American Policing and the Danger Imperative

Abstract

In spite of long-term declines in the violent victimization of U.S. police officers, the danger of police work continues to structure police socialization, culture, and behavior. Existing research, though attentive to police behavior and deviance that negatively affects the public, analytically ignores how the danger of policing engenders officer behavior that harms police themselves. Drawing on ethnographic observations and interviews in three U.S. police departments, this article describes how police are informally and formally socialized into the danger imperative—a cultural frame that emphasizes violence and the need for officer safety—and its effect on officer behavior. As a result of perception mediated through the danger imperative, officers engage in policy-compliant and policy-deviant behaviors to protect themselves from violence. Unfortunately, policy-deviant behaviors such as unauthorized highspeed driving and not wearing a seatbelt, though justified in the name of safety, lead to catastrophic car accidents that injure and kill both police and members of the public. This article concludes with discussion of how seemingly mundane policy deviant behaviors are a reflection of assumptions within police culture that undergird police practices that damage public wellbeing and perpetuate broader inequalities in U.S. policing.

1. Introduction

In testimony to the President's Task Force on twenty-first century policing, the president of the Fraternal Order of Police warned, “now more than ever, we see our officers in the cross-hairs of these criminals” (Canterbury 2015). However, trends in violence against police suggest that policing is growing safer over time. Felonious officer deaths have decreased for half a century (White et al. 2019) and recent analyses find no significant change in patterns of fatal or nonfatal assault on police officers (Maguire et al. 2017; Shjarback and Maguire 2019; Sierra-Arévalo and Nix 2020). Despite the measurable decrease in line-of-duty death and injury over the past half century, officers' concern with the mortal danger of their work continues to shape police socialization, culture, and practice (Ingram et al. 2018; Sierra-Arévalo 2019b). The understanding of policing as profoundly dangerous, in turn, encourages behaviors that damage the legitimacy of police, harm the public, and perpetuate inequalities in the criminal legal system (Stoughton 2014b).

Existing research, however, does little to elucidate the consequences of policing's preoccupation with danger beyond its deleterious effects on the targets of coercive control. Though a large and ever-growing body of research documents inequalities in policing and its negative effects on marginalized communities (Armenta 2017; Rios 2011; Stuart 2016), there is scarce consideration of how police culture and its related behaviors can also lead to unexpected, damaging consequences for police officers themselves. This analytic exclusion of officers likely underestimates the total costs of behavior oriented by the assumption of constant danger and frustrates consideration of how culturally mediated behavior aimed at keeping officers safe can, in fact, lead to officer injuries and deaths.

To elucidate the origins of such safety-enhancing behaviors and their often unseen negative externalities, I leverage insights from research on police, culture, organizations, and the social construction of risk to advance the concept of the danger imperative—a cultural frame that emphasizes violence and the need to provide for officer safety. Drawing on ethnographic observations and interviews with officers in three urban police departments in the US, I provide an empirical account of officers' formal and informal socialization into the danger imperative and how experience mediated through this frame encourages both policy-compliant and policy-deviant behaviors to ensure officer safety. I then highlight policy-deviant behaviors like unauthorized high-speed driving with neither emergency lights nor seatbelts which, though justified as necessary to ensure officer safety, contribute to the injury and death of police officers in high-speed car crashes. Following this analysis of the unintended and counterproductive behaviors shaped by the danger imperative, I consider the broader implications of policing's focus on officer safety for police and the public. Beyond the harm to citizens and police officers resulting from vehicle crashes, I argue that seemingly mundane policy deviant behaviors are a reflection of assumptions within police culture that also undergird policing practices that damage public wellbeing and perpetuate broader inequalities in the US criminal legal system.

2. Policing, Danger, and Deviance

Recognizing the marked historical changes in the structure and implementation of police work (Manning 2011), policing is still work that pits “ordinary men” against the “extraordinary strains…and threats” posed by suspects who may violently resist arrest by police (Westley 1970: xvii). ¹ Though policing ranks well below logging, mining, and construction in terms of worker deaths and death rates (Bureau of Labor Statistics 2014), “policing is unique in that injury and death come not just from accidents, but from job performance” (Moskos 2009: 1). Unlike the miner claimed by a tunnel collapse or the construction worker slain in a fall from a roof, officers contend with the extraordinary threat posed by violent assailants.

Given the centrality of crime and potentially dangerous suspects to the professional mandate of police (Manning 1978), it is unsurprising that officers' concern with violence pervades the policing literature, as does the emphasis on officer safety enshrined in decades of police procedure, policy, and law (Stoughton 2014b). Scholars for over half a century have noted the emphasis placed by police on the danger of their work and the role of danger in shaping police training, how officers interact with one another, and how they interact with the public (Skolnick 2011 [1966]; Van Maanen 1978b; Westley 1970). More recent scholarship confirms that danger continues to be a “ubiquitous cultural theme” in policing (Crank 2014: 160), and that the necessity of ensuring officer safety at all times is a pervasive philosophy “reinforced at all levels of the police organization” (Moskos 2009: 22). Today, the danger of police work continues to occupy a “prominent position within [the] occupational consciousness” of police officers (Loftus 2010: 13).

The emphasis on violence and the danger of patrol is disseminated to officers though formal and informal mechanisms within the police department. In the police academy, defensive tactics and weapons training make up the largest part of academy curricula (Reaves 2015), and these lessons are supplemented with “war stories” from senior officers about the danger recruits will face (Van Maanen 1978a: 297–98). The possibility—even inevitability—of violence is further emphasized in field training. This field training, in addition to assigning experienced field training officers (FTOs) to guide rookie officers, is a mechanism for the transmission of informal norms and practices that are necessary to reconcile inflexible regulation with the uncertainty and limited resources that officers encounter on patrol (Engelson 1999). Outside academy and field training, officers' broader occupational environment is replete with symbols and cultural artifacts that perpetuate the cultural salience of danger in police work. Rare but devastating line-of-duty deaths are commemorated through cultural artifacts like memorial walls, tattoos, and funeral pamphlets, amplifying the danger of policing across time and space (Sierra-Arévalo 2019b).

Together, these various facets of police socialization propagate the prominence of danger in police culture, in turn, shaping officers' understanding and practice of their work. Danger is a key driver of the loyalty and cohesion at the core of policing's esprit de corps, yielding the unwavering support between officers needed to ensure one another's safety on patrol (Bittner 1970: 63). This intragroup solidarity is part of the “working personality” of police and structures how police interact with the public, fostering suspicion of community members (Skolnick 2011 [1966]), and support for aggressive enforcement tactics associated with “traditional” police culture (Ingram et al. 2018). Unfortunately, though solidarity can serve protective functions in the face of danger, it is also implicated in decidedly problematic phenomena like the “blue wall” or “code of silence” that hampers police accountability and helps officers avoid punishment for wrongful acts (Skolnick 2002). What's more, strong in-group ties and a collective emphasis on danger can manifest in aggressive, enforcement-centric practices that characterize a “warrior” approach to policing linked to deviant behaviors ranging from rudeness to the use of excessive and illegal force (Ouellet et al. 2019; Skolnick and Fyfe 1994; Stoughton 2014a). The persistence of such damaging police deviance are part and parcel of departmental socialization designed to keep officers alive.²

However, narrowing analytic attention to these outward-facing acts of police deviance restricts our understanding of danger's effect on police practice to behaviors whose negative consequences directly affect public well-being. As a result, how the foundational preoccupation with violence might mute attention to other dangers in officers' occupational environment is unexplored by existing research on danger and police behavior. To better understand how police officers come to focus on and protect against particular dangers, as well as the inadvertent and even counterproductive consequences of supposedly safety-enhancing strategies, research on police, danger, and culture must be brought into conversation with research on the perception and construction of risk.

3. Risk, the Collective Construction of Danger, and Normalized Deviance

Despite my preference for the term “danger” up to this point, terminological clarity between “danger” and the more general term “risk” is warranted. There is no such field as “danger studies” but instead a constellation of managerially oriented fields encompassing risk assessment, management, and analysis. These fields, largely dominated by economists, define risk with an array of probabilistic curves of potential gains and losses (see discussion by Short 1984). In this quantified paradigm, human behavior and decision made under conditions of “risk” boil down to a balancing act of probability and potential harm.

What is missed in the world of probabilistic risk is that people do not dread or avoid probabilities; people fear things and what those things may cause. As Desmond (2006) discusses in his study of wildland firefighting, prior accounts of risk and individual risk-taking ignore the lived, visceral nature of decisions made when confronting imminent harm. Just as probabilities matter little to a firefighter trying to survive amidst billowing smoke and burning trees, the statistical rarity of death is immaterial to a police officer who is taught to “treat every individual they interact with as an armed threat and every situation as a deadly force encounter in the making” (Stoughton 2014a: 228). In this spirit, I use “danger” to capture the individuals and their actions that stand to do physical harm to officers—the suspects and violence at the core of policing's professional mandate.

This operationalization of danger is useful for considering how culture—the norms, values, and patterned action of individuals embedded in groups—influences collective understanding of what constitutes a threat and appropriate strategies for addressing it. Which dangers are to be confronted, ignored, or avoided (and how to do so) does not exist a priori and is instead shaped by a group's environment. Firefighters learn and practice the use of their bodies and axes to fight roaring flame; police officers socialized to survive their shift practice perpetual suspicion attuned to “signs indicating a potential for violence and lawbreaking” (Skolnick 2011: 41 [1966]). In short, the interplay of culture and context influences group members' perception, their collective definition of what defines danger, and subsequent action of group members aimed at protecting themselves and the group from harm (Douglas and Wildavsky 1983).

For police, the nexus of culture and occupational context is also embedded within the organizational environment of the police department. As with organizations more broadly, collectively defined understandings of danger among police are influenced by organizational norms and values that set bounds for what is a reasonable solution to an organizational problem (DiMaggio and Powell 1983; Vaughan 1996). As police address issues ranging from low-level misdemeanors to violent crimes, they must negotiate the tension between their need to keep themselves safe from violence and the organizational constraints of limited time, large call loads, and departmental policy (Lipsky 1980). Officers' shared conceptions of danger, the harsh realities of the street, and the demands of the police organization engender perceptual and behavioral strategies that officers believe allow them to meet the cultural, occupational, and organizational requirements of their work (Van Maanen and Barley 1984).

However, police officers balancing these varied demands amidst conditions of uncertainty will deviate from organizational rules and engage in “normalized deviance” to solve a collectively defined problem (Vaughan 1996). The threat of violence is one such problem that frustrates adherence to departmental policies:

Danger typically yields self-defensive conduct, conduct that must strain to impulsive because danger arouses fear and anxiety […] As a result, procedural requirements take on a “frilly” character, or at least tend to be reduced to a secondary position in the face of circumstances seen as threatening. (Skolnick 2011: 62 [1966])

Normalized deviance aimed at solving the exigencies of the moment, like all individual action, comes with downstream consequences often unforeseen by individuals whose decisions are a response to problems “located in one context of time and space” (Giddens 1986: 14). As a result, both sanctioned and deviant behavior focused on addressing a specific danger necessarily ignore others and leave individuals imperfectly protected from the full range of environmental dangers (Kasperson et al. 1988). And though prior research shows that officers alter their behavior in response to the threat of potential violence, research on police officers' normalized deviance and their behavioral responses to danger does not consider how normalized police deviance, in addition to breaking organizational rules, might unintentionally expose officers to danger other than that posed by violent individuals.

4. Conceptualizing the Danger Imperative

To address this gap in our understanding of danger and its effects on police behavior, there is need for a concept flexible enough to account for the danger on which officers focus their attention and action, that which is ignored or minimized, and the unintended consequences of these behaviors. The cultural frame, defined as a cognitive lens through which individual experience is filtered to “highlight certain aspects and hide or block others” (Lamont and Small 2008: 80), provides such flexibility. I propose that police officers employ such a frame—what I term the danger imperative—that emphasizes potential violence and the need to provide for officer safety at all times. Unlike past research that centers on officers' use of perceptual typologies to differentiate potentially lethal threats from harmless civilians or run-of-the-mill “asshole[s]” (Skolnick 2011 [1966]; Van Maanen 1978b), the danger imperative frame provides analytic leverage for understanding how officers are socialized to practice their work as dangerous in particular ways and not in others.

Though a cultural frame does not cause behavior, it filters perception as to make certain behaviors more or less likely given a set of social facts (Lamont and Small 2008). For police officers, the danger imperative's emphasis on violence mediates perception in a way that encourages particular safety-enhancing behaviors, (re)constructing an organizational culture and lived experience that highlight the mortal peril faced by officers on patrol. As officers engage in behaviors tailored to protect them from violence, their behaviors are observed, learned, and echoed by other officers socialized into shared appreciation for the need to ensure officer safety. Officers who perceive their world through the danger imperative are thus more likely to orient their attention and action to address violence rather than other environmental dangers, increasing exposure to these unmitigated threats.

In addition to skewing officers' attention toward violence and away from other dangers, perception mediated through the danger imperative can encourage individual behavior that conflicts with organizational rules. Similar to the adaptive use of violence to ensure personal safety in communities distrustful of police (Anderson 1999; Kirk and Papachristos 2011), perception that highlights the threat of violence can normalize deviant behavior seen as necessary to survive the “collectively constructed cultural reality” of dangerous police work (Vaughan 1996: 65). Viewing their occupational environment through the danger imperative, officers participate in and reproduce the cultural reality of work than can devolve into violence at any moment. In turn, deviant rule-breaking is relabeled as a necessary police practice that allows officers to navigate the constraints of their work and the police organization.

The remainder of this article presents evidence from multiple police departments to show how officers are socialized into the danger imperative and the behavioral consequences of perception mediated through this frame. Following a description of my field sites and methodology, I illustrate the formal and informal socialization processes that perpetuate the danger imperative among police. I then describe how perception attuned to the threat of violence encourages policy-compliant and policy-deviant behaviors designed to keep officers safe, including behaviors such as unauthorized high-speed driving and the eschewing of mandatory seatbelts that unintentionally contribute to line-of-duty deaths. Finally, I consider the interconnectedness of officer safety and public wellbeing, then discuss links between the danger imperative, police practice, and persistent social inequalities.

5. Field Sites and Fieldwork

Ethnographic observations and interviews were gathered across three urban police department in the US. Table 1 provides descriptive information on the police departments in Elmont (EPD), West River (WPD), and Sunshine (SPD).³ The three departments included in this study are located in unique regions of the US Elmont is a small city of less than 150,000 along the eastern seaboard, West River is home to some 400,000 on the west coast, and Sunshine is a city of approximately 500,000 in the arid Southwest. Although these departments vary in terms of size, they are all within the top 1% of local police departments in the US by number of sworn, full-time officers they employ: the Elmont Police Department has approximately 500 sworn officers, the West River Police Department close to 700, and the Sunshine Police Department has nearly 900. These sworn officers are primarily assigned to the patrol division of each department and are responsible for responding to calls for service concerning issues ranging from low-level disorder to serious violence. All three departments are overwhelmingly male, reflecting the gender imbalance of policing in the US writ large (FBI 2019b). Finally, these departments differ in their racial and ethnic makeup, with West River being the most diverse department and Sunshine being the least diverse of the three.⁴

Table 1. Department Characteristics

Data collection took place from 2014 through 2018, though I began interacting with the EPD as early as 2012. My observations of the EPD were spread across this time period, while my time with the WPD and SPD were condensed into intensive 2-month data collection periods in the summer of 2015 and 2016, respectively. An additional month-long revisit to West River took place in the summer of 2017 and phone calls, text messages, and emails were exchanged with officers across field sites throughout the study period. Access to both the EPD and the WPD were aided by my ongoing involvement with violence reduction efforts, allowing me to meet patrol officers as well as the police executives that would ultimately approve my ongoing contact with patrol officers. Access to the SPD was approved by departmental leadership only after I provided a letter of recommendation from a high-ranking WPD officer who could speak to my research activities in their department.

Ethnographic observations and interviews were predominantly gathered in the course of “ride-alongs” in which I accompanied officers on patrol. Table 2 shows observation hours and interviews collected in each police department. Ride-alongs covered all areas of each city, with rides in Elmont oversampling Thursday through Sundays—the highest call volume days of the week—and higher call volume shifts in light of the shortened, four-hour blocks to which the EPD restricts ride-alongs. I rode for full shifts in West River and Sunshine, staggering shifts to cover all days of the week and all times of the day; SPD shifts are ten hours long and WPD shifts range from ten to tweleve hours. Jottings of observations were recorded in a small notepad or in a notetaking application on a smartphone and expanded after each ride-along into more expansive and detailed field notes in OneNote.

Table 2. Data Collected per Field Site

Following the example of recent sociolegal scholarship (Hureau and Braga 2018; Sierra-Arévalo 2019a, 2019b), this study collected “ethnographic interviews” that follow the style of an “informal conversation […] infused with ethnographic elements” (Hureau and Braga 2018: 520; Spradley 1979). These ethnographic interviews were recorded via smartphone or audio recorder with participant consent and subsequently transcribed. In contrast to more formal interview methods with a predetermined protocol, this interview method allows for the content of interviews to be molded to the unique positionality or experiences of an individual and their immediate context. This fluid style of ethnographic interview is especially useful for studying police as they allow the researcher the flexibility to engage in informal conversations and build rapport, a decided advantage given police officers' weariness of outsiders (Rojek et al. 2012).

Analysis of observational and interview data followed an abductive approach in which data collection and analysis proceeded alongside and in conversation with one another (Timmermans and Tavory 2012). As observations and interactions were collected via field notes and recorded interviews, the author inductively reviewed these data and wrote research memos designed to “record interpretations and incipient patterns emerging from the concrete realities of the social worlds of research sites” (Lempert 2011:247; Glaser and Strauss 1999). Developing themes, new questions, and emergent theories from this analytic process informed the author's observational focus and questions during subsequent ride-alongs and interviews with police officers. Data from new observations and interviews were then reviewed and compared with existing data and research memos to further specify the focus of continued data collection and theory formulation. This recursive, back-and-forth between fieldwork, data, and theory supported “theoretical sampling” that allowed for in situ “testing” of initial theories with new ethnographic observations in the field (Stuart 2018: 220–22). This approach, combined with a multisite design that expands on classic and contemporary single-site studies (Brayne 2017; Skolnick 2011 [1966]; Stuart 2016; Westley 1970), enabled the refinement of research questions and theoretical development to explain common cultural understandings across disparate departments.

Even with formal permission from department leadership in hand, access and rapport were continually negotiated across my field sites. When I first met officers during lineup—a preshift meeting in which officers receive assignments and are appraised of activity during prior shifts—a lieutenant or sergeant would usually let me introduce myself to the assembled officers. I would explain to officers that I wanted to learn about policing firsthand and that I valued the perspective of line-level officers often missing from media coverage of policing issues. I also used this introduction as an opportunity to inform officers that I would not be sharing their names or the name of the department in any written work. Multiple officers expressed interest in my research and even went so far as to espouse thanks for being given an opportunity to share their perspective.

Of course, some officers were suspicious of me and my motives. When interacting with skeptical officers, I leveraged past experiences to build trust and rapport. For example, when officers probed about my family background or commented on how expensive my education must have been, I would use this as an opportunity to mention my working-class upbringing and my public-school education. This served as a way to align myself with the experiences of many officers as much as distance myself from the “anti-cop” ideology officers associated with my university, academia, and the news media. As I spent more time with officers, I was also able to share my own accounts of patrol. This story sharing, in addition to being a feature of police “canteen culture” and a mechanism for forming bonds between officers (Waddington 1999), provided evidence of my familiarity with police work by showing I had, to some degree, “been there” alongside officers (Geertz 1988). With regard to my positionality, being a young, heterosexual male allowed for relatively easy movement through the male-dominated, heteronormative environment of the police department. More instrumentally, my Spanish-English bilingualism made me a useful asset to officers who needed a translator while interacting with Spanish-speaking suspects, victims, and witnesses.

The following presentation of results weaves observations and interactions recorded in field notes with audio recorded interviews. Following the example of other ethnographic work that combines similar data (Contreras 2012; Sierra-Arévalo 2019a, 2019b), I denote interactions and conversations recorded in field notes with the use of italics and leave data from interview transcripts in plain text.

6. Formal Socialization and Policy-Compliant Behaviors

Formal, organizationally mandated training is an important mechanism for the socialization of police officers. For recruits, the police academy is their official introduction to the culture, regulations, and practice of policing. Following graduation and being “cut loose” on the street, periodic in-service training provides officers a refresher on underlying assumptions of dangerous police work and the tangible skills officers can (and should) use to stay safe on patrol. Such tactics comply with department policies that set the bounds for organizationally approved strategies to ensure officer safety.

6.1 Training Danger

Nowhere is the role of formal training in officers' socialization into the danger imperative clearer than in use-of-force training. During Sunshine in-service training, I observed officers go through scenarios in the PRISim judgment and use-of-force simulator, a “shoot-do not-shoot” simulation system that projects call-for-service video scenarios that range from a suspicious person to an armed hostage taker. Each video has multiple “branches” that can be selected by a training officer to dynamically alter scenarios that, per the PRISim's manufacturer, expose officers to the “judgement calls, indecision, sudden fear, partial understanding, blind side surprise and eye-blink response […] that condition[] the trainee for survival” (Cubic Defense UK 2019).

I observed Officer Stuart (SPD) take part in one scenario that was controlled by Officer Baker, the academy training instructor in charge of administering PRISim training. In the scenario, Officer Stuart arrived on scene to a landlord-tenant dispute between two men outside a one-story house. Raising his voice over the argument concerning rent, Stuart told both men to calm down and then instructed the tenant at the bottom of the steps to come with him and step away from the porch. As Stuart continued to call to the tenant, the cowboy-booted landlord barked, “Deal with this guy, because if you don't deal with him I will!” The landlord stormed back into the house and Stuart commanded him to come back out immediately. A few moments later the man reappeared on the porch, only this time carrying a shotgun. The tenant fled at the sight of the weapon and Stuart immediately unholstered his pistol and took aim at the landlord, yelling, “Drop the gun!” Less than two seconds later and without further command, Stuart fired several shots that struck the landlord.

The image on the screen froze, two blue dots showing where Stuart's infra-red pistol shots landed and where bullet holes would've blossomed had he been firing his department-issued Glock. I asked Officer Stuart what he was thinking when he chose to shoot:

He made the choice to go back in and back out with a gun. Him walking out with a gun is already a threat. I gave him a verbal command to drop it, he didn't, and at that point I would fear for my life.

Officer Baker, who had been in control of the PRISim scenario's branches, concurred with Stuart's explanation and viewed his decision to use lethal force as a reasonable (i.e., legal) response. This assessment was especially notable given that Baker, unbeknownst to Officer Stuart, had already selected a branch in which the landlord would not raise his weapon and would, instead, surrender and place his shotgun on the ground. Baker explained:

I don't have a real problem with that [shooting]. The idea on this scenario is that the guy doesn't pick the gun up at you, but then again, who walks in when two cops are standing out there, grabs a shotgun, and walks back out again? With a gun?

Though it was not the one selected, the other scenario branch—in which the landlord does raise his weapon—represents a potential outcome that officers must plan for and safeguard against at all times. That the landlord did not raise his weapon does not change that he could have raised it and fired at any moment. In Baker's eyes, this deadly possibility further justified Stuart's decision to use lethal force.

[The scenario] can go two different ways. When he pulls the shotgun up, how long does it take? Not even a half a second. It's about a quarter to a half a second for him to go from that position to getting that shotgun downrange … I won't critique that shot. That was reasonable.

Officers also learn to appreciate the deadly stakes of their work in training not explicitly geared toward the use of force, such as in training scenarios that use academy instructors to play drivers or suspects with whom trainees interact. After a ten-hour shift in late-July, I met Officers Alonzo and Diggler (SPD) for some Saturday night beers. I mentioned a video I had seen in which a South Carolina State Trooper shot an unarmed, Black motorist who leaned into his car to retrieve his driver's license. Officer Diggler, an academy instructor, suggested that this seeming overreaction is tied to academy training that teaches recruits that “Everybody wants to murder you.” ⁵

He described one such training scenario in which he played the driver of a vehicle that contained a sims gun (simulation pistol) in plain view. Though ostensibly designed as a simple car stop, the scenario's deeper goal is to teach trainees to be aware of the contents of all vehicles they stop, to address the presence of weapons, and the costs of not doing so. As Diggler explained, “If [the trainee] doesn't see [the gun] you kill the cop… if he doesn't see the gun, when he walks back to his car you get out and you fucking murder him…But the problem is that you put them into that mindset.” Alonzo, seated next to me, distilled this mindset concisely: if you do not heed your training and plan for violence, “You're gonna die.”

Virtual and real-world simulations teach officers to view their work through a lens that emphasizes the possibility of violence, be it in a confrontation with an armed suspect or in the course of a routine traffic stop. Though in a controlled environment, such training is vital to a socialization process that inculcates and perpetuates collective understanding of the unpredictable danger of patrol. Interpreting their work through the frame of the danger imperative, officers understand that survival on the street demands their assiduous preoccupation with violence and constant consideration of the worst possible outcome of interactions with the public.

6.2 Video and Vicarious Danger

Formal socialization into the danger imperative also uses body or dashboard camera video of real-life fights, shootings, and car stops “gone bad.” As one EPD officer explained, these videos are intentionally graphic, painting a grim picture for recruits:

We see a lot of videos of officers dying, officers calling out for help. Their mic is keyed in and we're hearing their last breaths, their last words. It hits home.

On a bitterly cold graveyard shift in mid-January, Officer Michaelson (EPD) had a similar explanation for the videos he was shown in the academy.

They show a bunch of videos of stops—dash cams, body cameras—to show you how things can go wrong really quickly. They show you videos to…I don't know…I guess to show you not to make any assumptions. To not let your guard down.

He provided a detailed description of a dashcam video of a traffic-stop he was shown years ago:

…it's a state trooper or a sheriff, I don't remember. But the guy gets out of his car, he's dancing in the middle of the street telling the officer to shoot him, and then he gets up in the officer's face screaming at him. If you get up in my face like that, being aggressive, I'm going hands on. But this officer didn't and the suspect goes back to his car and starts loading a rifle. The whole time the officer is screaming at him to stop, to get down, but the guy won't listen. Then he comes back with a rifle out, the officer is telling him to put it down, and then the guy starts firing. The officer tries to take cover but the suspect keeps shooting. He eventually reloads then comes around the back of the car and executes the officer. It was crazy.⁶

This video—described in a popular police magazine as “forever seared in the minds of police officers across the world” (Law Officer 2017)—is of the 1998 murder of Kyle Dinkheller, a deputy of the Laurens County Sherriff's Office in rural Georgia (see Figure 1). Multiple officers across the EPD, WPD, and SPD recalled seeing this video even though the incident occurred over two decades ago and many hundreds of miles away. Despite this temporal and geographic distance, Deputy Dinkheller's murder provides officers with a poignant and vicarious experience of the brutal speed with which the mundane parts of police work can snap to lethal violence. Though officers today have no direct connection to this more than 20-year-old murder, videos of such line-of-duty deaths emphasize the violence that all officers, regardless of where they patrol or when, must always be prepared to face.

Figure 1. Screenshot of Deputy Kyle Dinkheller's Dashcam Video. Note: Individual pictured is Deputy Dinkheller's killer, Andrew H. Brannan.

Officers are also shown videos during lineup as part of “lineup training” that periodically occurs before officers are sent out on patrol. During one lineup in West River, the lieutenant acting as watch commander announced she was going to show officers a video of an incident in southwestern city. She instructed the room, “Pay attention to issues of officer safety. What can we learn from it? Pay attention to everyone involved.”

What media accounts say began as a dispute in a Walmart bathroom between a customer and Walmart employee eventually culminated in both a suspect and an officer nonfatally shot, one suspect killed.⁷ The video was utter chaos, the audio a mixture of screams, grunts, curses, and “Get on the ground!” as punches, pepper spray, TASERs, and eventually firearms were used by officers to gain control of the situation. My field notes from that lineup read: “By the end [of the video] I couldn't even see if someone had been shot, much less who did the shooting.”

At the conclusion of the video, the lieutenant summarized the events shown, emphasizing that this incident could have been avoided if officers had properly utilized tactics learned in academy training: “Control the situation, keep people separated. These officers didn't and look what happened: a brawl.” She concluded lineup the way watch commanders always did before officers went out on patrol: “Be safe.” ⁸

Importantly, lineup training differs from academy training in that it does not focus entirely on recruits. From the newest officer still in field training to the saltiest veteran, everyone who will be answering calls that shift is present for lineup. The presentation of violent videos to a multigenerational police audience underlines that socialization into the danger imperative is not restricted to an officer's time in the academy. On the contrary, rookies and veterans alike receive formal, organizationally backed reminders that reinforce the necessity of interpreting their work through the danger imperative. These reminders make clear that the cost of not doing so is simply too high. As stated by Officer Gutierrez (EPD) while driving around the city between calls for service, “Always expect the unexpected. It could be an ambush. You never know. The smallest BS call could end up going bad. When you're not alert, that's when you get hurt.”

6.3 Policy-Compliant Tactics

In addition to encouraging the perceptual preoccupation with violence, officers' formal socialization into the danger imperative imparts organizationally sanctioned, policy-compliant tactics for ensuring officer safety. After giving a verbal warning to the driver of a sedan with overly tinted windows, Officer Landry (WPD) parked outside a mechanic's shop to document the stop on his in-car computer. As he wrote the boilerplate report, he explained several tactics he learned in training for use in pedestrian and vehicle stops. When stopping a pedestrian, for example, Landry assumes a proper “POI” (position of interrogation) akin to a relaxed boxing stance. In such a position, an officer's gun-side hip is angled away from the subject so the officer's firearm cannot be grabbed; the officer's hands are ready near their chest so they can quickly push, block, or punch. During vehicle stops, he considers factors like his vehicle's position in relation to the stopped car and asks himself, “Are your spotlights on? Do you know where you are? Do you have a cover unit?” ⁹

Another tactic used during traffic stops is to touch the trunk of the stopped vehicle before approaching the driver's window. As explained to me by Officer Menendez (WPD), this academy-trained technique is designed to make sure that the trunk is closed and that there is not a gunman waiting to ambush the approaching officer. When he could not think of a particular instance in which an officer had been attacked by someone hiding in a trunk, I asked whether he would admit that the probability of there being an armed gunman hiding in a car's trunk was likely very low. He responded:

It's low but that's how we're trained. We train for the worst-case scenario; we don't train to stop granny. We train to expect someone to have a gun, to try and hurt us. It doesn't happen often but it does happen. People do fight you.

EPD officers are also socialized to expect a gunman in the trunk of vehicles they stop. Officer Richardson (EPD) recounted that, in addition to learning to touch the trunk when first approaching a vehicle, his training emphasized the need to guard against a hidden gunman throughout the stop.

Especially when someone says their wallet or registration is in the trunk, you have them unlock the trunk and then you have them step back, you don't let them open the trunk. Because, if they open the trunk, you know Stockton [an academy training officer] is going to be in the trunk with a gun! [Laughs] That's how they train you. Even though it's highly unlikely, when I open a trunk I'm always thinking someone could be in there.

Should an officer be attacked during a vehicle stop, touching the trunk serves another important function: leaving an officer's fingerprint on the car. This is important, Richardson explained, because if someone “shoots and kills you and drives away… My fingerprint is on the car now… so if they find the car they'll have my print on that car.” Touching the trunk, then, does more than allow an officer to check for a potential threat; it provides a means for officers to ensure that, should the worst happen, their killers are brought to justice.

Together, these various policy-compliant tactics comprise a stock of safety-enhancing behaviors that officers incorporate into their practice to ensure officer safety during interactions with the public. Turning on your spotlight, keeping your gun side back, and touching the trunk are—when looking through the frame of the danger imperative—reasonable, even necessary tactics for protecting against potential violence. Even though officers admit that an armed gunman waiting in the trunk of a vehicle is unlikely, their perception of their environment through the danger imperative continuously circumscribes and guides their behavior in ways they believe will ensure they return home alive.

7. Informal Training and Policy-Deviant Behavior

Many of the safety-enhancing tactics used by police are relatively mundane and unlikely to harm the officers or citizens involved. Leaving a fingerprint on a car's trunk, keeping one's firearm away from a suspect, and letting dispatchers know where a stop is occurring are relatively low-cost, policy-compliant tactics for ensuring officer safety. Supplementing these formally taught techniques are informally learned behaviors that directly contravene departmental policy. Such policy-deviant behaviors are meant to help officers keep themselves and their fellow officers safe but, in fact, contribute to injuries and deaths on patrol.

7.1 Unauthorized Code 3, Code 2½, and Providing Backup

Though the threat of violence is front and center in police socialization and practice, the policing profession has increasingly come to recognize the risk of injury and death as a result of vehicle crashes during high-speed driving. As a result, contemporary police departments frequently employ policies to restrict high-speed vehicle pursuits and nonpursuit emergency driving (Walker and Archbold 2013). In what is known more commonly as driving “code 3,” officers exceed the speed limit and utilize their lights and sirens to quickly reach calls in which there is an immediate threat to life, such as a robbery or shooting in progress. The policies of Elmont, West River, and Sunshine restrict code 3 responses to when officers are expressly dispatched to an emergency call or, if officers are not assigned to that call, when officers request and receive a superior's approval.

Despite these policy restrictions, officers frequently drive code 3 without approval. Officer Morales (SPD) explained his department's policy to me in detail but also admitted it is no guarantee of officers' compliance:

By GOs, general orders, you're able to go code 3 if you come up on the radio and say, "I'll be en route code 3." But you cannot go code 3 to anything if there's already two units going code 3; it's only max two units code 3 to a scene. Same thing in a vehicle pursuit: you've got a lead car, back car, and that's it, going code 3, lights and sirens. Does that happen? [Laughs] No.

On the way to drop off Officer Roland on his walking beat in Elmont, Officer Vance explained that he flouts the code 3 policy in the interest of getting to serious “priority 1” calls quickly. “Yeah, you're supposed to [get approval], but I don't. There's different rules for it. If it's a serious call, a priority 1, I'll go code 3…I mean, if it's a priority 1 that's really far away I might come on the radio [to notify dispatch].” Chiming in from the back seat, Roland gave his own perspective, “I go [code] three to all priority 1 [calls]; I never ask for approval. I might get jammed up if I get in a 22 [vehicle collision] but that's on me.” I asked him why he would knowingly break the policy if he knew it might mean discipline in the event of a car accident. He responded, “Seconds matter if it's a hot call. Getting there fast can keep someone alive and you'd want your buddies to come back you up if it were you, right?”

In officers' view, that they could be the difference between life and death for another officer overrides the potential disciplinary consequences of driving code 3 without approval and the significant danger such driving poses to them, other officers, and other motorists. As officers navigate work viewed through the danger imperative, they reify the collective responsibility for ensuring officer safety and affirm unauthorized code 3 driving as a useful, policy-deviant strategy to stay safe.

In addition to driving code 3 without authorization, officers also use a technique—referred to as driving “code 2½”—that allows allow them to simultaneously race to fellow officers' aid while abiding by at least part of the code 3 policy. In contrast to code 2 driving that stipulates officers should obey all normal traffic laws, code 2½ is an informal middle ground in which officers exceed the speed limit but do not consistently use their lights and sirens. During a daytime shift in West River, Officer Mizel drove code 2½ to provide backup for an officer that was responding to a report of an assault on a homeless man. Though he was driving well above the speed limit, he did not leave his lights and sirens activated as this would be driving code 3 without approval. Instead, he used his lights and sirens selectively to make cars move over or to “break” an intersection when the light was red. Once traffic cleared or he was through an intersection, he would deactivate his lights and sirens. As Mizel explained, he knows that he cannot go code 3 but he still needs to provide backup quickly for his squad mates. “I can't go full lights and sirens but you'll drive a little more aggressively, go a little faster. You don't want to leave your teammates out there by themselves.”

In Mizel's view, policies designed to minimize high-speed car accidents do not change that other officers expect him to provide backup quickly nor his expectation of other officers to do the same for him. Unlike the cases I observed of officers using their lights and sirens to do little more than avoid waiting at a traffic light, responding to “hot calls” where violence is imminent or already in-progress is rooted in officers' shared understanding that they must protect one another on patrol. In order to meet this collective demand for officer safety while not explicitly driving code 3 without approval, Mizel drives at high speed without the lights and sirens designed to alert other motorists and prevent car accidents.

I asked Officer Cisneros, an EPD academy instructor, why officers willfully ignore policies designed to protect them. He explained that officers' in-the-moment assessment of safety-related exigencies produce tension between departmental policy and officers' collective need to protect one another:

I think officers to a degree are looking at the immediate, you know. I need to protect my brother or sister officers. We're the thin blue line. We're the ones that are on the front line. We're not behind a desk, we're at the front line. If this goes wrong, this is going to affect me and him [fellow officer]. I can make that change.

Despite potential punishment for flouting departmental policy and the danger of a catastrophic car crash, perception filtered through the danger imperative biases officers' behavioral calculations to focus on the threat of violence. In lieu of strict adherence to organizational rules, officers engage in policy-deviant behaviors like unauthorized code 3 and code 2½ driving to ensure they provide speedy backup to other officers from whom they expect the same aid. Unfortunately, this collective responsibility for officer safety and its associated policy-deviant behaviors, though intended to protect officers, places them in harm's way.

7.2 Seatbelts as Safety Concerns

In addition to the unauthorized code 3 and code 2½ driving, officers also engage in the policy-deviant behavior of not wearing their mandated seatbelt. This behavior specifically endangers officers, in large part because it exacerbates the danger of their frequent high-speed driving.

During one night on patrol, Officer Estacio (EPD) and I listened to the radio as officers in another district reported they were following a car that fled when they attempted to stop it for a traffic violation. “Let's see if we can get in the mix,” said Estacio, and immediately sped off toward where radio traffic indicated a potential chase was developing. Without wearing his seatbelt, Estacio accelerated to speeds upwards of 60 miles per hour through residential areas, past a hospital, park, and school. Within a few minutes a supervisor came over the radio and, when officers could not articulate probable cause that the driver was guilty of anything more than a traffic violation, denied permission to continue the pursuit.

After the would-be pursuit, Estacio explained why he, like most officers, does not wear his seatbelt.

I'd say 90% of us don't wear our seatbelts. It's just too much for us. Tactically, some places the cops wear the vests or the carriers that have all their gear in the front or on their chest, then it's easier to wear the seatbelt. But here, look at where my holster is [points to pistol holster that is obstructing the seatbelt buckle]: it's too much. I'd say only 10% actually wear [a seatbelt].

Implicit in this rationalization is the belief born of perception through the danger imperative that, at a moment's notice, an officer might need to defend their lives. In anticipation of this violent hypothetical, officers like Estacio leave their seatbelt off to improve access to their firearm.

This policy deviant solution to the problem of officer safety is not attributable to individual-level nonconformity alone. Instead, such policy-deviant behavior is perpetuated through informal interactions among officers that transmit and maintain the “hidden curriculum” of policing (Engelson 1999). In West River, Officer Garner explained that though she now wears her seatbelt after being reprimanded by a sergeant many years ago, a veteran officer informally taught her early in her career to not wear her seatbelt.

Garner: It's an officer safety issue. The seatbelt might prevent you from being able to get to your gun or your spray, or it could snag on your belt if you're trying to get out of your car quickly to chase a suspect.

MSA: And you were taught this in training?

Garner: Well, not officially. Department policy is we always wear our seatbelts, but unofficially we're told not to. First day on the street it was “Forget it.”

Other female officers eschew their seatbelts based on the same officer safety concerns voiced by male officers. For example, Officer Herrera (SPD) leaves off her seatbelt because, “…it takes a really long amount of time to unbuckle your seatbelt and jump out of the vehicle […] If someone's gonna shoot me, they're gonna shot me while I'm getting out of the vehicle.” Officer Willis (WPD) explained she does not wear her seatbelt because, “you don't want to get stuck in your car. If I need to get out quickly the seatbelt can get caught, or it can get in the way if I have to get to my weapon.” Similarly, seatbelt usage did not map neatly onto differences in age, race, or police experience. For example, Herrera is a twenty-six-year-old Latina with three years on patrol and Willis is a Black woman in her late thirties with nearly fifteen years of law enforcement experience. The link between officers' preoccupation with violence and their decision to break departmental policy requiring a seatbelt was present across multiple departments and officers of varying demographic characteristics.

Importantly, officers' eschewing of seatbelts is not simply born of irrationality or ignorance. On the contrary, officers are cognizant of the risks involved with high-speed driving and some actively enjoy its inherent dangers. Officer Doyle, a twenty-five-year-old officer in West River, was one such officer. When I asked him what his favorite part of police work was, he immediately answered, “Car chases. I love car chases.” Other officers spoke of an “adrenaline addiction,” the “rush” and “fun” of dangerous work like high-speed driving. Officer Sylvan, a seventeen-year EPD veteran, colorfully summarized his early-career mindset and that of young officers like Doyle whom he'd observed in his time on patrol: “Holy shit! I've got a huge fucking hard on of adrenaline, call to call, fucking fighting, getting shot at, driving 100 miles an hour […] That's as living as you get.”

Though he enjoys high-speed driving, Doyle is also aware of the potentially catastrophic consequences of this thrill-seeking behavior, especially without a seatbelt. He explained that though he might forget sometimes, he did try to remember to put his seatbelt on given the danger of a high-speed crash.

I try to put it on when I'm going code 3 [lights and sirens] […] Sometimes I forget but that's my own personal liability. I understand that if I got into an accident the department is not going to cover me. […] I definitely put it on during car chases when I'm going code 3. Anything where I'm going hella fast.

Nonetheless, Doyle mobilized the same violence-centric logic used by other officers when explaining that his seatbelt at times prevented him from quickly exiting his vehicle to address an armed suspect. Doyle's perception through the danger imperative highlights the necessity of immediate and unencumbered reaction to a perceived threat, simultaneously muting the danger of high-speed crashes sans seatbelt and amplifying that of potential violence.

There's times where I'll be driving and the next thing you know I'll be like, “Oh shit, that dude's got a fucking gun!” I'll stop [mimics tires screeching], try to get out—fuck. Stuck on the seatbelt […] I'd rather just be able to jump out on people, you know. If I have to, be able to jump out of this deathtrap of a car.

Doyle did wear his seatbelt during some of the ten hours I was with him on patrol. True to his words, however, he also left it off while driving nearly 100 miles per hour to support a developing car chase almost ten minutes away. He never caught sight of the vehicle being pursued by other officers.¹⁰

Finally, officer's policy-deviant behavior is encouraged by direct exposure to violent threats experienced through danger imperative. Near the end of a graveyard shift in Sunshine, I sat with Officer Frels as he finished paperwork that had accumulated over the course of his shift. Seeing that he had not worn his seatbelt throughout the shift, I asked Frels whether he thought being shot or being hurt in a car accident was more likely. He responded, “Um…statistically I would say car accident.” I pressed and asked why, then, he chose not to wear a seatbelt. He described an encounter in which a young man with a knife attempted to force Frels and another officer to kill him.

I am sitting at the light on Salado and Wayans […] all of the sudden I see out of the corner of my eye some guy running off the median towards my car. […] It's a kid, he has a knife in his hand, has his arms [held out to his sides], he's like, “Fuckin' kill me, pigs!” […] So I fucking draw [my pistol] and I said, “Don't come any fucking closer or you're dead!”[…] I have my seatbelt on at the time, my foot's still on the break, I'm in drive, so I'm stuck in a real shitty position to have to fucking react to this fucking kid […] So, I open my door and transition [my pistol] to my soft hand [non-dominant hand] […] I have to throw my car in park, then I have to take off my seatbelt, now my seatbelt is stuck on my off hand. So now that the car is in park I have to…switch back to my shooting hand, slide my other hand out, then put both hands back on my firearm.

The young man was eventually subdued by a backup officer's TASER and taken into custody. This encounter, though resolved without significant injury to officers, suspect, or bystanders, fundamentally affected Frels's subsequent behavior. To ensure he is able to address violence even during the routine procedures of a vehicle stop, Frels now ignores departmental policy requiring the use of his seatbelt while driving.

[…] in that situation, that seatbelt was a huge hindering [sic] for me […] I did not like feeling not being able to fully do anything because I was restrained by my seatbelt. So, I feel that one incident has tailored my thought and the way I do things. […] when I pull traffic, I'm running the license plate, I'm looking at the [computer] return, I'm trying to watch the driver, I'm trying to see how many people are in the car…there's a lot going on. So as soon as I stop that car, I like being able to not worry about the seatbelt or worry about taking the seatbelt off right before I stop the car […] Or I get to a hot scene, you know, I can throw the car in park and hop out right away, not having to worry about taking my seatbelt off beforehand or taking my seatbelt off when I get out.

Frels' behavior is especially telling given that the SPD participates in the “Below 100” program, a nonprofit initiative aimed at reducing line-of-duty deaths, especially deaths that might be prevented with seatbelt use and reduced driving speeds. Despite Below 100 academy training and printed materials posted in SPD facilities that implore officers to “Wear your belt […] Watch your speed.” (see Figure 2), Frels' understanding of his work through the danger imperative attenuates his response to the danger of high-speed crashes to ensure he can effectively respond to a violent threat. In their efforts to ensure officer safety by being ever-ready for a foot chase or firefight, officers in the EPD, WPD, and SPD consciously violate departmental policies designed to prevent their deaths.

Figure 2. Below 100 Posters Displayed in a Sunshine Police Department Facility.

7.3 Unintended Consequences of the Danger Imperative

Though intended to protect officers from the threat of violence, the policy-deviant behaviors described contribute to the grievous injury and death of police officers. Just before 3 am in West River, Officer Jenkins—a rookie officer who I met before he entered the WPD academy—responded at high speed and without lights and sirens to a report of suspicious persons. While racing to the location listed in the call text, he collided with another vehicle, lost control of his patrol car, and careened into a parked semitruck. The impact obliterated the front end of the patrol car and trapped him inside.

After being extricated by firefighters, he was rushed to a local hospital for emergency surgery and placed in a medically induced coma. At the time of this article's writing, Jenkins is able to open his eyes and sporadically track movement. His family has transferred him to a long-term medical facility and is raising money to pay for neurological rehabilitation not covered by his insurance. Despite the severe injuries sustained by Officer Jenkins in this crash, one officer I spoke with confirmed that he and other officers in the department continue to ignore departmental policy requiring seatbelt use, even when driving at high speeds.

In Elmont, Assistant Chief Altidore described a high-speed crash that left one officer comatose and another dead as they responded to a domestic violence call.

The reason why the officer died is because he wasn't wearing a seatbelt. He got ejected and his own car ran over him, rolled over […] It wasn't a pretty sight […] I was there. I arrived there, he was still hanging out the door. And the car was back upright, landed back on its four wheels. […] It was Reggie Tagliano. Reggie, you alright? You don't realize, no, he's not alright because the car just rolled over him … Had he been wearing his seatbelt, he would have survived that crash. Probably not the other officer who's a vegetable now […] but [Reggie] would have survived that. Definitely.

Though it is impossible to know exactly why Tagliano chose to not wear his seatbelt in this particular case, the consistent justifications of officers in Elmont, West River, and Sunshine for not wearing theirs suggest that Tagliano's death, though tragic, is unremarkable in the factors that contributed to it. Injuries like Jenkins's and deaths like Tagliano's are the unfortunate and unintended outcomes of behaviors stemming from officers' perceptions of their work through the danger imperative.

Because of the emphasis placed on protecting themselves from violence, officers make what they conceive as strategic choices to ensure their safety while preserving their ability to chase or fight. In these efforts, officers choose to disregard departmental policy restricting high-speed driving and requiring seatbelts that they believe create unacceptable officer safety concerns. These policy-deviant behaviors, though justified as a way for officers to stay safe from the threat of violence, increase the probability of injury and death during the high-speed driving that they engage in far more often than battles with armed suspects.

8. Discussion and Conclusion

This article joins a growing body of contemporary sociolegal research on the street-level practices of police officers and provides needed insight into the role of danger and threat in police behavior. My findings from the Elmont, West River, and Sunshine Police Departments show that the cultural frame of the danger imperative—the perceptual preoccupation with violence and the provision of officer safety—shapes officer behaviors intended to ensure survival on the street. Unlike policy-compliant behaviors like calling for backup or touching the trunk of a stopped vehicle, policy-deviant behaviors like unauthorized high-speed driving, code 2 ½ driving, and failure to use a seatbelt during high-speed driving, though intended to keep them safe from violence, lead to the injury and death of police officers.

As a complement to research that links officers' preoccupation with danger to deviant behaviors ranging from corruption to brutality and excessive force, this study also allows us to see that some supposedly safety-enhancing police behaviors endanger both the public and the police. Namely, officers' attempts to avoid violent victimization can also directly contribute to the vehicle crashes that represent the leading cause of accidental death in policing and the second most common cause of line-of-duty deaths overall (FBI 2019a). What's more, this high-speed driving also contributes to the injury and death of community members. One analysis finds that police pursuits between 1979 and 2013 claimed the lives of more than 5000 bystanders or passengers and injured tens of thousands more (Frank 2015). Though policy-deviant behaviors like not wearing a seatbelt or unauthorized high-speed driving may appear unimportant alongside high-profile police shootings, this article shows that even seemingly mundane officer behaviors shaped by danger have grave consequences for police and the public they serve.

The danger imperative's emphasis on violence in the line of duty also has implications for the link between police behavior, public well-being, and police legitimacy. For example, the danger imperative is closely related to the “warrior mentality” and its emphasis on aggressive enforcement. “Warrior” officers approach interactions under the assumption that any contact with the public could erupt into violence at any moment. Those looking to “maintain the edge” over suspects will engage in preemptive uses of force seen as necessary to ensure control over a situation and head-off resistance (Stoughton 2014a; Van Maanen 1974). Indeed, the vast majority of force used by police occurs before there is physical resistance from a suspect (Stoughton 2014b: 866–68). Even though such force is very often legal, within departmental policy, and in line with officer training, these dominance-based interactions are precisely those liable to escalate into violence that harms both police and the public (Stoughton 2014a: 229–30; Garrett and Stoughton 2017: 250–51). Even if an interaction does not result in physical violence, aggressive and antagonistic police behavior in the name of officer safety can still create interactions that reinforce distrust of police, further damage police legitimacy, and increase the risk of future encounters escalating (Gau and Brunson 2010).

Officers' preoccupation with violence and officer safety is also implicated in cases of excessive force. In keeping with the state's interest in allowing police to forcefully protect their lives (Harmon 2008), officers regularly cite safety concerns as a reason for their use of force. When deploying their TASERs, for example, officers will justify such force on the grounds that it enhances their safety, even in cases where fellow officers view the use of force as plainly excessive (Sierra-Arévalo 2019a). Similar justifications are used in cases of excessive lethal force, including rare cases in which an officer is convicted of a criminal offense, such as in the police killings of Walter Scott and Laquan McDonald (Crepeau and St. Clair 2018; Schmidt and Apuzzo 2015). More often, this justification is given in cases that do not result in criminal convictions, such as in the police killing of Philando Castille (Ingraham 2017), and in cases where no charges are filed, such as in the police killings of Michael Brown and Tamir Rice (Haag 2018; Sanburn 2014). Whether a use of force is found to be reasonable, criminal, against department policy, or some combination thereof, it is clear that the preoccupation with violence and officer safety is closely tied to officers' use of force, including the high-profile police killings that are at the heart of US policing's current crisis of legitimacy (Cobbina 2019; Weitzer 2015).

Crucially, neither aggressive, dominance-based policing, police violence, nor the consequences of such police action are equally distributed across the population. Instead, longstanding associations between race, place, violence, and criminality perpetuate the funneling of police into minority communities (Anderson 2012; Capers 2009). As a result, Black Americans continue to disproportionately experience police stops, searches, arrests, and uses of force (Epp et al. 2014; Gaston 2019; Kramer and Remster 2018), virtually guaranteeing that those most likely to be perceived as suspicious, violent, and criminal will be those most likely to encounter police primed for threat (Eberhardt et al. 2004; Goff et al. 2014). In turn, the concentration of police in Black communities contributes to decreased mental and physical health (Sewell and Jefferson 2016; Sugie and Turney 2017), lower educational attainment (Legewie and Fagan 2019), and disengagement from key social institutions and the labor market (Brayne 2014).

Of course, the ethnographic and interview data presented in this analysis cannot, in and of themselves, confirm the direct effect of officers' preoccupation with danger on behaviors like stops, searches, arrests, or uses of force, nor on deviance that ranges from low-level policy violations to serious misconduct like police brutality. To better delineate this cognitive-behavioral chain, researchers might built on research that finds culturally shared distrust of the public and support for aggressive patrol tactics predict both officers' use of force and complaints filed against them (Ingram et al. 2018). Such inquiry could employ survey methods to create individual and group-level measures of perceived danger, then combine these constructs with individual-level data on deviant and nondeviant officer behavior. In conjunction, cultural measures and fine-grained data on officer behavior would enable exploration of danger's individual and collective effects on the full range of discretionary police action. Knowing that police behavior is strongly tied to officers' local environment and the features of their interactions with citizens (Gaston 2019; Klinger et al. 2016), future research might also incorporate data on neighborhood demographics, crime, violence, and situational-level measures to better elucidate how the preoccupation with violence interacts with micro-level context to shape police behavior.

Though this article's findings can be generalized across three departments—an improvement over past single-site studies of police—it would be a mistake to extrapolate findings from the Elmont, West River, and Sunshine Police Departments to the approximately 18,000 state and local law enforcement agencies in the US (Reaves 2011). Future inquiry might look to apply the qualitative methods used in this study to smaller rural and suburban departments to assess whether and to what degree the danger imperative shapes officer behavior in these contexts. Variation across police organizations could also be captured with data on departmental training, policies, and state law, and then be combined with data on officer gender, age, experience, race, education, and occupational role (e.g., patrol, investigations, SWAT) to explore how officers' understandings of danger are structured by both individual and organizational characteristics.

Additionally, while past research makes clear that danger shapes police culture and practice outside the US (Fassin 2013; Perkins 2018; Reiner 2010), the contours of the danger imperative described in this article are undoubtedly tied to the particular context in which US police operate. Namely, the supply of civilian-owned firearms in the US—estimated at more 393 million (Karp 2018)—sets the working environment of US officers apart from that of police in other countries. From early ethnographic work that found US officers were more concerned with being shot than their British counterparts (Banton 1964: 111), to more recent analyses showing that US police are victimized by firearms at far higher rates than those in European nations (Sierra-Arévalo and Nix 2020; Zimring 2017: 86–87, 95–96), the danger of US policing and its place in police culture is indelibly shaped by firearms. As argued by Carlson (2020), the threat of firearms is central to how US police understand themselves, their work, and the public, in turn amplifying the particular “intensity and imagery” of US policing (401).

Finally, policing's pervasive preoccupation with violence and officer safety is implicated in the police profession's resistance to reform and its efforts to expand police power. Officer safety is often used by police advocates, especially police unions, as a shield against reform efforts. The New York City Police Benevolent Association, for example, has for years fought against the release of NYPD misconduct data on the grounds that these data will enable retaliatory violence against officers (Rayman 2020; Taggart and Hayes 2018). More broadly, police union contracts across the nation commonly require misconduct and disciplinary records to be destroyed after a period of time, further frustrating efforts to enhance transparency and accountability (Rushin 2017).

The danger of police work is also explicitly leveraged to expand police power and protections. In recent years, some have claimed that the United States is in the grip of a “war on cops” characterized by growing disrespect and violence directed at police (Mac Donald 2016). Though scholars have found no evidence of a significant increase in violence toward police in recent years (Shjarback and Maguire 2019; Sierra-Arévalo and Nix 2020; White et al. 2019), concerns over officer safety undergird new “Blue Lives Matter” laws that seek increased punishment for attacks on police (Craven 2017). This increased punitiveness is supported by officials at the highest levels of the federal government. In a speech to the Fraternal Order of Police, a police advocacy organization with more than 354,000 members, Attorney General William P. Barr claimed increasing violence against police officers and promised to introduce federal legislation to fast track the death penalty for those who kill a police officer (Barr 2019). While the ultimate outcomes of this rhetoric and proposed changes are as of yet unknown, evidence strongly suggests that these developments—rooted in policing's preoccupation with violence and officer safety—will further expand police power, increase legal punitiveness, and frustrate reform efforts.