The diagnostic state: Police discretion and the securitization of HIV/AIDS in the United States

Abstract

How do law enforcement agencies shape public health responses during epidemics? While scholars have examined the role of established medical institutions such as the Centers for Disease Control and the World Health Organization, the police remain understudied. Through a multimethod analysis of archival materials, national survey data, and comparisons across New York and San Francisco, I argue that police functioned as informal medical knowledge producers who transformed HIV/AIDS from a public health challenge into a security threat requiring punitive intervention. Vague federal guidance opened a discretionary space that officers filled with their own ideas rather than medical evidence. I introduce the concept of the “diagnostic state” to describe how the structures of policing allowed officers to define what counts as illness and how it should be managed. In the absence of clear protocols, officers relied on visual heuristics to determine infection status, effectively constructing an extralegal diagnostic apparatus that operated parallel to—and often in contradiction with—public health guidance. These patterns established precedents that foreshadowed COVID-19 responses, illustrating the ways that institutional authority can either reinforce or disrupt efforts to address epidemics.

Introduction

How do law enforcement agencies shape public health responses during epidemics? When HIV/AIDS emerged in the early 1980s, police departments faced an unfamiliar virus and no clear mandate for how to respond. Their role was undefined. So, their gaze turned inward. Keeping officers safe—and uninfected—became the overriding concern. Federal directives reinforced the institutional priority of occupational hazard and officer protection, a framework that effectively displaced broader public health concerns and the communities most affected by the epidemic. From the outset, the question was never how to protect the public from the virus, but how to protect officers from the public.

Much has been written on the confusion, rapidly shifting knowledge, and fragmented institutional responses to HIV/AIDS (Adetiba, 2025; Cohen, 1999). ¹ Initially, the virus outpaced understanding, leaving medical professionals, community organizers, and policymakers scrambling. Knowledge arrived in fragments, often laced with error, and policies took shape in the shadow of that uncertainty. These errors were structural but predictable outcomes of responding to a novel pathogen in a context where knowledge was contested. Since then, scholars working at the nexus of disaster studies, organizational behavior, the anthropology of epidemics, and the sociology of knowledge have converged around the question: how do institutions function when scientific understanding remains unsettled, when evidence is incomplete, public anxieties escalate, and policy decisions must be made before a consensus is reached? One consistent theme across these bodies of scholarship is that institutions often falter under such pressures, making consequential mistakes and perversely exacerbating the injustices they claim to address (Collier and Lakoff, 2015; Lakoff, 2008, 2017; Remes et al., 2021; Shapin, 1995).

Within carceral studies, the literature has primarily focused on how correctional facilities responded to HIV/AIDS, often underscoring the violent conditions that define life inside prisons. These accounts detail how wardens, correctional officers, and other prison administrators failed to manage the epidemic with adequate care or competence (Fleury-Steiner, 2008; Rubenstein et al., 2016). But police occupy a distinctly visible position in everyday life, patrolling streets and engaging in routine interactions that make policing central to public experience (Flavin, 1998). Without analyzing how the police functioned during the HIV/AIDS epidemic, we overlook a critical site where public health, criminalization, and institutional power collide and intersect.

In this paper, I document US police responses to HIV/AIDS in the 1980s and 1990s, and the institutional logics underlying their actions. Central to my analysis is the 1987–1988 National Institute of Justice (NIJ) report “The Law Enforcement Officer and AIDS.” Flavin (1998) demonstrated how NIJ publications rendered HIV/AIDS an occupational hazard lurking in every arrest, search, and public interaction. ² Building on these insights, I argue that “The Law Enforcement Officer and AIDS” offered more than safety advice; it made the police diagnosticians. Officers exercised literal diagnostic power: the authority to read bodies for symptoms, determine infection status, and dictate what interventions were necessary. Such authority was a cultural force—the power to fuse sexuality, race, and illness into a single interpretive frame, and conscript medicalization into a technology of moral assessment (Epstein and Timmermans, 2021).

First, I historicize the relationship between policing and public health, showing how the treatment of disease has always carried a punitive ethos. Diagnosis itself has never been purely medical; it emerges from negotiations between symptoms and social categories, scientific knowledge and cultural meaning, clinical observations and moral judgments (Brown, 1995). I argue that scientific uncertainty surrounding HIV/AIDS created the conditions under which police could layer their own diagnostic frameworks onto already overdetermined processes, producing categories that lived on the margins of medical discourse, animated by carceral logics, and dressed with the force of law. Second, I extend conversations on police discretion by revealing how vague federal guidance enabled officers to exercise diagnostic judgment in routine encounters. Third, I examine San Francisco and New York as opposing models of implementation—one championed as harm reduction, the other as disciplinary intervention—to foreground the pervasive structural tensions around policing disease and, more provocatively, to ask whether public health concerns and policing might be incompatible projects. Lastly, I highlight the enduring legacies of HIV/AIDS-era policing on contemporary health emergencies.

Methods

I employed a multimethod approach consisting of three components: archival ethnography, national survey data, and comparative analyses of New York and San Francisco.

Drawing on anthropological and historical approaches to bureaucratic reasoning and state institutions (Hull, 2012; Riles, 2006), I read federal policy documents, training manuals, legal testimony, and municipal debate transcripts not merely as sources of information, but as expressions of institutional life—records of how police departments learned, reasoned, and feared HIV/AIDS. Here, the archive functions more like an ethnographic field site, where officers and policymakers communicated the moral economies and knowledge systems that structured their understandings of risk and contagion. Reading across these documents reveals an institutional habitus (Bourdieu, 2018)—a patterned way of knowing—that systematically framed disease as disorder and infection as evidence of deviance. I reconstruct the affective and cognitive infrastructure of diagnostic policing, showing how discretion was produced not only in the streets but in the bureaucratic imagination of the state. All archival materials referenced were drawn from publicly accessible digital repositories maintained by the NIJ, the National Library of Medicine, the United Nations Office on Drugs and Crime, The New York Times, and Human Rights Watch.

To scale the archival findings from documents to institutions, I review national survey data from 1981 to 1997 (Burgess et al., 1992; Edwards and Tewksbury, 1996; Thompson and Marquart, 1998; Yearwood, 1992). Following the issuance of the NIJ report, Edwards and Tewksbury (1996) surveyed 49 state police academies and identified substantial inconsistencies. While 86% of academies offered either basic or in-service HIV/AIDS training, the quality and content varied dramatically. Seventy-one percent adopted specific policies for handling suspected or confirmed HIV/AIDS cases—a majority, but one that still left nearly a third of academies without formal guidance, despite national standards existing for over a year. These figures dropped sharply when it came to internal governance—only 25% had policies in place to address HIV-positive employees, with many pointing to federal guidelines, such as the NIJ report, as sufficient reason to forgo creating their own. The rationales for policy adoption varied as well. Over half (55%) of police academies reported implementing policies to comply with legal mandates—most often referencing Occupational Safety and Health Administration (OSHA) requirements. Thirty-five percent implement policies with the main goal of protecting officers, and just 10% emphasized ensuring fair treatment of citizens.

Thompson and Marquart's (1998) synthesis of survey data from 142 officers revealed that police lacked adequate knowledge about HIV/AIDS transmission. More than a third (37.1%) incorrectly believed unprotected heterosexual contact was the primary transmission method. Roughly 35% identified intravenous drug use, while 19.3% believed that male homosexual/bisexual contact was the leading source. Among the remaining officers, 7% cited “other” unspecified means and only 1.4% mentioned accidental bodily fluid exposure. Remarkably, zero identified contaminated blood transfusions or transplant procedures as possible routes. Additionally, Burgess et al. (1992) found that HIV/AIDS-related fear among officers registered at 5.2 on a 10-point scale. These national patterns of institutional variation, knowledge deficits, and officer anxiety provide essential context for understanding the archival materials and local practices examined in the following analysis. I reference these data periodically to highlight the dominant attitudes and prevailing norms across precincts, though my primary analytical focus remains on archival documents that capture how these dynamics unfolded on the ground.

Nevertheless, these data illuminate how discretionary authority was institutionally distributed, offering a national baseline from which to contextualize the comparative analysis of New York and San Francisco—two jurisdictions largely considered to be geographic epicenters of the US epidemic and, by the mid-1980s, became global symbols of crisis. Scholars have consistently examined these cities to understand state responses to HIV/AIDS (Arno and Hughes, 1989); I extend these conversations by focusing on the police, demonstrating that bureaucratic rationalities remained strikingly consistent across both cities despite their contrasting HIV/AIDS models. While discretion varied in local expression, it was structurally reproduced, suggesting that lauded differences masked deeper institutional continuities.

The diagnostic state: Preconviction medical abandonment

From its earliest days, HIV/AIDS was defined by punishment. The Reagan administration decidedly intensified police surveillance, arrests, and draconian sentencing through its war on drugs, sparking a militarized turn with fortified narcotics units and coordinated federal–local offensives targeting crack users and dealers (Williams, 2021). When HIV/AIDS emerged, and quickly became associated with intravenous drug use, policing escalated to another level, confirming that punishment, not rehabilitation, was the focus of the US penal system. As Hoppe (2017: 6, 12) writes, lawmakers were “already in the habit of proposing handcuffs and prisons as solutions to social problems […] paving the way for a new era of criminalization that targeted disease.” Given the way HIV/AIDS was narratively bound to both drug use and sexuality, the epidemic provided a new justification for the repression of queer communities. William F. Buckley Jr. infamously used a 1986 New York Times column to advocate branding the newly diagnosed with tattoos; more broadly, the Supreme Court's decision in Bowers v. Hardwick (1986) to uphold a Georgia sodomy law reinforced the legal basis for state intrusion into queer intimacy. ³ Armed with these legal and rhetorical covers, raids on bathhouses, nightclubs, and public spaces followed. Because certain groups were increasingly targeted, sickness and illness could be rendered as criminal matters, and police authority could now extend into the most intimate spaces and private lives.

But when, exactly, do police acquire medical authority? While officers have long engaged in informal diagnostic assessments—evaluating intoxication, mental illness, and contagion—HIV/AIDS policing seemed to rest on three enabling conditions: (1) contested scientific knowledge creating interpretive flexibility, (2) federal policy ambiguity enabling discretion rather than constraining it, and (3) pre-existing stigmatization of affected populations that allowed diagnosis to operate through familiar categories of deviance.

Epstein (1996) demonstrates that uncertainty in the biomedical domain often creates a vacuum that other institutions—political, cultural, or bureaucratic—quickly fill. As doctors struggled to understand how the disease spread, the police moved decisively into that space of public health knowledge acquisition and offered their own explanations. There was a struggle over the meaning of HIV/AIDS: different institutions trying to make sense of populations, racing to diagnose the causal factors, and desperate to arrive at clear answers and decisive solutions. But the question of who dies, has died, or will die was also a question of representation (Hall, 2006; Treichler, 2020). In what Epstein (2008) calls “the biological paradigm,” HIV/AIDS was a site where society worked out fundamental political questions about social justice, citizenship, and equality at both the biological and social levels. Science bleeds into statecraft, and statecraft bleeds into science, or as Reich (2025) observes, “biomedicine is politicized, just as governance is increasingly biomedicalized.” Knowledge splintered into competing interpretations, each bearing the imprint of its institutional origin and the cultural logics animating its claims.

In the chaos of the epidemic, when every new case raised more questions than answers, and when public terror outpaced medical knowledge, the police constructed their own medical realities, which the law then validated. Even as epidemiologists worked to destigmatize the virus and explained that handshakes, shared utensils, or routine contact posed no transmission risk, police continued to associate it with criminality and danger. In arrest reports, phrases like “appears to have AIDS” or “possible HIV infection” appeared directly alongside charges (Hoppe, 2017). Underlying these assessments were categorical assumptions of threat, risk, and pathology: Who looked sick enough to have AIDS? Who appeared dangerous enough to transmit HIV? Police work has always proceeded from these very logics of suspicion; officers learn to interpret social problems as matters of order and disorder (Ericson and Shearing, 1986; Najdowski et al., 2015). As Zacka (2017: 70) reminds us, officer training is more than a distant psychological exercise to be recalled when needed; rather, it becomes “embodied and embedded in a quasi-automatic mode of response,” a practiced intuition that shapes perception before conscious thought intervenes. ⁴ The result was a bifurcated regime of knowledge, one produced in clinics and laboratories, another generated on street corners and in police cruisers. Both claimed authority over the social meaning of the epidemic.

These punitive responses to public health emergencies represent what I call the diagnostic state—the way institutional and professional structures and norms shape the knowledge base producing our assumptions about the prevalence, incidence, treatment, and meaning of disease. Epidemics, as Rosenberg (2020) observes, have a dramaturgic form. They unfold along a plot line of increasing and revelatory tension, build toward a crisis that challenges the moral character of individuals and communities, and from there, reaffirm fundamental social values and modes of understanding (Rosenberg, 2020). Under the diagnostic state, the drama unfolds along predetermined lines—only some lives appear worthy of treatment, while others remain marked as threats. To understand how these scripts endure and achieve the status of common sense, we might pause to ask what kind of social activity diagnosis actually is.

As sociologists of diagnosis have long argued, the act of naming a condition is fraught precisely because it assigns social identities; it can stigmatize, criminalize, and sort populations into hierarchies of moral worth that extend far beyond clinical encounters (Jutel, 2009; Jutel and Nettleton, 2011). Brown (1990, 1995) famously urged scholars to ask different questions of diagnosis entirely, to ask who gets to define illness, through what institutional means, and toward what ends for those rendered objects of classification. Even health professionals, as Rivera-Cuadrado (2023) demonstrates in the context of COVID-19, construct their own understandings of risk through localized assessments of spatial exposure, the perceived inadequacy of protective gears like masks and gloves, and visceral perceptions of danger. If medical practitioners themselves construct risk situationally—through proximity, material conditions, and felt danger—the consequences of diagnosis become all the more acute when appropriated by those with little to no clinical training. Diagnosis, in other words, exceeds the identification of pathology; it mediates the relationship between biomedical knowledge and institutional power, while naturalizing those distributions as inevitable sociomedical outcomes. The diagnostic state, as I theorize it here, is a question of how state actors, namely the police, establish a “discursive monopoly” over the meaning of disease, crafting narratives that seek to override and subsume the counterpublics that challenge them. Foregrounding the police not only expands the sociology of diagnosis beyond the clinic, but insists that we interrogate different points of contact—the street encounter, the arrest, the raid—as scenes where fragments of medical discourse, moral judgment, and carceral logic all converge.

The processes of medicalization and diagnosis flattened categories and merged identities together (Conrad, 1992; Jutel, 2009; Jutel and Nettleton, 2011). Was a positive test a medical fact, a moral judgment, or a criminal sentence? The answer was all of the above. What emerged was a system where those deemed criminal by virtue of their sexual deviance forfeited the right to quality healthcare, what some refer to as “carceral medicine” (Pitts, 2019). But what happens when carceral medicine is practiced before conviction?

Scholarship on punishment focuses heavily on sentencing, incarceration, and postrelease supervision, but police exercised penal authority before courts became involved (Soss and Weaver, 2017). Visual readings of moral worth predetermined who should be left to suffer, which rendered questions of care irrelevant before they could even arise. When officers refused aid to someone “who looked like a junkie” or identified potential HIV-positive individuals by appearance, these judgments had no clinical basis (Leinen, 1992). Police training cannot reasonably encompass specialized medical education, but they made these determinations nonetheless. These practices constitute an untheorized mechanism of punishment, what I call preconviction medical abandonment. Preconviction medical abandonment describes how punishment operates outside formal legal processes, how the carceral state extends its reach not only through confinement but through the calculated withdrawal of care. The police enacted a street-level epidemiology that overrode public health measures with necropolitical calculations of deservingness (Mbembe, 2020). Diagnosis, here, collapsed into judgment; judgment collapsed into a (death) sentence. In a sense, the officer became a physician, judge, and executioner at once, not through any formal delegation of power, but through the interpretive latitude that naturalized indifference toward some lives.

Preconviction medical abandonment, too, belonged to a broader reconfiguration of everyday police work. By the 1990s, as Schrader (2019a, 2019b) demonstrates, American policing had also undergone a profound transformation, formally adopting militarized tactics, surveillance technologies, and the expansionist logics of the Cold War. Police had badges and no borders, their power stretched across jurisdictions and into new domains with little oversight or resistance. The mandate to “protect and serve” had become infinitely elastic, ready to encompass public health as yet another space where threats must be contained and populations managed through force. Order-maintenance became counterinsurgency by another name.

White (2023) further historicizes these practices with his analysis of infectious disease management in the twentieth century. White argues that international health regulations have been governed by what he terms “epidemic orientalism,” an organizing logic of the West versus the rest that constructs disease as something that originates elsewhere and threatens civilized societies, where certain populations become marked as sources of disease in need of isolation and control. What epidemic orientalism reveals at the scale of global health governance, we can observe in the federal directives that structured everyday police work during the epidemic. Policing has become central to how we manage epidemic threats—a configuration that demands critical examination, not only for what it aims to contain, but for what it quietly permits and perpetuates.

From occupational safety to social hygiene: The NIJ and the practice of selective knowability

Under Director James K. Stewart (1982–1990), the NIJ was heavily oriented toward drug enforcement, devoting approximately 60% of its research funding to examining trafficking patterns, developing enforcement strategies, and evaluating intervention programs (Stewart, 1989). Situated in the mid of the War on Drugs, these institutional priorities collided with the emergence of HIV/AIDS. As drug enforcement intensified, arrests of low-level users and sellers multiplied, and officers treated daily contact with drug users as daily contact with AIDS. To police departments, it was self-evident: if AIDS killed deviants, and they arrested deviants, they were arresting the infected. Because the NIJ viewed HIV/AIDS through drug arrests, the scale of the epidemic—for them—was as large as the War on Drugs made it.

By the mid-1980s, police departments were demanding answers. The NIJ decidedly established a clearinghouse to manage the deluge of inquiries from law enforcement agencies. Within two months, they received over 700 calls (Flavin, 1998). The volume made it clear that police wanted more than periodic updates and telephone consultation; they were looking for something to reference, something written for them. The Centers for Disease Control (CDC), for instance, staffed by epidemiologists and public health officials, addressed the general public with messages about harm reduction (Adetiba, 2025). The NIJ operated differently, staffed largely by current and former law enforcement personnel. Officers preferred guidance that came from within their own ranks (Schrader, 2019b). Officers preferred to hear from themselves.

In 1987, the NIJ issued “The Law Enforcement Officer and AIDS,” a document that purported to provide “all the knowledge” officers needed to protect themselves. ⁵ Distributed to every police department in the United States, it established federal guidance for what officials deemed an unprecedented occupational danger. Yet for all its authoritative status as the national standard, the report was relatively brief. At 50 pages, each section focused on the most basic contours of the disease: “History,” “Treatment of AIDS,” “Testing for AIDS,” “Bodily Fluids Which May Contain the AIDS Virus,” “Transmission of AIDS,” “AIDS Cases in the United States,” to name a few. ⁶ The 1988 addendum offered marginal additional content. Rather than contextualizing field encounters or elaborating on the complexities that officers might face, the revised edition largely appended CDC statistical updates with minimal analytical mediation between public health surveillance data and law enforcement practices. Its core instructions, however limited in scope, established a distinctive approach to risk.

The report instructed officers to operate under the presumption of universal infection: “A police officer or correctional facility officer will not necessarily know whether or not a person is infected with AIDS or any other communicable disease, especially during an arrest. Therefore, an officer should assume that any person they are arresting or dealing with is potentially infectious.” ⁷ In principle, universal precaution was sound public health guidance because it normalized protective measures as standard practice. The conclusion reinforced such reasoning: “We have all the knowledge we need to protect ourselves from AIDS both on and off the job […] we have learned that it is difficult to contract AIDS from being on the job.” ⁸

Yet what the NIJ failed to do was reconcile these claims in ways that officers could operationalize. The report never clarified transmission routes with precision, never distinguished between encounters that carried actual risk and benign contact, and never specified what “sensible precautions” might entail. For officers already uncertain about the disease—as the survey data made clear—the message that everyone should be presumed infectious, delivered along the assurances that transmission was unlikely, generated confusion rather than clarity. What did it mean to treat everyone as potentially infectious if the disease was difficult to contract? Which encounters warranted caution, and which did not? The NIJ had an opportunity to resolve institutional uncertainty by educating officers on the clinical realities of transmission. Instead, it deferred to abstraction: “Remember that some situations will arise not covered by guidelines or policies, and that caution and common sense should prevail.” ⁹

“Common sense” did considerable ideological work here. Survey data revealed that roughly one-third of officers feared contracting HIV through routine contact, despite epidemiologists repeatedly dismissing such concerns as baseless (Yearwood, 1992). But decades of militarized policing now allowed these anxieties to be enacted in the field without consequences (Go, 2024). As one police officer stated plainly: “I would not administer aid to anyone who looked like a junkie or a fag, no matter who they were” (Leinen, 1992).

The NIJ report ultimately was agenda setting; it made the construction of occupational risk a problem other policing institutions felt compelled to address over the next four years. In 1989, when the International Association of Chiefs of Police promulgated a model policy on “Communicable Disease,” it re-emphasized the NIJ's position that officers faced “potential exposure” but never established clear thresholds for possible risks (Flavin, 1998). Two years later, the OSHA mandated new protocols for “Occupational Exposure to Bloodborne Pathogens,” requiring comprehensive workplace safety measures, including protective equipment, exposure control plans, and postexposure medical surveillance (Flavin, 1998).

Even as these authorities enacted these precautions, they failed to answer a critical question: how many officers had actually died from occupational exposure? Researchers, too, disagreed about the prevalence of occupational transmissions; while Hammett et al. (1994) identified no deaths, Bigbee (1993) found evidence of seven exposure incidents between 1981 and 1991, with at least three resulting fatalities. The NIJ report arrived directly in the middle of these debates, but offered no clear resolution between them: “There have been several reported cases of occupational transmission of the AIDS virus to law enforcement personnel. However, these cases, which were reported by the police agencies [remain] under investigation by public health authorities and have not been confirmed as occupationally acquired.” ¹⁰ So which was it—a clear and present danger or statistical noise?

For years, federal and professional law enforcement organizations issued guidelines emphasizing severe occupational risks despite minimal—and contested—evidence that any officers had contracted HIV/AIDS. Police work does involve potential exposure to blood and bodily fluids during arrests, searches, and medical emergencies, but the NIJ treated possibilities as inevitabilities, as though every interaction carried equal risk. These institutional communications were structured around the most catastrophic imaginings of HIV/AIDS, particularly scenarios involving blood splattered into officer's mouths or needing to physically restrain someone profusely bleeding (Tomes, 2022). The emphasis on these perceived dangers ultimately fueled the adoption of spit hoods—mesh bags placed over a person's head to prevent them from spitting. ¹¹ First used in the United Kingdom and then adopted in the United States, officers used these bags regularly because spitting was believed to be a transmission route. ¹² Concurrent with broader trends in police militarization, departments had also been equipped with riot gear that came with sturdy face masks, which became increasingly visible in the 1980s and 1990s. Spit hoods and face masks initially emerged in response to concerns about hepatitis C, but they were broadened to encompass HIV/AIDS as the epidemic progressed.

Research on police responses to policy recommendations suggests that departments selectively adopt what resonates with their priorities and operational needs (Ericson and Shearing, 1986; Manning and Hawkins, 2023). Historically, police have been able to exercise such discretion because the policies they institute often hold a kind of legal opacity, or as Roberts (1998) notes, “a vagueness that invites opportunistic and undisciplined enforcement.” Proponents insist that vagueness is essential to policy design for police departments because it creates the conditions under which discretion can become a tool of governance (Goldstein, 1963). Discretion has long been theorized as central to modern policing, granting officers the latitude to interpret situations and make decisions in real time based on whatever information is available. Such information is oftentimes refracted through race, class, age, gender, and demeanor (Davis, 1998). Scholars have consistently shown that discretion exists on a spectrum: at one end, tightly constrained discretion produces relative consistency in enforcement practices; at the other, wide-ranging discretion leads to marked inconsistencies in how the law is applied, and who it applies to (Stranieri et al., 2000).

How does an officer determine if someone is high risk? What visible markers signal infection? When does suspicion justify precaution? “The Law Enforcement Officer and AIDS” never specified any criteria. Instead, the report created a professional obligation to act without standards of action, sustained by the principle that whatever transpired, whatever mistakes were made, institutional procedures would always default to defending officers because their safety was priority, and often, their sole concern. Never mind the fact that determining HIV/AIDS status requires medical testing that officers neither possessed nor were trained to conduct—diagnostic capacities entirely outside the scope of police work, rendering any such assessment fundamentally speculative. Police training manuals, informed by international frameworks like The United Nations Office on Drugs and Crime's Training Manual for Law Enforcement Officials on HIV Service Provision for People Who Inject Drugs and its accompanying case studies, told officers to identify “high-risk individuals” through appearance and behavior: effeminate men, anyone with track marks, sex workers, people “dressed provocatively” (Worden et al., 2024). ¹³ Citron et al. (2005) found that domestic curricula were inconsistent, often medically inaccurate, and steeped in psychiatric logics that characterized gay men as “immature, psychologically maladjusted, or developmentally arrested” well after the American Psychiatric Association removed homosexuality from the Diagnostic and Statistical Manual of Mental Disorders in 1973.

Medical discourse circulated in the worlds of policing but never as systematic knowledge. Infection rates and risk categories functioned as heuristics, as proxies for understanding but never specified how to respond. These contradictions were readily apparent in New York and San Francisco; since these cities were identified as geographic epicenters of the epidemic, these officers were portrayed as maximally vulnerable to infection. These departments were also positioned as national models and received intense media attention. Examining them illustrates how the NIJ report was interpreted at a departmental level, the practices it enabled, and how it reinforced ongoing enforcement activities as necessary components of epidemic control.

Reading bodies: Policing as diagnosis in New York and San Francisco

The diagnostic state, as outlined earlier, emerges when police authority extends into medical knowledge production under three conditions: contested scientific knowledge, federal policy ambiguity, and pre-existing stigmatization. San Francisco and New York offer a natural experiment in how these conditions materialized across contexts. Historians traditionally present these cities as opposing models of epidemic response: San Francisco as the progressive exemplar, characterized by integrated public health systems and the establishment of the world's first HIV/AIDS outpatient clinic, and New York as the punitive counterpoint, where city leadership explicitly instrumentalized the epidemic as the perfect justification for a law-and-order agenda they had long sought to implement (Johnson, 2022). Both cities, I argue, operationalized diagnostic policing as standard practice, differing not in whether officers exercised medical judgment without medical training, but in how each jurisdiction discursively legitimated their authority through dissonant articulations of what Fassin (2012) calls “the moral economy”—San Francisco constructing officers as reluctant health actors, New York casting them as guardians against biological and social contamination.

By the 1980s, the concept of “urban crisis” became inextricably linked to New York, with the city serving as symbolic representation of metropolitan collapse and societal degradation (Sugrue, 2014; Weaver, 2017). Homelessness, crack cocaine proliferation, escalating concerns regarding street-level disorder, and finally HIV/AIDS cohered into a “synergistic plague,” epidemiological, social, and economic disruptions that intensified one another's deleterious effects, generating urban conditions that city officials and civic elites characterized as not only unsustainable but existentially threatening to New York's future (Wallace, 1988). Mayor Edward Koch made his stance clear: “[…] plagues of street dealers [constitute] the transcendent urban problem of our time,” he contended, “a dreary symbol of human failure [evoking] a sense of despair and contempt for the ability of law and government to deliver them from crime and addiction” (Williams, 2021). Koch believed that New York had lost control of its own streets.

Structuralist accounts traced urban crisis to processes of deindustrialization, white flight, municipal fiscal austerity, and the systematic withdrawal of federal support for urban infrastructure and social programming, a systemic dismantling that reconfigured the political economy of American cities (Sugrue, 2014; Weaver, 2017). Cultural deficit theories, championed by neoconservative commentators and increasingly adopted by municipal leadership, located crisis instead within the supposed moral failures of urban populations themselves: a “culture of poverty” characterized by dependency, criminality, and resistance to bourgeois norms of behavior and familial respectability (Sugrue, 2014; Weaver, 2017). Despite their competing modes of social analysis, both narratives authorized a singular conclusion: New York was experiencing unprecedented crisis, and crisis demanded immediate and forceful state action.

State intervention in New York achieved its most concentrated expression through Operation Pressure Point (OPP). Launched in 1984 as a joint collaboration between the New York Police Department and the Drug Enforcement Administration, the operation targeted street-level drug markets across the Lower East Side through zero-tolerance enforcement (Zimmer, 1987). Officers were tasked with descending forcefully on communities already facing the most devastating effects of HIV/AIDS, neighborhoods overwhelmingly composed of Black, Latino, and low-income residents.

Koch asserted that OPP was the model intervention, elevating crime control as an epidemiological necessity. Declaring that “[…] it is clear that Operation Pressure Point is a tremendous success which may show the way for controlling the epidemic of dangerous drugs that threatens our nation,” Koch argued that crack cocaine accelerated the spread of HIV/AIDS through needle sharing and transactional sex work, and positioned the NYPD as the institutional force protecting New York from converging threats of drugs and disease (Williams, 2021). For Koch, more arrests meant fewer infections. OPP formalized the criminal legal system as central to epidemic control, promising that securing the city, its drug markets, and its streets would shield New York from HIV/AIDS.

Community-level political leaders and organizations in other neighborhoods had yet to see firsthand the civil liberties abuses wrought by OPP (Williams, 2021). But in a city reeling from drugs and disease, the overriding concern was managing problems at scale. Confronted with mounting crises, New York expanded OPP, believing it was a solution to the intersecting challenges of drugs, disease, and disorder. Neighborhood leaders supported the expansion of OPP-style initiatives, and in some cases actively demanded them (Williams, 2021). Councilmen approached police chiefs en masse to establish Pressure Point operations in their own districts. Simultaneously, the work of George Kelling and James Q. Wilson on “Broken Windows Policing” circulated throughout the NYPD, resonating deeply with officers who found theoretical validation for aggressive order-maintenance (Williams, 2021).

For marginalized communities in New York, OPP's expansion recalled a fresh history. Barely 15 years had passed since the 1969 Stonewall Inn raid, when police violently cleared the gay club, physically attacking customers, workers, and detaining 13 people whose clothing violated gender norms (Stein, 2019). Though no formal statute specified how many articles of “gender-appropriate” clothing one needed to wear, police enforced an informal “three-article rule,” conducting invasive searches and bathroom checks to determine people's sex. Here, police had repurposed old masquerade laws from the 1800s to criminalize gender nonconformity alongside explicit prohibitions on public same-sex intimacy.

HIV/AIDS witnessed the return of these aggressive tactics. Black and Latino sexual minorities reported experiencing police harassment at double the rate of white LGBTQ individuals (Flavin, 1998; Jonsen et al., 1993; Parker et al., 2018). Their accounts highlighted physical assault and pretextual stops predicated entirely on them “looking queer” (Gill-Peterson, 2013; Schulman, 2021). The scope extended beyond self-identified LGBTQ people; heterosexual individuals also reported harassment because police believed they were gay (Gill-Peterson, 2013; Schulman, 2021). Police stationed themselves outside queer clubs and bars, rendering them breeding grounds for disease that required surveillance (Royles, 2020). Their intensified presence had immediate consequences: needle exchange programs relocated, harm reduction clinics and outreach services closed (Royles, 2020). OPP now undermined the very mechanisms designed to contain the epidemic and save lives. The diagnostic state had returned, now dressed in the language of disease prevention rather than moral enforcement, but its effects on queer and gender-nonconforming people of color remained brutally familiar.

Similar tactics surfaced in San Francisco. In 1984, Mayor Dianne Feinstein sent undercover police “spies” into bathhouses to secretly monitor sexual activity under the pretense of protecting public health (Trout, 2021). When exposed by the San Francisco AIDS Foundation that September, the operation spawned fierce criticism over public health becoming moral policing. But the city would soon offer a justification for its practice, evidence that officers faced real danger. San Francisco offered the first official case of what the NIJ claimed was a serious, widespread concern: an officer who contracted HIV/AIDS on the job. ¹⁴ In 1992, a legal case involving Officer Thomas Cady surfaced, claiming that he contracted HIV during a 1984 arrest of a detainee living with the virus. Years later, Cady and his legal team successfully proved that his diagnosis was the result of occupational exposure.

Captain Mike Hebel, Cady's attorney, believed the decision would establish a precedent for subsequent incidents: “Now a police officer in San Francisco who is subject to a needle stick or any other type of bleeding […] from someone who is suspected of being HIV-infected, has clear authority to point to […] It is no longer murky.” ¹⁵ Law enforcement circles interpreted the ruling as validation of their fears, as concrete acknowledgment of the dangers they faced in the line of duty. But Cady himself viewed the outcome with profound ambivalence. Exhausted by the length of the legal process, he remarked angrily, “I thought they were going to wait until after I died.” ¹⁶ Cady also resigned from the San Francisco police force, saying that legal vindication could not address the deeper institutional toxicity he experienced.

Cady's testimony highlighted systematic misconduct within law enforcement. Despite holding the same badge as his colleagues, Cady reported enduring relentless verbal abuse based on his sexual orientation. He also broke rank to expose the countless scenes of violence he witnessed inside stations and city streets: “The [police] would beat [people] up, call an ambulance and then write out that they were resisting arrest […]. I would tell my superior and nothing would happen.” ¹⁷

Returning to Edwards and Tewksbury's (1996) findings, 75% of police training academies nationwide failed to implement HIV-specific employment policies addressing the rights or occupational protections of seropositive personnel, with administrative officials asserting that existing national guidelines rendered supplementary internal policies redundant. But their institutional rationalization collapses almost entirely when examined against the experiences of officers such as Cady. If three-quarters of police academies determined that vague federal directives obviated the necessity for concrete internal policies, the discretionary space created by NIJ guidance operated without regulation or constraint in terms of implementation. Burke (1994: 193) elaborates on the deep-rooted homophobia structuring police culture, accounting for why officers sustained these behaviors over time and why no antidiscrimination policies were adopted:

[…] the machismo sub-culture of the police and the role of the police as regulators of deviance all make it difficult for the police to accept a non-conformist orientation […] From a police point of view then, homosexuality would appear to represent part of the societal disorder that the police officer has dedicated his or her life to eradicating. Tolerating non-heterosexual officers (let alone condoning their recruitment) therefore would represent the most serious kind of contamination and the worst possible threat to the integrity of the Service.

When officers treated one of their own—someone with the same badge, performing the same duties—with such cruelty because of his sexual orientation, we might at least pose the question: how were those without a badge treated in the streets?

Measured against the actual course of the epidemic, the internal contradictions pervading the NIJ report once again come into sharp relief. The NIJ issued urgent warnings about severe occupational hazards in 1987; roughly seven years passed before the first case of an officer contracting HIV on duty could be substantiated. Meanwhile, officers regularly expressed panic about contact, but engaged in beatings that would necessitate precisely the kind of physical interaction they claimed to fear.

In 1983, before the federal government issued formal official directives on HIV/AIDS and policing, San Francisco Deputy Police Chief James Shannon expressed his primary concern: “We have a large homosexual population […] The police are concerned that they could bring the bug home and their whole family could get AIDS […] fear they might catch the disease from administering first aid, such as mouth-to-mouth resuscitation, to a victim of the disease.” ¹⁸ The San Francisco Police Department immediately began distributing gloves, masks, and protective equipment, warning officers to minimize physical contact with potentially infected populations.

When the NIJ report arrived four years later with the same instructions, it merely reinforced what Shannon and his colleagues had already implemented, providing institutional validation that allowed existing practices to continue uninterrupted. There was a recursive dialectic between the largest police departments and federal policy. These exchanges produced and reproduced shared logics about risk, crisis, and danger, ultimately shaping what the police were prepared to do—and refused to do—in response to HIV/AIDS.

Though New York and San Francisco were ostensibly distinct models, both rested on particular interpretive habits: read bodies, assess risk, act on suspicion. Policing became diagnosis, and diagnosis legitimized punitive action. Preconviction medical abandonment was protocol. More than police overreach or mission creep, what crystallized was a reconfiguration of authority itself, a diagnostic state in which the power to enforce became indistinguishable from the power to define and to know. The virus itself made no moral distinctions; the diagnostic state could not function without them.

Punitive intensification: The long durée of epidemic response

“Why do you have so many condoms?” Most people would say that carrying protection is responsible, a way of caring for your health and the health of others. But after she was handcuffed outside a convenience store, Maria began to see those foil packets differently—a possible conviction. One condom might be for personal use. Two could be a coincidence. But three? Five? Ten? What sort of illicit behavior might she be planning? The answer: nothing at all. The police just thought she “looked the part.”

In their 2012 report, “Sex Workers at Risk: Condoms as Evidence of Prostitution in Four US Cities,” Human Rights Watch interviewed over 300 people, including 200 current and former sex workers. ¹⁹ Many recounted the invasive police interrogations they endured over their possession of contraceptives. Some people stopped carrying protection altogether, deciding they would rather gamble with HIV/AIDS than handcuffs. Staying alive came with no good answers. ²⁰

Public health agencies distributed condoms as prevention tools. Yet police departments across multiple cities—New York, Los Angeles, Washington DC, San Francisco—treated condom possession as evidence of prostitution. Officers conducted searches explicitly seeking contraceptives, confiscated them, and cited quantities as proof of intent to engage in sex work. Even as public discourse proclaims the AIDS crisis “over,” future research must confront its unfinished histories—the enduring practices of criminalization and the reconstituted systems of governance that uphold them.

The institutional and conceptual frameworks developed around HIV/AIDS continue to shape global health interventions. During COVID-19, familiar dynamics recurred: vague CDC enforcement guidance, police tasked with implementing health measures, racialized enforcement of mask requirements and quarantine orders, and force rationalized as disease prevention. Sheriffs invoked constitutional authority to ignore mask directives while conducting mass arrests during the 2020 protests where social distancing was impossible (Demir and Cassino, 2024; Thusi, 2020). These contradictions mirrored the NIJ's position that officers should retain broad authority to protect themselves from occupational hazards, even when it undermined effective public health management.

When crises unfold, institutions often delegate authority to police. Efforts to improve epidemic response must account for police as institutional actors who shape, not merely implement, health policy. Charting a different course means re-examining which institutions manage crises, how discretion functions across populations, and whether public safety must route through criminal legal systems. Without these limits, future epidemic responses could repeat these patterns—treating marginalized communities as threats, criminalizing prevention methods, and viewing punishment as public health. What comes next depends on us. We can build systems of care or continue expanding structures of policing. Committing to alternatives requires more than policy adjustments. It requires us to reimagine what safety means, who deserves it, and how cities should pursue it.