CRIMINAL JUSTICE

1017

BEYOND THE LAW: AN AGENDA FOR POLICING REFORM

MEGAN QUATTLEBAUM & TOM TYLER

ABSTRACT

Legal discussions about how best to manage the use of force by the police have focused on possible changes in the legal standards through which the police are held accountable for their actions—standards established in Graham v. Connor. We argue that such changes are unlikely to change police conduct in desirable ways. The police are currently trained and equipped to manage all the problems they face through the threat or use of coercion even though this approach is a poor fit to most of the issues the police actually deal with in their everyday work. The consequence of this mismatch is that police actions provoke and intensify conflict in many settings, leading inevitably to instances of the overuse of force. A better solution to the problem of the police use of force is to focus beyond the law on the organization of policing. We propose four approaches for possible reorganizations of the police. The goal of the first two approaches—exiting the social welfare field and collaboration with nonpolicing agencies—is to limit police actions to those situations in which their willingness and ability to compel obedience via force is appropriate. The other two approaches—specialization and civilianization—aim to diversify the skill set of police so that some members of the department are trained, equipped, and able to be deployed to deal with the variety of problems that can be better handled through a “social welfare” skill set.

 Director of the Council of State Governments Justice Center.  Macklin Fleming Professor of Law, Professor of Psychology, Yale Law School;

Founding Director of The Justice Collaboratory. Thanks to Sam Breidbart and José Girón Arias for exceptional research assistance.

1018 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

CONTENTS

INTRODUCTION ............................................................................................. 1019 I. LITIGATION AND ITS LIMITS .............................................................. 1020 II. A MISMATCH BETWEEN OFFICERS AND THEIR JOBS ......................... 1024 III. AN AGENDA FOR THE NEXT WAVE OF POLICING REFORM ............... 1027

A. Changes in Police Deployment .................................................. 1028 1. Approach One: Exiting the Social Welfare Field ................ 1028 2. Approach Two: Collaboration with Nonpolicing

Agencies .............................................................................. 1033 B. Changes Within Policing ........................................................... 1035

1. Approach Three: Specialization .......................................... 1036 2. Approach Four: Civilianization ........................................... 1038

a. Paraprofessionals ......................................................... 1040 b. Civilian Experts ............................................................ 1043

3. Implementing the Changes .................................................. 1043 CONCLUSION ................................................................................................. 1044

2020] AN AGENDA FOR POLICING REFORM 1019

INTRODUCTION

This symposium follows the traditional legal approach of attempting to expand liability so that law enforcement officers will face consequences when they use force in an impermissible way. We suggest that effectively addressing the issue of police excessive force requires responding to a more fundamental reality: the current police organizational model (1) requires that officers address what we term “social welfare” problems but (2) fails to provide them with the training and equipment they need to do so.

Though drawing hard-and-fast lines is challenging in such a high-discretion profession, we define social welfare responsibilities to include those police tasks that are not undertaken for the purpose of preventing and investigating crime or apprehending people who have committed crimes. Examples include responding to private or family disputes (except those that involve domestic violence or other criminal activity); enhancing individual or community well-being; and abating poverty, addiction, mental illness, and their symptoms. Such tasks are themselves many and varied, and they include maintaining public order (by policing public drunkenness, public urination, aggressive panhandling, etc.), resolving family or neighbor disputes (loud music, troubles with children, etc.), enforcing traffic laws, and subduing difficult or disorderly persons (who may or may not be mentally ill). Managing these social welfare problems effectively requires, at a minimum, incredible patience and a steady demeanor; expertise in social work, psychology, and similar fields would also be highly beneficial. One thing that is not necessarily required—and that could make many such situations demonstrably worse—is the threat or use of physical force.

Despite this reality, law enforcement agencies are generally structured to employ a fleet of armed generalists who are recruited, equipped, and trained in ways that lead them to approach the wide variety of problems they confront through a framework of force-backed, command-and-control tactics. To address this mismatch, legal authorities need to step back from the focus on Graham v. Connor’s1 objective “reasonableness” standard. Instead, they need to ask whether policing occurs within an institutional framework that equips officers with the skills they need to address the social welfare problems that confront them. Alternatively, they need to ask whether police should be addressing social welfare problems to begin with.

The solutions proposed for the problem of police excessive force have tended not to address these underlying organizational issues. California, for example, now requires that police reasonably believe that force is necessary before they are allowed to deploy it.2 Our argument is that sending the same type of officers to manage the type of problems that the police typically encounter is likely to lead to problems with the use of force irrespective of which legal standard is deployed. What is needed is a reexamination of the way that agencies recruit, train, and deploy police officers.

1 490 U.S. 386 (1989). 2 See CAL. PENAL CODE § 835(a) (West 2020).

1020 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

In this Essay we propose and evaluate four, nonmutually exclusive approaches to organizational reform, all of which are already underway to varying degrees in different departments.

The first two approaches adopt the current model of police training while reducing or eliminating the role police play in performing social welfare tasks. Officers continue to be trained and deployed as they are now—that is, as armed officers trained to use force to compel compliance. How they are deployed changes. In the latter two approaches, the skill set of officers is differentiated and at least some officers expand their repertoire of strategies to emphasize skills more helpful in managing “social welfare” tasks. This can include having more unarmed officers or having more officers trained in specialized skills, such as empathy building and de-escalation. The approaches are as follows:

(1) Exiting the social welfare field: police departments could elect not to perform social welfare functions, turning those responsibilities over to other government agencies.

(2) Collaboration with nonpolicing agencies: police departments could collaborate with government agencies or nonprofits that have expertise in relevant areas.

(3) Specialization within law enforcement organizations: police departments could differentiate policing tasks, train officers differently depending upon which tasks they will be assigned, and then dispatch appropriately specialized officers. Some officers would be trained to perform “social welfare” functions.

(4) Civilianization within departments: police departments could draw in people in allied professions (e.g., psychology, social work) and include them in efforts to respond to “social welfare” problems.

Ultimately, we conclude that police departments will be best served by utilizing all four of these approaches in combination, with the precise mix of each depending upon local concerns and conditions. Our argument is that structural changes in advance of any particular police interaction are key to increasing the safety of officers and those they police.

I. LITIGATION AND ITS LIMITS

Thirty years ago, the Supreme Court decided Graham v. Connor. The Graham Court held that “all claims that law enforcement officers have used excessive force . . . should be analyzed under the Fourth Amendment and its ‘reasonableness’ standard.”3 Reasonableness, the opinion declared, would be “judged from the perspective of a reasonable officer on the scene, rather than

3 Graham, 490 U.S. at 395. Graham extended Tennessee v. Garner, which held that police use of deadly force must comply with the Fourth Amendment. See Tennessee v. Garner, 471 U.S. 1, 7 (1985).

2020] AN AGENDA FOR POLICING REFORM 1021

with the 20/20 vision of hindsight.”4 The majority listed three factors for courts to consider when determining the reasonableness of police conduct, including “the severity of the crime at issue, whether the suspect poses an immediate threat to the safety of the officers or others, and whether [the suspect] is actively resisting arrest or attempting to evade arrest by flight.”5 Importantly, the Court barred plaintiffs from arguing that police excessive force violates the Due Process Clause of the Fourteenth Amendment.6

A great number of people have criticized Graham.7 Among other things, they argue that Graham makes it unnecessarily difficult for plaintiffs to overcome

4 Graham, 490 U.S. at 396. 5 Id. 6 See id. at 388. Before Graham was decided, some courts held that the Fourteenth

Amendment’s due process guarantee protected citizens from undue police force. See, e.g., Rutherford v. City of Berkeley, 780 F.2d 1444, 1446 (9th Cir. 1986) (holding that plaintiff stated valid excessive-force claim when he alleged that arresting officer’s use of force violated substantive due process); Johnson v. Glick, 481 F.2d 1028, 1032 (2d Cir. 1973) (“[A]pplication of undue force by law enforcement officers deprives a suspect of liberty without due process of law.”).

7 See, e.g., Geoffrey P. Alpert & William C. Smith, How Reasonable Is the Reasonable Man? Police and Excessive Force, 85 J. CRIM. L. & CRIMINOLOGY 481, 486 (1994) (stating that Graham’s “reasonableness standard” has “forced police departments to create policies on the use of force that are unworkable”); Michael Avery, Unreasonable Seizures of Unreasonable People: Defining the Totality of Circumstances Relevant to Assessing the Police Use of Force Against Emotionally Disturbed People, 34 COLUM. HUM. RTS. L. REV. 261, 320-21 (2003) (arguing that Graham allows for overreliance on “split-second decisions” as justification for excessive force); Avidan Y. Cover, Reconstructing the Right Against Excessive Force, 68 FLA. L. REV. 1773, 1818 (2016) (“The Court’s rejection of due process as a constitutional limit on excessive force in Graham was mistaken.”); Rachel A. Harmon, When Is Police Violence Justified, 102 NW. U. L. REV. 1119, 1119 (2008) (describing excessive-force doctrine as “deeply impoverished, . . . indeterminate[,] and undertheorized”); John C. Jeffries, Jr., The Liability Rule for Constitutional Torts, 99 VA. L. REV. 207, 266 (2013) (arguing that Graham’s discussion of qualified immunity is “confusing and has worked much mischief”); Cynthia Lee, Reforming the Law on Police Use of Deadly Force: De-escalation, Preseizure Conduct, and Imperfect Self-Defense, 2018 U. ILL. L. REV. 629, 645 (“One problem with the Graham Court’s embrace of reasonableness is that racial stereotypes about Blacks and other racial minorities can affect perceptions of whether an officer’s use of force was reasonable.”); Cara McClellan, Dismantling the Trap: Untangling the Chain of Events in Excessive Force Claims, 8 COLUM. J. RACE & L. 1, 8 (2017) (describing how lower courts apply Graham differently); Osagie K. Obasogie & Zachary Newman, The Futile Fourth Amendment: Understanding Police Excessive Force Doctrine Through an Empirical Assessment of Graham v. Connor, 112 NW. U. L. REV. 1465, 1465 (2018) (“The Court’s doctrinal choice in Graham has contributed to the perpetuation of police excessive use of force in many communities of color.”); Seth W. Stoughton, Policing Facts, 88 TUL. L. REV. 847, 868 (2014) (arguing that Graham Court’s factual description of police work as involving “split-second judgments” is flawed); Aaron Kimber, Note, Righteous Shooting, Unreasonable Seizure? The Relevance of an Officer’s Pre-seizure Conduct in an Excessive Force Claim, 13 WM. & MARY BILL RTS. J. 651, 665 (2004) (arguing that Graham standard is

1022 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

qualified immunity,8 that Graham stymies the development of excessive-force law,9 that the reasonableness factors listed in Graham do not provide officers with adequate guidance,10 that Graham is too indeterminate,11 and that Graham makes dubious factual commitments about the nature and dangerousness of policing.12

The latter point deserves some discussion. The Court in Graham stated that “police officers are often forced to make split-second judgments—in circumstances that are tense, uncertain, and rapidly evolving—about the amount of force that is necessary in a particular situation.”13 But Professor Seth Stoughton has argued that when police officers use force, they rarely do so under those circumstances. The type of resistance suspects typically offer is nonviolent—a suspect will fail to obey an officer, pull away, or flee.14 Accordingly, the measures police most often deploy are usually not physical— an officer will shout, curse, or threaten to use force.15 On the rare occasions police officers do use force, they typically use very little—most often a push or a grab.16 This is supported by the fact that police-citizen encounters that involve the use of force seldom result in injury to the officer or the citizen.17 Thus, when a police officer uses force, they usually do so offensively to induce compliance, not defensively to protect their own safety or the safety of an innocent bystander.

The (1) level of force police use; (2) frequency with which police use force; and (3) offensiveness, not defensiveness, that characterizes police use of force all paint a different factual picture than that which the Court described in its opinions in Graham and Graham’s predecessor case, Tennessee v. Garner.18 As Stoughton puts it, “[T]he realities of police violence are such that the circumstances in which officers must make a truly split-second decision are highly unusual, which militates against the Supreme Court’s generalization.”19

“arbitrary because it focuses on how one can categorize the pre-seizure encounter, rather than looking at how that conduct affects the eventual use of force”).

8 See Cover, supra note 7, at 1807. 9 See id. at 1789-90. 10 See Harmon, supra note 7, at 1125-46. 11 See id. at 1129-30. 12 Stoughton, supra note 7, at 867-68 (finding that police use force primarily to induce

compliance, not to defend themselves). 13 Graham v. Connor, 490 U.S. 386, 396-97 (1989). 14 See Stoughton, supra note 7, at 867 (describing these types of actions as “by far the most

common types of opposition that officers must overcome”). 15 See id. (“Correspondingly, the most common officer responses involve the use of

nonphysical force: shouting, cursing, and threats of force.”). 16 See id. 17 See id. at 867-68. 18 471 U.S. 1 (1985); see also supra note 3. 19 Stoughton, supra note 7, at 869.

2020] AN AGENDA FOR POLICING REFORM 1023

The consequence, he argues, is that the excessive-force doctrine gives too much deference and leeway to police officers.20

In light of these criticisms, academics have proposed reforms that replace or modify the Graham standard.21 Their goal is to curb excessive force by making it easier for plaintiffs to prevail in § 1983 excessive-force cases. This strategy comes with problems of its own.

First, some of the candidates to replace Graham would arguably make it harder, not easier, for plaintiffs to recover. Take the Fourteenth Amendment’s Equal Protection Clause.22 Under current equal protection doctrine, a government action will be subject to tiered scrutiny only if the plaintiff shows that the actor specifically intended to harm a group.23 A showing of such subjective intent is exceedingly difficult to make.24

Another alternative is the Due Process Clause of the Fourteenth Amendment. Indeed, when the Court decided Graham, “the vast majority of lower federal courts” applied due process analysis to excessive-force cases.25 Under that test, courts were to consider whether the officer applied force “maliciously and sadistically for the very purpose of causing harm.”26 Proving the officer’s malicious subjective intent made recovery difficult for plaintiffs. In short, the leading alternatives that critics offer to Fourth Amendment analysis would arguably lead to fewer awards and less police accountability.

Second, even if legal reforms expand excessive-force liability, there are several reasons to think that police conduct may remain unchanged. For every $100 that plaintiffs receive in police civil rights cases, only $0.02 comes out of

20 Id. 21 See, e.g., Alpert & Smith, supra note 7, at 501 (advocating for redefinition of

“reasonable person” standard for use-of-force cases); Obasogie & Newman, supra note 7, at 1498 (“[T]he Fourteenth Amendment and its greater conceptual sensitivity to and awareness of group dynamics . . . might be a more appropriate vehicle through which to adjudicate matters concerning excessive police force.”).

22 The Fourteenth Amendment provides that: [n]o State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any State deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws.

U.S. CONST. amend. XIV, § 1, cl. 2. 23 Pers. Adm’r v. Feeney, 442 U.S. 256, 279 (1979) (“[Discriminatory purpose] implies

that the decisionmaker . . . selected or reaffirmed a particular course of action at least in part ‘because of,’ not merely ‘in spite of,’ its adverse effects upon an identifiable group.”).

24 See Serena Mayeri, A New E.R.A. or a New Era? Amendment Advocacy and the Reconstitution of Feminism, 103 NW. U. L. REV. 1223, 1293 (2009) (stating that feminist lawyers had long criticized the intent requirement of equal protection cases).

25 Graham v. Connor, 490 U.S. 386, 393 (1989); see also Johnson v. Glick, 481 F.2d 1028, 1033 (2d Cir. 1973) (establishing four-part due process test).

26 Johnson, 481 F.2d at 1033.

1024 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

the pockets of the offending officers.27 Thus, an individual officer almost never internalizes the monetary cost that she generates when she uses excessive force. She has little financial incentive to correct course.

Nor is it certain that the financial pressures of litigation will change the way municipalities regulate the conduct of their police officers. Some academics believe that government administrators are sensitive only to political costs, not financial costs.28 There are a slew of reasons that police misconduct may generate financial costs but fail to generate political ones. Voters, for example, may not be aware of police misconduct or the burdens it places on a municipality’s budget.

The foregoing discussion suggests that legal reforms aimed at abandoning or changing Graham are unlikely to solve the problem of police use of excessive force. In the following Part, we discuss another powerful reason to believe that the solution to this problem lies beyond the law. Police officers spend the vast majority of their time performing nonenforcement tasks, many of which we categorize as “social welfare” functions. Despite the social work nature of these functions, police training and equipment teaches officers to perform these tasks primarily with force. As a result of this mismatch, police-citizen interactions may be more volatile and violent, increasing the incidence of questionable uses of force.

A focus on litigation misses important questions: First, should police be performing social welfare roles? Second, if so, how might we train and equip them to better perform these roles? Third, is there a better way for society to address the social welfare problems that police today spend so much time struggling to resolve?

It might seem obvious that to remedy this mismatch we should retain the existing policing model and give officers more tools to address more problems. Our argument is that there are actually several alternative organizational models that might be applied to law enforcement, all of which are different ways of answering the questions we pose.

II. A MISMATCH BETWEEN OFFICERS AND THEIR JOBS

Police time-and-task studies consistently show that police officers spend very little time on “reactive crime-solving activities.”29 As one author of such a study concluded, “It is unfortunate for the Country that the police are imbued with this totally wrong perception of themselves. The police do perform social work. In fact, they perform more social work than they perform law enforcement.

27 See Joanna C. Schwartz, Police Indemnification, 89 N.Y.U. L. REV. 885, 885 (2014). 28 See Joanna C. Schwartz, How Governments Pay: Lawsuits, Budgets, and Police

Reforms, 63 UCLA L. REV. 1144, 1152-53 (2016). 29 See Bernard E. Harcourt & Tracey L. Meares, Randomization and the Fourth

Amendment, 78 U. CHI. L. REV. 809, 821-29 (2010).

2020] AN AGENDA FOR POLICING REFORM 1025

Regretfully, as social workers, most police are poorly trained and incompetent.”30

Another study showed that suburban police officers spent most of their time patrolling, performing administrative tasks like report writing, performing off- duty tasks, traveling en route to specific locations, and enforcing traffic violations.31 Notably, only the final task involves contact with members of the public.32 In another study, officers in Indianapolis were observed for eight-hour stretches.33 Out of those eight hours, officers spent an average of two hours and six minutes on “general patrol” (i.e., “[t]raveling with no particular destination and without focus on a specific problem”); one hour and ten minutes “en route” to a specific destination; twenty-nine minutes on administrative tasks (including report writing, equipment maintenance, processing evidence, and appearing in court); twenty-two minutes on “information gathering” (such as reviewing records and meeting with other police officers); and one hour and five minutes on personal business.34 Thus, the majority of their day (five hours and twelve minutes) was spent on these tasks. “Encounters”—defined as “all face-to-face contact with citizens that included verbal exchange or physical contact”—took only one hour and fifty-six minutes out of the average day.35 “Problem-directed” activities—including traffic enforcement, surveillance, crowd regulation, escort, investigations of suspicious circumstances, and searches for or pursuits of individuals—took only fifty-one minutes.36 A generous interpretation of these figures would suggest that police officers spend only one-fourth of their average day investigating crimes or apprehending people who have committed them.

The fact that police officers do not spend most of their time fighting crime does not mean that they instead perform social welfare tasks. In actuality they perform many administrative tasks. The point is that police officers are trained as generalists who deploy force to compel compliance when that skill set is not central to much of their daily jobs and when this focus in policing prevents officers from learning the skills that would enable them to perform social welfare tasks. Indeed, from 2011 to 2013, 48% of new policing recruits attended a “[s]tress-based,” military-style training academy.37 Such training “typically

30 John A. Webster, Police Time and Task Study, 61 J. CRIM. L. 94, 100 (1970). 31 John Liederbach, Addressing the “Elephant in the Living Room”: An Observational

Study of the Work of Suburban Police, 28 POLICING 415, 417 (2005). 32 Id. 33 Roger B. Parks et al., How Officers Spend Their Time with the Community, 16 JUST. Q.

483, 497-98 (1999). 34 Id. 35 Id. 36 Id. 37 BRIAN A. REAVES, U.S. BUREAU OF JUSTICE STATISTICS, NCJ 249784, STATE AND LOCAL

LAW ENFORCEMENT TRAINING ACADEMIES, 2013, at 1 (2016), https://www.bjs.gov/content /pub/pdf/slleta13.pdf [https://perma.cc/8XCX-VN6L].

1026 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

involves intensive physical demands and psychological pressure.”38 And for all recruits, the topics emphasized in training academies—which run for an average of twenty-one weeks—make clear that police departments value expertise in using force more than a facility for social welfare tasks. For example, recruits spend an average of 168 hours on “weapons, defensive tactics and the use of force,” which includes seventy-one hours on firearms and sixty hours on self- defense.39 Although nearly all (97%) of academies required some training in community policing, recruits spent approximately forty hours on the topic—less than half of the time they spent on self-improvement.40 And while more than 90% of academies included training on mental illness, the ten hours spent on this topic was significantly less than the hours dedicated to report writing.

We thus recruit, hire, and train police officers for the job we imagine them to have rather than for the one they will actually be called upon to do. It remains the case that police officers are generalists, primarily trained in and capable of deploying a variety of levels of force. From domestic disputes to traffic violations to public drunkenness to reports of shots fired, the police who are sent to patrol the streets or respond to calls are generally the same armed officers and have received largely identical training. Consequently, they approach situations in similar ways: by emphasizing their capacity and willingness to use coercion.

Of course, as long as police carry firearms, they need to know when and how to use them. We do not take issue with training that trains police to use firearms. Rather, our point is that police training is not useful for the many other functions that the police normally perform. Because of the focus on force in police training, an officer’s conception of a “reasonable” course of action usually involves using force. Thus, the entry of the police into any situation defines that situation as one in which the potential use of force is an issue.

Indeed, when armed officers appear on the scene, their presence may have the effect of exacerbating, rather than calming, the situation. This may be especially true in communities that distrust the police. A consequence is the escalation of conflict, leading to increasing uses of force. In fact, studies suggest that coercion is not a particularly effective mechanism for obtaining compliance and that such requests lead to resistance and defiance.41 As McCluskey notes, “[T]he coercive power that police bring to bear on a citizen in the form of commanding, handcuffing, arresting and so on, has a minimal impact on citizen’s compliance

38 Id. 39 Id. at 5. 40 Id. at 5-7. This training covered topics such as “identify[ing] community problems

(77%), the history of community-oriented policing (75%), interacting with youth (62%), using problem-solving models (61%), environmental causes of crime (57%), and prioritizing crime and disorder problems (51%).” Id. at 7.

41 JOHN D. MCCLUSKEY, POLICE REQUESTS FOR COMPLIANCE: COERCIVE AND PROCEDURALLY JUST TACTICS 108 (2003) (“The higher the level of coercive action displayed by police, the less likely targets are to comply.”).

2020] AN AGENDA FOR POLICING REFORM 1027

decision.”42 This is true because “[f]or every one unit increase [in] the index of coercion citizens are about twice as likely to rebel against the self-control request.”43

This persistence of a “one person fits all needs” model is not inevitable; in other professions, role differentiation has become the norm. Every call for medical help does not lead to a doctor stepping out of an ambulance, but virtually every call for law enforcement helps produce an armed officer. There are many paramedics but few parapolice officers.

This “one model fits all” approach is costly both to local budgets and to American law enforcement officers and the communities they serve. It is financially costly because the overwhelming majority of the activities undertaken by law enforcement officers do not require the ability to deploy force.44 Hence, the additional training and certification needed to be able to deploy deadly force is a costly and underutilized extra skill set. This model is also costly to communities and to law enforcement officers because the mismatch between officer skills and tasks can, as described above, have the perverse effect of making conflict more likely. And because officers are not adequately trained in the type of social work skills that would allow them to address interpersonal issues, they are not able to effectively meet many of the community needs with which they are confronted.45 Over time, this promotes legal cynicism in minority communities, which feel dominated and constrained, and heightens stress among officers who experience tension when dealing with civilians whom they perceive to be hostile.46

III. AN AGENDA FOR THE NEXT WAVE OF POLICING REFORM

What do we want law enforcement officers to do and who are the right officers to do that job? Clear answers to these questions, rather than a litigation strategy, will make law enforcement safer for officers and the public.

We see two primary lines of response yielding four possible approaches for reform. The first two approaches involve changes in police deployment but not police skills. Most radically, the police could exit the social welfare field. Americans could seek to reduce police officers’ functions down to the smallest, hard core of violent and other serious crime problems (armed robbery, murder, terrorism, cybercrime, public corruption, and the like) for which we see no other possible response. To make this exit effective, states would be required to

42 Id. at 100. 43 Id. at 108. 44 See Webster, supra note 30, at 100 (“[Patrolmen] spend more than 50 percent of their

time on administrative tasks. Traffic, police initiated events, social service, and crimes against property account for about 47 percent of their time. Less than three (3) percent of the Baywood patrolman’s time is spent on dispatches concerning crimes against persons.”).

45 Id. 46 Monica C. Bell, Police Reform and the Dismantling of Legal Estrangement, 126 YALE

L.J. 2054, 2068-72 (2017).

1028 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

decriminalize minor offenses against public order. This would result in the deployment of officers in situations in which serious crimes were involved.

Presumably, other state and local agencies would need to be staffed, funded, trained, and empowered to address those social welfare problems that police officers would no longer handle. Therefore, a more modest version of this approach would be for law enforcement agencies to pull back from social welfare tasks without fully exiting the field. This would require collaboration with nonpolicing agencies.

The second primary line of response would be to rethink how we recruit, equip, and train police officers such that not all of them are armed and many of them are civilian specialists. Under this approach, we would seek to ensure that police officers do a better job at their inevitable—and perhaps even desirable— social welfare functions. We suggest two models for such a reorganization: specialization within law enforcement organizations and civilianization.

Because the first line of response would require the most dramatic changes, we discuss it first.

A. Changes in Police Deployment

1. Approach One: Exiting the Social Welfare Field

Recently, some activists and scholars have asserted that “it is a flawed endeavor to use the criminal justice system to address the manifestations of social disadvantage”47 and have advocated for a “shrinking and refining [of] the police footprint” such that police officers no longer take on a social welfare role.48 Some (but not all) concede that, “[a]s coercive agents, the police do have a role to play. They remain a vital means of resolving immediate crises, of promoting a sense of safety, [and] of responding to specific instances of criminality.”49 But they worry that “routing rehabilitation and social services through the police [may] perversely widen the carceral net and reify the ‘culture

47 KATHERINE BECKETT & STEVE HERBERT, BANISHED: THE NEW SOCIAL CONTROL IN URBAN AMERICA 153 (2009); see also VICTOR M. RIOS, PUNISHED: POLICING THE LIVES OF BLACK AND LATINO BOYS 20 (2011) (criticizing “trend to use crime-control metaphors and material resources to solve non-criminal, everyday social problems”).

48 Bell, supra note 46, at 2149; see also KATE HAMAJI ET AL., CTR. FOR POPULAR DEMOCRACY, LAW FOR BLACK LIVES & BLACK YOUTH PROJECT 100, FREEDOM TO THRIVE: REIMAGINING SAFETY & SECURITY IN OUR COMMUNITIES 1 (2017), https://popular democracy.org/news/publications/freedom-thrive-reimagining-safety-security-our- communities [https://perma.cc/N3HR-8JUC] (profiling twenty-five community organizations that are advancing “invest/divest” campaigns “that seek to divest resources away from police and prisons towards communities and their development”).

49 STEVE HERBERT, CITIZENS, COPS, AND POWER: RECOGNIZING THE LIMITS OF COMMUNITY 140 (2006).

2020] AN AGENDA FOR POLICING REFORM 1029

of control.’”50 They conclude that police officers should not be involved in social welfare tasks.51

Even if we were to successfully limit police officers to being specialists in crime fighting, the extensive nature of our penal codes means that they would still be engaged with many problems that we suspect these same activists and scholars would view as “manifestations of social disadvantage,” including public drunkenness, public urination, loitering, possession of drugs for personal use, and failing to pay the fare on public transportation. Thus, successfully reorienting policing in this way would likely require the decriminalization of these sorts of behaviors.

Moreover, as the scholars who advocate shrinking the footprint acknowledge, it is not clear which governmental or societal actors could be counted on to take over police departments’ discarded social welfare tasks. “The use of [police] to deal with difficult, awkward, or troubling situations primarily reflects the lack of alternatives. In the absence of adequate low-income housing, shelter beds, drug and alcohol treatment programs, and inpatient care facilities,” it is tempting to deploy police officers to try to “make the problem go away.”52

This is not to suggest that the game is not worth the candle. It is self-evident that it is best to, for example, provide adequate community health resources to those with mental or behavioral health problems before they come into contact with the criminal justice system. Studies suggest that doing so will be cheaper and more effective than relying on police and jails to address these problems.53

Indeed, because the current model uses expensive sworn officers for all police functions in the community, there are no low-cost uses of the police. If the police did not have to perform low-level order and maintenance functions performable by other city agencies, significant resources would become available to fund more social services. Consider the situation in medicine. For example, if every ambulance contained a doctor, medical costs would rise dramatically. Instead, by building initial contact around paramedics, the system is able to allocate

50 Bell, supra note 46, at 2147. 51 For similar arguments, see HERBERT, supra note 49, at 139 (“[I]f community

regeneration is a priority, why foreground the police in any such effort? As agents of coercive force, the police do not make especially good community builders. . . . In short, it is better to constrict, not expand, the role of the police in any project aimed at community regeneration.”); and ALEX S. VITALE, THE END OF POLICING 16-17 (2017).

52 BECKETT, supra note 47, at 83. 53 See, e.g., DAVID CLOUD & CHELSEA DAVIS, VERA INST. OF JUSTICE, TREATMENT

ALTERNATIVES TO INCARCERATION FOR PEOPLE WITH MENTAL HEALTH NEEDS IN THE CRIMINAL JUSTICE SYSTEM: THE COST-SAVINGS IMPLICATIONS 1-4 (2013), https://www.vera. org/downloads/Publications/treatment-alternatives-to-incarceration-for-people-with-mental- health-needs-in-the-criminal-justice-system-the-cost-savings-implications/legacy_ downloads/treatment-alternatives-to-incarceration.pdf [https://perma.cc/K3AU-VRKQ].

1030 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

resources more efficiently.54 Once this idea is accepted, the key issue is generally considering cost-versus-quality trade-offs.

Building community organizations pays off. For example, studies suggest that providing housing to individuals who are homeless can be more cost-effective than permitting them to cycle into the criminal justice system.55 Moreover, early research suggests that local nonprofits may cause crime rates to decline, with every new organization formed in a given city of 100,000 people leading to a 1% drop in violent crime and murder there.56 Similarly, crime rates decline when people have jobs, so organizations that focus on providing employment opportunities lower the rate of crime.57 This makes investing in the formation of such organizations—rather than focusing our attention exclusively on police departments—a compelling value proposition from the crime-prevention perspective.

For the exit-the-field approach to result in savings, however, administrators would have to shift resources from the police to the social service organizations that would replace them. This has proven to be easier said than done. While the crime rate today is substantially lower than it was in the past, police forces have actually gotten larger.58 In 1992, the ratio of sworn officers to civilians was 2.23.

54 Does this approach undermine service delivery? There is some evidence that having a doctor in every ambulance would improve medical outcomes. See Akihito Hagihara et al., Physician Presence in an Ambulance Car Is Associated with Increased Survival in Out-of- Hospital Cardiac Arrest: A Prospective Cohort Analysis, PLOS ONE, Jan. 8, 2014, at 1, 1. The challenge for emergency medical services (“EMS”) is similar to that of police calls. The majority of calls are not for true medical emergencies, so sending a doctor would have no additional value. At a system level, whatever loss in medical care occurs through the absence of a doctor in some situations is counterbalanced against the system’s capacity to use freed- up resources to fund more EMS units and cut response times. Funds not spent paying doctors to go to nonemergency situations can be reallocated to improve the system of medical care. Id.

55 See, e.g., Angela Ly & Eric Latimer, Housing First Impact on Costs and Associated Cost Offsets: A Review of the Literature, 60 CANADIAN J. PSYCHIATRY 475, 482 (2015) (“Most studies have observed decreases in justice costs. Homeless people are often arrested for crimes associated with survival strategies, such as entering private property or sleeping on a park bench. Also, a positive association between more severe psychiatric symptoms and nonviolent crimes has been observed. By providing housing to homeless people and support to stabilize mental health symptoms, a decrease in police contacts, arrests, detentions, and court appearances can be expected.” (footnotes omitted)).

56 PATRICK SHARKEY, UNEASY PEACE: THE GREAT CRIME DECLINE, THE RENEWAL OF CITY LIFE, AND THE NEXT WAR ON VIOLENCE 53 (2018).

57 Aaron Chalfin & Justin McCrary, Criminal Deterrence: A Review of the Literature, 55 J. ECON. LITERATURE 5, 36 (2017).

58 NATHAN JAMES, CONG. RESEARCH SERV., R45236, RECENT VIOLENT CRIME TRENDS IN THE U.S. 1 (2018), https://fas.org/sgp/crs/misc/R45236.pdf [https://perma.cc/LS6A-R6UA].

2020] AN AGENDA FOR POLICING REFORM 1031

In 2017, it was 2.40.59 This raises the question of what justifies maintaining large police departments in an era of low crime if the police are only or primarily agents of crime control. Because police agencies have a lot of political clout, diminishing their budgets is likely to be difficult.

The strongest and best arguments for this kind of shift would thus involve a substantial downsizing of police departments and a direct reinvestment of the savings generated into supportive community programming that is designed to address the problems that police departments used to tackle. Such a shift would be in line with the ideas of local divest/invest campaigns, which have been launched in jurisdictions across the country.60 The goal of such campaigns is to “divest resources away from police and prisons toward communities and their development.”61 Such programs thus seem like a good model for downsizing law enforcement and reinvesting in social service provisions that occur outside of the criminal justice system.

But we should be mindful of the fact that while police officers, community antipolice activists, and the academic literature resist and critique social welfare policing, that model has proved broadly popular. Indeed, the “social work” aspects of modern-day policing are often equally strongly pressed by communities as necessary and resisted by rank-and-file officers, who may perceive them as unexciting, undesirable, and implicitly feminine.62 A study of young people living in three high-crime Philadelphia neighborhoods found that “[d]espite the often unfavorable dispositions the youth . . . had toward the police,” when asked “what they would do to reduce crime and disorder in the city, most responses centered on improving or augmenting law enforcement.”63 In part, the authors argue, this seeming inconsistency “can be explained as youth advocating for the types of responses to crime that they know from experience”; “because they are exposed to saturation policing on a regular basis [they] . . . perceive it as the logical response to crime.”64 Certainly, community residents form their preferences in light of present-day realities. A reform proposal that would eliminate social welfare tasks from police departments’ lists of responsibilities would thus be grounded in the assumption that communities would be at least as happy if social welfare functions were performed by nonpolice agents of the state.

59 DUREN BANKS ET AL., U.S. BUREAU OF JUSTICE STATISTICS, NCJ 249681, NATIONAL SOURCES OF LAW ENFORCEMENT EMPLOYMENT DATA 15 (2016), https://www.bjs.gov/content /pub/pdf/nsleed.pdf [https://perma.cc/WD8G-KA3U].

60 HAMAJI ET AL., supra note 48, at 97. 61 Id. at 1. 62 HERBERT, supra note 49, at 97. 63 Patrick J. Carr, Laura Napolitano & Jessica Keating, We Never Call the Cops and Here

Is Why: A Qualitative Examination of Legal Cynicism in Three Philadelphia Neighborhoods, 45 CRIMINOLOGY 445, 461 (2007).

64 Id. at 468.

1032 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

Still, it is worth remembering that during the mid-century professional era, when police departments did self-consciously attempt to limit themselves to a crime-only mission, they were not particularly popular in their communities. Having social welfare activities performed by other government agencies could thus theoretically make local government more popular but would deprive the police of an opportunity to present themselves to the community in a more benevolent and caring light. Today’s police leaders may understand this, which could explain why many of them seem to seek out social welfare activities as a way of earning community legitimacy. For example, a document entitled Next Steps for Reform, released by the Chicago Police Department in the wake of the police shooting of Laquan McDonald, features on its cover pictures of police officers barbequing, playing chess, and shopping with members of the community, many of them children.65 The not-so-subtle message is that the activities pictured—fixing a child’s scooter, zipping a toddler’s jacket—are within the central activities of policing.66

Another issue for an exit-the-field strategy is what functions the police will exit. One way of reasoning about the question of whether policing agencies should ever engage in social welfare tasks is to ask oneself what one sees as the balance of outcomes that police departments should be aiming to achieve. We might then authorize police officers to undertake all activities that are (1) legally permitted and (2) democratically endorsed and that (3) demonstrate success in achieving the desired balance of outcomes.

We imagine that most people would answer the question about what the goals of policing should be by focusing on public safety. We agree, and we would add that in our view safety should mean freedom from predation perpetrated by private individuals and government actors. Importantly, there will be times when these goals will be in tension. But “[p]ublic safety is not just the reduction of harm, the maintenance of order, or promotion of security. Rather, safety requires freedom from personal victimization, community disenfranchisement, and government overreach.”67 We would thus judge police departments based upon their capacity to maximize individual and community safety and well- being while minimizing the use of punishment and physical force.

Taking these as our goals for policing agencies, we imagine that they would engage in some social welfare activities in order to achieve said goals. If, for

65 See generally CHI. POLICE DEP’T, NEXT STEPS FOR REFORM (2018), https://policy.chicagopolice.org/cpds-next-steps-for-reform/ [https://perma.cc/L6JX-94R3].

66 Other police departments also highlight their social welfare activities as a point of pride. MPD 2.0: A New Policing Model, a promotional book distributed by the Minneapolis Police Department, highlights officer involvement in the Police Athletic League; their creation of “Little Free Libraries”; and their distribution of hats, blankets, and sleeping bags to the needy during the winter months. See generally JANEÉ HARTEAU, MINNEAPOLIS POLICE DEP’T, MPD 2.0: A NEW POLICING MODEL (2015) (ebook).

67 Our Approach, YALE LAW SCH.: JUSTICE COLLABORATORY, https://law.yale.edu /centers-workshops/justice-collaboratory/justice-forward [https://perma.cc/QN86-GSJ7] (last visited Apr. 20, 2020).

2020] AN AGENDA FOR POLICING REFORM 1033

example, setting up a Saturday night basketball program in a neighborhood that offers few opportunities for young people to engage in social activities on the weekends proves to be a successful way of reducing the incidence of violence in a neighborhood, would we want police departments to be in the business of setting up such programs? We say a conditional “yes,” though for us the answer depends on what type of police officers staff the program and whether their involvement is a stalking horse for covert surveillance of the individuals who show up to play the game or a sincere and straightforward effort at community engagement and trust building.68 Often what is needed is support, not coercion.

This example demonstrates that the desirability of involving police officers in social welfare tasks is strongly related to the degree with which officer skills and training match that role. Because officers are not typically recruited for having such skills or trained around this type of skill set, departments likely do not have a deep pool of officers who can effectively engage in these types of community contact.

Our preferred strategy for reform is one that devolves some social welfare tasks away from policing agencies while also ensuring that more police officers are better recruited, trained, and equipped to take on those tasks.69 Below, we outline some steps that police departments could take to better align officer skills and training with the social welfare tasks we, as a society, currently ask those officers to perform. Each of these reform possibilities aims to move away from the model of having an armed, generalist police officer perform the full range of tasks currently assigned to police departments.

2. Approach Two: Collaboration with Nonpolicing Agencies

Collaboration with nonpolicing agencies occurs when officers avoid police involvement in everyday low-level crimes—crimes that do not immediately threaten the health or safety of people in the community or crimes involving special populations such as the mentally ill. However, unlike exiting the field, officers at the same time develop active partnerships with organizations in the community and facilitate their ability to address such problems.

The argument for collaboration is that law enforcement agencies are unlikely, particularly in the near term, to wholly abandon their social welfare role. And even with specialized civilians on staff, it is unlikely that police departments will be able to address internally all aspects of the myriad social welfare problems they confront. As Professor Wesley Skogan notes, “[I]f community policing is the police department’s program, important parts of it will fail. . . . Problem

68 Empirical research shows that the way people react to gestures of reconciliation by the police depends upon whether people perceive the motives of the police as being sincere. See Tom R. Tyler, Phillip Atiba Goff & Robert J. MacCoun, The Impact of Psychological Science on Policing in the United States: Procedural Justice, Legitimacy, and Effective Law Enforcement, 16 PSYCHOL. SCI. PUB. INT. 75, 76, 89 (2015).

69 One way to promote this goal is to train all officers in de-escalation tactics, procedural- justice approaches, and implicit bias. See id.

1034 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

solving takes sustained, government-wide commitment to the program, and many American cities do not succeed in developing this commitment.”70

Thus, reform-oriented departments may also choose to collaborate with other sectors of civil society. Examples of such cooperation could include “joint ventures with other government departments, national and international nonprofit organizations, and private-sector companies.”71 One concrete example of such collaborations has occurred in the area of responses to individuals with mental illness.72 So-called “police-mental health collaboration” models include “co-responder teams,” in which a “[s]pecially trained officer and a mental health crisis worker respond together to mental health calls for service,” and “mobile crisis teams,” which engage mental health professionals to respond to the scene of calls for police service in the hope of stabilizing those in crisis and avoiding unnecessary jail or emergency room intakes.73 Mobile crisis teams typically do not include a law enforcement officer, though the teams will show up in response to law enforcement requests for their assistance.

But we should be mindful that “[m]aking inter-organizational cooperation work can be one of the most difficult problems facing innovative departments.”74 Indeed, some authors find that “[t]he mental health care system itself appears to be a barrier for progress regarding this issue of policing the mentally ill. Social service agencies often refuse to admit intoxicated or psychotic persons referred by police.”75 Many of these problems have ready solutions, including the use of designated drop-off sites, a no-refusal policy, and a streamlined and consistent intake process.76 What is needed is the political will to encourage coordination and cooperation. The success of collaboration also depends upon the existence of viable alternative agencies with which the police can cooperate. In small communities, limited resources may lead the police to be the only group that is funded to a level such that it can address problems, leaving it without agencies to collaborate with.

70 Wesley G. Skogan, Why Reforms Fail, 18 POLICING & SOC’Y 23, 30 (2008). 71 Christopher Stone & Jeremy Travis, Toward a New Professionalism in Policing, 16 J.

INST. JUST. & INT’L STUD. 11, 27 (2013). 72 Abigail S. Tucker, Vincent B. Van Hasselt & Scott A. Russell, Law Enforcement

Responses to the Mentally Ill: An Evaluative Review, 8 BRIEF TREATMENT & CRISIS INTERVENTION 236, 244 (2008).

73 COUNCIL OF STATE GOV’TS JUSTICE CTR., POLICE-MENTAL HEALTH COLLABORATION PROGRAMS: A DIFFERENT WAY OF POLICING 9-12 (2016) (PowerPoint), https://csgjustice center.org/law-enforcement/webinars/police-mental-health-collaboration-programs-a- different-way-of-policing/ [https://perma.cc/52GK-YQNA] (discussing different mental health responsibilities of teams within law enforcement units).

74 Skogan, supra note 70, at 30. 75 Tucker, Van Hasselt & Russell, supra note 72, at 246-47. 76 Id. at 247.

2020] AN AGENDA FOR POLICING REFORM 1035

B. Changes Within Policing

The advantage of the exit-the-field strategy and of collaboration with outside agencies is that they do not test the ability of the police to adapt to a new role. The police stay as they are, and social welfare functions are performed outside the department. This is clearly the preference of many police leaders who complain that police are not equipped to manage functions such as assisting people who are mentally ill and in crisis. If the police focused on their crime- fighting role, they could maintain their current mission, and selection, training, and supervision could remain relatively unchanged.

But relying exclusively on a solution that involves having the police do less misses an opportunity to make beneficial changes to the character of policing agencies themselves. It also assumes that resources are available outside of the police department to manage the problems the police do not address. An alternative is for the police to repurpose themselves. We propose that, in order to properly orient themselves to a social welfare mission, police departments must engage in some combination of specialization and civilianization. We take these up in turn.

We note that, although such reforms have been implemented at the margins of police departments, what we are proposing is a comprehensive cultural and structural shift that will bring to an end the era of exclusive reliance on armed policing generalists. We imagine police departments in which many officers are civilians—some with specialties in psychology, investigation, or other fields— only a minority of whom are equipped for and tasked with the deployment of deadly force.

We note further that if police departments are more explicit about their social welfare goals and role, this will open the door to strategic thinking about how best to deploy their resources and abilities to achieve those goals. For example, Professors Jennifer Wood and Laura Beierschmitt have argued that the same insight that underlies so-called “hot spots” policing—that crime is concentrated in particular places—could also form the basis of police engagement in “hot spots of vulnerability” for behavioral health crises.77 That is, instead of merely engaging in “reactive and generic responses to calls” for service, police might engage in “focused interventions tailored to high risk groups and environments,” perhaps in partnership with other governmental agencies.78

In considering the idea of reorganizing the police so that some officers take on specialized skills, it is important to recognize that this assumes both that these retrained police can implement the new skills and that this is a good idea organizationally. One might argue that a cautionary example is American

77 Jennifer D. Wood & Laura Beierschmitt, Beyond Police Crisis Intervention: Moving “Upstream” to Manage Cases and Places of Behavioral Health Vulnerability, 37 INT’L J.L. & PSYCHIATRY 439, 440 (2014) (“[W]e stress the importance of establishing a data-driven approach to crime and disorder reduction in areas of the city we term ‘hotspots of vulnerability.’”).

78 Id. at 445.

1036 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

firefighting. As the number of fires has declined (by 57% between 1980 and 2013), the number of firefighters has not declined.79 On the contrary, the number of firefighters has increased 50%.80 How has this been justified? Firefighters have taken on tasks such as emergency medical response. This is the reason that fire engines respond to calls for medical assistance. Two issues are raised: The first is whether firefighters with expanded training are good emergency responders in nonfire situations. The second is that, even if firefighters are good emergency responders, this has created a situation in which expensively trained firefighters are performing tasks that might be performed by emergency-services workers (who lack the expensive training and skills required to be good at fighting fires).

1. Approach Three: Specialization

Police officers have always been asked to take on a dizzying array of responsibilities, and this trend escalates as society becomes more complex. They are called to the scene of bank robberies and to intervene when businesses wish to remove people who are panhandling from the sidewalks; they are called when there is a mass shooting and when parents feel they have lost control of their teenage children; and they are asked to prevent car accidents by ticketing reckless drivers and to prevent crime by patrolling neighborhoods in their squad cars or on foot. Very few individuals possess the skills to do all of these things well. Indeed, the fact that individual police officers get so many things right should humble all of us who have never spent a day in their shoes.

Still, we think it is a mistake on the level of departmental planning and leadership to expect generalist police officers to do all of these things and to do them well. As we have endeavored to show, the most glaring flaw in having police officers perform social welfare tasks is that they frequently lack the relevant skill set. Specialization is one way of addressing this problem. Police departments might differentiate policing tasks, train officers for the specific tasks to which they will be assigned, and dispatch appropriate types and levels of police to different situations and problems. This concept is already recognized in many departments that have SWAT teams, but it can also be implemented with units that are less force oriented.

Specialization has made some inroads in police departments. Police officers in New York were being trained in family crisis intervention as early as 1966.81 And today, Crisis Intervention Training—a forty-hour curriculum that teaches

79 Fred S. McChesney, With Fires, Why Are There Far More Firefighters?, WASH. POST, Sept. 4, 2015, at B01.

80 Id. 81 Morton Bard & Bernard Berkowitz, Training Police as Specialists in Family Crisis

Intervention: A Community Psychology Action Program, 3 COMMUNITY MENTAL HEALTH J. 315, 315 (1967).

2020] AN AGENDA FOR POLICING REFORM 1037

police officers the specialized skill of recognizing signs of mental illness and co- occurring disorders and how to de-escalate crisis situations—is widespread.82

“This first generation effort proves that police awareness and decision- making can be influenced in ways that balance public safety” against other objectives.83 “Positive results have been observed for outcomes[,] including reduced policing costs, decreased injuries to officers[,] and greater numbers of transports to mental health services.”84 But it remains the case that these specialization efforts have by-and-large focused in the mental health arena.85 Moreover, “the capacity of police to link people to behavioral health services is largely constrained to crisis situations.”86 In part because police leaders have failed to explicitly acknowledge law enforcement’s social welfare role, they have not been able to train, equip, or deploy their officers effectively to prevent those situations from reaching crisis stage in the first instance.87

One key to successful specialization is triage. Many requests for police assistance come through 911 calls. Those calls are often relayed to officers without much effort to define the nature of the problem; send appropriately trained officers; or even tell those officers dispatched that the call has special features, such as involving mental illness.88 One key antecedent to making

82 The model was developed in Memphis in 1988, and CIT International Inc. states that it is now used in forty-nine states and four countries. CIT INT’L, INC., CIT PROGRAM OVERVIEW (2017), http://www.citinternational.org/resources/Documents/CIT%20Program%20 Overview.2017.pdf [https://perma.cc/8M4Q-7GFG]. A competing model, Mental Health First Aid, is an eight-hour course that the National Council for Behavioral Health says has been completed by over 19,000 first responders, some of them police officers. See U.S. Reach, MENTAL HEALTH FIRST AID, https://www.mentalhealthfirstaid.org/algee-ometer/ [https://perma.cc/8UNV-QEAH] (last visited Apr. 20, 2020).

83 Wood & Beierschmitt, supra note 77, at 440. 84 Id. (citations omitted). 85 It is unsurprising that so much effort has been focused here. “Police officers routinely

provide the first line of crisis response for situations involving persons with mental illnesses. Individuals with severe mental illness generate no less than 1 in 10 calls for police service. These calls for service are common and constitute an estimated 7% of all police contacts.” COUNCIL OF STATE GOV’TS JUSTICE CTR., supra note 73, at 5.

86 Wood & Beierschmitt, supra note 77, at 443. 87 Some police departments do deploy a case management model in which certain

officers—often in collaboration with mental health professionals—carry a caseload of individuals who have had repeated interactions with law enforcement and work with them to develop solutions to avoid those interactions in the future. In particular, case management officers aim to keep people connected to mental health services and other community resources. COUNCIL OF STATE GOV’TS JUSTICE CTR., supra note 73, at 13.

88 S. REBECCA NEUSTETER ET AL., THE 911 CALL PROCESSING SYSTEM: A REVIEW OF THE LITERATURE AS IT RELATES TO POLICING 9 (2019), https://www.vera.org/downloads /publications/911-call-processing-system-review-of-policing-literature.pdf [https://perma.cc/WCQ4-7NSJ] (“Because there is no standardized protocol for police call- taking, the information gleaned during the call [from 911 dispatcher to police] may not . . . be optimized to give responders the information they need before arriving at the emergency.”).

1038 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

specialization work, then, is to build up the capabilities of 911 call centers to categorize the types of calls received and the manner of intervention that the situation requires.

Of course, we recognize that in the nearly three-quarters of American police departments that have fewer than twenty-five officers on staff, the possibilities for specialization will be limited.89 Moreover, in big and small departments, developing greater specialization within the police role will not, in and of itself, resolve the structural challenge that we have identified with law enforcement staffing.90 Sending sworn, armed officers to handle social welfare tasks is unnecessary, dangerous, and expensive—even if they are specialists. Police departments of all sizes can and must also explore the possibility of the civilianization of many positions.

2. Approach Four: Civilianization91

Some of the myriad tasks we assign to police officers self-evidently call for the intervention of agents of the state who are armed and trained for physical confrontation—resolving a hostage crisis, responding to an active shooter situation, and the like. But it is less clear why traveling five miles over a posted speed limit, for example, obviously invites an interaction with someone who carries a gun on her person.

“Civilianization refers to a law enforcement agency’s hiring of nonsworn personnel to replace or augment its corps of sworn officers, typically with the aims of reducing costs and improving service.”92 Notably, civilianization has tended to come along with specialization.93 That is, civilians are hired by police departments not as generalists but to perform specific functions with respect to which they may have special expertise.94 As Professor Brian Forst argues, “[I]t has become increasingly clear that civilians tend to perform certain specialized roles more effectively than sworn officers, who are selected and trained as generalists and then rotated from one assignment to the next accordingly.”95

89 OFFICE OF CMTY. ORIENTED POLICING SERVS., U.S. DOJ, FINAL REPORT OF THE PRESIDENT’S TASK FORCE ON 21ST CENTURY POLICING 29 (2015), https://cops.usdoj.gov /pdf/taskforce/taskforce_finalreport.pdf [https://perma.cc/RZ4Y-QPDS] (“Half of all law enforcement agencies in the United States have fewer than ten officers, and nearly three- quarters have fewer than 25 officers.”).

90 See supra Part II (arguing that mismatch between officer training, which focuses primarily on situations requiring lethal force, and actual tasks of officers, which include more social welfare tasks, results in excessive police force).

91 Civilianization can be a problematic term in the policing context, but we use it here because it is the one widely used to describe the phenomenon we discuss.

92 Brian Forst, The Privatization and Civilianization of Policing, in 2 CRIMINAL JUSTICE 2000: BOUNDARY CHANGES IN CRIMINAL JUSTICE ORGANIZATIONS 19, 23 (2000).

93 Id. (“Civilians are employed as communications specialists, criminalists . . . , computer specialists, lawyers, and a host of other support positions.”).

94 See id. 95 Id. at 24-25.

2020] AN AGENDA FOR POLICING REFORM 1039

We note that civilianization might proceed along two paths discussed in turn below: the employment of paraprofessionals and/or the employment of experts. And we are supportive of both. We define paraprofessionals as workers who are less comprehensively trained than their professional counterparts but who have specialized training that allows them to perform a narrow set of functions at least as well if not better than the professionals.

Experts, by contrast, are very highly trained in their area of specialized knowledge. We think that police departments could make use of both types of civilians, employing, for example, paraprofessionals to take on some routine tasks (like traffic enforcement) and experts (like psychologists) to perform complex tasks (like crisis intervention) more ably than their generalist counterparts.

Civilianization has made significant inroads in policing organizations. While estimates suggest that civilians made up 7% to 8% of police employees in the 1950s, they represented 22% of full-time employees of American municipal police departments by 2008.96 But the civilianization’s progress has been limited in two respects: First, civilians tend to supplement, not replace, sworn officers.97 Indeed, civilianization is promoted on the grounds that it frees up sworn officers “for the critical work of policing the streets.”98 Second, civilians tend to be deployed solely in administrative and operational positions that do not involve contact with the public.99

“[A]dvocates argue that civilians can be cheaper, more efficient, and easier to hire and lay off than sworn officers.”100 Indeed, cities have undertaken studies to explore the cost savings that would result from hiring civilians instead of sworn officers. The Houston Police Department estimated that hiring a civilian

96 Megan Alderden & Wesley G. Skogan, The Place of Civilians in Policing, 37 POLICING 259, 260 (2014) (summarizing research on percentage of civilians in policing from 1950 through 2008). The progress of civilianization has also seen some reversals. In Chicago, for example, civilian personnel dropped from 15% of the police department in 1993 to 6% in 2010. See CITY OF CHI. OFFICE OF THE INSPECTOR GEN., REVIEW OF OPPORTUNITIES FOR CIVILIANIZATION IN THE CHICAGO POLICE DEPARTMENT 8 (2013), http://chicagoinspector general.org/wp-content/uploads/2013/01/IGO-Opportunities-for-Civilianization-within- CPD-Final-1-23-13.pdf [https://perma.cc/33KN-MX8E]. But the country overall seems to be holding at a rate of civilianization of approximately 21%. See BRIAN A. REEVES, BUREAU OF JUSTICE STATISTICS, NCJ 248677, LOCAL POLICE DEPARTMENTS, 2013: PERSONNEL, POLICIES, AND PRACTICES 2 (2015), https://www.bjs.gov/content/pub/pdf/lpd13ppp.pdf [https:// perma.cc/D973-FJKE].

97 REEVES, supra note 96, at 1 (presenting data from Bureau of Justice Statistics showing that number of sworn police officers has increased simultaneously with increase in number of civilian employees in police departments since 1987).

98 Forst, supra note 92, at 37. 99 Id. 100 Alderden & Skogan, supra note 96, at 259.

1040 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

employee instead of a sworn officer saves over $50,000 per officer;101 the Chicago Inspector General estimated savings of between 16% and 41% per position through civilianization.102 Civilians are cheaper not only because their salaries tend to be lower than those of generalist officers but also because their fringe benefit rates are significantly lower and because civilians do not cost as much to insure.103 In particular, any change that involves personnel who are not armed saves money.

But we suggest that support of civilianization does not have to hinge on potential cost savings. There are also reasons to think that sending unarmed professionals to do many of the jobs now performed by armed generalists would pay great community-trust dividends. We detail the possibilities below.104

a. Paraprofessionals

One variant of civilianization is the creation of “parapolice” officers. Described as “a middle ground between the sworn officer and the civilian,” these individuals typically wear uniforms but are not authorized to carry weapons or use force.105 They have been deployed for tasks such as assisting victims of sexual assault, domestic violence, and child abuse; investigating crimes; and performing public outreach.106

Notably, paraprofessionals may be—but are not necessarily—specialists in the specific tasks they perform.107 Their range of tasks will certainly be narrower than that of today’s armed, generalist officers, who may be called to respond as readily to a hostage crisis as to a report of a homeless individual panhandling in front of a neighborhood store. Paraprofessionals will not be expected or allowed to use physical force or effectuate arrests.

In a report outlining proposed staffing changes, the Houston Police Department identified five questions that it felt must be addressed before a position could be “civilianized.” These are:

1. Does the position require law enforcement duties (i.e., powers of arrests, use of force, statutory requirements, carrying a firearm)?

101 HOUS. POLICE DEP’T, PROPOSED OPERATIONAL STAFFING ENHANCEMENTS FOR THE HOUSTON POLICE DEPARTMENT 84 tbl.44 (2014), http://www.houstontx.gov/police /department_reports/operational_staffing/Staffing-Recommendations-for-the-HPD-Final.pdf [https://perma.cc/8WSW-GFQZ].

102 CITY OF CHI. OFFICE OF THE INSPECTOR GEN., supra note 96, at 55. 103 See id. (“[T]he City would achieve savings due to the more generous fringe benefits

that sworn officers receive and the non-salary compensation provided to sworn officers and not civilians.”).

104 See infra Section III.B.2. 105 Forst, supra note 92, at 59. 106 Id. 107 Id.

2020] AN AGENDA FOR POLICING REFORM 1041

2. Are the skills, training, experience, or credibility of a sworn officer required to fulfill the duties of the position?

3. Would assigning a [sworn] officer to the position in question be helpful in developing their leadership skills?

4. Would assigning [a sworn] officer be helpful for other reasons (e.g. assignment would be helpful in developing the skills or knowledge of civilian staff)?

5. Can the requirements of the position by [sic] fulfilled by a specially trained civilian?108

“In analyzing a position for the potential of civilianization, questions 1-4 must be answered ‘no,’ and the last question must have a ‘yes’ answer.”109

We think that these conditions would allow civilians to address many social welfare tasks. Enforcement and routine patrol (walking a beat or driving around a neighborhood) take up significant amounts of police officer time and seem to us like tasks particularly well suited to be undertaken by civilian officers. Indeed, a study conducted in England examined the deterrent effect of having “Police Community Support Officers”—civilian members of the police staff who are uniformed but unarmed and who hold few arrest powers—rather than traditional constables conduct randomized patrols in crime “hot spots” in one city.110 They found that the “crime reduction effect of extra patrols in hot spots is not conditional on ‘hard’ police power. Even small differences in foot patrols showing the ‘soft power’ of unarmed paraprofessionals . . . were causally linked to both lower counts of crimes and a substantially lower crime harm index score.”111 Although we recognize that the cultural context is different, such results suggest that “soft power” patrol is a concept that merits exploration in the United States.

So what needs to happen in order for police departments to embrace this type of civilianization? We identify two criteria that typically must be present in an industry for paraprofessionalization to occur: First, there must be some way to certify and train paraprofessionals. Paraprofessionals must have a way of proving their credentials by demonstrating some kind of license or certification of training. Certification programs with standardized curricula help legitimize

108 HOUS. POLICE DEP’T, supra note 101, at 70. 109 Id. 110 Barak Ariel, Cristobal Weinborn & Lawrence W. Sherman, “Soft” Policing at Hot

Spots—Do Police Community Support Officers Work? A Randomized Controlled Trial, 12 J. EXPERIMENTAL CRIMINOLOGY 277, 277 (2016) (studying “whether crime-reduction effects of increased police patrols in hot spots are dependent on the ‘hard’ threat of immediate physical arrest, or whether ‘soft’ patrols by civilian (but uniformed) police staff with few arrest powers and no weapons can also reduce crime”).

111 Id. at 278.

1042 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

paraprofessionals to potential consumers.112 Second, the relevant existing professionals must not be strongly organized against the change.113 One story of paraprofessionalization—the success story of paramedics—helps to illustrate the importance of these criteria.

Widespread acceptance of paramedics has occurred because serious attention and effort has gone into the job of training and certifying these workers. 114 In the 1970s and 1980s, a national organization of paramedics began accrediting training programs and determining appropriate curricula.115 There was also a legislative effort to nationalize emergency-response standards. “The EMS Services Development Act of 1973 . . . authorized grants to develop a comprehensive EMS system throughout the country, for feasibility studies and planning, for the establishment and initial operation of EMS systems, and for the expansion and improvement of current systems.”116

Paraprofessionalization in medicine could thus occur because proponents of paramedicine were able to develop a training and certification regime for the role. Moreover, doctors and nurses were not organized in such a way as to provide significant resistance to the introduction of paramedics to their field.117 These existing professionals did not view the introduction of paraprofessionals as a threat to their jobs or their professional identities.118

It seems to us that policing, like medicine, is poised for a shift in its mode of service delivery that would involve the increased use of paraprofessionals. The challenge will be in helping police unions to see that remaking the police

112 For example, the moves from standard curriculum to degree residency program to national accreditation standards helped legitimize paramedics. Dennis Edgerly, Birth of EMS: The History of the Paramedic, J. EMERGENCY MED. SERVICES, Oct. 2013, at 43, 43 (tracing development of paramedic training from “first nationally recognized curriculum for EMS” in 1969 through first paramedic residency program in 1972 to contemporary accreditation requirements).

113 Indeed, the threat of “deprofessionalization” is likely to prompt unionization on the part of the ununionized professionals. See Joseph A. Raelin, Unionization and Deprofessionalization: Which Comes First?, 10 J. ORGANIZATIONAL BEHAV. 101, 106 (1989) (“[U]nionization emerges when an established profession has already begun to face serious deprofessionalizing encroachments . . . .”).

114 See generally Edgerly, supra note 112 (tracing origin of paramedic services to Johnson Administration’s study—”Accidental Death and Disability: The Neglected Disease of Modern Society”—which recommended establishment of best practices for emergency response services).

115 Id. 116 Manish N. Shah, The Formation of the Emergency Medical Services System, 96 AM. J.

PUB. HEALTH 414, 419 (2006). 117 Edgerly, supra note 112, at 43 (“Paramedics were taught by nurses and physicians who

were interested in emergency medicine and had visions of what it could look like in an out- of-hospital setting. Many, however, had never worked in the sometimes harsh prehospital setting or in the back of a moving ambulance.”).

118 Id.

2020] AN AGENDA FOR POLICING REFORM 1043

workforce does not have to represent an existential threat. Civilian workers deserve a fair living wage, good benefits, and other workplace protections no less than sworn officers. Ideally, police unions would be engaged in ensuring that these new positions are constructed with the well-being of the employees in mind. A key to making this change is to emphasize that these new employees are not replacing officers who are engaged in the type of policing activities that many officers feel are their true calling as professionals, as such tasks reasonably require the presence of someone who is trained and able to deploy force.

b. Civilian Experts

The use of paraprofessionals is not only economically efficient but also can be informationally efficient when those paraprofessionals possess specialized knowledge about the limited range of tasks they are hired to perform.119 For example, paraprofessionals might be trained to help victims of domestic violence or sexual assault navigate the legal process.120 If they are trained to address this type of trauma, they may prove better than their generalist counterparts at completing this specific set of tasks. By the same logic, we support the employment of civilian experts as well as paraprofessionals.

3. Implementing the Changes

We recognize that the path toward civilianization and specialization is likely to be bumpy, particularly given the historically insular culture of policing. Observers of police culture characterize it as involving “[a]n exaggerated sense of mission towards [the police] role”121 and a “strong sense of togetherness with colleagues,”122 as well as placing a high premium on masculinity.123 They note that police culture “displays remarkable continuity . . . and continues to exert considerable influence over the day-to-day police work.”124 Indeed, in a survey of 472 civilians from nineteen different policing agencies, Professors Megan Alderden and Wesley Skogan found that “feelings of acceptance by sworn members of the organization”125 played a key role in their work morale.

Still, such changes are not only possible but also necessary if police departments are to continue performing social welfare tasks. As we have endeavored to show, continuing to rely on armed, generalist officers to perform

119 See Forst, supra note 92, at 25 (listing paraprofessional civilian specialists as “crime scene and forensic lab technicians, information system and database specialists, lawyers, planning and research specialists, budgeting and finance specialists, administrators, and clerks”).

120 Id. at 59. 121 Bethan Loftus, Police Occupational Culture: Classic Themes, Altered Times, 20

POLICING & SOC’Y 1, 1 (2010). 122 Id. at 12. 123 Id. at 7. 124 Id. at 3. 125 Alderden & Skogan, supra note 96, at 278.

1044 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

such functions is the worst of the available options. It sends people who are highly and expensively trained to use force into settings in which other skills are sorely needed and in which the capability to deploy force may serve to aggravate rather than calm.

CONCLUSION

Like the police, legal scholars have a traditional way of thinking based upon their own training. That way of thinking views problems in policing as caused by poor legal standards for holding the police accountable.126 From this perspective, the problem of police accountability lies in the Graham standard.127 If this standard were changed, it is hoped and believed, the problem of officer shootings could be mitigated.

We argue that the factors that lead to officer shootings lie in the background.128 Officers are chosen, trained, and evaluated against the framework of being able to effectively deploy force. Unfortunately, this skill set is limited in several ways. In particular, many of the activities that police officers are involved in are social welfare tasks, and having the ability to effectively deploy force is irrelevant to being good at those tasks.

Consider an example. California recently passed a law encouraging officers to use de-escalation techniques and crisis-intervention methods before deploying lethal force.129 The law explicitly moves from a standard of reasonable use to one of necessity.130 This proposal accepts the current framing of policing: that every event will be responded to by an armed and sworn police officer. It assumes that the appropriate change is that those officers will have expanded training to have a set of skills beyond those based upon force. And

126 Alpert & Smith, supra note 7, at 486 (“The current standard [for judging the excessiveness of police force] has forced police departments to create policies on the use of force that are unworkable.”).

127 Id. (“Since Graham, however, the objectivity assessment for police use of force has become a ‘guided tour’ with a different guide for each tour (i.e., the expert witness). . . . Identifying what is ‘reasonable’ in a given encounter is a difficult challenge which must move beyond the limits of our present method of analysis.”).

128 See supra Part II (arguing that mismatch between officer training, which focuses primarily on situations requiring lethal force, and actual tasks of officers, which include more social welfare tasks, results in excessive police force).

129 CAL. PENAL CODE § 835(a) (West 2020); see also Michael Martin, Law Professor on California’s New Police Use-of-Force Law, NPR (Aug. 24, 2019, 5:13 PM), https://www.npr.org/2019/08/24/754052321/law-professor-on-california-s-new-police-use- of-force-law [https://perma.cc/9J9D-8JXE] (summarizing California standard for police force, which asks “whether the officer put themselves into an unnecessarily dangerous situation and then used force to address the danger that they should’ve avoided in the first place”).

130 Martin, supra note 129 (“Previously, police could use lethal force under reasonable circumstances. Under the new law, an officer can use lethal force only when necessary.”).

2020] AN AGENDA FOR POLICING REFORM 1045

officers will be required to use those alternative approaches before deploying force.

The argument that officers should be trained in and mandated to use alternatives to force-based strategies is an important first step in seeking to address police shootings. We believe that mandating change will have little impact in the absence of new training since the police would still be bringing the same skill set to situations governed by a different legal rule. But our suggestion is that this is only one of a variety of models that might be considered to address the problem of police shootings.

Our argument is that “accurately diagnosing the policing crisis is central to the practical project of reforming policing.”131 We think that an accurate diagnosis of today’s challenges would include the fundamental mismatch between police-department personnel and many of the roles they are asked to assume. And we do not think that anything short of a radical reconceptualization of what police do and who does the job will suffice to address the problem.

Of course, operationalizing police reform in the United States is a notoriously challenging endeavor. The United States is home to 18,000 separate law enforcement agencies, half of which have fewer than ten officers and nearly three-quarters of which have fewer than twenty-five officers.132 One of the few things they all share is a strong tradition of localized decision-making about criminal justice priorities.133 Aspiring reformers must, therefore, ask how their proposals will be implemented in such diverse contexts and without the capacity for centralized policymaking.

But the fact that police departments across the country are different from one another is not an excuse for their failure to change in light of evidence that current practices are not working. All of these police departments—big and small, urban and rural—primarily deploy armed generalists to do the job of street policing, a practice we are not convinced is justified in any context. Many commentators see the variation in policing policy and practice across the country as positive, noting that it allows police departments to “meet local priorities.”134 They therefore resist a common approach to policing reform across departments.135 But whatever its value, localism is a poor excuse for failing to

131 Bell, supra note 46, at 2066. 132 OFFICE OF CMTY. ORIENTED POLICING SERVS., supra note 89, at 29. 133 See, e.g., LAW ENF’T LEADERS TO REDUCE CRIME & INCARCERATION, FIGHTING CRIME

AND STRENGTHENING CRIMINAL JUSTICE: AN AGENDA FOR THE NEW ADMINISTRATION 1 (2017), http://lawenforcementleaders.org/wp-content/uploads/2017/02/LEL_Agenda_for_ a_New_Administration.pdf [https://perma.cc/UMR4-PBAN] (“[E]ach community is best- suited to set its own crime-fighting practices . . . .”).

134 Debo P. Adegbile, Policing Through an American Prism, 126 YALE L.J. 2222, 2226 (2017).

135 Id. (“Although there is no uniform approach, many police leaders have recognized that avoidable uses of force erode public trust and in turn make communities less, not more, safe.”).

1046 BOSTON UNIVERSITY LAW REVIEW [Vol. 100:1017

keep up with evidence-based best practices.136 “Citizens should be entitled to professional performance from U.S. police officers wherever they find them. Not only should the definition of professional performance be constantly evolving, but the public—itself mobile across the country—should expect police officers everywhere to keep up with these developments.”137

Communities are experts in their own sense of safety, and they are loudly communicating to police departments that the way they enforce the law today perversely makes many Americans feel less secure. In Detroit in 1967, Los Angeles in 1992, Ferguson in 2014, and so many other places, a substantial minority of Americans have been raising their voices again and again in protest against the way they are policed. To answer them appropriately, policing must not only be reformed but also reimagined.

136 Stone & Travis, supra note 71, at 30 (“This kind of coherence implies the development of national norms of how the police respond to situations, particularly to criminal activity, public disorder, political dissent or even a traffic infraction.”).

137 Id. Indeed, one might argue that in the post-Ferguson era of viral online videos, we have learned that we have a national police force whether we like it or not. Officers are judged by the actions and omissions of their colleagues from across the country. Police departments and police leaders thus have their own reasons to prefer a regime in which all of their colleagues across the country are bound by—or at least striving toward—a set of best practices for the profession.

Copyright of Boston University Law Review is the property of Boston University, School of Law and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.