The breadth of the criminal law and the unfettered discretion it creates are among the most significant challenges facing American criminal justice today. These twin problems have a particularly corrosive effect on policing, where they lay the foundations for many of the most prominent flashpoints for community anger, including intensive police surveillance, arbitrary enforcement, racial discrimination, and the cavalier use of pretextual authority. This paper considers the potential of the so-called “focused deterrence” strategy to counteract these dangers by channeling police discretion along more principled routes than it usually follows. Although the focused deterrence model is most commonly understood as a powerful strategy for combating crime, I argue that it should also be understood as a powerful strategy for advancing the rule of law. At its core, the focused deterrence model aims to make the exercise of discretion more deliberate and transparent, redirecting slack criminal justice authority towards the most significant community problems rather than squandering it on the haphazard enforcement of whichever violations of the law happen to come to official attention. By establishing deliberative forums for accomplishing that goal, well-executed focused deterrence initiatives help to combat the arbitrary and excessive use of overbroad legal authority. More substantively, the model’s commitment to deterrence provides an important source of focus and constraint. Properly understood, deterrence is not just a mechanism of crime prevention but a moral ideal rooted in the right to self-defense. If we take that ideal seriously, as many focused deterrence initiatives have, it rules out the opportunistic use of legal authority to prevent crime at all costs, including the cost of disregarding the rights and dignity of suspected offenders. In all of these respects, focused deterrence initiatives have the potential to help reassert the rule of law in an environment where it has become precarious. I demonstrate that this potential is not an accidental byproduct of an initiative designed for other purposes but a consequence of its intellectual genealogy—a legacy of its forgotten roots in an earlier era’s efforts to understand and tame police discretion.

Too much law amounts to no law at all.

William J. Stuntz1

  • 1. William J. Stuntz, Collapse of American Criminal Justice 3 (2011).


The reach of American criminal law is staggering. It is now possible to end up in jail for what seems like an unlimited range of conduct: for spitting in public, walking on the wrong side of the street, failing to return library books, and riding a bicycle without a bell, among many other things. There is a risk of jail, moreover, not only for doing the things that lawmakers want to put a stop to, but also for enabling them: not only burglarizing homes but also possessing burglars’ tools, not only using drugs but also hosting a party where drugs may be used, not only speeding but also using a radar detector, not only engaging in gang-related violence but also appearing in public with fellow gang members. In each case the risk of jail is over-determined, for what the criminal law covers, it covers repetitively: it bans carjacking as well as the robbery and kidnapping that comprise it; it bans mail fraud as well as the particular kinds of scams that might be conducted through the mail; and it bans robbery as well as its more easily proven analogues, such as theft while possessing a weapon and theft that risks harm to others.2

Most famously, the criminal law of recent years has authorized remarkably harsh punishments for all this misbehavior: five years in prison for two sugar packets worth of crack cocaine, a life sentence for theft if the thief has already been to prison twice, and fifteen years for possession of a single bullet by a career criminal. After conviction, probationers and parolees are subject to pervasive restrictions on their behavior: who they associate with, where they live, when they leave their homes, and whether and where they work.3

Whatever we might say about each of these provisions in isolation, in the aggregate they shape the character of American criminal justice in fundamental ways. Most importantly, they provide criminal justice officials—the police, prosecutors, and community corrections staff who cannot possibly enforce all of these provisions to the hilt—with enormous discretionary power to pick and choose among them. To take the most famous example: when the traffic code becomes so complex that few people can drive three blocks without violating it, police have a nearly unlimited reservoir of legal authority to pull motorists over when they want to check for fugitives and contraband (or, for that matter, when they want to pursue any other goal ancillary to the overt purpose of the traffic code).4 “[C]riminal law does not function as law,” William Stuntz famously wrote. “Rather, the law defines a menu of options for police officers and prosecutors to use as they see fit.”5

The arbitrary and intensive use of this massive reservoir of discretion contributes to many of the most significant concerns about American criminal justice today, including racial discrimination by police and prosecutors, the atmosphere of unrelenting surveillance on many urban streets, the cavalier use of pretextual authority, and overuse of the criminal sanction.6 The need to control the exercise of discretion stands out as one of the most important priorities for criminal justice reform today. More visible problems like excessive use of force, police killings, illegal arrests, and unconstitutional search and detention practices obviously deserve the central place they have occupied in the contemporary agenda for police reform. Alongside those problems, however—and in some respects interconnected with them—the reform agenda also needs to grapple more than it has with the problem of unfettered police discretion and the legal environment that produces it. It needs strategies for channeling police discretion in morally defensible ways.

This paper finds one such strategy in an unlikely place: in the widely celebrated “focused deterrence” model of policing, which originated in the efforts of the Boston Police Department and its collaborators to tackle the wave of youth violence that overtook the city in the 1990s.7 The heart of the focused deterrence model lies in efforts to identify people, places, and behaviors at which the full force of the overbroad criminal law should be directed—for example, a problem gang that should be targeted for strict enforcement of parole conditions and outstanding warrants, priority prosecution for federal crimes, and even intensive enforcement of quality of life rules like noise ordinances and jaywalking laws.8 Focused deterrence has most often been understood as a strategy for controlling crime rather than advancing justice, and some commentators have worried that it may actually undermine justice by intensifying proactive police surveillance in potentially troubling ways.9 More to the present point, the focused deterrence model seems to deliberately and unapologetically exploit the massive reservoir of discretion that the expansive reach of the criminal law has created. At a moment when the rule of law has deteriorated so badly, what can possibly recommend an approach to policing that thrives on the pretextual use of slack legal authority?

Although focused deterrence emerged out of a legal phenomenon that is and ought to be a source of concern, I will argue that its best examples represent an important strategy for addressing those concerns. The focused deterrence model has the potential to channel criminal justice discretion along more principled routes than it usually follows, making decisions about how to use discretionary authority more transparent, reflective, and restrained than they otherwise would be; in the process, it has the potential to combat the inscrutable, arbitrary, and excessive use of legal authority that over-criminalization threatens to produce. The criminologists who have mainly discussed the focused deterrence model have often neglected that potential; instead they have emphasized its possible impact on crime control. Even if that impact proves inconsistent or unknowable, the focused deterrence model would still be worth pursuing because it can potentially help bring an increasingly lawless domain of practice under the rule of law. At its best, it represents a judicious variety of proactive policing that reduces the risk that criminal justice authority will be used in arbitrary, inscrutable, and rights-violating ways. By making this argument, I hope to add my voice to those who believe this approach to policing still belongs on the contemporary police reform agenda, despite the fact that this agenda has evolved dramatically since the time when the Boston Police Department’s pioneering focused deterrence project first took shape.10

At a more general level, I hope to extend our understanding of the moral concerns that contemporary police reform can and should pursue. When the criminal law reaches as broadly as it does today, the bare fact that the law authorizes police intervention provides insufficient warrant for the decision to intervene; as several influential scholars have observed, there is more to good policing than mere lawfulness.11 But there is also more to good policing than merely procedural justice—the leading alternative to mere lawfulness that these and other scholars have offered as a guiding ideal for good policing.12 What exactly these two perspectives leave out is not easy to define, but I will suggest that one of the missing elements involves the choices that police make about how to use their discretionary authority to invoke the substantive criminal law. That topic was central to policing scholarship during the last major crisis of police legitimacy a half century ago, but it has not been a major part of the contemporary debates about police reform. By returning to that older body of scholarship and tracing its legacies in contemporary practice, I hope to revive and extend an important set of normative concerns.

Those concerns, I suggest, can be understood as aspects of our commitment to the rule of law—not the familiar, narrow version of the ideal that equates it with mere lawfulness or rule-following,13 but a more complex version that concedes that discretion cannot and should not be eliminated, yet remains committed to legal values like the reduction of arbitrariness, transparency, and respect for rights.14 At their best, focused deterrence projects illustrate how contemporary policing can realize this complex but important ideal. They represent a promising variation on a half-century old theme: the development of administrative strategies for channeling the massive discretion that front-line criminal justice practitioners must exercise—in this case through deliberation among an organized group of practitioners about how to handle a specific community problem rather than the broad administrative rulemaking advocated in the past. By making this case, I aim to reinterpret the focused deterrence model in terms of its significance for the rule of law.15

My argument unfolds in several stages. Sections II and III begin by tracing the historical roots of focused deterrence back to an earlier effort to understand and tame criminal justice discretion. First, Section II revisits the rich body of scholarship on the problem of criminal justice discretion that emerged in the 1960s and demonstrates the contribution it made to Herman Goldstein’s influential concept of “problem-oriented policing.” Next, Section III explains the connection between problem-oriented policing and focused deterrence and explicates the logic of the focused deterrence model in more detail. By tracing this two-stage historical evolution, I aim to show that the potential of the focused deterrence model to channel criminal justice discretion is not an accidental byproduct of an initiative designed for other purposes, but a legacy of its intellectual genealogy.

With this historical background in place, Sections IV and V turn to a close examination of contemporary focused deterrence practice (as outlined in prescriptive literature about the general model and evaluation reports about individual initiatives) in order to understand its potential to advance important ideals associated with the rule of law. Section IV describes how the logic of the focused deterrence model combats the excessive and arbitrary use of discretionary authority, though it also describes the failure of some individual projects to follow that logic consistently. Section V then argues that the commitment of these projects to deterrence (rather than incapacitation or any other mechanism of crime prevention) provides an important source of focus and constraint. Properly understood, deterrence is not just a mechanism of crime prevention but a moral ideal rooted in the right to self-defense.16 That ideal demands a commitment to rule of law values that many focused deterrence projects have actually exhibited in practice. Once again, some individual initiatives have not lived up to (or even recognized) that commitment; but many have, and by explicating it more clearly I hope to add to the prescriptive literature about the focused deterrence model.

Section VI concludes, summarizing the potential contribution that the focused deterrence approach can make to the police reform agenda today.

II. Discretionary Justice and the Roots of Problem-Oriented Policing

Today, the breadth of the criminal law and the troubling discretion it produces are associated especially with the important work of William Stuntz and those he has influenced, but this contemporary work continues a much older dialogue that dates back more than half a century. Although that older body of work never engaged with certain important aspects of the problems it considered,17 it did delve more deeply than most contemporary scholarship into administrative strategies for reasserting the rule of law. Those strategies survive, albeit sometimes in half-forgotten and weakened forms, in the problem-oriented policing model that eventually emerged from those inquiries—and through that general model in the Boston Gun Project and its offshoots.18

This body of work was set in motion in the mid-1950s by the American Bar Foundation’s (ABF ’s) criminal justice survey, the first major ethnographic study of criminal justice administration in the United States. By providing a close look at what front-line actors like police, prosecutors, and judges actually did in several cities, the ABF survey revealed how little the law on the books shaped their daily work. Discretion was deep and pervasive; none of these officials could be described as ministerial agents of the criminal law.19

This discretionary character of front-line work resulted from the nature of the criminal law. Legislatures passed broad gambling statutes that, read literally, covered the most innocuous church lottery or social card game, but they would never tolerate full enforcement. The profound vagueness of major legal categories like “disorderly conduct,” which accounted for one-eighth of all arrests reported to the FBI when the ABF studies were conducted,20 allowed those laws to serve as all-purpose repositories of authority on the streets. Misguided or archaic laws remained on the books long after they had become defunct, still available for opportunistic and undisciplined use.21 Some critics called for legislatures and city councils to narrow the scope of these laws and resolve their ambiguities,22 but the close look at front-line practice provided by the ABF surveys convinced most observers that a large dose of discretion was unavoidable. Some of the vague and overbroad laws that underwrote criminal justice discretion should certainly be clarified or repealed, but the problems that front-line officials had to contend with were too varied and unpredictable to fit entirely into unambiguous and narrow laws.23

A.    Guiding the Use of Discretion: From Rulemaking to Problem-Oriented Policing

What was troubling was not that discretion existed, but that it had been ignored. As lawmakers, administrators, judges, and academics glossed over the need to exercise discretion at the front lines of criminal justice practice, individual police officers and prosecutors were left to wing it—to rely on their own instincts or the informal norms of their occupational subcultures in order to fill in the massive gaps left by their overbroad legal instructions.24 That arrangement made the exercise of discretion in criminal justice idiosyncratic and inscrutable, and it posed a serious barrier to democratic accountability. The problem was particularly pronounced in policing, which attracted the lion’s share of commentary in subsequent years. A decade after his work as a staff researcher for the ABF survey, Herman Goldstein clearly formulated the problem:

Confronted each day by frequently recurring situations for which no guidance is provided, the individual officer either develops his own informal criteria for disposing of matters which come to his attention—a kind of pattern of improvisation—or employs informal criteria which have, over a period of years, developed within the agency of which he is a part. . . . [T]he challenge is to devise procedures which will result in police officers employing norms acceptable to society, rather than their personal norms, in their exercise of discretion.25

Legislators, judges, and criminal justice officials showed no ability or enthusiasm to provide the guidance needed. Lawmakers and judges lacked effective tools to shape important regions of front-line police work, and police managers squandered their attention on the minutiae of administrative procedures rather than the substance of front-line work.26

While the first problem—lawmakers’ lack of effective tools—seemed intractable, the second—police managers’ misplaced attention—seemed correctable in principle. A diverse chorus of criminal justice officials and scholars soon urged police agencies to develop administrative guidelines for the use of police authority: when to release a juvenile delinquent to his parents; when to order a crowd to disperse; when to use force; how to use confidential informants; and so on.27 These guidelines were often framed as a solution to the “selective enforcement” problem the ABF surveys had drawn attention to28 by providing a set of principles that police should rely on to decide when they should and should not enforce the criminal law.

Although the chorus itself has gradually faded into the background, Goldstein’s part in it contributed important themes to his enormously influential concept of problem-oriented policing.29 Years before the Crime and Delinquency article first formulated problem-oriented policing explicitly,30 Goldstein had already argued that police should develop guidelines for handling the most commonly recurring community problems they encountered, conceiving of those guidelines as devices for shaping the discretion police exercised in their daily work.31 Vague and overbroad statutes authorize far more legal intervention than the police ever could or should carry out, but it made more sense to narrow their reach problem-by-problem than statute-by-statute:

Because the law, in reality, is a means to an end, selective enforcement must be examined in the context of the various objectives of the police. Whatever guidance a police administrator provides his officers in deciding whether to arrest for disorderly conduct, for example, should be contained in broader policies on such problems as street gatherings, domestic disputes, and public inebriates, rather than in a specific policy or rule on the enforcement of the disorderly conduct statute.32

More than most other members of the “guidelines” chorus, Goldstein recognized that a decision not to invoke the law rarely meant a decision to do nothing at all. Useful guidelines needed to consider not only when the police should make an arrest but also when they should turn to non-legal strategies—to the mental health service providers, social workers, bar owners, detoxification centers, mediators, and regulatory bodies they sometimes relied on to help handle the community problems they encountered, as well as the informal tools they employed on their own initiative (such as ordering a drunk patron out of a bar or temporarily putting a gun in safekeeping). The importance of these non-legal tools had been another important finding of the ABF survey, and the failure to recognize it contributed to the overuse of the criminal sanction.33 In this respect, the guidelines Goldstein called for were not just manuals for selective enforcement; they also outlined the possible alternatives to legal authority in cases when law enforcement was inappropriate.

In short, Goldstein viewed guidelines as a method of systematizing police experience with commonly recurring community problems, refining and disseminating what the best practitioners already knew about the use of various legal and non-legal tools for handling those problems.34 Problem-oriented policing took shape as a framework for developing and using these problem-focused guidelines. Today, the Center for Problem-Oriented Policing’s “Problem Specific Guides” (covering problems like “Disorderly Youth in Public Places,” “Domestic Violence,” and “Chronic Public Inebriation”) are direct descendants of the “broader police policies” Goldstein had advocated for in the 1960s and 1970s to guide the use of police discretion (covering topics like “street gatherings, domestic disputes, and public inebriates”).35

B.    The Consequentialist Turn in Problem-Oriented Policing

As the early focus on shaping discretion evolved into the later focus on resolving community problems, Goldstein’s writing lost some of its emphasis on the rule of law concerns that originally inspired it. He increasingly described criminal justice interventions as “means to an end,” suggesting to many readers that they could be evaluated according to their success at achieving that end (for example, whether a crackdown on gambling really did reduce other forms of crime, whether letting the complainant decide to arrest the man who had abused her reduced domestic abuse, and whether patterns of traffic enforcement were well-tailored to minimize congestion and traffic accidents).

Over time, the burgeoning literature about problem-oriented policing has devoted almost all of its attention to crime prevention mechanisms and effects.36 The most sophisticated work in this genre has richly expanded the possibilities of policing practice, but it has applied a narrowly consequentialist evaluative lens to those innovations—i.e., it has almost exclusively asked whether problem-oriented policing initiatives have reduced crime.37 Lawrence Sherman, one of the most influential policing scholars of the past three decades, explicitly pressed Goldstein and his interpreters in this direction, urging them to place even more emphasis on rigorous social science evaluation of the long-term impact that police interventions had.38 More sympathetic commentators accepted a wider range of social scientific methodologies than Sherman did, but they implicitly stressed the same imperative to evaluate crime prevention impacts.39

This approach to the study of problem-oriented policing loses sight of the significant concerns about the rule of law that shaped its evolution. The early concerns about unfettered police discretion lay first and foremost in the abuse and idiosyncratic pressures it was subject to—the temptation to let convenience or a desire to log overtime determine whether to make an arrest, the inclination to defer to the most vocal and powerful members of the community and slight the interests of the marginal and voiceless, the possibility of retaliating against those who legitimately questioned police authority, and the scope for prejudice to drive police decision-making. It is true that by surrendering to those pressures the police might squander an opportunity to prevent crime more effectively.40 But the more immediate objection arose from the injustice of letting morally irrelevant or morally abhorrent factors determine who would be subjected to the burdens of police authority—or, more simply, from allowing the exercise of discretion to remain mysterious, and therefore unaccountable.41

Although problem-oriented policing and the guidelines movement that it evolved from undoubtedly aimed to overcome the “means over ends syndrome,” Goldstein’s paradigms of that syndrome were not front-line decisions that aimed to treat people fairly at the expense of long-run crime control. They were police reforms that obsessed over internal administrative rules: standards for uniforms and equipment, shift change times, procedural rules for filing reports, organizational structures, and so on. Police management had become too focused on these bureaucratic minutiae while neglecting the work front-line officers did in the community. The measure of that front-line work has at least as much to do with justice and the rule of law as with crime control.42

By revisiting the historical roots of problem-oriented policing, we can more easily recognize its submerged potential to help reassert the rule of law. Goldstein’s model emerged out of his efforts to grapple with the problem of police discretion—to make the use of discretionary authority less arbitrary, less intrusive, more reflective, and more transparent—and it still provides an important framework for accomplishing that goal. By encouraging front-line officers and police managers to reflect on the recurrent situations they encounter in their work, problem-oriented policing stimulates explicit, principled deliberation about how police should respond to those situations (and perhaps prevent them from arising in the first place), rather than leaving the response to unexamined, unaccountable habit. The necessary deliberation is not just about what will reduce crime, but about what is fair and consistent with other well-founded community values.

That is a different interpretation of the point of problem-oriented policing than the one that dominates contemporary scholarship, but recent discussion has not ignored it entirely. Michael Scott, in particular, has argued that contemporary accounts of problem-oriented policing have often put too much emphasis on crime prevention:

[C]laims about the police’s capacity to single-handedly reduce crime, disorder and fear at the community or higher level are simply not warranted. The greatest promise of problem-oriented policing may be that it is the approach most likely to maintain the delicate balance between freedom and order, and minimize the likelihood that police actions will undermine their legitimacy in society. This is so largely because the problem-oriented approach rejects the very excessive reliance on the enforcement of criminal law, and the use of force that accompanies it, that so often leads to abuse and consequent erosion of public trust in the police.43

Sixteen years later, these observations are even more compelling. The more thoroughly consequentialist interpretation of problem-oriented policing arose during an era of high-crime rates, when concerns about public safety reached crisis levels and pressure mounted for the criminal justice system to take aggressive action.44 It is perhaps not surprising that the finer points of the rule of law in the exercise of police discretion faded from view in that context. Now that criminal justice overreach has overtaken crime as a major policy concern, we might reasonably look more closely at the contemporary offshoots of problem-oriented policing for hints about its continuing potential to help resolve the problems that set its development in motion.

III. The Boston Gun Project

Nearly four decades after Goldstein first formulated the idea of problem-oriented policing, the Boston Gun Project45 stands out as its most distinguished application.46 The project and its descendants illustrate the potential for problem-oriented policing to channel criminal justice discretion in constructive and defensible ways, and by doing so to address some of the concerns about the rule of law that ultimately gave birth to it.

The Boston Gun Project took shape in 1995 during a surge in youth violence in the city. Boston police, Harvard researchers, community outreach workers, state and federal prosecutors, and community corrections officers spent months meeting and analyzing criminal justice records to understand more clearly what kind of incidents comprised the wave of youth violence. This Working Group undertook a detailed review of every known youth homicide in the city over the past four years, drawing on the street knowledge of police, outreach workers, probation officers, and the offenders they interviewed to understand how and why each incident had unfolded. Over the course of several months, the group gradually developed a picture of the youth violence problem. It was not the “huge, amorphous, and deeply rooted” problem portrayed in popular discussion about youth violence in the 1990s, according to project director David Kennedy, but a clearly defined and highly structured one: a few dozen active youth gangs entangled in a longstanding cycle of vendettas. “[F]or all practical purposes, this was Boston’s youth violence problem: a few score groups, already well known, made up of shockingly active offenders, many of them already well known, hurting each other along already very well-known vectors of group vendetta.”47 The intervention that the Working Group developed to tackle this problem came to be known as “Operation Ceasefire.”

The fact that known, high-rate, well-organized offenders were responsible for most of Boston’s youth violence was an indictment of existing criminal justice strategies, but it was also an opportunity: these offenders had parole conditions and outstanding warrants, they were suspects in cold cases that could be reopened, and they regularly violated a wide range of new laws—everything from trespassing and public drinking to driving without a license.48 Whenever law enforcement officials chose to do so, they could immediately bring coordinated attention to any given group in a devastating way, gathering together all the available discretionary authority and directing it towards a precisely defined target: violence committed by the small number of groups responsible for most of Boston’s serious crime. “It was narrow—don’t shoot,” Kennedy recalls.

The normal frame said, don’t be in gangs, don’t commit crimes, don’t sell drugs, don’t carry weapons, don’t violate your probation, don’t drink and drug. Turn your life around. Go back to school, get a job. Go forth and sin no more. This cut to the chase: Don’t hurt people. Say any more—don’t carry guns, don’t sell drugs, don’t recruit kids into your gang—you couldn’t back it up. There was just too much of it, and too little of us. Draw those lines in the sand, they would cross them, and it would be obvious that the cops had lied. They were used to being lied to, lying to them was normal: The next time you drop a hot urine I’m going to violate you, don’t let me see you back in my court, young man. Our side lied all the time. It was like we were training them to ignore us. This was focused, limited. We started to call it “focused deterrence”: focused on one problem, focused on the core offenders.49

Within the larger universe of violent groups, just a few stood out at any given moment. “Most groups didn’t kill anybody on any given day, month, year; those who were hot stood out very clearly.”50 At the beginning of Ceasefire, the active group was the Vamp Hill Kings. Having identified their target, officials served outstanding warrants, reopened cold cases, ordered drug tests, enforced curfews, raised bail requests, pursued federal cases, fast-tracked state cases, and even initiated deportation proceedings. The Kings eventually stopped fighting, and when another group started up, the process repeated.51 In this way, the Working Group followed outbreaks of violence from one group to the next, deciding which one would receive concentrated criminal justice attention at any point in time by continually assessing its best intelligence about which gangs were currently active and using that intelligence to target its discretionary authority.

As enforcement came down hard on one group after another, the Working Group explicitly communicated what it was doing to its targets. “We’re here because of the shooting,” criminal justice officials announced. “We’re not going to leave until it stops. And until it does, nobody is going to so much as jaywalk, nor make any money, nor have any fun.”52 Much of this communication took place in special forums convened for the purpose. Community corrections officers and other law enforcement required or invited specific offenders to attend open meetings with themselves, police, streetworkers, prosecutors, and community leaders. Law enforcement officials enumerated the powers at their disposal and what would lead them to use those powers.53 This communication was important because so much of the enforcement was pretextual. “The papers say this was a drug operation, and it was,” officials explained of one crackdown. “[B]ut it really happened because of the violence, and it’s violence that will draw anything similar in the future.”54

The Working Group brought this message to a wide audience. Although the enforcement crackdowns were only launched against groups that had broken the no-violence rule, the forums were convened for gangs that had not recently been active to tell them what response they could expect if they became active. In particular, once a violent group had been quieted, the Working Group tried to warn its rivals that they would be next unless they put down their guns. In that way, the message traveled across the city’s gang landscape along the existing network of vendettas.55

Widely celebrated, the Boston Gun Project inspired dozens of imitators over the next two decades. Many of them were sponsored by the federal Department of Justice, first under the research-heavy Strategic Approaches to Community Safety Initiative and later under the U.S. Attorney-led Project Safe Neighborhoods. I will refer to these projects collectively as “focused deterrence” initiatives. Some followed Boston’s lead closely, others elaborated the model in innovative new directions, others fell down in the fact of intractable implementation challenges, and some just seemed to miss the point. Underneath these variations, however, all of them were deliberately exploiting the discretion inherent in today’s criminal law.

The police, prosecutors, and community corrections officers involved in these initiatives did not create that discretion; it is a feature of the legal landscape they inhabit. The most productive question to pose about their work is not whether overcriminalization is desirable but, given its existence, how the discretion it generates can most legitimately be used. Because of the diverse forms the Gun Project’s descendants have taken over the past two decades, they illustrate both the promise and perils of the focused deterrence model as an administrative device for channeling police discretion, and thereby for advancing the rule of law in a legal environment where it has become precarious.

IV. Reasserting the Rule of Law

Like all well-crafted problem-oriented policing initiatives, the focused deterrence model channels criminal justice discretion by making decisions about how to use it more reflectively and transparently—by making those decisions the focus of explicit choices in collaborative working groups devoted to the resolution of significant community problems, rather than tacit choices by isolated individuals. That deliberative process and the practices that emerge from it have the potential to advance rule of law values in several ways. To clarify that potential, this section and the next one will consider how well focused deterrence initiatives have addressed the main concerns that recent critics of overcriminalization have emphasized. Those concerns capture some of the most important aspects of the rule of law that are threatened by the contemporary criminal justice regime.

The most basic problem with the broad reach of the criminal law is the way it facilitates excessive and arbitrary use of the criminal sanction. By definition, overbroad laws prohibit more behavior than their underlying rationale really requires.56 The most direct concern about such laws is that they may ensnare behavior that doesn’t really warrant punishment, or at least not such severe punishment: they produce injustice by punishing too many people, too harshly. The use of drug laws as a proxy for drug-related violence illustrates the problem.57 As Stuntz observes:

Drug crimes were not solely proxy crimes; a large fraction of incarcerated drug offenders . . . were suspected of drug crimes and nothing else. But the laws authorizing their punishment were designed with violent offenders in mind. So nonviolent drug offenders were, in effect, punished both for the crimes they committed and for the violence of the drug markets in which they participated. . . . The use of drug crime as a (partial) proxy for violence amounted to a sentencing enhancement, and a dramatic one at that, for black drug crime.58

In this way an overbroad legal tool designed to tackle a more specific problem (drug market violence, rather than drug markets themselves) spills over the boundaries that were supposed to contain it.

Overinclusive laws also contribute to arbitrary punishment—to differences in treatment that result from accidents of fate, caprice, and other morally irrelevant factors. “When the criminal justice system overcriminalizes,” Douglas Husak observes, “no discernable principle distinguishes those persons who are punished from those who are not.”59 For example, when Rudolph Giuliani served as a federal prosecutor in New York City, he randomly selected one day a week when all street-level drug dealers would be prosecuted in federal court, where they would likely receive much more severe sentences than they would have received in state court. Husak comments:

Despite the notorious difficulties implementing a principle of proportionality, no person would contend that the same criminal behavior becomes more serious and should be punished more harshly because it happens to be perpetrated on a Tuesday rather than on a Wednesday—especially when notice is deliberately withheld about the date on which the longer sentences will be imposed.60

The concern here is not that these randomized crackdowns are ineffective; they may be very effective.61 The concern is that such a strategy deliberately imposes far more severe punishments on some people than others for morally irrelevant reasons.62 The example provides a paradigmatic case of the arbitrary element in the enforcement of overbroad laws, illustrating a pervasive phenomenon in an especially vivid way.

A major virtue of the focused deterrence model lies in the way it combats these twin problems of excessive and arbitrary enforcement. The central thrust of the strategy is precisely to redirect discretionary criminal justice authority towards the most significant acts of violence (or other significant community problems) and away from the haphazard enforcement of whichever violations of the law happen to come to official attention. Early in the Boston Gun Project, community outreach workers complained about the heavy federal sanctions that prosecutors intended to pursue:

In one particularly tense exchange, an older Streetworker asked, “Is everybody who might sell drugs on the street facing these huge Federal penalties? My son, God forbid, might choose to do a little of that; is he going to be exposed?” “No,” [U. S. Attorney Don] Stern replied. “This is about violence. Only the key players in the most violent groups have to worry, and they’ll get fair warning, just like Intervale did.”63

That response reflects the outcome of the Working Group’s deliberations, which aimed to decide in a principled and explicit way how available discretion should be used. Each focused deterrence project begins with a detailed analysis of the nature of the crime problem in a particular community in order to identify the most significant threats to the community’s viability64: what those problems are and who exactly are responsible for them. Participants undertake close scrutiny of relevant incident reports, pool the existing street knowledge of field officers, and conduct new investigations to answer these questions.65 At its best, the result provides a more informed picture of who contributes to significant community problems than criminal justice officials normally consult. Cincinnati Police Chief Thomas Streicher drew this contrast as he reflected on his department’s own experience with focused deterrence: “Do you want to go out every day and arrest 50 crack heads with 50 crack pipes and fill up the court system? . . . Or do you want to bring in three tenths of 1% of the population that were responsible for the homicides?”66

Recent efforts to apply the focused deterrence model to overt drug markets illustrate the pattern most vividly. By systematically scrutinizing officials’ initial assumptions about who sustains those drug markets, they have repeatedly discovered that some of those assumptions are flawed. Kennedy summarizes the experience of law enforcement in High Point, North Carolina, where street drug markets had become a major concern in several neighborhoods, particularly the city’s West End neighborhood:

Marty Sumner and Larry Casterline started figuring out who the drug dealers in the West End were. The High Point police had been doing sweeps there, month after month, for years; Marty Sumner and the narcotics guys knew there were hundreds of dealers. But as they’d promised, they hunkered down and did their homework. Patrol and narcotics officers reported who they thought were active. The police department reached out to probation and parole officers, asked them the same question. They pulled incident reports for the West End violent and drug crimes, identified offenders, checked and cross-checked their associates in criminal history databases . . . Marty Sumner assigned each apparent dealer to a narcotics investigator. You’re going to be the expert on this person, he said. Figure it all out. Is he really dealing? How much? Inside or outside? How extensive is his record? Is he violent? . . . The answer was not what they expected. The answer was: Sixteen dealers . . . Larry Casterline [said,] “This exercise helped officers realize that they may have been directing enforcement action toward individuals who lived in and around the drug market but who were not actually involved in it.”67

The whole thrust of focused deterrence initiatives is to avoid the excessive and arbitrary use of criminal justice authority by focusing discretionary enforcement on those offenders that careful investigations have determined to be most responsible for significant community problems.68

Even within that targeted group, focused deterrence aims to minimize the use of enforcement authority. The early literature on criminal justice discretion and the idea of problem-oriented policing that emerged from it both stress the wide range of alternatives to arrest that police have available to them. In his most recent work, Kennedy stresses the massive burden that American law enforcement has imposed on black communities, and he embraces a moral and strategic imperative to combat crime in ways that minimize it.69 In High Point, police pursued that goal by “banking” cases against the majority of the sixteen active dealers they identified in the West End neighborhood. They conducted surveillance and undercover buys that would have justified an indictment, but then held these cases in abeyance, convening a “call-in” where authorities warned the dealers that they would be arrested immediately on the banked evidence unless they stopped dealing. The strategy aimed to achieve effective deterrence by transforming the very low risk of arrest for any given drug sale under conventional policing practices into a near certainty. By doing so, however, it also sought to minimize the use of enforcement authority by making the threat so credible that it would never need to be carried out.70

In the meantime, the decision to bank viable cases demonstrated restraint to the community. “[I]t was a graphic and concrete way to show the community, dealers, and their families that the views they had of law enforcement as conspiring to harm the community and control young black men is wrong,” Kennedy maintains.71 One NAACP leader in nearby Winston-Salem seemed to agree, telling the audience at an open community meeting: “I never would have believed that the police would hold our young men in their hands, able to put them in prison, and not do it.”72 A federal public defender involved in the High Point initiative added: “[I]f this helps turn things around, it’s going to be the start of treating people better. Of talking to people first, and arresting as a last resort.”73 To the extent that strategic use of discretion allows officials to forgo enforcement even where it would be justified, it not only avoids the application of overbroad laws to cases where their underlying rationale does not apply, it also restricts the use of legal authority even among cases to which that rationale does apply.74

In High Point and other drug market interventions, not all cases are banked. The most serious offenders are arrested and prosecuted immediately. The decisions about which cases warrant forbearance once again raise concerns about arbitrary and excessive enforcement. “What about everybody else you didn’t treat this way,” one community member wondered aloud at a Providence call-in.75 The question seemed especially apt in Providence. Officials had identified 104 dealers driving the targeted drug market, but seventy-one were arrested immediately simply because they were not from the neighborhood, and twenty-six more were arrested because they had a record of violence or guns. Only seven dealers remained to invite to the call-in—”The Lucky Seven,” as participants came to call them.76 That label perfectly captures the concern about arbitrary enforcement: that nothing but luck determines who will suffer the full force of the criminal law. Should a dealer’s neighborhood of residence determine whether he is invited to the call-in for a warning rather than arrested? Should any history of violence or illegal gun possession disqualify a dealer from the call-ins?

The large literature on the implementation of the focused deterrence model has recognized the significance of these questions. The U.S. Department of Justice “Drug Market Intervention Implementation Guide” calls for extensive deliberation about the criteria that will be used to decide which cases will be banked and which pursued immediately, observing: “Defining specific criteria will allow you to respond to questions about why one person was offered a second chance and another person was not.”77 The Project Safe Neighborhood “Chronic Violent Offender” strategy handbook similarly encourages focused deterrence working groups to articulate clear and explicit criteria to reduce the potential for bias, and it calls for a comprehensive effort to gather relevant evidence to determine which offenders or groups satisfy those criteria.78 This process of articulating explicit and defensible criteria and applying them systematically is what distinguishes focused deterrence initiatives from more casual exercise of criminal justice discretion. By providing a forum for doing that—a forum whose whole purpose is to deliberate about the best use of discretionary authority—focused deterrence combats the arbitrary and excessive use of the criminal law.

There are many ways in which this process can fall short of the ideal that defines it. Most simply, projects may fail to develop or use any clear criteria at all. In Los Angeles, Operation Ceasefire directed police and prosecutorial discretion toward two youth gangs throughout the project, and published reports seem to indicate that those targets were chosen largely based on community input at an early open meeting. A few days after that meeting, a particularly brazen shooting in the heart of the territory of one of the two gangs became a “triggering event” that mobilized enforcement action against the gang’s members.79 Shootings by different gangs the week after this triggering event received no additional attention from the Ceasefire working group.80 Instead, the two gangs originally selected as targets remained the focus of attention for the life of the project, based only on the views expressed by community members at the early project meeting, as well as the brazen shooting by one of the two gangs that happened to take place shortly thereafter. It is hard to distinguish that sequence of events from the crackdowns police have always mounted in response to community pressure. At their best, focused deterrence projects make deliberate and informed decisions about where discretionary authority should be directed rather than falling back on the potentially arbitrary forces of habit or unexamined community pressure.81

The mere existence of enforcement criteria is not the only necessary condition in this process. Explicit criteria do little to advance the rule of law if those criteria are inconsistent with community values or substantively indefensible. (Giuliani’s “Federal Day” relied on an explicit criterion, but one that was obviously morally irrelevant.) Reasonable people will often disagree about whether any particular criterion is appropriate, so controversy can never be eliminated from these judgments.82 In these cases transparency and democratic accountability are particularly important, since the appropriate substantive criteria are contested.

In many projects, however, it seems clear that administrative convenience and data availability exert too much influence on case selection. To identify the most serious offenders in a neighborhood, some projects have relied entirely on easily available criminal history files, assuming that the worst offenders are the ones with the longest rap sheets. More sophisticated projects have recognized that these data may provide an incomplete perspective, missing recent offenses that still linger in the court system and giving too much weight to old offenses, and failing to account for the varying seriousness of each offense.83 Others rely uncritically on criminal justice labels, for example, by treating all “gun” crimes the same without distinguishing between possession and use.84 More complex algorithms for identifying serious offenders calculate weighted point totals, assigning a seriousness score to each prior offense and summing all offense scores for each offender (or occasionally applying a more complex weighting system). These approaches strive to avoid bias and discrimination,85 but they may introduce new problems. Formulaic “weights” may understate the challenge of commensurating different crimes;86 organizational records may provide a distorted picture of the reality they purport to describe;87 and point systems may serve to justify the police practices that created the data they rely on in the first place.88 These problems indicate that the aura of objectivity surrounding data-driven decision-making may be exaggerated, and it is surely correct to insist on subjecting its conclusions to a reality check.89 The need for judgment cannot be eliminated; professional and community input can be disciplined but not ignored.

Other projects have used the availability of federal charges in deciding whom to arrest and prosecute in order to maximize “impact.”90 Since federal laws are especially notorious for overbreadth, that criterion is probably perverse. For years the Boston Gun Project was presented as a more nuanced alternative to Project Exile, which effectively pursued full enforcement of certain federal gun laws in the jurisdictions that adopted it.91 The zero-tolerance spirit animating that initiative is diametrically opposed to the discretionary spirit of the Gun Project. As I have stressed throughout, the fundamental premise of focused deterrence as an application of problem-oriented policing is that vague and overbroad sources of legal authority cannot and should not be enforced to the letter; they demand situated judgments about whether the motivation behind them will be well-served in a particular case. The main virtue of the focused deterrence model—the forum it provides to make those situated judgments in an informed way—is undermined when a working group adopts any such rigid rule.

Because these judgments about how discretion should be channeled can be momentous—including a near-certainty of a decades-long prison term—they demand extraordinary care. The information they rely on will not typically be vetted by a court according to traditional rules of evidence and due process (though of course the cases they eventually lead to will be). Informal talk about the basis of these quasi-judicial judgments can be jarring. “We almost always know who did it,” Kennedy says of the triggering incidents that drove decision-making in Boston. “[W]e may not be able to prove it, to take it to a prosecutor, but we know.”92 The source of such knowledge has included jailhouse informants, streetworkers, community members, and vice officers.93 As many participants and commentators have recognized, there is a moral imperative to scrutinize the quality and relevance of this information, even if the appropriate standards remain unclear.94 The rules of evidence for formal adjudication are extensive; those for the exercise of discretionary judgment are almost nonexistent. Because they make the basis of those judgments explicit, focused deterrence initiatives have the potential to catalyze reflection on this crucial domain of practice, and thereby make the exercise of discretion less arbitrary.

V. The Morality of Deterrence

Critics of overcriminalization have also raised a deeper and more substantive concern than those I have considered so far, and by examining it briefly I hope to clarify another valuable feature of the best focused deterrence projects. The concern I have in mind is best (and most apocalyptically) expressed by Markus Dubber. For Dubber, the staggering reach of modern American criminal law creates the conditions of a police state, expanding discretion so broadly that criminal justice officials have nearly unconstrained authority.95 In that environment, sovereign power more closely resembles the power of the patriarch over his servile dependents than the power of a legitimate government over free and autonomous subjects.96 At best, in today’s America, this potentially boundless power is subject to norms of efficiency: it doles out punishment in a way that will maximize crime prevention—particularly through selective incapacitation of the dangerous, but potentially through opportunistic deterrence and rehabilitation as well. “Offenses simply lay the foundation for an assessment of dangerousness,” Dubber writes. “In their very malleability lies their value. It’s this malleability that makes room for the discretionary dangerousness assessments at the heart of the system.”97 In this way overcriminalization promotes the advance of utilitarian ideals and undermines justice. It is fundamentally antagonistic to respect for rights and dignity, viewed as substantive limits on the state’s authority over its subjects.98 Powerful sovereigns may decide to exercise mercy from time to time, but the optional exercise of mercy is no substitute for the systematic recognition of individual rights. This danger arises even for impeccably democratic sovereigns, since a democratic majority can ride roughshod over individual rights almost as easily as an autocrat can.

Some may be unmoved by this concern. Yes, they will say, that is exactly what we hope to accomplish: case-by-case assessment of dangerousness in order to prevent crime as effectively as possible. (The popularity of this view explains why democratic accountability alone cannot address this concern.) It is not hard to find this utilitarian perspective in the literature about focused deterrence, which is often unapologetic in its desire to harness any available authority to eliminate significant community problems. Although it is sometimes qualified in practice, the idea that the end justifies the means often lurks beneath the surface of projects launched to combat major crises like rampant youth homicide. Regardless, these initiatives have been deliberately engineered to maximize deterrence, and for that reason it may appear that they cannot avoid treating people like “tigers . . . in a circus,” as Andrew von Hirsch memorably put it—“as beings that have to be restrained, intimidated, or conditioned into compliance” —rather than as real moral agents.99 Even the appeals to conscience commonly included in focused deterrence call-ins (typically conveyed by those who belong to the offenders’ families and communities) are a deliberate part of a strategy to maximize the combined force of formal and informal sanctions.100 The recent interest in procedural justice among focused deterrence projects is also often justified and motivated, and therefore bounded, by a desire to maximize compliance and respect for the law; it, too, often takes shape as a pragmatic accommodation to maximize crime prevention rather than an independent moral ideal rooted in respect for persons.101

All of that is to say that focused deterrence hardly seems to galvanize individual rights against the corrosive bath of overcriminalization. And yet, in practice some of the most prominent focused deterrence projects have refused to exploit all available discretion in the name of effective crime control, and particularly to employ prototypically utilitarian strategies (such as selective incapacitation of the dangerous) that other moral frameworks condemn. For example, in April 2007, several Cincinnati politicians used criminal history databases to assemble a list of “likely killers” and called on the city’s emerging focused deterrence initiative (the Cincinnati Initiative to Reduce Violence, or CIRV) to direct concentrated enforcement at the members of the list.102 Civil rights advocates immediately criticized the proposal,103 and the CIRV working group agreed. Kennedy’s response is hard to parse: “It’s not our list, criminologists have spent generations trying to do this kind of prediction, it doesn’t work, we have nothing to do with this, we said, over and over.”104 The rationale expressed here seems narrow: the best predictions criminologists have been able to make simply are not accurate; assessments of dangerousness do not work. When Kennedy stresses that “it’s not our list,” he seems to criticize the methods used to generate it—perhaps because it was based entirely on criminal history files, without access to the situated knowledge of professionals and on-the-ground community members in the working group.105 It is tempting to conclude that CIRV members had no principled reason to reject the likely killers list, just reasons to want to fine-tune it. Perhaps it is nothing but the “huge outcry”106 that led CIRV to reject this particular attempt to use malleable laws to assess and incapacitate the dangerous; perhaps the concern for fairness arose from pragmatism rather than principle. Perhaps it simply reflects the humane sensibilities of the people who happened to be involved in this particular project, rather than anything essential in the focused deterrence model.

Perhaps. But I want to consider the possibility that the logic of focused deterrence requires a commitment to individual rights that is inconsistent with the purely utilitarian tendencies Dubber identifies, as well as the broader corrosion of limits on state authority that makes those tendencies possible. At their best, focused deterrence projects are not just efforts to incapacitate especially dangerous individuals (though they can and have degenerated into that on many occasions). To the extent that they truly aim to generate deterrence, they can and should be viewed as efforts to establish a particular kind of societal self-defense that is entirely consistent with respect for individual rights—and indeed that demands it.107

Three decades ago, Warren Quinn maintained that “a practice of punishment is at its moral core a practice of self-protective threats,” and he argued that these threats are perfectly consistent with a deontological, rights-respecting morality.108 Quinn imagined a fantastically sophisticated device that automatically inflicts punishment on anyone who commits a crime with at least as much accuracy as our current system of criminal justice. It would be a legitimate act of self-defense to activate that device as a deterrent—we can use force to defend our rights, so surely we can activate a device that threatens to use force to do so. (With an important caveat: since the kind of self-defense one can employ against an attack on one’s rights is limited, the kind of punishment the device can legitimately inflict is limited in the same way; this is the source of the proportionality constraint on deterrent punishments.) Quinn argued that the right to make and carry out deterrent threats is on the same moral footing as the right to activate such an imaginary device. Neither one ignores the rights of the people it threatens; both are tailored to the complex structure of those rights.109

Focused deterrence projects should be viewed as this kind of self-protective threat. At their core, they deliver a specific threat about the sanctions that will follow unless the targeted offenders cease offending, and they aim to create an organizational structure that will quickly and automatically make good on that threat if they do not. Quinn’s argument is that this is precisely the form that morally defensible crime prevention should take. Unlike selective incapacitation and other utilitarian strategies, focused deterrence does not require (as Quinn puts it) “any neglect of rights for the sake of effects.”110 This interpretation makes sense of CIRV’s rejection of the “likely killers list,” as well as the limited and controversial role of many other purely predictive practices in the history of the focused deterrence model. It also makes sense of the common and growing concern for the rights and dignity of would-be offenders that many focused deterrence projects have exhibited, despite their unapologetic embrace of the need for swift and certain deterrent punishment.

In these respects, the focused deterrence model can and should reflect a commitment to substantive moral values other than crime control. Those values provide principled guidance for the legitimate use of discretionary authority.

A.    Deterrence and Fair Notice

Consider, for example, how the logic of deterrence has led many of these projects to make enforcement practices more transparent to the suspected offenders they target. Lack of transparency in this respect is a major problem associated with overcriminalization. Because it is typically impossible to enforce overbroad laws to the letter, officials must devote their limited attention to a subset of potential cases, and potential lawbreakers rarely know what they must do to stay out of trouble. In this respect, overbreadth is a species of vagueness, and like all forms of vagueness it undercuts the aspect of rule of law ideals that is sometimes called “fair notice.” As Stuntz has observed, however, this concern about the unfairness of vague laws is simultaneously a concern about deterrence:

Broad criminal codes ensure inconsistency. Broad codes cannot be enforced as written; thus, the definition of the law-on-the-street necessarily differs, and may differ a lot, from the law-on-the-books. . . . [A] signal that cannot be seen is a very poor signal. Law-on-the-street, the sum of millions of arrest and prosecution decisions by thousands of police officers and prosecutors, seems designed to minimize visibility.111

In his final book, Stuntz returned to the muddied message that overbroad laws convey, and he illustrated the problem with an example from Boston. In 2007, when local officials came to believe that a gang called the Lucerne Street Doggz had carried out dozens of shootings and six murders in just two years, federal prosecutors aimed to incapacitate the group by making maximum use of laws against buying and selling drugs and guns.112 Stuntz complained that the message conveyed by such prosecutions is indecipherable.

Recall the Lucerne Street bust. The gun and drug charges filed in that case may track gang members’ history of criminal violence, but only in the aggregate. The gang was targeted because of the many acts of violence its members committed, and (at least in part) because of the many acts of violence members of other Boston gangs committed. But no gang member knows which particular acts of violence prompt such targeting. From the point of view of any individual offender, the odds that any particular shooting will lead to a later drug trafficking prosecution must be very low.113

As criticisms of the drug laws themselves and the way they are ordinarily used, this critique rings true, but this particular example is poorly chosen. Although Stuntz does not discuss the connection, the Lucerne Street Doggz case was an explicit offshoot of the Boston Gun Project, which had fallen apart in 2000 but had been revived six years later under newly-appointed Police Commissioner Edward Davis. Properly conducted, the focused deterrence strategy explicitly addresses the concern that pretextual enforcement sends a muddied signal. A central point of the “call-ins” and related communications strategies is to clarify the rationale for pretextual enforcement. Stuntz worried that “no gang member knows which particular acts of violence prompt such targeting,”114 but the Boston Gun Project explained that connection explicitly and repeatedly.

For Kennedy, like Stuntz himself, that strategy flows from a concern for effective deterrence—from a recognition that sanctions only deter if people know of them and believe them115—but it also relates to the concern about fairness. Explicit statements about how discretion will be used provide the fair notice component of due process ideals. Kennedy recognizes as much in several places: “because of the explicit communication with gang members, the approach seemed to be fundamentally fair. ‘Here’s how the game’s going to be played,’ gang members were told.”116 The same idea underlies U.S. Attorney Don Stern’s promise, quoted earlier, that targets of concentrated enforcement would “get fair warning.”117 Community members reinforced this idea. Tracy Lithcutt, head of the streetworkers, told gang members: “If you don’t hear what’s being said to you today . . . it’s on your heads.”118 The idea has deep roots in the focused deterrence model. The “Operation Scrap Iron” initiative that the original Boston Gun Project grew out of began with a campaign to put targeted gang members on notice, as youth workers, probation officers, and police took pains to convey why they were there and what would happen if violence continued.119 From the beginning, a central thrust of most of these projects has been to articulate how discretion will be used explicitly and publicly—not just within focused deterrence working group but to the wider world as well. These efforts to communicate how discretion will be used aim to maximize deterrence, but in the process they advance the rule of law as well. Not every deterrence strategy accomplishes both goals simultaneously,120 but these examples illustrate how it is possible to do so.

B.    Morality’s Constraints

While respect for individual rights and effective crime prevention sometimes coincide, at other times they may point in different directions. In those cases, the morality of deterrence rules out some of the crime prevention strategies that some focused deterrence initiatives have been tempted to employ. In particular, to show proper respect for individual rights, deterrent threats and the punishments that carry them out must be motivated by and proportioned to their target’s own potentially wrongful behavior.121 We do not treat a person as a mere means to an end when we place him under a threat to compel him to respect our rights, but we do use him—and thereby violate his rights—if we punish him in order to deter others.122 The justification for deterrent threats (and for the punishments that carry out those threats, if it comes to that) can never be vicarious.

That constraint raises significant questions for the focused deterrence model, which has used various forms of collective responsibility as a core crime prevention tactic. Some of those tactics may be legitimate, but they require more careful moral scrutiny than they have often received. One of the most innovative post-Boston focused deterrence projects was launched in Lowell, Massachusetts, in the fall of 2002. Cambodian and Laotian gangs were a major source of youth violence in the city, but they proved resistant to the techniques that had been developed elsewhere. Unable to exert much direct influence over the young gang members themselves, the project searched for other sources of leverage over their behavior. Eventually officials hit on the strategy of strictly enforcing anti-gaming laws against older relatives, who often ran illegal gambling businesses. Until the youth violence stopped, gambling would receive intensive enforcement. As one overview of the project summarized it: “Lowell officials applied an additional lever to adult offenders involved in illegal gaming activities with the goal of having the adults exert influence over youths involved in street violence.”123

Described in that way, the Lowell project seems ethically troubling. The severe threat of intensive enforcement directed at adult gambling was not justified by the serious character of gambling itself but by the seriousness of youth violence. The “pulling levers” strategy implies that officials will mine their discretionary authority to impose the harshest possible sanctions, but gambling itself hardly seems to warrant such a draconian response. In that respect the project seems to use the adult gamblers as a mere means to an end.124

More detailed descriptions of the project provide a different perspective. As the national Project Safe Neighborhoods evaluation describes it, the Working Group’s understanding of Lowell’s East Asian gang problem evolved as officials developed better intelligence sources about these uncommonly secretive organizations. It became clear over time that the neighborhoods’ youth gangs were closely connected with the gambling operations. Gang members appeared suspiciously often in neighborhood gambling dens and casinos, and some of them seemed to work as runners for the illegal businesses. Eventually it became clear that at least some of the gangs were actually part of the gambling businesses, which regularly called on them to collect gambling debts and otherwise protect their business interests. The “gambling elders” were not simply legally vulnerable community members who happened to have leverage over young gang members. They supported and even encouraged the gang members’ activities directly, and in that respect they shared at least some moral responsibility for the violence itself. That responsibility makes the crackdown on their activity easier to justify. Moreover, the crackdown itself was more measured than the “pulling levers” metaphor seems to imply. In practice, it might amount only to parking a police car outside an illegal casino, coupled with a warning that “when violence erupts, no one makes money.”125

Without this background, Lowell’s strategy seems problematic. It would be illegitimate to prosecute a no-holds-barred crackdown on petty gambling by a gang member’s family to try to force the family to pressure the gang member to change his ways: that would be a clear case of the kind of vicarious punishment that any plausible morality condemns. It is worrisome that the focused deterrence literature has often recounted this oversimplified version of Lowell’s experience as if it were an accurate description of events in that city—and, under that description, as something worth emulating. That version of the Lowell story can usefully serve as a hypothetical case in which stern deterrence might be effective but would nevertheless be unjustified.

The actual chain of events in Lowell, by contrast, represents a sophisticated effort to grapple with the complex issue of group responsibility. That issue has a central place in many focused deterrence initiatives, which often target the group structure of criminal offending.126 Lowell’s experience provides an especially exotic example of the form that structure can take; the internal dynamics of gangs themselves (and group-based offending more generally) provide a more straightforward and pervasive example. As Kennedy puts it in a discussion of Operation Ceasefire’s prototype:

It was deterrence aimed at the group, the gang. Criminal justice is almost entirely about individuals. A drug crew can get some besotted juvenile to kill someone; when he does, he, not the drug crew, gets arrested. Short of exotic RICO and conspiracy prosecutions, which are in practice exceedingly rare, groups get no formal legal attention at all. But gangs were driving things.127

The point here is not just a Machiavellian tactical claim that interventions against people who happen to have leverage over criminals of interest may be effective. Kennedy seems to take pains to register a moral concern about the way so many of the people who drive community violence escape responsibility, as in the case of the drug crew members who “get some besotted juvenile to kill someone” with impunity.128 Formal legal strategies for enforcing group responsibility, such as RICO statutes, have so far proven unsuitable (particularly when the group’s responsibility for violence is less direct than Kennedy’s example of the “besotted” youth implies). Perhaps the discretion granted to law enforcement officials can be legitimately used to fill this void, as long as those who wield it make an earnest attempt to sort out meaningful distinctions in moral responsibility. Those distinctions are often very difficult, but the complexity of the hardest cases need not paralyze judgment in simpler ones. There is a clear distinction, for example, between targeting an uncle simply because he can exert influence on his nephew and targeting a fellow gang member who did not pull the trigger but might, and who approves of it. That form of group-based responsibility—for example, warning an entire group that “gun use by any one member of the gang would result in legal problems for all members,”129 and telling them if “[y]ou see one of your boys picking up a gun,” they have a responsibility to “tell him to put it down”130—may sometimes be justified.131 It is, in any case, a far less sweeping and more refined strategy than the blunderbuss assessments of dangerousness that discretionary justice ordinarily contributes to.

VI. Conclusion

The breadth of the criminal law and the massive discretion it creates are among the most significant challenges facing American criminal justice today. They enable arbitrary and excessive punishments, they undermine the system’s transparency and accountability, and they chip away at the protections that individual rights ought to provide. They threaten to fundamentally undermine the rule of law.

And yet it is hard to see how we can put the genie back in the bottle. Many contemporary critics of overcriminalization aim to develop principles of restraint that can be used to identify legislative overreach, but all of them recognize the significant obstacles to any effort to define and enforce principled limits to the scope of the criminal law.132 Historical examples of successful legal efforts to narrow the scope of the criminal law are few and far between, and the examples that exist do not seem to offer much hope.

The focused deterrence projects I have described in this paper, along with the broader idea of problem-oriented policing they apply, represent a different response to the problem of overcriminalization—a response based on administrative strategies for improving the way discretion is exercised rather than a legislative effort to eliminate it. Ultimately they aim to influence the ethical commitments of enforcement officials rather than the content of the law itself. That characteristic makes it possible to narrow the law’s application in a more tailored, contextually informed way than is possible through legislative reform.133

Even at its best, the focused deterrence model obviously does not “solve” the problem of overcriminalization. Discretion still exists, and prosecutors, police, and other enforcement officials can still abuse it. As I have repeatedly illustrated, they have sometimes done so in the name of the focused deterrence model itself. The hope is simply that the proliferation of focused deterrence projects will redirect more decisions about how enforcement discretion should be used down the structured, deliberative channels those projects establish. Every case brought through the local Operation Ceasefire is potentially a case that will not be brought in a more cavalier way. U.S. Attorneys’ capacity to bring federal gun charges is limited, and in jurisdictions where much of that capacity is devoted to cases nominated through the deliberations of a focused deterrence working group, less remains to be harnessed in more haphazard ways. The High Point police department’s narcotics unit is finite, and capacity directed to the West End through the drug market initiative cannot be squandered elsewhere.

The existence of the focused deterrence alternative can also ground the critique of less principled uses of discretion. The targeted use of discretionary authority to pursue identifiable serious offenders—determined to be so based on deliberation and investigation—puts the more cavalier use of that authority that prevails in so many cities in stark relief. We can more clearly articulate what is wrong with blunderbuss stop-and-frisk initiatives and intensive drug market crackdowns by pointing to the focused deterrence model as a viable alternative. (That was exactly what happened in Cincinnati, where Streicher and many of his staff gradually came to reject many of the department’s old ways of using discretion as the focused deterrence model took hold).134

In some of these respects, the focused deterrence model represents a return to the “rulemaking” model that emerged from the ABF-inspired critique of overcriminalization in the 1960s. Both models rely on administrative practices to channel criminal justice discretion out of a belief that legal tools can never do the job alone.135 The earlier debate about police rulemaking, however, clarified many of the limitations of the strategy. Ronald Allen, one of rulemaking’s most forceful critics, argued its limitations at length, and we do not need to agree with all of his arguments to share his conclusion that police rulemaking is unlikely to resolve all of the concerns about the rule of law that inspired it.136 Focused deterrence projects address those concerns in a way that avoids many of rulemaking’s limitations. Their relevance to those concerns and their potential to overcome those limitations should not be surprising, since the focused deterrence model descends from the problem-oriented policing strategy into which rulemaking evolved.

Most important, the case-by-case deliberation that the focused deterrence model emphasizes has the potential to avoid the rigidity of rulemaking. Despite the best intentions, the guidelines that some police agencies developed in the 1970s often seemed to become just another form of law, and the effort to use rulemaking to channel discretion faced the same obstacles as legal reform itself. 137 Situated deliberation about how discretion should be used is different. There may be no exact rule for distinguishing “slight,” “petty,” and “trifling” violations of the law from those that deserve serious attention, but discretionary judgment can make those distinctions easily.138 The focused deterrence model encourages the people who make those discretionary judgments to do so more consistently by making them more explicit, transparent, and informed. The grounded analyses it calls for make it possible to use legal discretion in a defensible, tailored way, enacting complex values that general legal or administrative codes can never adequately capture.

More subtly, but relatedly, the focused deterrence model may avoid another important barrier to rulemaking: the fact that criminal justice officials are reluctant to embrace a self-imposed effort to restrict their authority more narrowly than the legislature requires.139 Focused deterrence initiatives influence police decision-making in a way that may not feel quite so limiting. They are not ex ante restraints on police authority that potentially disarm officers before it is clear what is really at stake. They take shape as an effort to develop principled responses to specific community problems through careful deliberation about the best course of action with a range of affected interests—police, prosecutors, parole and probation agents, community members, service providers, and sometimes even (though arguably not as often as they should) the defense bar. Abstract debates about “tying the hands of the police” give way to specific questions about how particular people can legitimately be treated, and the answers must be justified deliberatively to a working group rather than made unilaterally by an isolated officer.

In these ways, focused deterrence has the potential to uphold the demands of legality in a world where full enforcement has become chimerical. For a critic like Dubber, the massive overbreadth of modern criminal law has undermined legality altogether; law is replaced with police—the unconstrained will of the sovereign state. Those alternatives do not exhaust the field. There is legality short of mechanical enforcement;140 indeed mechanical enforcement itself is another kind of tyranny.141 True justice lies not in reviving a simplified ideal of the rule of law but in developing an ethic of discretionary justice. That is precisely what the best focused deterrence projects strive to do.

Focused deterrence has been celebrated as a dramatically effective strategy for combating violence and other crimes that tear communities apart. I have tried to argue that it also has the potential to become a dramatically effective strategy for restoring the rule of law. Just as communities plagued by violence can turn to the focused deterrence model for guidance, communities plagued by an arbitrary and excessive system of criminal justice can do the same.142 Some of the guidance it provides is reasonably clear: establish a working group with a mandate to identify significant community problems, understand who and what is responsible for those problems, and decide systematically how to use the broad authority that is available to intervene. The deliberation that takes place in these forums brings the exercise of criminal justice discretion out of the shadows, subjecting it to explicit and collaborative scrutiny. That process has great potential to make the exercise of authority more consistent with rule of law values.

Beyond that basic advice, however, many accounts of focused deterrence provide too little guidance about how working group deliberations can best realize that potential. In some cases, the advice they provide may even be counterproductive. The task is undoubtedly difficult, and the necessary guidance may always remain complex and indeterminate, but in this paper I have tried to begin to articulate the considerations that need attention and to identify the kinds of practices that merit further scrutiny. Working group deliberations cannot focus exclusively on the instrumental question of which strategies will most effectively reduce crime; they also need to consider whether those strategies use discretionary authority consistently, transparently, and in a manner that respects the rights and responsibilities of the people involved. Focused deterrence initiatives may need to reject practices that some participants consider to be effective tools for crime prevention because those practices undermine the rule of law. Judgments and distinctions like these have received too little attention in a literature dominated by concerns about crime control. Those concerns are understandable, but the neglect of rule of law values makes focused deterrence vulnerable to critics who view it as unethical.143 Future work should aim to provide as much guidance about how the focused deterrence model can advance the rule of law as it does about how it can maximize crime prevention.

The most promising aspect of the focused deterrence model lies in its ability to provide both kinds of guidance simultaneously. At its best, focused deterrence embodies a morally sensitive crime prevention strategy rooted in the right to self-defense, worked out over decades of reflection on the problem of criminal justice discretion.

  • 2. I take the examples of the breadth of the criminal law from Douglas Husak, Overcriminalization (2008); Stuntz, supra note 1, at 81; Erik Luna, The Overcriminalization Phenomenon, 54 Am. U. L. Rev. 703 (2005); William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505 (2001) [hereinafter Pathological Politics]; William J. Stuntz, Substance, Process, and the Civil-Criminal Line, 7 J. Contem. Legal Issues 1 (1996); Lindsey Reiser, Police Suspend Practice of ‘Pretextual Stops’ in Phoenix Neighborhoods, CBS 5 (Apr. 11, 2012, 12:21 AM), (
  • 3. On harsh punishments, see David Kennedy, Don’t Shoot 47 (2011); Marc Mauer, Racial Impact Statements as a Means of Reducing Unwarranted Sentencing Disparities, 5 Ohio St. J. Crim. L. 19, 20 (2007). See generally James Q. Whitman, Harsh Justice 56–59 (2003).

    For a brief account of parole and probation conditions, see Joan Petersilia, When Prisoners Come Home 81–83 (2003) and Joan Petersilia, Community Corrections: Probation, Parole, and Prisoner Reentry, in Crime and Public Policy 499, 512–13 (James Q. Wilson & Joan Petersilia eds., 2011).

  • 4. David A. Harris, “Driving While Black” and All Other Traffic Offenses: The Supreme Court and Pretextual Traffic Stops, 87 J. Crim L. & Criminology 544, 558–59 (1997).
  • 5. Stuntz, supra note 1, at 4. Students of the police had made the same basic point a half-century ago. Cf. James Q. Wilson, Varieties of Police Behavior 83–89 (1968); Egon Bittner, The Police on Skid-Row: A Study of Peace Keeping, 32 Am. J. Soc. 699, 702–04 (1967).
  • 6. Pervasive discretion also makes criminal justice institutions resistant to reform, as front-line workers have considerable freedom to interpret central directives in ways that those who issued them did not foresee. For historical examples, see Christopher Agee, Streets of San Francisco 71 (2014) and David Rothman, Conscience and Convenience (2002). For a contemporary example, see Katherine Beckett et al., The End of an Era? Understanding the Contradictions of Criminal Justice Reform, 664 Annals Am. Acad. Pol. & Soc. Sci. 238 (2016).
  • 7. David Kennedy et al., Reducing Gun Violence: The Boston Gun Project’s Operation Ceasefire 1 (2001).
  • 8. See David M. Kennedy, Pulling Levers: Chronic Offenders, High-Crime Settings, and a Theory of Prevention, 31 Val. U. L. Rev. 449 (1996–1997).
  • 9. The most explicit version of this concern comes from Elizabeth Griffiths and Johnna Christian, who recently wrote that “[f]ocused deterrence is thus premised, at least in part, on the heightened scrutiny associated with police surveillance and the state’s willingness to sanction aggressively; in Foucault’s terms, gang factions are expected to self-regulate as they are increasingly transformed into the objects of ‘panoptic gaze.’” Considering Focused Deterrence in the Age of Ferguson, Baltimore, North Charleston, and Beyond, 14 Criminology & Pub. Pol’y 573, 574 (2015) (internal citations omitted). Related concerns seem implicit in Tom Tyler, Jonathan Jackson, and Avital Mentovich’s recent critique of “proactive policing,” a category that apparently includes at least some versions of the focused deterrence model; see their article: The Consequences of Being an Object of Suspicion: Potential Pitfalls of Proactive Police Contact, 12 J. Empirical Legal Stud. 602, 611 (2015).
  • 10. See, e.g., Saul Green & Richard Jerome, City of Cincinnati Independent Monitor’s Final Report 31–32, 53–55 (2008); Harold Pollack, A Crime and Policing Expert Critiques Black Lives Matter’s Police-Reform Plan, N.Y. Mag. (Aug. 24, 2015, 12:12 PM), (
  • 11. See, e.g., Tracey L. Meares, The Good Cop: Knowing the Difference Between Lawful or Effective Policing and Rightful Policing—And Why It Matters, 54 Wm. & Mary L. Rev. 1865 (2013); Tracey L. Meares et al., Lawful or Fair? How Cops and Laypeople View Good Policing (Yale Law Sch. Pub. Law, Working Paper No. 255, 2014).
  • 12. For leading statements on procedural justice, see generally Tom Tyler, Why People Obey the Law (2006) and Tom Tyler & Yuen Huo, Trust in the Law (2002). Procedural justice provides the theoretical foundation for the final report of the President’s Task Force on 21st Century Policing. Tom Tyler et al., 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 (2015).

    Procedural justice scholars have made important contributions to policing by carving out a place for concerns about fairness and legitimacy that might otherwise have simply been ignored, and their driving concern for dignified treatment embodies an important moral ideal. See, e.g., Jeremy Waldron, The Rule of Law and the Importance of Procedure, 50 Nomos 3, 13 (2011). By itself, however, a procedural conception of justice remains incomplete and even counterproductive. It is wrong for police to invoke their legal authority too readily even if they do so in a procedurally fair way, and one-sided attention to fair procedures can exacerbate rather than alleviate the substantive injustice that accompanies them. Robin West recently expressed these concerns eloquently: “[E]xcessively precious procedures in the face of grotesque substantive law from which there is truly no exit, even with all the procedure in the world, can be a massive insult to dignity.” Such procedures can also be demoralizing, lending undeserved legitimacy and resilience to substantively unjust practices, making them “all the more invulnerable to change, whether through politics, revolution, or subterfuge.” The Limits of Process, 50 Nomos 32, 40, 42 (2011).

  • 13. As expressed, for example, by Friedrich Hayek: “Stripped of all technicalities [the Rule of Law] means that government in all its action is bound by rules fixed and announced beforehand.” Friedrich Hayek, The Road to Serfdom, in The Collected Works of F.A. Hayek, Vol. 2: The Road To Serfdom 112 (2007); see also the discussion of Hayek, infra note 141.
  • 14. See, e.g., Ronald Dworkin, Political Judges and the Rule of Law, in A Matter of Principle 9 (1985); Margaret Jane Radin, Reconsidering the Rule of Law, 69 B.U. L. Rev. 781, 812–19 (1989); Frederick Schauer, Rules and the Rule of Law, 14 Harv. J. L. & Pub. Pol’y 645, 691–94 (1991); Lawrence Solum, Equity and the Rule of Law, 36 Nomos 120, 120–21 (1994). More concretely, Kathleen Noonan, Charles Sabel, and William Simon develop a complex interpretation of rule of law values in public administration that goes well beyond simple rule-following. See generally Kathleen Noonan et al., Legal Accountability in the Service-Based Welfare State, 34 Law & Soc. Inquiry 523 (2009); Charles Sabel & William Simon, The Duty of Responsible Administration and the Problem of Police Accountability, 33 Yale J. Reg. 165 (2016).
  • 15. I do not claim that my interpretation describes what the participants themselves consciously had in mind. A good interpretation tries to clarify the meaning of some practice, but that enterprise is necessarily critical. If the interpretation simply restates what the participants already think they are doing, it does not make their practices any clearer. As Charles Taylor put it:

    [I]f the explanation is really clearer than the lived interpretation then it will be such that it would alter in some way the behavior if it came to be internalized by the agent as his self-interpretation. In this way a hermeneutical science which achieves its goal, that is, attains greater clarity than the immediate understanding of agent or observer, must offer us an interpretation which is in this way crucially out of phase with the explicandum.

    Charles Taylor, Interpretation and the Sciences of Man, in Philosophy and the Human Sciences 16, 27 (1985). As Taylor notes in another essay, this kind of interpretation is fundamentally normative. The claim that an interpretation of a practice is a good one is more than a descriptive claim that it captures what those engaged in it actually do; it is also a normative claim that the interpretation enables the practice to become “less stumbling and more clairvoyant.” Social Theory as Practice, in Philosophy and the Human Sciences, supra, at 90, 111. I have discussed this approach to interpretive work in police research. See David Thacher, Policing is Not a Treatment: Alternatives to the Medical Model of Police Research, 38 J. Res. Crime & Delinq. 387 (2001) [hereinafter Policing is Not a Treatment]; see generally David Thacher, The Normative Case Study, 111 Am. J. Soc. 1631 (2006); David Thacher, The Perception of Value: Adam Smith on the Moral Role of Social Research, 19 Eur. J. Soc. Theory 94 (2016).

  • 16. See generally Quinn, infra note 108; infra Section V.
  • 17. In particular, older works did not consider the political economy that generates the expansion of criminal law. Cf. Pathological Politics, supra note 2, at 510 (describing the tacit alliance between prosecutors and legislators and the marginalization of judges that contributes to overcriminalization).
  • 18. The best contemporary legal scholarship relating the problem-oriented policing model to the problem of criminal justice discretion appears in administrative law rather than criminal law. See, e.g., Sabel & Simon, supra note 14. Further back in time, Debra Livingston’s rich analysis of administrative strategies for shaping police discretion draws from problem-oriented policing and engages directly with the legacy of the ABF surveys, but it focuses more specifically on the order maintenance role of the police rather than police discretion in general. See Debra Livingston, Police Discretion and the Quality of Life in Public Places: Courts, Communities, and the New Policing, 97 Colum. L. Rev. 551 (1997) [hereinafter Quality of Life]; Debra Livingston, Gang Loitering, the Court, and Some Realism about Police Patrol, 1999 Sup. Ct. Rev. 141 (1999).
  • 19. Ronald Allen, Police and Substantive Rulemaking: Reconciling Principle and Expediency, 125 U. Pa. L. Rev. 62, 63 (1976); see generally Discretion in Criminal Justice (Lloyd Ohlin & Frank Remington eds., 1993) (providing an important retrospective and extension of this literature, with comprehensive references to the original work); see also Samuel Walker, Origins of the Contemporary Criminal Justice Paradigm: The American Bar Foundation Survey, 1953-1969, 9 Just. Q. 47 (1992). The ABF survey’s conclusions were soon reinforced by the foundational ethnographies of police work, particularly the research conducted by Egon Bittner, supra note 5, and those he influenced, especially Wilson, supra note 5.
  • 20. Fed. Bureau of Investigation, Uniform Crime Reports for the United States 93 (1958).
  • 21. See generally Quality of Life, supra note 18; William O. Douglas, Vagrancy and Arrest on Suspicion, 70 Yale L. J. 1 (1960).
  • 22. See, e.g., Norval Morris & Gordon Hawkins, The Honest Politician’s Guide to Crime Control 2–3 (1972); Herbert Packer, The Limits of the Criminal Sanction (1968) 292–95; Douglas, supra note 21, at 14.
  • 23. As Herman Goldstein put it decades later:

    Because the ABF reports described in great detail the specific incidents that the police handled, they drew attention to the broad range of conduct commonly lumped under the umbrella of crime. Within categories of specific crime labels (like assault, burglary, and certainly disorderly conduct), they illustrated with unusual clarity the endless number of unpredictable variables that distinguished one incident that the police handled from another (e.g., the presence of mental illness, the involvement of alcohol, the relationship between the victim and the offender, the age of the alleged offender, the prior record of the individuals involved). The police . . . often improvised in their responses to take note of the variables, using criteria that, depending on who reviewed them, might be praised or condemned and with results that were equally mixed. The varied responses, viewed neutrally, carried a strong message. They illustrated the need for flexibility in handling the infinite variety of situations that are brought to police attention.

    Herman Goldstein, Confronting the Complexity of the Police Function, in Discretion in Criminal Justice, supra note 19, at 23, 33–34; see also Quality of Life, supra note 18, at 575, 636–38.

  • 24. See Herman Goldstein, Police Policy Formulation: A Proposal for Improving Police Performance, 65 Mich. L. Rev. 1123, 1126–28 (1966); see also Walker supra note 19, at 68–69.
  • 25. Goldstein, supra note 24, at 1127–28.
  • 26. On the limited powers of legislators and judges, see Joseph Goldstein, Police Discretion Not to Invoke the Criminal Process: Low-Visibility Decisions in the Administration of Justice, 69 Yale L. J. 543, 568–69 (1959); Wayne R. LaFave & Frank J. Remington, Controlling the Police: The Judge’s Role in Making and Reviewing Law Enforcement Decisions, 63 Mich. L. Rev. 987 (1965). On police managers, see Herman Goldstein, Policing a Free Society 98–100, 309–11 (1977); President’s Comm’n on Law Enforcement and the Administration of Justice, Law Enforcement Policy: The Police 16–17, 42–61 (1967); Goldstein, supra note 24, at 1127.
  • 27. See, e.g., Kenneth Culp Davis, Discretionary Justice 52–57, 219–29 (1969); Kenneth Culp Davis, Police Discretion 98–112, 139–58 (1975); accord Wayne R. LaFave, Controlling Discretion by Administrative Regulations: The Use, Misuse, and Nonuse of Police Rules and Policies in Fourth Amendment Adjudication, 89 Mich. L. Rev. 442 (1990); see also President’s Comm’n, supra note 26, at 18–27; Anthony G. Amsterdam, The Supreme Court and the Rights of Suspects in Criminal Cases, 45 N.Y.U. L. Rev. 785, 813–15 (1970).
  • 28. See Walker, supra note 19, at 61, 66–67; see also Goldstein, supra note 26 (describing the problems inherent in selective enforcement).
  • 29. In Confronting the Complexity of the Police Function, Goldstein traces problem-oriented policing to the ABF surveys, as an effort to take seriously the many dimensions of the police function’s complexity that the ABF surveys uncovered. Goldstein, supra note 23, at 63–64; cf. Walker, supra note 19, at 67–68.

    The “guidelines” movement has not disappeared entirely. George Kelling has repeatedly drawn from this tradition in his influential work on the order-maintenance role of the police. See George Kelling, Broken Windows and Police Discretion 25–45 (1999); George Kelling & Catherine Coles, Fixing Broken Windows 178–93 (1996) (framing explicitly the order-maintenance role as an application of problem-oriented policing). See, e.g., Kelling & Coles, supra, at 274. The guidelines movement also survives in contemporary work on the police use of force. See, e.g., Samuel Walker & Carol Archbold, Critical Incident Policies, in The New World of Police Accountability (2014). In general, however, police rulemaking has been anemic and discretion remains extravagant. See Barry Friedman & Maria Ponomarenko, Democratic Policing, 90 N.Y.U. L. Rev. 1827, 1843–44 (“[R]ules adopted by democratic bodies to govern policing tend to be few and far between.”).

  • 30. Herman Goldstein, Improving Policing: A Problem-Oriented Approach, 25 Crime & Delinq. 236 (1979).
  • 31. See Goldstein, supra note 24, at 1126–28; Herman Goldstein, Administrative Problems in Controling the Exercise of Police Authority, 58 J. Crim. L. Criminology & Police Sci. 160, 163 (1967).
  • 32. Goldstein, supra note 26, at 95.
  • 33. See id. at 71–92.
  • 34. Goldstein repeatedly wrote about the use of guidelines as a tool for police “to systematize their experience so that it can be effectively communicated to new officers through training programs and to others, like judges . . . .” Goldstein, supra note 24,at 1133.
  • 35. Goldstein, supra note 26, at 95.
  • 36. See, e.g., Anthony Braga, Problem-Oriented Policing and Crime Prevention (2008); Crime Reduction and Problem-Oriented Policing (Karen Bullock & Nick Tilley eds., 2012); Anthony Braga & David Weisburd, Focused Deterrence and the Prevention of Violent Gun Injuries: Practice, Theoretical Principles, and Scientific Evidence, 36 Ann. Rev. Pub. Health 55 (2015).
  • 37. See, e.g., Anthony Braga et al/, Problem-Oriented Policing, Deterrence, and Youth Violence: An Evaluation of Boston’s Operation Ceasefire, 38 J. Res. Crime & Delinq. 195 (2001); Anthony Braga et al., Problem-Oriented Policing in Violent Crime Places: A Randomized Controlled Experiment, 37 Criminology 541 (1999); David Weisburd et al., Is Problem-Oriented Policing Effective in Reducing Crime and Disorder?, 9 Criminology & Pub. Pol’y 139 (2010).
  • 38. Lawrence Sherman, Problem-Oriented Policing, by Herman Goldstein, 82J. Crim. L. & Criminology 690, 705–07 (1991) (book review).
  • 39. Malcolm Sparrow, Governing Science, in New Perspectives on Policing 16–22 (Nat’l Inst. Just. & Harv. Kennedy Sch., Exec. Sess. on Policing & Pub. Safety, Jan. 2011); Nick Tilley, Whither Problem-Oriented Policing, 9 Criminology & Pub. Pol’y 183 (2010), 191–92. Tilley has usefully framed Problem-Oriented Policing as “the open-minded pursuit of ethical and effective solutions to recurrent problems.” Tilley, supra 39, at 183. The underdevelopment of the ethical dimension in this formulation is what I mean to emphasize.
  • 40. See, e.g., Goldstein, supra note 24, at 1137.
  • 41. The concerns change when the police turn to non-legal interventions, but moral ideals like justice, liberty, and equality remain important in those cases as well. Even the most anodyne-seeming environmental interventions often implicate important social values other than effectiveness, see, e.g., R. A. Duff & S. E. Marshall, Benefits, Burdens and Responsibilities, in Ethical and Social Perspectives in Situational Crime Prevention 17, 19 (Andrew von Hirsch et al. eds., 2000), and therapeutic alternatives to law enforcement obviously do so as well.
  • 42. See Policing is Not a Treatment, supra note 15, at 391.
  • 43. Michael Scott, Problem-Oriented Policing: Reflections on the First 20 Years 129 (2000).
  • 44. Goldstein opened Problem-Oriented Policing with his worries about whether “a somewhat theoretical concept for improving policing” would seem frivolous in the shadows of the drug crisis that most urban areas were struggling with at the turn of the decade, and it is hard not to wonder how his presentation was shaped by a felt need to speak to the crisis of public safety in the late 1980s. Herman Goldstein, Problem-Oriented Policing xi (1990).
  • 45. The National Institute of Justice funded the Gun Project work as a problem-oriented policing initiative focused on youth violence. See David Kennedy, Don’t Shoot 30 (2011).
  • 46. In 2000, Scott described the gun project as “the best example of high-level problem analysis.” Scott, supra note 43, at 118. A decade later, Tilley described that initiative and its offshoots as “[perhaps] the best examples we have so far” of analytically sophisticated problem-oriented policing. Tilley, supra note 39, at 192.
  • 47. Kennedy, supra note 45, at 43.
  • 48. David Kennedy, Pulling Levers: Chronic Offenders, High-Crime Settings, and a Theory of Prevention, 31 Val. U. L. Rev. 449, 461–62 (1997).
  • 49. Kennedy, supra note 45, at 54.
  • 50. Id. at 55.
  • 51. Id. at 60–70. But see David Kennedy et al., Reducing Gun Violence: The Boston Gun Project’s Operation Ceasefire 33–34 (2001). The two accounts seem to give slightly different chronologies for the Vamp Hill Kings and Intervale Posse crackdowns, presumably because in practice the two initiatives overlapped.
  • 52. Kennedy et al., supra note 51, at 27–28. The quotation comes from a description of an earlier project that Ceasefire was modeled on, but Ceasefire itself adopted the same basic message.
  • 53. To make the point vivid, they drew attention to exemplary cases like Freddie Cardoza, a gang leader sentenced to nineteen years in federal prison under an armed career criminal law after he was stopped on the street with a single bullet (he was not even carrying a gun). Id. at 35–38.
  • 54. Id. at 40.
  • 55. Kennedy, supra note 45, at 58.
  • 56. Husak, supra note 2, at 154.
  • 57. They are convenient tools for combating drug-related violence because the drug laws have become essentially strict liability offenses. Conviction is nearly automatic, unlike the violent offenses they serve as pretexts for. See Markus Dubber, Victims in the War on Crime 34 (2003); Stuntz, supra note 1, at 269–74. “Proxy laws” as Stuntz describes them are a species of overbroad laws: the implicit criterion that defines the subset of cases that ought to be pursued is something like “those in which we believe the offender has committed a hard-to-prove violent offense as well.”
  • 58. Stuntz, supra note 1, at 273.
  • 59. Douglas Husak, Overcriminalization, in A Companion to Philosophy of Law and Legal Theory 624 (Dennis Patterson ed., 2010).
  • 60. Husak, supra note 2, at 28–29 (internal citations omitted).
  • 61. See Jan Eeckhout et al., A Theory of Optimal Random Crackdowns, 100 Am. Econ. Rev. 1104 (2010); Lawrence Sherman, Police Crackdowns: Initials and Residual Deterrence, 12 Crime & Just. 1 (1990).
  • 62. For Bernard Harcourt, this randomization strategy has its own virtues. See Against Prediction 237–39 (2007). Most commentators, however, seem to share Husak’s concern. Husak himself concedes that the strategy is perfectly legal, but its moral propriety is a different question. Husak, supra note 2, at 29.
  • 63. Kennedy et al., supra note 51, at 40.
  • 64. The criterion of “what is the greatest threat to community viability” is Kennedy’s. Kennedy, supra note 45, at 225–30. One can imagine other criteria for choosing problems. See, e.g., Goldstein, supra note 26, at 95–100. Many focused deterrence initiatives do not seem very reflective about this issue, but youth violence is so obviously significant that perhaps it does not require much reflection.
  • 65. For detailed accounts of the most important analytic approaches, see Tim Bynum et al., Chronic Violent Offender Strategies: A Project Safe Neighborhoods Strategic Intervention (2002); Scott Decker et al., Gun Prosecution Case Screening: A Project Safe Neighborhoods Strategic Intervention (2006); John Klofas et al., Crime Incident Reviews: A Project Safe Neighborhoods Strategic Intervention (2006).
  • 66. Kennedy, supra note 45, at 246.
  • 67. Id. at 167–68. The implementation guide developed by the Bureau of Justice Assistance for drug market interventions like High Point stresses the importance of defining and identifying significant and active dealers. Natalie Hipple & Ed McGarrell, BJA Drug Market Intervention Implementation Guide and Lessons Learned 5, 9–11 (2009).
  • 68. The explicit, deliberative process that focused deterrence initiatives rely on does not simply correct inaccurate assumptions, such as the hangers-on mistaken for high-rate drug dealers or the mere “crack heads” who have nothing to do with serious violence. It also encourages principled decisions about the kinds of problems that deserve concentrated criminal justice attention in the first place. Crackdowns that make intensive (and even draconian) use of all available legal tools are common in criminal justice, but in the normal course of events, community pressure and traditional departmental practices determine how this substantial reservoir of authority will be deployed. Focused deterrence initiatives reopen that question for deliberate reflection. (In that respect it perfectly illustrates Goldstein’s early idea that default uses of discretion arise out of police subculture and community pressure, while guidelines might make it more principled.) For example, Kennedy compares Ceasefire’s intensive response to gang violence with the common response to an assault against a police officer. Kennedy, supra note 45, at 56. In Boston, the Working Group raised the question of whether similar crackdowns might be directed at other targets. “You know what happens when someone kills a cop,” Working Group members told one gang. “We go after everybody involved, we never back off, we never stop. That’s what we’re going to do if you guys hurt somebody. That’s what we’re going to do if somebody hurts you. Hurting you is like hurting a cop. It’s all off-limits now.” Id. at 65–66.
  • 69. Kennedy, supra note 45, at 140–48.
  • 70. Mark Kleiman stresses this aspect of the High Point intervention in When Brute Force Fails 45–48 (2009).
  • 71. David Kennedy, Deterrence and Crime Prevention 154 (2009).
  • 72. Id. at 154–55.
  • 73. Kennedy, supra note 45, at 176.
  • 74. In other words, even where it would be authorized by a narrower analogue to the existing overinclusive law—one that is “no more extensive than necessary to achieve its objective.” Husak, supra note 2, at 153.
  • 75. Kennedy, supra note 45, at 193.
  • 76. Id. at 192.
  • 77. Hipple & McGarrell, supra note 67, at 13.
  • 78. For example:

    One of the initial decisions that needs to be made is selection of the criteria that will be used to determine who is a chronic violent gun offender. What does it take to be included in the group of individuals who are determined to be ‘impact players’? Whatever criteria are selected should be applied in a standard and consistent fashion. In addition, a broad spectrum of data sources on offenders should be examined. The exclusive use of intelligence sources may omit some individuals who are deserving of being included with this group of offenders.

    Bynum et al., supra note 65, at 6.

  • 79. George Tita et al., Reducing gun violence: Results from an intervention in East Los Angeles 15–16 (2011).
  • 80. Id. at 18.
  • 81. Community input can and should play a role in those decisions, but the spontaneous demands of those who happen to attend a one-time community meeting cannot determine their outcome alone. See generally David Thacher, Equity and Community Policing: A New View of Community Partnerships, 20 Crim. Just. Ethics 3 (2001).
  • 82. I advance one moral framework for approaching these judgments in the next section.
  • 83. See, e.g., Bynum et al., supra note 65, at 36.
  • 84. But cf. id. at 7.
  • 85. Id. at 6, 8, 26.
  • 86. For example, even if gun possession offenses receive lower scores than use offenses, does it really make sense for three possession offenses to outweigh a single shooting offense? Should an offender ever be able to qualify for targeted attention based on a long string of possession offenses alone? For a general discussion of commensurability and valuation in non-criminal law, see Cass Sunstein, Incommensurability and Kinds of Valuation: Some Applications in Law, in Incommensurability, Incomparability, and Practical Reason 234 (Ruth Chang ed., 1997).
  • 87. See Harold Garfinkel & Egon Bittner, “Good” Organizational Reasons For “Bad” Clinic Records, in Studies in Ethnomethodology 186, 195–96 (1967). The focused deterrence literature specifically notes how plea bargaining may affect the meaning of criminal history files. See Bynum et al., supra note 65, at 7.
  • 88. Sarah Brayne, Stratified Surveillance: Policing in the Age of Big Data (2015) (unpublished Ph.D. Dissertation, Princeton Univ. Dept. of Sociology) (on file with author) (“Despite the stated intent of the points system to redress some of the inequalities and bias in police practices, data-driven surveillance can actually exacerbate pre-existing inequalities in insidious ways because it puts individuals already under suspicion under new and deeper forms of surveillance, while appearing to be objective, or, in the words of one captain, ‘just math.’”).
  • 89. As recommended in Bynum et al., supra note 65, at 13. For general critiques of technology- and data-driven policing, see David Sklansky, The Persistent Pull of Police Professionalism, in New Perspectives on Policing 8–12 (Nat’l Inst. Just. & Harv. Kennedy Sch., Exec. Sess. on Policing & Pub. Safety, Mar. 2011); and Petra Bartosiewicz, Beyond the Broken Window, Harper’s (May 2015), (
  • 90. See Bynum et al., supra note 65, at 1, 19.
  • 91. See generally Daniel Richman, Project Exile and the Allocation of Federal Law Enforcement Authority, 43 Ariz. L. Rev. 369 (2001).
  • 92. Kennedy, supra note 45, at 35.
  • 93. See, e.g., Kennedy, supra note 71, at 116; Klofas et al., supra note 65, at 17 (using jailhouse focus groups to identify important drug houses).
  • 94. One U.S. Department of Justice guide insists: “Given that this information will form the basis of enforcement and prosecution efforts, it is critical that it be accurate.” Bynum et al., supra note 65, at 13. In the text I want to stress the need to press harder on what the norm of “accuracy” requires. Greater skepticism should undoubtedly apply to evidence favoring enforcement than evidence favoring lenience. Concerns undoubtedly multiply when both the accuracy and moral relevance of a particular piece of information is in question. How should a working group treat a youth worker’s claim that a flagged offender has recently re-enrolled in school, a probation official’s claim that an offender has joined back up with the group that originally got him in trouble, or a community leader’s claim that an overlooked offender is a danger to the community?
  • 95. See Markus Dubber, The Police Power: Patriarchy and the Foundations of American Government 175–76 (2005) [hereinafter Police Power]. Dubber focuses particularly on so-called “possession” offenses, which are a central plank of overcriminalization. Dubber, supra note 57, at 14–16, 32–97; see also Husak, supra note 2, at 44.
  • 96. See generally Police Power, supra note 95. A similar perspective underlies E.P. Thompson’s famous account of the rule of law: “[T]here is a difference between arbitrary power and the rule of law. We ought to expose the shams and inequities which may be concealed beneath this law. But the rule of law itself, the imposing of effective inhibitions upon power and the defence [sic] of the citizen from power’s all-intrusive claims, seems to me to be an unqualified human good.” Whigs and Hunters 266 (1975).
  • 97. Dubber, supra note 57, at 21.
  • 98. For that conception, see especially Rawls’s famous statement: “Each person possesses an inviolability founded on justice that even the welfare of society as a whole cannot override.” A Theory of Justice 3 (1971). Ronald Dworkin argues that the rule of law requires respect for substantive rights in Political Judges and the Rule of Law, in A Matter of Principle, supra note 14, at 32.
  • 99. Andrew von Hirsch, Censure and Sanctions 11 (1996).
  • 100. Kennedy, supra note 71, at 119–20.
  • 101. See Austin Sarat, Authority, Anxiety, and Procedural Justice: Moving from Scientific Detachment to Critical Engagement, 27 Law & Soc’y Rev. 647, 656 (1993) (reviewing Tom Tyler, Why People Obey The Law (1990)); David Thacher, Perils of Value Neutrality, 44 Res. Soc. Org. 317 (2015).
  • 102. Kennedy, supra note 45, at 243.
  • 103. See, e.g., Jessica Brown & Jane Prendergast, Group Offers ‘Likely Killers’ List, Cin. Enquirer, Apr. 23, 2007, at A1.
  • 104. Kennedy, supra note 45, at 243.
  • 105. Elsewhere he and others do seem to tout their ability to predict just who will become involved in violence. If anything, however, clinical judgment based on this kind of ineffable situated knowledge has proven less successful than actuarial instruments in prediction. See, e.g., William Grove et al., Clinical Versus Mechanical Prediction: A Meta-Analysis, 12 Psychol. Assessment 19 (2000).
  • 106. Kennedy, supra note 45, at 243.
  • 107. Although the specific argument about the justification and moral limits of deterrence that I will develop may be controversial, any reasonable account of deterrence must recognize some moral limits—or so its leading modern theorists concede. See Franklin Zimring & Gordon Hawkins, Deterrence: The Legal Threat In Crime Control 32–50 (1973); Johannes Andenaes, The Morality of Deterrence, 37 U. Chi. L. Rev. 649, 650 (1970). In that respect, while some of the specific substantive claims I make about the ethical values that should guide focused deterrence projects may be arguable, some moral guideposts are unavoidable.
  • 108. Warren Quinn, The Right to Threaten and the Right to Punish, 14 Phil. & Pub. Aff. 327, 336 (1985). He continues: “We cannot object to certain deterrent threats of punishment that stand against us because others have a right to try to protect themselves from us by these means.” Quinn’s work (collected in Morality and Action (Philippa Foot ed., 1993)), has had a major influence on the resurgence of deontological moral theory. See Thomas Nagel, The Value of Inviolability, in Morality and Self-Interest 102 (Paul Bloomfield ed., 2008); F. M. Kamm, Non-Consequentialism, the Person as an End-in-Itself, and the Significance of Status, 21 Phil. & Pub. Aff. 354 (1992) (reviewing Shelly Kagan, The Limits of Morality (1982) and Warren Quinn, Actions, Intentions, and Consequences: The Doctrine of Doing and Allowing, 98 Phil. Rev. 287 (1989)).
  • 109. Quinn, supra note 108, at 337–47. Quinn also writes:

    Innocent actions that do not menace others are morally protected. This protection consists in the fact that we may not in general attempt to prevent them coercively or frustrate them violently. Violations of important moral rights are, on the other hand, morally exposed. That is, we may try to prevent or frustrate them by means that would, in other contexts, violate their agents’ rights.

    Id. at 346. He continues: “Morality . . . designs variances in some of our rights so that these rights will not interfere with a range of defensive strategies.” Id. at 349. In effect, Quinn insists that properly crafted deterrence need not treat its objects as “rightless.” Contra Police Power, supra note 95, at 126 (“The ‘criminal classes’ were the proper objects of policing because they, through their criminal act, had become rightless.”)

    The right to threaten punishment follows fairly easily from the right to activate the device that Quinn imagines, but the right to actually carry out the threat once it has been made is harder to justify. I will not try to summarize Quinn’s difficult argument here, but the basic idea—“[i]f the urgency of self-protection makes moral room for threats it also makes moral room for punishment”—seems plausible. Quinn, supra note 108, at 370. Phillip Montague provides a simpler analysis of the same problem:

    [I]f establishing a system of punishment is justified, then creating certain positions that system requires, with their attendant responsibilities, is justified, and the people who occupy those positions are certainly justified in fulfilling their responsibilities—that is, presumptively justified in participating in the punishment of individuals.

    Phillip Montague, Punishment as Societal Self-Defense 72 (1995).

  • 110. Quinn, supra note 108, at 373.
  • 111. Pathological Politics, supra note 2, at 521–22.
  • 112. Stuntz, supra note 1, at 269–70.
  • 113. Id. at 273.
  • 114. William J. Stuntz, Unequal Justice, 121 Harv. L. Rev. 1969, 2025 (2008).
  • 115. See, e.g., Kennedy et al., supra note 51, at 39; Kennedy, supra note 71, at 123–27.
  • 116. Kennedy et al., supra note 51, at 29.
  • 117. Id. at 40.
  • 118. Kennedy et al., supra note 45, at 65.
  • 119. “Throughout the operation, [Youth Violence Strike Force] officers, probation officers, and Streetworkers had told gang members directly why the crackdown was occurring and what it would take to make it stop.” Kennedy, supra note 51, at 27–28.
  • 120. For example, Giuliani’s randomized “federal day” crackdowns may or may not have been an effective deterrence strategy, but they clearly did not provide fair notice to the people they affected; see Husak, supra note 59. Some focused deterrence projects have also failed in this respect. Boston made a deliberate decision not to warn the Vamp Hill Kings (the first gang targeted in the original Gun Project) before taking action—before serving old warrants, stepping up probation enforcement, fast tracking state prosecution, referring cases for federal prosecutions, and initiating deportation proceedings. Officials believed that this initial crackdown needed to take place before issuing a warning “to build credibility;” the Vamps were invited to their first call-in several weeks later. Id. at 34. Los Angeles apparently never communicated its enforcement plans with the two gangs the project targeted. Some cities have decided explicitly not to notify the people who wind up on their high-priority “chronic violent offender” lists. These decisions may or may not have been effective ways of maximizing deterrence, but they clearly failed to provide fair notice. When the decision to concentrate discretionary enforcement on particular people and groups can carry such momentous consequences, and when such concentrated enforcement represents a major departure from the status quo, these decisions invite the concerns about fair notice that the best focused deterrence initiatives help to resolve.
  • 121. E.g., the “morally exposed” actions that Quinn describes in an earlier quotation, not the “morally protected” innocent actions he describes in the same passage.
  • 122. See Quinn, supra note 108, at 345–46.
  • 123. Edmund McGarrell, Focused Deterrence and “Pulling Levers”, in Encyclopedia of Criminology and Criminal Justice 1667 (Gerben Bruinsma & David Weisburd eds., 2014). Kennedy gives a similar account: “In Lowell, Massachusetts, authorities working with Harvard researcher Anthony Braga addressed street shootings by young Asian gang members by raiding the gambling enterprises of their elders—and telling the elders that the shootings were bringing the unwanted attention. The shootings stopped entirely.” Kennedy, supra note 72, at 117.
  • 124. Moreover, it seems to do so in racially inequitable ways: low-level gambling offenses in the city’s Southeast Asian communities will receive attention when similar offenses elsewhere do not.
  • 125. Anthony Braga et al., Understanding and Preventing Gang Violence: Problem Analysis and Response Development in Lowell, Massachusetts, 9 Police Q. 20, 40 (2006); accord Jack McDevitt et al., Lowell, District of Massachusetts: Case Study 6, a Project Safe Neighborhoods Strategic Intervention 12–14 (2007).
  • 126. See, e.g., Andrew Papachristos & David S. Kirk, Changing the Street Dynamic: Evaluating Chicago’s Group Violence Reduction Strategy, 14 Criminology & Pub. Pol’y 525 (2015). Kennedy’s most detailed analysis of the focused deterrence model emphasizes that the capacity to impose group responsibility is a distinctive advantage of strategies that rely heavily on discretion. Kennedy, supra note 71, at 121–22.
  • 127. Kennedy, supra note 45, at 53.
  • 128. Id.
  • 129. Philip Cook et al., Gun Control, in Crime and Public Policy 281 (James Q. Wilson & Joan Petersilia eds., 2011).
  • 130. Kennedy, supra note 45, at 67.
  • 131. The topic is too complex to treat in any detail here. Useful touchstones for thinking about collective responsibility include Joel Feinberg, Collective Responsibility, 65 J. Phil. 674 (1968); John Kekes, Collective Responsibility as a Problem for Liberalism, 20 Midwest Stud. Phil. 416 (1995); Dennis Thompson, Moral Responsibility of Public Officials: The Problem of Many Hands, 74 Am. Pol. Sci. Rev. 905 (1980).
  • 132. See, e.g., Husak, supra note 59, at 62–63; Luna, supra note 2, at 729–31; Stuntz, supra note 1, at 301–02.
  • 133. “Law enforcement, criminal justice, is a massive, one-size-fits all architecture,” Kennedy observes. “It’s the same for everybody, everywhere, all the time. This was focused, customized, particular, highly specific.” Kennedy, supra note 45, at 54.
  • 134. See note 66 and the associated text above. Cf. James Forman, Jr. & Trevor Stuntz, Beyond Stop and Frisk, N.Y. Times (Apr. 19, 2012), The same comparison has taken a more abstract and general turn in Sabel & Simon, supra note 14, which unfavorably contrasts what they call “assertive policing” with the focused deterrence model as an elaboration of problem-oriented policing.
  • 135. Both also run up against the challenge of acknowledging and justifying selective enforcement. In his earliest reflections on guidelines, Goldstein observed that “[t]he whole thought of trying to defend a policy of selective enforcement is a bit frightening.” Herman Goldstein, Police Discretion: The Ideal Versus the Real, 23 Pub. Admin. Rev. 140, 145 (1963). More than four decades later, the troubled Baltimore focused deterrence project ran into debilitating controversy over concerns about selective prosecution. As one Mayoral spokesman put it: “[H]aving enough information to indict somebody and then not actually doing so is not something that this group felt was appropriate.” Justin Fenton, A Look At What Works In Fighting City Crime: Baltimore Joins Network To Discuss Safety Strategies, Balt. Sun (June 24, 2009), (
  • 136. See Allen, supra note 19. Here I depart from the specific tool emphasized by Friedman & Ponomarenko, supra note 29, at 1832, 1884. Rulemaking has an important role to play, particularly in the tactical areas that Friedman and Ponomarenko focus on, but by itself it cannot adequately channel the discretion embodied in the substantive criminal law.
  • 137. As Goldstein put it two decades after the guidelines movement’s peak: “The intent, on the part of the original proponents [of written guidelines for the use of discretion], was to view the development of policy as a way to encourage responsible thought about a difficult issue and to view the end product as a conclusion of this process—as a statement of the way in which a given police agency thinks about a problem—which can then serve as guidance for its personnel. The tendency in practice, however, is to treat anything that is written down as rules, which then are interpreted as edicts. When this occurs, the desired flexibility is lost, and any necessary diversion from the rules to meet the variety of situations the police confront returns the police to acting in a sub rosa manner.” Goldstein, supra note 23, at 55.
  • 138. Nicholas Parrillo, Against the Profit Motive 288 (2013). Parrillo describes the rise of discretionary administration in the nineteenth and early twentieth centuries as a solution to the problem of rigid rules across a wide range of regulatory domains: “No rule could describe with perfect accuracy the conduct that lawmakers intended to proscribe, so lawmakers commonly wrote an overly broad rule whose sharp edges could be ‘sanded off ’ through selective nonenforcement.” Id. at 40. For example, as Congress became concerned that law enforcement by U.S. Attorneys had become “excessively, perversely rule-bound” by the end of the nineteenth century, it took steps “to stop officers from standing on the letter of the law and instead focus on its amorphous ‘spirit’,” allowing “the scope of statutory criminal liability to remain broad while trusting in street-level discretion to remedy its overbreadth, case by case.” Id. at 288.
  • 139. As Goldstein acknowledged, the typical police official is often more than tolerant towards overbroad laws: “His attitude is often that the law should be left on the books; it may come in handy sometime.” Goldstein, supra note 135, at 144; accord Allen, supra note 19, at 106–07.
  • 140. See the references in note 14 above.
  • 141. Parrillo, supra note 138, at 40 (“[W]ithout these kinds of forbearance, alien governance could become so radically divorced from prevailing moral sentiment or so practically unbearable that it would lose even the grudging cooperation of the populace”). The most prominent dissent from this perspective comes from Hayek, who insisted on rigorous enforcement of the law as written: “It may even be said that for the Rule of Law to be effective it is more important that there should be a rule applied always without exceptions than what this rule is. Often the content of the rule is indeed of minor importance, provided the same rule is universally enforced.” Hayek, supra note 13, at 117. This commitment reflected Hayek’s emphasis on the economic realm, and his fervent hope for a predictable field of action in that domain: “Like the laws of nature, the laws of the state provide fixed features in the environment in which [a person] has to move; though they eliminate certain choices open to him, they do not, as a rule, limit the choices to some specific action that somebody else wants him to take.” F. A. Hayek, The Constitution of Liberty, in The Collected Works of F.A. Hayek, Vol. 17: The Constitution of Liberty 221 (2007). Whatever the merits of this view in the economic realm, it makes less sense for criminal justice. Perhaps there is no moral reason to hold a farmer responsible for the damage his cows have done to a neighbor’s unfenced plot, rather than hold the neighbor responsible for failing to erect a fence; in that case it may matter more that the rule that fixes liability be clear than what that rule is. See Ronald Coase, The Problem of Social Cost, 3 J. Law & Econ. 1, 2 (1960). But the same cannot be said in the domain of criminal law. It matters morally whether criminal law defines robbery as taking something by force or threat of force or, alternatively, theft while in possession of a weapon. As it happens, consistent enforcement of either rule will sometimes be obtuse. Intelligent application of both rules together may be necessary to truly identify those who are morally blameworthy to the same degree—and it is identifying that group of equally blameworthy people that must be the goal for criminal law enforcement.
  • 142. The organization of Kennedy’s most recent account of focused deterrence seems to imply this understanding. The book begins with an account of the riots that erupted in Cincinnati after police shot Timothy Thomas in 2001, Kennedy, supra note 45, at 1, and it concludes with an account of Cincinnati’s transformation under a U.S. Department of Justice consent decree that mandated problem-oriented policing. Id. at 232–66. I have only tried to provide philosophical and historical context for this new and promising account of what focused deterrence involves.
  • 143. See supra note 9. The focused deterrence project in Chattanooga, Tennessee recently stalled in the face of serious conflicts with the county District Attorney after he raised ethical concerns about devoting special attention to cases the working group had nominated. Shelly Bradbury & Zack Peterson, District Attorney Neal Pinkston Says He Won’t Give VRI Cases Priority, Chattanooga Times Free Press (Nov. 8, 2015),
    news/local/story/2015/nov/08/district-attorney-neal-pinkstsays-he-wont-giv/334595/ ( Whether or not those concerns are legitimate or sincere is impossible to judge from a distance. What seems clear, however, is that focused deterrence projects that take the high moral ground by embracing rule of law values will be in an especially strong position to address concerns of that kind.