“The Right Kind of Crazy”: How Patrol Officers Police the Boundaries of Mental Illness through Hybridized Strategies

Abstract

How to better police mental illness is an evergreen component of criminal justice reform agendas. The Los Angeles Police Department (LAPD) has, like many departments, adopted specialized strategies designed to improve these encounters by tasking officers with both care and control responsibilities. These hybridized policing strategies are illustrative of a larger trend of managing social marginality through institutions that increasingly destabilize the penal/welfare state binary. This article draws from fieldwork with the LAPD to analyze how patrol officers construct the category of “mental illness” and deploy hybridized strategies. The analysis focuses on the inflection points that shape how a subject is categorized and the call’s disposal to understand how policing from the “murky middle” of state governance unfolds on the ground. Findings show how officers strategically invoke the pressure of time and the power of place to construct this category and deploy specialized resources when resolving trouble case, or “5149 and a half,” calls. Here, hybridized strategies function to manage social marginality through a governance of problem solving that appears uninterested in doing either care or control. The article concludes by reflecting on the project of hybridizing care and control to police mental illness specifically and social marginality more broadly.

INTRODUCTION

It is late one weekday morning as two Los Angeles Police Department (LAPD) patrol officers respond to a 911 call for service placed by a man who has reported being worried about his sister. The patrol officers arrive at the residence, which the siblings share, and knock on the door. The man opens the door for the officers, who stand at the threshold to listen as he explains why he has called 911. “I’m worried about her, and I just don’t know what to do. She’s been up for days, and I just don’t know how to help her,” he says, running his hand through his hair. “All right, sir. Can we come in?” The man steps aside and lets the officers into the front room, where his sister is vacuuming the room’s front-right corner in tight circles. “Your brother says you’re not sleeping, and he’s worried about you. What’s going on?” one of the officers asks the woman. She starts to explain—yes, she has been up for several days, but that is not exactly unusual for her: “Sometimes, I just have stretches of sleeplessness happen to me.” She is shouting so that the officers can hear her over the vacuum, which she has continued to push in a small semi-circles throughout the exchange. “I’M FINE,” she shouts. “Okay, I see. Ma’am, can you turn off the vacuum?” the same officer asks. She does so reluctantly. “I don’t even know why he called you guys,” she exclaims, turning to her brother with an accusatory glance and then back to the officers. “Look, I need to finish cleaning the house before I go, and I’m already late!” She turns the vacuum back on and continues to circle it over and over the same patch of carpet.

The brother sighs loudly. “She just won’t listen to me,” he says.

I watch this interaction unfold through a one-way mirror while standing in a narrow corridor sandwiched between two identical conference rooms in the LAPD’s headquarters in downtown Los Angeles. Behind me, two more patrol officers are responding to a call for service involving a woman in a park who 911 callers have described as exhibiting odd behavior. I overhear her answer to an officer’s opening how-are-you-doing query before I refocus on the vacuum call. “I’m very busy. I’m working on a secret mission. The Supreme Leader and me are working to save the planet,” she says. It is the final day of the LAPD’s Mental Health Intervention Training. The participants—all LAPD patrol officers—are working in pairs to respond to three different “scenario-based skills training” role-playing exercises.

Each scenario allows participants to use skills learned during the training about drawing from the context of a call and eliciting enough relevant information from subjects to decide how to best dispose of the situation. In each scenario, the officers should determine that the call involves someone who has a mental illness or is in a mental health crisis and that the department’s Mental Evaluation Unit (MEU) should be contacted for additional guidance. Over the next couple of hours, I watch a dozen pairs of officers respond to the different scenarios from the sidelines. Several MEU members, who have spent the last three days facilitating the training series for thirty-some LAPD patrol officers, are there to observe and evaluate how participants perform on the exercises. Officer Cruz, the MEU member watching the vacuum call alongside me, editorializes for my benefit: “They should ask her why cleaning the house is so important to her and what she is late for.”

The responding officers look at each other for what feels like a long moment but is just several seconds. One says loudly: “Okay, ma’am. You want to clean the house and get on with your day. We get it, but your brother is worried about you, so here we are. We want to make sure you are okay.” The other officer jumps in: “Can you turn off the vacuum so we can talk and get you on your way?” “I GUESS,” she shouts. “Thank you, ma’am. That’s great. Can I ask you what’s so important to get to? Where are you heading after you get the house cleaned?” “Good,” Officer Cruz tells me, nodding his head. “If they keep prompting her with the right questions, she’ll tell them she is running late for a poker tournament and that she has just decided to skip work so she can go to the casino instead. We’ll see what they decide to do once they get those details from her.”

I ask him what the officers should do next. He explains that the responding officers should consider her impulsive behavior in conjunction with the history of sleeplessness that she has already volunteered at the beginning of the interaction. If they collect all the relevant details by asking the right questions and correctly piece them together, they should decide that the woman might be experiencing a manic episode and that it is possible that she is unable to take care of herself—or what would be considered “gravely disabled” under California’s civil commitment law, Welfare and Institutions Code (WIC) § 5150. ¹ As Cruz explains, “we’d want the officers to then call the MEU for assistance in the field with this call to see if this is a 5150 situation. She might be bipolar, and we’d be able to help the officers figure out whether or not she should be transported to County [hospital] for a possible psych hold.” ²

Since the late twentieth century, US police departments have increasingly implemented specialized trainings and resources to help officers more effectively recognize, de-escalate, and respond to situations that, like the ones included in the MEU’s scenario-based skills training exercises, involve someone who has a mental illness or is experiencing a mental health crisis (Deane et al. 1999). The regularity with which officers encounter mental illness in the field reflects the increased role of carceral institutions in managing this social problem. According to the “trans-institutionalization” literature, the primary responsibility for managing the institutionalized “mentally ill” shifted from the state hospital system to the criminal justice system as a result of a constellation of changes to mental health and criminal justice policies during the late twentieth century (Gilligan 2001). As a result of these changes, police have come to function as de facto “streetcorner psychiatrists” (Teplin and Pruett 1992), who routinely encounter situations involving mental illness and then funnel people with mental illnesses into jails and prisons, which serve as the “new asylums” (Torrey 1997).

In this way, the arc of trans-institutionalization is consistent with a robust punishment literature that analyzes how the social problems that were once within the province of the welfare field have been reallocated to the penal field during the late twentieth-century punitive turn (Garland 2001; Wacquant 2009). People with mental illnesses, previously managed through the welfare state’s hospital system, are now managed through the penal state’s criminal justice system, where its control imperatives impose significant harm. Criminal justice empirics reflect these realities. For example, people with mental illnesses are disproportionally incarcerated in US jails and prisons (Bronson and Berzofsky 2017), where they then face additional harm to their well-being (Reiter and Blair 2015). Nearly 25 percent of the people that police have killed since 2015 were experiencing a mental health crisis (Washington Post). The criminal justice system’s institutions may be on the frontlines of managing mental illness, but they do so at a high cost.

From a policy perspective, specialized policing strategies that make the role of “streetcorner psychiatrists” explicit have proliferated across the country and are designed to remediate these harms through trainings and resources that prioritize identifying people as having a mental illness and diverting them to treatment. As suggested in the opening vignette, applying for an emergency psychiatric hold, or a “5150” in LAPD slang, is one of the primary mechanisms that police use to facilitate access to care services, and, for LAPD officers, calling the MEU to discuss the appropriateness of this option for calls involving mental illness is mandated by departmental policy (LAPD 2015). As a result, these specialized models make facilitating access to behavioral treatment services for people with mental illnesses, which is already a regular part of policing, an institutionalized responsibility (Lamb et al. 1995).

In this way, specialized policing strategies are emblematic of a larger shift in the state’s governance of social marginality through institutions that destabilize the penal/welfare state binary by hybridizing care and control. Increasingly, scholars have troubled framings that, like the trans-institutionalization literature, conceptualize the relationship between the penal and welfare state as zero-sum. Instead, studies identify how the two fields intersect to govern social marginality (Beckett and Western 2001; Soss, Fording, and Schram 2011). For example, a growing body of ethnographic studies of the criminal justice system finds that the penal state’s institutions—from the police (for example, Stuart 2016; Herring 2019; Pifer, 2019), courts (for example, Tiger 2012; Kohler-Hausmann 2018, Zozula 2019), and corrections (for example, Miller 2014; Halushka 2019; Sue 2019; Phelps and Ruhland 2021)—are involved in delivering the social services that have traditionally been understood as being within the welfare state’s province. The prevalence of strategies hybridizing care and control suggests that the state’s approaches to managing social marginality should be mapped along a continuum that allows for a “murky middle” where punishment and welfare institutions converge (Brydolf-Horwitz and Beckett 2021, 94). I engage with the literature on hybridization by placing specialized strategies like the LAPD’s MEU on this continuum and examining how patrol officers decide to resolve calls for service that, like the one described in the opening vignette, implicate policing from this murky middle.

I draw from fieldwork with the LAPD to analyze how patrol officers decide when to call the MEU for assistance with a call for service. Under LAPD policy, officers must call the MEU’s Triage Desk, which is staffed 24/7 by a specially trained LAPD officer or a County Department of Mental Health clinician, whenever they encounter someone that they believe may have a mental illness. The LAPD uses California’s Lanterman-Petris-Short Act, better known by the WIC section number that informs LAPD slang for “person with mental illness”—as its practical definition of mental illness. Thus, patrol officers are required to call the Triage Desk for assistance in the same situations where WIC § 5150 would empower them to apply for an emergency involuntary psychiatric hold: when the subject “is a danger to others, or to himself or herself, or gravely disabled” as a result of a “mental health disorder.”

As a result, a call to the Triage Desk implicates a two-step process, in which officers first categorize a subject as having a mental illness and then categorize the call itself as one that falls within the MEU’s purview. One purpose of this article is to analyze how the categorization process unfolds during individual police-subject encounters as patrol officers navigate layers of LAPD hierarchy, the law, and the context of each call to determine whether a call for service involves, as one patrol officer described it to me during fieldwork, “the right kind of crazy” that should be managed by the MEU. ³ Another purpose of this article is to analyze what this category can reveal about the types of problems that are envisioned as properly solvable by the MEU. How do officers decide which calls for service should be managed via hybridization, and what do those decisions reveal about the governance that is at hand in the murky middle?

I show how officers resolve “trouble case” calls where, as in the opening vignette, it is not readily apparent whether the MEU’s specialized resources should be invoked to evaluate if the subject is a “5150” who should be diverted to emergency psychiatric treatment (Llewellyn and Hoebel 1941, 29). My analysis of these trouble case calls, which I term a 5149 and a half, makes two contributions. First, I add empirical depth to our understanding of how patrol officers construct the material reality of mental illness by identifying how subjects are filtered in—and out—of the category of “mentally ill” as officers draw from the context of trouble case calls. Second, I show how patrol officers decide which trouble case calls should be resolved by invoking the MEU. In both decision points, I find that patrol officers strategically invoke the pressure of time and the power of place to inform the construction of this category and the utilization of specialized resources. Patrol officers leverage these contextual considerations to strategically navigate the process of “burden shuffling” trouble case calls to the MEU’s murky middle (Seim 2017, 452). As I show in the article, these decisions are much less about how officers define or identify mental illness in any medical sense and much more about how they construct which trouble case calls can be solved via the MEU. I argue that these hybridized strategies thus function to manage social marginality through a governance of problem solving that is shaped more by time and space constraints than by the desire to do either care or control.

This article thus adds dimension to our understanding of the “murky middle” of state governance by considering how frontline workers construct the category of mental illness to deploy a hybridized policing strategy that makes the management of social problems beyond crime its goal. Frontline workers invoke contextual factors to construct the category of mental illness and decide which problems can be solved by shuffling the burden to an institution that lies somewhere along the state’s expanded governance continuum and, alternatively, which problems are envisioned as being “unshuffable.” It also demonstrates the policy limitations of reform strategies that ask frontline workers to hybridize care and control to “better” manage social marginality.

THEORETICAL CONSIDERATIONS

The issue of how police can better navigate their role as “streetcorner psychiatrists” (Teplin and Pruett 1992) has been one of policing’s most chronic challenges for the last forty years (Engel 2015). Increasingly, police departments have implemented specialized trainings and programs that are designed to help officers better understand, de-escalate, and respond to calls that involve a mental illness or mental health crisis (for example, Borum et al. 1998; Deane et al. 1999; Steadman et al. 2000; Hails and Borum 2003). These strategies have long been the focus of numerous studies evaluating their efficacy as policy reforms intended to accomplish goals like diverting people with mental illnesses away from the criminal justice system and reducing the likelihood that police encounters will result in officers using force. For example, policing scholars evaluating various types of specialized strategies and training have focused on outcome metrics like officer perception (for example, Ritter et al. 2010; Morabito, Watson, and Draine 2013; Bonfine, Ritter, and Munetz 2014) and their effect on the outcomes of police encounters with people with mental illness or on officer ability to identify mental illnesses (for example, Janus et al. 1980; Watson et al. 2009; Ellis 2014; Morabito and Socia 2015; Peterson and Densley 2018; Rogers, McNiel, and Binder 2019).

Outside the policing policy literature, specialized policing strategies are also of increasing interest to welfare and punishment scholars examining the role of police and the penal state more broadly in managing social marginality. For example, in their typology of policing social marginality, Steven Herbert, Katherine Beckett, and Forrest Stuart (2017) identify three different approaches that are distinguishable, in part, by the degree and character of care they deliver alongside their coercive police power. In describing each of the three approaches, they identify and describe a particular specialized policing strategy as emblematic. For example, broken windows policing strategies in New York and Seattle typify the “aggressive approach,” the LAPD’s Safer Cities Initiative typifies “coercive benevolence,” and the Seattle Police Department’s Law Enforcement Assisted Diversion (LEAD) Program promises a new form of “officer-harm reduction.” Chris Herring (2019) has recently identified “complaint-oriented policing” as an additional form of policing social marginality utilized by the San Francisco Police Department to regulate homelessness.

Substantively, the policing strategies cited and described to articulate the differences between the models are focused on responding to the problems of poverty, many of which have also been analyzed by scholars interested in how policing strategies provide insight into postindustrial governance (for example, Davis 2006; Vitale 2008; Beckett and Herbert 2009; Sharp 2014; Laniyonu 2018). But how might this typology be used to analyze specialized policing strategies focused on other forms of social marginality that have come to be managed by the criminal justice system during the late twentieth century, such as mental illness? While there are instances where policing urban poverty intersects with the policing of behavioral health, as when Seattle police officers make a prebooking diversion to a LEAD case manager who can facilitate access to methadone treatment (Herbert, Beckett, and Stuart 2017, 15), the focus of the crisis intervention trainings and specialized response units that have been evaluated by policing scholars is on managing mental illness in general. Where do models like the LAPD’s MEU, which is designed to be deployed in response to all LAPD encounters with people with mental illness, fall on the care and control continuum of approaches to policing marginality?

In implicating these questions, specialized policing strategies are also of broader analytical interest to scholars focused on understanding the hybridization of care and control in the criminal justice system’s management of the socially marginalized (for example, Miller 2014; Stuart 2016; Tiger 2012; Kohler-Hausmann 2018; Halushka 2019; Herring 2019; Pifer, 2019; Zozula 2019; Phelps and Ruhland 2021). Research in this area suggests that punishment and welfare are enmeshed such that the penal state’s core actors are involved in the project of caring for and controlling marginalized populations (Beckett and Western 2001; Soss, Fording, and Schram 2011). For example, Marco Brydolf-Horwitz and Beckett (2021) identify Stuart’s (2016) work on “therapeutic policing” in Los Angeles’s Skid Row as the sort of policing strategy that exists in the interinstitutional “murky middle” of state approaches to managing social marginality. Stuart describes the LAPD’s Safer Cities Initiative as a campaign designed to govern the urban poor by leveraging the threat of punishment to coerce Skid Row’s residents into diversionary programs like housing shelters and recovery organizations.

More recently, Chris Herring (2019) has shown how officers tasked with policing homelessness in San Francisco sometimes reclassified calls for service involving a homeless person to another bureaucracy, such as a social welfare, medical, or sanitation agency. In this way, Herring shows how police engage in a process of “burden shuffling” (Seim 2017) the problem of homelessness to other frontline workers, which is designed to protect police time and, as a byproduct, facilitate access to social services for the urban poor. Policing strategies in the murky middle implicate the power of the police to facilitate access to the types of social services resources that the welfare state has typically managed. Existing research shows how police use “coercive benevolence” (Stuart 2016) or “burden shuffling” (Herring 2019) to facilitate this access to care, but how do police use specialized policing strategies that make the provision of care a defining feature to navigate this space? In some ways, specialized policing resources focused on mental illness take the hybridization of care and control that Brydolf-Horwitz and Beckett (2021) argue exists in the murky middle even further by equipping them with trainings and resources that teach them mental health intervention techniques or provide, or even pair, them with mental health clinicians (Hails and Borum 2003).

This hybridization is especially explicit in the specialized model at the heart of this article. The LAPD’s model is characterized as what policing scholars describe as a multilayered, co-response model. The MEU’s multiple layers refer to the several components that are housed within the unit, including a Triage Desk that is staffed 24/7 to provide patrol officers with guidance about how to manage calls in real-time, a fleet of System-wide Mental Assessment Response Teams (SMART) that respond to calls in the field, a Case Assessment Management Program that can follow up with individuals through a non-emergency visit, and the Training Unit that facilitates the department’s Mental Health Intervention Training. The MEU is characterized as a co-response model because its members are all specially trained LAPD officers or County Department of Mental Health clinicians and because its SMART units pair an officer and a clinician into a single mobile unit to assist patrol officers and divert people with mental illnesses from custody to treatment by making mental evaluations and referrals in the field. Given this arrangement, I place the LAPD’s MEU on Brydolf-Horwitz and Beckett’s (2021) continuum to frame an analysis of the types of calls for service that are managed by this unit’s literal pairing of care and control agents into a hybridized unit. What, for example, might the MEU’s deployment to some calls but not others reveal about the types of problems that are managed in the murky middle?

As the call for service described in the opening vignette demonstrates, patrol officers are the primary responders to the vast majority of calls for service involving someone who has a mental illness, but they are required to call the MEU’s Triage Desk after stabilizing the scene for guidance whenever they encounter someone they believe may have a mental illness (LAPD 2015). Thus, understanding the MEU’s function as a case of policing in the murky middle also requires understanding how patrol officers decide who has a mental illness and when to call the Triage Desk. Policing literature has long established that the police, like the general public, are able to correctly identify someone as having a mental illness in a general sense (Bittner 1967). Yet we know much less about the processes through which police officers construct the category of “mentally ill” or how people are filtered in—or out—of that category during encounters. This is likely because most contemporary research on this genre of specialized policing strategies is designed to evaluate whether a particular training improves the accuracy with which police can identify mental illness—an epistemological framing that takes mental illness as an objective category with an essential basis for diagnosis.

Instead, I employ a modified social constructionist approach to move the analysis beyond whether individual officers correctly identify individuals as having a mental illness to examine how and why they categorize some people but not others as “mentally ill” (for example, Link et al. 1989). Research in this tradition acknowledges the existence of an objective basis for diagnosis, but it is focused on uncovering the social processes that inform how and why some people are sorted into ostensibly medicalized categories of mental or medical abnormality (for example, Noll 1991; Trent 1995; Dorr 2006; Metzl 2009). This is an especially important framing for the present study because it moves the analysis beyond whether or not individual officers can correctly identify individuals as having a mental illness to examine how and why they categorize some people but not others in this way. As such, my study is less concerned with whether or not officers “get it right” when applying the label of “mentally ill” and more with examining factors that shape when and how they apply that label in order to invoke—or avoid—specialized resources. This decentered approach prioritizes examining how patrol officers shape the meaning of mental illness for deploying the specialized strategies at the heart of this case study.

While this framing may be novel in the policing context, the broader punishment and society literature has already mobilized a constructionist approach to show that the on-the-ground meaning of criminal justice policies predicated on these hybridized medical-social categories is especially dependent on how frontline workers do categorization work (for example, Miller and Radelet 1993; Rhodes 2004; Pifer 2016). For example, Lorna Rhodes (2004) shows how custody staff renegotiate and delineate the boundary between “mad” and “bad” in solitary confinement when deciding who requires care and who requires punishment against the carceral logics that dominate in maximum custody. In the policing context, earlier work has identified some of the factors that help shape why officers deem some encounters a “serious police problem” requiring formal disposal (Bittner 1967, 279), but these studies are more interested in identifying the factors that predict how a call involving someone with a mental illness will be resolved rather than the factors that shape how officers bound the category of mental illness itself (for example, Teplin and Pruett 1992). For example, officers consider the extent of the bureaucracy (Rock, Jacobson, and Janopaul 1968) or the amount of time (Matthews 1970) that would be involved in pursuing hospitalization when deciding how to dispose of an encounter with someone who is in psychiatric distress. What factors are involved in constructing the category of mental illness itself?

The present research is interested in centering categorization work as a critical component of understanding how a particular policing strategy unfolds on the ground. The LAPD’s specialized resources are only deployed when a patrol officer invokes them; unpacking how officers categorize a person as having a mental illness or not is thus critical to understanding how policing from the murky middle functions on the ground. What inflection points shape how people come to be categorized within the scope of the MEU’s purview? And what does this categorization process reveal about the function of hybridized policing strategies in managing social marginality?

RESEARCH SITE, DATA, AND METHODS OF ANALYSIS

My data were collected during fieldwork conducted with the LAPD. Over the course of five months, I participated in both “ride-alongs” and foot patrols facilitated by ten sergeants and one lieutenant during watches in three of the four LAPD bureaus—the Valley, Central, and South bureaus—that spanned both day and night. Under LAPD guidelines, ride-alongs must be facilitated by a supervisor—this was usually, but not always, the patrol sergeant on duty for any given watch—so while I was not able to always observe the initial moments of an encounter between an officer and a subject, the supervisors that facilitated my ride-alongs agreed to monitor incoming calls for service and respond proactively to calls that could be especially relevant to my study. In addition, LAPD policy requires a supervisor to be present for all calls involving someone who may have a mental illness to ensure that departmental guidelines, including those that mandate a call to the MEU’s Triage Desk, are followed. Together, the cooperation of the officers who facilitated my ride-alongs and the LAPD’s (2015) Department Manual maximized my ability to observe calls that involved persons who perhaps had a mental illness and see how patrol navigated these calls.

One additional piece of LAPD policy is worth noting here. When a call for service comes through to an officer’s in-car computer, dispatch automatically attaches a reminder to the incident detail for any and all calls for service that dispatch believes, based on the 911 call, could involve a person who may have a mental illness; this reminder reads in all capital letters “Contact MEU upon Scene Stabilization.” During my study, this reminder proved to be in no way determinative of how an encounter would unfold or whether patrol would call the MEU. I observed officers categorize calls as within the purview of the MEU that did not have this tag in the incident detail, and I observed officers categorize calls that did have this tag as outside the purview of the MEU. Some of the trouble case calls I analyze in my findings section featured this reminder from dispatch, while others did not, but I do not attach any real analytical significance to this tag because I observed that the ultimate categorization of a subject as mentally ill or not and the decision to contact the MEU or not was entirely negotiated by the responding officers. I note the presence or absence of the tag in my findings section. Methodologically, though, this reminder served to prompt the patrol supervisor who was facilitating my ride-along to respond to calls, even if they were outside of an assigned area, to serve as the supervisor for the responding officers, which helped to maximize my ability to observe calls of interest. In this way, I think that my presence as a researcher may have altered how patrol supervisors utilized the tag. It is entirely possible that patrol supervisors would not have responded to some calls but for my presence as a ride-along.

My university’s internal review board agreement allowed me to document my observations of how LAPD officers engaged with individuals believed to be mentally ill and interact with officers, but it prohibited me from interacting with the members of the community I encountered while on a ride-along for ethical reasons. As a result, I did not systematically collect data about the subjects of police encounters beyond what I could observe or glean during a conversation with the officers about the call. I did not conduct formal interviews with the LAPD officers, but I always spoke informally with the officers involved with the call about their work and about the specific interactions I was observing. These conversations are an important source of data that supplement and help to contextualize my observations of the calls because I was able to ask the officers to reflect on their experiences and explain their choices. This is consistent with my study’s focus on understanding why and how officers categorize people as having a mental illness or not rather than on evaluating whether or not they made the “correct” choice.

I also attended the complete series of training classes facilitated by the MEU’s Training Unit that is mandatory for all LAPD patrol officers. This Mental Health Intervention Training (MHIT) is designed to provide small groups of officers with educational and tactical information about policing people with mental illnesses and includes lectures by MEU officers and clinicians, presentations by community groups, small group activities, role-playing exercises, and field trips. I had intended to attend this training in a strictly observational capacity, but I ultimately occupied a more participatory role given the training’s focus on active learning and the small size of the training group (approximately thirty patrol officers attended the training series that I observed). I did not participate in the group discussions, but I did passively participate in most of the break-out small group activities and completed one video role-playing scenario where I was given a training gun and asked, with “my” partner, to deescalate a call for service involving a suicidal man with a gun in his garage.

I do not think my presence as a researcher in these activities altered the course or content of the training series, likely because the MEU’s Training Unit hosts LAPD outsiders frequently. Most importantly, for this present study, the training allowed me to have informal conversations with patrol officers from across the city about their work and to observe each officer’s performance on the training’s “final exam” activity on the last day of the training: a series of three “calls for service” role-playing exercises where officers work in teams and MEU members evaluate (1) the participants’ ability to apply the mental health intervention techniques that have been learned during the training and (2) how participants identified when to contact the MEU for assistance during a call for service. Observing these exercises, one of which I described in this article’s opening vignette, helped me to conceptualize the analytical category of a trouble case call as one that required responding officers to do ad hoc categorization work to determine if this person and this call merited MEU intervention by soliciting additional information. I used this ambiguity to analyze my ethnographic field notes to understand how patrol officers use context to construct the meaning of “mental illness” for deploying the MEU’s specialized policing resources.

For this article, I drew from both the observational data I collected during calls for service and training classes as well as the open-ended conversational dialogues that I had with LAPD officers during training downtime or while driving in the squad car during calls for service, in stations before and after watches, in jail while processing an arrestee, or in an emergency room awaiting medical or psychiatric admission. Collectively, I spent approximately 120 hours in the field for this study and took notes during informal conversations with eighty-four different LAPD officers. At the beginning of each day in the field, I made sure, as part of the informed consent discussion, to ask permission to take notes of my observations and then again when starting a conversation with an officer. Permission was always granted, and, in my opinion, I do not think my taking notes altered the course of a call or had a chilling effect on the officers’ willingness to speak with me, which is perhaps because the act of taking notes in the field to produce a report later on is very familiar to officers. My data was thus recorded in what David Snow and Leon Anderson (1987, 1344) call the “stepwise fashion.” I first made mental and jotted notes in a reporter’s notebook while in the field and then made a more detailed audio recording in which I narrated all of my recollections and conversations related to my research questions immediately after exiting the field, which were later transcribed. I then expanded these jottings into a detailed and extensive field narrative following each day’s observations. As a result, the interactions, conversations, and exchanges that I report in the following sections are my best efforts to reconstruct the dialogue I overheard or engaged in while in the field. While my use of the step-wise method to record data ensures that the spirit of the dialogue is accurate, I never employed a digital recorder while in the field, and, thus, while I do report quotations, perfect transcription is impossible. I also note that the ethnographic field notes do not include any identifying information about the participants, and I use pseudonyms for people and places to protect their privacy.

Data analysis was conducted through an iterative cycle of coding my field notes, reading the literature, and memo writing (Emerson, Fretz, and Shaw 2011; Timmermans and Tavory 2012). I subjected my field notes to multiple rounds of open coding and thematic analysis during a memoing process. During this iterative process, I coded first for the recurrent and significant themes that emerged from the data. Then, after several rounds of open coding and a two-stage memoing process, I built a more structured set of coding categories designed to reveal how specialized policing strategies are deployed and how mental illness is constructed on the ground by the LAPD officers in my study. Finally, while I observed and interacted with officers who work across the city, I have not made any strong claims about my findings’ generalizability to the entirety of the LAPD—which has approximately nine thousand sworn officers—or any other jurisdictions. Instead, my article is designed to provide a deep look at how the patrol officers in my study decided to use, or not use, the MEU’s specialized resources. This question is worth asking because the LAPD’s approach to policing mental illness is considered a model by other law enforcement agencies and advocates alike, making it an important object of study in understanding how specialized policing strategies function on the ground.

FINDINGS: RESOLVING THE 5149-AND-A-HALF TROUBLE CASE THROUGH TIME AND PLACE

To present my findings, I describe a selection of four 5149-and-a-half trouble case calls that I observed in the field to demonstrate how officers in my study filtered individuals in—and out—of the category of “mentally ill” during patrols when deciding whether to deploy the MEU’s hybridized policing strategy. I define a 5149 and a half as those calls for service where it is not readily apparent whether or not a subject is a person with mental illness, and, thus, it is unclear whether or not the MEU should, under LAPD policy, be contacted. I focus on how 5149-and-a-half trouble case calls are resolved because their ambiguity allows me to identify factors that seemed especially important as officers decided which “burdens” should be “shuffled” to the MEU (Seim 2017). In the sections that follow, I report two findings about the process of categorizing and disposing of trouble case calls: the first is on the role of efficiency concerns and the second is focused on how officers invoked place. For each finding, I analyze one trouble case call where the patrol officers deployed the MEU and one where the patrol officers did not to show how each factor can function to move a subject in as well as out of the category of “mentally ill.”

The Pressure of Time

Efficiency concerns always percolated in the background of constructing “mental illness” since any call involving someone who may have a mental illness should, under LAPD policy, involve the MEU. While the MEU is designed, in part, to save patrol officers time in the field by allowing them to shuffle the burden of a time-consuming “5150” call to one of the MEU’s specialty teams, involving the MEU imposes its own set of time demands on patrol officers. The process of deploying the MEU begins with a mandatory call to the MEU’s Triage Desk. At the bare minimum, this call involves a checklist of questions that unfolds through an extended game of literal and figurative telephone played between the MEU member working the unit’s Triage Desk and the patrol officers at the scene. During this call, the Triage Desk will also check the LAPD databases to determine if the subject has a history of contact with the police and the Los Angeles Department of Mental Health databases to identify any case managers, psychiatrists, or treatment centers involved in the subject’s mental health care. Ultimately, the phone call to the Triage Desk is designed to determine if the subject may meet the criteria for a WIC § 5150 hold. If so, the patrol officers must then either wait for a SMART unit to evaluate the subject in the field or, if a SMART unit is not available, transport the subject themselves to an emergency psychiatric facility where a doctor assesses whether an emergency psychiatric hold is justified.

The year before I started my fieldwork, the Los Angeles Times reported that LAPD officers tasked with transporting someone to be evaluated for an involuntary psychiatric hold spend, on average, two hours in the waiting rooms of one of the county’s three public psychiatric emergency departments (Sewell 2015). When I attended the MHIT series, the officers I spoke with placed the average wait time at four hours, and some described being off the streets and tied up with a WIC § 5150 hold for an entire “watch” or shift. Officer resentment at the wait and being off the streets dealing with an undesirable call was palpable. During my time in the field on ride-alongs, patrol officers navigated their categorization of any given situation involving someone who may have a mental illness in the shadow of this time-consuming process and strategically decided when to avoid or deploy the MEU’s specialized policing resources—and all its procedures—to maximize patrol efficiency.

In this excerpt from a Friday afternoon ride-along in downtown Los Angeles, the pressure of time was eventually revealed as an instrumental force in how the subject was (almost) categorized and the MEU’s resources leveraged. A caseworker at a shelter had called 911 to report that Terry, one of her clients, had made concerning statements to his counselor during a session and had become increasingly agitated afterward, yelling and slamming doors down the dormitory hallway. Dispatch had put out a call for service that tagged the incident as one in which the officers should, upon scene stabilization, contact the MEU, and Sergeant Hunt and I responded to a patrol unit’s request for a supervisor.

Once inside, the officers interviewed Terry’s caseworker and counselor in the dormitory hallway and learned that Terry had become upset after realizing that he had lost his wallet sometime the day before. As the caseworker explained, Terry had spent the morning shouting concerning statements while pacing up and down the hallway. The counselor added that Terry had stopped taking the mood stabilizers prescribed for his bipolar disorder about two weeks ago. Given this, Sergeant Hunt decided that a Triage Desk call was appropriate, but a call to the Triage Desk alone is not dispositive of the categorization process. During the phone call, officers, in consultation with the MEU’s expertise, must still determine whether or not the subject fits any of WIC § 5150’s criteria. For this particular call, the question was not whether or not Terry had a mental illness but, rather, whether the LAPD should transport Terry to the nearest county hospital for an emergency psychiatric evaluation. The call to the MEU’s Triage Desk is designed to help patrol officers in the field make this determination, and, thus, this trouble case call with Terry illuminates the nuance that is involved in resolving a 5149 and a half and the role that factors like the pressure of time can play in how specialized policing resources are deployed.

After Sergeant Hunt’s decision that the MEU needed to be contacted, one of the responding patrol officers placed a call to the Triage Desk, while his partner and Sergeant Hunt continued to gather information to relay to the Triage Desk. “He’s been saying things like ‘there’s no hope anymore’ and ‘there’s no point in going on anymore” and that he’s ‘just going to go out today and see what happens,’” his caseworker and counselor explained to the officers. Meanwhile, Terry remained in his room, telling no one and everyone that he was a Vietnam veteran with no criminal record. “Why did you call the police?” we heard Terry shout. Sergeant Hunt asked Terry to come out of his room so that the officers could interview him. In Terry’s case, the extent of the MEU’s involvement came to hinge on how the patrol officers framed their description of the situation. If they constructed Terry as a danger to himself in their assessment of the situation to the Triage Desk, the patrol officers would need to either wait for a SMART unit to respond in person to assess Terry or transport Terry themselves to the nearest emergency psychiatric facility and wait for him to be admitted. If Terry was not deemed a danger to himself, the officers could leave without taking any official action since no one had alleged that Terry had committed a crime. This additional layer of categorization means that officers seeking to avoid the time-consuming procedures associated with a potential WIC § 5150 hold need only strategically navigate their description of the situation to avoid triggering one of the law’s three prongs. As a result, officers can navigate the determinative phone call to the Triage Desk by strategically framing their description of the situation to get the result they want.

In Terry’s case, the patrol officer tasked with calling the Triage Desk relayed a condensed description of the scene. His narrative started with the case worker’s initial phone call to 911, describing to the Triage Desk that shelter staff were concerned about a resident who had been slamming doors and yelling. Next, he described the scene at the shelter when the officers arrived, telling the Triage Desk that Terry was agitated and upset that the police had been called. “He says he’s a Vet and that his record is clean,” the officer told the Triage Desk as the MEU officer presumably checked Terry’s claim against the LAPD database. The patrol officer then summarized the information gleaned from the interviews with Terry and the staff. Terry was upset not just because he had lost his wallet but also because it meant that he would have to start the process of applying for benefits all over since he no longer had a valid ID. The officer paused to listen to the Triage Desk. “No, he didn’t say he was going to kill himself,” he responded.

This was true. No one on scene had reported that Terry had mentioned suicide or used the word “suicidal” during the interview, and Terry was adamant that he was just upset, but the patrol officer’s explicit operationalization of WIC § 5150’s “danger-to-self” prong as requiring an affirmative declaration of suicidality did not capture the complete context of the call. Staff had reported that Terry had made concerning statements that could be suggestive of a desire to self-harm. “Hold on,” Sergeant Hunt told the officer. “Let’s get the caseworker over here.” Terry’s caseworker repeated her story, and the officer dutifully relayed her more detailed description of Terry’s comments and behavior to the Triage Desk. The Triage Desk determined that Terry should be taken into police custody so a SMART unit could respond and further evaluate him. The second patrol officer escorted Terry around the corner to handcuff him, and Terry became even more agitated, shouting, in repeated sequence: “Why is this happening?” “Why am I in cuffs?” “I have no criminal record!” “I’m a vet—an American hero!” To his counselor, Terry said, “I’ll never trust you again.”

The counselor apologized. “I had no choice,” he explained and then more quietly and perhaps more to himself than to Terry: “This is wrong.” After Terry quieted, we exited the shelter; one patrol officer placed Terry in a squad car, while the second officer completed the phone call with the Triage Desk. “Do you have a gun,” he asked Terry, ducking his head into the backseat of the squad car. “No.” At the end of the phone call, the Triage Desk instructed the patrol officers to transport Terry to the division’s station to wait for a SMART unit to conduct an assessment. Sergeant Hunt and I then left the scene to continue patrolling downtown some forty-five minutes after the LAPD first arrived at the shelter. “So, what do you think,” he asked me as we crossed the street to his car to continue his shift. This was my second ride-along with Sergeant Hunt, and we had an established rapport. “I don’t know,” I answered with a shrug and a headshake. “What do you think?” “Yeah, they [patrol] didn’t want to deal with this. Did you hear the way he was answering the MEU’s questions? Yeah, Terry didn’t say he wanted to kill himself, but that wasn’t the whole story. That’s why I jumped in to grab the caseworker so that she could repeat what she told us, and it could be repeated to the MEU.”

Remarkably, Sergeant Hunt and I saw Terry walking down the street later that afternoon. Sergeant Hunt called out to him from his window, and Terry stopped to tell us what had happened since we had last seen him in handcuffs in the back of the squad car. He had been released from the police station after a SMART unit determined that he did not fit the criteria for an involuntary psychiatric hold. Sergeant Hunt asked Terry if he understood why his caseworker had called the police, and Terry said: “I get it. This all started with me making a joke that got twisted up.” Terry was in good spirits by then, joking that his only complaint now was the length of the walk back to the shelter. Officers know that calls for service involving the mentally ill will likely be time-consuming, especially if the subject ultimately does need to be admitted to an emergency psychiatric facility for an evaluation under WIC § 5150.

Time simmers in the background of every call involving a subject that may have a mental illness, and, here, only Sergeant Hunt’s intervention as the supervisor on scene prevented it from prevailing in the responding officers’ categorization of Terry. Of course, this is, in part, precisely why the LAPD’s policies require a supervisor’s presence on calls like these. And, as it turned out, the patrol officers were right that Terry was not a danger to himself, but they arrived at this conclusion based less on the context of the call and more because of their desire, as Sergeant Hunt put it, to avoid “dealing with it” and the time dealing with it would take. Dealing with Terry would mean spending part, or perhaps all, of their watch waiting for the SMART unit or a doctor to take custody of Terry, which is patrol time not well spent.

Other times, however, the pressure of time prevailed in categorizing subjects, as in this excerpt from a ride-along with Sergeant Wood in South Los Angeles. Around 10:30 p.m. and some six hours into the shift, Sergeant Wood received a request for a supervisor on a call for service involving someone who might be, as the incident log description noted, mentally ill. Sergeant Wood radioed in to confirm that he would respond, and, when we arrived, we found two patrol officers on the sidewalk in front of a house on a quiet, dark residential street adjacent to a major surface street running parallel to a main Los Angeles freeway. A woman in handcuffs sat on the curb, while a man watched from the house’s dark front porch, and the patrol officers briefed Sergeant Wood about their initial interviews. The man had called 911 because the woman was dancing on his front lawn and refused to leave. This odd behavior explained the incident log’s note to contact the MEU upon scene stabilization, and its disposal would hinge on how the officers explained why the woman was dancing down a residential street late on a Friday night. As the call unfolded, Sergeant Wood worked hard to move the woman, who said her name was Elizabeth, away from this category—and prolonging the LAPD’s responsibility—to protect his officers from spending their watch off the streets and in a hospital waiting room. In this trouble case call, the pressure of time was explicit.

After the briefing, Sergeant Wood instructed the officers to request a rescue ambulance so that paramedics from the Los Angeles Fire Department (LAFD) could evaluate Elizabeth and perhaps identify a medical, rather than a mental health, explanation for her behavior. This would allow him to “burden shuffle” Elizabeth from the LAPD to the LAFD (Seim 2017). When the ambulance arrived some ten minutes later, the paramedics escorted Elizabeth to the ambulance to check her vitals. However, to Sergeant Wood’s irritation, they soon returned with Elizabeth in tow and placed her inside the responding officers’ squad car. There was nothing to suggest that she needed immediate medical attention, which meant that, under the LAFD’s policy, they were not required to transport her to the hospital. They went back to their ambulance.

Sergeant Wood, however, was determined to insulate his officers from the task of transporting Elizabeth to the hospital in their squad car and waiting there until she was admitted to the emergency room for a psychiatric evaluation. He was already down a unit after two of his officers had been tasked with transporting a man in a mental health crisis to the hospital for an emergency psychiatric evaluation at the beginning of his watch, and, some five hours later, they had still not returned. Losing another unit to transport Elizabeth to the hospital, Sergeant Wood made clear to his officers while awaiting the rescue ambulance’s arrival, was simply not an option. Elizabeth began to protest from the squad car that she had done nothing wrong, and Sergeant Wood turned to question her directly for the first time. The interview would determine how Elizabeth would be categorized and whose burden she would ultimately become: “How are you feeling tonight?” “I’m feeling good, man,” she said. “Oh, so you’re feeling good? What did you take?” Sergeant Wood motioned for the paramedics to come back. If Sergeant Wood could establish that Elizabeth’s behavior resulted from drugs, the LAFD would, with patrol supervision, transport Elizabeth to the hospital. “Listen, I’m just dancing, having a good time,” Elizabeth countered. She gave a small dry laugh and rolled her eyes, “I’m Bob Marley.” “Oh, so you think you’re Bob Marley?” Sergeant Wood’s tone was sharp as he folded his arms. “Yeah, I think I’m Bob Marley,” Elizabeth challenged, her tone sarcastic.

Sergeant Wood turned his back to Elizabeth and, ignoring her tone, called the paramedics back over. “Did you guys check her out for drugs? She’s altered. She thinks she’s Bob Marley. She’s your problem.” Sergeant Wood walked back to his car and began instructing his officers on how to dispose of the call. “That’s going in the report,” he told his officers about Elizabeth’s Bob Marley comment. He instructed them to make sure the paramedics transported Elizabeth and to make sure that they got back onto the street as soon as possible. For Sergeant Wood, this outcome represented the most efficient way to dispose of the call. Though the patrol officers would stay with Elizabeth until she was admitted to the emergency room, she would be admitted through the ambulance bay rather than through the front doors, through which those who arrive in a squad car must be processed. This would significantly expedite the speed with which she would be admitted and relinquished from the LAPD’s custody. Perhaps on another night, if he had had more patrol officers in the field, the call would have unfolded differently, but, on that night, Sergeant Wood navigated the encounter through the pressure of time.

The Power of Place

Place also functioned as a central mechanism in the process of categorizing an individual as mentally ill. Specifically, officers invoked their preexisting knowledge of the particular place where a call for service was unfolding to help determine if a subject fit within the criteria delineated by WIC § 5150 and then if the MEU should be contacted (for example, Herbert 1997). As the following pair of trouble case calls demonstrate, the power of place can function in the background of encounters with varying degrees of explicitness to shift people in or out of the category of mental illness. For example, while on a ride-along in South Los Angeles, Sergeant Reyes and I arrived at a school housed in a small two-story stucco building, which looked more like an apartment building than a campus, to supervise two patrol officers responding to a call for service about a student from the school’s administration. We reviewed the incident detail displayed on the computer screen in Sergeant Reyes’s patrol car before joining the patrol officers, who had just arrived on the scene on the sidewalk outside the school. The incident detail described that a Brea Academy staff member had called 911 asking for officer assistance with a teenage male student, who, the caller had said, was threatening staff with a stapler. A final note on the incident detail, displayed in all capital letters, reminded the responding officers to “Contact the MEU upon Scene Stabilization,” which is, as I noted in the previous section, an automatic, but not determinative, tag attached by dispatch.

As we walked toward the school’s front door, Sergeant Reyes seemed to recognize the building as he shared with me that his division receives frequent calls for service from its staff. The Brea Academy is, he told me, a “school for special kids, you know, the ones with problems that regular schools can’t deal with.” In fact, the school we had arrived at was one of the dozens of special education schools in Los Angeles that educate students with disabilities who cannot be accommodated in a mainstream public school setting. In some cases, these schools will specialize in serving a specific type of disability, such as students with learning disabilities, intellectual disabilities, visual impairments, or the nebulous category of emotional disturbance, while others serve a diverse population of students with different qualifying conditions. It was unclear to me which student population the Brea Academy served, and the LAPD officers involved with the call also did not know, nor were they familiar with the Individuals with Disabilities Education Act that governs these sorts of schools. ⁴ However, regardless of what types of disabilities this school accommodated, Sergeant Reyes’s characterization of the place as a school for students with “problems” set the tone for the LAPD’s encounter with Brendan, the fifteen-year-old student at the center of this particular call for service and became instrumental in framing the subsequent interaction, categorization, and ultimate disposal.

We found Brendan at the back of the school’s interior courtyard facing the stucco wall, his hands cuffed behind his back while Officer Williams conducted a pat-down and stood nearby to observe Officer William’s interview with Brendan. Meanwhile, his partner, Officer Perez, interviewed a school staff member who stood in a nearby doorway just out of earshot. Officer Williams finished the pat-down and instructed Brendan to turn around: “Do you have a diagnosis?” “I don’t know,” Brendan mumbled at the ground. He refused to make eye contact, and it looked as though he had recently been crying. Officer Williams continued, his tone sharper: “Do you see a doctor?” Brendan seemed to sense this shift in tone and looked up: “Yes.” “Do you take medication?” “Yeah, for anxiety.” Officer Williams began to probe on WIC § 5150’s first prong, danger to self: “Have you ever tried to hurt yourself?” Brendan responded: “I was just playing basketball on the roof.” The officer asked: “Were you trying to jump off?” Brendan countered: “No! I was mad, so I was throwing the ball at the railing.” “What made you upset?” “My head was hurting.”

Officer Williams then shifted his line of questioning to WIC § 5150’s second prong, danger to others: “Did someone here make you upset?” Brendan responded: “No! I grabbed a stapler and was playing with it, but I didn’t want to hurt anyone.” Just then, Officer Perez finished his interview with the staff member and joined us to ask Sergeant Reyes if the MEU should be contacted. “Absolutely. Let’s get a SMART unit out here or have them meet us at the station,” Sergeant Reyes decided. Officer Williams and Sergeant Reyes walked Brendan to a small conference room, while Officer Perez called the MEU’s Triage Desk from the courtyard to describe the scene and relay the information they had learned during their field interviews. Officer Perez used his cell phone to relay a condensed description of the call to the MEU’s Triage Desk: “Well, right now, he’s calm. When we showed up, he was just walking around. But the school says he’s emotionally disturbed, and he got aggressive with a teacher, waving a stapler at her.”

Meanwhile, Officer Williams and Sergeant Reyes continued to interview Brendan, who was now sitting but still handcuffed in a chair, and they learned that he had an open juvenile delinquency case and that he did, in fact, know his diagnosis—attention-deficit/hyperactivity disorder. Officer Perez popped in, the phone still to his ear, to ask Brendan if he hears voices (he answered no), and then he returned to announce that the MEU would be sending a SMART team to assess whether Brendan should be taken to a hospital, where a doctor would determine if he met the criteria for an involuntary psychiatric hold under WIC § 5150. Brendan started to cry at the prospect of a hold. He had plans with his dad that weekend and was worried that a hold would mean spending the weekend in the hospital instead or, worse, getting sent back to juvenile hall to serve the rest of his sentence. Officer Williams used the sweatshirt on Brendan’s lap to wipe at his tears and explained that he was not in trouble, even though he was in handcuffs. The handcuffs were only on for “his safety and ours,” Officer Williams explained. Some forty-five minutes later, a plain-clothes SMART team arrived to begin their assessment and relieve the patrol officers from the call. Shortly after the SMART unit arrived, Sergeant Reyes and I walked back to his car, and I asked him what he thought would happen to Brendan: “I don’t know, probably take him to County [hospital] for a hold? This school, you know, I’ve been here at least twice before. Last time, I walked in, and they [the staff] had a big girl pinned in the office, and she was banging her head on the desk. Places like this, you gotta just call the MEU and have them deal with it.”

Patrol officers like Sergeant Reyes quickly become familiar with the places in their division that, like the Brea Academy where this call unfolded, serve a specialized, or “problem,” population. That familiarity manifested—sometimes quite explicitly, as in Sergeant Reyes’s remarks about the school—in the process of categorizing individuals like Brendan as a “problem” for the MEU or for patrol to manage. Here, Sergeant Reyes’s immediate decision to contact the MEU after Officer Williams’s initial interview with Brendan seemed based more on the place at which the call unfolded than on whether the specifics of Brendan’s behavior in the classroom and on the basketball court rendered him a danger to self or others. In fact, Officer Prez had not yet even debriefed Sergeant Reyes about his interview with the Brea Academy staff member. The decision to call the MEU seemed, for Sergeant Reyes, a foregone conclusion as soon as he recognized the Brea Academy as the place where this call was unfolding. Even though his decision to request that a SMART unit respond to assess Brendan was very likely the “right” decision given the LAPD’s mandate that the MEU be contacted whenever a subject is suspected of having a mental illness, especially given the fact that Brendan did eventually reveal a diagnosis to the patrol officers, place nonetheless dominated the encounter to the sublimation of context.

The next excerpt, which draws from a Friday night ride-along in downtown Los Angeles with Sergeant Vega, also highlights how the categorization of a place percolates in the background of a call and bleeds into the process of categorizing the person, though to a very different outcome. Around 10:30 p.m., we responded to a request for a supervisor made by two patrol officers who had picked up a call for service reporting that a woman was running naked in the streets of downtown Los Angeles. The incident detail included the standard reminder to “Contact the MEU upon Scene Stabilization.” As the MHIT instructors teach when they introduce the laws that govern the LAPD’s encounters with the mentally ill, running in traffic due to a mental illness is a textbook example of behavior that would authorize officers to apply for an involuntary psychiatric hold under WIC § 5150. Such a person is both a danger to self since they might be struck by a passing car and a danger to others since those in nearby cars could be injured in an accident caused by striking, or trying to avoid striking, the person. As we drove to the scene, I thought the call, based solely on its incident log description, would be a textbook 5150 for the officers, but, instead, the power of place rendered it a 5149-and-a-half call.

When we arrived on the scene, a woman was already sitting in the back of a squad car, mostly wrapped in a blanket and swinging her legs back and forth over the curb. The two patrol officers who had initially responded to the call for service briefed Sergeant Vega. The woman had been running naked down the street when they arrived on the scene, but they had been able to detain her in the back of their squad car without incident. They were waiting for a supervisor to question her since it was unclear what explained her behavior. As bits and pieces of Mary’s story emerged, the cause of her nakedness and late night run through the streets shifted between mental illness or drug use. Place would soon emerge as the central fulcrum upon which this call’s disposition—and the categorization of Mary—would pivot.

A patrol officer asked Mary if she had been using drugs, and she responded that, yes, she had been smoking methamphetamine earlier. The officers seemed to think that this explained her behavior, and I asked why. Sergeant Vega explained that meth overheats the central nervous system: “She got high and then got hot, so she tore her clothes off.” Sergeant Vega deemed the call a “medical emergency” situation, which, under the LAPD’s guidelines, meant that Mary could be transported to the nearest county hospital in a rescue ambulance and that the police would accompany her until she was admitted through the ambulance bay. A patrol officer turned to request an ambulance, but when Sergeant Vega asked Mary when she had used the drugs, she responded that she had smoked the meth some six hours ago. This detail undermined the sergeant’s certainty that this was just a medical emergency. He explained that six hours was just too long ago for the high to have caused her to still be hot and to explain her nakedness. “Maybe she’s crazy, too?” he wondered rhetorically to me as an aside.

Mary interjected to explain that she had taken her clothes off “for just a second” at the nearby mission where she had been staying and that someone had stolen them. Mary offered no explanation for why she had taken them off in the first place or for why she then started running through the streets upon realizing that they were missing, but her revelation that she had been staying at the mission earlier in the night seemed to stop the sergeant from following up on his suspicions that she may have also had a mental illness. In fact, Sergeant Vega and I had also just come from that mission, which is one of the oldest in Los Angeles’s Skid Row area, where he had decided to give me a brief tour to fill part of the downtime between responding to calls. The mission’s courtyard stays open all night to provide a place to sleep for unsheltered individuals who may be resistant to the idea of a formal shelter placement, and, as Sergeant Vega explained during our ad hoc foot patrol, to the rules, structure, and sobriety requirements that shelters enforce. Sergeant Vega’s construction of the courtyard as filled with addicts and the surrounding Skid Row area as a home to a bustling narcotics trade tilted Mary back into the category of an addict.

Soon, an LAFD rescue ambulance arrived, and its two paramedics jumped out to assess Mary and then to help her onto the stretcher that would take her to the county hospital. The streetlamp fully illuminated her face, and, even though it was still somewhat obscured by her hair, Sergeant Vega exclaimed that not only did he recognize her from his time patrolling the downtown area but that he also remembered her as a repeat naked runner. The paramedics loaded Mary into the ambulance, and, as we walked back to the sergeant’s car, he directed one of the patrol officers to ride with Mary to the hospital and the other to follow in their squad car. The MEU was not contacted, and, as Sergeant Vega told me, she would likely be back on the streets by that morning. Mary, he predicted, would not see a mental health professional during her time at the county hospital, despite the possibility that a mental illness was functioning alongside or even instead of the meth as the cause of the night’s naked run, which under WIC § 5150 could render Mary a danger to herself or others. Here, Mary’s association with the mission, and the meaning of that place for Sergeant Vega, had rendered drug use and mental illness as competing, rather than potentially co-existing, categories. Place, rather than the full context of the call and the unresolved ambiguity of why Mary had been running through the streets, worked to make sense of Mary and whether or not specialized policing resources would be mobilized.

DISCUSSION AND IMPLICATIONS

The issue of how police can better navigate their role as “streetcorner psychiatrists” (Teplin and Pruett 1992) is an evergreen component of criminal justice reform conversations (Engel 2015). Adopting specialized policing strategies like the one at the heart of this case study represents a common policy mechanism to remediate the harms associated with “traditional” police responses to people with a mental illness or in a mental health crisis. In Los Angeles, the LAPD’s MEU and departmental policies are considered a gold standard for how to “better” police mental illness (Office of the Independent Monitor of the LAPD 2009; Council of State Governments Justice Center 2010; LAPD 2016). Under the LAPD’s model, patrol officers are required to call the MEU’s Triage Desk during any encounter that involves someone who may have a mental illness in order to receive real-time guidance from the MEU’s members and perhaps even field assistance from a SMART unit to save patrol officers time, divert people with mental illnesses from jail to treatment, and prevent encounters from escalating (O’Neill 2015). Evaluating the efficacy of these specialized strategies by analyzing, for example, whether or not specific programs can move the needle on metrics like arrest or use of force in a jurisdiction is critical for informing policing policy, but so too is a qualitative understanding of how they unfold on the ground. This article takes a step toward the second question by examining how LAPD patrol officers make choices about whom they categorize as “mentally ill” and when to call the MEU for assistance. The answers to these questions shed light on how police, and the broader criminal justice system, manage social marginality through strategies that hybridize care and control.

My analysis of a series of trouble case calls shows how officers construct “the right kind of crazy” for deploying the MEU by considering the pressure of time and relying on the power of place. ⁵ The fact that officers invoke contextual factors during this process is inevitable given the law and policy that oversee how the LAPD polices “the mentally ill.” California’s WIC § 5150 outlines the three types of scenarios where LAPD officers should categorize someone as having a mental illness and call the MEU’s Triage Desk for assistance. WIC § 5150’s three prongs—danger to self, danger to others, or gravely disabled—are broad and require an individualized operationalization during each and every call for service. This ambiguity of law is compounded by the fluid nature of the category of mental illness itself; officers must decide, for example, that someone is not only a danger to self but also that it is also due to a mental illness. One may certainly be a danger to oneself if, as a MEU officer explained during the MHIT training I observed, he decides to go into a rival gang’s territory alone, but it is because of criminal activity, not because of a mental illness, that he presents a danger to himself. ⁶ In making these choices—about whether WIC § 5150 applies, about who has a mental illness, and about when to call the MEU—patrol officers exercise immense power to shape the material reality of mental illness for deploying the very specialized policing strategies designed to ensure that “the mentally ill” are better policed.

In this article, I identify time and place as key mechanisms through which patrol officers construct the boundaries of mental illness and trace how this process informs how officers decide whether to invoke the MEU. These two processes are inextricably linked—how officers determine whether an individual has a mental illness also determines whether they should contact the MEU—yet they could also function independently. In part, these mechanisms served a constitutive function when officers used time and place to filter an individual in and out of the category of mental illness, as when the power of Skid Row rendered Mary and her naked run the product of drug use rather than mental illness. Mary may have been a danger to herself or others by running through the dark streets of downtown Los Angeles, but by identifying the cause as a medical, rather than mental, illness, Mary did not embody the sort of problem envisioned as solvable via the MEU’s specialized resources. Yet these factors also served a strategic function for officers in constructing mental illness as part of the decision to invoke—or avoid—the MEU, such as when the patrol officers sought to avoid extended MEU involvement in Terry’s case by narrowly operationalizing WIC § 5150’s “danger to self’ prong because categorizing him as being outside the MEU’s purview would have saved them time.

However, though my analysis frames time and place as separate mechanisms, I do not mean to oversimplify the process through which mental illness is constructed. There is likely much to learn by destabilizing my analysis to consider, for example, the relational aspect of these mechanisms. For instance, while I analyze Elizabeth’s story as an example of the institutional pressure of time in constructing mental illness, the encounter may also reveal an equally powerful relational mechanism at work. In fact, when the paramedics first arrived at the call, they approached me first, and we engaged in small talk for several long minutes until we all realized that they thought that I was the subject of the call. The paramedics’ initial assumption that I was the subject of the late night call in South Los Angeles could perhaps bridge the context of the call itself with its subject’s demographic characteristics by pivoting the analysis to my relationship to the place where the call unfolded. In this counter analysis, it could be the “discordant” relationship between my presence as a white female on a dark street corner in a historically Black neighborhood at night that prompted the paramedics to ask me: “what seems to be the trouble tonight” when they first arrived (see Frohmann 1997).

The data here do not permit this level of analysis as I could not interact directly with the subjects of police encounters, but it does suggest that the relational, or even hierarchal, quality of time and place likely matters in constructing mental illness and that it merits further study with data that systematically collect demographic data and personal characteristics from the subjects of the police interactions. This particular vignette also suggests another important set of considerations that likely inform this process as well. Patrol officers, as the encounter with Elizabeth shows, navigate calls for service within a larger bureaucratic field populated by other frontline workers (see, for example, Emerson 1983; Lara-Millián 2014; Seim 2017) and with their past experiences with encountering mental illness in the field in mind (for example, Pifer, 2019). Though not the focus of this article, they certainly matter to how police officers bound the category of mental illness and how they use the MEU’s hybridized resources.

My findings, despite the limitations of my data and the particular focus of my analysis, offer new insight into the mechanics and meaning of the categorization and call disposal processes associated with hybridized policing resources. Here, officers invoked time and space to bound the category of mental illness and the calls that fall within the MEU’s purview. In this way, the MEU serves as another bureaucratic site that patrol officers must navigate as they seek to resolve trouble case calls through a process of “burden shuffling” (Seim 2017). However, not all trouble case calls need to be shuffled somewhere. Under the LAPD’s policy, the MEU only needs to be deployed when patrol decides that a subject may meet the WIC § 5150’s criteria. Thus, my findings shed light not just on how patrol officers shuffle burdens among other frontline workers but also on how they decide what trouble case calls can be solved by some piece of the state’s continuum of managing social marginality and which problems should simply be released by deciding to use neither care nor control. In this way, my analysis of trouble case calls shows that, for patrol officers, the “right kind of crazy” are those burdens that can be shuffled somewhere and the wrong kind represents an unshuffable burden that is beyond the sphere of state governance. ⁷ To be the “wrong kind of crazy” is to be a problem that is perhaps beyond the ability, or interest, of the state’s problem-solving infrastructure.

The categorization and disposal of the trouble case calls analyzed in this study comports with one of the major contributions of policing research: time and place matter to how police decide to dispose of calls involving not only people who have a mental illness but also a host of other police decisions. For example, Steven Herbert’s (1997) fieldwork with the LAPD in the early 1990s showed that officers measured any given incident against their understanding of the law, space, and time in controlling through territoriality. The fact that police officers in my study made decisions about how to resolve the 5149-and-a-half calls I observed through the same processes that the policing literature consistently identifies as salient in police decision making suggests that police are likely to remain police, even when reforms seek to make their function as “streetcorner psychiatrists” intentional by developing specialized strategies that ask them to hybridize care and control.

Thus, in some ways, my findings are not novel, but when they are considered within my larger interrogation of when and how specialized policing resources are deployed, they enable reflection on the project of policing from the “murky middle” as a means of managing social marginality (Brydolf-Horwitz and Beckett 2021). In addition, these findings help contextualize the MEU as a policy reform designed to improve the department’s policing of mental illness because its resources hinge on how patrol officers categorize someone as having a mental illness and when they decide to invoke its specialized policing strategies. Understanding how they exercise this power provides insight into the meaning of mental illness for policing and how—and even if—specialized policing strategies unfold. And, as my findings show, patrol officers make these choices through a process that is, given that they are police officers, entirely unsurprising.

As new approaches to managing social marginality emerge from the murky middle, researchers have paid careful attention to how the hybridization of care and control unfolds on the ground and to what consequences. This hybridization is literal and explicit in my field site, as when a patrol officer calls the MEU’s Triage Desk or when a specially trained police officer and Department of Mental Health clinician arrive on the scene in a SMART unit. At their core, they ask the police to do both, but my analysis of how patrol officers adjudicate trouble case calls suggest that they are likely uninterested in doing either. Perhaps then, when the state hybridizes care and control, it ultimately achieves neither.