You are hereSkip Navigation Links > Home > Professionals > Ontario's Youth Action Plan > Roots of Youth Violence > Volume 5 > Community Development Strategies

Review of the Roots of Youth Violence: Literature Reviews

Volume 5

Community Development Strategies23


During his election campaign, Canadian Prime Minister Stephen Harper repeatedly emphasized a tough law-and-order platform based on the premise that “our safe streets and healthy communities are increasingly under the threat of gun, gang, and drug violence” (Harper, personal webpage). Harper’s solution: more policing, mandatory minimum sentences for gun-related offences and longer periods of incapacitation for repeat and violent offenders.

The implementation of these measures is based on a “reactive” crime-fighting strategy, whereby the imposition of tough, draconian sanctions is the direct response to crimes that have already taken place. While much of policing has traditionally been based on this reactive paradigm, there has increasingly been an emphasis on enacting “proactive” strategies that seek to promote change before crimes occur.

One proactive approach that has garnered significant public attention has been the implementation of community-based crime prevention strategies. While there has been a lot of ambiguity surrounding the precise meaning of “community-based crime prevention,” and what falls under this general rubric (Welsh and Hoshi, 2002:165), the common denominator of these approaches is that they all emphasize the importance of fostering community-based collaborations in the prevention of crime. This is based on the assumption that neighbourhoods themselves are criminogenic (Sampson and Wilson, 1995; Shaw and McKay, 1942) and so, in order for policies and programs to be effective, the broader social environment has to be targeted.

The present paper provides an elaborate review of the community-based crime prevention initiatives that have been implemented in an effort to reduce youth violence. Particularly, this paper will evaluate the research that has been conducted over the past decade, and will critically assess the methodological rigour of the various studies conducted on violence prevention programs. Given that most of the evaluations have, to date, been conducted in the United States, this review will primarily be structured around the American literature.

The present review will be structured in the following manner: first, an overview of the theoretical underpinnings of community-based crime prevention strategies will be offered. Next, the existing body of literature will be summarized in order to provide the necessary context. Building on these reviews, the literature from the past decade will be reviewed thematically. Finally, this review will end with a synthesis of the literature, outlining future directions for study in the area of community-based crime prevention programs.

The Theory of Community-Based Crime Prevention

The implementation of a community-based approach for reducing youth violence is based on a rich theoretical tradition that can be traced back to the famous “Chicago School” studies of the early 20th century (Welsh and Hoshi, 2002: 166–67; Rosenbaum, 1988: 326). These studies emphasized the importance of neighbourhood composition in the determination of crime rates, and specified that communities with high levels of transience, an ethnically heterogeneous population, and a large proportion of young males are most likely to be characterized as high-crime societies (Cullen and Agnew, 2006: 87). Shaw and McKay (1942) found that, over time, Chicago’s crime rates clustered within specific geographical regions of the city regardless of the ethnic composition of those who inhabited the area (Cullen and Agnew, 2006: 87). Shaw and McKay concluded that there was a high level of social disorganization endemic to these communities, and that this disorganization mediated the relationship between the individual and crime. While the theory of social disorganization has been refined and wrapped in a new cloak over time (e.g., Sampson and Wilson, 1995), the central premise has remained largely the same: socially disorganized communities have lower levels of informal social control, which ultimately erodes the shared norms and beliefs in the community. These low levels of informal social control combine to create a highly criminogenic society, so in order to prevent crime, programs have to strengthen social ties and create an enhanced sense of community (Cullen and Agnew, 2006: 87).

A slightly different theoretical approach to studying community-based crime prevention argues that, similarly to social disorganization theory, increasing the levels of community cohesion is an important prerequisite for preventing crime. Popularized by the theory of broken windows, the focus of this line of research is not to address the structural inequalities that ultimately lead to crime; rather, the focus is on the more proximal characteristic of creating social order, which can be attended to by altering the physical characteristics of the environment. The broken windows thesis is predicated on the assumption that crime is about the “breakdown of social control,” and the first step in the dissolution of social control is the presence of social disorder (Cullen and Agnew, 2006: 457). Using their famous broken windows” analogy, Wilson and Kelling (1982) argued that the presence of social disorder (e.g., drunks, loiterers, prostitutes) is like a broken window, and if it is not fixed, then “all of the other windows will soon be broken” (31). Crime prevention efforts are therefore necessarily centred on fixing these broken windows, which is tantamount to cleaning the seemingly socially disordered elements of the community and making them appear orderly. The re-exertion of social control is usually achieved in the form of various types of policing, although the broader community is also often included (e.g., neighbourhood watches, anonymous tip lines), with the end result being greater levels of social cohesion and a redefined sense of community.

Although the broken windows theory operates from a more conservative framework to prevent crime than the social disorganization theory, both theories emphasize that in order for crime prevention programs to be effective, the inclusion of community preventative efforts is absolutely essential.

Previous Reviews on Community-Based Crime Prevention Strategies

The theoretical appeal of community-based crime prevention strategies, and the precipitous costs associated with youth violence has brought forth an emergence of a myriad of different community-based crime prevention programs. Now part of the popular vernacular, these programs are part of the crime prevention literature more generally, a $4 billion dollar per year industry in the United States alone (Sherman, 2002: 1).

Despite the costs associated with implementing these programs, they have remained surprisingly resistant to evaluations and evidence of their effectiveness (Welsh and Hoshi, 2002; Rosenbaum, 1988). Dennis Rosenbaum (1988), in his earlier synthesis of the crime prevention literature, was strongly supportive of the theoretical rationale for these programs, but was adamant about the need for an accompanying body of research to support their efficacy. When addressing why there is not a consensus on “what works,” Rosenbaum pointed directly to the quality of the evaluation research, and stated that “unless evaluations are funded adequately and conducted by trained researchers, they should not be conducted at all because the results are usually untrustworthy” (381).

The inadequacies of the literature were re-emphasized by Welsh and Hoshi (2002) who (writing some 14 years later) stated that “community-based crime prevention does not, at the present time, demonstrate evidence of proven effectiveness in reducing crime” (189). Although they are similarly critical of ideologically driven programs that do not attempt to evaluate their effectiveness, Welsh and Hoshi (2002) do cite a number of encouraging trends in the evaluation literature, and emphasize the need to abandon the “nothing works” philosophy in favour of a more rigorous understanding of effective practices and principles (Welsh and Hoshi 2002: 189).

Neither of these reviews, importantly, questions the underlying theoretical framework upon which violence prevention programs are based. Instead, these reviews redirect our attention towards empirically substantiating the theoretical significance of these programs in an objective, methodologically correct and rigorous manner (Welsh and Hoshi, 2002: 189–90; Rosenbaum, 1988).

The Current Review

The purpose of this review is to examine the recent literature that has been published in the area of community-based crime prevention, with a specific emphasis on the methodological advancements that have been made over the course of the last decade. Focusing only on those programs that were formally evaluated, this literature will be organized into the following sections: community-based gun prevention, community-based gang prevention programs, community-based school prevention programs, and community empowerment programs. Each section will be considered in turn.

Following a synthesis of the literature, a detailed discussion will be offered that summarizes the current state of knowledge with respect to community-based crime prevention strategies, emphasizing future directions for research and policy.

Community Based Gun Prevention Programs

Gun-related violence constitutes a serious public health problem. In the United States, 32,000 Americans died of gunshot wounds in 1997, which was greater than the number of people who died from AIDS or liver disease in the same year (Cook and Ludwig, 2000: 15). Or, to place gun-related violence in a more historical context, over one million Americans have been killed since 1965, which is greater than the total number of American soldiers killed in all the foreign wars of the 20th century (Cook and Ludwig, 2000: 15).

Given the exorbitant costs associated with gun violence, a lot of impetus has been directed toward ways in which gun-related violence can be prevented. One specific area of investigation has been the implementation of gun buy-back programs in the United States. In their review of the evaluation literature, Welsh and Hoshi (2002) found that the efficacy of gun buy-back programs has been plagued by a series of serious methodological weaknesses, and as a whole, the body of scientific research has failed to support the belief that these types of programs are an effective means of preventing criminal gun violence (189). Kuhn et al. (2002) concurred with Welsh and Hoshi’s conclusions (143), and found that in the case of Milwaukee, gun buy-backs typically yield different types of handguns than those used in either homicides or suicides (145). The tentative conclusion from these studies, then, is that while gun buy-backs do create a great deal of media exposure (Kuhn et al., 2002: 145.), they have not been empirically validated as a reasonable strategy to prevent gun-violence.

Another community-based prevention strategy for dealing with gun-related violence has been a series of locally based American programs, which have been funded by the National Institute of Justice (NIJ). These programs are all based on the seminal findings of Operation Ceasefire, a collaborative community-based intervention between the Boston Police Department and researchers from Harvard University. Recognizing the important interactions between workers in a wide variety of law enforcement, social service and other professional capacities (Braga et al., 2001: 218), Operation Ceasefire was centred around the “pulling levers” philosophy, a deterrence strategy where the emphasis was placed on a small number of chronically offending gang youth. The implementation of this large, interdisciplinary framework resulted in significant reductions of youth violence and homicide in Boston (Braga et al., 2001: 219), and the program has generally been deemed a solid success (Roehl et al., 2006: 1). Despite the absence of a “true experimental design,” the study utilized robust evaluation methods and supported the underlying theory that, when properly implemented, various groups can work together to decrease youth violence.

Operating at roughly the same time as Operation Ceasefire, a multidisciplinary team of researchers and community agencies in Atlanta, Georgia were also evaluating a preventative youth violence initiative called Pulling America’s Communities Together (PACT), mandated to implement “a problem-solving approach to juvenile gun violence” (Kellerman et al., 2006: 3). The PACT program was predicated on the assumption that gun violence was the end result of a series of illegal activities and, in order to prevent gun violence, the earlier stages of illegal activities needed to be interrupted. Primarily targeting the purchasing and selling of illegal handguns, the program ran for five years (1995 to 1999), and over the duration of the study, the number of homicides in Atlanta fell 18 per cent (Kellerman et al., 2006: 23). The authors were careful to note, however, that the sharp decline in the homicide rate was accompanied by a more general decline in the homicide rate at both the state and national level. The authors did not statistically adjust for these macro-level shifts in the homicide rate, and as such, the evidence could not explicitly link the change in Atlanta’s homicide rate to the impact of the PACT program.

Funding for Atlanta’s PACT program ended in 1999, and the Strategic Approaches to Community Safety Initiative (SACSI), an intensive violence prevention program funded by the U.S. attorney’s office, replaced PACT. Organized, around the principles of the Operation Ceasefire initiative, SACSI comprised a multi-agency collaborative endeavour between community agencies, the police, and the research community with the specific mandate of using best-known practices to reduce violent crime at the local level (Kellerman, et al., 2006: 29; Roehl et al., 2005: 1).

The distinguishing characteristic of the SACSI programs was the fact that they were based on collaborations between the police, the broader community and academic researchers, with the specific goal of empirically evaluating the extent to which these programs successfully decreased the levels of violence. The SACSI program, in other words, addressed the harsh criticism of earlier reviews (e.g., Rosenbaum, 1988), and allocated resources to evaluate their initiative.

The SACSI programs, in total, consisted of 10 geographically and demographically diverse cities, evaluated in two separate phases (see table 1 for a summary of the 10 cities evaluated). Throughout the period of evaluation, the violent crime rate immediately dropped in most cities (Roehl et al., 2006: 4), and the body of research, considered in its entirety, reveals that preventative policies integrating various community partnerships can have a positive impact on the crime rate (Roehl et al., 2006: 78–80).

While the SACSI programs were not designed to specifically answer youth violence, any initiative aimed at reducing violent crime rate necessarily has to focus on youth, given that they comprise the highest risk of violent offending and victimization (Farrington, 1986). Three programs were specifically designed around the prevention of youth gun-related violence, and as such, each of these programs will be reviewed below.

The Strategic Approach to Community Safety Initiative (SACSI) began in St. Louis in the fall of 2000 and continued through 2003. The need to prevent gun-related violence in St. Louis was great, given the fact that St. Louis has traditionally had homicide rates that ranked in the top 10 nationally, and within the confines of the city, homicide rates had traditionally clustered around seven hot spots. (Decker et al., 2005: 8). Furthermore, the homicide rates for Black males aged 15 to 19 exceeded 380 per 100,000, and for Black males reached 600 per 100,000. For these groups, firearms accounted for over 95 per cent of the deaths among this group (Decker et al., 2005: 6). The St. Louis SACSI project, therefore, consisted of a series of preventive interventions, policing, and deterrence strategies to decrease the gun-related homicide rates in this specific demographic. Using a saturation enforcement model, selected aggressive police enforcement, and an emergency-room intervention, the homicide rate in Atlanta fell from its 10-year average of 175 to 113 in 2002 and 68 in 2003 (Decker, 2005: 49). While these findings are striking (and were much greater than the relative homicide rate drops in other jurisdictions), these studies’ findings are limited by the short time period following the implementation of the program, and the absence of robust time series models that could account for potential biases in the data (e.g., period effects, cohort effects).

Initiated in Winston-Salem North Carolina, and targeted directly at the issue of youth violence, the SACSI program established a working group to study the issue of youth violence, and from this, an intervention was designed and evaluated. The intervention consisted of a continued outreach program by police and community agencies that attempted to bring together a diverse group of community representatives. Specifically, the intervention services provided through SACSI included job training, job placement, mentoring, family-based services, and after-school activities (Roehl et al. 2006: 13). The evaluation of the Winston-Salem SACSI program revealed a 58 per cent decrease in juvenile robberies and a 19 per cent decrease in juvenile incidents in target neighbourhoods (Roehl et al., 2006: 20). However, while these decreases were significant, at least one-fifth of the targeted group committed at least one violent act (Easterling et al., 2002: 13). Furthermore, without a control group, the authors could not make definitive statements about the counterfactual (Easterling et al., 2002: 13). They conclude by recommending that future studies should 1) adopt a reasonable model of behaviour change to guide the development of strategies; 2) be more strategic in the choice of leverage points that could produce behaviour change; and 3) maintain a culture of strategic thinking (Easterling et al., 2002: 13–14).

Finally, the SACSI program began in Rochester, New York in the spring of 2001 and continued through 2004. Emphasizing that the process of implementing a collaborative intervention is “less linear than a Request for Proposals might suggest,” (Klofas et al., 2007: 141), the end result was the continuation and reinforcement of a collaborative team of criminal justice professionals to address the problem(s) associated with youth violence (Klofas et al., 2007: 142). The research team was also granted full access to Rochester’s Crime Analysis Unit, and the researchers were able to gain a wide range of “census, school, and community data that provide a rich and comprehensive understanding of the dynamics of youth violence” (Roehl et al., 2006: 38). This research found that there was a statistically significant reduction in the homicide rate in Rochester during the preceding 12 months, and that this trend was particularly true for young, Black males who are at particular risk for being involved in gun-related violence (Klofas et al., 2007).

Table 1. SACSI Community-Based Gun Prevention Programs in 10 U.S. Cities

U.S. City Focus of SACSI Initiative, and years of study Type of Intervention Methodology and Statistics Results and Conclusions
Albuquerque Focus on reducing homicide, firearms violence, and aggravated assault Emphasis onprevention and onthe policing (arrest, enforcement deterrence models, and saturation patrols) -Number of homicides and violent crimes reported annually - Homicide rates remained the same
- Violent crime decreased slightly
Atlanta Focus on homicide, firearms violence Emphasis onpolicing (arrest, enforcement, and prosecution), and prevention strategies (jobs, job training, awareness campaigns) -Number of homicide crime rates reported annually -Homicide rates dropped from 38.5 to 33.8
- Violent crime rates dropped significantly
Detroit Focus on homicide and violent crime rates Emphasis on policing (arrest, enforcement, and prosecution), and prevention strategies -Number of homicides and violent crime rates -Homicide rate decreased from 54 annually to 40 annually
- Violent crimes decreased from around 2700 annually to around 2000 annually
Indianapolis Focus on homicide and firearms violence Emphasis on policing (arrest, enforcement, and prosecution), and prevention strategies (jobs, job training, awareness campaigns) -Number of homicides, aggravated with a gun and armed robberies reported weekly from January 1997 June, 2001. -53 per cent decrease in gun assaults vs. 19 per cent city wide
-44 per cent decrease armed robberies in target neighbourhood vs. 8 per cent city wide
Memphis Focus on rape and sexual assault Emphasis on prosecution, parole and probation strategies, and prevention strategies -Number of forcible rates reported annually -Forcible rapes decreased from 938 annually to 480 annually
New Haven -Focus on reducing firearms violence Emphasis on arrest, enforcement, suppression, deterrence, public awareness -Number of homicides, assaults, and armed robberies reported annually -Violent crime rates dropped 24.4 per cent in target areas compared with 15 per cent in all cities
-Homicide rates dropped 35 per cent compared with 20.4 per cent in other cities
Portland -Focus on violent crime among 15-24 year olds -Emphasis on community policing, prevention (job training), arrest, suppression, enforcement, seizures, saturation patrols (gun prosecution), -Number of homicides, rapes, robberies, and aggravated assaults reported annually -Homicide decreased 51.7 per cent in target city and 36.3 per cent country wide
-Violent crime rates decreased 36.2 per cent in target city vs. 26.6 per cent in all cities
Rochester -Focus on youth and the prevention of firearm violence -Emphasis on community policing, child and family services, policing (arrest, enforcement), on deterrence, and on saturation patrols -Number of homicides, and violent rates reported annually -Increased from 19.6 annually to 20.2 annually (homicide)
-Violent crimes dropped from around 897 annually to around 800 annually
St. Louis -Focus on reducing homicide and firearm violence -Emphasis on prevention (awareness), home visits, gun prevention, and enforcement and suppression strategies -Number of homicides and violent crimes reported annually -Homicide rate dropped from 39.0 annually to 32.5 annually
-Violent crime rate dropped from approximately 2500 annually to 2000 annually
Winston-Salem -Focus on violent crime under 18 Emphasis on prevention strategies (employment, job training and on enforcement strategies and home visits -Pre-test/post-test comparisons of arrest records in 4 targeted neighbourhoods -Significant decrease in the homicide rates
-Juvenile incidents down 19 per cent, juvenile robberies down 58 per cent

The 10 SACSI initiatives funded by the National Institute of Justice (NIJ) have all morphed into Project Safe Neighborhood (PSN), which incorporates enforcement, supervision, prevention, and intervention strategies in all local U.S. Attorney districts throughout the country (Roehl et al., 2006: 77). By extending the PSN policies nationally (but still implementing them at the local level), the standardization of SACSI’s community-based collaborations and integration can be fostered, while allowing enough flexibility to ensure that the strategies can be tailored to the unique individual context of the community (McGarrell, 2005: 21).

Project Safe Neighborhood is the latest version of community-based prevention programs designed to stop gun-related crimes before they happen. These PSN programs, funded by the National Institute of Justice, have increasingly stressed a community-based collaboration, and the continued importance of evaluating these programs with trained researchers. While the findings have not been unanimous, and while some methodological questions persist, these programs have generally resulted in decreased levels of gun-related violence after the initiation of the various programs. These PSN results are encouraging, and there is enough scientific evidence to justify the continued implementation and evaluation of these types of programs.

Community-Based Gang Prevention

Similarly to the literature on gun-related violence, gang-related violence has been perceived as a largely American phenomenon. The National Youth Gang Survey (NYGS), for example, reports that in large cities (over 50,000 people), approximately 80 per cent of police agencies report the existence of gang-related problems (Egley and Ritz, 2006: 1). In terms of prevalence, the NYGS estimates that there are approximately 760,000 gang members in the United States, representing some 24,000 gangs (Egley and Ritz, 2006: 1).

While studies have documented the existence of a gang-related problem with Canadian youth (see Leschied and Heng, 2007; Wortley and Tanner, 2006: 1; Westmacott, Stys and Brown, 2005), the size and scope of the problem is still largely unknown. The consensus in the current research is that gangs do constitute a growing concern in Canada (Leschied and Heng, 2007; Wortley and Tanner, 2006; Westmacott, Stys and Brown, 2005), and future evaluations are needed in order to determine the conditions under which these gangs exist.

The Canadian and American literature, considered together, has warranted a significant amount of attention with respect to ways in which community-based prevention programs can be used to decrease the rates of gang violence. These community-based gang prevention efforts can be traced back to Thrasher’s (1936) seminal study of gang membership in New York. By providing a group of high-risk and underprivileged youth access to a Boys Club, Thrasher hypothesized that the negative characteristics of the criminogenic environment could be circumvented and these children would ultimately be less likely to engage in delinquent and truant behaviours. Thrasher (1936) failed to find that the Boys club was associated with a decrease in juvenile delinquency; however, he did offer a pioneering framework for preventatively addressing juvenile delinquency by focusing on the community.

Welsh and Hoshi (2002) reviewed the community-based gang prevention literature, and separated the research into two distinct categories: gang prevention and gang intervention programs. Gang prevention programs, according to Welsh and Hoshi, were programs that focused on discouraging high-risk children and youth from gang membership, while gang intervention programs were designed around targeting youth already involved in gang activity (Welsh and Hoshi, 2002: 172). Based on the scientific literature, they found that the gang intervention programs used “moderately strong” (179) scientific methodology, and that there is evidence that programs designed to reduce cohesion among youth gangs comprise a “promising community-crime prevention modality” (179). These findings are much stronger than the gang prevention studies, and in particular, Welsh and Hoshi point to the efficacy of Operation Ceasefire in preventing gang-related violence.

Operation Ceasefire, to reiterate, was a collaborative endeavour between the police, academic researchers from Harvard University, and workers from a wide range of social service disciplines. Their efforts were primarily centred on a deterrence-based strategy called “pulling levers,” and focused on both gun-related crimes and gang-related behaviours. These two behaviours are positively and strongly associated with one another, and the focus on preventing both led to statistically significant reductions in both gang-and gun-related violence in Boston’s youth (cited in Welsh and Hoshi, 2002: 179).

Operation Ceasefire’s success at reducing both gun-related and gang-related violence lends support to the need for broad-based community interventions. In fact, research has shown that both prevention and suppression programs, implemented in isolation, have had little impact on preventing gang-related violence (Esbensen, 2000), and therefore, gang prevention and intervention efforts require the collaboration of resources in a manner that seeks change on both the individual and societal level.

This review of the community-based gang prevention literature will focus on a number of evaluations that have taken place since the original publication of the Operation Ceasefire study in 1988 (see table 2 for a summary of the recent literature), and will then synthesize the current state of the literature.

Foremost among the community-based gang prevention initiatives has been the Gang Violence Reduction Program (GVRP) in Chicago’s Little Village neighbourhood. Originally piloted in the early 1980s, the GVRP has long employed a multi-modal strategy in order to reduce and prevent gang-related violence on Chicago’s south side. The program, collaborating between the police and broader community agencies, employed a variety of integrated community strategies including community mobilization, suppression, intervention, and development. Using the language of the researchers, the program was designed to represent “an interorganizational and community approach to the youth gang problem in its most violent form” (Spergel and Grossman, 1997). The evaluation of the program itself (2003; 1997) specifically targeted gang members from two competing gangs, who accounted for approximately 70 per cent of the region’s gang violence (Westmacott, 2005:16–17). This sample of at-risk youth received extensive contact with a variety of program staff an average of four times per week, and logistic regression analyses reveal that youth who had more contact with workers had less involvement with gang activity relative to those who had less contact with workers. In addition, the overall violent crime rates for these youth were significantly lower when compared with the control groups (Spergel, 2007; 2003; Westmacott et al., 2006: 16). The authors were careful to note one caveat: while the targeted group did have reduced levels of legal involvement, the program did not affect macro-level change in gang activity.

Targeting at-risk youth in order to reduce gang-related activity has become a popular strategy for preventing youth violence in the community more generally. Similar to the GVRP, the Building Resources for the Intervention and Deterrence of Gang Engagements (BRIDGE) was designed around a preventative framework for youth aged 12 to 19, and was implemented in the City of Riverside, California starting in 2001. The program consisted of a wide range of programs, including job training, free driver training, and individual counselling (Westmacott et al., 2005: 15), and the pre-test/post-test evaluation did not find support for the reduction of gang-related activity and crime (Seachrest et al., 2003). When combining the outcomes of the program with the expense, the evaluators concluded that the program should be given low priority for continued funding (Westmacott et al., 2005: 15).

Table 2. Community-Based Gang Prevention Programs

Author Name Title of Study Type of Intervention Methodology and Statistics Results and Conclusions
Spergel, 2007;
Spergel et al.,2003;
Choi et al.,2000;
Spergel and Grossman, 1997
The Little Village Gang Violence Reduction Project in Chicago -Community strategies including: community mobilization, gang suppression, gang intervention, personal development -Time series, before after, multivariate regression, logistic regression, comparison with control group
-Measuring contact with youth workers, self reported behaviour, and crime patterns
-More counselling leads to less offending
-Compared with control groups, crime was reduced
-Overall gang membership in the community was not affected
Seachrest et al., 2003 Building Resources for the Intervention and Deterrence of Gang Engagement (BRIDGE) in Riverside, California -Targeted at risk youth with job training, driver training, and counselling -Pre test/post test evaluation -No statistical support for the efficacy of the program
-arrest rates were actually higher post program
Esbensen and Osgood,1999;
Ramsey et al., 2003
Gang Resistance Education and Training (GREAT) program, Arizona -9 week training program for at risk youth from local police -Cross sectional posttest design with statistically robust weighting techniques -Moderate effects (effect size) on pro-social behaviours and self reported behaviour
Esbensen, 2001 Gang Resistance Education andTraining (GREAT) program, Arizona -9 week training program for at risk youth from local police -Longitudinal follow-up of randomly selected participants from the GREAT program with a randomly selected control group -Failure to replicate cross sectional results; students fail to show find consistent behavioural or attitudinal differences
Williams, Curry and Cohen, 2002 Gang Prevention Programs for female adolescents: an evaluation in Boston, Colorado, and Seattle -Support groups, personal growth, conflict resolution, and self-esteem enhancement -Process evaluation, interviews, observations
-Pre-test/post-test design
-All the programs faced a series of problems regarding implementation and inadequate data. Results not trustworthy

Unlike the BRIDGE program and the BVRP program, the Gang Resistance Education and Training (GREAT) program recruited youth from a school-based sample. While the authors’ sample selection process was methodologically unique, and allowed for a greater pool of potential participants, the project similarly targeted youth who were at risk for future gang membership. Located in Arizona, the GREAT program consisted of a nine-week school-based program that was designed to 1) reduce gang activity; and 2) to educate young people about the consequences of gang membership and involvement (Esbensen and Osgood, 1999: 198). Results from this cross-sectional survey suggest that: “students who participated in the GREAT program reported significantly more prosocial behaviors and attitudes than those who did not participate in the program” (Esbensen and Osgood, 1999: 216). The authors are careful to note that their evaluation was very short-term, and therefore, the group of at-risk children was not yet at an age to comprise the most serious risk for gang violence (Esbensen and Osgood, 1999: 217).

Esbensen (2001) addressed the limitations of the cross-sectional evaluation of the GREAT program by extending his analysis to include a longitudinally based quasi-experimental design where students from six of the cities where the GREAT program was implemented were randomly selected and compared with a randomly selected control group. Consisting of a pre-test, and two follow-up periods (one- and two-year), the study failed to find consistent behavioural or attitudinal differences between participants and non-participants across the data points (Westmacott et al., 2005: 6). The authors stress the importance of addressing social and structural considerations in the implementation of individually based prevention programs (Westmacott et al., 2005: 6).

The literature on the efficacy of gang-prevention research, unlike the research on gun-prevention programs, does not have a lot of empirical validity based on the studies that have been recently completed. The modest effects of the GREAT program (Esbensen and Osgood, 1999; Ramsey, Rust and Sobel, 2003) were not replicated in the longitudinal study of the GREAT program (Esbensen, 2003). However, the Little Village Gang Violent Reduction Project in Chicago did find consistently positive outcomes, which indicate that the at-risk youth did experience a decrease in arrests for violent crime. This program did not affect macro-level change in gang membership, and the associated costs of the programs might outweigh the statistical benefits.

School Based Community Initiatives

The GREAT program, designed as a gang resistance strategy targeting at-risk youth in high schools, provides an example of how the school is used as a medium for implementing community-based crime prevention strategies for high-risk youth. These schools have, in fact, increasingly become an important setting for these types of programs. In Ontario, the government has recently made an annual investment of $20 million for the Community Use of Schools (CUS), which is predicated on the assumption that “schools in Ontario are recognized as hubs for community activity and will be affordable and accessible in order to support the goals of a healthier Ontario” (Community Social Planning Council of Toronto, 2007: 6).

This review will therefore consider recent programs that have integrated school-community collaborations in order to prevent youth violence. These studies, in addition to the GREAT program, considered above, will help to inform the current state of the school-community collaborative literature.

In a program funded by the Canadian Department of Justice, 21 Vancouver-based high-risk youth under the age of 12 were selected from the Community-Based Crime Prevention Program, a larger five-year-based initiative. Utilizing an interdisciplinary team of professionals that included classroom teachers, special education teachers, family counsellors, and community-based workers, the study found that, using the classroom as the primary setting for a community-level prevention program, this group of at-risk youth did not experience a decrease in their level of violence. Sibylle Artz did note, however, that there was a 41.2 per cent drop in the overall school violence rate during the course of the study (Artz, 2001).

In another prevention program recognizing the importance of both the school and community in predicting levels of violence, Rollin et al. (2003) studied the effectiveness of a community-based violence prevention program on a sample of 80 high-risk grade eight students in three of Florida’s public schools (Rollin et al., 2003: 404). Utilizing multivariate analysis of variance techniques (MANOVA) and t-tests, the authors found that the children receiving the mentorship program had significantly reduced numbers of suspensions and school-based infractions relative to those who did not. Despite these encouraging findings, the authors are careful to note that the study had problems with missing data, observational biases (i.e., the Hawthorne Effect), and a limited amount of statistical power (Rollin et al., 2003: 414). These limitations all serve to increase the probability of committing a type-1 error, and their presence therefore limits the extent to which these results can be trusted.

Table 3. Community-Based School Violence Prevention Programs

Author Name Title of Study Type of Intervention Methodology and Statistics Results and Conclusions
Rollin et al., 2003 School-based violence prevention model for at-risk grade eighth youth (North Florida) -Mentorship program where a group of at-risk grade eight youth received one to one mentorship from an adult mentor -Study used a control group study to examine the difference between those receiving the mentorship program vs. those who did not.

-MANOVA, t-test were used to examine the differences
-The group receiving the mentorship had significantly fewer suspensions and other school based infractions

-The study did not, how ever have a large sample size, which means there is apotential problem with type-1 errors. There were also problems with missing data
Artz, 2001 -Community based approach for dealing with chronically violent under 12 year old children (Vancouver Island) -A group of high risk 12 year old children received support from a wide variety of professionals over a 5 month period -Not stated -The 21 at risk youth did not experience decreased levels of violence, although the broader Community Based Violence Prevention Initiative did find a significant reduction in violence
Casella, 2002 Where public meets the pavement: Stages of public involvement in the prevention of school violence (New York State) -A qualitative and theoretically-based article where the author discusses the relationship between school-based violence and the import of the broader community -Semi-structured qualitative interviews, and a theoretical framework -Schools need funding and collaborations with community based agencies in order to address the underlying structural determinants of school-based violence
Esbensen and Osgood, 1999;
Ramsey et al., 2003
Gang Resistance Education and Training (GREAT) program, Arizona -9-week training program for at-risk youth from local police -Cross sectional posttest design with statistically robust weighting techniques -Moderate effects (effect size) on pro-social behaviours and self reported behaviour
Esbensen, 2001 Gang Resistance Education and Training (GREAT) program, Arizona -9-week training program for at risk youth from local police -Longitudinal follow-up of randomly selected participants from the GREAT program with a randomly selected control group -Failure to replicate cross sectional results; students fail to show find consistent behavioural or attitudinal differences
Segawa, et al., 2005 Evaluation of the Effects of the Aban Aya Youth Project in Reducing Violence among African American Youth Using Latent Growth Mixture Modeling Techniques , Chicago Illinois -Aban Aya youth project was a longitudinal efficacy trial designed to compare three interventions: School-community, Social Developmental, and Health Enhancement -Longitudinal randomized trial of 552 high-risk African-American adolescent boys in 12 high-risk classes (Chicago)

-Pre-test given at the start of grade five, and followed up at the end of grades five, six, seven, and eight

-Latent growth modelling was used to identify clusters
-Results reveal that there are three distinct “clusters” of youth in terms of growth modelling, with a “high-risk” subgroup emerging as a distinct group.

-These results suggest that there are distinct subgroups of children/youth within a high-risk sample/community
Jagers, et al., 2007 An examination of the role of mediators in the prevention of violent behaviours in the Aban Aya Youth Project, Chicago Illinois Aban Aya youth project was a longitudinal efficacy trial designed to compare three interventions: School-community, Social Developmental, and Health Enhancement Longitudinal randomized trial of 552 high-risk African-American adolescent boys in 12 high-risk classes (Chicago)

-Pre-test given at the start of grade five, and followed up at the end of grades five, six, seven, and eight

-Mediators were used to see whether communal values, self-efficacy, and empathy determined violent behaviours
-The youth in the treatment group experienced a greater growth in empathy relative to the control groups, which mediated the development of violent behaviour

A more statistically robust framework was used to examine the impact of the Aban Aya Project, a community-based prevention program for at-risk African-American adolescents residing in high-risk Chicago neighbourhoods. Using a longitudinal randomized design (LRT), the researchers used latent class growth mixture models (i.e., multilevel models) on five waves of self-report data (pre-test and four follow-up tests, given to students through the end of grade eight). By adequately dealing with a nested data structure (e.g., observations nested within the individual over time), and with missing observations (Segawa et al., 2003: 129), the researchers found that, using a violence scale ranging from zero to 21 (Segawa et al., 2003: 133), three distinct classes of children emerged. These three classes (low-risk, medium-risk, and high-risk) emerged regardless of the treatment condition, leading the authors to conclude that the program had differential effects on different clusters of individuals, and that different subgroups may, in fact, be responsive to different treatment modalities.

In a more recent study using data obtained from the Aban Aya Youth Project (AAYP), Jagers et al.(2007), examined whether changes in violent behaviours over time were mediated by the following intervening variables: community values (i.e., social cohesion), empathy, and violence avoidance efficacy (i.e., ability to exert control and goal achievement) (Jagers et al., 2007: 172). The results from their longitudinal mediation analyses suggest that youth in the treatment condition experienced a greater growth in empathy relative to the control group, and that this growth mediated the development of violent behaviour (Jagers et al., 2007: 177). And, while the study did not find that either community values or self-efficacy was significant for the AAYP group, self-efficacy was associated with the reduced likelihood of violent behaviour (Jagers et al., 2007: 177).

The results from these two studies suggest that the AAYP intervention was most effective for the group of high-risk boys (Segawa et al., 2005), and that empathy mediates the relationship between the AAYP intervention and youth violence (Jagers et al., 2007). Irrespective of the specific findings, these studies, taken together, represent how a longitudinal school-based design can be implemented in order to investigate the effect of a broader community-based cultural intervention.

The need to consider the school concurrently with the broader community was emphasized by Ronnie Casella (2002) in her qualitative work entitled “Where the policy meets the pavement: stages of public involvement in the prevention of school violence.” Using a series of semi-structured interviews from two high schools in New York State, Cassella argues that addressing school-based violence is dependent on “community-based initiatives [that are] just now becoming institutionalized in schools” (Casella, 2002: 349). This argument is based on the assumption that community-based organizations and the school are both fundamental in violence prevention programs, and that with the appropriate allocation of resources, the collaborative efforts of the social service agencies, police departments, and the school can combine to positively affect the structural determinants of violence.

Community Mobilization and Empowerment

Whereas the previous sections were devoted to evaluating community-based prevention that focused on harm reduction strategies (e.g., gun violence and gang violence), this section will assess some of the literature loosely classified under the umbrella of community mobilization and empowerment. While this terminology is shrouded by the lack of a clear definition (Welsh and Hoshi, 2002: 168–169), it is useful to consider this as a qualitatively distinct category, because the emphasis of these programs is on empowering the community as a means for improvement (Lai, 2005). And, while many of the previous evaluations did have elements of their program designed to empower members of the community (e.g., Segawa et al., 2005), these programs are considered separately because community empowerment was the primary focal point of these studies (see table 4 for an overview of these programs).

In their article entitled “Engaging Community Residents to Prevent Violence,” Bowen, Gwiasda and Brown (2004) describe how the Institute for Community Peace (ICP) was founded to address concerns over escalating rates of youth violence in the United States. The ICP views violence as a “complex phenomenon arising from individual, systemic, and societal factors” (Bowen et al., 2004: 357) and found that few resources were directed towards primary prevention. Based on the assumption that communities can be engaged to prevent violence, and can break the cycle of violence (Bowen et al., 2004: 357), the ICP presents a theory of change including creating safety, understanding violence, building communities, promoting peace, and building democracy and social justice. The authors claim that their program led to “dramatic decreases in community homicide rates, vast physical improvements....and the active engagement of community residents” (Bowen et al., 2004: 362). However, the authors do not support these claims with any description of their sample, their methodology, or the type of program they reviewed. So, while it is plausible that the ICP did, in fact, lead to decreased homicide rates and a safer community, there is no evidence to substantiate these findings. Given the fact that there was a general decline in the homicide rates throughout the country during the time in which this “study” was conducted, it is impossible to discern whether, with any degree of certainty, these findings have any merit, or whether they were part of a more general trend.

Table 4. Community Empowerment Programs

Author Name Title of Study Type of Intervention Methodology and Statistics Results and Conclusions
Bowen et al., 2004 Engaging Community Residents to Prevent Violence -The ICP (Institute for Community Peace) present a theory of change including creating safety, understanding violence, promoting peace, and building communities -No overview of the methods, statistics, sample, or the location of the sample were provided -Authors claim that their program led to “dramatic decreases in community homicide rates and vast physical improvements,” although no data was described and no evidence was provided
Lai, 2005 Responding to Asian Pacific Islander youth violence: Lessons learned from a community mobilization strategy, 2005 -Implementation of a community mobilization effort called the Community Response Plan to identify risk and protective factors for API youth -Not fully explicated, although they have cited the commitment to being “data driven” -Still tentative, as there has not been an investigation of the associated effect(s) on reducing violence Future evaluations are required
Andres Hyman et al., 2007 Oppression and Empowerment: Perceptions of Violence Among Urban Youth, New Haven, Connecticut -Case studies using community-building and improvisational role-playing -Case studies and role-playing -Youths view violence through the lens of “social ills,” and by using role playing, a sense of “hopelessness” was expressed

The National Council on Crime and Delinquency (NCD) implemented a federally funded community mobilization effort, which, beginning in 2001, was designed to prevent youth violence in Oakland, California’s Asian Pacific Islander (API) youth population. The API youth, while relatively small in terms of their overall proportion of the population, have experienced vast increases in terms of youth violence, and as a larger group, the Asian Pacific Islanders suffer from economic and linguistic disadvantages (Lai, 2005: 163–64). In an effort to address these issues, a multi-year mobilization effort called the Community Response Plan (CRP) was implemented by a group of community leaders to coordinate data analysis in order to identify risk and protective factors associated with the API. The initiative successfully facilitated a greater level of collaboration between various agencies, and the authors recommend a continued growth of the community response plan (CRP), asserting that while community mobilization requires a lot of investment and effort, it can successfully meet the larger goals of violence prevention (Lai, 2005: 178). While this may be true, future work needs to evaluate whether the theory is, in fact, associated with decreases in violence, and whether this costly intervention actually meets its mandate.

Andres-Hyman et al. (2007), in their paper entitled “Oppression and Empowerment: Perceptions of Violence among Urban Youth,” examined the implementation of a community-based empowerment program in New Haven, Connecticut that was designed to build community consensus around issues of youth violence. Using community-building techniques and improvisational theatre techniques, the authors found that youth view violence through the “lens of social ills” (154), and in particular, through economic stratification and hopelessness (154). In addition, they also found that role-playing “realistic” situations helped facilitate the expression of thought and perspective in the youth that would have otherwise been difficult to articulate, and they believe that improvisational role-play is a potential research tool that can help build community empowerment.

These studies, considered together, exemplify the type of inadequate research that frustrated Dennis Rosenbaum 20 years ago. Based largely on an ideological belief about the world rather than on any established methodological rigour, these studies leave the reader wanting. In particular, Bowen, Gwiasda and Brown’s (2004) evaluation of the Institute of Community Peace is emblematic of research that offers powerful conclusions without any accompanying evidence.

Future research evaluating community mobilization and empowerment must address these limitations, and if they expect to receive substantial amounts of public money, then there has to be some demonstration of efficacy.

Conclusion and Future Directions

When Dennis Rosenbaum initially synthesized the community-based crime prevention literature in 1988, he was disheartened by the general lack of quality in evaluation research (Rosenbaum, 1988: 381). Similarly, Welsh and Hoshi (2002) expressed their disappointment regarding the scientific rigour of the body of evaluation literature, taken as a whole (189). Despite these seemingly pessimistic conclusions, both reviews emphasized the redeeming features of community-based crime prevention, and were optimistic that future research could rectify the limitations outlined in their respective studies.

Assessing the community-based crime prevention literature over the past decade, this review found that, there is indeed, reason for optimism as many advancements have been made. In particular, Operation Ceasefire’s collaboration between the police, the broader community, and academic researchers has provided a strong framework that has become standard policy in American federally funded projects. The implementation of this “evidence-based” practice has resulted in evaluations that are grounded in methodology, and as a result, some programs have proved to be successful in preventing crime.

While more research has abided by these evaluation standards, there are still many programs, based on ideological assumptions, that are resistant to validating the efficacy of their programs. In addition, there is still too much research that fails to utilize appropriate control groups and proper statistical analyses that are essential in minimizing type 1 error.

Future research needs to rectify these limitations and continue to build on the methodological advancements that have been made over the past decade. With “36.2 per cent of global mortality due to interpersonal violence occurring among young adults aged 15-29 years” (Leschied, 2007: 41-42), it is important to understand how we can implement effective interventions to reduce youth violence. Time is of the essence.


Andres-Hyman, R. C., A. Forrester, I. Achara-Abrahams, M. L. Lauricella and M. Rowe. (2007). Oppression and empowerment: Perceptions of violence among urban youth. Journal of Community and Applied Social Psychology, 17, 147–158.

Artz, S. (2001). Community Based Approach for Dealing with Chronically Violent Under Twelve Year Old Children. Ottawa: Canada Department of Justice.

Bowen, L. K., V. Gwiasda and M. M. Brown. (2004). Engaging community residents to prevent violence. Journal of Interpersonal Violence, 19(3), 356–67

Braga, A., D. Kennedy and A. Piehl. (1999). Problem-oriented Policing and Youth Violence: An Evaluation of the Boston Gun Project. Final report submitted to the National Institute of Justice. Cambridge, MA: John F. Kennedy School of Government.

Choi, S. E., S. F. Grossman, A. P. Jacob, A. M. Spergel, E. M. Barrios, R. V. Sosa, et al. (2000). Combating gang violence in Chicago’s Little Village Neighborhood. On Good Authority, 4(2), 1–4.

Comey, J. and S. Miller. (2002). Project Safe Neighborhoods: America’s network against gun violence. U.S.A Bulletin 50(1): 11–15.

Community Social Planning Council of Toronto. (2007). Research Bulletin: Breaking down the Barriers: Opening up Ontario’s Schools for Expanded Community Use.

Cook, P. J. and J. Ludwig. (2000). Gun Violence: The Real Costs. New York: Oxford University Press.

Cullen, F. T. and R. Agnew. (2006). Criminological Theory: Past to Present. New York: Oxford University Press.

Decker, S. H., G. D. Curry, S. Catalano, A. Watkins and L. Green. (2005). Strategic Approaches to Community Safety Initiative (SACSI) in St. Louis. Rockville, MD: National Institute of Justice.

Easterling, D., L. Harvey, D. Mac-Thompson and M. Allen. (2002). Evaluation of SACSI in Winston-Salem: Engaging the Community in a Strategic Analysis of Youth Violence. Rockville, MD: National Institute of Justice.

Egley, A. and C. E. Ritz. (2006). Highlights of the 2004 National Youth Gang Survey. U.S. Department of Justice: Office of Juvenile Justice and Delinquency Prevention.

Esbensen, F. (2001). The national evaluation of the Gang Resistance Education and Training (G.R.E.A.T.). In J. Miller, C. L. Maxson and M. W. Klein (Eds.), The Modern Gang Reader (2nd ed., pp. 289–302). Los Angeles: Roxbury

Esbensen, F. A. and D. W. Osgood. (1999). Gang Resistance Education and Training (GREAT): Results from the national evaluation. Journal of Research in Crime and Delinquency, 36(2), 194–225.

Esbensen, F. A. (2000). Preventing Adolescent Gang Involvement. Bulletin. Washington: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention.

Farrington, D. P. (2003). Methodological quality standards for evaluation research. Annals of the American Academy of Political and Social Sciences, 587, 49–68.

Farrington, D. P. (1986). Age and crime. Crime and Justice, 7, 189–250.

Harper, S. (2008). Office of the Prime Minister website. Available online at:

Howell, J.C. (2000). Youth Gang Programs and Strategies: Summary. Washington: Office of Juvenile Justice and Delinquency Prevention: National Youth Gang Center.

Hyman, R. C., A. Forrester, I. Achara-Abrahams, M. L. Lauricella and M. Rowe. (2007). Oppression and empowerment: Perceptions of violence among urban youth. Journal of Community & Applied Social Psychology, 17, 147–158.

Jagers, R. J., A. A. Morgan-Lopez, T. L. Howard, D. C. Browne and B.R. Flay. (2007). Mediators of the development and prevention of violent behaviors. Prevention Science, 8(3), 171–179.

Kellermann, A. L., D. uqua-Whitley and C. S. P. Parramore. (2006). Reducing Gun Violence: Community Problem Solving in Atlanta. Rockville, MD: National Institute of Justice.

Klofas, J. M., C. Delaney and T. Smith. (2007). Strategic Approaches to Community Safety Initiative (SASCI) in Rochester, NY. Rockville, MD: National Institute of Justice.

Kuhn, E. M., C. L. Nie, M. E. O’Brien, R. L. Withers, G. J. Whitemute and S. W. Hargarten. (2002). Missing the target: A comparison of buyback and fatality related guns. Injury Prevention, 8, 143–146.

Lai, M. H. (2005). Responding to Asian Pacific Islander youth violence: Lessons learned from a community mobilization strategy. Crime and Delinquency, 51(2), 158–179.

Lenoir, J. (2002). Using community resources in gun violence reduction initiatives. U.S.A Bulletin, 50(1), 30.

Leschied, A. W. and C. L. Heng. (2007). Effective Programming in Ontario Youth Justice: The Challenge Posed by Gangs, Reingegration from Custody to Community and Short Term Stays in Custody. Working Paper. The University of Western Ontario.

Mann, R. M., C. Y. Senn, A. Girard and S. Ackbar. (2007). Community-based interventions for at-risk youth in Ontario under Canada’s Youth Criminal Justice Act: A case of the “runaway” girl. Canadian Journal of Criminology and Criminal Justice, 49, 37–73.

McGarrell, E. F. (2005). Strategic Problem Solving and Project Safe Neighborhoods. School of Criminal Justice, Michigan State University: Working Paper.

Ramsey, A. L., J. O. Rust and S. M. Sobel. (2003). Evaluation of the gang resistance education and training program. Education, 124(2), 297–309.

Roehl, J., D. P. Rosenbaum, S. K. Costello, J. R. Coldren, A. M. Schuck, L. Kurnard, et al. (2006). Strategic Approaches to Community Safety Initiative (SACSI) in 10 U.S. Cities: The Building Blocks for Project Safe Neighborhoods. Rockville, MD: National Institute of Justice.

Rollin, S. A., C. Kaiser-Urley, I. Potts and A. H. Creason. (2003). A School-Based Violence Prevention Model for At-Risk Eighth Grade Youth. Psychology in the Schools, 40(4), 403–416.

Rosenbaum, D. P. (1988). Community crime prevention: A review and synthesis of the literature. Justice Quarterly, 5(3), 323–395.

Sampson, R. J. and W. J.Wilson. (1995). Toward a theory of race, crime and urban inequality. In J. Hagan, J. and R. Peterson (Eds.) Crime and Inequality (pp. 37–54). Stanford: Stanford University Press.

Sechrest, D. K., P. J. Schram, G. Bichler, S. Tibbetts and J. Worrall. (2003). Second Biannual Evaluation Report for the Riverside County Probation Department Juvenile Crime Prevention Act Program. San Bernardino, CA: California State University, Center for Criminal Justice Research.

Segawa, E., J. E. Ngwe, Y. Li and B. R. Flay. (2005). Evaluation of the effects of the Aban Aya Youth Project in reducing violence among African American adolescent males using latent class growth mixture modeling techniques. Evaluation Review 29(2), 128–148.

Shaw, C. R. and R. McKay. (1942). Juvenile Delinquency and Urban Areas. Chicago: University of Chicago Press.

Sherman, L. W. (2000). Gun carrying and homicide prevention. Journal of the American Medical Association, 283(9), 1193–1195.

Sherman, L. W., D. P. Farrington, B. C. Welsh and D.L. MacKenzie. (2002). Preventing crime. In L. W. Sherman, D. P. Farrington, B. C. Welsh and D. L. MacKenzie (Eds.), Evidence-Based Crime Prevention (pp.1–12). New York: Routledge.

Spergel, I. A. (2007). Reducing Gang Violence: The Little Village Gang Project in Chicago. Chicago: Alta Mira Press.

Spergel, I. A. and S. F. Grossman. (1997). The little village project: A community approach to the gang problem. Social Work, 42(5), 456–470.

Spergel, I. A., K. M. Wa, S. F. Grossman, A. Jacob, S. E. Choi, R. V. Sosa, et al. (2003). The Little Village Gang Violence Reduction Project in Chicago. Chicago: University of Chicago, School of Social Service Administration.

Thrasher, F. M. (1936). The Boys’ Club and juvenile delinquency. American Journal of Sociology, 41, 66–80.

Welsh, B. C. and A. Hoshi. (2002). Communities and crime prevention. In L. W. Sherman, D. P. Farrington, B. C. Welsh and D. L. MacKenzie (Eds.), Evidence-Based Crime Prevention (pp.165–197). New York: Routledge.

Westmacott, R., Y. Stys and S. L. Brown. (2005). Selected annotated bibliography: Evaluation of gang intervention programs. Research Branch: Correctional Services 4.

Wilson, J. Q. and G. L. Kelling. (1982, March). Broken windows. Atlantic Monthly, 249(3), 29–38.

Wortley, S. and J. Tanner. (2006). Criminal Organizations or Social Groups? An Exploration of the Myths and Realities of Youth Gangs in Toronto. Unpublished paper. Toronto: Centre of Criminology, University of Toronto.

23 This chapter was written with the assistance of Steve Cook, Ph.D candidate, Department of Sociology, University of Toronto.


Volume 1. Findings, Analysis and Conclusions

Volume 2. Executive Summary

Volume 3. Community Perspectives Report

Volume 4. Research Papers

Volume 5. Literature Reviews