structures whose original logic has eroded but whose daily routines continue undisturbed. The nuclear family as a normative reference point, the nation-state as an autonomous economic actor, lifetime employment as a career model: each persists in institutional language and bureaucratic procedure long after its empirical foundations have crumbled.

The mechanism is not conspiracy but inertia. Institutions encode their purposes in calendars, job descriptions, software systems, and statutory frameworks. Each of these artifacts has its own maintenance schedule. The annual budget cycle continues; the reporting templates get filled out; the committees convene because the bylaws require quorum. No one needs to actively defend the institution—they simply need to not actively dismantle it. And active dismantling is costly, controversial, and rarely anyone's job.

What looks from outside like institutional life is often what James C. Scott might call legibility maintenance: the institution survives because it remains the unit through which the state, the market, or the profession reads social reality. Even when its substantive work has migrated elsewhere, its categories continue to organize data collection, resource allocation, and credentialing. The form becomes the infrastructure of measurement, which makes its removal a measurement crisis.

Consider the persistence of certain academic departments whose original disciplinary boundaries no longer track how research is actually conducted. The intellectual work has reorganized into interdisciplinary centers and informal networks, but the department remains because it is the unit of hiring, tenure, and budgeting. Eliminating it would require redesigning systems that touch every part of the university.

Zombie institutions thus reveal a counterintuitive principle: institutional survival does not require ongoing justification. It requires only the continued absence of a coordinated effort to end it. Since coordination is expensive and the institution's quiet existence imposes diffuse costs on no specific party, the equilibrium tilts strongly toward persistence.

Takeaway

Institutions do not need to be useful to survive—they need only to escape being actively killed. Persistence is the default state of any arrangement that has been encoded into infrastructure.

Every institution, in operating, produces a constituency. Salaries are paid, contracts are awarded, reputations are built, identities are formed. Over time, this constituency develops interests that are formally independent of the institution's stated mission. They are invested in the form, not the function. When the function evaporates, the constituency remains—and mobilizes.

This is the dynamic that Mancur Olson described in his analysis of distributional coalitions: concentrated beneficiaries face strong incentives to defend an arrangement whose costs are diffused across the broader public. A licensing board protecting an obsolete profession, a subsidy program serving a vanished economic sector, a treaty organization without a coherent adversary—each survives because the people inside have nowhere else to deploy their specific human capital, and their loss would be visible while the public's gain would be invisible.

What deepens this dynamic is that constituencies do not merely defend institutions—they reconstruct the narrative of what those institutions are for. Threatened with abolition, they generate new rationales: educational missions, security functions, cultural preservation, equity mandates. The institution becomes a kind of empty vessel into which successive justifications are poured. The original purpose is no longer the binding commitment; the constituency is.

This narrative plasticity explains why purely functional critiques rarely succeed. Demonstrating that an institution fails at task X is met with the response that its real value lies in task Y, and when Y is debunked, task Z is produced. The institution behaves, in Lakatosian terms, like a research program with an infinitely flexible protective belt. The hard core being defended is not any specific function but the existence of the institution itself.

Recognizing this pattern reframes reform strategy. Successful dismantlement rarely comes from proving uselessness—it comes from either co-opting the constituency through transition arrangements or finding a moment when the constituency itself is weakened, distracted, or divided. Function is the language of reform; constituency is its actual battlefield.

Takeaway

An institution's beneficiaries are not its servants—they are its true substrate. Once a constituency forms, the institution's purpose becomes negotiable, but its existence does not.

Even when institutions do finally collapse, they rarely disappear. They leave deposits—legal categories, professional identities, physical infrastructures, mental habits—that constrain whatever rises in their place. Successor arrangements must build on these sediments, and the sediments do not weather evenly. Some layers prove remarkably resistant, channeling new institutions into the grooves of the old.

Consider how feudal land tenure shaped subsequent property law centuries after feudalism's political collapse. Or how Cold War security architectures continue to structure intelligence cooperation, alliance commitments, and threat assessment frameworks decades after the strategic context that produced them has vanished. The institutions died; their categories survived and now organize problems they were never designed to address.

This is what we might call institutional archaeology: the study of how dead institutions act on the living through the artifacts they leave behind. The artifacts include statutory law, bureaucratic taxonomies, professional credentials, and—perhaps most powerfully—the implicit theories of how social problems should be sorted that practitioners absorb during training. Each of these constrains the imagination of reformers, because they define what counts as a recognizable institutional move.

Path dependence theorists, from Paul Pierson onward, have shown that early institutional choices generate increasing returns: each subsequent choice is cheaper if it conforms to the established pattern and more expensive if it departs. By the time the original justification fades, the surrounding ecosystem has adapted to the institution's existence, and uprooting it requires reconfiguring everything that grew around it. The cost of departure compounds with time.

The implication for institutional design is sobering. Founders are not merely solving present problems; they are laying down archaeological strata that will shape problem-solving for generations. The most consequential decisions are often invisible at the moment of choice—a category here, a jurisdictional line there—because their consequences compound only through their later use as the unquestioned starting point of subsequent design.

Takeaway

Dead institutions govern the living through the categories they bequeath. What we treat as the natural shape of a problem is often the fossilized solution to a forgotten one.

Institutional persistence is not a failure of rationality. It is the predictable output of how institutions actually work—as cognitive scaffolding, as constituency-generators, and as depositors of categorical sediment. The question is not why obsolete forms survive, but why we ever expected them to do otherwise.

This reframing carries practical weight. Reformers who treat obsolescence as a sufficient condition for change will lose, repeatedly, to institutions whose constituencies have time, focus, and embedded advantage. Effective institutional change requires identifying the specific mechanism of persistence at work—inertial drift, constituency defense, or archaeological constraint—and designing intervention to match.

More fundamentally, recognizing these dynamics should make us humble about the institutions we build today. Whatever we create will outlive its purpose, develop defenders we cannot foresee, and leave deposits that shape arrangements we cannot imagine. Designing for institutional mortality—building in sunset clauses, succession mechanisms, and graceful disassembly—may be the most underappreciated craft in institutional design.