2

The Landscape of Proactive Policing

The previous chapter provided a broad historical context and specific definition of proactive policing. But, having delineated a broad concept of proactive policing, the committee also notes that the various programs and interventions undertaken in the name of proactive policing differ greatly in terms of both what the police do and the theoretical models that inform their activities. Any description of proactive policing is made even more difficult by the fact that police activities span a wide array of responsibilities, many of them shared with state or federal law enforcement. For example, the committee decided not to examine proactive policing approaches to white collar crime, which are primarily carried out by federal law enforcement agencies. Similarly, we did not consider law enforcement efforts to deal with organized crime, international drug trafficking, or trafficking in human beings. The committee decided that while such activities are often proactive and may involve local law enforcement, they were not a part of the landscape of proactive policing that has come to be associated with municipal policing in American cities, which was the main focus of our discussions. This means, to a great extent, that the policing strategies reviewed in this report refer to those public, frontline policing strategies that have been applied to prevent or reduce ongoing, street-level crime and disorder harms.

In focusing on this range of proactive policing, we faced an additional problem. How could such a broad array of approaches be linked in ways that would help to draw broader conclusions about the broad mechanisms underlying prevention? The committee decided that an approach that identified what the key logic models of prevention were at the outset would pro-

vide for the greatest insights into understanding whether proactive policing was effective and whether and how it affects communities, the lawfulness of policing, and racial disparities and racial bias. Taking this approach meant that we recognized at the outset that policing in the real world would not conform simply to the prevention models we identified. In the real world of policing, practices may draw upon a variety of logic models for prevention. This makes sense when the goal is preventing crime rather than identifying the underlying theoretical mechanisms that create preventive outcomes. What this means in practice is that specific programs carried out in policing often fall across the categories defined by the committee.

The committee identified four broad approaches to crime prevention that summarize the directions that proactive policing has taken over the past few decades: place-based approaches, problem-solving approaches, person-focused approaches, and community-based approaches (see Table 2-1). While the police practices described in this report may include elements of multiple models of prevention, it is generally the case that they develop primarily as a response to the insights of one logic model in particular. For example, hot spots policing and predictive policing developed primarily in response to the insights underlying the logic model of place-based prevention (described below), whereas community-oriented policing and procedural justice policing rely primarily on a logic model emphasizing the key role played by communities in crime prevention. This does not mean that specific programs do not also draw from other logic models of prevention. Rather, it is possible to think about the broad directions of proactive policing in reference to these categories and, more generally (as we do in later chapters), to draw broader conclusions about why programs or practices have the impacts observed.

The place-based approach seeks to focus policing resources more efficiently and effectively by capitalizing on the concentration of crime incidents at certain locations, or microgeographic places, within a department’s entire jurisdiction. Policing strategies that take a place-based approach include hot spots policing, predictive policing, and use of closed circuit television (CCTV).

A second approach, referred to here as the problem-solving approach, seeks to take a scientific approach to diagnosing the problems that underlie a pattern of crime incidents. After identifying the causes of these problems, it attempts to tailor solutions to the problems by addressing their causes, thereby preventing (or reducing) future crime. Strategies that take this approach include problem-oriented policing and third party policing.

The third approach focuses on deterring crime by capitalizing on the insight that a small proportion of the crime-committing population commits a disproportionate share of the crimes. Strategies that employ this person-

focused approach include focused deterrence; repeat offender programs; and stop, question, and frisk (SQF).

The fourth approach, which we call the community-based approach, focuses on involving the community in defining the key problems of policing and on fostering the community’s role (as understood by a strategy’s logic model) in maintaining order and public safety. Strategies that take a community-based approach include community-oriented policing, procedural justice policing, and broken windows policing.

These four approaches have different implications for the outcomes of policing, whether those outcomes be crime control, a community’s evaluation of its police, the lawfulness of policing, or potential disparities or bias in the application of policing. To understand why and how these approaches have been used in actual policing programs and interventions, we will ask three questions for each approach:

1. What factors underlay its emergence as a proactive policing approach?
2. What are the main types of policing practices (here called strategies) that use this approach?
3. What is the underlying logic model, and the evidence for that model, that informs strategies that adopt this policing approach?

Before applying the conceptual framework and its taxonomy of policing approaches and strategies to the real world and the research literature about it, two caveats are in order. First, as already noted, actual policing programs and implementations of proactive practices often incorporate elements that fall under two or more of the approaches as defined above; even more frequently, they combine elements from several strategies, as these are defined in this chapter. To aid comprehension, we reserve the terms “approach” and “strategy” for the taxonomic elements of the framework summarized in Table 2-1. We reserve “logic model” for the rationale underlying an approach or a strategy implementing an approach. Second, although the committee has adopted terminology in common use in the research literature and in policing practice, we recognize that the strict characterizations given in this report will sometimes conflict with how these terms are used in one study or another. For purposes of our discussion, the committee has interpreted whatever terminology the original authors used into the terminology of our conceptual framework.

STRATEGIES FOR A PLACE-BASED APPROACH

Policing has always had a geographic or place-based component, especially in how patrol resources are allocated for emergency response systems

TABLE 2-1 Four Approaches to Proactive Policing

(Sparrow, Moore, and Kennedy, 1992). In order for officers to respond quickly to citizen calls, police organizations developed geographically based systems that took into account the crime levels in particular areas. Under the standard model of policing, which emphasized shortening response times, police resources were organized using macrogeographies, which refers to areas the size of patrol officers’ beats, an organization’s precincts, or other relatively large administrative areas. In contrast to the standard model, proactive place-based policing (see Weisburd, 2008) focused on smaller, “micro” units of geography, often termed “crime hot spots.” Such a hot spot might be a single building or address; street segments or the faces of a street block; or clusters of addresses, block faces, or street segments with common crime problems.

Since the 19th century, scholars have found evidence that crime is more prevalent in some places than others (Guerry, 2002; Quetelet, 1842; Mayhew, 1968). However, research emerging in the late 1980s showed that this concentration of crime occurred at a very microgeographic level. Place-based proactive policing developed in response to this growing body of evidence (Sherman, Buerger, and Gartin, 1989; Sherman and Weisburd, 1995; Weisburd and Green, 1995). Its logic model was based on the research findings that crime incidence was highly concentrated in crime hot spots. As Sherman and Weisburd (1995, p. 629) remarked in the first large-scale test of effectiveness of hot spots policing in Minneapolis, Minnesota, if “only 3 percent of the addresses in a city produce more than half of all the requests for police response, if no police are dispatched to 40 percent of the addresses and intersections in a city over one year, and, if among the 60 percent with any requests the majority register only one request a year (Sherman, Buerger, and Gartin, 1989), then concentrating police in a few locations makes more sense that spreading them evenly through a beat.”

Important to the development of place-based policing are theoretical perspectives that also emerged during this period (Braga et al., 2011; Weisburd and McEwen, 1997; Weisburd and Telep, 2010). Key to the standard model of police patrol had been the idea that opportunities for crime were common throughout the urban landscape (see Repetto, 1976). But with the entry of economists into the analysis of crime (Becker, 1968; Ehrlich, 1973; Cook, 1986), the assumption that the crime rate was somehow determined by the number of “offenders” was challenged. The economic theory of crime conceptualized criminal behavior as a choice available to everyone, influenced by the perceived costs and benefits of available criminal opportunities. The crime rate, from this perspective, is determined both by the potential payoff to exploiting an opportunity (amount of “loot” in the case of property crime) and by the probability of arrest and punishment. The availability of attractive criminal opportunities

is thus one determinant of crime and is itself heavily influenced by private behavior of potential victims.

Routine activities theory (Cohen and Felson, 1979), situational prevention (Clarke, 1995), and crime pattern theory (Brantingham and Brantingham, 1993) challenged the idea that criminal opportunities were unending and raised the question of whether specific places have characteristics that attract or generate crime. These perspectives, which are often termed “opportunity theories” (see Cullen, 2010; Wilcox, Land, and Hunt, 2003), suggested that reduction of crime opportunities at specific places would likely prevent crime without displacing it to other locations.¹ Using this theoretical background, advocates of place-based policing argued that traditional objections to targeting microgeographic hot spots—objections that assumed crime displacement—would be unlikely to offset the crime prevention gains generated by focusing policing on hot spots.

The underlying logic model of place-based policing—that police can capitalize on the strong concentration of crime at microgeographic places—has been confirmed in a large number of studies over the past few decades (see Andresen and Malleson, 2011; Braga, Papachristos, and Hureau, 2014; Brantingham and Brantingham, 1999; Crow and Bull, 1975; Curman, Andresen, and Brantingham, 2015; Pierce, Spaar, and Briggs, 1988; Roncek, 2000; Sherman, 1997; Sherman, Buerger, and Gartin, 1989; Weisburd and Amram, 2014; Weisburd et al., 2004; Weisburd and Green, 1995; Weisburd, Morris, and Groff, 2009; Weisburd, Maher, and Sherman, 1992; Weisburd, Groff, and Yang, 2012; Weisburd, 2015). These studies confirmed that microgeographic concentrations of crime do not necessarily conform to traditional ideas about crime and communities. In particular, neighborhoods that are considered troubled often have discrete locations that are free of crime, and crime hot spots do occur in neighborhoods that are generally viewed as advantaged and not crime prone (see, e.g., Weisburd, Groff, and Yang, 2012). A number of studies also suggested that hot spots of crime are often stable over long periods of time (see, e.g., Weisburd, Groff, and Yang, 2012; Andresen and Malleson, 2011).

Hot Spots Policing

Sherman and Weisburd (1995) developed the strategy of hot spots policing in the Minneapolis Hot Spots Patrol Experiment. Hot spots policing covers a range of police responses, but they all focus resources on locations where crime incidents have been highly concentrated. By focusing on microgeographic locations with high concentrations of crime, hot spots policing

___________________

¹ The four main dimensions of opportunity theory are (1) motivated offenders, (2) suitable targets, (3) guardianship, and (4) accessibility/urban form.

aims to increase the general deterrence of police actions, in this case by increasing perceptions of the certainty of enforcement action (Durlauf and Nagin, 2011). There may also be a specific deterrent impact of hot spots policing, if offenders who are arrested because of increased patrols are thereby dissuaded from future offending. In addition to specific and general perceptual deterrence, police can also alter the situational opportunities that exist at hot spots by altering the environmental design of places (see, e.g., Clarke, 1997), engaging “place guardians” such as building managers or store owners (Eck and Weisburd, 1995), and engaging communities at the hot spots (Weisburd, Davis, and Gill, 2015).²

Once a hot spot is identified, police may implement a range of tactics appropriate to the particular type of hot spot to prevent crime in the given microarea, and these tactics often incorporate elements typical of one or another of the other three proactive policing interventions discussed above (refer to Table 2-1). In 2008, the Police Executive Research Forum (PERF) conducted a survey on hots spots policing that was distributed to its general members.³ The results of the survey indicate that when police engage in hot spots policing, they implement aspects of general patrol/enforcement strategies, an offender-oriented strategy, problem-oriented and community-oriented strategies, or a general investigative strategy.⁴Table 2-2 shows the use of each policing practice by the principal type of crime associated with

___________________

² As noted earlier, actual policing practice often combines elements from two or more of the approaches. A hot spots policing practice that seeks to engage the community could easily become a hybrid of place-based and community-based approaches.

³ PERF agencies represent an important and influential group of the nation’s largest police forces. To be eligible for PERF general membership, one must be the executive head of a state or local police agency that has 100 or more employees and/or serves a jurisdiction of at least 50,000 persons. The survey discussed here was completed by 191 PERF agencies, representing a response rate of 63 percent. “The responding agencies were predominantly large, with a mean of 997 officers and a median of 315. Their service populations averaged nearly 460,000 and had a median size just below 161,000. Eighty-three percent of the [agencies] were municipal agencies, while the remainder consisted primarily of county [agencies] (13 percent).” The U.S. agencies in the sample represented all four primary regions of the United States (as defined by the U.S. Census Bureau), and their jurisdictions accounted for 21 percent of the country’s population in 2006 (Koper, 2014, pp. 126–127).

⁴ A general patrol/enforcement strategy can include such practices as directed patrol, traffic stops and field interviews, order maintenance, foot patrol, overtime saturation patrol, fixed police presence, and use of mobile suppression or saturation units. An offender-oriented strategy may consist of interventions that target known offenders, execute warrant services, and check on probationers and parolees. Problem-oriented and community-oriented strategies include problem analysis and problem solving, intervening at problem locations, community policing partnerships, and multiagency task force operations. General investigative strategies consist of interventions such as surveillance, decoy operations, buy-bust/reverse stings, and the use of technologies like surveillance cameras or gunshot detection systems (Koper, 2014).

TABLE 2-2 Percentage of Responding Agencies Using a Proactive Policing Practice, by Principal Crime Type Associated with a Hot Spot

NOTE: n = 176 agencies.

SOURCE: Koper (2014, p. 130).

a hot spot.⁵ As these results illustrate, there is often overlap between tactics used in hot spots policing and tactics typically associated with the other proactive policing approaches in Table 2-1. Box 2-1 describes a hot spots policing program in Sacramento, California, further demonstrating that police departments often use a range of tactics from different approaches at hot spots.

Predictive Policing

Predictive policing is a strategy for proactive policing that uses predictive algorithms based on combining different types of data to anticipate where and when crime might occur and to identify patterns among past criminal incidents. Predictive policing tends to focus on geospatial predic-

___________________

⁵ The underlying logic model of hot spots policing does not limit police to implementing only these practices. For example, the police could engage in community policing (foot patrol being one tactic often associated with community policing, another being door-to-door getting-to-know-you police interventions) or procedural justice.

tion of crime activity, and many police departments who adopt predictive policing approaches use computer software to generate maps of predicted crime activity.⁶

Predictive policing takes data from disparate sources (both real-time crime data and frequently other noncrime data) and identifies patterns in the aggregated dataset. Police then use those patterns to anticipate, prevent, and respond more effectively to future crime. Predictive policing overlaps with hot spots policing but is generally distinguished by its reliance on sophisticated analytics that are used to predict likelihood of crime incidence within very specific parameters of space and time and for very specific types of crime.

Predictive methods can be used to predict crime incidence by type, predict offenders, predict perpetrators’ identities, or predict victims of crime. For geospatial prediction of crime activity, many police departments use computer software to generate maps of predicted crime activity. Methods used to identify likely perpetrators of past crimes use available information from crime scenes to automatically link suspects to crimes; methods predicting potential victims of crimes identify at-risk groups and individuals, such as those in proximity to at-risk locations, individuals at risk of victimization, and individuals at risk of domestic violence (Perry et al., 2013).

Making predictions is only half of predictive policing; the other half is carrying out interventions that act on the predictions (Perry et al., 2013). Police combine predictions (and crime analysis more generally) with strategies and tactics at predicted locations. For example, in Shreveport, Louisiana, the police department used monthly predictions of locations of future crimes to drive a strategic decision-making model that included increasing officer awareness of hot spots in roll call and using predictions to implement a broken windows intervention (Hunt, Saunders, and Hollywood, 2014).

Predictive policing as a strategy for proactive policing has its origins in the National Institute of Justice’s first predictive policing symposium, held in 2009 in Los Angeles. Participants at that meeting identified numerous

___________________

⁶ According to Ferguson (2012, p. 265), predictive policing is a “generic term for any crime fighting approach that includes a reliance on information technology (usually crime mapping data and analysis), criminology theory, predictive algorithms, and the use of this data to improve crime suppression on the streets.” Ratcliffe (2014, p. 4) defines predictive policing as “the use of historical data to create a spatiotemporal forecast of areas of criminality or crime hot spots that will be the basis for police resource allocation decisions with the expectation that having officers at the proposed place and time will deter or detect criminal activity.” However, predictive policing methods may at times also focus on predicting individuals who may become offenders or on predicting perpetrator identities using regression and classification models that include risk factors, statistical modeling to link crimes, and computer-assisted queries and analysis of intelligence and other databases (Perry et al., 2013).

potential applications of predictive policing, but the primary actual use was the description of the time and location of future incidents in a crime pattern or series. For example, police in Richmond, Virginia, used predictive policing methods to analyze random gunfire incidents. Using this analysis to make predictions, they were able to anticipate the time, location, and nature of future incidents (Uchida, 2009).

There are a number of companies that sell commercial predictive policing software (e.g., PredPol and HunchLab 2.0) as well as one program funded by the National Institute of Justice and available without charge.⁷ These software programs require access to real-time crime data (and, sometimes, other noncrime data) that are geocoded, reliable, and fit for the analytic purpose. The software must also have an appropriate algorithm that can produce viable predictions when needed and produce them in a format that is easily translated to operational personnel. Beyond the software, in order to implement predictive policing, a decision-making system in the operational environment capable and willing to make resource allocation decisions based on the predictions is necessary, along with the adoption of appropriate tactics tailored to the crime problem (see Ratcliffe, 2014). We note, however, that a software program is not necessary to produce results akin to those produced by predictive policing software programs; with sufficient knowledge and under the right circumstances, a well-trained crime analyst could perform the activities of a dedicated software program.

Because the concept of predictive policing is relatively recent, there is a lack of clarity with regard to both the specifics of operationalization of these definitions and the specifics of the police strategies applied (Santos, 2014). The effectiveness of predictive policing is difficult to establish because, to be a bona fide new policing strategy, it may require combining two components. The first is a software algorithm or prediction regime that is able to better predict future criminality than any existing alternative mechanisms (such as current software for crime mapping and/or the abilities of a crime analyst). Second, predicted grids should incur an operational response that is identified specifically with predictive policing.

Closed Circuit Television

CCTV is a surveillance technology comprising one or more video cameras connected in a closed circuit to a monitoring system. A CCTV system for proactive policing usually includes a number of cameras that can pan, tilt, and zoom in various directions; a mechanism to convey the real-time images to a monitoring location; a range of other elements that store,

___________________

⁷ The PROVE software utility is available at https://www.hunchlab.com/tools/prove [July 2017].

display, or otherwise monitor the camera live feed; and a human element whereby someone monitors the images either in real time or in response to an incident (Ratcliffe, 2006). Though CCTV may be used reactively, the committee examines here the uses of CCTV for proactive policing; that is, when CCTV is used to view suspicious situations or disorders, to which police might be able to respond before the situation deteriorates into a crime incident.

How cameras are monitored varies significantly among police departments. When used proactively, CCTV cameras are actively monitored, requiring a person who watches the camera feed and can deploy personnel to the incident in real time. Some CCTV systems, such as in the town of Malmö, Sweden, are actively monitored only during weekend nights from midnight to 6 a.m. (Gerell, 2016). Some systems may also have such high camera-to-operator ratios that doubt is cast as to the level of “active” monitoring actually taking place (Smith, 2004; Piza et al., 2015).

CCTV cameras are used to increase the risks for offenders of committing crime and specifically comprise a formal surveillance mechanism that enhances or replaces the role of police or security personnel (Welsh and Farrington, 2008; Clarke and Eck, 2005). In other words, prevention occurs if a potential offender is aware of the camera and makes the decision that the risk of capture outweighs the benefits of the imminent offense (Ratcliffe, 2006). CCTV cameras placed overtly are hypothesized to generate a general deterrence mechanism that increases the perceived risk of capture among the general potential offender population, should a crime be committed. They also raise the possibility of specific deterrence by which any offenders who are captured through use of the camera scheme are dissuaded from future offending. This perceptual deterrence is therefore rooted in the certainty, severity, and celerity of punishment, where “deterrence is maximized by sanctions that are perceived as inexorable, burdensome, and expeditious” (Apel, 2016, p. 59). CCTV aims to heighten the last of these: perception of the celerity of enforcement action.

STRATEGIES FOR A PROBLEM-SOLVING APPROACH

Herman Goldstein argued in 1979 that the police could be more effective in reducing crime if they took a more “problem-oriented” approach. Goldstein noted that the police had become so concerned with the means of policing that they had neglected the goals of policing. He called on police to refocus on those goals, which in his view could be defined as solving problems in communities (Goldstein, 1979, 1990). The logic model of problem solving assumes that if the police focus on specific problems, they will be more successful at reducing crime and other community problems. Strategies for a problem-solving approach, such as problem-oriented

policing and third party policing emerged from Goldstein’s work.⁸ These problem-solving strategies seek to identify causes of problems and draw upon innovative solutions with attention to assessment.

Problem-solving strategies, with their analytic focus, often incorporate policing practices characteristic of other approaches. For example, in practice there is often overlap of problem-solving practices with practices typical of a place-based approach and a person-focused approach. As will be discussed below, interventions that primarily take a community-based approach often explicitly include elements characteristic of the problem-solving approach as well.

Problem-Oriented Policing

Problem-oriented policing is an analytic method for developing crime reduction tactics. This strategy draws upon theories of criminal opportunity, such as rational choice and routine activities, to analyze crime problems and develop appropriate responses (Clarke, 1997; Braga, 2008; Reisig, 2010). Using a basic iterative process of problem identification, analysis, response, assessment, and adjustment of the response (often called the scanning, analysis, response, and assessment [SARA] model), this adaptable and dynamic analytic method provides a framework for uncovering the complex mechanisms at play in crime problems and for developing tailor-made interventions to address the underlying conditions that cause crime problems (Eck and Spelman, 1987; Goldstein, 1990). Depending on the nuances of particular problems, the responses that are developed—even for seemingly similar problems—can be diverse. Indeed, problem-oriented policing interventions draw upon a variety of tactics and practices, ranging from arrest of offenders and modification of the physical environment to engagement with community members.

Historically, most police departments engaged in incident-driven crime prevention strategies. These departments sought to resolve individual crime incidents instead of addressing recurring crime problems (Eck and Spelman, 1987). In his seminal article that challenged existing police policy and practice, Herman Goldstein (1979) proposed an alternative: police should search for solutions to the recurring problems that generated repeated calls. Goldstein described this strategy as the “problem-oriented approach” and envisioned it as a departmentwide activity. He intended for problem-oriented policing to also address the problem of unguided police discretion

___________________

⁸ Proactive partnerships with other organizations (such as code or liquor enforcement agencies, schools, probation, and private businesses), situational crime prevention, and crime prevention through environmental design are also commonly used as practices for a problem-solving approach. These are generally included in our review as third party policing practices.

(which could give rise to negative consequences such as improper use of force, ineffective crime reduction procedures, corruption, and discriminatory practices) and the “means-over-ends syndrome” (meaning an overemphasis on means, without appropriate attention to the goals, or ends).

Goldstein also emphasized from the outset that police engaged in problem-oriented policing should focus their efforts on problems the community cares about and that practices typical of a community-based approach should be among those considered in trying to reduce any given problem. Eventually the Office of Community Oriented Policing Services (COPS Office) in the U.S. Department of Justice formally hybridized problem-oriented policing and community-oriented policing by making problem-oriented policing a key element of its community-oriented policing strategy (discussed below). Thus, like many of the strategies discussed in this chapter, problem-oriented policing is in practice often implemented in practice in conjunction with practices characteristic of other policing approaches.

Problem-oriented policing requires police to be proactive in identifying underlying problems and to develop an array of tactics to address the problem, not just a particular police tactic (Goldstein, 1990; see also Weisburd et al., 2008, p. 10). However, research suggests that it is often difficult for police officers to fully implement a problem-oriented policing strategy (Cordner, 1998; Eck and Spelman, 1987; Clarke, 1998; Braga and Weisburd, 2006). Indeed, the research literature is filled with cases where problem-oriented policing programs tend to fall back on traditional methods and tend to have weak problem analysis components (Buerger, 1994; Capowich and Roehl, 1994; Cordner and Biebel, 2005; Read and Tilley, 2000). Box 2-2 describes a problem-oriented policing program in the Jacksonville, Florida, Sheriff’s Office.

Third Party Policing

Third party policing draws upon the insights of problem solving but also leverages “third parties” who are viewed as significant new resources for preventing crime and disorder. The argument for third party policing asserts that the police cannot successfully deal with many problems on their own. Thus, the failures of the standard model of policing are inherent in the limits on police powers, and crime prevention requires police engagement with third parties. Using civil ordinances and civil courts or the resources of private agencies, police departments engaged in third party policing recognize that much social control is exercised by institutions other than the police (e.g., public housing agencies, property owners, parents, health and building inspectors, and business owners) and that crime can be better managed through coordination with these institutions, using means other than the criminal law.

Mazerolle and Ransley (2006, p. 192) suggested that the growth of third party policing may be part of a larger transformation of governance in Western democracies away from state sovereignty and control and toward “networks of power.” Meares (2006, p. 207) similarly suggested that third party policing bears similarities to certain forms of civil regulation (e.g., of accountants, lawyers, employers, and sports leagues) and that, given the pervasive forms of civil regulation today, it is not surprising that third-party efforts are becoming common in the enterprise of street crime control.

Again, there is often overlap in practice between third party policing and the other strategies examined by the committee. The focal point of third party policing can be people, places, or situations. Third party policing efforts are sometimes directed specifically at categories of people—such as young people, gang members, or drug dealers—and at other times at specific places, such as crime hot spots (Mazerolle and Ransley, 2006,

p. 197). For example, police in southern California used regulatory policy to promote responsible management among operators of nuisance motels (Bichler, Schmerler, and Enriquez, 2013). In Oakland, California, police implemented the “Beat Health” third party policing program to abate drug and disorder problems (see Box 2-3). Still other third party policing programs may seek to engage business improvement districts in crime prevention activities, such as coordinating private security services to complement public security (Cook, 2011; Cook and MacDonald, 2011).

Third party policing interventions (and the problem-solving policing approach more generally) could also include strategic partnerships with private security entities. For example, there are now approximately 1,000 private Business Improvement Districts in the United States, funded by special assessments on owners within the boundaries of the district, that supplement public services.

STRATEGIES FOR A PERSON-FOCUSED APPROACH

In the standard model of policing, the primary goal of police was to identify and arrest offenders after crimes had been committed. But beginning in the early 1970s, research evidence began to suggest that the police could be more effective if they focused on a relatively small number of chronic offenders (Pate, Bowers, and Parks, 1976). Similar to the research showing a concentration of crime at certain microgeographic locations, Wolfgang, Figlio, and Sellin (1972) found that a large proportion of crime was committed by a small proportion of offenders: just 6 percent of the juveniles they studied were responsible for 55 percent of juvenile arrests. These findings—the existence of a substantial, identifiable group of chronic offenders—were replicated in a series of other studies of criminal behavior (see, e.g., Farrington and West, 1993; Howell et al., 1995; Blumstein, Farrington, and Moitra, 1985).

These studies led to innovations in policing based on the logic model that crime prevention outcomes could be enhanced by focusing policing efforts on the small number of offenders who account for a large proportion of crime. From this perspective, the standard model of generalized investigation and prevention was deficient because it spread resources too broadly across the general criminal population. Specific deterrence could be gained by focusing on very high rate offenders who are responsible for a large part of the crime problem, and general deterrence would be enhanced by the message that high rate offenders are the focus of concentrated police activities. Such programs, at least in their development, rely not on the social or demographic characteristics of offenders as a method of allocation of police resources but rather on official data about crime.

Focused Deterrence

Offender-focused deterrence, also known as pulling levers, is a strategy that attempts to deter crime among a particular offending population and is often implemented in combination with interventions typical of a problem-solving approach (Braga and Weisburd, 2012). Focused deterrence allows police to increase the certainty, swiftness, and severity of punishment in innovative ways. The first focused deterrence intervention, Operation Ceasefire, was implemented to reduce youth homicide in Boston during the mid-1990s (Braga et al., 2001; Kennedy, Piehl, and Braga, 1996). This well-known program was designed to prevent violence by reaching out directly to gangs, saying explicitly that violence would no longer be tolerated, and backing up that message by “pulling every lever” legally available when violence occurred (Kennedy, 1997). Box 2-4 gives a more detailed description of Operation Ceasefire. Central to the strategy is direct interaction with offenders and communication of clear incentives for compliance and consequences for criminal activity. Most offender-focused deterrence interventions target various criminally active groups and networks, including gangs and drug crews.

Focused deterrence interventions target specific behaviors by the relatively small number of chronic offenders who are viewed as highly vulnerable to criminal justice sanctions. The strategy aims to directly confront offenders—for example, by telling them that continued offending will not be tolerated and informing them how the system will respond if they violate behavior standards. An important aspect of the strategy is often the use of face-to-face meetings with offenders. McGarrell and colleagues (2006) suggested that these types of direct communication, followed up with appropriate enforcement responses to continuing violations, may cause offenders to reassess the risks of committing crimes. It is likely that other complementary crime-control mechanisms are at work in a focused deterrence strategy (see, e.g., Braga, 2012). Focused deterrence typically is incorporated in a hybrid intervention or program with elements of both the problem-solving and community-based approaches. According to Braga and Weisburd (2012, pp. 349–350), “the emphasis is not only on increasing the risk of offending but also on decreasing opportunity structures for violence, deflecting offenders away from crime, increasing the collective efficacy of communities, and increasing the legitimacy of police actions.”

There have also been examples of focused deterrence applied to street drug markets and individual repeat offenders. In High Point, North Carolina, the Drug Market Intervention aimed focused deterrence at eliminating public forms of drug dealing such as street markets and crack houses by warning dealers, buyers, and their families that enforcement was imminent (Kennedy and Wong, 2009; Corsaro et al., 2012; Saunders et

al., 2015). This intervention targeted individual “overt drug markets” and established a joint police-community partnership that identified individual offenders and notified them of the consequences they faced if they continued dealing in drugs. This partnership also provided support services for these individuals through a community-based resource coordinator, while conveying the message that there now was an uncompromising community norm opposed to drug dealing.

In a focused deterrence intervention in Chicago, parolees who had been involved in gun- and gang-related violent crimes and were returning to one of the highly dangerous neighborhoods selected for the intervention were required to attend “offender notification forums.” The forums informed

them that as convicted felons, they were vulnerable to federal firearms laws with stiff mandatory minimum sentences. The forums also offered social services and included talks by community members and ex-offenders (Papachristos, Meares, and Fagan, 2007). The communications process at these forums was intentionally designed to promote positive normative behavior change by engaging the parolees in ways they were likely to view as procedurally just, rather than simply threatening.

Stop, Question, and Frisk

SQF has become one of the most controversial proactive policing strategies because police directly interact with citizens, using intrusive police powers. The legal authority to perform individual SQF is grounded in the landmark 1968 Supreme Court decision Terry v. Ohio. Terry v. Ohio and related decisions have concluded that police may stop a person based upon a “reasonable suspicion” that they are about to commit, are in the process of committing, or have committed a crime.⁹ If a separate “reasonable suspicion” that the person is armed and dangerous exists, the police may conduct a frisk of the stopped individual. Given this standard, although situational factors are also relevant, Terry v. Ohio stops cannot be conducted lawfully without reference to the behavior of the individual being stopped. When carried out as a proactive policing strategy, an SQF program relies upon the legal authority granted in Terry v. Ohio and its progeny to engage in frequent stops in which suspects are questioned about their activities, frisked if possible, and often searched, usually with consent.¹⁰

Stops, frisks, and arrests, whether reactive or proactive, are subject to the same legal standards. Traditionally, stops, frisks, and arrests are tools police use reactively as a means to address a particular crime they witness or have reported to them or to investigate specific suspicious behavior. In this context, harmless or ambiguous conduct often will not justify the resources that would be necessary to address it, and officers leave such conduct unaddressed rather than intrude on individuals. By contrast, as a proactive policing strategy, departments often employ coercion more expansively and to promote forward-looking, preventative ends. This strategic use of Fourth Amendment doctrine is legal (Whren v. United States 517

___________________

⁹ The Supreme Court has not ruled as to whether Terry v. Ohio can be used to investigate a completed misdemeanor, and it has suggested that it might not be permissible. However, Terry can be used as the legal justification for police to investigate a completed felony (United States v. Hensley, 469 U.S. 221 [1985]); see also Navarette v. California (572 U.S. ___ [2014]).

¹⁰ Police may perform a frisk (or pat-down) on an individual if, during a lawful stop, they have reasonable suspicion that the person is armed and dangerous. A frisk is a limited search of the person’s outer clothing for the purpose of discovering weapons (Terry v. Ohio, 392 U.S. 1 [1968]).

U.S. 806 [1996]). See Chapter 3 of this report for a full discussion of the legality of SQF. Nevertheless, in this way, some deterrence-oriented proactive strategies generate incentives for officers to conduct more frequent and intrusive, and therefore liberty-reducing, searches and seizures, aided by the rules developed by the U.S. Supreme Court for reactive policing, than reactive policing would generate.¹¹ Today, police executives regard SQF as an important crime prevention tool (see, e.g., Terkel, 2013).

SQF programs often involve blanketing areas within a city with pedestrian stops to reduce violent crime, as was the case in Philadelphia (see Box 2-5). It is often assumed in these programs that such stops play a key role in deterring potential offenders, as it raises the probability of being stopped and searched by the police. Other cities have used SQF programs in an attempt to change perceived risks of engaging in particular crimes, such as gun and drug crimes.

Although we have categorized SQF as a strategy for a person-focused approach because of the legal requirement that police focus on the behaviors of specific people to undertake a stop, SQF has also been used as a proactive policing tactic aimed at controlling and preventing crime at specific places. For example, Weisburd, Telep, and Lawton (2014) found that SQF in New York City had been implemented as a type of hot spots policing tactic, where SQF stops were concentrated on specific high-crime streets. Kansas City, Pittsburgh, and Indianapolis have used SQF practices to address gun crime at hot spots (see Sherman, Shaw, and Rogan, 1995; Cohen and Ludwig, 2003; McGarrell et al., 2001).

Some scholars argue that the SQF strategy has negative consequences for communities (see Chapter 5 of this report; see also Fagan et al., 2010), and it has been criticized for targeting the young, non-Whites, and specific neighborhoods (see Gelman, Fagan, and Kiss, 2007; Ridgeway, 2007; Stoud, Fine, and Fox, 2011; see also Chapter 7 of this report). In New York City, a court found the SQF program of the New York Police Department (NYPD) to be unconstitutional and restricted the NYPD’s use of the strategy. See Chapter 3 of this report for a discussion of SQF’s legality.

STRATEGIES FOR A COMMUNITY-BASED APPROACH

The community-based approach seeks to enlist and mobilize people who are not police in the prevention of crime and the production of public safety. However, in this approach, the focus is generally not on specific actors such as business owners but the community more generally. While community-based strategies may incorporate practices typical of the other

___________________

¹¹ Proactive strategies that emphasize narrowly focused deterrence are unlikely to have this effect.

proactive policing approaches, such as problem-solving or place-based policing, their key orientation is toward the community. In some cases, community-based strategies rely on enhancing the community’s ability to engage in collective action to do something about crime. This is often referred to as the “collective efficacy” of the community (Sampson and Raudenbush, 1999). In other cases, the strategy seeks to change community evaluations of the legitimacy of police actions (Tyler, 2004). These objectives are often intertwined.

Much of 20th century police reform attempted to assign the police the

core responsibility for crime control (Fogelson, 1977; Kelling and Moore, 1988; Walker, 1977), whereas community policing reformers sought to encourage the clientele of the police to become “coproducers” of crime control, dealing not only with the immediate concerns of a specific incident but also with underlying issues that may aggravate crime problems. As the 2004 National Research Council report on policing stated, “community policing may be seen as [a] reaction to the standard models of policing. . . . While the standard model of policing has relied primarily on the resources of the police and its traditional law enforcement powers, community policing suggests a reliance on a more community-based crime control that draws not only on the resources of the police but also on the resources of the public” (National Research Council, 2004, p. 233).

The impetus for community-based policing strategies came in part from conflicts between the police and the public that emerged in the 1960s and 1970s, especially among non-White and disadvantaged communities (see Chapter 1). The approach’s logic model developed from a growing research base that suggested that the community was key to crime control (Reiss and Tonry, 1986; Skogan, 1992). One early indication that community involvement was important for controlling crime came from a large-scale study conducted in the late 1970s of rapid response to emergency calls to the police (Spelman and Brown, 1981). Although the study overall found that increasing police response time would not lead to crime reductions, the researchers also concluded that citizen willingness to call the police was key to any potential crime prevention gains. Similarly, a series of studies in the 1970s and 1980s pointed to the importance of citizen cooperation in increasing police effectiveness (see, e.g., Reiss, 1971; Spelman and Brown, 1984).

Community-Oriented Policing

At its outset, community policing did not originate as a proactive approach to solving crime problems. In its original formulation, its proponents sought to give greater priority to a wide range of order maintenance and public service functions that had not been given priority during the “professional” reform era (Goldstein, 1987; Greene and Mastrofski, 1988; Kelling and Moore, 1988; Rosenbaum, 1994). While it may be argued that service to citizens was always an important part of American police work (see, e.g., Wilson, 1968), community-based policing legitimated a set of roles for the police that had previously been unrecognized or underappreciated, especially in the way that governments measured police performance (crime, arrest, and clearance rates). In short, community-based policing at the outset did not necessarily define crime reduction, at least in terms of traditional measures, as a central element of its success (see, e.g., Klockars,

1988; Skolnick and Bayley, 1986). However, crime control became a key goal of community policing over time, making it attractive to national, state, and local community leaders sensitive to the high political priority crime control had assumed in the 1980s and 1990s.

As a strategy focusing on a community-based proactive crime prevention approach, community-oriented policing tries to address and mitigate community problems (crime or otherwise) for the future and to build social resilience, collective efficacy, and empowerment to strengthen the infrastructure for the coproduction of safety and crime prevention. These objectives reflect a variety of program theories (variants of the approach’s logic model as stated in Table 2-1) about the crime-prevention mechanism at work in community-oriented policing. For example, with practices such as neighborhood watch or police–citizen patrols, increased guardianship may create a deterrent effect. Guardianship may also be the result of building collective efficacy in neighborhoods, so that citizens feel empowered to apply informal social controls to risky behavior, suspicious incidents, or unsupervised youth. Skogan (1986, 1990) discussed community-oriented policing as playing an important role in reducing fear and thereby lowering the chances of citizen withdrawal and isolation—two factors that, when left unchecked, may lead to further crime and disorder (see also the discussion below of broken windows policing).

Community-oriented policing has been described as both a philosophy of policing and an organizational strategy (National Research Council, 2004; Greene, 2000) in which police agencies embrace a vision of their function that is larger than just reacting to and processing crime (Skogan and Hartnett, 1997). This vision generally entails the inclusion by police agencies of community groups and citizens in coproducing safety, crime prevention, and solutions to local concerns. Despite its longevity as a reform—it dates back more than three decades—there is still considerable variation in how community-oriented policing is defined. Nevertheless, a degree of consensus seems to have formed around treating it as an organizational strategy that embraces three core processes and structures (Skogan, 2006b): (1) citizen involvement in identifying and addressing public safety concerns, (2) the decentralization of decision making to develop responses to locally defined problems, and (3) problem solving. Each of these three elements of community-oriented policing could be implemented independently. What gives problem solving and decentralization a community-oriented policing character is when these elements are embedded in the community engagement (often called “partnership”) element. The inclusion of problem solving as an element again points to the overlap across the committee’s four approaches to proactive policing (refer to Table 2-1).

Early manifestations and research on community-oriented policing focused on tactics such as foot patrol, neighborhood watch, and community

meetings or newsletters. However, as noted above, the definition expanded to include practices from the problem-solving approach. More recently, community-oriented policing has encompassed such notions as building collective efficacy and empowerment (see Sampson, 2011); procedural justice and legitimacy (see Tyler, 1990);¹² and efforts to increase police accountability through citizen review boards, body-worn cameras, and improved complaint processes.

In 2014, the Major Cities Chiefs Association (MCCA), whose membership represents the largest police agencies in North America and the United Kingdom, conducted an electronic survey to better understand the community policing practices of its members. Of the 75 North American member agencies, 42 responded to the survey. Table 2-3 shows the number and percentage of departments who reported that they engaged in specific community-oriented policing practices. Some of these practices fall under the committee’s definition of proactive policing, but others do not. It should be noted that this list is not exhaustive of the sorts of tactics and activities that have been characterized as “community-oriented policing” (see Roth, Roehl, and Johnson, 2004).

Departments define and deploy what this committee means by a strategy of community-oriented policing in different ways; some view it as the responsibility of a special community-policing unit, while others view it as an organizational philosophy. Many agencies do both. (See Box 2-6 for a description of community-oriented policing in Chicago.) According to the 2014 MCCA survey, responding departments allocated personnel to perform community-oriented tasks using centralized, decentralized, or hybrid structures. Seventeen percent of agencies (n = 7) used a centralized structure where only full-time community policing officers were deployed to conduct community-oriented policing activities; 21 percent (n = 9) used a decentralized structure, which considered community-oriented policing exclusively a part of patrol officer duties; and the majority of respondents (62%; n = 26) used a hybrid structure with a combination of dedicated full-time staff, patrol officers, and special units engaging in activities aimed at community-oriented policing objectives (Scrivner and Stephens, 2015, p. 9).

Procedural Justice Policing

A more recent organizational innovation with a focus on the community-based approach is procedural justice policing. Like community-oriented policing, this strategy also assumes that the police cannot succeed in their efforts to control crime without the support of the public. However, in its efforts to change the public’s relationship with the police, procedural

___________________

¹² Procedural justice and legitimacy are discussed in the next section.

TABLE 2-3 Prevalence of Use of Community-Policing Practices by North American Police Agencies Responding to the 2014 MCCA Survey

NOTE: MCAA = Major Cities Chiefs Assocation, POP = problem-oriented policing.

SOURCE: Scrivner and Stephens (2015, p. 10).

justice policing focuses on how the police treat the public as individuals in everyday encounters.¹³ Whereas community-oriented policing often focuses on giving the community the outcomes that it wants (e.g., more safety, more noncrime services, greater responsiveness to personal needs), procedural justice focuses on giving citizens police decision processes that manifest demonstrations of police fairness and regard for a person’s dignity. Fair and considerate police processes are presumed to render even unpleasant outcomes (an arrest or citation) less objectionable to the person on the receiving end.

Also, unlike community-oriented policing, procedural justice policing does not seek to enlist the public in coproductive activities during these routine encounters but rather seeks to impress upon the citizen and the

___________________

¹³ Conceivably, this strategy could include many other occasions when police and public interact, such as neighborhood association meetings attended by the police.

wider community that the police exercise their authority in legitimate ways.¹⁴ According to the logic model of this strategy, when citizens feel that policing is legitimate, they are more inclined not only to defer to police

___________________

¹⁴ Legitimacy, within the context of the procedural justice literature, is “a property of an authority or institution that leads people to feel that authority or institution is entitled to be deferred to and obeyed” (Sunshine and Tyler, 2003, p. 514).

authority in the present instance but also to collaborate with police in the future, even to the extent of being more inclined not to violate the law. That is, procedural justice policing is based upon the idea that police shape the evaluative judgments citizens make about police performance (i.e., whether it is effective, fair, lawful), and these evaluations shape general orientations toward the police (i.e., police legitimacy). Legitimacy then shapes the behavior of citizens in terms of law abidingness, cooperation with authorities, and engagement in the community (see Chapter 5 of this report; see also Tyler, 2003). Therefore, when police engage in activities that promote procedural justice, they are presumed to enhance their perceived legitimacy not only among those who experience police contacts directly but also from a broader communication of their actions to the community more generally.

Procedural justice policing tries to encourage four main characteristics of police behavior that are viewed as affecting perceptions of police legitimacy: (1) Do they provide opportunities for voice, allowing members of the public to state their perspective or tell their side of the story before decisions are made? (2) Do they make decisions in ways that people regard as neutral, rule-based, consistent, and absent of bias? (3) Do they treat people with the dignity, courtesy, and respect that they deserve as human beings and as members of the community? (4) Do people believe that their motives are trustworthy and benevolent—that is, that the police are sincerely trying to do what is good for the people in the community?

The key to understanding the procedural justice strategy is that its elements focus on how people experience policing: whether they feel they have voice, whether they think the procedures are neutral, whether they feel respected, and whether they infer that the police are trustworthy. Trustworthiness is the key to accepting discretionary decisions, according to this logic model. The argument underlying the strategy is that the way people perceive these features of police action shapes whether people do or do not judge the police to be legitimate.

In deciding to include procedural justice as a proactive policing strategy, the committee recognized that many of the behaviors connected to procedural justice may also more generally be seen as a standard part of democratic policing (Nagin and Telep, 2017). Although the committee agrees with this position, it recognizes that procedural justice policing has been presented by its advocates not only as “good police behavior” but also as a strategic approach to policing that should increase police legitimacy, citizen compliance, order, and safety in police-public encounters and should reduce crime in the long run (see, e.g., Tyler, Goff, and MacCoun, 2015).

While it may be true that treating people with respect and fairness can be seen as part of overall good practice, in procedural justice policing

the police modify their actions to consciously and deliberately mold the attitudes of the community in advance of events that might create conflict or crisis. Under the logic model informing this strategy, the police are instructed to do this in order to proactively influence what happens later. For example, the strategy may aim to create a climate in which the public is more willing to defer to police authority, in which people more willingly obey the law and help to solve crimes, and in which the public accepts that the police are acting with good intentions and should be given the benefit of the doubt in ambiguous situations. Sometimes such activities may come as a response to citizen calls for police service, and in this sense they may be seen as reactive. But they also may occur in community meetings or with other proactive contacts with the police. Moreover, even in responses to citizen requests for police service, under procedural justice policing police officers should seek to apply procedural justice principles not only to initiators of police responses but also to bystanders and offenders.

In this sense, the police act proactively by engaging in many types of actions designed to build a “reservoir of trust” in the community. Whatever the fit of procedural justice with democratic principles, procedural justice policing seeks to develop longer-term gains in terms of police legitimacy and crime. Advocates of this approach argue that an overarching focus on the principles of procedural justice is key to prevention and other outcomes. Indeed, it is sometimes presented as an alternative to other proactive policing strategies that focus on short-term crime-prevention gains:

Until recently, procedural justice practices were not explicit objectives of particular policies and programs but rather were simply observed in their “natural” state as the product of discretionary choices made by individual police officers in specific police–public interactions (Mastrofski et al., 2016). The report of the President’s Task Force on 21st Century Policing (2015) argued that procedural justice is an important aspect of building trust and legitimacy in communities, and therefore the task force called on departments to adopt procedural justice as a guiding principle. This call

has led to the development of a larger number of programs and policing interventions that explicitly promote procedural justice.¹⁵

Box 2-7 describes how the King County, Washington, Sheriff’s Office has implemented principles of procedural justice in its work.

Broken Windows Policing

Another strategy of the community-based approach uses a very different logic model for the problem of crime control. Broken windows policing sees the key to crime prevention as operating in the informal social controls within communities (Wilson and Kelling, 1982). Its focus, accordingly, is on how the police can reinforce and enhance such social controls, especially where informal social controls have become weak (see Weisburd et al.,

___________________

¹⁵ Procedural justice policing can follow both an internal and external model. Internal procedural justice refers to practices within an agency and the relationships officers have with their colleagues and leaders. It follows the logic model that those officers who feel respected by their organization are more likely to bring this respect into their interactions with the public. External procedural justice focuses on the ways officers interact with the public and how the characteristics of those interactions shape the public’s trust of the police (Tyler, 1990; Sunshine and Tyler, 2003; Haas et al., 2015).

2015). It shares with community-oriented and procedural justice policing a concern for community welfare and envisions a role for police in finding ways to strengthen community structures and processes that provide a degree of immunity from disorder and crime in neighborhoods. Unlike community-oriented policing, this strategy does not emphasize the coproductive collaborations of police and community as a mode of intervention; rather it focuses on what police should do to establish conditions that allow “natural” community entities to flourish and promote neighborhood order and social/economic vitality.

The concept of broken windows policing developed out of a Police Foundation study, the Newark Foot Patrol experiment (Police Foundation, 1981). The police officers walking patrol in the study were engaged in activities (e.g., closing down a bar early after being called twice to end brawls in that same bar) that might be seen as part of the policing task in the standard model of policing. However, from this study’s results and drawing on earlier studies by Zimbardo (1969) and Zimbardo and Ebbesen (1969), Wilson and Kelling (1982) identified a link between social disorder and crime and suggested that the police ought to pay attention to many problems that may be seen as peripheral to the police function under the standard model. The broken windows hypothesis held that “untended” behavior (e.g., abandoned property, unruly youth behavior) could lead to the breakdown of community controls and that serious crime developed because the police and citizens did not work together to prevent urban decay and social disorder (Weisburd and Braga, 2006b, pp. 14–15).

According to Wilson and Kelling (1982, p. 31), “at the community level, disorder and crime are usually inextricably linked, in a kind of developmental sequence.” The broken windows logic model posits an indirect pathway from disorder to crime through increased fear and, subsequently, the breakdown of informal social controls in the community. The fear-of-crime literature at the time provided fairly consistent support for a strong linkage between disorder and fear (Garofalo, 1981; Garofalo and Laub, 1978; Hunter, 1978; see Hinkle, 2013 for a review), and early studies in this area can in some sense be seen as supporting the broken windows logic model.

Diminished informal or community social controls are thus a key component of the logic model underlying the broken windows concept of crime control. Wilson and Kelling (1982) argued that disorder problems, and the resulting increased levels of fear, lead to withdrawal from the community. This withdrawal takes two forms: people moving away and the remaining residents becoming less likely to intervene in community affairs. ‘‘Untended’’ behavior also leads to the breakdown of community controls (Weisburd et al., 2015):

The nature of police “broken windows” interventions varies from informal enforcement tactics (warnings, rousting disorderly people) to formal or more intrusive ones (arrests, citations, SQF), all intended either to disrupt the forces of disorder before they overwhelm a neighborhood’s capacity for order maintenance or to restore afflicted neighborhoods to a level where community sources of order can now sustain it. The two most commonly implemented (separately or in combination) forms of broken windows policing have been the use of aggressive policing that uses misdemeanor arrests to disrupt disorderly social behavior to prevent crime (often referred to as “zero tolerance”¹⁶) (see Taylor, 2006; Cordner, 1998; Eck and Maguire, 2006; Skogan, 2006b; Skogan et al., 1999) and the use of problem-oriented or community-oriented policing methods to address disorderly conditions that might contribute to crime (see Kelling and Coles, 1996). Box 2-8 describes a zero tolerance version of broken windows policing that was implemented in New York City.

A broken windows strategy may also be used in conjunction with other proactive policing strategies. For example, in Jersey City, New Jersey, officers used aggressive order maintenance as a tactic to reduce violent crime at hot spots (Braga et al., 1999). Similarly, a broken windows strategy was used in Los Angeles, through the Los Angeles Police Department’s Safer Cities Initiative, to target homeless encampments in the downtown “skid row” area that were believed to be linked to high rates of street crime and disorder. The tactics implemented for this initiative included breaking up encampments, issuing citations, making arrests, and maintaining a visible police presence in the area (Berk and MacDonald, 2010).

___________________

¹⁶ However, Kelling and Sosa (2001) argued that the term “zero tolerance” is often used derisively to describe broken windows policing interventions in which officers consistently use discretion and routinely assess the circumstances surrounding offenses, and therefore, they argue, the use of the term may be inaccurate for these interventions.

THE DIFFUSION OF PROACTIVE POLICING ACROSS AMERICAN CITIES

To what extent have these four proactive policing approaches spread across the landscape of American policing? To answer that question, the committee drew on data collected from the National Police Research Platform (NPRP), PERF, and MCCA. The PERF and MCCA surveys have already been described earlier in this chapter.

Overall, it is clear that many departments claim to be using multiple proactive policing innovations. The NPRP, the most comprehensive and representative survey gathering this information, uses a diverse national sample of approximately 100 municipal police and sheriff’s agencies, of which the majority are agencies that have between 100 and 3,000 sworn officers. Between October and December 2013, the NPRP conducted a survey of its participating agencies, asking knowledgeable persons within the organization to indicate whether specific innovations had been adopted, whether department policy regarding an adopted innovation had been established

and, if so, in what year.¹⁷ Seventy-six of the 100 police agencies completed the questionnaire.¹⁸ Interestingly, the survey results suggest that there is very wide use of proactive policing in medium-to-large police agencies in the United States. Mastrofski and Fridell (n.d., p. 3) reported that three-quarters of the responding departments adopted at least 8 to 10 “innovations.”¹⁹Table 2-4 lists the findings relevant to proactive policing.

The most commonly employed proactive policing innovation according to this survey was community-oriented policing, which more than 90 percent of agencies claim to be employing, supported by formal policy. Using the taxonomy adopted for this report, 9 of 10 local law enforcement agencies with more than 100 sworn officers reported in 2013 that they had adopted community-oriented policing with supporting formal policies (Mastrofski and Fridell, n.d.). Community-oriented policing became popular among police leaders in the 1990s (Roth, Roehl, and Johnson, 2004)

___________________

¹⁷ The median number of sworn officers for the entire NPRP was 274; the median for the 2013 department-characteristics survey was 255.

¹⁸ Although 24 NPRP agencies did not complete the survey, the profile of survey respondents did not differ markedly from the total NPRP sample (Mastrofski and Fridell, n.d., p. 1).

¹⁹ The 2013 NPRP survey designated the following as innovations: evidence-based policing, video recording (CCTV), CompStat, broken windows policing, early intervention systems, problem-oriented policing, procedural justice policing, hot spots policing, crime analysis, and community policing (Mastrofski and Fridell, n.d., p. 3).

and was especially attractive because of the availability of federal grants, issued by the COPS Office, to support community-oriented policing programs (Reisig, 2010, p. 20). The popularity of this strategy has seemingly been sustained despite declining funding in the latter part of the 2000s.

Perhaps surprising, given the relatively later emergence of procedural justice policing on the American police reform agenda, an almost equal number of departments (89.5%) claim to have implemented practices for this strategy in their department. While the depth of involvement and commitment to these strategies cannot be gauged by the surveys, the data suggest that police agencies across the United States are concerned about police legitimacy (as defined in the procedural justice logic model) and view community-based policing interventions as key to their work.

Ninety-one percent of departments surveyed claimed to use hot spots policing, again pointing to very high penetration of this strategy in American policing. Problem-oriented policing was also widely noted, with about 82 percent of responding NPRP departments claiming to use this strategy. Use of broken windows policing was claimed by 79 percent of NPRP respondents.

PERF conducted the Future of Policing Survey in 2012. The survey instrument was distributed to 500 police departments across the country, and nearly 200 police departments responded. While the PERF Survey was directed at its membership, which generally consists of larger and more progressive police agencies, the results provide a picture of the use of proactive policing strategies similar to the NPRP results (see Table 2-5). In this case, community-oriented policing, problem-oriented policing, and directed patrols/focused deterrence were the strategies most commonly used. Targeting known offenders and hot spots policing were also common, with almost

TABLE 2-5 Prevalence of Use of Proactive Policing Strategies by Percentage of Agencies Responding to the 2012 Future of Policing Survey (n = 200)

SOURCE: Police Executive Research Forum (2014, p. 50).

80 percent of departments claiming to use these strategies. Not surprisingly, predictive policing, which is a newer innovation, was less commonly employed.²⁰ Although the agencies affiliated with PERF do not constitute a representative sample of all U.S. police agencies or of any subset thereof (e.g., large agencies), they may serve as a good indicator of likely trends in the use of strategies among larger police agencies (see Koper, 2014, p. 126).

The largest and most representative of the surveys to provide information on proactive policing is the Law Enforcement Management and Administrative Statistics (LEMAS) Survey, administered by the Bureau of Justice Statistics (BJS). BJS collects data from a representative sample of local police departments and provides national estimates on a variety of agency characteristics. The survey is completed every 3 years. Table 2-6 displays the 2007 survey findings on hot spots and community policing,

___________________

²⁰ In addition, to date, most departments implementing predictive policing must first purchase predictive policing software. This upfront cost may slow adoption of the strategy.

TABLE 2-7 Police Departments in 2013 with Community-Policing Mission Components

SOURCE: Reaves (2015).

and Table 2-7 presents the 2013 survey findings on community policing.²¹ One advantage of the LEMAS survey is that it allows one to look at the variability in policing strategies by department size—and the overall story that emerges from these findings is that the claimed use of proactive strategies declines as the size of departments declines.

The prevalence of SQF is not examined by any of the above surveys, possibly because few departments created formal policies or structures to implement it, or possibly because of the controversy surrounding use of this strategy.²² However, one relevant survey data source, the 2011 BJS Police-Public Contact Survey, found that of the 62.9 million people ages 16 and older with one or more police contacts in 2011, 7.3 percent (4.59 million) reported the contact was an involuntary street stop or arrest or other involuntary contact (not an involuntary traffic stop).²³ Among those individuals reporting an involuntary contact, 19.1 percent (72,083 indi-

___________________

²¹ BJS uses the term “community policing,” which corresponds with the committee’s use of the term “community-oriented policing,” as both emphasize collaboration with communities, support through agency management structures, and problem solving (see Reaves, 2010, p. 26).

²² Though SQF is not a formal strategy in most police departments, it is used by all police departments in response to reasonable-suspicion observations or calls for service.

²³ The Police-Public Contact Survey does not identify the police department with which the person interacted.

viduals) reported being searched or frisked (Langton and Durose, 2013, pp. 2, 11–12). Between 2003 and 2010, reported SQF stops in New York increased almost four-fold from 160,851 to about 600,000 (Weisburd, Telep, and Lawton, 2014). At its peak in 2011, the NYPD reported 685,000 SQFs (for a population of 8.5 million).²⁴ Philadelphia and Los Angeles also saw substantial increases in pedestrian stops made by the police in the first decade of the 21st century. In Philadelphia, police reported 250,000 stops (in a city of 1.5 million) in 2009, double the number in 2007. Los Angeles reported 244,038 stops (in a city of 3.85 million) in 2008, double the number of stops in 2002 (Jones-Brown, Stoudt, and Moran, 2013).

These data tell us that many of the proactive policing approaches described in this chapter are not isolated programs used by a select group of agencies but rather a set of strategies that have been diffused across the landscape of American policing. Although the surveys are informative and present a general picture of American policing, especially among large departments, they do not offer a complete picture. For example, it is not known with certainty what motivates police organizations to embrace these innovations. One hypothesis is that these adoptions are motivated by “technical” concerns, such as a desire to reduce crime and to create and maintain safe communities (Mastrofski and Uchida, 1993; National Research Council, 2004, pp. 308–312). Police departments may also be motivated by federal funding incentives or in response to federal litigation (see Chapter 3 for a discussion of the U.S. Department of Justice’s litigation strategy).²⁵ Still another perspective, sometimes termed “institutional theory,” suggests that the motivation is the pursuit of legitimacy among one’s peers and support from an organization’s stakeholders. According to this hypothesis, police leaders may adopt strategies in the absence of hard evidence that they work in a technical sense (or even in the face of evidence that they do not work) simply because they perceive that their peers, especially high-visibility leaders in the field, are touting those strategies or using them. And this motivation, according to institutional theory advocates, can account for the rapid diffusion of certain police innovations in the past few decades (Weisburd et al., 2003; Willis, Mastrofski, and Weisburd, 2007). Where institutional pressures are strong for adoption, there can be a tendency to garner the benefits of “being on board” with the innovation without having fully implemented it.

___________________

²⁴ That figure declined to 191,851 SQF incidents in 2013, and further declined to 22,565 SQF stops in 2015, as a result of court challenges and a changing political environment. See http://www.nyclu.org/content/stop-and-frisk-data [May 2017]. Chapter 3 of this report discusses Floyd v. City of New York (2013).

²⁵ As described above, for example, community-oriented policing was an especially attractive innovation for police departments because of the availability of federal grants, issued by the COPS Office, to support community-oriented policing programs (Reisig, 2010, p. 20).

These surveys of police agencies also do not collect information relevant for determining with confidence the fidelity with which each strategy was implemented, how frequently it is actually used, or the scope of its use (how many people use it) within the department (Maguire and Mastrofski, 2000). Moreover, it is unclear whether departments consistently report their practices across surveys. Systematic data are lacking on how many resources (e.g., staffing levels) are devoted to each proactive strategy; also lacking are systematic data on how they are targeted.

Further complicating researchers’ ability to estimate the prevalence of proactive policing approaches is that the standard way to calculate staffing levels for a given proactive strategy is to tally the number of officers assigned to a unit charged with that strategy. However, the problem with this estimation is that usually the officers assigned to engage in that proactive strategy are also charged with engaging in many other activities, and agency records do not readily distinguish proactive-program efforts from other efforts. For example, officers assigned to specialist community-based policing or problem-solving units also may have responsibilities for responding to calls for service (reactive policing). And officers whose basic job assignment is traditional reactive patrol in the same neighborhood may also take opportunities during their discretionary time to engage in community policing and problem solving.

Even thoroughly researched proactive projects (e.g., Skogan, 2006b, pp. 59–64) do not provide much information on the “dosage” of staff time and activities (Mastrofski and Willis, 2010, p. 83). In the few instances where detailed time-management studies have been executed via systematic observation by researchers, the finding is that although community-policing specialists spent more time on community-policing and problem-solving activities than generalist patrol officers in the same department, the norm for community-policing specialists remained the traditional, reactive encounter (Parks et al., 1999; Smith, Novak, and Frank, 2001). Surveys that simply gather department staffers’ general impressions of how much officers engage in a given strategy are vague or even misleading. New methods are therefore necessary to determine the prevalence, scope, and frequency of the use of various policing innovations throughout law enforcement agencies in the United States. We return to this issue in our concluding chapter, where we discuss the committee’s recommendations for new data collection on proactive policing.

CONCLUSION

Each of the four approaches to proactive policing identified by the committee is derived from a different logic model, each focusing on a different method for preventing crime and disorder. A place-based approach seeks to

capitalize on empirical findings about the concentration of crime in small microgeographies. A problem-solving approach assumes that when the police focus on solving specific problems, rather than applying broadly defined generalized strategies, greater crime-prevention gains will be achieved. In a person-focused approach, empirical data on the concentration of crimes among a small part of the criminal population form the key element of the logic model. And finally, with a community-based approach, the importance of the community in solving crime problems is the primary logic model of prevention. In practice, these approaches often entail overlapping police strategies and programs in the field, an issue that we will turn to in later chapters, as the committee assesses the impacts of proactive policing that are more difficult to isolate and examine. One conclusion that can be drawn from reviewing these approaches is that they are, overall, used widely in American policing. The widespread use of proactive policing practices makes careful assessment of their consequences for crime, communities, legality, and bias and discrimination particularly important.