Durham Youth Gang Strategy
Building the Evidence - Evaluation Summaries 2012-ES-32
Gang activity in the Durham Region was first recorded in 1989. In 2006, Durham police data indicated a growing number of self-identified, gang-involved youth in the Pickering/Ajax/Oshawa region. Ten street gangs were known to operate in Durham and approximately 21 of Toronto's street gangs had at least one member living in the Durham region. Police and criminal justice stakeholders were concerned that youth at the Brookside Youth Centre in Durham had reported learning from each other about gang-related activities.
Durham's gangs range from neighbourhood gangs to highly structured gangs connected to criminal organizations like the Hell's Angels.
In response to the growing gang problem in Durham, the National Crime Prevention Centre (NCPC), provided $1.5 million to launch the Durham Youth Gang Strategy (DYGS). The program started on April 2008 and operated until March 2011.
2. Project Description
The DYGS is a comprehensive program developed by the Durham Family Court Clinic (DFCC) and the Murray McKinnon Foundation (MMF), in collaboration with community agencies to support prevention and intervention activities. The DYGS is based on the Office of Juvenile Justice and Delinquency Prevention (OJJDP)'s Comprehensive Gang Model, also known as the Spergel Model. It takes a three-pronged approach that encompasses prevention, intervention and suppression activities.
The Spergel Model presumes that gangs become chronic and serious problems in communities where key organizations are inadequately integrated and there are not sufficient resources available to target youth at risk of gang involvement and youth involved in gangs. To address these problems, the Comprehensive Gang Model calls for community institutions – including law enforcement, social welfare agencies, and grass roots organizations – to work together to achieve a more integrated, team-oriented approach.
A one-to-one approach was used to implement therapeutic interventions such as cognitive behavioural therapy, solutions-focused therapy and group work. Youth were paired with positive role models (trained professionals) to encourage a healthy adult relationship. The youth worker met with the youth and parents weekly or biweekly, helped them develop psychosocial goals and advocated on their behalf to receive necessary resources. Another component of the program was to get youth involved in their community in a meaningful way by becoming an educator or mentor for other youth.
Upon graduation, certain youth would be asked to be role models. The program also aimed to strengthen and develop partnerships by inviting social service agents as well as lawyers, judges, police and non-traditional services to community forums. These served to educate the public on gangs in general, such as the consequences of joining a gang, and to discuss the local gang problem.
The primary components of the program include:
Mobilizing local residents, youth, community groups, civic leaders and agencies to plan, strengthen, or create new opportunities and linkages to existing organizations for gang-involved and at-risk youth; coordinating programs and services as well as the functions of staff within and across agencies.
Providing programs and social services to gang-involved youth and those at high risk of gang involvement; also, using DYGS workers to actively engage gang-involved youth.
Providing and facilitating access to educational, training and employment programs or services targeted to gang-involved youth and those at high risk of gang involvement. DYGS workers assist youth involved in the program by advocating for and supporting them through opportunities for educational, training, recreational and employment programs that may not have otherwise been available.
The DYGS staff works closely with the police, probation case managers, child protection workers, and other service providers to provide youth with close support and encouragement.
Organizational change and development
Facilitating organizational change and development to help community agencies better address gang problems through a team problem-solving approach consistent with the philosophy of community and problem-oriented policing. Developing and implementing policies and processes that result in the most effective use of available and potential resources within and across agencies.
The primary goals of the program were to help youth increase their awareness of the consequences of gang involvement, to encourage a less positive attitude toward gangs, and to decrease the risk factors that make them prone to joining gangs.
DYGS targeted youth, between 12 and 18 years of age, at risk of being involved in gang activity, already involved in gang activity, or attending high-risk schools. DYGS also worked with the parents of at-risk youth, parents in high-risk communities, and adults and community agencies, organizations and services.
According to the pre-test surveys (n = 72), the program engaged youth and young adults aged 13 to 20 years old. Of those who attended the DYGS program, 92% were males, 97% reported English as their first language, and 92% were Canadian citizens. Half of all participants identified themselves as Black (52%), while 21% identified themselves as mixed-heritage, and 21% as White.
Over three-quarters of the youth (85%) had been charged with a criminal offence. When youth were asked how many charges they had, their answers ranged from one to 50 charges. Break and entry was the most significant charge (25%), as well as assault (25%). Other common criminal charges included unarmed robbery (15%), theft (8%) and narcotics (8%).
Since the program began in April 2008, DYGS received 341 referral-based inquiries and 108 referrals were made for gang-prevention programming. Of these 108 referrals, 95 youth at risk of gang involvement and youth involved in gangs had enrolled in the DYGS program. The other 13 youth were discharged after the referral was made due to a number of factors. Of these 95 youth who received services, 72 youth consented to participate in the evaluation (a response rate of 76%) and completed the pre-survey instruments.
In addition, DYGS provided school-based Gang Awareness Workshops to 3,095 students in high-risk schools in the Durham Region, educating youth about the risks of youth joining gangs. DYGS reached community members by providing three Youth Gang Forums. A total of 1,014 community stakeholders attended these three forums, along with 60 members of the general public interested in the presence of gangs in the Durham Region.
3. Evaluation of the Program
An independent process and outcome evaluation was conducted over a three-year period (2008-2011) to determine if the DYGS had achieved its objectivesFootnote 1. The evaluators employed a single group repeated measures design. Youth were tested at three points: pre-program, mid-program (three months) and post-program (six months). Focus groups were also conducted with youth, parents and stakeholders to qualify the quantitative results.
The evaluators determined that it was not feasible to have a comparable control group during the first year of the evaluation study. To control for threats to validity, the evaluators assessed how risk levels and relevant treatment contributed to change in the outcomes of interest. Groups of youth with lower risk levels and lower dosage levels acted as proxy comparisons.
Descriptive statistics were used to provide general information about demographics, risk levels and frequency of activities to measure dosage. The evaluators attempted to develop a fidelity tool to identify key program activities that needed to be implemented according to the work plan. The fidelity tool was also used to increase internal validity. The evaluators tracked external events that could have impacted the outcomes in order to factor these events into the statistical analyses. A tracking tool was developed to document and incorporate program dosage.
Semi-structured interviews were conducted with DYGS program staff and selected youth to provide an in-depth account of implementation issues. Interviews with staff were conducted at the program's office location. Interviews with youth were conducted in the community. They ranged from 30 to 60 minutes in duration.
During semi-structured interviews, the evaluators used a list of open-ended questions with each respondent. Based on the respondents' answers, the evaluator adapted the questions accordingly. The interviews were audio-taped, transcribed, and then imported into software for the qualitative analysis which identified themes through a grounded theory approach.
Outcome measures focused on assessing changes in anti-social attitudes towards law, the courts, the police, criminal activity, and peer connections to gangs. The measures examined changes in levels of aggression, substance abuse and emotional connections to family members. Behaviour-related measures included re-arrest rates, gang involvement and gang delinquency.
The research instruments developed used questions from the Comprehensive Gang Strategy, OJJDP survey and the Eurogang survey. The instrument package included a number of standardized and non-standardized tools.
Youth were given the choice of self-administering the survey for privacy reasons, in which case, they were instructed to put the completed survey in an envelope and seal it. They were also given the option to complete their survey with a DYGS worker.
Of the 72 youth who completed pre-test surveys, 34 (47%) completed the mid-test (six months), 28 (39%) the post-test survey (typically 12 months after the pre-test), and no youth completed the follow up survey (18 months after the pre-test).
4. Evaluation Findings
The DYGS was able to reach its intended target group. Of the 51 youth in the program who were deemed at risk of gang involvement, scores indicated that all youth had met the minimum risk score to be eligible for the program. Six youth (12%) were deemed low-risk, and more than half of the youth at risk of gang involvement (63%) were deemed moderate-risk at entry into the program. Thirteen youth (25%) were deemed to be at high risk for gang involvement.
The average cost per youth was $12,720 or $385 per week, based on 95 participants, with an average of 33 weeks in the program. Of the 108 referrals plus 341 referral-based general inquiries, there were a total of 449 total referrals whose average cost per referral based inquiry was $2,691.
Awareness and Beliefs
Youths' perception of criminal activity
Youths' positive perceptions of criminal activity and delinquent peer relationships were reduced during the program. However, evidence suggests that the change process was slow. Qualitative interviews suggest that although engagement was an important factor for the success of youth in the program, building trust with staff took time and required substantial energy and commitment by staff to foster the positive relationships that helped facilitate the change.
Perceptions of the police
Youths' perceptions of the police, the courts and the laws remained unchanged.
Attitudes towards criminal activity
For youth who completed the program, the results show a significant reduction (p < .05) in positive attitudes towards criminal activity with a moderate effect size (ES = 0.30). Again, the data suggests that youth perceptions of criminal activity did not change quickly, as this change occurred six months after being in the program. The Post-hoc Pairwise Comparisons analysis shows that there was a significant mean difference between pre and post measures (MD = -3.348, SE = 1.02, p < .010).
Risk and Protective Factors
Based on self-report, there was a significant increase in the number of youth attending school at the post-test. In qualitative interviews with DYGS staff, they indicated that it was difficult to place youth involved in the DYGS program in schools, especially in cases where youth experienced disciplinary action. DYGS staff stated that they worked diligently to place the youth in alternative programs despite these challenges. The findings support a positive effect for these efforts given the positive change in the youths' school attendance at the post-test.
Youth interaction with family members
The results show a significant increase in youths' closeness with their family members at the completion of the program with moderate effect (ES = 0.39, p < .05). Post-hoc Pairwise Comparisons analysis shows that the significant mean difference change occurred between pre- and mid- measures (MD = -2.565, SE = .913, p < .05) and between pre- and post- measures (MD = -3.261, SE = .894, p < .05).
This suggests that youths' connections with their family members changed significantly within the first six months of their involvement with the program. Although DYGS was unable to recruit parents into the program, it seems that youths' perceptions of their families improved even without parental participation. This is a surprising finding, especially since family closeness scores were not significantly predicted by factors related to the youths' level of family support, weekly attendance or the number of program sessions completed. The reasons for this positive change in family closeness remain unclear.
The findings suggest that the program was not able to contribute to an increase in youth employment opportunities by the end of the program. Simply asking youth if they had employment may not have been a viable way to determine if the program had helped youth find employment. Consequently, additional indicators of employment were developed during the evaluation. At the post-test, 33% (six of the 19 youth) reported receiving employment training during the six month time period.
Peer association with gang members
The results show a significant reduction in youths' closeness with peers involved in gangs (p < .05) with a moderate effect (ES = 0.38). Post-hoc Pairwise Comparisons analysis shows that the significant mean difference change occurred between pre- and post- measures (MD = -2.056, SE = .739, p < .05), suggesting that youth disconnection with peers involved in gangs also changed slowly while in the program.
Post-hoc Comparison analysis also showed a significant mean difference for youth at risk of gang involvement compared to youth involved in gangs (MD = -2.270, SE = .703, p < 0.05). This supports the need for a differential response when working with both youth at risk of gang involvement and those involved with gangs.
Findings demonstrate that the program did not have an impact on gang involvement. These results must be viewed with caution given the small sample size available at the post-test. These findings suggest that membership in Durham Region gangs may be more stable than transient, making it potentially difficult to help youth leave a gang once they are involved.
Drug and alcohol use
Overall, youth reported that their drug of choice was marijuana. A statistically significant finding of less drug use was found from pre-test to mid-test but no difference was found from pre-test to post-test. Differences in alcohol use by youth at the end of the program were not statistically significant.
To measure potential changes in delinquency, a pre- and post-test assessment of youth with prior charges (at least six months prior to completing the surveys) was conducted. The change in the number of charges at the post-test was not significant. However, assaults dropped from 18% to 7% and break-and-enter dropped from 18% to 1%. These findings were triangulated with the interviews with the youth which suggested that additional charges while in the program were mostly due to breaches of probation.
The findings of this evaluation are subject to a number of limitations. Most importantly, the evaluators were not able to create a feasible comparison group. However, the team used multivariate analysis to address possible threats to internal validity related to confounding variables. The lack of a comparison group reduced the confidence related to isolating the program's contribution to the outcomes being assessed. The sample size was also too small to administer the 18-month (after the pre-test) follow-up tests after program implementation.
5. Lessons Learned and Recommendations
The differential trajectory of youth at risk of gang involvement and youth involved in gangs highlights the need to provide two separate programs. The dual target group presumes that the project will be able to provide a tailored program to address the varying needs of these two groups. Only one program was provided for the two different target groups.
Separately tracking and documenting the progress of youth at risk of gang involvement and youth involved in gangs would provide a better understanding of the unique experiences and needs of these target populations. This would allow the monitoring of programs developed specific to the unique needs of the youth involved in the program.
DYGS staff reported that a number of sessions with youth have included text-messaging and emails between staff and youth, but these were not being reported as “participation” given that this was not face-to-face contact. Incorporating new technology into program implementation and evaluation should be considered. The nature of the “texting” however should demonstrate a link between the “treatment” and “risk” factors identified by case managers.
Results of the process evaluation demonstrate that 12 youth per worker is an appropriate caseload to provide weekly sessions with each youth for the program duration (one year). With three staff, each with a caseload of 12 youth, the maximum number of youth included in the program should not exceed 36 youth for any given year.
Based on the success of the community and public forums to increase community members' knowledge about youth gangs, staff should develop strategies to engage public and community members and give them the opportunity to participate in local youth gang initiatives.
Staff should conduct a needs assessment to uncover the potential barriers to parental participation, especially since special strategies were needed to engage the parents of youth involved in gangs.
With partners working together on this project, data management for evaluation purposes was challenging. Different agencies were collecting individual data on their clients, which then needed to be combined for evaluation purposes. A centralized case management system should be established for programs that include multiple agencies to facilitate a seamless data collection and management system between the partners involved in the program to establish a reliable process for both collecting and analyzing data in an on-going manner and for evaluation.
Qualitative interviews should be conducted with youth and with other relevant key stakeholders to explore their perceptions and experiences of the program. The integration of qualitative research is key to understanding the unique experiences of the participants and for triangulating findings with quantitative findings to provide a richer context of the program.
Tracking and documenting risk levels, dosage, frequency of treatment and “relevant treatment” would provide data for conducting multivariate analysis. This type of data and analysis enhances program attribution. This is especially necessary when a comparison group is not available.
Attrition is a major issue with volatile populations. Incentives for participation should be implemented to encourage youth to both participate in the evaluation and remain involved in the evaluation so that the there is sufficient power to report on statistical differences in the outcomes.
Accurate completion of the fidelity checklist is encouraged so that external program factors that can influence the outcomes are considered in the analysis. The fidelity checklist could ensure that other organizations/agencies will be able to replicate the program.
It is recommended that when measuring employment, other indicators, such as employment training, and changes in employment-related skills be primary measures. These are more reasonable measures especially when these outcomes require changes in the socio-economic domain.
Obtaining official police data to validate offending patterns is vital in crime-related evaluations. Although a police representative participated in the NCPC evaluation advisory committee, this did not guarantee access to police data. Higher level agreements between the Department of Public Safety and provincial partners is required to ensure access to police data for NCPC program participants.
The results of the program evaluation demonstrate that the DYGS program has made gains in addressing the issue of youth gangs in the Durham Region.
The program, based on the OJJDP (Spergel model), has contributed to favourable changes in reducing youths' perceptions of criminal activities, attitudes toward criminal activity, peer associates, interaction with family members and school attendance.
The program did not favourably contribute to more of the intermediate and long-term outcomes related to delinquency, drug abuse and gang exiting for involved youth.
For more information or to receive a copy of the final evaluation report, please contact the National Crime Prevention Centre by e-mail at firstname.lastname@example.org.
If you wish to register for the NCPC mailing list to receive information from the Centre, please visit the subscription page at: http://publicsafety.gc.ca/cnt/bt/mlng-lst-eng.aspx
- 1 Saini, M. (2011) Evaluation of the Durham Youth Gang Strategy. MAS Information and Evaluation Group. Toronto, Canada.
- Date modified: