Towards a framework for preventing community violence among youth

Abstract

This article, in an effort to assist the selection and deployment of evidence-informed strategies, proposes a new conceptual framework for responding to community violence among youth. First, the phenomenon of community violence is understood in context using a new violence typology organized along a continuum. Second, the need for a new anti-community violence framework is established. Third, a framework is developed, blending concepts from the fields of public safety and public health. Fourth, evidence from systematic reviews and meta-analyses concerning community violence is summarized and categorized. Finally, an anti-violence framework populated with evidence-informed strategies is presented and discussed.

Keywords:

• Violence prevention
• youth violence
• community violence
• evidence-based
• evidence-informed

Introduction

In 2012 alone, 95,000 children and young people under the age of 20 were murdered, constituting almost 20% of all homicides (UNICEF, 2014). 90% of these young victims lived in low- and middle-income countries, and especially in Latin America and the Caribbean where the highest homicide rates for children and adolescents are found. Boys are at particular risk for homicide, accounting for 70% of victims less than 20 years of age, particularly during late adolescence (ages 15–19), when 57% of all violent deaths among young victims occur. Violent death at any age is brutal and tragic, but particularly so for children, as their innocence and powerlessness belies personal culpability and heightens our sense of unfulfilled potential and lost opportunity.

This article primarily concerns the violence, particularly lethal violence, which occurs among youth under the age of 20 and in community settings. While young people are the main focus, this content is also relevant for those in young adulthood and beyond. This article draws heavily from a comprehensive study of community violence reduction strategies recently completed by the author and Christopher Winship, the Diker-Tishman Professor of Sociology at Harvard University, “What Works in Reducing Community Violence: A Meta-Review and Field Study for the Northern Triangle” (Abt & Winship, 2016). That study’s meta-review synthesized the results of 43 systematic reviews and meta-analyses of violence reduction programs, aggregating over 1400 individual programmatic evaluations in order to identify ‘what works’ in reducing community violence.

Systematic reviews use predetermined methods to identify, select, appraise, and combine the results from individual studies in a clear, unbiased, and systematic manner. Meta-analyses combine quantitative data from individual studies using established statistical techniques. Both are designed to overcome limits to the external validity of individual studies (Killias & Villetaz, 2008). These techniques have grown popular as scholars recognize their usefulness for identifying themes and patterns across large numbers of empirical studies (Makarios & Pratt, 2012), and are being employed with increasing frequency in medicine, education, criminal justice, and other fields (Aos, Miller, & Drake, 2006).

Only causal evidence, i.e. evidence identifying a causal relationship between an intervention and its intended effect, was considered in the underlying report. The same is true here. While an emphasis on rigorous evidence, i.e. causal evidence drawn from high quality experimental or quasi-experimental designs, has advantages in terms of accuracy, reliability, and transparency, there are at least three significant limitations associated with such an approach. First, it is important to remain mindful that public policy cannot be based exclusively on science. Evidence and data should be used to improve public policy decisions, not replace them entirely (Robinson & Abt, 2016), which is why the phrase ‘evidence-informed’ will be used here instead of the more recognized term ‘evidence-based’. Second, the vast majority of rigorous research has been conducted in high-income settings, so conclusions drawn from such studies may be of limited external validity when applied in low- or middle-income settings. Third, focusing on causal evidence creates a bias in favor of the programmatic interventions capable of generating such evidence. Evaluating institutions or systems is a much more complicated exercise where establishing causality may be difficult or even impossible. Nevertheless, a sound understanding of the causal evidence, albeit mostly from high-income counties and programmatic evaluations, can enhance understanding and improve decision-making with regard to community violence prevention.

Community violence, defined and described

The term ‘community violence’ is best understood in context alongside other forms of violence. While the World Health Organization (WHO) (Krug, Mercy, Dahlberg, & Zwi, 2002), United Nations (UNODC, 2013), and others divide violence into discrete categories, these schemas generally fail to conform to actual violence as witnessed or experienced on the ground. In reality, violence comes in many forms, all containing similarities and dissimilarities, each independent from and dependent on the others. To better capture this complexity and provide meaningful guidance to practitioners in field, violence may be better understood according to a set of six attributes described along an admittedly imperfect continuum.

First, violence varies in its lethality or capacity to cause serious physical injury – a shove versus a fatal shooting. Second, it occurs in different settings – in the privacy of one’s home or on a public street. Third, the number of individuals involved may be few, as with a dispute between neighbors, or many, as with conflicts among gangs. Fourth, violence may be as spontaneous as a bar brawl or as methodically planned as an assassination. Fifth, it may be expressive of emotions, including anger, or instrumental in its aim of achieving a particular goal. Sixth and finally, incidents of violence may occur as frequently as domestic disputes or as rarely as formally declared wars between states.

These six attributes are strongly, but not perfectly, associated with each other. To capture these associations we collapse them into a single dimension along a continuum (Figure 1). Obviously, this continuum is neither entirely complete nor perfectly accurate, but viewing violence along a continuum is a helpful means of understanding the relationships between different forms of violence while avoiding overly simplistic categories.

Figure 1. Typology of violence continuum.

The typology in Figure 1 describes six forms of violence: violence between family members and/or intimate partners in the home; violence involving students at school; violence between and among community members; violence committed by gang members; violence committed by organized criminal groups; and violence between nation states, i.e. war. These categories are intended to be illustrative, not exhaustive, and do not single out specific populations, i.e. men, women, children, or the elderly. Community violence is emphasized because it is the focus of this paper. Organized and state violence are de-emphasized here because they are less likely to involve children or young people directly, although there are notable exceptions, e.g. child assassins and soldiers.

At one end of the continuum, violence is interpersonal, i.e. generally occurring between individuals known to one other. Individual incidents occur frequently but are rarely lethal or cause permanent physical injury. It is unplanned, disorganized, emotional, and impulsive in nature. This violence is traditionally viewed a private matter, occurring between family members, intimate partners, schoolmates, or friends. If addressed by public institutions, it will likely involve a wide array of public health stakeholders with limited law enforcement participation, if any. Bullying is one example of violence at this end of the continuum.

At the opposite end, violence occurs between groups, often large in size, where individuals are generally not known to one another. Unlike bullying, this violence occurs infrequently but is highly lethal, often resulting in significant numbers of casualties. It is planned, organized, and instrumental. This violence is a generally state matter and traditionally the province of law enforcement and military institutions. Formally declared conflicts between states exemplify the violence at this end of the continuum.

In the middle of this continuum lies community violence, the focus of this article. Community violence, particularly homicide, occurs primarily in public settings. It is interpersonal, i.e. taking place between individuals and small groups that may or may not know one another. It is loosely planned at best and generally impulsive in nature. That said, the impact of community violence is nevertheless severe, often resulting in death or disabling injury. Its perpetrators and victims are usually, but not exclusively, young men and boys from disadvantaged backgrounds and communities. Community violence may result from disputes or from conventional forms of street crime, e.g. robberies, and implicates both the public health and public safety fields as well as multi-disciplinary, multi-sector responses.

As noted previously, all forms of violence are interconnected. The contagion between different forms of violence is an important subject worthy of serious exploration but lies beyond the scope of this article.

Establishing the need for a community violence framework

While it is clear that all violence is interrelated, it is equally clear that the differing characteristics of various forms of violence necessitate differing approaches – there is no universal strategy for violence prevention, nor should there be. For instance, a key component of any strategy is deciding the number and type of partners to be mobilized. In this regard, responses to violence will vary greatly – a response to bullying may involve coalitions of educators and parents, while addressing organized criminal violence typically demands the coordination of law enforcement groups. Community violence is perhaps unique in the breadth of stakeholders who may contribute to an effective response, including children and parents; community, business, and faith-based leaders; social service and health providers; along with law enforcement and criminal justice agencies.

This broad range of partners is appropriate given that community violence is a pervasive, persistent, and complex socioeconomic phenomenon. Understanding it requires a multi-disciplinary approach. Addressing it demands a multi-sector response. In order to properly organize any collective response, a framework is necessary to coordinate the activities of individual components so that they help rather than hinder one another in pursuit of a common goal: diminished community violence. According to the National Academy of Sciences, ‘As the global community recognizes the connection between violence and failure to achieve health and development goals, a resource such as an evidence-based framework could more effectively inform policies and funding priorities locally, nationally, and globally’ (Carroll, Perez, & Taylor, 2014).

In order to be useful, a framework must be theoretically sound but also grounded in the empirical reality of the problem it seeks to address. It must also have practical utility for implementation in the field. As criminologist Lawrence Sherman has noted, ‘Crime should be classified in whatever way supports crime prevention’ (Sherman, 2012). In short, an effective framework for responding to community violence must clearly articulate a reasonably complete, accurate, and useful description of both the problem and its solution.

Blending public safety and public health: a proposed framework

While many fields have made contributions to the study and practice of community violence prevention, public safety and public health outpace the others by a significant margin. A framework for preventing community violence should therefore look to these disciplines first for organizing insights and principles. There are many differences between them, and the two fields occasionally compete for attention and resources, but fortunately public safety and public health professionals have become increasingly collaborative, drawn together by their shared interest in promoting peace and reducing violence. In fact, the future success of violence prevention depends in part on the continued strength of this critical partnership.

In describing the challenge posed by community violence, a framework can begin by drawing from a number of complementary criminological perspectives, most fundamentally rational choice and routine activities theory. Rational choice theory posits that criminals are self-interested rational decision-makers who weigh the costs and benefits of their conduct, albeit imperfectly, before acting (Cornish & Clarke, 2008). Routine activities theory holds that crime occurs when likely offenders meet suitable targets in the absence of capable guardians (Figure 2) (Cohen & Felson, 1979), and where the likelihood of such convergence is considered a function of the routine activity patterns of all concerned.

Figure 2. Routine activities theory.

Next, the routine activities framework is amended to reflect that community violence occurs at the confluence of many factors, perhaps best summarized using the well-known journalistic and investigative trope of the five Ws and one H, i.e. who, what, when, where, why, and how. If community violence is ‘what happened’, then such violence is a function of place (‘where it happened’), time (‘when it happened’), people (‘who was involved’), the motivations of those involved (‘why it happened’), and behavior (‘how it happened’), all of which can be crudely formulized as follows and as visualized in Figure 3:

Figure 3. Community violence theory, full version.

With the additional elements in place, this expanded formula can account for any number of criminological theories concerning the causes of crime and violence. Biological, psychological, and developmental theories influence people, or p_e. Social, cultural, and environmental theories impact either the social or physical environment, i.e. place or p_l. All of these theories affect motivations and behavior, i.e. m and b. Lastly, this framework can be consolidated by merging time into place and motivation into people (Figure 4). Community violence, in this simplified equation, becomes a function of places, people, and behaviors:

Figure 4. Community violence theory, simplified version.

Critically, understanding community violence in this manner reflects how such violence actually behaves in the real world. One of the most powerful criminological findings from the past two decades is that community violence is sticky, clustering tightly in specific places, among specific people, and around specific behaviors. In Boston, 1% of youth aged 15–24 were responsible for over 50% of city-wide shootings, and 70% of total shootings over a three decade period were concentrated in ‘hot spots’ covering approximately 5% of the city’s geography (Braga & Winship, 2015). In five Latin American cities, 50% of homicides occur in 1.59% of blocks (CAF, 2014). In most metropolitan areas, .5% of the population is responsible for 75% of the homicides (Muggah, 2015). Given that all the ‘available empirical and theoretical evidence suggests that crime is concentrated at a small number of high-risk places during high-risk times and generated by a small number of very risky people’ (Braga, 2015), the evidentiary foundation for this conclusion is strong enough to warrant organizing policies around it. In addition to places and people, community violence also concentrates around high-risk behaviors, including (1) carrying a firearm (typically illegally) (Koper & Mayo-Wilson, 2006), (2) being under the influence of alcohol (WHO, 2010), and (3) belonging to a gang or otherwise violent group of individuals (Decker, 1996).

In addition, understanding community violence in terms of places, people, and behaviors is more easily grasped and readily implemented by practitioners than other conceptual frameworks. Hot spot and problem-oriented policing strategies, for instance, have been disseminated around the globe, familiarizing many law enforcement agencies with at the least a rudimentary understanding of place, people, and behavior-based strategies.

To capture potential responses to the challenges posed by community violence, a framework can look to the public health field, which generally organizes anti-violence efforts into primary, secondary, and tertiary prevention as shown in Figure 5. Primary prevention addresses risk factors associated with violence in the general population. Secondary prevention focuses on sub-populations with risk factors for future violence either as victims or perpetrators. Tertiary prevention attempts to intervene with those already engaged in violent behavior.

Figure 5. Public health model, original version.

This model has many advantages. First, it classifies efforts by risk level with the understanding that as risk levels increase, fewer individuals are implicated. Second, it emphasizes prevention, a crucial component of a collective anti-violence response that has been traditionally underappreciated and underutilized. Third, the model is familiar to most public health practitioners and many others in the field, making it accessible and easy to use.

The public health model also has a number of disadvantages. First and foremost, it ignores law enforcement, the traditional institution charged with responding to crime and violence. This alone renders the model incomplete. Law enforcement is an essential partner in any community violence prevention strategy, and violence prevention efforts will be inhibited if police and prosecutors view their role as purely reactive. Violence prevention should be viewed and defined broadly in order to include law enforcement efforts to stop violence before it begins. Secondly, the public health model has yet to provide a clear explanation to practitioners of how tertiary prevention operates in the context of violence prevention. Under tertiary prevention, the model tends to conflate prevention and rehabilitation – two practices that are generally conducted separately in the field. Third, in practice the model has generally placed an exaggerated emphasis on primary prevention, an important but not necessarily dominant element of an evidence-informed, multi-sector response for preventing community violence.

For these reasons, the public health model should be modified when used in relation to community violence (Figure 6). First, suppression should be added to account for the role of law enforcement. Suppression prevents violence via deterrence and incapacitation, generally but not exclusively through threats of arrest and incarceration. Second, tertiary prevention and rehabilitation should be separated into discrete categories. While they both concern those engaged in violence, tertiary prevention focuses on those currently in residing the community, whereas rehabilitation generally assists former offenders reentering society after they have been adjudicated and/or imprisoned.

Figure 6. Public health model, revised version.

Having provided a conceptual framework for understanding both the problem of and solutions to community violence, we now combine our criminology-based theory of community violence with our modified public health model, mapping places, people, and behaviors against prevention (primary, secondary, and tertiary), suppression, and rehabilitation, creating a grid with a total of fifteen sections (Table 1). This framework is theoretically sound, reasonably complete, informed by the evidence, and implementable in practice.

Table 1. Anti-community violence framework, empty.

	Primary prevention	Secondary prevention	Tertiary prevention	Suppression	Rehabilitation
Places
People
Behaviors

As noted previously, different forms of violence require different strategies. This framework is carefully constructed in order to address community violence among youth, especially lethal violence. Addressing other forms of violence will require different strategies, partners, and frameworks. As discussed further below, an approach carefully constructed to address one form of violence cannot be casually transferred to other forms of violence, such as family violence or violence perpetrated by organized crime, without significant modification.

Understanding ‘what works’ in community violence prevention

An effective anti-community violence framework should be populated with strategies informed by the best and most rigorously generated evidence, data, and information available. In addition, selected programmatic strategies should be suitable for and tailored to the local context in which it will operate. Finally, in order to be effective, programming must be soundly implemented. In some analyses, well-implemented interventions outperform poorly implemented ones, even if the latter have stronger, more evidence-informed designs (Lipsey, 2009). Tension between these principles is unavoidable, and mitigating such tension is a complex matter of professional judgment. This section explores the first of these principles; the latter two are examined in the discussion.

Since its introduction in the late 1990s, the term ‘what works’ has been understood in criminology to mean programming that has been demonstrated to be effective according to causal evidence generated from well-designed quasi-experimental or preferably experimental evaluations (Sherman et al., 1997). As noted previously, this paper draws significantly from a recent report (Abt & Winship, 2016) that analyzed 43 separate systematic reviews and meta-analyses synthesizing such evidence concerning community violence. Conclusions from that report are summarized and elaborated upon below in order to provide initial guidance on how best to populate the proposed framework.

Place-based primary prevention

Urban renewal strategies are associated with reduced crime and violence as well as improvements in police legitimacy and collective efficacy, but the number and quality of studies supporting these findings are limited (Cassidy et al., 2014). CPTED, or Crime Prevention Through Environmental Design, seeks to prevent crime through manipulation of the physical environment. Multiple reviews found only modest impacts on crime and especially violence for such strategies (Cassidy et al., 2014; Farrington & Welsh, 2002; Farrington et al., 2007; Welsh & Farrington, 2009). Neighborhood watch programs yielded similarly modest effects (Bennett et al., 2006).

In order to strengthen these programmatic strategies, particularly for urban renewal and CPTED, the focus of these efforts could be narrowed to the specific locations where most violence occurs, i.e. hot spots, in effect elevating them in focus from primary to secondary or tertiary prevention. Urban revitalization and environmental crime prevention efforts are worthwhile for a multitude of reasons, but if the intended purpose is violence prevention specifically, their focus should be restricted to those micro-locations that generate the greatest amounts of such conduct.

People-based primary prevention

Researchers disagree as to the impacts of vocational training on criminal behavior, reflecting a broader uncertainty in the field as to the effectiveness of stand-alone employment, vocational, and training programs (Aos et al., 2006; Visher, Winterfield, & Coggeshall, 2005) for violence reduction. Youth mentoring receives a similarly mixed assessment (Fagan & Catalano, 2013; Jolliffe & Farrington, 2007). As with place-based primary prevention, such programs could improve performance with additional focus on youth at the greatest risk for violence. In addition, pairing such programming with proven strategies like cognitive behavioral therapy (CBT) (examined further below) could also enhance effectiveness.

More promising were school and especially family-based interventions, particularly when such efforts employed CBT. Early childhood programs such as the Perry Preschool program and the Nurse Family Partnership have especially strong and well-established effects (Fagan & Catalano, 2013). Making sure these programs serve the schools and families most impacted by violence will further strengthen anti-violence outcomes.

Behavior-based primary prevention

Juvenile curfews, gang prevention, and gun buyback strategies all demonstrated no impacts on crime or violence (Adams, 2003; Gravel, Bouchard, Morselli, & Descormiers, 2012; Makarios & Pratt, 2012).

Place-based secondary prevention

No systemic reviews examining place-based secondary prevention interventions were identified. An exploration of why systematic causal evidence is available for certain framework categories but not others would be worthwhile but is beyond the scope of this article.

People- and behavior-based secondary prevention

CBT uses clinical psychological techniques to alter the distorted thinking and behavior of criminal and juvenile offenders. A Campbell Collaboration review strongly reinforced, with 58 studies, 19 of which were randomized controlled trials, what numerous others had previously found: CBT works (Lipsey, Landenberger, & Wilson, 2007). CBT has been effective in reducing recidivism among juvenile and adult offenders, in institutional or community settings, as part of a broader program or as a stand-alone intervention. Few interventions can match its reliability and versatility. CBT was associated with a relatively large 25% average decrease in recidivism, but when the most effective types of CBT were used, recidivism declined 52%. These most positive results were not an outlier – approximately 1 in 5 of the interventions studied produced such effects or better.

Several family-based secondary prevention programs, including the well known Family Functional Therapy, Multi-Systemic Therapy, and Multidimensional Treatment Foster Care interventions, have strong records of demonstrated of effectiveness in reducing aggressive behavior, delinquency, and contact with the criminal and juvenile justice systems (Fagan & Catalano, 2013). These programs often prominently feature CBT, among other techniques. It should be noted, however, that one systematic review has questioned the effectiveness of Multi-Systemic Therapy (Littell, Campbell, Green, & Toews, 2005).

While no systematic reviews have been conducted, numerous studies indicate that violence can be reduced substantially by regulating the availability and use of alcohol (WHO, 2010), either by managing the hours, prices, and locations at which it is sold, providing treatment for alcohol abusers, or by improving the management of environments where it is served.

Place-based tertiary prevention

No systemic reviews examining place-based tertiary prevention interventions were identified.

People- and behavior-based tertiary prevention

A common prominent feature of strategies in this area is an intense focus on those at the highest risk for violence, paired with an equally intense focus on a narrow range of behavior, usually firearm-related shootings and homicides. Focused deterrence had the largest direct impact on crime and violence, by far, of any intervention examined in the meta-review. Focused deterrence involves the identification of specific offenders and offending groups, the mobilization of a diverse group of law enforcement, social services, and community stakeholders, the framing of a response using both sanctions and rewards, and direct, repeated communication with the individuals and groups in order to stop their violent behavior. In another Campbell Collaboration review, 9 of 10 focused deterrence interventions substantially reduced crime and violence, with homicide reductions ranging from 34% to 63% (Braga & Weisburd, 2012). Since publication of that review, additional studies have documented more examples of focused deterrence success (Corsaro & Engel, 2015).

The use of street, gang, or youth workers to quell violence has a mixed record of effectiveness and the strategy’s results appear quite sensitive to implementation. No systematic review has been performed, but the best-known application of the strategy, CureViolence, has been evaluated numerous times, earning uneven results (Whitehill, Webster, & Vernick, 2012; Wilson & Chermak, 2011) while drawing support from some quarters (Butts, Roman, Bostwick, & Porter, 2015) and criticism from others (Kennedy, 2011; Papachristos, 2011). ‘Street mediation’ strategies offer significant promise but also some risk. Further systematizing and professionalizing the approach, clarifying streetworkers’ relationship to formal institutions such as law enforcement, and evaluating additional streetworker models (Los Angeles’ Urban Peace Institute and Providence’s Institute for the Study and Practice of Nonviolence are both noteworthy, longstanding efforts) could strengthen the understanding and performance of these efforts.

Place-based suppression

Hot spots policing, which focuses police attention on the small geographic areas where crime frequently concentrates, has consistently demonstrated modest to moderate impacts on crime and violence (Braga, Papachristos, & Hureau, 2014). Disorder policing, also known as broken windows policing, addresses physical and social disorder in neighborhoods in order to prevent crime and violence. While this strategy has moderate crime benefits, problem- and community-oriented applications demonstrated stronger results, without triggering community resistance, than aggressive ‘zero tolerance’ versions (Braga, Welsh, & Schnell, 2015). Community policing leverages partnerships with residents and the community in order to reduce crime and disorder, but surprisingly has little discernable impact on crime and violence, although it did positively affect citizen satisfaction, perceptions of disorder, and police legitimacy (Gill, Weisburd, Telep, Vitter, & Bennett, 2014).

People-based suppression

Problem-oriented policing uses analysis to tailor law enforcement responses to specific public safety problems, yielding a modest to moderate reductions in crime and violence (Weisburd, Telep, Hinkle, & Eck, 2010). Interestingly, problem-oriented policing appears to improve the performance of other policing strategies, such as hot spots, disorder, and community-oriented policing, and may have greater impact in a supporting rather than leading role. It should be noted that problem-oriented strategies span the spectrum of place-, people-, and behavior-based approaches.

Behavior-based suppression

Targeted firearms enforcement has demonstrated moderate effects in reducing gun crime and violence (Koper & Mayo-Wilson, 2006; Makarios & Pratt, 2012). Conversely, aggressive drug enforcement appears to have minimal impacts (Mazerolle, Soole, & Rombouts, 2006) and may actually increase violence by destabilizing drug markets thereby increasing competition and violence among drug sellers (Werb et al., 2011).

Place-based rehabilitation

No systemic reviews examining place-based rehabilitation interventions were identified.

People-based rehabilitation

There is a substantial body of evidence demonstrating that well-designed, well-implemented recidivism reduction programs employing a risk/needs/responsivity framework are effective (Lipsey, 2009; Smith, Gendreau, & Swartz, 2009). The evidence is equally clear that surveillance, deterrence, and discipline strategies are ineffective at best in reducing recidivism among youth (Petrosino, Turpin-Petrosino, Hollis-Peel, & Lavenberg, 2013; Wilson, MacKenzie, & Mitchell, 2005). While boots camps have no impact, Scared Straight programs actually cause harm in that they are associated with modest increases in juvenile recidivism. Restorative justice programs appear to have modest impacts on offender recidivism, when both the victim and offender affirmatively consent to participate in the intervention (Latimer, Dowden, & Muise, 2005; Sherman, Strang, Mayo-Wilson, Woods, & Ariel, 2015).

Behavior-based rehabilitation

Multiple systematic reviews indicate that drug treatment and drug courts can significantly and positively impact recidivism (Holloway, Bennett, & Farrington, 2006; Mitchell, Wilson, Eggers, & MacKenzie, 2012; Mitchell, Wilson, & MacKenzie, 2007).

Populating the framework with ‘what works’

A framework populated with the strategies described above is provided in Table 2. When interventions with minimal, mixed, null, or negative effects on community violence are removed from the framework as in Table 3, a clearer picture of ‘what works’ begins to emerge. Once the places, people, and behaviors responsible for generating most community violence in a given jurisdiction have been identified, this framework can provide a helpful roadmap for identifying, selecting, and coordinating programmatic responses.

Table 2. Anti-community violence framework, populated.

	Primary Prevention	Secondary Prevention	Tertiary Prevention	Suppression	Rehabilitation
Place	Urban renewal			Hot spots policing
				Disorder policing
	CPTED
	Neighborhood watch
				Problem-oriented policing
				Community policing
People	Family-based therapy	CBT	Focused deterrence	Problem-oriented policing	Recidivism reduction
		Family-based therapy
					Restorative justice
			Streetworker programs
					Boot camps
	School-based programs				Scared straight
	Vocational training
	Mentoring
Behavior	Family-based therapy	CBT	Focused deterrence Streetworker programs	Problem-oriented policing	Drug courts and treatment
		Family-based therapy
	School-based programs			Firearms enforcement
		Alcohol regulation
				Drug enforcement
	Juvenile curfews
	Gun buybacks
	Gang prevention

Note: Interventions with modest, mixed, null, or negative effects on community violence are indicated in italics.

Table 3. Anti-community violence framework, populated with ‘what works’ only.

	Primary Prevention	Secondary Prevention	Tertiary Prevention	Suppression	Rehabilitation
Place			(Urban renewal) (CPTED)	Hot spots policing
				Disorder policing
				Problem-oriented policing
People	Family-based therapy	CBT	Focused deterrence	Problem-oriented policing	Recidivism reduction
		Family-based therapy
			(Streetworker programs)
	School-based programs
Behavior	Family-based therapy	CBT	Focused deterrence (Streetworker programs)	Firearms enforcement	Drug courts and treatment
		Family-based therapy
		Alcohol regulation
	School-based programs
				Problem-oriented policing

Note: Interventions that, if revised, could have significantly improved effects are indicated in parentheses.

For instance, with micro-locations where violence generally concentrates, place-based suppression strategies such as hot spots policing can make an immediate difference, while hot spot urban renewal and CPTED prevention strategies can improve transportation, housing, lighting, and vegetation in the immediate area so that reductions in violence are more likely to be sustained over time.

For those most likely to perpetrate or be victimized by violence in and around hot spots, focused deterrence initiatives supported by street or youth workers could send a clear message that violence will not be tolerated. CBT can train young men to control their anger and solve their disputes peaceably. Weak programming can be strengthened rather than eliminated by incorporating CBT elements. Family-based therapy can assist parents in keeping their children on the right path.

For the most dangerous behaviors, targeted police patrols can engage repeat offenders and take guns off the street. Regulations on alcohol can change when, where, and how liquor is sold, making it harder for violent people in hot-spot areas to drink to excess. Finally, as to gangs, focused deterrence and streetworker programs have proven capable of influencing gang behavior.

Making the case for accumulation, concentration, and coordination

Examining anti-violence efforts across all categories, it becomes clear that only a few demonstrate clear and substantial positive effects on crime and violence, and even the most powerful interventions are incapable of reversing high rates of community violence by themselves. Given this, success requires the accumulation of individually modest but collectively robust programmatic effects. Risk and protective factors for violence are cumulative by nature (Office of the U.S. Surgeon General, 2001), so a strategy that builds impact over multiple programs makes similarly good sense.

As noted previously, crime and violence generally concentrate in and around a small number high-risk places, people, and behaviors. It follows that programmatic interventions targeting these concentrations are more likely to be effective than those that do not. Across the spectrum of anti-violence programming, it is well established that interventions focusing on the highest risk places, people, and behaviors generate the strongest effects. This is true of interventions relating to policing (Braga, 2015), gang reduction (Gravel et al., 2012), youth firearm violence reduction (Petrosino et al., 2015), youth violence prevention (Matjasko et al., 2012); and adult and juvenile recidivism reduction (Hollin, 1999; Lipsey & Cullen, 2007).

Similarly, in public health terminology, interventions targeting indicated and selected populations tend to outperform those addressing universal ones. In a systematic review of 41 youth violence interventions, researchers found that effectiveness increased as the interventions moved from primary through secondary to tertiary prevention (Limbos et al., 2007). Of the 15 studies that were randomized controlled trials, two of six (33%) primary, three of seven (43%) of secondary, and both (100%) of the tertiary prevention interventions were effective in reducing violent behavior among youth.

Accumulating and concentrating effects will fail if crime simply ‘moves around the corner’. Fortunately, a robust body of rigorous evidence clearly establishes that when crime and violence are targeted, displacement to surrounding areas is minimal.

[O]ver 30 years of research evidence on this topic … suggests that crime relocates in only a minority of instances. More commonly, it has been found that the opposite, a diffusion of crime reduction benefits in nearby areas not targeted by interventions, occurs at a rate that is about equal to observations of displacement. (Johnson et al., 2014)

A corollary to the case for concentration is the need for coordination among selected programs. Unfortunately, there is little practical guidance for policymakers on how to identify the right mix of interventions and how to coordinate them effectively (Abt, 2014). For instance, the case for ‘comprehensive’ approaches to community violence prevention is decidedly mixed. Some reviews have found comprehensive or holistic approaches to be ineffective due to the inherent implementation challenges associated with getting numerous participants and organizations ‘on the same page’ (Gravel et al., 2012; Matjasko et al., 2012). Others claim such approaches work because they capitalize on the strength and diversity of multiple stakeholders (Makarios & Pratt, 2012; Petrosino et al., 2015). Complete comprehensiveness, while laudable in theory, is unlikely to be achievable in practice. The best case for multi-disciplinary collaboration recognizes that the capacity to coordinate is a finite resource like any other, and one to be used judiciously.

In order to achieve significant reductions in community violence, resources should be amassed and aligned where they will be most effective. Accumulating, concentrating, and coordinating efforts is intuitive, backed by strong evidence, and perhaps most importantly, economically and administratively feasible. Public and private institutions responding to violence lack the capacity to act everywhere, but they can collaborate where it matters most.

Discussion

The conceptual framework proposed here has several advantages over previous models. First, community violence – the challenge to be addressed – is contextualized in relation to other forms of violence. Second, the framework integrates theories and models from criminal justice and public health, the two most dominant fields in violence prevention. In doing so, the framework accounts for both the problem of community violence as well as potential policy solutions – an advancement over models that describe only one or the other. Third, the framework is informed by the most rigorous evidence currently available, in that the model reflects the concentrated nature of the phenomenon and synthesizes programmatic evidence of effectiveness. Fourth, by consolidating and adapting previous well-recognized theories and models, the framework is relatively readily understood and applied. In short, the proposed framework satisfies the previously-stated criteria of completeness, accuracy, and usability.

This conceptual framework also has a number of limitations. First, the proffered violence continuum is admittedly and necessarily imperfect. This can and should be improved upon as scholars and practitioners focused on other forms of violence contribute their insights and expertise.

Second, the framework itself is intended for community violence only. Violence in homes and schools operates differently than in the community, as does organized violence between gangs, criminal organizations, or states. For example, while the evidence is clear with regard to the limited displacement of community violence, the same cannot be said for sophisticated criminal organizations such as transnational drug cartels, which have repeatedly demonstrated the capacity to relocate or otherwise respond to targeted interventions. Relatedly, the framework does not speak to crime generally, but instead focuses squarely on violence, particularly lethal violence. Finally, the framework primarily addresses youth, but with adaptation the framework could have value in addressing community violence more generally, as its concentrated nature tends not vary with age.

Third, while the populated framework represents a summary of the best evidence currently available concerning anti-community violence programming, the limits of this evidence and of evidence-informed policy generally must be kept in mind. Within the field of community violence prevention, significant evidentiary gaps remain, and the scientific process is an iterative one, meaning that our understanding of community violence and how best to prevent it must be continually updated and refined. In addition, the evidence relied upon here was drawn overwhelmingly from high-income settings, raising important questions concerning its generalizability or external validity. While emerging evidence indicates that while crime and violence appears to concentrate similarly across contexts (Jaitman & Ajzenman, 2016), low and middle-income nations often face capacity challenges when attempting to implement evidence-informed programming.

Fourth, the framework speaks primarily to the identification, selection, and organization of evidence-informed programs, but understanding the evidence is only one of several components of success in reducing violence. Programs must be carefully selected, adapted, and implemented according to local circumstances. Doing this effectively requires a sound understanding of the local context along with an intervention’s essential elements of effectiveness. If the context allows for adoption of the essential elements, implementation can proceed with nonessential components altered as necessary. If the context requires compromising of one or more essential elements, a more contextually appropriate choice should be selected. Additionally, the programmatic focus of the framework offers little in terms of institutional strengthening, which is another important component of any long-term anti-violence strategy.

Fifth, the framework does not address program implementation, a critical component of intervention success (Hollin, 1999; Lipsey & Cullen, 2007). When implementing, careful attention should be paid to the quantity, intensity, and/or dosage of a given treatment. The U.S. National Research Council (2013) has concluded that with regard to youth at risk for violence and criminality, ‘Whatever the specific mechanism, the appropriate focusing of more intense (and costly) interventions on higher risk adolescents produces a greater reduction in subsequent offending and limits the negative effects of unwarranted intensive intervention on less serious offenders’. This finding reinforces the previously made case for the accumulation and concentration of programmatic effects.

With these advantages and disadvantages in mind, how might be the theoretical framework advanced in this article be adopted? First, every effort – whether at the local, state, or national level – should begin with analysis, the goal of which is to identify the places, people, and behavior among which violence concentrates. Next, a common understanding of the concentrated nature of violence must be built by transparently sharing this information with civic, community, and criminal justice stakeholders. Procedural justice principles (Tyler, 2006), increasingly used in the United States and elsewhere to improve the legitimacy of criminal justice institutions, can offer helpful guidance in this area.

Once a common understanding of the problem has been achieved, the effort may turn to solutions. Using the framework, stakeholders would identify a balanced set of contextually appropriate programmatic interventions, including both suppression and some form of prevention (preferably secondary or tertiary), in each of the categories of place, people, and behaviors. Especially initially, care must be taken to avoid ‘policy sprawl’, e.g. including extraneous partners who contribute little but drain coordination capacity. Ideally, the selected interventions will bring together a small set of like-minded multi-disciplinary partners who can focus exclusively on preventing community violence where it concentrates most. In terms of measuring performance, it makes sense to begin with homicide, as it is the most costly, comparable, and reliably measured form of violence.

Once selected, the implementation of interventions must maintain fidelity to the essential elements of effectiveness while adapting said interventions to the local context. Care must be taken to ensure the delivery of high-dosage treatments to high-risk places and people. Efforts should be monitored and evaluated on an ongoing basis, using the best data and most rigorous methodologies available under the circumstances.

Finally, an ongoing effort should be conducted to coordinate with other anti-violence efforts represented along the violence continuum. Regions plagued by high rates of violence generally suffer from numerous forms of violence, necessitating a set of separate but loosely connected strategies. While emphases may vary, given the diversity of and contagion between different forms of violence, policymakers cannot afford to focus on only one type of violence to the exclusion of all others. Community violence prevention practitioners should therefore meet semi-regularly with their colleagues working to prevent other forms of violence, maintaining situational awareness and seizing opportunities to collaborate when possible. For instance, in El Salvador, efforts to prevent community violence should be aligned with efforts to limit the violence perpetrated by gangs, cartels, and other criminal organizations. Similarly, community violence practitioners should be aware of the linkages between early exposure to family and intimate partner violence and violence perpetrated later in life.

Conclusion

Community violence among youth is a complex and persistent social phenomenon. Responding to such a challenge requires a sophisticated and nuanced understanding of the problem as well as the collective capacity for solutions. Such an understanding is necessarily imperfect and constantly evolving, but can nevertheless be advanced via the development and adoption of theoretically sound, evidence-informed, and practically implementable frameworks such as the one proposed in this article.

Disclosure statement

No potential conflict of interest was reported by the author.

Funding

This work was supported by Know Violence in Childhood.