Introduction

Calls for police reform abound in the United States. Excessive uses of police force—too commonly ending in fatalities—have prompted widespread protests and calls for new approaches. That much of this police violence is visited on people of color only intensifies calls for fundamental retraining of officers and for a new framing of the police role.

Although many well-publicized police uses of force have occurred in comparatively small cities, police-community tensions remain especially heightened in urban areas, where class- and race-based inequalities are endemic. Because the police are regularly called into areas beset by various forms of socioeconomic disadvantage, they often represent the most visible symbol of state authority in such neighborhoods. When the police use their coercive force in a manner residents consider illegitimate, tensions commonly emerge, and sometimes boil over into civil unrest. As a result, a key question for urban police departments is how best to confront the unavoidable issues that stem from various manifestations of social marginality.

Criminological, sociological, and legal research has excelled at demonstrating the direct and indirect harms caused by policing in marginal urban communities (Beckett and Herbert Reference Beckett and Herbert2009; Rios Reference Rios2011; Goffman Reference Goffman2014; Stuart Reference Stuart2016). Because the majority of the literature focuses on what does not work, scholars have been slow to offer alternative policing models that can lessen community tensions, and perhaps even provide some form of relief for those who suffer from poverty and addiction.

Our goal in this article is to facilitate that discussion by assessing three current approaches to the policing of urban social marginality. We are particularly focused on police approaches to insecurely housed individuals who have substance abuse and/or mental health issues. Such individuals have been the focus of much political and police attention; visible homelessness and the sometimes unpredictable behavior associated with it have compelled different efforts to enroll the police to address such so-called disorder. Our comparison of police approaches to these individuals allows us to draw systematic attention to what we believe to be a promising new police reform—the relatively nascent model we refer to as officer-led harm reduction. To understand and assess this emerging model most effectively, we contrast it with two less promising approaches. Each of these models can be distinguished by at least three characteristics: the official purposes of police contact and sanctions, the political and organizational partners involved, and the underlying ideologies of crime and crime control. Each approach also reflects key differences in the way that the police manage the often uneasy balance between their coercive and supportive roles.

The first model we discuss is aggressive patrol. This has been packaged in different variants in US cities—as “broken windows,” as “order maintenance,” as “zero tolerance,” as “stop and frisk,” as what Beckett and Herbert (Reference Beckett and Herbert2009) label “banishment.” Although these approaches vary slightly, they all leverage the police's ability to intrude and arrest, and reflect a faith in deterrence to curb low-level criminality and to assist local business interests in cleaning up areas riven by the disorder ostensibly created by the impoverished. A second approach might be termed coercive benevolence or therapeutic policing (Stuart Reference Stuart2014 Reference Stuart2016). Here, officers use the threat of arrest to try to compel individuals to avail themselves of various social services that might alleviate their poverty and/or reduce their dependence on controlled substances. Most evident in Skid-Row-type areas, coercive benevolence tries to marry the police capacity for violence with social work's capacity to heal. Like aggressive patrol, coercive benevolence rests on a neoliberal ideology that treats marginal individuals as cost-benefit actors who, through improper choices and behaviors, are responsible for their current condition.

Officer-involved harm reduction differs from these approaches in several important ways. Rather than pursuing arrests to enhance deterrence, officers use their discretion to channel troubled people who do not pose a public safety threat out of the criminal justice system and toward harm-reduction-oriented services. Importantly, the programs to which people are referred do not require abstinence, and the threat of criminal sanction is not used to compel participation. Moreover, the provided services are based on the principles of harm reduction and on the idea that poverty and marginality are caused by structural failings, rather than individual ones. The harm-reduction approach holds that many social problems cannot be solved through increased criminalization, and that significant improvements in individual and community well-being may occur in the absence of abstinence.

In our assessments of these contrasting approaches, we stress three overarching points. The first is that police efforts to reduce the ill effects of social marginality, particularly that represented by visible homelessness, are necessarily limited. This is because marginality arises from structural factors the police cannot address. In the first section, we review those factors, and argue for an appropriate level of caution in assessing any police effort to address the ill effects of poverty and addiction. Our second argument focuses on police culture and organization. Powerful currents inside police departments predispose officers to see the downtrodden in particular ways, and to favor particular patrol strategies. Any meaningful police reforms will need to address these longstanding practices and preferences. Much the same could be said about the wider political environment in which the police operate, which is the focus of our third argument. Just like officers, powerful political actors often expect the police to act in particular ways in response to alleged increases in so-called disorder. For this reason, any police reform efforts will likely fail unless key political actors are supportive for long enough to allow change to take hold.

Of course, calls for police reform are hardly new. One can write the history of policing as a succession of one approach following another, with each designed to remedy the defects of its predecessor (Walker Reference Walker1977). The changes that any one reform accomplishes are highly variant. That many reforms only marginally succeed illustrates our arguments that the internal politics of the police matter significantly, as do the wider political dynamics in which departments are enmeshed. Given their discretionary power, patrol officers often succeed admirably in resisting change. Similarly, shifts in the wider politics of policing can push and pull departments in rapidly changing directions, and thereby thwart any given reform before it can become established. For these reasons, a promising option like officer-involved harm reduction will remain vulnerable to usurpation.

We move through five sections. In the first, we review the various structural factors that generate social marginality, and explain their significance to the police role. Precisely because these factors are structural, they limit just how much the police can address social disadvantage and the problems that attend it. We use the subsequent sections to review the three different approaches in turn—aggressive patrol, coercive benevolence, and officer-involved harm reduction. We use the final section to review our arguments, and to assess whether and how officer-initiated harm reduction can gain the widespread embrace that we believe it deserves. Although this approach cannot stand alone as an adequate response to contemporary calls for police reform, we see it as a notable step in the right direction, if it can gain the necessary political traction.

The Police and Social Marginality: Structural Determinants

There is little question that the police must regularly confront the various manifestations of social marginality. Their common tours of duty include a disproportionate number of stops in less advantaged neighborhoods. They are frequently hailed to place pressure on the socially marginal who generate the various forms of disorder that commonly cause consternation, such as trash, graffiti, used needles and condoms, and human waste. The police are also commonly asked to address behaviors like sleeping in public and aggressive panhandling. Certainly, evidence of social marginality is starkly visible and notably consequential in US cities. Its ill effects are felt most strongly by the marginal themselves, who suffer from insecure housing and poor employment prospects, and who commonly confront struggles with mental illness and/or substance abuse. It is also consequential to business owners, who fear the loss of commerce stemming from seeming disorder on or near their doorsteps, as well as service organizations that house or otherwise assist the impoverished. But it is also highly significant to the police. Indeed, inequality very strongly shapes the workaday lives of police officers. Any full understanding of urban policing thus requires at least a basic understanding of the dynamics of urban inequality.

In the first instance, such inequality can be understood as an inevitable distillation of the capitalist space economy (Harvey Reference Harvey2007 Reference Harvey2009; Smith Reference Smith2008). A constituent component of capitalism is its differentiation between those with some control over the production process from those with none. This differentiation, of course, generates wage gaps of often gargantuan scope. Such gaps have increased in recent years, given the decline in middle-class industrial jobs, and the rise of a service sector that generates wide distinctions between workers with more versus less highly valued skills (Harrison and Bluestone Reference Harrison and Bluestone1990; Stiglitz Reference Stiglitz2013).

The resultant income inequality is instantiated through a range of sometimes connected processes. The capitalist housing market is most critical. It creates distinct living environments for different class fractions. The cultural celebration of the suburban ideal helped to generate demand for bucolic neighborhoods that attracted the well-heeled (Jackson Reference Jackson1987), and bequeathed a poorer quality of life for the unfortunates who could not escape central cities (Dreier, Mollenkopf, and Swanstrom Reference Dreier, Mollenkopf and Swanstrom2014). Suburbanization thus compelled a sharp reduction in the tax bases of cities that also bore a growing poor population. More recently, however, rapid gentrification is transforming many urban areas once abandoned by the suburbia-bound. Yet the poor again suffer, displaced by rising rents and notable shifts in the residential and commercial landscape (Smith Reference Smith1996; Lees, Slater, and Wyly Reference Lees, Slater and Wyly2008). Importantly for the police, the poverty-stricken individuals who do remain in such neighborhoods are regularly decried as a form of disorder that officers are expected to address.

Racial dynamics magnify the class issues underlying segregation. Although it is no longer legal for government and real estate agents to engage in the explicit process of “redlining” urban neighborhoods—drawing boundaries around areas where undesired minorities were meant to remain—various informal processes reinforce segregation. For example, realtors still steer individuals toward certain neighborhoods, just as banks use lending practices to reinforce longstanding patterns of segregation (Orfield and Ashkinaze Reference Orfield and Ashkinaze1991; Massey and Denton Reference Massey and Denton1998; Wyly and Holloway Reference Wyly and Holloway1999). Communities with large numbers of people of color are often politically marginalized, and are less equipped to resist the siting of such hazardous projects as landfills, incinerators, and toxic-waste sites (Bryant and Mohai Reference Bryant and Mohai1992; Pulido Reference Pulido2000).

The lack of low-income housing also critically shapes the realities that urban police officers must face. In the United States, visible homelessness was an unsurprising consequence of a sharp reduction in federally funded public housing (Wolch and Dear Reference Wolch and Dear1987). As this was occurring, many institutions for the mentally ill were shuttered, principally because they witnessed widespread malpractice (Scull Reference Scull1977; Lamb Reference Lamb2001). These two developments created a large population of disadvantaged and insecurely housed individuals. When those individuals frequented urban public space, they generated concerns about how they might influence commerce and tourism (Mitchell Reference Mitchell1997; Gibson Reference Gibson2004).

The withdrawal of public support for housing and mental illness is consistent with the rise of neoliberalism, commonly understood as an ideology that trumpets the role of market forces over government interventions in the economy (Larner Reference Larner2000; Brenner and Theodore Reference Brenner and Theodore2002; Harvey Reference Harvey2007). Neoliberal logic has legitimated the loss of public housing, because the private housing market is presumed to provide options for everyone, including the poor. It is hardly an accident that the increased popularity of neoliberal thinking has accompanied increased urban inequality.

Yet the state has hardly disappeared in the United States in the neoliberal era (Herbert Reference Herbert1999; Western and Beckett Reference Western and Beckett1999). This is most evident with mass incarceration, itself a strong influence on many urban neighborhoods. Since the late 1970s, the United States has more than quadrupled its rate of incarceration, and now leads the world. People of color are notably overrepresented in the incarcerated population, and this profoundly influences the neighborhoods from which they come (Clear Reference Clear2009). These neighborhoods lose residents who possess some measure of earning power, and their spouses and children lose potential familial support (Western Reference Western2006). Further, those individuals return to their neighborhoods postincarceration with reduced levels of employability (Pager Reference Pager2007), political incorporation (Uggen and Manza Reference Uggen and Manza2002), and health (Massoglia Reference Massoglia2008). In this fashion, policing and incarceration work to exact continuing punishment and inequality, for both individuals and their neighborhoods.

Our goal here is not to provide a thorough explanation of urban socioeconomic differentiation, but to situate the police within the wider contexts in which they operate. Even a cursory look at the history of policing demonstrates that the police response to social marginality is highly contradictory, reflecting both coercive and supportive goals. The fact that the police simultaneously perform such seemingly contradictory functions can be traced to their role as political actors who must negotiate the demands of a variety of local and extra-local interests (Huey Reference Huey2007; Stuart Reference Stuart2016). Some of the loudest of such demands come from dominant economic, racial, and other social groups who call on the police to uphold existing inequalities by supervising and regulating those designated as problematic, dangerous, or simply a nuisance. For example, many US police departments grew directly out of the “slave patrols” instituted in the antebellum period. As Withersmith (Reference Withersmith1974, 20) describes: “Throughout all of the [southern] states during the antebellum period, roving armed police patrols scoured the countryside day and night, intimidating, terrorizing, and brutalizing slaves into submission and meekness.” After emancipation, as modern capitalism took shape, the police turned their coercive attention to the new industrial working class that presented new challenges to the economic order. By the late nineteenth century, the police regularly mobilized to defeat workers' strike efforts and unionization attempts (Harring Reference Harring1983).

At the same time, however, the police have played a more accommodative, supportive, and perhaps even compassionate role. As the police history by Eric Monkkenon (Reference Monkkenon2004) shows, early police forces were also heavily tilted toward the provision of various social services. The police helped run soup kitchens, inspect boilers, standardize weights and measures, find lost children, take censuses, regulate health standards, provide ambulances, and supply overnight lodgings in police stations. In some cities, local police stations housed more migrant workers, mendicants, and “hobos” per night than local shelters and charity organizations.

The history of policing can thus be read as a constant see-sawing between coercive and supportive roles in response to inequality's various effects. As a result, we focus squarely on this tension in our analysis of the three contrasting approaches in the pages to come. We also acknowledge that the police simply do not possess the capacity to alter the underlying determinants of social marginality. This reality warrants limited expectations of the role of the police in responding to social marginality. Indeed, the police are poorly served by political forces that expect officers to solve social problems they cannot meaningfully address. That said, police responses to poverty that are attentive to wider structural dynamics can generate some positive, if limited, improvements in the quality of life of the impoverished. This explains our support of officer-involved diversion and harm reduction.

The value of officer-involved harm reduction is best understood by first considering two other approaches to the policing of social marginality: aggressive patrol and coercive benevolence. Aggressive patrol is the most longstanding approach, and the most hegemonic. It also comports most closely with established normative scripts and accepted practice within police departments. For this reason, it will prove a challenge for departments to shift toward a more progressive vision of policing like officer-involved harm reduction.

Approach I: Aggressive Patrol

Toward the beginning of the twentieth century the police role shifted toward crime prevention, a trend noticeably accelerated by the rise of the professional movement. This movement sought to craft a more particularized role for the police to help enhance their status, namely, as that social agency most uniquely poised to respond to and help reduce crime. The police worked to create a more effective bureaucratic structure and to develop sophisticated technologies (Fogelson Reference Fogelson1977; Walker Reference Walker1977). These would together help achieve a key goal: to apprehend criminal offenders so regularly that widespread deterrence would take hold. All potential criminals would see arrest and punishment as so certain that they would desist.

Faith in the police as crime-fighting agents of deterrence is central to the legitimation of aggressive patrol. Because of their capacity to exercise legitimate coercive force (Bittner Reference Bittner1970), the police are called to instances of alleged criminal wrongdoing. Their coercive force enables them to effectuate arrests when appropriate, and to set deterrence in motion: the threat of arrest and subsequent jailing will presumably dissuade individuals from criminal misbehavior. Here, crime is understood as arising principally from certain individuals whose behavior can be shaped by the criminal justice system. Aggressive patrol, and the arrests it generates, can thereby be legitimated as a key means not only to reduce crime, but also to solve other problems related to poverty and inequality.

Aggressive patrol comes in different guises, few of which lack some measure of controversy. Some more benign forms are ones sensibly connected to repeated patterns of criminal behavior in particular locales. In such alleged problem-solving exercises (Cordner and Biebel Reference Cordner and Biebel2003), the police analyze crime patterns in a focused area, target a well-designed intervention, and withdraw once the dynamic improves. Importantly, such interventions might include more than just arrests; they could also involve such changes to the physical environment as improved lighting and more open architecture, and the development of various forms of surveillance (Welsh and Farrington Reference Welsh and Farrington2009). Yet such targeted and hence limited efforts are less common than other forms of aggressive patrol.

In the contemporary moment, many such efforts take some ideological refuge in the theory of broken windows. This well-known theory suggests that police efforts to reduce disorder will also reduce more serious crimes (Wilson and Kelling Reference Wilson and Kelling1982). Such disorder comes in the forms of unwanted street-level behaviors commonly engaged in by homeless individuals and others on the social margins. Also known as “order maintenance” or “quality of life” policing, this approach encourages cities to criminalize behaviors considered disorderly, such as sitting or lying in public places or engaging in aggressive panhandling, among others. Facing the suburban flight of high-income earners and the depreciation of the urban core, municipal business interests increasingly call on police departments to deploy such tactics to assist them in revitalizing downtowns, maximizing the profitability of retail spaces, and providing boosts to the tourism and convention industry (MacLeod Reference MacLeod2002; Vitale Reference Vitale2008). In Seattle and elsewhere, mere presence in particular places can constitute a crime (Beckett and Herbert Reference Beckett and Herbert2009). Through a variety of maneuvers, individuals can be banned from particular areas, and can be arrested if they appear in those places. Again, it is the police's capacity to effectuate an arrest that presumably will deter individuals from engaging in unwanted behaviors.

Aggressive patrol can also be used to uncover evidence of criminal wrongdoing. Stop-and-frisk exercises illustrate this philosophy. Here, the police use their authority to search individuals. This can happen when the police discern probable cause of a crime, when they can claim a threat to their physical safety, or when they ask an individual to consent to a search. In stop-and-frisk exercises, the police are typically looking for illegally acquired guns or drugs. Hence, this strategy can be legitimated as a deterrent to serious crime (Gelman, Fagan, and Kiss Reference Gelman, Fagan and Kiss2007). Similarly, the police can engage in so-called investigatory traffic stops (Epp, Maynard-Moody, and Haider-Markel Reference Epp, Maynard-Moody and Haider-Markel2014). In such instances, officers use any means they can to pull over motorists they consider suspicious, commonly after a minor traffic violation. After initiating the stop, the police then work to convince the motorist to allow a search of the vehicle in hopes of discovering contraband.

Aggressive patrol can thus be legitimated as a deterrent to criminal offending. From this perspective, even if it inconveniences some individuals, it serves public safety. It also commonly serves as a welcome exercise for officers. For many, aggressive patrol provides the satisfaction of demonstrating competence by securing an arrest. It also requires officers to be very proactive, which many enjoy (Herbert Reference Herbert2001). Further, it comports with the typical police understanding of criminal behavior: that it is the result of a few “bad apples” who pollute an otherwise law-abiding community (Herbert Reference Herbert2006). The police thereby commonly see themselves as using their arrest power to deter those bad apples from misbehavior. In these and other ways, aggressive patrol aligns well with common subcultural currents within the police world.

It also aligns well with cultural currents external to the police. It is now commonplace to turn first to the police whenever concerns emerge about crime or disorder, due to a hegemonic belief that officers can respond quickly and effectively. It is thus common, for instance, for business groups to develop strong and ongoing relations with police departments to give them better influence over police behavior and policy. In Seattle, for example, an official in one such group expressed satisfaction with spatial restrictions on individuals. As she acknowledged, such restrictions made police arrests welcomingly easy, and hence accentuated officers' capacity to place pressure on unwanted individuals in areas heavily trafficked by consumers and tourists (Beckett and Herbert Reference Beckett and Herbert2009).

In its various manifestations, aggressive patrol thus endorses a strong and presumably efficacious role for the police in addressing criminal or criminogenic behavior by the socially marginal. The police welcome opportunities to use their coercive power to make arrests or otherwise deter bad apples from troubling the law-abiding. Citizens concerned about the ill effects of alleged disorder look to the police to use their power effectively. However, there are considerable critiques one can level at aggressive patrol. Two such critiques deserve particular scrutiny: that it results in illegitimate disparities in arrests and law enforcement attention, and that it ignores, and even exacerbates, the underlying dynamics that generate much crime and disorder in the first place.

Racial and socioeconomic disparities are a consequence, in part, of officers' decisions about where to concentrate their attention. Particularly if they are responding to particular community concerns, officers commonly focus on specific areas where those concerns are presumably most present. For their part, officers are often keen to find areas that they consider target rich (Herbert Reference Herbert2006), places with higher odds of effectuating an arrest. Such areas commonly house poorer individuals, who spend a disproportionate amount of their time in public space, and thus are more visible to officers. The association between poverty and race means that racial disparities in arrest patterns are widespread. Further, officers often focus on particular crimes, such as selling crack cocaine, which can increase the odds of racial disparities in arrests even in the absence of intentional racial profiling (Beckett et al. Reference Beckett, Nyrop, Pfingst and Bowen2005; Beckett, Nyrop, and Pfingst Reference Beckett, Nyrop and Pfingst2006). In short, aggressive patrol focused on the socially marginal will often increase the likelihood of notable class- and race-based disparities in both arrests and subsequent jail and prison terms. Among the consequences of such unwelcome disparities is a decline in the perceived legitimacy of the police and of the criminal justice system more broadly. When citizens experience themselves as singled out for police attention or otherwise treated unjustly by officers, they lose faith in the police (Tyler and Huo Reference Tyler and Huo2002; Epp, Maynard-Moody, and Haider-Markel Reference Epp, Maynard-Moody and Haider-Markel2014). This can lead affected citizens to be less willing to comply with officer requests in the future, and even to withdraw from public life more broadly (Lerman and Weaver Reference Lerman and Weaver2014).

Thus, by reducing citizen compliance with police requests, aggressive patrol can be counterproductive. Such requests might include information about the commission of crimes. If information flow is reduced, then so is police efficacy. Aggressive patrol can be counterproductive in another sense, as well. If disorder results ultimately from one or more of such underlying conditions as chronic unemployment, insecure housing, mental illness, and substance abuse, then aggressive patrol can work to exacerbate such problems. If individuals are fearful of unwanted police attention, they might make themselves less visible in particular places, including places that host much-needed social services. If access to such services is reduced, then the ill effects of marginalization can fester and even worsen. In addition, regular trips to jail only add to the stresses of everyday life, and make the route back to the social mainstream more arduous. This is especially true if individuals accumulate fines and warrants, and thereby become perpetually vulnerable to police intrusion.

Despite these problems, the adoption and apparent success of broken windows policing in the United States has had transnational consequences. In the wake of the implementation of a particularly aggressive form of broken windows policing in New York City, firms such as Giuliani Partners and the Bratton Group began to influence regional and municipal policing practices and crime policies around the globe. For example, after being declared America's “top cop” by Time Magazine in 1996, New York Police Chief William J. Bratton stepped down from public office and formed his own private consulting company. This consulting group, the Bratton Group, has worked most extensively in South America, where he has advised public officials on both crime policy and campaigning on an anticrime platform (Lifsher Reference Lifsher2001). Similarly, upon retiring from the mayor's office, Rudy Giuliani formed Giuliani Partners, a consulting organization staffed by many of his top aides from city hall and former police commissioner Bernard Kerik (Mitchell and Beckett Reference Mitchell and Beckett2008).

These prominent firms promoted an aggressive version of the so-called New York model, often with unfortunate consequences that mainly included advancing neoliberalism and the private interests that supported it (Mitchell and Beckett Reference Mitchell and Beckett2008; Campesi Reference Campesi2010; Muller Reference Muller2013). The work of these and other transnational security consulting firms coincided with a push by the conservative think tank, the Manhattan Institute, to share ideas for urban renewal with leaders in Latin America. For example, after opening a new office in Santiago, Chile, the Manhattan Institute began the Inter-American Policy Exchange, a new initiative to promote zero-tolerance policing, educational reform (i.e., vouchers), and probusiness tax reform in Latin American cities (DePalma Reference DePalma2002). Critics have argued that the main legacy of the involvement of these transnational firms has been the clearing of desirable, high-rent neighborhoods for gentrification and urban renewal (Campesi Reference Campesi2010; Muller Reference Muller2013).

It is thus clear that aggressive patrol is not an effective or just strategy for addressing the manifestations of social marginality in the United States or elsewhere. Unsurprisingly, widespread critiques of aggressive patrol have helped birth alternative approaches. In the next two sections, we consider two such approaches: coercive benevolence and officer-initiated harm reduction.

Approach II: Coercive Benevolence

At first glance, coercive benevolence and aggressive patrol appear quite indistinguishable. Indeed, both modes employ stop and frisks, intensive enforcement of quality-of-life laws, and a neoliberal logic whereby the police address poverty by coercing the socially marginal to make what police consider better choices. Closer examination, however, reveals that coercive benevolence is characterized by a divergent set of community partnerships and seeks an ultimate end that departs significantly from aggressive patrol.

First, the community-police alliances found in this mode involve a far greater degree of input and influence by social service organizations and homeless shelters that stress intensive rehabilitation. Such organizations have grown exponentially since the drafting of the Clinton administration's 1993 Continuum of Care plan. This plan redirected funding away from basic emergency shelters toward nonprofits addressing the individual pathologies deemed responsible for homelessness, including substance abuse, mental illness, inadequate life skills, and deficient workforce training (Sparks Reference Sparks2012). To remain competitive for both public and private funding, these service organizations have replaced the traditional model of “three hots and a cot” with a more interventionist, rehabilitative, and “postsecular” (Cloke, May, and Johnsen Reference Cloke, May and Johnsen2010, 23) model centered on intensive residential recovery programs. As a result, rather than primarily banish, incapacitate, or eliminate problematic populations, coercive benevolence uses police interventions and legal punishments to help “improve poor people” (Katz Reference Katz1997, 13) by trying to compel them to engage in self-help and reform. This mode of control is typically found in Skid Row districts and other urban areas that are characterized by a large number of precariously housed individuals, homeless shelters, and recovery organizations.

At its core, coercive benevolence seeks to alter the behavior of marginal populations by presenting them with powerful ultimatums: enter a rehabilitative program or face arrest, enter secure housing or go to jail, look for employment or suffer monetary fines. The police enforce these ultimatums through two primary mechanisms. First, through de jure policy, the police create diversion programs in partnership with local shelters and recovery organizations. These programs combine the stick of legal punishments with the carrot of rehabilitative services. Second, through de facto patrol interventions, officers attempt to compel wayward citizens toward self-reflection and self-reform. With diversion programs in place, officers utilize moments of discretionary enforcement to shepherd individuals toward rehabilitative spaces and, ultimately, toward approved lifestyle choices.

This model is most evident in Los Angeles's Skid Row district. In 2006, the Los Angeles Police Department launched the Safer Cities Initiative (SCI)—one of the most aggressive zero-tolerance policing campaigns to date—in Skid Row. Alongside intensive enforcement of civility laws and stop-and-frisk tactics, the department simultaneously created three programs designed to address the poverty, addiction, and mental health of Skid Row denizens. The first program, Streets or Services (SOS), provides those arrested on misdemeanor charges—such as sitting on the sidewalk or loitering—with the option to have their charges dropped by enrolling in a twenty-one-day residential recovery program at a participating shelter. Arrestees who choose this option proceed from booking to a caseworker consultation, and then into the program, all within a matter of hours. The second program, called Homeless Alternative to Living on the Street, allows thousands of individuals receiving citations to work off their fines by performing community service—typically basic janitorial or manual labor—in a partner organization. According to an assistant city attorney overseeing the program: “The work isn't the primary issue. … The overarching mission is to get people into services while they're there.” Adhering to this logic, the third program, the Homeless Court Program (HCP), provides an alternative, more lenient adjudication system for homeless arrestees. Held monthly at a partner facility, HCP may only be accessed by individuals who have participated in a rehabilitation program for at least ninety consecutive days and have not been rearrested or ticketed in the previous six months. Individuals can access HCP only after obtaining a letter of support from a case manager at an approved facility detailing the applicant's positive rehabilitation progress.

With such programs in place, coercive benevolence unfolds through officers' routine patrol behaviors. In interviews, Skid Row officers indicated that they envision their daily tasks as undergirded by the presumptive need to alter the incentive structure that propels deviance and allows individuals to become comfortable living in poverty (Stuart Reference Stuart2014). Indeed, envisioning diversion programs as effective, officers overwhelmingly cite destitute residents as unwilling to “take advantage” of the neighborhood's rehabilitative programs. In the words of a veteran Skid Row officer: “With the concentration of services … there is really no excuse for most people. … This means that about 99 percent of the people that remain down here [on the streets of Skid Row] are here because they choose to be here.” This view animates officers' understanding of their own role, articulated succinctly by a senior officer: “Most people have pretty much communicated to us that they aren't willing to make good choices in their lives. So, our job is to help them make the right choice.”

Officers attempt to coerce marginal citizens toward the right choice through various strategies. The most obvious of these is to activate the legal sanctions that will channel people toward diversion programs. More regularly, however, officers deploy the logic of diversion programs to bring about self-reform and rehabilitation through more indirect mechanisms. For example, officers strategically enforce quality-of-life laws to create graduated levels of punishments that increase over time. For officers, the “Skid Row lifestyle” is learned and cumulative. Officers anticipate that if they can make prolonged residence more costly, individuals will reconsider their unwillingness to live productive, sober lives. One means to do this is strategically to enforce quality-of-life laws—particularly laws prohibiting sitting, lying, or sleeping on the sidewalk—to force people to move constantly. By doing so, officers intend to harden neighborhood spaces against opportunities to consume controlled substances. In the words of one Skid Row officer: “It's pretty hard to spark your crack pipe while you're busy walking.” Officers also selectively enforce quality-of-life ordinances to constrict the availability of vital resources. Interestingly, this tactic is primarily targeted at those they consider do-gooders who travel to the neighborhood to pass out charitable items such as food, clothing, or blankets. By prohibiting such philanthropy, officers attempt to limit Skid Row inhabitants' ability to sustain themselves physically without self-reform or program participation. Taken together, each of these strategies is designed with a common goal: reward the right choice, make the bad choice costly.

It is easy to imagine why city leaders celebrate coercive benevolence despite its hefty price tags—SCI, for example, costs the city roughly $6 million annually for an impact area of only fifty blocks. In a national climate in which police aggression and punitiveness have become major points of contention, coercive benevolence seems palatable: it pairs rehabilitation with punishment to help reduce crime and to address the ill effects of poverty with ostensibly compassionate interventions.

Yet when we cut through the rhetoric and subject coercive benevolence to systematic analysis, several concerns emerge. First, the model appears to have little to no effect on crime. Despite the implementation of the SCI, for example, crime rates in Skid Row continue to mirror those found in areas where the model is not practiced (Blasi and Stuart Reference Blasi and Stuart2008). Second, the fact that punitive measures have been transformed into methods of intake into recovery programs appears to have amplified officers' willingness to conduct pedestrian stops, issue expensive citations, and make arrests for relatively mundane behaviors. In the first year of SCI alone, officers made roughly 9,000 arrests and issued 12,000 citations—all in a neighborhood with a population hovering somewhere around 13,000 (Blasi Reference Blasi2007). Finally, diversion programs animated by coercive benevolence are largely ineffective in improving individuals' material, social, and psychological conditions. According to the LAPD, the SOS program by itself referred an average of thirty individuals per month into rehabilitation programs. But fewer than 10 percent of individuals enrolled in SOS twenty-one-day programs ever graduate. The median length of stay is only three days, and nearly a third of participants abscond within the first twenty-four hours (Blasi Reference Blasi2007). Much attrition is caused by the severe and coercive regulations within the organizations. Participants are subject to expulsion for a number of infractions, including verbal abuse, consumption of alcohol or drugs, possession of drug paraphernalia, failing to check in throughout the day, or leaving a dormitory without permission.

Perhaps more critical than its failure to achieve its goals, coercive benevolence has broader epistemological effects that inhibit our ability to target the underlying and durable causes of social marginality. Embracing a neoliberal logic, coercive benevolence reframes marginality in individualized, cost-benefit terms: disadvantaged individuals have weighed the potential risks, rewards, and costs of becoming productive citizens, and willfully embraced marginality because the benefits outweigh the losses (Stuart Reference Stuart2016). By manipulating behavioral incentives, coercive benevolence aims to reprogram marginal populations to make the right choice and, through collaboration with intensive recovery programs, transform them into sober, self-governing, and responsible citizens. In the process, the net of state surveillance is widened, as the police target an expanding range of behaviors believed to impede correct lifestyle decisions.

Approach III: Officer-Led Harm Reduction

Neither aggressive patrol nor coercive benevolence, then, meaningfully addresses the drivers of urban inequality. In addition, both models reinforce primary cultural circuits within the police world, most notably the idea that officers can use their powers to deter the unwanted behavior of misdirected individuals who make poor choices. That coercive benevolence does so within the structure of cooperative partnerships with social service organizations has minimal impact on police practice or efficacy. We therefore see greater merit in a newer reform movement, officer-led diversion guided by harm-reduction principles. A new program in Seattle that seeks to actuate this idea has shown promising results (Collins, Lonczak, and Clifasefi Reference Collins, Lonczak and Clifasefi2015), and the program is being adopted in cities across the United States (Pendley Reference Pendley2015).

Seattle's efforts grew out of concerns about the role of race in policing. Across the United States, racial disparities in drug law enforcement have been the source of significant controversy, and this problem was especially pronounced in Seattle (Mitchell and Beckett Reference Mitchell and Beckett2008, 56, Table 10). The racially disparate impact of the Seattle Police Department's drug enforcement practices thus became the subject of lengthy, complex, and controversial litigation. At the same time, Seattle officials found that aggressive enforcement tactics did not eradicate open-air drug markets, particularly downtown, and the persistence of visible drug activity triggered significant community pressure to do something about drug market activity. By the late 2000s, no one was satisfied with the status quo, including the SPD itself. As Sergeant Sean Whitcomb, then-spokesman for the SPD, put it: “Officers are frustrated arresting the same people over and over again. We know it's not working” (quoted in Green Reference Green2011, B1). Others agreed.

For years, adversarial relations between the SPD and its critics stymied dialogue about how to improve drug law enforcement. In 2005, however, this impasse was broken. The resulting collaboration led to the creation and implementation of Seattle's Law Enforcement Assisted Diversion (LEAD) program. LEAD is the first known prebooking diversion program for people arrested on narcotics and prostitution charges in the United States. Launched in October 2011, LEAD is the product of a multiyear collaboration involving a wide range of organizations, including the Defender Association's Racial Disparity Project, the Seattle Police Department, the ACLU of Washington, the King County Prosecuting Attorney's Office, the Seattle City Attorney's Office, the King County Sheriff's Office, Evergreen Treatment Services, the King County Executive, the Washington State Department of Corrections, and others.

In 2013, Beckett conducted a process evaluation of LEAD to determine whether the program was operating as intended, and to identify any implementation challenges. Toward this end, various data were collected and analyzed. These included observations of LEAD-affiliated SPD and Department of Corrections officers and sergeants, as well as case managers, as they conducted LEAD-related work, review of foundational documents, observations of the LEAD operations work group and policy group meetings, and interviews with a wide range of LEAD stakeholders and participants. In what follows, we draw on these data to describe LEAD's evolution and operating principles, and to identify some of its strengths and ongoing challenges.

At the outset, LEAD partners spent significant time developing a protocol to outline the procedures by which police officers provide referrals, and by which LEAD clients are engaged by social service providers. Under the protocol, when an eligible individual is arrested for a low-level drug offense or prostitution, a trained police officer may elect to refer that individual to a LEAD case manager rather than pursuing a booking into jail. Not every low-level drug offender is presumptively eligible for LEAD because several criminal background exclusions may apply. These exclusions notwithstanding, officers retain a high degree of discretion over the referral process. For example, even individuals with more serious criminal backgrounds can be referred to LEAD at the recommendation of the arresting officer. In addition, SPD officers may elect not to refer those who are eligible for LEAD to the program. The rationale for this degree of discretion is that officers possess uniquely deep knowledge about those they regularly encounter, and are therefore well situated to determine if someone can safely and beneficially work with case managers.

In a referral, a police officer makes an arrest, then contacts a LEAD case manager, who then goes to the police precinct to conduct an initial screening. In most cases, the officer quickly relinquishes custody of the referred person to the caseworker. The arrest record is still forwarded to the city or county prosecutor, who may file charges. However, the presumption is that no charges will be filed as long as the individual completes both an initial screening and a full intake assessment with LEAD case managers within thirty days. Although LEAD was originally conceived as a prearrest diversion program for low-level drug offenders, it expanded to include sex workers, to help ensure significant participation by women who suffer from addiction and/or extreme poverty. This appears to have been successful, as approximately half of all LEAD clients are female.

LEAD stakeholders sought to hire case managers who are accustomed to working in an intensive and hands-on manner with their clients. LEAD stakeholders refer to this orientation as the “guerilla approach” to social work, highlighting case managers' willingness to do everything from tracking down recalcitrant clients in dark alleys to accompanying them as they complete paperwork, keep appointments, and apply for services and housing. LEAD stakeholders also hired case managers who are comfortable with a harm-reduction philosophy. That is, LEAD case managers are trained to meet clients where they are, to help those clients to identify and achieve their personal goals. Notably, abstinence may or may not be among their clients' objectives, especially in the short term.Footnote ¹

Upon referral, individuals meet with a LEAD case manager who conducts an initial intake assessment and endeavors to connect the client with his or her most needed services. After this assessment, the individual can leave, but is asked to return to the LEAD office for the intake interview. After that interview, the individual is a LEAD client. If the intake assessment is not completed within thirty days, the individual can be prosecuted for the originating offense. In this sense, the possibility that charges will be filed does serve to induce potential clients to complete the intake interview. However, as long as the intake interview is completed, prosecutors do not use the threat of filing charges to sanction subsequent nonparticipation or noncompliance.

LEAD clients then work with case managers to create an individually tailored intervention plan. Such plans may include assistance with housing, treatment, education, job training, job placement, licensing assistance, transportation, small business counseling, child care, or other services. Dedicated LEAD funds pay for these services, although public resources are sometimes accessed, wherever possible without displacing others in need.

Several core operating principles guide LEAD's provision of services. First, LEAD adheres to a nondisplacement principle: an individual's status as a LEAD client cannot allow him or her to move to the top of a waiting list for social services. For example, if a client seeks methadone treatment, LEAD monies will pay for that treatment until that client moves through the waiting list for publicly funded treatment. This nondisplacement principle maximizes the likelihood that LEAD will benefit the community as a whole, not just program participants.

Second, LEAD follows a harm-reduction approach, which seeks to increase individual and community wellness without focusing exclusively on sobriety. Harm reduction assumes that overcoming drug addiction is a long and arduous process in which setbacks are common, and meaningful improvements may occur without abstinence. Moreover, case managers emphasize assisting clients to identify their own goals and supporting them to meet those objectives.

Participation in LEAD thus does not require abstinence. Instead, caseworkers work to ensure that clients will use various forms of support to reduce harm to themselves and others. To help ensure this outcome, LEAD participants' eligibility for services and benefits are not time-delimited. Moreover, the LEAD protocol does not authorize any formal or punitive sanctions for any form of noncompliance. At regularly held work group meetings, police officers, case managers, and prosecutors share information to effectuate informed decisions about LEAD clients. In particular, these meetings help prosecutors weighing whether to charge LEAD clients for alleged offenses subsequent to their enrollment.

In short, LEAD seeks to improve public safety and public order by reducing drug use and delivery, and by stanching the quality-of-life problems associated with open-air drug and sex markets. By enabling police officers to divert potential arrestees, and by providing intensive case management services and resources guided by harm-reduction principles, LEAD seeks to prevent arrests from leading to additional criminal justice intervention.

A recent outcome evaluation suggests that LEAD has been remarkably effective in reducing recidivism. Specifically, the findings reveal “statistically significant reductions for the LEAD group compared to the control group on average yearly criminal justice and legal system utilization and associated costs” (Collins, Lonczak, and Clifasefi Reference Collins, Lonczak and Clifasefi2015, 5). For example, LEAD clients spent thirty-nine fewer days in jail than similarly situated arrestees who did not enter LEAD. Similarly, the odds that a LEAD client was sentenced to prison in the first year after enrollment in LEAD were 87 percent lower than for non-LEAD clients. Unsurprisingly, these reductions meant significant cost savings relative to non-LEAD clients (Collins, Lonczak, and Clifasefi Reference Collins, Lonczak and Clifasefi2015). An evaluation of potential benefits to clients' social, health, and psychological well-being is under way.

In addition to these notable recidivism reductions, LEAD has generated other positive effects. One of these derived from stakeholder interactions. Many stakeholders were surprised that, despite diverse motivations and lingering resentments, they could find common ground. Many noted that their assumptions about other organizations were upended due to their participation in LEAD. As one supervisor at Evergreen Treatment Services put it:

In interviews, many stakeholders made it clear that the transformation of perspectives and institutional relationships was illuminating and rewarding. But it has also created a space through which to pursue other meaningful criminal justice reforms. Here, a defense attorney describes one example:

Many stakeholders thus reported that their perspectives and relationships had been transformed, and that interest in criminal justice reform had increased as a result of their participation in LEAD. The fact that city and county authorities are actively considering the adoption of safe consumption facilities is one recent expression of this increased openness to reform ideas (Beekman Reference Beekman2016; Martin Reference Martin2016).

Finding such common ground has been critical to LEAD's success, in large part because of the challenge of securing buy-in from police officers. LEAD's harm-reduction approach asks officers not to arrest a drug law violator, but to refer him or her instead to a program that does not require abstinence. Moreover, officers who do make referrals often see LEAD clients still on the streets. That people can remain in LEAD even if they are publicly visible and actively using drugs is anathema to many officers. As a SPD Lieutenant explained:

In short, LEAD asks officers to understand addiction through a structural and trauma-informed lens, and to de-emphasize accountability and disincentives as a means of addressing risky behaviors. In other words, LEAD attempts to transform police culture in a fairly fundamental way. The Seattle experience suggests that effecting such change by securing officer support for a harm-reduction program is an ongoing and challenging process, one with potential to pay significant dividends over time, but also one that could be derailed without significant political support.

Although uniformly positive about LEAD as an alternative to the drug war, stakeholders readily identified a number of structural challenges that threaten to undermine LEAD's potential impact. For example, even with access to resources, caseworkers struggled to find client housing. Case managers agreed that their clients' criminal records restricted housing access, but also noted a lack of affordable housing for clients who still use drugs. More generally, LEAD stakeholders said that existing capacity constraints can impede the success of any social service intervention. One put it this way:

In sum, LEAD operations in Seattle show that a broad range of stakeholders can collaborate to institutionalize meaningful drug policy reforms that do not rely on criminal justice or therapeutic coercion. Moreover, this collaboration is itself productive; it appears to yield a variety of dividends, including the potential for transformation of police culture and the creation of a new openness to reform ideas. Still, expanding LEAD in Seattle and institutionalizing it elsewhere requires careful attention to several challenging tasks: continued support from police who make referrals, a sustainable revenue stream, and access to affordable housing and suitable services. These substantial long-term challenges notwithstanding, the first years of LEAD's operations provide compelling evidence that officer-involved harm reduction can be developed, and can generate a variety of notable benefits. Compelled by this evidence, cities from Albany, New York to Santa Fe, New Mexico have launched LEAD, and many others—including Baltimore and Atlanta—are planning to do so soon (Associated Press 2016).

Conclusion

We are not Pollyannaish enough to believe that the widespread adoption of officer-involved harm reduction will solve the legitimacy crises of US urban police departments. Nor do we believe that such efforts would eliminate the manifold and punishing consequences of widespread inequality. Such inequality will persist absent meaningful structural reforms that increase the availability of affordable housing, well-paying low-skilled jobs, health care, and services for mental illness and substance abuse. Further, we recognize that reforming police approaches toward individuals who are insecurely housed and saddled with substance abuse and/or mental health issues might do little to change their practices with respect to young men of color. Officer-led harm reduction is thus not a wholly sufficient police reform.

In an ideal world, housing, health care, addiction treatment, and mental health services would be available to all who need them, mass incarceration would not exist, and there would be no need to ask the police to play a prominent role in the provision of social services. Unfortunately, this ideal world is hard to imagine in reality. We argue that in this context, officer-led harm-reduction diversion can serve a variety of positive ends, including enhanced service delivery, criminal justice diversion, and the transformation of a police culture that historically promoted a more adversarial approach to the socially marginal. Rather than recruiting officers to use their arrest and other coercive powers to seek to deter criminal and disorderly behavior, officer-led harm reduction asks them to take a more capacious approach. Officers are encouraged to recognize the underlying drivers—both structural and biographical—of much criminal behavior, and to partner constructively with caseworkers who can guide individuals toward helpful services. In the process, officers must rework common cultural norms that emphasize the centrality of coercion and deterrence, and embrace the idea that significant harm can be reduced even in the absence of abstinence.

As we have noted, the police will necessarily be overly involved in the lives of the poor. A police approach that fundamentally sees each individual as a rational actor will inevitably stress coercive tactics designed to impel better choices. This core assumption helps explain the failure of aggressive patrol and coercive benevolence. Officer-involved harm reduction invites officers to see poor neighborhoods and poor people in a different fashion, and gives those officers an opportunity to challenge cultural norms that bear reconsideration.

Yet the police do not bear sole responsibility for their well-worn emphasis on deterrence. When concerns about criminality emerge, it is customary for political leaders and others to turn to the police for an ostensive solution. In Seattle, for example, during the same period that LEAD was demonstrating initial success, concerns about disorder in one particular area downtown led to the “nine and a half block” strategy. This initiative, endorsed strongly by police and political leaders alike, emphasized the value of making large numbers of arrests and maintaining visible patrol. When there is a widespread sense that something must be done, the police and their coercive powers are hailed and put to inevitably limited effect.

For this reason, it is important that LEAD and similar initiatives retain widespread and consistent political support. Officer-led harm reduction cannot solve all the problems of poverty and addiction, and thus its successes will be constrained, and likely not even visible. Absent concerted political support, officer-led harm reduction might be susceptible to efforts to resort to the time-honored tradition of aggressive patrol.

In these political machinations, it would also be helpful to remember that the police are limited in their ability to meaningfully impact the manifestations of urban inequality. As the LEAD story illustrates, even when officers are willing to participate in steering the disadvantaged to services, those services might not always be available to the degree necessary. More generally, absent access to housing and jobs, the socially marginal will remain largely in the same status, and thereby commonly act in a manner that attracts police attention. Ideally, the social support system would be robust enough to sharply reduce police focus on the downtrodden. With enough jobs, supportive housing, and mental health services, the police's involvement with the poor would be reduced. But that day may never arrive, especially in a neoliberalized United States, and thus the police's role will likely remain a significant one. The question, then, is just what role they will play.

It is understandable why many urban residents struggle to see the police as legitimate. This is, in part, a function of police patrol practices that disproportionately concentrate coercive force on certain groups of people, especially young men of color. But it is also likely a function of widespread disaffection with social marginality itself. Because the police are the most obvious symbol of state power, they unwittingly attract scorn from those who suffer under the weight of structural forces that constrain their opportunities.

Thus, we strongly endorse efforts to embrace officer-involved diversion guided by harm-reduction principles. It is clearly a superior alternative to aggressive patrol or coercive benevolence, in no small part because it works to transform unhelpful cultural norms within police departments. We are thus inclined to believe that a move in this direction will likely enhance the legitimacy of the police in the eyes of those who are often on the receiving end of the coercive sticks that officers wield. At the same time, such legitimacy will likely be enhanced even further if social marginality were itself the focus of the same concerted attention often directed at crime and disorder. If our collective goal is to enhance the legitimacy of the police, then reducing the prevalence and ill effects of urban inequality is likely the best path to ensure that outcome.