Managing the “hot spots”: Health care, policing, and the governance of poverty in the US

Abstract

Health care systems in the United States are experimenting with a form of surveillance and intervention known as “hot spotting,” which targets high-cost patients—the so-called “super-utilizers” of emergency departments—with intensive health and social services. Through a calculative deployment of resources to the costliest patients, health care hot spotting promises to simultaneously improve population health and decrease financial expenditures on health care for impoverished people. Through an ethnographic investigation of hot spotting’s modes of distribution and its workings in the lives of patients and providers, we find that it targets the same individuals and neighborhoods as the police, who maintain longer-standing practices of hot spotting in zones of racialized urban poverty. This has led to a convergence of caring and punitive strategies of governance. The boundaries between them are shifting as a financialized logic of governance has come to dominate both health and criminal justice. [health care, chronic illness, governance, policing, poverty, United States]

In recent years US health care systems have adopted “hot spotting,” an innovative new strategy in health care delivery. Relying on population health surveillance and intervention, hot spotting identifies the highest-cost patients in the population and directs substantial care and resources toward helping them. The targets of hot spotting are the so-called super-utilizers of health care services, and the intervention provides them with intensive medical care and social services. The aim is to improve health and decrease costs (Finkelstein et al. 2020; Fleming et al. 2019). Super-utilizers are defined according to an economic metric. They are a small segment of patients whose care accounts for a significant proportion of national health care spending, owing to their frequent hospitalizations and trips to the hospital emergency department (ED). In the US, 5 percent of patients account for more than 50 percent of costs (Cohen 2014). Sometimes simply referred to as “the 5 percent,” high-cost patients have become the object of intensive research and policy making intended to reduce their burden on health systems and governments (Noonan 2020).

People who become high-cost patients typically have multiple chronic illnesses and, in many cases, what medical discourse describes as the “comorbidities” of substance use and mental health problems. Beyond these medical classifications, super-utilizers face conditions of structural violence and racialized poverty, primarily homelessness, food insecurity, lack of income, and exposure to violence and trauma. These patients’ repeated hospitalizations and trips to the ED are usually driven by chronic illnesses that they cannot manage because they lack resources and suffer from extreme precarity. In the settings that we studied, the problem of super-utilizers is structural and intractably bound up with a set of social problems connected to concentrated urban poverty.

Hot spotting—the intervention to manage super-utilizers’ costs and health—is not only a form of delivering medical care but also a mode of governing poverty; it is both a strategy of resource distribution and a mode of regulating poor people’s behavior. In bringing ethnographic attention to hot spotting as a technique of governance, we find that it provides lifesaving humanitarian interventions while operating within the racialized structures of violence that produce continual life crises. The institutional rationality of hot spotting and the encounters of care that it produces illustrate the often-contradictory role of medicine in the lives of poor people: both caring and coercive, it intertwines care and violence.

As a governing strategy, hot spotting incorporates previous modes of intervening in the population. In the late 20th century the US and European states expanded their punitive approaches to governing urban poverty. They deployed police, jails, and prisons as their preferred mode of population management (Alexander 2010; Davis 2003; Fassin 2013, 2017; Gilmore 2007; Wacquant 2009). Theorists have interpreted the dramatic expansion of the state’s carceral apparatus and mass incarceration as a response to the social dislocations generated by neoliberal governance: the exclusion of racialized minorities from the wage labor system, deepening poverty and socioeconomic inequalities, and the dismantling of social welfare programs. In this account, forces of economic exclusion and welfare retrenchment generated “surplus populations” (Gilmore 2007) or “problem populations” (Wacquant 2009), which the state increasingly captured through carceral institutions and managed through a lens of racialized criminality. Further, theorists have argued that under a regime of mass incarceration, the carceral system itself has been evacuated of logics of care and rehabilitative aspirations, giving way to a total logic of punishment and incapacitation (Davis 2003; Fassin 2017; Simon 2013). For instance, ethnographic work in housing projects in Paris reveals the forms of imagination and everyday practice through which policing units create a relationship of “war” with poor urban residents, justifying violent tactics, inciting arrests for minor violations, and driving the trend toward mass incarceration (Fassin 2013).

At the same time, anthropologists have documented how medical institutions and care have become central to the state management of poverty, in which such institutions often play an ambiguous role as simultaneously ameliorative and coercive (Bourgois and Schonberg 2009; Garcia 2010; Knight 2015; Stevenson 2014; Sufrin 2017). While on the one hand medical institutions have worked to direct humanitarian resources to situations of extreme need (Farmer 2005), on the other hand they have played a regulative role. They have done so by imposing coercive norms of behavior through conceptions of health and security (Foucault 2009; Ong 2003; Rose 2007) and by demanding compliance with medical care as a prerequisite for ethical subjecthood (Whitmarsh 2013). A range of anthropological work supports the view that medical institutions have extended their role in managing segments of the urban poor in the US, whether they be unemployed, homeless, mentally ill, substance users, or immigrants and ethnic others (Bourgois and Schonberg 2009; Hansen, Bourgois, and Drucker 2014; Knight 2015; Ong 2003). For example, research has described how the medicalization of homelessness has shifted policy makers away from regimes of criminalization and toward ostensibly gentler frames of treatment and reform, yet in doing so they have reproduced disciplining techniques of individual responsibilization (Hanssmann et al. 2021; Willse 2015).

Hot spotting’s object of intervention—the high-cost patient—is a figure of threat and an object of containment. Though emergent, this figure is continuous with previous racialized constructions of high-intensity “users” of state services. In each case, such figures represent a dangerous burden to society and must be restrained and reformed. Scholars have warned that “we may see ‘high utilizer’ join ‘welfare queen’ and ‘gangbanger’ in the pantheon of demonized subjects for ‘endangering our national health care budget and the health of worthy citizens who are not bringing health problems on themselves’” (Ehlers and Krupar 2017, 40). Super-utilizers are figured as uncontrolled overconsumers of health care. Why are health care systems strained? Why does health care spending continue to rise? Because, argue the proponents of hot spotting, a small segment of impoverished people cannot self-regulate. They cannot manage their illnesses or their bodies. According to the prevailing discourse on high-cost patients, the accumulation of chronic illness in certain segments of the population threatens access to care for others who do self-manage and consume responsibly, and it imperils the national economy as a whole. “Why do we need health care hot spotting?” asked an affiliate from the Camden Coalition, the nonprofit organization that pioneered health care hot spotting. “Perhaps the most compelling reason is the now well-documented percentage of our economy that is consumed by health care expenditures. We are now spending roughly 17 percent of our GDP and a fifth of all personal income on health care.”¹ Reducing the economic burden of severely chronically ill people living in poverty is the central rationale for expanding care and humanitarian intervention for some of the sickest and most marginalized people.

Our analysis begins with an overview of our methods and a brief genealogy of health care hot spotting. We describe how it arose from and overlaps with policing practices in US cities. We then present ethnographic findings centered on a person we call Randall, who received the hot-spotting intervention.² We argue that hot spotting’s economic and calculative rationality cuts across criminalizing and medicalizing approaches to managing poverty and reveals shifting fault lines among these domains. We conclude by suggesting that economic investment and return are becoming a reigning logic in the governance of poverty, generating hot spots as sites of interest for both policing and health care and decentering normative assessments of deviance, illness, and social problems.

Ethnography of hot spotting and participants

Research for this article was conducted as part of a team-based ethnographic project based at two public hospitals in California cities. From 2015 to 2017 two ethnographers were located at each hospital, where they observed hot-spotting teams. They also interviewed patients and providers (N = 123) at various health care organizations and associated community services. Fieldwork was conducted at one of the sites by the first author (MF), who is white. MF also drew from the larger ethnography to compare related cases and validate interpretations.

Health care hot spotting goes by a variety of names, including super-utilizer programs and complex care management. The latter was generally the term used among the health care providers at our sites, although some hospital leaders emphasized the terms hot spotting or hot spotters. We use hot spotting to emphasize the continuities of the practices we observed with their histories, as described in the next section. The teams making this intervention consist of nurses, social workers, community outreach workers, data analysts, and primary care physicians. They seek to provide targeted individuals with health care services, to coordinate a range of social services for them, to improve their “self-management” skills, and to reduce their hospital visits. We shadowed team members as they worked in the hospital, interacted with patients, visited patients in the community, and coordinated with shelters and social services. We also spent time with patients at their homes; places where they slept on the street; at detox facilities and mental health facilities; and at medical, legal, or social services appointments.

In the hot-spotting program where this research was conducted, the targeted patients frequented hospitals primarily to treat heart failure, diabetes, end-stage renal disease, and chronic pulmonary disorders. See Table 1 for patient characteristics.

Table 1.

Patient characteristics from the California-based health care hot-spotting program where primary fieldwork took place. The left-hand column represents demographics for all patients (N = 140) enrolled in the program. The right-hand column represents data from a medical chart review, listing health conditions for a representative subset of study participants (n = 41) at the field site.

Male (63%)	70% had four or more chronic-disease diagnoses
Black (55%)	Mean of 16 prescriptions
Latinx (16%)	58% diagnosed with depression
White (16%)	72% diagnosed with substance use disorder
Asian/Pacific Islander (11%)	83% had been homeless at least once
Other (2%)

Hot spotting as a technique of governance

Hot spotting is a technique of surveillance and intervention, one that identifies areas of intensity within a biopolitical field—“hot spots”—and then directs substantial resources to them, in a targeted way, to modulate the field or population as a whole. While police hot spotting dispatches law enforcement to “high-crime” areas, health hot spotting distributes care to a population’s most expensive patients. Our ethnography reveals that police and medical hot spotting have been deployed within the same zones of racialized urban poverty, often targeting the same individuals and neighborhoods. The health hot spots where many of our participants lived were already problematized as criminal hot spots. Not only were they living with a multitude of unmanaged chronic diseases, but many also had concurrent histories of being incarcerated and targeted for policing.

Hot spotting draws on the history of disease mapping, which began in late 18th and early 19th centuries in European states. National statistics made it possible to bring together population data and spatial measures to identify zones of elevated disease incidence and prevalence (Foucault 2009; Hacking 1990). The spatial mapping of disease was bound up with the new modes of modern power that developed increasingly fine-grained techniques of observing and controlling populations. According to Michel Foucault (1977), for instance, disciplinary power—state techniques of confining and controlling individual bodies—developed alongside the conceptual imaginaries of the plague as a particular kind of disorder that was located in towns and neighborhoods and that required people to spatially isolate themselves and to present their bodies and homes for inspection. Smallpox, by contrast, was managed through techniques of biopower, which treated the disease as an endemic feature of the population, one to be managed via statistical knowledge of aggregate life (Foucault 2009). Foucauldian genealogies of state power posit a close connection between pastoral and disciplinary modes of intervention, in which states wield the capacity to “make live” as a means of containing and controlling the population.

Over the 19th and 20th centuries, hot spotting developed as a mode of surveillance and intervention. This new practice focused on identifying and intervening in spatiotemporally bounded zones of illness. It was developed largely in the context of fighting infectious disease and in connection to a range of disciplinary and biopolitical techniques aimed at containing outbreaks. Medical hot spotting’s spatial units have varied widely. They have included neighborhoods, cities, or entire regions, such as West Africa or Southeast Asia, which are represented as global hot spots of disease emergence (Lessler et al. 2017).

In the 1980s and 1990s hot spotting was also developed by US policing agencies as a data-driven approach to law enforcement. This approach generates maps of crime hot spots to identify and contain people thought to be committing offenses in the areas with the highest crime rates. Under Commissioner William Bratton, the New York City Police Department initiated the most prominent police hot-spotting program, known as CompStat. These efforts drew on the much-critiqued “broken-windows theory” of policing, in which police intensively manage minor crimes in “disordered” neighborhoods. The broken-windows thesis posits that serious crimes are precipitated by minor violations, such as loitering, writing graffiti, public drinking, and jumping turnstiles; these create an environment of “urban decay” that authorizes violent crime (Harcourt 2001). Hot spotting gave the city government a politically neutral, technological justification to hyperpolice poor, largely Black and Latinx neighborhoods. It helped propel the trend toward mass incarceration, and it justified aggressive stop-and-frisk practices. During this period, amid its milieu of new surveillance technologies and intensive “order-maintenance policing,” there emerged the infamous figure of the “superpredator,” typically a young Black man identified as a high risk for committing significant crimes and requiring extensive surveillance and carceral capture.

Hot spotting first made the jump from policing to health care in Camden, New Jersey, around 2009. Politicians and media have long held up Camden as a paradigmatic case of America’s “urban problem.” It was once a great industrial city, whose shipbuilding and manufacturing industries gave rise to a large labor market. Today, it is one of the nation’s poorest cities, and it is often featured at the top of lists ranking cities by violent crime. The city has been constructed in the public imaginary as “the poster child of postindustrial decay” (Hedges 2010). These representations of disorder also render the city as a site for experimentation in governance, including for-profit infrastructure projects and the emergency management of budget crises (via “takeovers” of city government by state-appointed boards from 2002 to 2010). The city continues to be a site for testing new police surveillance methods, such as the use of video and microphones for mapping gunshots (Wiig 2018).

A physician named Jeffrey Brenner is widely credited with developing the health care hot-spotting method. Brenner, a white 31-year-old from the suburbs of Philadelphia, was working at a family medicine practice in Camden when he was asked to be a “citizen member” of a police reform commission. Since he worked on the reform commission, Brenner learned about the use of police hot spotting in New York City and created his own crime map of Camden based on data he pulled from EDs at local hospitals. After the police union rejected the commission’s proposal to enact hot spotting, Brenner had the insight to repurpose hot spotting for health care delivery. That is, in rejecting Brenner’s technique, the police inadvertently led Brenner to push for the health care system’s wholesale adoption of hot spotting and its data-intensive mapping practices.

“So he took what he learned from police reform and tried a Compstat approach to the city’s health-care performance—a Healthstat, so to speak” (Gawande 2011). Brenner founded the Camden Coalition, a nonprofit network of service providers that has received millions of dollars in federal funding, to develop and implement hot spotting. Crime hot spots, Brenner found, are also hot spots of illness and, especially, health care spending. Structural violence generates not only what policing identifies as criminality but also what a doctor comes to see as an extremely disordered use of medical services. Here, he could identify the 5 percent of patients driving more than half of spending on the health care safety net. In Camden, Brenner’s coalition found that 13 percent of the population accounted for 80 percent of the city’s health care costs.³ Moreover, the costliest 1 percent of patients accounted for 29 percent of spending (Gross et al. 2013).

Health care hot spotting gained wider prominence in the medical field with the publication of Atul Gawande’s (2011) New Yorker article “The Hot Spotters,” featuring Brenner’s work in Camden. Since then, health hot-spotting programs have spread to public hospitals throughout the country. This has happened in large part because new models of health care financing, implemented by the Affordable Care Act of 2010, have incentivized health care systems to invest in cost-saving interventions.

The programs we studied in California were explicitly modeled on Camden’s hot-spotting model. The medical director of one of the programs said that, after reading Gawande’s article, he thought, “We need a hot-spotting program!” As the program got started, he consulted regularly with Brenner and the Camden Coalition.

In Camden and elsewhere, health care hot spotting has expanded from its original meaning as a form of spatial mapping to encompass a mapping of clusters of intensity within a biopolitical field, clusters that need not exist in strictly spatial terms. This can be seen in a video presentation in which the Camden Coalition affiliates described how to use clustered heat maps to conceptualize hot spots within “virtual” space (see Figure 1). As the video narrator says,

Figure 1.

Heat map representing hot spots as “virtual neighborhoods,” which are composed of many nonspatial forms of data, including costs. (By permission of the Camden Coalition of Healthcare Providers, 2014)

A hot spot map . . . creates virtual “neighborhoods” of hierarchical similarity between them based on the values in a large number of variables [including cost]. . . . Hot spotting finds clusters of extremes with a comprehensible coherence where intervention can have a disproportionate positive impact on a subpopulation of manageable size.⁴

The demand for hot spotting—and cost reduction, or debt mitigation—has given rise to a robust private-sector market offering digital tools to surveil the population and identify targets for care. These digital tools use big data sets and predictive analytics to identify targets who are most likely to be affected by the intervention in ways that reduce costs. Thus, these tools offer the promise of maximizing the impact of narrow, bounded investments in the population, and they use algorithmic prediction technologies that resemble those driving the burgeoning market for predictive policing tools.⁵

As hot spotting moves across the domains of policing and medical care, we can understand it as a mobile and mutating technique of governing—one that retains a core commitment to targeting high-risk hot spots as a mode of managing populations, yet one that is also flexible about the problems it addresses and the interventions it deploys. Camden and its urban crises functioned as a laboratory for developing governing modes that proliferate to urban centers throughout the US, including in California. As Brenner said, “The innovation that we are figuring out here in Camden is going to mean better health for everyone.”⁶

Encountering the hot spots

On a late Thursday afternoon, I (MF) accompanied Philip, a community outreach worker on the hot-spotting team, as he visited a patient at a single-room occupancy (SRO) hotel. I had met the patient, Randall, a few weeks earlier, after he was admitted to the ED. Randall was a 54-year-old African American man whose main health problem, at this point, was congestive heart failure, although he also had diagnoses of uncontrolled hypertension, schizophrenia, and “active substance use disorder.” When I had first met Randall, he continually gasped for air and seemed to drift in and out of consciousness as we spoke. Now, however, his health had stabilized, and he seemed to be doing much better.

During interviews, I learned that Randall had spent his adult life in and out of prison, serving one long stint of nearly 16 years. In our broader ethnographic study, we learned that many people who become “super-utilizers” at these particular hospitals also had a long history of encounters with the carceral system. Among people who in the last 12 months had visited the ED at least four times or been hospitalized at least three times, about 71 percent had been convicted of crimes or had recently spent time in jail, according to one study of a large Medicaid population (Vickery et al. 2018). We’ll return to this point later.

After the last time Randall was released from prison, he was homeless. His health declined precipitously when he developed heart failure. He ended up in the ED 30 times in one year. As Randall recounted,

I was stressed and I was sick, and I was denied medical on the streets [because he lacked insurance]. I was getting weaker and weaker living outside in the street, sleeping in abandoned cars. It was no good for my health. I didn’t know at the time that I had a health problem. I just started getting weak, and it didn’t make sense for me to go nowhere because I wasn’t going to get no help in a clinic, so I went to the emergency room at [the county hospital], because they don’t deny you.

During this time, he also began hearing voices and had several involuntary holds at the psychiatric hospital. Each time, he was released to the street with a long list of medications. Without a place to live, Randall frequently lost his medication and had difficulty getting them refilled.

The extremely high costs of Randall’s health care triggered a red flag at the hospital. This put him on the radar of the hot-spotting program, which later made him eligible for additional support, including housing.⁷ The hot-spotting team arranged for Randall to be discharged straight to the SRO rather than to the streets. Randall was provided with what they refer to as “wraparound care,” meaning care that is attentive to a patient’s medical, social, and psychological needs. One of the team nurses told me Randall had a “well-defined mental health history which includes a diagnosis of schizophrenia and severe substance abuse disorder, but he is not interested in working with [us] on any of these problems.” He was interested only in working on his “really bad CHF [congestive heart failure].” The first time they recommended psychiatric treatment, Randall “went ballistic” and refused. Within a month of being at the SRO, Randall’s heart condition was significantly better. The first time I saw him at the SRO, he was walking around briskly, doing laundry down the hall. In the hospital his body had been very swollen from the poorly controlled CHF, but now he looked thinner. “All my medication is perfect. I am feeling much better,” Randall said, making the OK sign with his hand. “I am so much better. God healed my body.”

That day we stood in Randall’s room chatting with Philip, the community outreach worker. He and Randall discussed a plan to register Randall in a food bank program. Philip was instructing Randall on how to notify the food bank of his needs and how to take the bus there. As they spoke, Randall became increasingly angry that he would need to take a 30-minute bus ride each way. “But this is medical!” he loudly exclaimed, describing his need for groceries. He had a pained look on his face. He was trying to convince Philip that his hunger pangs should be regarded as a medical issue. He was doing this because Philip had a stack of taxi vouchers in his bag, but he could give them to patients only for medical appointments. Randall wanted one to make the trip to the food bank easier. Philip explained that Randall would have to settle for a bus ticket. Randall replied, this time nearly shouting, “I know you got them, come on! This is medical. This is my stomach that needs this!” Philip and the other providers on the hot-spotting team certainly agreed with Randall that access to food is “medical.” Randall’s multiple medical conditions required him to take upwards of 20 pills a day, and his inconsistent access to food limited his ability to take these medications, since some needed to be taken with food. When he was hungry, Randall would lose track of his complex medication schedule and miss doses. Just a couple of weeks before, Randall was having trouble breathing, and his poorly controlled heart failure was making his legs swell. He told Philip on the phone that he was “starving” and that if his breathing got any worse, he was going to call 911. Philip went to the food bank for Randall and brought him several bags of canned foods, bread, pasta, and a few fruits and vegetables. When Randall was first enrolled in the program, before he was intensively working with the team, he was carrying around a plastic bag full of about 30 prescription pill bottles collected from his various hospital visits. This predicament of going hungry while being extensively medicated is common among so-called super-utilizers, who are often caught in a cycle of extreme deprivation. For them, emergency medical care is the only option for crisis management.

Philip refused to give Randall the taxi voucher because, according to the terms of the hot-spotting intervention, Randall had to learn how to sustain himself—even though he had minimal resources and even if it took a lot of resources to help teach him. At issue was not the cost of the vouchers but the possibility that Randall might become dependent on the team, which Philip and the other providers wanted to avoid. The success of hot spotting hinges on teaching chronically ill people how to self-manage with minimal resources, as well as improving their access to basic resources.

In their attempts to stabilize patients’ health and life conditions, the hot-spotting teams in this study helped patients find access to food, housing, income, and other basic resources. This was a core strategy of their care practices. They arranged for transportation, accompanied Randall to his many medical appointments, and worked with him extensively so that he understood the information and directives given by doctors. The team guided Randall through the long legal process of applying to the Social Security Income (SSI) program on the basis of his medical diagnoses, a process that included sitting with him during three-hour waits at the Social Security office. They had frequent contact with Randall, visiting and calling him on the phone, first at the places where he slept on the street and then at the SRO once they got him housed. Under a rubric of providing medical care, these health care providers took into account a range of socioeconomic conditions generating patients’ crises. Pursuing a strategy of managing chronic illness and reducing hospital costs, they sought to mitigate the worst effects of poverty.

Medicalized public assistance

Anthropologists have argued that, in the aftermath of US welfare reform in the 1990s, biomedicine dramatically expanded its role in managing the lives of poor people (Hansen, Bourgois, and Drucker 2014). In this account the state medicalized poverty by making diagnoses of chronic illness and disability virtually the only avenue for poor people to access public benefits. While social welfare has contracted, more people began receiving federal disability income on the basis of approved medical diagnoses, suggesting that people who previously received welfare cash assistance transferred to disability benefits (Wong 2016). Through ethnographic case studies of impoverished people in New York City, Hansen and colleagues demonstrated that the medicalization of poverty has produced the “permanently disabled pathological patient” as a new category of stigmatized subject (Hansen, Bourgois, and Drucker 2014). The super-utilizer represents a new iteration of this stigmatized subject, one who is not only permanently disabled and pathologically dependent but who also, in overconsuming care, acutely threatens health care systems and the economy. While the medicalization of poverty and the expansion of disability benefits was accompanied by efforts to exclude “benefits cheaters” (Hansen, Bourgois, and Drucker 2014, 82), who represent an unnecessary cost burden to the state, hot spotting by contrast seeks to reduce the state’s cost burden by including particular individuals in the targeted expansion of services.

After a lifetime of being subjected to severely life-limiting circumstances and being excluded from medical care, other than at the ED, patients like Randall now find themselves given access to life-sustaining resources, including housing, food, transportation, and mental health care. Randall became eligible for such assistance because he was identified as part of a health care hot spot—a cluster of unreasonably expensive, uncontrolled chronic disease.

A central aim of the hot-spotting intervention was to engage people in regular primary care, which involved an often-fraught process of convincing them to accept the biomedical narrative of their symptoms and to buy in to the normative expectations of disease self-management (Fleming et al. 2021). Randall, like many patients, sometimes resisted the demands of living under chronic disease diagnosis, which required that he adhere to medications, follow diet restrictions, and moderate substance use. Other times, he used his heart failure diagnosis to argue for more resources.

When Randall demanded taxi vouchers to get his groceries, he was clearly appealing to the logic of medicalized public assistance by claiming that his need for food—his hunger—and his need for transportation were properly medical issues. He was, in fact, well aware that his housing and survival still depended on his diagnoses and on his appeal to institutional resources based on his medical needs. For instance, when a team nurse visited to check on his medication adherence, Randall insisted on discussing his precarious housing situation. “If I end up homeless and get sick again,” he said, “that will take my life.” In later interviews he described how he learned to engage the medical system as a survival tactic. “I may not look it, but I’m educated,” Randall sometimes said. Hospital staff often remarked that some patients who frequent the ED to access resources they want or need are highly strategic in how they present their health complaints. The team recognized Randall’s skill at navigating bureaucratic structures, including the health care and legal systems. They were impressed by his ability to advocate for his needs, to make use of legal help for his SSI appeals, and to deal with the hostile management company at his SRO. “He’s no dummy. Not by a long shot,” an outreach worker said.

Yet they also saw him as too narrowly focused on his financial goals and as sometimes overly demanding. Over time, team members began to comment that Randall withheld certain information (largely about mental health symptoms and substance use). He did so, they said, so he could focus on his two main priorities: improving his heart condition and becoming financially secure through disability benefits. The team wanted to work with Randall on his “behavioral health” issues, but he was uninterested and stayed largely focused on getting his SSI application approved. As a social worker said, “A big motivation for being in this [continuing to collaborate with the team] was the money piece. Without that, he is not engaged.” Further, they often said Randall’s style of requesting resources reflected an attitude of entitlement. “I don’t know where Randall’s sense of entitlement came from,” a community outreach worker told me. On a separate occasion, a nurse said, “He’s really pretty committed to his crack cocaine. He’s getting kind of bossy with people. ‘Get me a this or that.’ Arrogant, entitled.”

To address Randall’s behaviors and attitude, the hot-spotting team members adopted an “individual-reparative strategy” (Hanssmann et al., 2021; Willse 2015). But they supplemented this strategy with one that understood Randall as part of a population situated within socioeconomic structures. When the team first began meeting patients in their communities, they often reported being “shocked” and overcome by the degree of deprivation, violence, and disorder the patients faced. By understanding individual patients as part of a population subject to socioeconomic structures, the team members viewed the patients’ health conditions as an injustice caused by poverty and exclusion—as a social wrong that they could begin to address by gathering resources for the patient under the sign of healing.

A physician described the patients as “the most marginalized of the marginalized.” A team manager said,

Mostly they’re medically fragile and additionally fragile because they’re living in the tiny branches, you know. They just don’t have good support structures. [. . .] With our patients I think I’ve learned more about racism. It’s eye-opening. It’s really eye-opening. They get worse care, worse services. Lots of them not seeing a doctor for years. They don’t have the trust, and they have no reason to have it.

Minimalist intervention

To understand how hot spotting modulates the costs of managing impoverished people, we must consider the legal provisions that define the bare minimum of humanitarian provisioning in the US. Under a 1986 federal law titled the Emergency Medical Treatment and Labor Act (EMTALA), hospitals are required to treat people who come to the ED presenting with an emergency medical condition, defined as a condition that, without treatment, will likely lead to serious impairment or death. The hospital is required to examine anyone who walks through the door, and if the person is acutely ill enough, it must provide care regardless their ability to pay. While the state may abandon certain people to live without income, shelter, or regular health care, they are legally entitled to resources—regardless of their economic means and citizenship status—if they are facing an immediate risk of death, as defined by legally codified guidelines.

EMTALA was intended to prevent EDs from denying or delaying care for impoverished people and to combat practices such as the “wallet biopsy,” in which hospitals assess patients’ ability to pay before providing lifesaving care, or “patient dumping,” in which medically unstable patients are discharged or transferred (Friedman 2011). The law marks the redline of what society deems morally acceptable, keeping people from being denied care and left outside to die or give birth on the street (Hoffman 2006). In this sense, EMTALA is a minimalist provision, creating a right to health care under only the most dire circumstances. It has turned EDs into a “reluctant safety net” (Hoffman 2006) or contemporary “almshouse” (Malone 1998). The ED is, in many instances, the only institution that does not turn away people who are in dire need.

A humanitarian focus on meeting only the most basic needs for survival may work to maintain a population just above the threshold of death. It thereby defers actions that would lead to a greater flourishing, constituting a minimalist biopolitics (Redfield 2005). The condition of being a super-utilizer arises from being caught in structural violences that continually push certain people to the threshold of death, where they then receive acute intervention, which repairs them to only a minimal level of survival. They are then put back into the conditions of violence, where the cycle begins again. Hot spotting, then, is a mode of interrupting this cycle through calculative governance. It deploys care and humanitarian resources to support the life chances of sick and marginalized people in limited and targeted ways.

EMTALA is one of the largest federal mandates to provide services to have gone unfunded (Friedman 2011); costs instead fall on states and local health care systems. A significant driver of financial costs for maintaining this minimalist intervention are people with multiple chronic illnesses who repeatedly come close to death when those illnesses are acutely exacerbated. Hot spots are defined medically but produced economically through chronic diseases that rage out of control, raising the question of how states and health care organizations can change peoples’ health behaviors to contain costs and inculcate disease self-management.

Criminalized management

In the latter decades of the 20th century states ramped up the punitive management of poverty. Criminalization, above all, marked conditions of extreme socioeconomic precarity—like those experienced by patients such as Randall. The police would now coercively manage those who were unemployed, homeless, or struggling with substance use or mental health problems (Herring, Yarbough, and Alatorre 2020; Wacquant 2009). Many of the high-cost patients targeted for health care hot spotting live in the same zones of racialized poverty targeted by police hot spotting, and they have extensive histories of carceral involvement. For instance, by virtue of experiencing homelessness and having a substance use problem, some patients were repeatedly arrested. Others experienced lengthy prison sentences earlier in their lives and continued to struggle with unemployment and poverty after their release.

With the rise of mass incarceration, the carceral system largely abandoned its rehabilitative ambitions and therapeutic approaches (Simon 2013). Today, contact with the criminal justice system is itself a risk factor for poor health (Golembeski and Fullilove 2005). For the participants in our ethnography, encounters with police and jails often meant a step backward in managing their illness. In particular, stays in the county jail disrupted much of their care, since patients often lost their medications, missed appointments, or had substance use relapses after being discharged.

Randall spent long stretches of time in prison, and his criminal record made him vulnerable to rearrest and reincarceration. This put him at a distinct disadvantage when the SRO management company used the threat of arrest to evict him from the hotel. Randall was evicted after he was accused of stealing an internet router and cables from the building’s lobby. The management claimed to have security footage showing Randall taking the items, but Randall denied that it was him. Randall believed the SRO evicted him because the hot-spotting team had submitted a formal complaint about bedbugs in his room, forcing the SRO to pay for fumigation. At one point, Randall’s body was covered with bedbug bites, which turned into open sores. During a visit to his SRO, he showed me the sores on his shoulder and back and described how he had been pouring bleach on the sores because “bedbugs don’t like bleach, but it burns really bad.” He described seeing bugs coming out of the sores when he did this.⁸ After receiving the letter from the hot-spotting team confirming that Randall’s room was infested with bedbugs, the SRO manager asked Randall to sign a paper saying that the management company was not responsible for fixing the problem. In Randall’s account, after he refused to sign, they accused him of stealing the router. He said, “They didn’t want to do none of that, and now they don’t want me here anymore.”

After his eviction, Randall moved in with a family member in a low-rise apartment complex on Nolan Avenue. Another hot-spotting patient already lived in the complex. Several other patients lived within a block of Randall’s new home. When Philip learned that Randall moved to Nolan Avenue, he was immediately concerned. He called Nolan Avenue “crack kingdom” and described it as a major center for cocaine sales and sex work.

During a meeting in a small conference room at the hospital, the team discussed this change in Randall’s living situation. A nurse expressed great concern and said the apartment complex, along with several others near Nolan Avenue, had many residents who were “heavily involved in drugs,” including some very sick patients. She asked, “Is there a way that we can notify the landlords about the drug activity that is going on in their apartment buildings? Maybe they can do something about it.”

The team manager responded, “We don’t want to go the legal route [the police], but we are concerned about what is happening there. A lot of our patients are coming from right there. It’s a public health issue.” They considered the possibility of contacting the director of public health for the county and discussed how they might word the email.

An outreach worker who twice visited Randall at his new place said that, both times, he was “standing outside on the corner with a bunch of guys. He was very down and depressed. He lives with four or five people, and they are all doing evil things, including him.” He said the roommates were selling crack cocaine out of the apartment and Randall appeared to be high on both occasions. In later discussions the team members told me they were wary of going to Nolan Avenue to do outreach. They felt uneasy among the neighborhood’s gangs and drug dealers.

They were also concerned about encounters with the police, who intensively surveilled and managed the neighborhood. As a manager said, “Outreach workers get pulled over by the cops while they’re trying to outreach. It happens a lot. It’s actually very scary because of what’s been happening with the shootings. [. . .] The police are an issue. I’m sure every single one of them has been pulled over.” The hot-spotting program deliberately employed outreach workers from the same neighborhoods as the patients. This shared background was the basis for connecting with patients and establishing trust relationships. “Most of the outreach workers,” said a team member, “they’re young African American men who lived in the community.” But being African American and from the community also made them subject to police surveillance and management. Outreach workers were pulled over by the police several times traveling specifically to the area around Nolan Avenue. On one occasion, a police officer approached the outreach worker’s car with his gun unholstered. In response the manager ordered large magnets displaying the name of the county health care system. The outreach workers could put them on their vehicles in hope that the police would be less likely to target them.

After the team spent nearly a year trying to persuade Randall to see a psychiatrist, he agreed to an assessment. The psychiatrist prescribed him the antipsychotic Risperadone and other medications. Shortly after, Randall’s psychiatrist contacted the hot-spotting team because Randall did not show up to follow-up appointments. The psychiatrist suggested that the team do outreach. Philip checked on Randall, and I accompanied him. We drove about 15 minutes from the hospital and found him standing in the parking lot of his low-rise apartment complex. Randall had lost a lot of weight. His face was gaunt, and it took me a moment to recognize him. He didn’t say much at first. Just that he was tired. He had his own room in the apartment, but, he said, “I can’t rest.” Philip asked if Randall had been taking his psych meds. Randall said he didn’t like how the meds made him feel. The pills made him “rock back and forth all day long,” and when he stopped taking them, the rocking stopped. Randall’s voice began to sound more upset, and he said he would not go back to the psychiatrist because “they want me to take medicine!”

Standing in the parking lot, Philip asked Randall, “Why don’t you have a phone anymore?” Philip explained that he could easily get one from the “Obama phone” kiosk outside the medical center. “I don’t want another phone,” Randall said. The last one, he explained, was “bugged,” and people were listening to him. “It’s better to be unplugged,” he said, because it’s easier to “keep people from spying on me.” Philip said nothing in response. He gave Randall a piece of paper with his next appointments written on it and a handful of bus tickets.

Histories of racist housing policy have resulted in concentrations of poverty within neighborhoods such as Nolan Avenue (Rothstein 2017). Across policing and medicine, these zones and the people residing within them become acute problems for governance through surveillance practices. Health care hot spotting identifies the most expensive patients in the population, thereby bringing attention to the same apartment complexes, city blocks, and neighborhoods where intensive “broken-windows policing” takes place. In dispatching health care into the hot spots, hospital workers were themselves subject to police surveillance and management. As we discuss below, however, medicalized and criminalized forms of managing hot spots were not strictly opposed. Their entanglements may reveal how new arrangements of poverty management are taking shape today.

Chronic disease, criminality, behavior

Among police, hot spots are constructed as zones of exception. In these zones, officers have expanded jurisdiction for search, seizure, and surveillance, and this is justified by the area’s high crime rate. Similarly, identifying hot spots of high-cost illness justifies medical institutions’ exceptional spending on documenting and intervening in patients’ lives. In addition to justifying exceptional effort and expenditure, hot spots are sites of exception where the normal boundaries between care and coercion may become blurred, and where care and punishment are deployed in a more deeply imbricated manner.

The next time I saw Randall, he was sitting upright in a hospital bed in the ED, with one wrist and one ankle each handcuffed to the bed. A police officer stood next to the bed, arms crossed. Randall was close to tears. He had been arrested for possession of cocaine and had spent the previous month in county jail awaiting a court date. Earlier that day, Randall was taken from the jail to the courthouse for a hearing, where he fell ill. His chest hurt, and he felt as though he could not breath. His public defender called an ambulance, and he was taken to the hospital. Until the episode at the courthouse, Randall had stayed out of the hospital for months.

After Randall’s arrest, his health care team had been in frequent contact with his public defender, trying to devise a strategy to keep him out of prison. Randall’s public defender wanted to leverage the support of the care team to argue for a lighter sentence owing to his severe illness. The team attested that Randall would be participating in an outpatient drug-treatment program. The judge set an exceedingly high bail—$115,000—and the team was not optimistic.

A few days later Randall was discharged from the hospital and sent back to jail. His court hearing was rescheduled for the next week. The social worker from the team accompanied Randall to his court hearing. “He got five years’ probation. He was not happy,” she reported. Given the high bail, the rest of the team was surprised that the judge let him off without prison time. In the meantime Randall skipped his appointment for probation orientation. The social worker said, “I made him call his probation officer and apologize for missing the orientation. And he was not happy about that.”

“The probation officer was not happy?” the team manager asked.

“Randall was not happy about me making him do that,” the social worker replied.

That same week, Randall’s SSI had been approved, and he was slated to receive a $15,000 back payment for the time that his application was held up in appeals. As one of the team members told me, “This will be both a blessing and a curse,” because, in his view, Randall was likely to spend the money on a drug binge. In a meeting the social worker said, referring to the money, “I think it’ll be gone quickly. And I think that he will be picked up [arrested] again. There might already be a warrant out for him for missing the probation orientation last week.” They speculated that he would end up in jail or sick in the hospital. They thought death was a very likely outcome of an inevitable drug binge.

One of the community outreach workers said,

He also really needs to go to [the drug-treatment program]. He is still using. I went out there to meet with him. We had this really cool plan to go to [the program] for drug treatment, but now he is saying I’m not dealing with [the program]. I don’t have time for that. I need to take care of my SSI.

The team decided to collaborate with Randall’s probation officer to come up with a plan that would empower a third party to take control of his finances (e.g., that he be assigned an institutional “payee”) and to compel him to participate in the drug-treatment program. The team manager explained,

So we went and visited his probation officer and learned about all these resources that he has through probation [. . .] housing resources, substance abuse treatment paid for front to back. It’s almost good that he has this probation. We can also ask them to mandate that he has a payee. I think that these folks can be our friends. They have the power to help get him into treatment. They have some control. I don’t mean to police him and take away his control. But what is going to happen next? Because we think that he is going to spend all the money on drugs.

The hot-spotting team interjected a care logic into the legal proceedings to argue that Randall would be better off receiving medical treatment than punishment. Conversely, the team also made use of the probation officer’s capacity for coercive and punitive control to force Randall to accept third-party management of his finances and to try to get him into drug treatment. This collaboration was justified, they said, by Randall’s extremely high risk of being hospitalized and of having poor health outcomes. Hot spotting, then, is not a simple transposition of carceral logics into a medical arena, but rather the formation of a medico-legal field of intervention, centered on behavior and its economic costs.

This attention to behavior and behavioral aberrance is common to both health care workers and the police in their practice of hot spotting. Both policing and chronic disease management converge around what we might call a philosophy of behaviorism, a philosophy in which behavior becomes a locus of knowledge and discipline (Foucault 1977; Harcourt 2001). As the designation “chronic” suggests, chronic diseases are defined through their temporality: they are bodily disorders that persist through time—for months, years, and in most cases, the rest of one’s life. They are not cured; rather, they are managed. People living with chronic disease diagnoses are subject to an intensified demand for self-management. They are expected to expend great energy to continually attend to and manage their bodies. Seen through the medical lens, the development of uncontrolled chronic illness is a problem of aberrant self-management and unhealthy lifestyle choices, and it is therefore seen as a kind of behavioral deviance. In this way, health hot spots are also sites of dysfunctional self-regulation and aberrant behavior; they are dysregulated bodies registered in financial terms.

Within this flattened field of behavioral intervention, readmission to the hospital becomes increasingly blurred with carceral recidivism. For instance, in a conversation with a patient who had several recent arrests for substance use and disorderly conduct, I asked her what she’d been up to. “Just trying to be good,” she said. I asked what “being good” meant, and she replied, “Staying out of the hospital,” which in her case entailed not only keeping her blood sugar under control and going to dialysis appointments, but also refraining from using substances.

Shortly after his SSI was approved, Randall stopped talking to the team members, and I never saw him again. According to Philip, Randall “fired” them, saying, “You do nothing for me. I don’t need you.” Randall planned to come for one last appointment with the primary care doctor associated with the hot-spotting team, so he could “tell Dr. Thomas to his face, ‘You’ve done nothing for me.’” Randall believed the doctor wrote a letter to the probation officer saying he needed a payee and blamed him for losing control over his SSI income. (The team members were uncertain if Dr. Thomas was responsible for the letter.) Randall angrily complained to Philip about having a payee. According to Philip, Randall had added, “Maybe it’s good I have a payee.’” Recounting this, Philip said, “Maybe Randall has a schizoid personality. It’s like he is two different people. He starts off all negative about something and then changes his mind completely.” Philip reflected that, while Randall continues to resist psychiatric treatment, “he is pretty much doing everything else independently.” He manages his heart medications and goes to medical appointments. In the 12 months after that last meeting, Randall had visited the ED twice, both times for shortness of breath, and he was treated and discharged. This was a vast improvement from the previous year, when he had 30 ED visits and long in-patient stays. From the perspective of calculative management, Randall’s stabilized heart condition meant tens or even hundreds of thousands of dollars less in state spending.

Governing by investment

The scenes of care in hot spots appear to represent a narrow and targeted reversal of the social abandonment that led to the problem of super-utilizers. But in deploying care and social welfare in a calculative way, it supports neoliberalism and its financialized logic of poverty management. Neoliberal social assistance, as it is practiced in the health care safety net, is conceptualized as an “investment” in the population, as a strategic and targeted deployment of basic resources, one that promises to generate a return on investment for the state or health system in the form of cost savings. Here, neoliberal governance generates targeted expansions of welfare, and its investments animate care and assistance for people who have long been denied resources for survival.

As we discuss elsewhere (Fleming et al. 2019), an ethos of care enacted by frontline health care providers often exceeded the bare financial rationality of hot-spotting programs, an ethos that was nonetheless imbricated with institutional logics of economic maximization. The health care providers on the hot-spotting teams were acutely aware that their mission was to reduce the costs of the most expensive patients, but they simultaneously deployed a social justice or humanitarian ethos as they took a reparative approach to the life conditions of the patients. The health care providers in this study were attentive to the structural determinants of health for marginalized people and understood their work as an intervention into the negative health effects of poverty. They enacted what can be seen as a form of “structurally competent” medicine (Metzl and Hansen 2014). The providers portrayed the patients’ health conditions as a societal injustice, and they often invoked a language of marginalization, inequality, and health disparity.

Hot spotting is surging as a governing mode in part because it ostensibly binds a care and justice ethos with an economic logos; it promises to distribute resources according to financially driven, cost-effective principles, and by doing so, to produce a moral effect: caring for the highest-need and sickest people, and improving the quality of care for all. A medical director told me, “The business case aside [the financial argument for providing care], my take on this from an ethical or moral standpoint, it’s incumbent on us to direct the most resources to the patients who are experiencing the most suffering. And that’s really the folks we’re taking care of.” In this account the financial rationality of hot spotting results in care for those who are suffering the most. This logic signals a neoliberal moral economy in which a financialized distribution of care maximizes the social good.

Across the street from the hot-spotting clinic, in an economically excluded, largely Black neighborhood, there was a billboard that represented another example, or mutation, of governance by investment (see Figure 2). The patients and health care providers would see this sign each time they went to the clinic. The sign shows how an investment rationality of governing—together with an emphasis on the economic costs of managing the poor—has become a kind of common sense regarding how those living in poverty are to be managed. It addresses the viewer as someone who would make a simple, rational investment (#DoTheMath) in the fate of the child—that is, choose the cheaper option. At the same time, it makes a moral argument about the injustice of prioritizing criminalization over education, again raising the amalgam of financial and moral logics at play. This representation of the problem also raises a counterfactual: What if it were cheaper to incarcerate rather than educate? Held within this imagery is the implication that if the “math” showed that it was cheaper to incarcerate this individual, perhaps it would make political sense to do so.

Figure 2.

This image from the California Endowment’s #DoTheMath campaign was on a billboard near the hot-spotting offices as well as other locations throughout California. (By permission of the California Endowment)

The health care providers we worked with embraced the financial-moral amalgam. This may simply be a strategic move for garnering resources for what they see as justice-oriented work. The health hot-spotting programs successfully delivered significant resources to people who had been excluded and subjected to state violence. They did so by representing them as investable, as figures of human capital, capable of producing a payoff. This, however, also has the effect of supporting the legitimacy of financialized management of the poor and downgrading the moral argument for care. Meanwhile, the “letting die” of marginalized populations is still practiced and justified within a neoliberal framework. We might therefore predict that when it is no longer cost-effective to invest in those residing in hot spots, they will once again be abandoned or incarcerated.

This possibility is further underscored by the fact that hot spotting proliferates a new category of stigmatized subject: the so-called super-utilizer, represented as a gathering threat to the nation’s health and economy. Medicalized constructions of the super-utilizer, the “frequent flier,” and the “chronically homeless” are becoming entrenched categories in the moral discourse on poverty, alongside older tropes of welfare queens and superpredators, all of which are figures of threat, in the form of unruly behaviors or uncontrolled consumption of public resources.

Lastly, as the result of a demand to manage marginalized populations more efficiently, states are rethinking the relationship among medicine, social services, and criminal justice. When the state’s primary purpose is to maximize the impact of financial investments, the lines between domains of state action become blurred. Punishment, care, and social services (and even education) may be resolved into and made comparable with a singular register of economic calculation. A major area of health policy innovation today pushes for better coordination across these sectors, under the rubric of cost savings. For instance, California’s Whole Person Care initiative, composed of Medicaid and state-matched funds, has devoted $3 billion to projects that integrate surveillance, infrastructure, and intervention across health care, social services, and criminal justice. The aim is to identify and manage the highest-cost users of these systems, people who typically have chronic disease. The financial burden of chronic illness has become a key justification for expanding medical institutions’ scope into carceral settings and rethinking criminal justice practice (Simon 2013).

As we witness the unfolding of criminal justice reform alongside ongoing health care reform, ethnographic attention should remain on institutional strategies of tracking and managing the most structurally vulnerable people—and the continual remaking of established categories of need, deservingness, and criminality. As medical care and carceral practices are currently and explicitly being realigned and as these sectors increasingly converge to manage patients with high-cost medical conditions, it will be important to attend to the divergent ethical orientations, techniques of intervention, and political ends that each sector pursues. This call is ever more urgent given the many racial and social justice protests taking place in US communities to radically shift investments and priorities among policing, health care, and social services. The demands put forward by these protests may in some ways align, and clash with, the financial rationality that undergirds poverty-management techniques like hot spotting.