Introduction

A rigorous, comparative evaluation of the CBVP demonstration was not possible due to the varied strategies employed by each of the five cities, the absence of geographically and age specific data about violence, and the study’s inability to control for the variety of external influences that may have affected crime trends in the program target areas (e.g., police actions, competing services from other agencies, increased gang activity, etc.). Conducting rigorous and controlled evaluations of community-based crime-reduction efforts requires the ability to address these factors from the very beginning. Programmatic funding should be tied to strict guidelines designed to support research goals, including a singular intervention model across sites and mandatory data collection and submission procedures.

On the other hand, greater research control over routine program activities in cross-site evaluations inevitably hinder each site’s ability to adjust for changing circumstances and would likely frustrate the designers and managers of local programs. Allowing for too much local flexibility, however, prevents cross-site evaluations from generating defensible results. The findings presented in this report suggest that the CBVP demonstration program may have led to positive changes in the communities involved, but evaluators were unable to tie those changes to the demonstration. Instead, the evaluators were forced to rely upon staff interviews and direct observations to build anecdotal support for the study’s main conclusions.

BROOKLYN

The CBVP project in Brooklyn focused on a section of Crown Heights, a rapidly-gentrifying neighborhood with a history of gang violence and crime. The nonprofit Center for Court Innovation (CCI) implemented the project through its Crown Heights Community Mediation Center, including an adaptation of the “Cure Violence” public health strategy for violence reduction. The program, “Save Our Streets (SOS) Crown Heights,” relied on an array of activities aimed at changing community norms and “interrupting” the transmission of violent behaviors. A key, but controversial, feature of the Cure Violence model is its minimal coordination with law enforcement. Staff members in the SOS program tried to maintain a collegial relationship with the local precinct, but the Cure Violence strategy hinges on outreach workers and violence interrupters who are deemed “credible messengers” for the young people most at risk of violence. Staff members in the program were often former gang-involved and previously incarcerated people.

The SOS staff worked closely with partners in local schools, churches, other community organizations, and local businesses to spread anti-violence messages and connect with young people. Their goal was to spread awareness about the consequences of violence throughout the neighborhood. Program staff focused specifically on the young people most involved in gun violence (as victims and perpetrators), through mapping patterns of violence and providing critical incident responses, hospital visits, intervention in disputes, etc. Case managers connected those young people with education, counseling, housing assistance, and other services. With CBVP funds received in 2010, the SOS catchment area focused on a 40-square block area and the young people (mostly 16-25 years old) from that area who met key criteria for the risk of violence.

When the Crown Heights program first began to work in the community, the staff encountered considerable mistrust from neighborhood residents—ironically because of suspicions that the program was too closely aligned with police. Over time, the program developed more legitimacy with residents through perseverance and its continual neighborhood presence.

The SOS strategy hinges on the neighborhood compatibility and credibility of its outreach workers and violence interrupters. Identifying and selecting the right staff was a constant challenge. The most effective staff members were people with direct, personal experience with gangs and gun violence. Senior staff suggested to researchers that the program could be more effective if staff members had greater access to professional support to deal with their own life challenges, past experiences, and job-related stresses. Recognizing the difficulties faced by staff members and the complexity of their work, the Brooklyn program reduced its worker caseload expectations midway through the CBVP project grant. The program also increased its focus on public responses to shootings and worked to strengthen the consistency of its approach, balancing the need for a quick response with the need to ensure that all relevant stakeholders were involved.

The project’s extensive use of posters, billboards, and public meetings seemed to pay off. Community residents expressed considerable awareness and appreciation of anti-violence messages. One frequent frustration—lack of sufficient social services—sometimes reflected lack of knowledge about available resources, and SOS events helped to bridge this gap. Some program staff found it challenging to shift activities and tone between their “crisis response” efforts after shootings to their more ongoing, less dramatic public education work. The fact that the SOS network successfully connected youth with jobs, education, and other services—even though this was often beyond the scope of SOS’ formal tasks—is evidence of the staff’s dedication and position of respect in the community.

Despite its many challenges, the Brooklyn project managed to develop a strong presence in the Crown Heights neighborhood during the CBVP project. Some people with knowledge of the effort attributed the program’s success to the skills, support, and clarity of vision provided through ongoing training for SOS staff—including time that staff members spent with personnel from the Cure Violence headquarters in Chicago and other New York sites implementing the same model. Simultaneously, the need for model fidelity and close coordination with other New York Cure Violence sites limited the nimbleness of the Crown Heights program.

The evaluation project examined the effectiveness of the Brooklyn program in several ways, including analyses of violent crime trends in Crown Heights after 2010 compared with an area near Crown Heights that was not served by Cure Violence. This study failed, however, to detect a consistent pattern of declining crime that could be attributed to the CBVP project. Homicides and violent crime arrests declined in Crown Heights and the comparison neighborhood. Non-gun arrests fell more in Crown Heights than in the comparison area, but there could be many reasons for this difference. Towards the end of the grant period, local residents apparently believed that they were better able to handle disputes without violence and they appreciated the support and persistence of the SOS staff. The evaluation, however, did not find strong evidence of impact in Brooklyn, either from the analysis of crime trends or from the quasi-experimental outcome evaluation using household surveys.

DENVER

Denver used its CBVP grant to implement a program called the Gang Reduction Initiative of Denver (GRID). The GRID project drew many key principles from the Comprehensive Gang Model (CGM), including a focus on employment and educational pathways out of violence for at-risk youth, increased social supports for youth in gangs, and the assistance of law enforcement for suppression when required (loosely based on the focused deterrence model). Facing the need to coordinate with other organizations already involved in this sector, Denver ultimately applied a more eclectic model with some elements of the CGM and some additional elements—police work (suppression), gang intervention, and primary/secondary prevention (including case management and outreach done by several other agencies).

Denver relied on the CBVP project to leverage about $10M in additional services related to violence prevention, although the framework of how the numerous activities fit together was not always explicit. The GRID project focused on working in three communities. A smaller team focused more closely on 20-25 families with multiple people involved in the justice system and/or gangs, and attempted to offer case management and other services in a way that accounted for the family dynamics.

The fluidity of the CBVP effort in Denver led to implementation difficulties. For example, the call-in meetings—a key feature of focused deterrence—initially stumbled due to lack of precision in the convening and messaging. Agencies involved tried their best to apply the suppression tactics—such as penalizing entire groups for actions of some members—and to communicate the strategy to up-and-coming youth in gangs. GRID, working with numerous partner agencies, attempted to combine social services and outreach with suppression (law enforcement threats). This approach was not always successful. Nonetheless, the hybrid approach resulted in much better working relationships across agencies—particularly police, social services, and nonprofits—and increased sharing of sensitive information.

GRID reached over 3,000 youth during the project period. Primary activities included conflict mediation, crisis response, counseling, victim services, and mentoring. Close coordination with outreach staff, halfway houses, police, and probation was also essential. Separate intervention teams offered tailored services to juveniles (14 to 21 years old) and people reentering the community from prison. The project also offered educational and employment opportunities and mental health services through other city agencies with some of its grant funding.

The available indicators of Denver’s effectiveness were mixed at best. Homicides in Denver dropped from 2004 to 2014, but the largest declines occurred simultaneously with program implementation in 2011. Violent incidents increased slightly as did gang arrests. Because arrests reflect police decisions and tactics, however, rising arrests do not necessarily suggest an effect (positive or negative) of GRID. No other implementation data were available for analysis. During the project period, Denver shifted from an emphasis on violence-intervention (drawing on focused deterrence and mediation) to a broader array of preventive services. The diffused approach made it difficult to track activities and effects, and there was less accountability on individual organizations for results. The lack of geographically specific data about program implementation exacerbated these problems. On the other hand, the GRID program seemed to make a difference in how city agencies in Denver worked with one another. High-level involvement across organizational sectors catalyzed new and improved relationships where previously there had been no contacts or where key partners had been openly antagonistic.

OUTCOME EVALUATION: BROOKLYN AND DENVER

Household surveys measured attitudes and perceptions of violence among residents in Brooklyn and Denver, and compared changes in those perceptions with two matched comparison areas in each city. The first survey in both cities took place in 2012, approximately one year into CBVP programming; the second survey took place in 2014, after nearly three years of programming. Surveys were designed to detect shifts in attitudes and to test for their association with program activities.

The results in Brooklyn failed to show positive changes on several key indices. In fact, neighborhood safety appeared to be worse in 2014 than 2012 in the program area. The analysis suggested that perceptions of violence had changed in the opposite direction than that intended by the program. At the same time, there were positive changes in the comparison area—for reasons unknown to the research team.

The Denver survey results showed some positive outcomes, with improvement in key scores in the program area but not in the comparison area. Even these improvements, however, were not statistically significant, although the analysis did detect increased knowledge of the program in the CBVP site.

The mixed results likely reflect the challenge of measuring changes in attitudes toward violence after a relatively brief and modest program intervention. Given that the identified problem was an entrenched “normalization” or tolerance of violence, it is not surprising that there would be limited, detectable evidence of change after just two years. Surveys focusing on attitudes also grapple with the subjective nature of perceptions. The mere existence of a violence-prevention program may heighten residents’ awareness of the extent of, or consequences of violence—which may alter resident attitudes and perceptions negatively, even if the program itself is beginning to have positive effects. Additional research that uses a longer timeframe to capture differences in program exposure and that may account for awareness of violence versus personal inclinations toward violence could provide more meaningful results.

OAKLAND

Oakland’s CBVP demonstration project operated from 2010 to 2014 and addressed gang violence, particularly among African-American young men. Oakland’s Department of Human Services led implementation of the program, building on prior programs that focused on gun violence and reentry services. In 2010, Oakland was part of a California Cities Gang Prevention Network and had its own Gang Prevention Task Force. For the CBVP project, Oakland blended elements of two established program models: Cure Violence (Chicago) and focused deterrence. City officials named their combined program “Oakland Unite.”

The Oakland Unite program identified targeted populations by both geographic and individual criteria related to various risk factors, gang affiliation, prior gun violence experience, and age. The target areas shifted during the program in response to new violence data. The program applied five central strategies: street outreach; crisis response; reentry and job support; public education; and “call-ins” (according to the focused deterrence model). The DHS coordinated numerous nonprofit groups to implement these strategies and relied heavily on inter-agency collaboration to keep them operating.

The outreach component involved “credible messengers” who could encourage young people to avoid violence, especially retaliatory violence. Outreach workers liaised with families and communities, managed case records, and coordinated with other agencies. Consistent with the Cure Violence model, the Oakland Unite program also used crisis response teams to engage young people in the aftermath of a violent incident. Public education campaigns promoted anti-violence messages and encouraged pro-social attitudes. A “Messengers for Change” campaign became widespread in print and media, as well as through public events. A “call-in” component—drawn from the focused deterrence model—offered positive incentives (food, stipends) to gang-involved young people and threatened legal consequences for those persisting in violent actions.

Homicides in Oakland declined during the program period. Shootings fell 15 percent and homicide dropped 30 percent from 2012 to 2014. This came after a spike in violent crime in the first part of program implementation (2011 and 2012), particularly in robbery. Many government officials gave credit to the “single voice” influence of the Oakland Unite program (especially the call-ins), but the available data did not allow for a more rigorous analysis of the program’s effects on crime. An analysis of the intensity of street outreach activities (amount and frequency) suggested that there was an association between intense outreach and greater reductions in shootings and homicides. All evidence, however, was suggestive rather than conclusive.

WASHINGTON, DC

The CBVP demonstration project in Washington, DC was called Creating Solutions Together (CST). Staff operated in the Columbia Heights and Shaw neighborhoods of central Washington from 2010 to 2013. The project’s goal was to reduce violent crime involving youth from the target areas. The nonprofit organization was known as the Columbia Heights / Shaw Family Support Collaborative (or “The Collaborative”) at the start of the grant period and led the development and implementation of the program. (The organization was later renamed Collaborative Solutions for Communities.) The CST program built on the foundations set by the Gang Intervention Project (GIP), which the Collaborative also implemented in the same area more than five years prior to the CBVP grant.

The GIP mapped shootings and gang activities, and worked with police and other organizations to persuade youth to embrace non-violent lives. The CST program retained this core component, including mediating gang disputes and “critical incident responses” by coordinated teams of outreach workers and police officers. The program also focused on preventing retaliatory violence. More broadly, CST tried to shift community norms regarding violence and youth perceptions of the costs of violence, and to support life-course alternatives for youth already involved in violence.

The program involved extensive outreach work, especially focusing on hard-to-reach youth who were not already part of their family-based social services. Outreach work allowed the Collaborative to build trust with youth, form a better understanding of gang dynamics, and connect youth with services and supports. The CST program worked closely with school and hospital personnel, often triangulating information about at-risk youth and strategizing tactics to ensure that youth stayed involved in pro-social activities, including GED classes, vocational skills training, and therapy or counseling. Case managers at the Collaborative integrated family-oriented services with interventions tailored for young people already involved in violence. Through CST, the Collaborative also formally certified outreach workers and trained dozens of other Washington, DC nonprofit and city government agencies in youth violence prevention skills.

Violent crime trends in the targeted areas generally showed a decline from 2006 to 2014, but it was not possible to infer a direct effect of program activities. Many of the positive indicators of falling crime in the program neighborhood were also observed outside the program area, though the decline did appear to be steeper in the targeted areas. The visibility of gang violence declined in the program area, however, and many staff members expressed confidence that their efforts had made a difference for neighborhood youth and for the safety of the entire community.

NEWARK

The City of Newark’s CBVP demonstration project began in 2013 and was called Newark United Against Violence (NUAV). Newark’s Office of Reentry and the Newark Police Department shared project leadership roles—an arrangement that enabled creative collaborations and also generated some obstacles due to divergent styles. As in other cities, Newark drew from several other established violence-prevention models, primarily drawing on focused deterrence, public health (Cure Violence), and hotspot policing. These components were linked with an assortment of other social and economic inclusion services. Previously, Newark had implemented several other programs targeting gun violence and gang activities in the city. The police component, led by the NPD, involved using more detailed violence data to deploy officers to “hot spots.” The idea was that officers would use both community policing methods (building relationships with individuals and organizations) and also apply focused deterrence tactics of group-level law enforcement penalties for individual violations. While the NPD often accomplished elements of both approaches, at times they were in contradiction, and sometimes there were insufficient or misaligned police resources for the method that was needed.

The NUAV program also held community “mobilization” events that served as a platform for local leaders to convey messages against violence and to remind youth about the potential consequences of gun violence. Like other programs, NUAV relied heavily on outreach workers, who found and engaged the most at-risk youth, connected them with services, and spread a message of non-violence. Outreach workers also had a mediation and crisis response role; they learned about local dynamics in order to be able to intervene before, during, or after a violent confrontation. Finally, NUAV partnered with two nonprofit organizations—the Newark Conservancy and Newark Community Solutions—that provided job training, apprenticeships, case management, counseling, and other social services. NUAV focused its programming in the South and Central Wards of Newark and on young adults 18-30 years old who had some involvement in violence. The program also accepted some youth who were referred through the youth court, which worked closely with Newark Community Solutions.

Newark, like other cities, found it challenging to combine multiple strategies that were sometimes incompatible. This was most evident in the role of the police. Hotspot policing, on the one hand, involves greater presence and involvement of the police in the community. Mediation and violence interruption, on the other hand, involves local residents with past system experience and street credibility, and often requires some separation from police involvement. Focused deterrence involves even larger roles for police and prosecutors, with an emphasis on certain groups or individuals rather than a geographic location. In some cases, these three strategies complemented one another, or affected different youth. In other cases, these differences led to disagreements and difficulties in coordinating strategies among various organizations and staff, most notably in the area of sharing information about gang-related individuals or incidents with police. The lack of clarity in the overall vision and model for the program also posed challenges for consistent messages and branding in the public education and community mobilization activities.

Due to the timing of electoral cycles, NUAV also faced implementation challenges due to staff turnover in city offices, including turnover of the original project team, and some shifts in local budget allocations. As a result, there were gaps in activities and documentation, and some services—most notably, the green jobs initiative—ended earlier than planned, which disappointed participants.

It was nearly impossible to assess the effects of Newark’s efforts under CBVP. The City maintained no data on crime specific to the program area or in a timeframe needed to conduct a meaningful analysis of program efforts. Newark experienced a dramatic decline in homicides before and during program implementation, but this was observed at the city level and across areas without any NUAV activities. Program staff believed that collaboration across Newark agencies had improved during the CBVP effort and that community perceptions of the program had evolved from skepticism to appreciation. It was not possible, however, for the evaluation to ascertain whether these positive developments contributed to increased public safety.

Final Thoughts

John Jay College’s evaluation of OJJDP’s CBVP Demonstration project found a number of positive outcomes over the course of the project. Each city involved in CBVP worked in earnest throughout the grant period and did as much as possible with their given resources. It is not possible to say that one city took a better approach than another (i.e., choosing to replicate a known model vs. adapting a model to the city’s circumstances). All cities were able to demonstrate at least some successes and believed that they were making a positive impact on the violence in their communities. Some cities were better able to document their successes than others, however, and some cities appeared more capable of continuing their successes after the cessation of OJJDP funding.

The findings presented in this report suggest that the CBVP demonstration program may have led to some positive changes in the communities involved, but those changes are based on program staff interviews and evaluator observations alone. When the study analyzed crime data from each city, and when the survey-based outcome evaluation component is considered, it is not possible to identify consistent effects of CBVP on youth violence and public safety.

_______________

Recommended Citation

Tomberg, Kathleen A. and Jeffrey A. Butts (2016). Street by Street: Cross-Site Evaluation of the OJJDP Community-Based Violence Prevention Demonstration Program. New York, NY: Research and Evaluation Center, John Jay College of Criminal Justice, City University of New York.

[ references ]