Disability benefit schemes in advanced welfare states share a disconcerting characteristic. Despite continuous reforms, growing budgets, and explicit commitments to equity and efficiency, they produce remarkably similar patterns of dysfunction. High initial refusal rates, massive appeal volumes, chronic processing delays, frequent rights disruptions, and widespread non-take-up appear with striking regularity.

These symptoms persist across vastly different institutional contexts regarding welfare regime type, generosity levels, degree of medicalization, and administrative traditions. This raises a fundamental analytical question: if these problems reproduce themselves despite contextual diversity and repeated reform attempts, are we facing mere implementation failures or dysfunctions structurally embedded in the very paradigm that governs these systems?

The problem is not the benefits themselves. It is how we manage access to them.

I analyzed five major disability benefit schemes chosen for the diversity of their institutional architectures: France's AAH, Belgium's ARR, the UK's PIP, US SSI, and Australia's DSP. Despite belonging to very different welfare state models, they share essential structural characteristics. All are non-contributory or weakly contributory financial benefits, conditioned on disability assessment, targeting long-term incapacity, administered through medical-administrative procedures, and subject to thresholds, criteria, controls, and periodic reassessments.

This functional proximity allows us to test a central hypothesis: if similar dysfunctions appear in such diverse institutional contexts, they stem less from national particularities than from a shared logic of conception.

The comparative analysis reveals a remarkably homogeneous dysfunction syndrome. Initial refusal rates are systematically high. In the UK, PIP has experienced refusal rates exceeding 50% during certain periods. In the US, the majority of SSI applications are rejected at first decision. In Belgium and Australia, over 60% of applications are similarly rejected at initial examination. This initial severity is not marginal; it constitutes a structural element of how these systems function.

These initial refusals are followed by a second equally systematic phenomenon: massive volumes of appeals and re-examinations, accompanied by high rates of favorable decisions on appeal. In the UK, tribunals rule in favor of applicants in the majority of contested cases. In the US, the applicant's journey frequently becomes an almost standardized sequence: refusal, appeal, hearing, acceptance. This combination of high initial refusals and frequent appeal successes reveals a structural failure of initial assessment. The system cannot, from the outset, correctly assess eligibility according to its own criteria.

Processing delays are chronically long in all cases. These do not constitute mere administrative inconveniences but major factors of precarity. For months, even years when appeals are included, people remain in extreme financial and social uncertainty. These delays themselves produce new administrative burdens: follow-up requests, repeated document submissions, complaints, procedural escalations. In other words, delay becomes an autonomous producer of administrative activity.

This regularity demands robust theoretical explanation. The root cause is a management paradigm founded on the logic of filtering and suspicion rather than access and support. These regimes operate according to an implicit mission that has supplanted their explicit one: managing access to rights through rationing rather than guaranteeing access to social protection. This inversion of purposes is neither explicitly recognized nor politically debated; it is inscribed in choices that appear technical but carry profound normative and political consequences.

The contribution of systems analyst John Seddon allows us to understand how this normative logic translates operationally into self-produced administrative overload. Seddon radically distinguishes real demand—the initial expression of a legitimate need—from failure demand, which he defines as activity generated by the system's inability to respond correctly to that demand at first contact.

In disability benefit schemes, failure demand takes multiple forms: applications reintroduced after refusal, administrative and judicial appeals, revision requests, repeated document transmissions, corrective exchanges, mediations, ping-pong between offices and litigation. This activity does not correspond to new needs but to repairing the effects produced by inadequate initial decisions, poorly designed procedures, or structurally inoperative criteria.

Seddon's empirical work shows that in comparable public service systems, between 40% and 80% of activity can consist of failure demand. This observation illuminates the central paradox of the regimes studied: administrations in chronic overload despite growing resources and relatively stable demand volumes. The overload does not primarily stem from social demand but from how the system treats—or fails to treat—that demand.

The filtering logic transforms the system into a machine for producing failure demand. By massively refusing initial applications, the system mechanically triggers repeated applications and appeals. By fragmenting procedures, it multiplies rupture points. By standardizing evaluation at a distance from concrete situations, it increases the probability of error. Each error feeds a correction cycle that consumes time, resources, and administrative energy without improving the response provided to real needs.

Political scientist Deborah Stone's work provides decisive analytical rupture. She shows that disability policies cannot be understood from their declared purposes alone. She introduces a fundamental distinction between the stated purpose of benefit schemes—guaranteeing income, responding to objectively observable needs, protecting vulnerable people, ensuring equitable treatment—and their real operative purpose: drawing boundaries between those who deserve help and those excluded from it, determining conditions of claim legitimacy, and setting the symbolic price of assistance.

Stone demonstrates that disability is not a stable category susceptible to objective measurement. It is intrinsically ambiguous across multiple dimensions: severity, duration, functional impact, interaction with environment, relation to work and social participation. This categorical instability explains the continuous proliferation of criteria, scales, standardized assessment tools, and periodic re-examinations. Administration attempts to determine what, by nature, resists determination.

This attempt produces empirically observable effects: presumption of refusal, arbitrary thresholds, decisional inconsistency, differentiated treatment of visible and invisible disabilities, heightened suspicion toward psychosocial or fluctuating disorders, and the obligation for people to "perform" their disability according to administratively receivable narratives.

The logic of suspicion allows the system to fulfill an implicit rationing function without explicitly assuming exclusion choices. It transforms a social right into a conditional favor, subject to repeated demonstrations of legitimacy. This logic constitutes the normative foundation upon which the dynamics of empirically observed dysfunction are built.

Economist Nicholas Barr's analysis completes this framework by highlighting the central role of transaction costs. Any redistribution policy generates costs related to defining eligibility, collecting information, control, execution, and dispute resolution. In disability benefit schemes, these costs are particularly high due to the intrinsic ambiguity of the categories mobilized.

Barr shows that heavily targeted schemes often appear more efficient because analysis is limited to benefit expenditures. However, when we integrate transaction costs—administrative, legal, human, and social—apparent efficiency collapses. Resources devoted to verification, re-examinations, and litigation can exceed savings realized through exclusions.

Failure demand constitutes a concrete expression of these transaction costs. Each appeal, each re-examination, each reintroduced file mobilizes resources that could have been devoted to resolving real needs. Delays and unjustified refusals moreover produce secondary costs: worsening health problems, recourse to emergency services, lasting social exclusion. These costs do not disappear; they are displaced toward other segments of the welfare state.

The economics of transaction costs thus reveals a central illusion of the current paradigm: the belief that administrative filtering would allow expenditure control. In reality, it redistributes and amplifies them while degrading service quality and system legitimacy.

The paradigm shift required rests on a decisive conceptual displacement: income security must no longer depend on administrative recognition of disability. As long as access to subsistence income is conditioned on medical-administrative evaluation, the system is mechanically led to produce suspicion, filtering, instability, and failure demand.

Current regimes confuse two radically distinct functions: guaranteeing minimum income allowing dignified living, and compensating functional limitations and disability-specific costs. This confusion transforms disability evaluation into an existential issue. Recognition or refusal determines not only access to complementary aids but the very possibility of meeting one's needs. Under these conditions, control logic becomes structurally inevitable, and administrative error becomes socially destructive.

Dissociating these functions allows us to defuse this dynamic. Income security can then be conceived as an unconditional foundation, independent of disability status. Disability evaluation then ceases to be a sorting mechanism for access to subsistence and can refocus on its proper purpose: identifying specific needs and organizing adapted responses.

The alternative paradigm can be formalized as a three-level architecture, each responding to a specific purpose and obeying its own logic.

The first level consists of unconditional income guarantee, ensuring everyone a material foundation of security. Exact modalities may vary—basic income, negative income tax, minimum income guarantee—but functional characteristics are essential: universality, stability, and absence of conditionality based on disability. This level fulfills an existential function: protecting against poverty and material insecurity, independent of individual trajectories, work capacities, or administrative statuses.

Contrary to received wisdom, this universality does not necessarily constitute budgetary drift. In reality, a massive portion of current budgets is already absorbed by a simple subsistence function. By integrating hidden costs—those of filtering bureaucracy, legal appeals, incessant re-examinations, and social consequences of non-take-up—the net cost of such a foundation would prove comparable, even lower, than current schemes. By freeing administration from its "resource guardian" function for this basic level, we eliminate the failure demand identified by Seddon.

The second level is compensation in the strict sense: compensating functional limitations and disability-related additional costs to enable effective social participation. At this level, evaluation is not only legitimate but necessary. Needs vary considerably according to situations: technical aids, personal assistance, environmental modifications, communication aids, mobility, cognitive or psychosocial support. Targeting is pertinent here because it aims not to sort those entitled to income but to adjust responses to specific needs.

However, the nature of evaluation changes radically. It becomes functional, contextual, and problem-solving oriented, not categorical or punitive. The central question is no longer "Is this person disabled enough?" but "What obstacles does this person encounter and what resources can remove them?" This displacement reduces administrative conflict. Decisions no longer determine access to subsistence but the nature and intensity of supports. Errors, while still possible, no longer have existential effects.

The third level concerns support services. This involves moving beyond the benefit logic to enter a logic of responding to situations. Services—whether concerning employment, housing, health, training, or social participation—must be organized around resolving problems experienced by people, not around abstract eligibility criteria.

This alternative architecture is not purely theoretical. It finds partial realization in Nordic systems, which, while not exempt from tensions, present significant structural differences compared to previously analyzed regimes. In these countries, access to basic income is less heavily conditioned on exhaustive disability recognition. Compensation and services occupy a more central place, with substantial investment in personal assistance, accessibility, and support. Evaluations are more oriented toward functional needs than toward proof of incapacity.

Observable effects are consistent with the systemic analysis developed here: fewer appeals, fewer rights disruptions, greater pathway stability, and proportionally lower administrative burden devoted to managing failure demand. These systems are not perfect, but they demonstrate that other institutional equilibria are possible and sustainable.

The question is no longer how to improve existing disability benefit regimes but whether it is still reasonable to persist in a conceptual framework that systematically produces the effects it claims to combat. To this question, the analysis developed here provides a clear answer: transformation is not merely desirable; it appears necessary if we want to preserve the credibility and viability of the welfare state itself.

The stakes extend far beyond administrative efficiency alone. The current paradigm imposes considerable human costs: administrative humiliation, prolonged precarity, health deterioration, and erosion of trust in collective solidarity. By transforming solidarity into a conditional favor requiring repeated proofs of legitimacy, these systems undermine the very legitimacy of the welfare state they are supposed to preserve.

The alternative proposed does not arise from an abstract ideal but from paradigmatic recomposition founded on systemic logic. By dissociating income security from disability evaluation, and by articulating a three-level architecture—universal income foundation, targeted functional compensation, problem-solving oriented support services—it becomes possible to neutralize failure demand production at its source.

Beyond the disability field, the proposed analysis illuminates a more general tension of the contemporary welfare state. In a context of budgetary constraints, demographic aging, and fragile solidarities, the temptation is great to protect redistribution through control. This article shows that this strategy is self-destructive: it transforms social uncertainty into administrative uncertainty, and protection into ordeal. Conversely, universal foundations combined with targeted but non-existential schemes offer a more robust path for reconciling efficiency, justice, and legitimacy.

The issue is therefore not solely technical or budgetary. It is fundamentally political and normative. It involves deciding whether social protection regimes should continue functioning as suspicion devices, organizing access to rights through ordeals and ex post correction, or whether they can be rethought as institutions of trust, oriented toward security, continuity, and response to real needs.