Cover Story

Unruly Turf

The Role of Interagency Collaborations in Reducing Gun Violence

By George Tita, K. Jack Riley, Gregory Ridgeway, and Clifford Grammich

Father Gregory Boyle, right, talks with former gang member Leonard Maldonado about tattoo removal to get a job. Boyle, a Jesuit Catholic priest, heads Jobs For A Future/Homeboy Industries, an employment referral center and economic development program for at-risk and gang-involved youth in the Boyle Heights area of Los Angeles.
George Tita is an assistant professor in the Department of Criminology, Law, and Society at the University of California, Irvine. Jack Riley is director of RAND Public Safety and Justice and of the Homeland Security Center within the RAND National Security Research Division. Also at RAND, Gregory Ridgeway is an associate statistician, and Clifford Grammich is a research communicator.

The Boston Gun Project has been hailed as a national model for reducing youth violence. Shortly after Boston launched the project in 1996, the number of homicides committed by juveniles in the city fell by about two-thirds. Could the Boston success be replicated elsewhere?

The project blended three ingredients of success that seemed transferable to other cities. First, a coalition of community leaders, criminal justice professionals, clergy, and researchers had collaborated on the project’s design, implementation, and evaluation. Second, the project offered youth offenders both “carrots” and “sticks”; it balanced social services (such as job training and substance abuse treatment) with tough punitive measures (such as stricter enforcement of parole and probation regulations). Third, the project was dynamic, meaning that the coalition members continually adjusted the balance of carrots and sticks as conditions warranted.

We at the RAND Corporation attempted to replicate the Boston success in a very different city: Los Angeles. With support from the National Institute of Justice, we instigated and evaluated a community-based effort to reduce gun violence among youth in a working-class area of the city east of downtown. We expected the project, when transplanted to Los Angeles, to retain the core elements of the Boston project. However, we expected that the type of problems addressed and the nature of the interventions pursued might differ from those in Boston, given the different community characteristics in Los Angeles and its much greater decentralization of criminal justice authorities. Following many months of planning and analysis, the antiviolence intervention in Los Angeles, called Operation Ceasefire, began in October 2000.

Some things worked well, and some things could have worked better. On the positive side, Operation Ceasefire confirmed that collaborative community groups, from Boston to Los Angeles, could reduce gun violence by prodding decentralized criminal justice agencies to concentrate their resources in a single geographic area. In this crucial respect, the Los Angeles project succeeded. The coordinated community effort reduced youth violence. The coalition members also succeeded in utilizing newly generated data to design effective interventions.

On the negative side, there were built-in constraints that prevented the Los Angeles project from fully replicating the Boston model and thus its likelihood of success. In Los Angeles, no city agency or community group took ownership of the project. No meaningful opportunity arose to make the project a dynamic one over the long term. And no funds were provided for social services to balance the criminal punishments. In the end, the project could not sustain itself.

In Los Angeles, we learned several lessons for future efforts to replicate the Boston Gun Project. For such efforts to succeed beyond a trial period, they need to include these elements of success:

  • Crime data should be analyzed carefully when designing interventions.
  • Social services should balance law enforcement efforts to the extent possible.
  • City leaders should provide concrete forms of sup- port to the municipal agencies involved.
  • City leaders should hold the municipal agencies involved accountable for the results.
  • Cost data should be collected to determine whether the projects merit continuation.

What Made Los Angeles Unique

At first, it was unclear whether the Boston approach would work in Los Angeles. There were concerns that the Los Angeles Police Department (LAPD) would not welcome a project developed elsewhere. It was unknown if an intervention designed for predominantly African American youth in Boston would be appropriate for a predominantly Latino city. And while the Boston Gun Project had been implemented citywide, a citywide effort was impractical in Los Angeles given its enormous size.

Therefore, we identified a smaller intervention area within the city. We chose the LAPD Hollenbeck division, a 15-square-mile area east of downtown with a population of about 200,000. Hollenbeck has an 81 percent Latino majority consisting primarily of people of Mexican heritage. Unlike other inner-city Los Angeles neighborhoods, where the population has shifted from a black majority to a Latino majority in recent years, Hollenbeck has had a Latino majority for decades. Hollenbeck gangs are among the oldest in the city, with some intergenerational gangs tracing their roots back to before World War II.

Perhaps the most unique characteristic of Hollenbeck came as a big surprise to many of the people working there. A detailed analysis of Hollenbeck homicide files revealed that gang issues, such as disputes over turf and “respect,” were the primary motivating factors for fully half of the homicides — 90 out of 180 — committed there between 1995 and 1998. An additional 45 homicides involved gang membership plus other factors, such as arguments, domestic altercations, or drug debts. A grand total of only 38 of the homicides — 21 percent — involved any kind of drug motive, such as a dispute over drug debts, the quantity or quality of drugs, or the robbery of a drug dealer (see Figure 1). Very few of the homicides were motivated by disputes over drug sales territory.

Figure 1

These findings drew incredulous responses from the working group members. Many had assumed that gang murders in Hollenbeck typically involved drug dealing. “These kids are being killed because of [dope]!” one officer insisted. But careful reanalysis of the data confirmed our findings and pointed to an important characteristic of the violence in this neighborhood: The links among gangs, drugs, and violence had been overdrawn. Although the gang members may sell drugs and may kill or be killed selling drugs, the motivation for the killings in Hollenbeck is not likely to stem from gangs fighting for market control. Rather, the local Latino gang members are more likely to engage in expressive acts of violence over issues of turf and respect.

The unique aspects of the violence in Hollenbeck pointed us toward unique strategies. Violence against others typically consisted of attacks against members of other gangs entering rival territories. Thus, each gang concentrated its activity, either protecting its own turf, preying on rivals on adjacent turf, or attacking rivals on other nearby turf where the gangs had overlapping social connections. Because the violence was spatially concentrated, it could possibly be mitigated by focusing on a small area, maybe no more than several square blocks.

“Nothing Stops a Bullet Like a Job”

As in Boston, the Hollenbeck initiative drew upon the support of area churches, notably those in the East Los Angeles Deanery of the Catholic Archdiocese. At the first meeting of the working group, 14 of the 17 people attending were priests. They represented a long tradition of “street intervention” in the area on behalf of local youth. For instance, Homeboy Industries is a local employment referral center established by a Jesuit priest and driven by the principle that “nothing stops a bullet like a job.” (See table of participating community groups.)

Having this structure of support behind the initiative was vital to the community’s acceptance of any role played by law enforcement agencies, given concerns about previous police interventions that had relied exclusively on gang suppression. The working group intended to balance law enforcement responses with social service programs. The goal was simple: Increase the cost of violent behavior to gang members while increasing the benefits of nonviolent behavior.

The spatial concentration of gang activity became the basis of the intervention. A vast network of rivalries among 29 area groups designated by the LAPD as “criminally active street gangs” offered several focal points for the intervention (see Figure 2). This “network analysis” of gangs also indicated an important structural break in the neighborhood that allowed for a natural experiment. The San Bernardino Freeway divides Hollenbeck into north and south. With one minor exception, no gang has any rivalries that cross the freeway. We focused the intervention in the southern portion of Hollenbeck, called Boyle Heights, because this was the area of the most intense gang rivalries and of the majority of violent crimes. We could then compare the results of the intervention in Boyle Heights with crime trends elsewhere in Hollenbeck.

Figure 2—Network Map of Hollenbeck Gang Rivalries

The working group spent several months considering a plan to quell gang violence in the wake of any triggering event that might lead one gang to retaliate against another. Because the violence was gang driven, the working group designed a strategy that would leverage the collective structure of the gang itself. As in Boston, the working group in Los Angeles designed an intervention based on a model of “collective accountability” — one seeking to hold all members of a gang accountable for the act of any single member. Among other features of the plan, it was to include

  • increased LAPD patrols in the immediate geographic area of the triggering event
  • deployment of officers from specialized police units to the broader neighborhood
  • additional police patrols in public parks
  • more-stringent enforcement of housing codes for properties used by gang members
  • more-stringent enforcement of public housing eligibility rules prohibiting possession of drugs, firearms, and other contraband
  • more-stringent enforcement of parole and probation conditions
  • more-stringent serving of outstanding warrants on gang members who had committed prior offenses
  • referral of gun law violations to federal prosecutors
  • dynamic and rapid application of these intervention elements after each violent incident to ensure that perpetrators and victims understood there were consequences for violent behavior.

At the same time, many of the agencies that were planning to impose tougher law enforcement measures were planning to offer social service programs as well. Parole officers and city agencies were planning to offer job-training opportunities. Probation officers were planning to offer tattoo removal and substance abuse treatment. Hospitals and community groups were planning to offer similar services. The working group spread the news throughout the community that the intervention would be launched in the wake of a triggering event — or immediately after a gang member committed a violent act and law enforcement officials had reasonable certainty about the perpetrator and the associated gang.

Many of these carefully laid plans, however, were overtaken by events. At an October 2000 meeting of the working group, representatives of the community organizations (who had been the most vocal advocates of the social service programs) urged the local LAPD captain to launch the law enforcement components of the initiative immediately. Given the very long time needed to implement social services such as job training, it did not appear feasible to coordinate the services with law enforcement. Of more urgent concern, violent crime seemed to be escalating rapidly. Events involving two gangs — The Mob Crew (TMC) and Cuatro Flats — were especially troubling (see rivalry highlighted in Figure 2).

The following weekend, a brazen “walk-by” shooting occurred in the heart of TMC territory. Five Cuatro Flats members exited a van, ran around a corner, and opened fire on a group of TMC members. Two people died: a 19-year-old TMC member and a 10-year-old girl who had been riding her scooter down the street and was hit by a stray bullet. This became the triggering event for Operation Ceasefire. It was launched the next day.

Six Months of Progress

The intervention as implemented differed in several ways from the intervention as planned. Most obviously, the social service programs were not widely available.

More generally, the intervention was not dynamic. The working group members did not continuously reprioritize their efforts and reallocate resources after each violent incident. Instead, the members focused almost exclusively on the two gangs involved in the triggering event. Subsequent shootings involving other gangs did not receive the level of attention associated with Operation Ceasefire. Consequently, the intervention never created a consistent perception that violent behavior would provoke an immediate response. The intervention became even more spatially concentrated. Rather than operating throughout Boyle Heights, it became confined largely to the five police reporting districts where TMC and Cuatro Flats were most active.

Nevertheless, violent crime and gang crime decreased in the aftermath of the intervention not only in the targeted area and in other areas of Boyle Heights but also throughout the larger Hollenbeck area. Figures 3 and 4 show the striking trends. Within just six months, violent crimes (including homicides, attempted homicides, robberies, assaults, and kidnappings) fell by 28 percent both in Boyle Heights and throughout Hollenbeck. Gang crimes (including gang-related violent crimes, firearm discharges, vandalism, and graffiti) fell by 40 percent in Boyle Heights and 32 percent throughout Hollenbeck. And gun crimes (including any of the above crimes that involved the use of a gun) fell by 23 percent in Boyle Heights and 33 percent throughout Hollenbeck. Even though homicides increased elsewhere in Los Angeles at this time, homicides decreased in Boyle Heights and Hollenbeck.

It is intriguing that the intervention appears to have helped reduce gang crimes involving guns both on the turf of the targeted gangs and on the turf of their rivals. In general, gang crime decreased more rapidly in Boyle Heights than in the remainder of Hollenbeck. Violent crime, however, decreased by nearly identical rates in both areas, and gun crime fell even more rapidly in the remainder of Hollenbeck than in Boyle Heights. These results suggest that focused law enforcement against two gangs helped to quell the activities of their rivals as well.

What Worked

Perhaps the most important success of Operation Ceasefire was the ability of the working group, using data analysis and collaborating with many different agencies, to craft a well-designed intervention. This process made it possible for each organization to contribute resources sufficient for making the collective effort a success.

We learned that it is possible for diverse criminal justice agencies — involving police officers, prosecutors, and probation officers — to work together effectively. The experience confirmed that each agency has unique resources that, when pooled, make each agency more effective in curbing violence than it otherwise could be alone.

The working group also rallied community support in ways that exceeded our expectations. Tailoring the intervention against an activity, such as gun violence, rather than an affiliation, such as gang membership, helped to garner community support. In turn, the support helped the working group to enlist a county supervisor to secure grants from the county probation department for hiring an intervention specialist for the project. Community support also led the city attorney’s office to dedicate a prosecutor and a community organizer to the project.

What Could Have Worked Better

As noted, the program never became dynamic. One reason for this was the disbanding of LAPD gang units in the wake of the city’s Rampart scandal in which several LAPD gang unit members had been accused of planting evidence. The disbanding of the units meant that many knowledgeable officers were no longer available to participate in Operation Ceasefire. It was difficult for the new gang unit officers to become familiar with the 29 gangs operating in the Hollenbeck area.

The working group members never truly assumed ownership of the project. One objective had been to create a self-sustaining mechanism by which members of the working group could continually address problems of violence in the community. We at RAND did not succeed in transferring ownership of the project from RAND to the working group.

One reason for this failure was the frequent rotation of agency personnel with whom the working group collaborated. During the 18 months of the most intensive planning, RAND researchers interacted with a half dozen LAPD captains. Such turnover may have been inevitable given the larger disruptions within the LAPD at the time, but the high turnover did not foster the stability needed to help the project sustain itself beyond the period administered by RAND. Changes in the political leadership of the city presented further obstacles to sustaining the initiative.

Another obstacle to sustainability was the lack of resources for any single agency to maintain the collaboration. Agency budgets do not encourage interagency collaboration. Agency personnel are not evaluated for their performance on such efforts. As one LAPD participant in Operation Ceasefire noted, “No one ever thought this was a bad idea. In fact, it makes sense. But departmental resources were never made available to implement the model in the intended way.”

Operation Ceasefire did not provide funds to any participating organization other than RAND. Agency personnel, including the LAPD officers and county probation officers who devoted the most time to the project, had to maintain their regular duties outside the project. Without people devoted exclusively to the project, it lacked accountability for success or failure.

Without participants being evaluated for their performance on the project, they did not see it as a primary responsibility. Failure was not seen as something that would endanger one’s career.

Los Angeles Police Chief William Bratton, left, listens to the advice of former gang member Jonathon Ortega, supervisor of Homeboy Industries, during a November 2002 kickoff of a campaign to fight graffiti in Los Angeles.


City leaders need to support criminal justice agencies involved in interagency collaborations such as Operation Ceasefire. Simultaneously, city leaders need to hold the agencies accountable for the collaborations. The reality is that most agencies have tight budgets. Almost none of the agencies can redirect resources for collaborations. And few agency directors are evaluated on how well they collaborate with other directors. To build future collaborations, city leaders need to ensure that support and accountability measures are built into agency budgets and evaluations.

Cost information is also needed. One reason why there are so few collaborations is that they are considered cumbersome and expensive. Most evaluations of collaborations to reduce crime have focused almost exclusively on how much crime has been reduced. Calculating the true costs in staff time and overhead expenses is necessary to determine whether such interventions merit replication or continuation.

Those who seek to replicate the Boston Gun Project need to analyze local crime data to define the unique causes of violence in an area. Only then can an appropriate response be crafted. Other neighborhoods, even within Los Angeles, could require different interventions than those used in Hollenbeck. In Hollenbeck, gang violence stems from inter-gang rivalries. Elsewhere in the city, gang violence could stem from intra-gang rivalries, as when former gang leaders return from prison and wish to reclaim their leadership roles.

The Hollenbeck working group was fully committed to social service programs, but it had far less access to these kinds of resources than did Boston. We suspect that municipal commitments to social services will almost always lag behind those for law enforcement. Nevertheless, broader efforts to attack the root causes of violence — such as joblessness, substance abuse, and limited education — need to be combined with law enforcement. The combined efforts would stand the greatest chance of producing permanent as well as immediate reductions in youth violence.

Related Reading

The California Wellness Foundation’s Violence Prevention Initiative: Findings from an Evaluation of the First Five Years, Peter W. Greenwood, Jeffrey Wasserman, Lois M. Davis, June A. Flora, Kim Ammann Howard, Nina Schleicher, Allan Abrahamse, Peter D. Jacobson, Grant Marshall, Carole Oken, Eric Larson, James Chiesa, RAND/MR-1342.0-TCWF, 2001, 46 pp.

Diverting Children from a Life of Crime: Measuring Costs and Benefits, Peter W. Greenwood, Karyn E. Model, C. Peter Rydell, James Chiesa, RAND/MR-699-1-UCB/RC/IF, 1998, 86 pp., ISBN 0-8330-2623-2.

“From Boston to Boyle Heights: The Process and Prospects of a ‘Pulling Levers’ Strategy in a Los Angeles Barrio,” George Tita, K. Jack Riley, Peter W. Greenwood, in Scott Decker, ed., Policing Gangs and Youth Violence, Belmont, Calif.: Wadsworth, 2003, pp. 102-130.

Reducing Gun Violence: Results from an Intervention in East Los Angeles, George Tita, K. Jack Riley, Greg Ridgeway, Clifford Grammich, Allan F. Abrahamse, Peter W. Greenwood, RAND/MR-1764-NIJ, 2003, 75 pp., ISBN 0-8330-3475-8.