Fighting Gangs: What Doesn’t Work

Print More

Getting kids to avoid or withdraw from gangs is among the most difficult tasks in youth work, and a fresh set of evaluations of federally funded anti-gang programs shows why.

After spending $13.3 million and 10 years on demonstration projects in five cities, the U.S. Office of Juvenile Justice and Delinquency Prevention found that almost nothing had a significant impact.

But there were plenty of lessons, some of which are being applied to new efforts. Among the key problems: not enough use of youth outreach workers, difficulties of community-based organizations acting as the “lead” agencies in the projects, and – a familiar one to anyone trying to implement a program model in different locations – failure to adhere to the program model.

This particular model tested the theory of Irving Spergel, a professor of sociology at the University of Chicago, who hypothesized that gang-related violence could be reduced if community agencies and groups engaged in coordinated efforts to assess, suppress, intervene and offer alternative opportunities to gang members.

In the mid-1990s, after years of rising violent crime – much of it attributed to gangs – OJJDP launched a demonstration of the model, calling it the Comprehensive Community-Wide Approach to Gang Prevention, Intervention and Suppression Program.

The model focused on five key strategies: mobilization of citizens, groups and agencies; social intervention using outreach workers, case managers, police and probation officers; academic and job training opportunities for gang members; suppression activities such as arrests, surveillance and sharing of information; and organizational changes that integrate the model’s strategies and reallocate resources among participating organizations.

At each site, a lead agency and a steering committee provided overall direction and support.

The program activities included arrests, tattoo removal, case management, job training, counseling and recreation.

Phelan Wyrick, OJJDP’s gang program coordinator, describes the model this way: “We know who you are. We know you’re out there. On the one hand, if you step over a line, we’ve got our eye on you and we will hold you accountable. But on the other hand, we want to offer you some alternatives and some choices.”

OJJDP funded five demonstration sites from 1995 to 1999: Bloomington-Normal, Ill.; Mesa, Ariz.; Riverside, Calif.; San Antonio; and Tucson, Ariz. Each site served about 100 to 250 at-risk youth, most of whom were Latino or African-American males from ages 14 to 17. Most described themselves as gang members or associates, and reported having been arrested on drug or violence charges.

In August, the federal government released the set of five 300- to 500-page final evaluation reports for the sites. The reports were compiled between 2000 and 2004 by national and on-site teams of researchers, led by Spergel.


“It’s a mixed bag,” Wyrick says. “Where the programs were implemented better, you got better results, and where they were poorly implemented, you didn’t get good results.”

Four of the five sites showed no significant impact by the programs on overall gang-related crime, although there were some reductions in specific areas. For example, in Bloomington’s Project OZ, led by Youth Impact Inc., frequent contact by probation and parole personnel seemed to reduce property offenses by male gang youth who had a lot of prior property offenses.

Program failures were generally attributed to poor implementation of the model design. Other common problems included poorly developed steering committees, an inappropriate mix of strategies and the dominance of a single agenda. For instance, in Mesa’s Gang Intervention Project (MGIP), led by the Mesa Police Department, less-delinquent youth were inappropriately included in the demonstration due to an overly “strong commitment to the social-intervention approach,” according to the evaluation.

Only Riverside – whose project was led initially by the University of California-Riverside and later by the Riverside Police Department – implemented the model’s components at a “good” level, according to evaluators, and showed positive results for youth overall. Youth in that program were three times as likely to show a reduction in serious-violence arrests as youth in a comparison group. The program did not reduce youths’ gang membership.

The evaluations gave OJJDP insight into at least three key components of the model: the efficacy of lead agencies, the use of outreach youth workers and the importance of alternative opportunities for gang members. “We’ve tried, to the extent possible, to incorporate some lessons learned” into subsequent gang-demonstration initiatives launched by the agency, Wyrick says.

Lead Agencies

One type of organization emerged as the most effective lead agency. “The police generally did a better job than the social agencies,” Spergel says. “They had more resources. They were more committed.”

When the nonprofit Our Town Family Center served as the lead agency for the Tucson Gang Project, for example, “they just didn’t end up having the kind of ‘oomph’ needed when it came down to it,” Wyrick says. “No police chief is really going to bend his or her will to that of a community nonprofit.”

As a result of that finding, OJJDP narrowed requirements for subsequent model demonstrations to use local government units as lead agencies. “We felt they also had the kind of weight to encourage the collaboration that’s required,” Wyrick says.

Outreach workers

The evaluations say most of the sites failed to use street youth outreach workers effectively, as called for by the model. San Antonio’s Gang Rehabilitation Assessment and Service Program, led by the city’s police department, did use them extensively for service provision.

“This outreach component is an important piece,” Wyrick says. “The role of that person is to be out in the community, building and gaining the trust of young people who are gang-involved, who aren’t necessarily knocking on the door of the local service center.”

Spergel, who has worked extensively with Mexican-American gangs in Chicago, says some of the most effective outreach workers are former gang members. (See “Retired Gangsters Gang Up on Youth,” November 2000.)


Riverside developed a six-month employment-training program that offered a $150 stipend and a network of employers to youth completing the program. The site also had representatives from two public school systems on its steering committee. Spergel says the creation of such opportunities is key to enticing older gang members – many of whom have matured to the point of wanting to settle down and support their families – out of gangs.

“Some of [the youth], even that are really deep into the gang lifestyle, are looking for a road out,” Wyrick says. “If they can get a real opportunity and get a chance, if done right, they’ll take it.”

Spergel believes that if implemented correctly, his model – often called the Comprehensive Gang Model – will deliver a broad array of adaptable services to youth most at risk for participating in violent gang crime.

The federal government appears to agree. In 2000, OJJDP funded four sites, under the renamed Gang-Free Schools and Communities Program, in an effort to follow Spergel’s model again. Data from those sites – Miami, Houston, Pittsburgh and East Cleveland, Ohio – are being collected for evaluation.

The reports are available free at, or for $15 each by calling (800) 851-3420.