The Organization of Abandonment: Homelessness and the Institutional Production of Premature Death by Wayne Martin Mellinger

 The Organization of Abandonment:

Homelessness and the Institutional Production of Premature Death

Wayne Martin Mellinger, Ph.D.

This is a chapter from the book I am developing entitled Abandonment in Real Time.

Abstract

This essay develops a unified theoretical framework for understanding homelessness as a sociology of abandonment. Drawing on ethnomethodology, interactional sociology, Dorothy Smith's institutional ethnography, and Loic Wacquant's analysis of advanced marginality, I argue that abandonment is not a policy failure but an institutional accomplishment — produced through the ordinary, routine work of homeless services bureaucracies. The framework is organized across four analytic levels: the structural context of urban growth regimes and surplus population production; the institutional organization of the homeless system as death regime, including its documentary practices; the interactional mechanisms through which abandonment is accomplished in real time; and the cumulative trajectories and forms of situated knowledge that result. These levels are analytically distinct but causally linked, each producing the conditions for the next. Together they constitute what I call the Layered Moral Economy of Oppression: a framework for understanding how abandonment is produced not through exceptional cruelty but through the accumulated weight of ordinary institutional practice.

I. Introduction: The Puzzle of Institutionalized Abandonment

Across cities in the United States, homelessness has become a persistent and highly visible feature of urban life. In many cities, extensive networks of services now exist to assist unhoused people — outreach teams, shelters, treatment programs, hospitals, case management systems, and coordinated entry infrastructures. These institutions are staffed by professionals who overwhelmingly express genuine concern for those they serve. And yet premature death among the unhoused remains common. In cities like Santa Barbara, California — among the wealthiest in the country — people die of exposure, untreated infection, and drug toxicity in the presence of a $30 million annual service infrastructure. They die not in the absence of care but surrounded by it.

How can systems organized around care produce outcomes that include exposure, abandonment, and early mortality? How can competent professionals acting in good faith participate in processes that, cumulatively, place vulnerable people at lethal risk? This essay develops a theoretical framework for answering these questions. Drawing on Harold Garfinkel's ethnomethodology, Erving Goffman's interactional sociology, Dorothy Smith's institutional ethnography, and Loic Wacquant's analysis of advanced marginality, it proposes a sociology of abandonment: an account of how ordinary institutional practices — rather than explicit intent to neglect or injure — organize exposure to harm.

This essay advances a sociology of abandonment to explain how contemporary homeless service systems can operate simultaneously as institutions of care and as organizational environments in which premature death becomes routine. The analysis argues that abandonment is produced through the ordinary institutional practices by which unhoused persons are rendered administratively legible, morally classified, temporally paced, and interactionally managed across multiple sites of service provision. Within this system, individuals move through a recognizable career of disposability in which encounters with outreach teams, shelters, hospitals, treatment programs, and bureaucratic assessments gradually reorganize them from potential beneficiaries of assistance into cases defined by noncompliance, high barriers, or service resistance. These classifications are stabilized through documentary practices that constitute the administrative person, while the interactional organization of care — what I call the machinery of ordinary cruelty — transforms delay, referral, and procedural neutrality into cumulative exposure to lethal environments. The result is a death regime in which overdoses, untreated illness, and exposure are rendered tragic but expected outcomes of homelessness itself. Yet those most exposed to these processes frequently develop an epistemology of abandonment — a situated knowledge of how institutional sorting operates — even as such knowledge is routinely discredited within official accounts. By integrating these mechanisms, the essay proposes a theoretical framework for understanding homelessness not simply as a condition of deprivation but as a patterned institutional accomplishment through which abandonment is produced in real time.

The argument is organized across four analytic levels, each of which produces the conditions for the next. At the structural level, urban growth regimes and housing markets generate surplus populations for whom homelessness is not a biographical misfortune but a structural location. At the institutional level, the homeless system operates as a death regime whose organizational logic is reinforced by documentary practices that constitute the administrative person before any face-to-face encounter occurs. At the interactional level, three mechanisms — the machinery of ordinary cruelty, abandonment in real time, and conditional dignity — translate institutional logics into lived harm. At the level of trajectories and knowledge, these accumulated encounters produce both the career of disposability and the epistemology of abandonment: a situated understanding of the sorting system that unhoused people develop precisely as it is being used against them. Together these four levels constitute what I call the Layered Moral Economy of Oppression — the emergent framework through which abandonment is produced not by any single actor or decision, but by the aggregate weight of individually defensible institutional practices whose collective effect is lethal. The essay proceeds through these four levels in sequence before a conclusion that draws them into a unified theoretical statement.

LEVEL I: STRUCTURAL CONTEXT

II. Urban Growth Regimes, Housing Markets, and the Production of Surplus Populations

Any adequate sociology of homelessness must begin at the structural level, not because structure determines interaction — it does not, in any mechanical sense — but because the institutional field within which abandonment is accomplished takes its shape from broader political-economic arrangements. Contemporary homelessness in American cities is not a residual problem, a temporary displacement, or a symptom of individual pathology. It is a structural feature of advanced capitalism's production of surplus populations.

Following Marx's analysis of the reserve army of labor and its elaboration by Mike Davis, Ruth Wilson Gilmore, and Matthew Desmond, we can understand the unhoused as constituting a specific fraction of what Wacquant calls the precariat — those expelled from stable labor markets and rendered superfluous to the reproduction of capital. The deindustrialization of American cities, the systematic defunding of affordable housing, the gutting of the social safety net since the 1980s, and the punitive turn in social policy have together produced a population for whom homelessness is not a temporary condition but a structural location. In cities like Santa Barbara, where median rents exceed any wage available to service workers, the production of homelessness requires no special explanation: it is the predictable output of a housing market in which shelter is a commodity rather than a right.

The concept of an urban growth regime — developed by Harvey Molotch and elaborated by subsequent urban scholars — designates the coalition of interests, typically real estate developers, local government actors, and growth-oriented business associations, that organizes land use decisions in American cities. Growth regimes are structurally oriented toward the maximization of exchange value in land and housing, which means they are structurally oriented against the preservation of affordable housing stock, the regulation of speculative development, and the interests of residents for whom housing is use value rather than investment. In high-cost coastal cities, the operation of growth regimes over several decades has systematically eliminated affordable housing, displaced lower-income populations from urban cores, and produced both the homelessness crisis and the political conditions under which that crisis is managed through criminalization rather than housing production.

Understanding homelessness as structurally produced by growth regime politics matters for the sociology of abandonment because it establishes that the population entering the homeless services system is not a random cross-section of the urban poor who happened to become homeless. It is a population that has been produced by specific political-economic arrangements, concentrated in specific locations, and subjected to specific forms of criminalization and social control that precede any encounter with the service system. The institutional apparatus that manages homelessness does not encounter a pre-given population of troubled individuals; it inherits a population that has already been subjected to decades of structural violence before the first intake form is completed.

Structure alone, however, cannot explain the differential outcomes within the homeless population. It cannot account for why some unhoused people receive services and others do not, why some are housed quickly and others cycle through the system for years, why some die and others survive. For these questions we require an analysis of institutions — of the specific organizational fields that have been constructed to manage the surplus population, and that in practice reproduce it. This is the work of the second analytic level.

LEVEL II: INSTITUTIONAL ORGANIZATION

III. The Homeless System as Death Regime

The concept of a death regime — borrowed and modified from Achille Mbembe's analysis of necropolitics — designates a set of institutional arrangements that systematically produce premature death among a specific population, not through direct killing but through the differential allocation of the conditions of life. A death regime does not murder; it manages. It does not exterminate; it sorts. But its outputs — measured in reduced life expectancy, elevated mortality from preventable causes, and deaths that occur at rates far above those of housed populations — are, in aggregate, lethal.

The homeless services system in American cities operates as a death regime in this sense. The evidence is not difficult to locate. The life expectancy of chronically homeless individuals is approximately thirty years below the national average. Mortality from drug overdose, cardiovascular disease, exposure, and untreated infection occurs at rates orders of magnitude higher among unhoused people than among the housed. In Santa Barbara County, as in counties across California, homeless deaths have increased consistently over the past decade even as the institutional infrastructure ostensibly dedicated to addressing homelessness has expanded. More services, more deaths: the paradox that defines the death regime.

The death regime concept shifts analytical attention from the individual to the institutional. The question is not why did Carlos die under that bridge. It is how the institutional environment is organized such that Carlos could die under that bridge in a city with a $30 million annual homeless services budget — and why his death, when it came, was recorded as a tragic consequence of homelessness rather than as the outcome of specific institutional practices and decisions.

The death regime operates through three primary institutional mechanisms. The first is triage abandonment: the systematic exclusion of those deemed most difficult to serve — those with serious mental illness, active substance use, prior criminal records, or histories of service refusal — from the programs most likely to produce stable housing. This exclusion is accomplished through categorical schemes that define eligibility in ways that structurally reproduce the population most at risk of dying. The second mechanism is temporal abandonment: the imposition of waiting times, processing delays, and bureaucratic deferrals that transform urgency into administrative routine, producing conditions under which treatable situations become life-threatening ones. The third is documentary abandonment: the requirement that people demonstrate eligibility through documentary performances that many are structurally unable to produce, thereby constituting them as administratively absent — present in the flesh but invisible to the system.

These mechanisms are not dysfunctions. They are features. They reflect the institutional logics of a system organized around accountability to funders, compliance with regulations, and management of liability — logics that are structurally indifferent to the survival of the population they nominally serve. The death regime does not require administrators who wish people harm. It requires only administrators doing their jobs within an institutional apparatus whose organizing logics produce lethal outcomes as a routine byproduct.

IV. Documentary Constitution and the Administrative Person

Institutions do not interact with persons directly. They interact with cases. The death regime accomplishes its sorting work not only through explicit eligibility criteria but through the documentary practices by which unhoused people are first translated into administratively processable cases. Dorothy Smith's institutional ethnography provides the central analytic tool here: the concept of the text as mediating artifact. For Smith, texts — forms, assessments, case notes, eligibility determinations — do not merely record social reality. They constitute it. They organize what counts as a case, what counts as need, what counts as compliance, and what counts as deserving. The unhoused person who appears at a service organization arrives not simply as a body in need but as a potential documentary subject — someone who must be translated into an administratively legible case before the institutional machinery can engage.

I call this process the documentary constitution of the unhoused: the set of practices through which a person's situation is translated into the categories, forms, and records that make it actionable within the institutional system. This translation is not innocent. It requires that the person's experience be reshaped to fit the available categorical schemes — that their substance use be characterized in specific diagnostic terms, that their housing history be rendered in formats compatible with database entry, that their behavioral patterns be interpreted through the lenses of mental health classification. In each of these translations, something is lost: the particularity of the person's situation, the relational context of their difficulties, the structural origins of their predicament.

What emerges from the documentary constitution process is what I call the administrative person: the version of the individual that exists within institutional records and that has determinate effects on what services they can access, how workers interact with them, and ultimately whether they are housed or abandoned. The administrative person may bear little resemblance to the person as experienced by themselves or by those who know them outside institutional contexts. A man who drinks because he lost his son and his job in the same year appears in the database as a chronic homeless individual with co-occurring substance use disorder and non-compliance with treatment recommendations. The loss, the grief, the structural circumstances are absent. The administrative failures are present.

Documentary constitution belongs analytically at the institutional level — prior to the interactional encounter — because it structures what workers see before they ever meet the person in front of them. The file precedes the face. A case manager encountering a new client does not meet that person as a stranger; they meet them as an already-constituted administrative person whose previous non-compliance, service refusals, and diagnostic history are present in the record before the first word is spoken. This pre-structuring of the encounter is one of the primary ways in which institutional logic reaches into interactional practice, shaping what is attended to, what is asked, and what interpretations are available.

One of the central mechanisms through which documentary constitution produces abandonment is what I call the eligibility paradox. The most vulnerable members of the unhoused population — those with the most severe psychiatric symptoms, the most chaotic living situations, the most extensive histories of trauma — are frequently those least capable of producing the documentary performances required to establish eligibility for services. They cannot consistently produce identification documents. They cannot complete multi-page intake forms. They cannot attend scheduled appointments. They cannot maintain the kind of stable biographical narrative that case management requires. The result is that the population most urgently in need of services is systematically constituted as ineligible for them — not because they lack need but because they lack the documentary capacity to prove it. Those who can perform eligibility are processed; those who cannot are abandoned before the interactional machinery is ever engaged.

LEVEL III: INTERACTIONAL MECHANISMS

V. The Machinery of Ordinary Cruelty

If documentary practices constitute the administrative person at the institutional level, interactional practices constitute the moral person — the version of the unhoused individual produced through face-to-face encounters with service workers, outreach staff, and institutional gatekeepers. These encounters are the sites at which abandonment is ultimately accomplished: where people are turned away, assessed as noncompliant, categorized as difficult, or simply not helped. They are also the sites that most resist analysis precisely because they appear, from within, as ordinary professional practice. Assistance is rarely refused outright. Instead it is delayed, redirected, or made contingent upon additional steps — each of which is locally reasonable, each of which contributes to a cumulative pattern of deferral.

I use the term the machinery of ordinary cruelty to designate the interactional mechanisms through which harm is produced without intention, without malice, and frequently without awareness. Cruelty here does not arise from malicious intent. It emerges through the routine execution of procedures that distribute responsibility across multiple actors and moments. Each participant experiences their action as locally justified — caseworkers must follow procedures, shelters must enforce rules, treatment programs must manage limited beds — yet the cumulative effect is exposure. The cruelty is ordinary in two senses: it is routine, unremarkable, embedded in the normal functioning of the institutional encounter; and it is accomplished through ordinary — that is, normal, professional, well-intentioned — interactional practices. The analytic task is not to explain why individual workers harm people but to describe the interactional organization through which harm is produced as a routine byproduct of competent professional practice.

The first mechanism is the normalcy assessment: the practice, embedded in intake and triage interactions, of evaluating the unhoused person's behavior against implicit standards of what a cooperative, treatment-ready client looks like. This assessment is not performed explicitly; it is accomplished through the management of conversational space — through what questions are asked and what answers are treated as adequate, through how affect is read and interpreted, through whether the person's account of their situation is taken up as a warrant for service or treated as evidence of unreliability. A person who speaks too quickly, appears too agitated, or whose account of their circumstances is internally inconsistent — all common features of trauma response, psychiatric disturbance, or simple exhaustion — is likely to be assessed as not ready for services, regardless of their objective need. The normalcy assessment is the interactional operationalization of triage abandonment: it is where institutional eligibility criteria are applied in real time through the practical management of conversational encounters.

The second mechanism is what Garfinkel called the documentary method of interpretation: the tendency to read each new piece of evidence in light of an already-established characterization of the person. Once an individual is categorized in the institutional record as non-compliant, service-resistant, or chronic — categories produced at the documentary constitution level — each subsequent interaction is filtered through this lens. A missed appointment confirms non-compliance. An expressed frustration confirms service resistance. A relapse confirms the chronic characterization. The person cannot escape their administrative identity because every action they take is interpreted through it. The documentary method of interpretation produces a self-fulfilling prophecy in which the administrative person colonizes the actual person, making it increasingly difficult for workers — however well-intentioned — to encounter the individual in front of them rather than the case in the file.

The third mechanism is moral accounting: the practice of making the receipt of services contingent on the performance of moral qualities deemed appropriate to the role of service recipient. These qualities include gratitude, compliance, optimism about recovery, and willingness to accept institutional definitions of the problem and its solution. Unhoused people who are angry, who challenge institutional authority, who refuse offered services in favor of their own assessments of their needs, or who simply fail to perform the affective register expected of a grateful recipient are likely to find services withheld, withdrawn, or made contingent on behavioral change. Moral accounting is the interactional mechanism through which conditional dignity — analyzed in Section VII — is produced and enforced moment by moment. It is the practice through which the institutional question of whether this person is deserving is answered in real time through the management of the service encounter.

VI. Abandonment in Real Time: Institutional Pacing and Temporal Management

Abandonment is not only accomplished through categorization and interactional assessment. It is accomplished through time. The temporal organization of the homeless services system constitutes one of its most consequential and least analyzed features — a mechanism of harm that is structurally invisible precisely because delay and waiting appear as administrative necessities rather than as practices with lethal effects. I call this dimension abandonment in real time: the production of abandonment not through explicit exclusion but through the imposition of institutional temporalities that are calibrated to the rhythms of the organization rather than the urgencies of the person.

The temporal architecture of homeless services is organized around a sequential processing logic: intake assessment, referral, waitlist placement, case assignment, service plan development, and then, eventually, housing search. Each step requires the completion of the previous one; each step introduces its own waiting period. For a person navigating this sequence from a position of acute crisis — sleeping outside in November, managing untreated withdrawal, carrying a psychiatric history that makes sequential appointment-keeping structurally impossible — the temporal architecture of the system is not a pathway to housing. It is a series of barriers, each of which has a formally neutral bureaucratic rationale and a practically lethal effect.

The concept of abandonment in real time is intended to capture something that purely structural or categorical analyses miss: the specific temporal experience of being processed by a system that moves at institutional speed while your body moves at biological speed. A person with an untreated foot infection who must wait three weeks for a medical intake appointment is not simply experiencing delay. They are experiencing the progressive worsening of a condition that will require hospitalization by the time the appointment arrives — if they survive to keep it. The infection does not wait for the appointment. The appointment does not accelerate for the infection. The gap between institutional time and biological time is where abandonment occurs.

Abandonment in real time is also accomplished through temporal dislocation: the imposition of a future-oriented temporal framework on a population whose survival imperatives are organized around the immediate present. Case management requires that clients maintain appointment schedules, complete paperwork within deadlines, attend workshops on specified dates, and plan for housing over timelines measured in months. These requirements presuppose a temporal orientation — a capacity to organize behavior around future contingencies — that is the product of stable housing, regular meals, adequate sleep, and freedom from acute threat. Chronic homelessness systematically destroys precisely these conditions. The person who fails to keep their case management appointments is not demonstrating irresponsibility; they are demonstrating the rational prioritization of immediate survival over institutional compliance. The system interprets this prioritization as evidence of non-compliance and records it as such, adding another layer to the documentary constitution of the undeserving case.

The temporal dimension of abandonment is further amplified by what might be called the urgency inversion: the institutional tendency for those in most acute need to receive the least timely response. Emergency shelter systems operate on a first-come-first-served or lottery basis that systematically disadvantages those too ill, too impaired, or too exhausted to arrive early and wait. Coordinated entry systems, designed to prioritize the most vulnerable, routinely have waitlists measured in years for the most intensive interventions. The most acute crises generate referrals to services with the longest waiting periods. Urgency, at the institutional level, produces not faster response but longer queues. The urgency inversion ensures that abandonment in real time falls most heavily on those for whom the gap between institutional time and biological time is most lethal.

VII. Conditional Dignity and the Moral Ordering of Care

The concept of conditional dignity designates a specific organization of the care relationship: one in which the dignity accorded to the recipient of care — the acknowledgment of their full human status, the recognition of their right to respectful treatment, the taking-seriously of their own accounts and assessments — is made contingent on their compliance with institutional expectations. This conditionality is distinctive of the homeless services encounter in ways that it is not, for instance, of the medical encounter, where patients are not ordinarily required to demonstrate worthiness before receiving care. In the American homeless services system, dignity is not a baseline; it is a reward for compliance.

Conditional dignity is not simply rudeness or disrespect, though both occur. It is a specific moral ordering of the care relationship that has determinate structural conditions and determinate effects. Its structural conditions include the legal-bureaucratic organization of eligibility, the institutional construction of the unhoused as a morally suspect population, the funding regimes that require demonstration of program effectiveness, and the professional frameworks through which service workers are trained to think about their clients. Its effects include the withdrawal of substantive help from those who refuse to perform the required moral register, the production of shame and self-blame among those who internalize institutional evaluations, and the generation of oppositional survival strategies among those who refuse such internalization.

The moral ordering of care in the homeless services context reflects what I call the deserving/undeserving binary: the persistent, historically deep division of the poor into those whose poverty is attributable to circumstances beyond their control and those whose poverty is attributable to character deficits. This binary, which has structured poor relief in Anglo-American contexts since at least the Elizabethan Poor Laws, has been elaborated in contemporary homelessness services through a series of more specific categorical distinctions: between those with mental illness and those with substance use disorders, between the chronically homeless and those newly homeless, between those who are housing ready and those who require treatment first. Each distinction carries a moral valence that organizes the practical allocation of resources and the quality of interactional treatment.

The conditionality of dignity is enforced through interactional practices that are formally neutral but morally consequential. The service worker who adopts a flat affective register with a client who has missed three appointments, who provides information in a minimal and technically adequate but relationally cold manner, who responds to expressions of frustration with procedural formality rather than human engagement — this worker is not violating professional norms. They are enacting them. The conditional dignity system does not require overt hostility; it requires only the systematic calibration of warmth, responsiveness, and relational engagement to the client's compliance performance.

The paradox at the heart of conditional dignity is that the population most in need of genuine relational recognition — those whose histories of trauma, abandonment, and institutional violence have most severely damaged their capacity for trust and their willingness to submit to institutional authority — is precisely the population least able to perform the compliance that would earn them that recognition. The system rewards those who need relational warmth least and withholds it from those who need it most. This inversion is not incidental. It is structural. It follows from the organization of care around institutional logics of management rather than human logics of need.

LEVEL IV: TRAJECTORIES AND KNOWLEDGE

VIII. The Career of Disposability

The three preceding levels — structural, institutional, and interactional — do not operate on discrete individuals in isolated encounters. They operate on people over time, across multiple encounters, in multiple institutional settings, across years and sometimes decades. The fourth analytic level attends to this temporal accumulation: to what happens to a person's social identity, institutional legibility, and life chances as the weight of structural, institutional, and interactional processes accumulates across a biographical trajectory. I call the result the career of disposability.

Goffman's concept of the moral career — the patterned sequence of changes in a person's self-concept and social identity as they move through institutional processing — provides the foundational analytic tool. I develop this concept to track not only changes in self-concept but changes in institutional constitution: the progressive transformation of a person from a potential beneficiary of assistance into someone whose abandonment has been organizationally normalized. The career of disposability is not a deliberate institutional project. It is an emergent effect of the accumulated weight of institutional encounters, each of which makes a small contribution to the progressive constitution of the person as someone whose survival is not a priority.

The career of disposability is accomplished through what I call the ratchet mechanism: a one-directional institutional logic in which negative evaluations accumulate and are very difficult to shed, while positive performances are discounted or treated as baseline. The first shelter turn-away for rule violation; the first case note recording non-compliance; the first treatment provider who diagnoses service resistance; the first coordinated entry assessment that scores the person as ineligible for permanent supportive housing; the first arrest for camping on public property — each deposits a layer of institutional evaluation that becomes part of the administrative person and structures every subsequent encounter. The ratchet turns in one direction only. Sobriety is baseline; relapse is documented. Appointment attendance is expected; missed appointments are recorded. The documentary career accumulates negative valence while positive performances disappear into the unmarked background of adequate compliance.

Over time, the accumulated ratchet produces what I call administrative death: the institutional classification of a person as someone for whom no further interventions are available or indicated. Administrative death precedes biological death — sometimes by years. It is accomplished through the bureaucratic closure of cases, the formal documentation of service exhaustion, and the informal institutional consensus that an individual is not ready for housing or services. Once a person achieves administrative death, they remain physically present in the institutional environment — they appear on outreach lists, they occasionally appear at service sites — but they have been constituted as beyond the reach of institutional intervention. When they die, their death confirms the institutional account: they were someone for whom nothing could be done.

The concept of administrative death is important precisely because it shifts analytical attention from the moment of biological death to the institutional processes that precede it. The question is not why Marcus died on the corner of State Street and Haley on a Tuesday in February. The question is how the institutional career he had been traversing for nine years constituted him as someone whose dying was, in a specific institutional sense, already accomplished before his body stopped breathing. The answer lies not in Marcus's choices but in the ratchet, the eligibility paradox, the normalcy assessment, the urgency inversion, and the conditional dignity system — the accumulated machinery through which his career of disposability was built, one encounter at a time.

IX. The Epistemology of Abandonment

The career of disposability produces not only a trajectory of progressive institutional exclusion but a distinctive form of knowledge. People who have traversed the homeless services system across multiple years, who have been sorted, assessed, referred, waitlisted, discharged, and re-enrolled across dozens of institutional encounters, develop a sophisticated practical understanding of how the sorting system operates. They know which shelters enforce sobriety requirements and which do not. They know which outreach workers will advocate for them and which will not. They know which documentation they need and which eligibility performances are required. They know, with considerable precision, how the institutional machinery works — and they know it from inside, as subjects of its operation rather than as its operators.

I call this situated knowledge the epistemology of abandonment. The term is intended to capture both the form and the content of what is known: a practical, embodied, interactionally-grounded understanding of how institutional sorting produces abandonment, developed through direct experience of being sorted. This knowledge is real knowledge. It is accurate, refined through repeated encounter, and often theoretically sophisticated in ways that formal academic analysis is not. When a man who has been homeless for seven years explains that the coordinated entry system is designed to document need rather than address it, he is not simply venting frustration. He is offering an institutional analysis that is, in many respects, more precise than the analyses produced by program evaluators who have never slept outside.

The epistemology of abandonment is, however, a peculiarly powerless form of knowledge. It is knowledge that cannot easily be converted into institutional leverage. It is knowledge that is routinely discredited within official accounts — attributed to paranoia, to the cognitive effects of substance use or mental illness, to the characteristic unreliability of homeless testimony. The mechanisms through which this discrediting operates constitute the second dimension of the epistemology of abandonment: not only the situated knowledge that unhoused people develop but the structured organization of knowledge-production that renders that knowledge inadmissible.

This structured organization operates through the individualization of causation: the institutional and journalistic practice of explaining each homeless death by reference to the individual circumstances of the person who died. Carlos died because of his fentanyl addiction. Maria died because she refused shelter. James died because of untreated schizophrenia. Each of these accounts may be locally accurate in the sense that the immediate cause of death was indeed overdose, or exposure, or cardiac arrest following years of untreated illness. But they systematically deflect attention from the institutional practices that constituted the conditions of dying. They render structural causation invisible by foregrounding individual causation — and in doing so, they work directly against the epistemology of abandonment by converting what unhoused people know about how the system produces their abandonment into evidence of individual pathology.

The dispersal of accountability compounds this effect. The homeless services system is organized as a loosely coupled network of agencies, each with responsibility for a specific piece of the service system and none with responsibility for the whole. When a person falls through the cracks between agencies, or dies while cycling between programs, the question of who is responsible is genuinely difficult to answer — not because no one is responsible but because responsibility has been organized to be undeterminable. This organizational design ensures that the institutional knowledge held by unhoused people — knowledge of exactly which cracks exist and how people fall through them — cannot be converted into accountability, because the institution whose accountability would be warranted does not exist as a singular, locatable entity.

The sociology of abandonment must take a position with respect to the epistemology of abandonment. This means insisting on analytic frameworks that make structural and institutional causation visible, refusing the individualization of causation, and taking seriously the knowledge that unhoused people develop through direct institutional experience. It means attending to what people who are being abandoned know about how abandonment works, and treating that knowledge not as anecdote or complaint but as primary sociological data about the operation of the institutional complex. The epistemology of abandonment is the knowledge the death regime produces in its subjects — knowledge that is, in the most literal sense, hard-won.

X. Conclusion: Toward a Sociology of Abandonment

The theoretical framework developed in this essay offers a specific way of seeing: a way of looking at the homeless services system and seeing not the provision of care but the production of abandonment; not the management of a social problem but the reproduction of premature death; not the failure of well-intentioned institutions but the success of an institutional apparatus organized around logics that are structurally indifferent to the survival of those it processes. This way of seeing is organized across four levels that are analytically distinct but causally linked — each level producing the conditions for the next, and each requiring its own conceptual vocabulary.

The four levels described in this essay are not parallel tracks. They are causally linked: each produces the conditions for the next, and each is required to account for the full mechanism of abandonment. Structural forces generate surplus populations who enter the institutional complex already bearing the weight of accumulated dispossession. Institutional logics then constitute those populations documentarily — producing administrative persons before any face-to-face encounter occurs. Interactional mechanisms translate those documentary constitutions into lived sequences of delay, moral assessment, and withheld recognition. And the cumulative weight of these encounters produces, over time, both a career of disposability and the hard-won, largely discredited knowledge of how the sorting machine works. No single level is sufficient. The harm is produced in the movement between them.

This is what the Layered Moral Economy of Oppression names¹: not a system in which individual actors are cruel or negligent, but one in which harm is the emergent property of the system as a whole. The harm is produced at the intersection of levels, in the gaps between organizations, in the accumulation of individually defensible decisions whose collective weight is lethal. The layering is not incidental to this outcome — it is constitutive of it. It is precisely what makes the system so difficult to reform and so easy to defend. At each level, the logic is coherent; at no level does anyone appear to be doing something unreasonable. And so people die, and their deaths are absorbed into an institutional account that renders the system blameless.

What, then, is the purchase of this framework? Three things, at minimum. First, it shifts the object of analysis from individual homeless people — their pathologies, their choices, their refusals — to the institutional practices that produce abandonment. This shift has direct implications for accountability: if abandonment is institutionally produced rather than individually chosen, then institutional actors bear institutional responsibility, and institutions can be held accountable in ways that individual biography cannot be. Second, the four-level architecture provides a vocabulary precise enough to distinguish the different mechanisms through which abandonment is accomplished at different registers, and therefore precise enough to identify specific sites of potential intervention — not generic policy reform but targeted disruption of specific mechanisms at specific levels. Third, and perhaps most importantly, the framework takes seriously the people who are dying — not as cases but as persons, and not as the authors of their own abandonment but as the subjects of an institutional apparatus whose operation they understand, often with considerable clarity, from the inside.

The sociology of abandonment is, at its core, a refusal of the epistemology of abandonment: a refusal to accept that homeless deaths are the natural, inevitable outcome of individual misfortune, and an insistence that they are the preventable outcome of institutional practices whose organization can, in principle, be changed. Whether they will be changed is a political question. Making that visible — converting what appears as natural fact into a question of institutional accountability — is the first task of a sociology of abandonment.

People die under bridges in wealthy cities not because they are fundamentally different from people who die in hospital beds, not because their lives are worth less, and not because nothing could have been done. They die because of how the institutions charged with their care are organized, what logics govern those organizations, and what practices those logics require across four interlocking levels of social life. A sociology of abandonment insists on this, and in insisting on it, refuses to let the deaths be ordinary. Understanding this dynamic does not in itself resolve the problem of homelessness. It does, however, clarify how systems organized around care can participate in the reproduction of abandonment — and it opens toward forms of care that do not merely manage vulnerability but fundamentally reorganize the conditions under which abandonment becomes possible.

Note

¹ The layered architecture of the Layered Moral Economy of Oppression bears conceptual affinity with Charles Goodwin’s notion of laminated action (2018), in which multiple semiotic fields — gesture, gaze, artifact, prior text — are simultaneously active and mutually constitutive in any moment of co-operative action. The intellectual resonance is real: in both frameworks, meaning or harm emerges from the intersection of analytically distinct layers rather than from any single layer operating alone. The critical distinction, however, is temporal. Goodwin’s lamination is synchronic — multiple fields active in a single interactional moment. The LMEO’s layering is diachronic: the levels accumulate across time and across a biographical trajectory, with each level producing the conditions for the next. It is this temporal depth — the ratchet of encounters across years, the documentary career that precedes every new encounter — that distinguishes the LMEO as a framework for understanding abandonment as a biographical and institutional process rather than a moment of interaction.

References

Davis, Mike. 1990. City of Quartz: Excavating the Future in Los Angeles. London: Verso.

Desmond, Matthew. 2016. Evicted: Poverty and Profit in the American City. New York: Crown.

Garfinkel, Harold. 1967. Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.

Garfinkel, Harold. 2002. Ethnomethodology's Program: Working Out Durkheim's Aphorism. Lanham, MD: Rowman & Littlefield.

Gilmore, Ruth Wilson. 2007. Golden Gulag: Prisons, Surplus, Crisis, and Opposition in Globalizing California. Berkeley: University of California Press.

Goffman, Erving. 1961. Asylums: Essays on the Social Situation of Mental Patients and Other Inmates. Garden City, NY: Anchor.

Goffman, Erving. 1963. Stigma: Notes on the Management of Spoiled Identity. Englewood Cliffs, NJ: Prentice-Hall.

Mbembe, Achille. 2003. "Necropolitics." Public Culture 15(1):11-40.

Molotch, Harvey. 1976. "The City as a Growth Machine: Toward a Political Economy of Place." American Journal of Sociology 82(2):309-332.

Smith, Dorothy E. 1987. The Everyday World as Problematic: A Feminist Sociology. Boston: Northeastern University Press.

Smith, Dorothy E. 1990. Texts, Facts, and Femininity: Exploring the Relations of Ruling. London: Routledge.

Smith, Dorothy E. 2005. Institutional Ethnography: A Sociology for People. Lanham, MD: AltaMira.

Tsemberis, Sam. 2010. Housing First: The Pathways Model to End Homelessness for People with Mental Illness and Addiction. Center City, MN: Hazelden.

Wacquant, Loic. 2008. Urban Outcasts: A Comparative Sociology of Advanced Marginality. Cambridge: Polity.

Wacquant, Loic. 2009. Punishing the Poor: The Neoliberal Government of Social Insecurity. Durham, NC: Duke University Press.

Goodwin, Charles. 2018. Co-operative Action. Cambridge: Cambridge University Press.

Wolch, Jennifer and Michael Dear. 1993. Malign Neglect: Homelessness in an American City. San Francisco: Jossey-Bass.