Title: Reducing Gun Violence: The Boston Gun Project's Operation Ceasefire Series: Research Report Author: David M. Kennedy, Anthony A. Braga, Anne M. Piehl, and Elin J. Waring Published: National Institute of Justice, 2001 Subject: Program evaluations; Juveniles: gangs, juvenile violence; Law enforcement: problem oriented policing 60 pages 150,000 bytes ------------------------------ Figures, charts, forms, and tables are not included in this ASCII plain-text file. To view this document in its entirety, download the Adobe Acrobat graphic file available from this Web site or order a print copy from NCJRS at 800-851-3420 (877-712-9279 for TTY users). ------------------------------ U.S. Department of Justice Office of Justice Programs National Institute of Justice RESEARCH REPORT Reducing Gun Violence: The BOSTON Gun Project's Operation Ceasefire ------------------------------ U.S. Department of Justice Office of Justice Programs 810 Seventh Street N.W. Washington, DC 20531 John Ashcroft Attorney General Office of Justice Programs World Wide Web Site http://www.ojp.usdoj.gov National Institute of Justice World Wide Web Site http://www.ojp.usdoj.gov/nij ------------------------------ Reducing Gun Violence: The Boston Gun Project's Operation Ceasefire Developing and Implementing Operation Ceasefire David M. Kennedy Anthony A. Braga Anne M. Piehl Measuring the Impact of Operation Ceasefire Anthony A. Braga David M. Kennedy Anne M. Piehl Elin J. Waring September 2001 NCJ 188741 ------------------------------ National Institute of Justice Sarah V. Hart Director Lois Felson Mock Program Manager Anthony A. Braga is a senior research associate in the Program in Criminal Justice Policy and Management of the Malcolm Wiener Center for Social Policy, John F. Kennedy School of Government, Harvard University, and a visiting fellow at the National Institute of Justice, U.S. Department of Justice. David M. Kennedy is a senior researcher in the Program in Criminal Justice Policy and Management of Harvard University's John F. Kennedy School of Government and was the director of the Boston Gun Project. Anne Morrison Piehl is an associate professor of public policy in the John F. Kennedy School of Government, Harvard University. Elin J. Waring is an associate professor of sociology at Lehman College and the City University of New York Graduate School. This research was sponsored by grant award number 94-IJ-CX-0056 from the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. Findings and conclusions of the research are those of the authors and do not necessarily represent the official position of the U.S. Department of Justice. The National Institute of Justice is a component of the Office of Justice Programs, which also includes the Bureau of Justice Assistance, the Bureau of Justice Statistics, the Office of Juvenile Justice and Delinquency Prevention, and the Office for Victims of Crime. ------------------------------ Foreword This Research Report is part of the National Institute of Justice's (NIJ's) Reducing Gun Violence publication series. Each report in the series describes the implementation and effects of an individual, NIJ-funded, local-level program designed to reduce firearm-related violence in a particular U.S. city. Some studies received cofunding from the U.S. Department of Justice's Office of Community Oriented Policing Services; one also received funding from the Centers for Disease Control and Prevention. Each report in the series describes in detail the problem targeted; the program designed to address it; the problems confronted in designing, implementing, and evaluating the effort; and the strategies adopted in responding to any obstacles encountered. Both successes and failures are discussed, and recommendations are made for future programs. While the series includes impact evaluation components, it primarily highlights implementation problems and issues that arose in designing, conducting, and assessing the respective programs. The Research Reports should be of particular value to anyone interested in adopting a strategic, data-driven, problem-solving approach to reducing gun violence and other crime and disorder problems in communities. The series reports on firearm violence reduction programs in Boston, Indianapolis, St. Louis, Los Angeles, Atlanta, and Detroit. ------------------------------ Contents Foreword Introduction Part I. Developing and Implementing Operation Ceasefire by David M. Kennedy, Anthony A. Braga, and Anne M. Piehl Youth Homicide in Boston Applying Problem-Oriented Policing The Working Group and the State of the Art in Boston Proceedings of the Working Group Key Themes Unraveling the Nature of Youth Violence Through Research Bad Implications and Bad Times Designing Ceasefire Applying the Strategy Effectiveness of the Strategy Reflections on Problem-Oriented Policing Conclusion Notes Part II. Measuring the Impact of Operation Ceasefire by Anthony A. Braga, David M. Kennedy, Anne M. Piehl, and Elin J. Waring Study Design Findings Discussion Notes ------------------------------ Introduction Problem-oriented policing holds great promise for creating strong responses to crime, fear, and public safety problems. It aspires to unpack such problems and frame strategic responses using a variety of approaches. Through a process of problem identification, analysis, response, evaluation, and adjustment of the response, problem-oriented policing has been effective against an array of crime, fear, and order concerns. The Boston Gun Project was a problem-oriented policing initiative expressly aimed at taking on a serious, large-scale crime problem: homicide victimization among youths in Boston. Like many large cities in the United States, Boston experienced an epidemic of youth homicide between the late 1980s and early 1990s. Homicide among persons ages 24 and under increased by 230 percent--from 22 victims in 1987 to 73 victims in 1990--and remained high well after the peak of the epidemic. Boston experienced an average of 44 youth homicides per year between 1991 and 1995. Sponsored by the National Institute of Justice and directed by David M. Kennedy, Anthony A. Braga, and Anne M. Piehl of Harvard University's John F. Kennedy School of Government, the Gun Project involved-- o Assembling an interagency working group of largely line-level criminal justice and other practitioners. o Applying quantitative and qualitative research techniques to assess the nature of and dynamics driving youth violence in Boston. o Developing an intervention designed to have a substantial near-term impact on youth homicide. o Implementing and adapting the intervention. o Evaluating the intervention's impact. The Boston Gun Project Working Group began meeting in January 1995. By the fall of that year, the Project's basic problem assessment had been completed and the elements of what is now known as the Operation Ceasefire intervention mapped out; implementation began in early 1996. The two main elements of Ceasefire were a direct law enforcement attack on illicit firearms traffickers supplying youths with guns and an attempt to generate a strong deterrent to gang violence. To systematically address the patterns of firearms trafficking identified by research, the Working Group: o Expanded the focus of local, State, and Federal authorities to include intrastate firearms trafficking in Massachusetts in addition to interstate trafficking. o Focused enforcement attention on traffickers of the makes and calibers of guns most used by gang members. o Focused enforcement attention on traffickers of guns that had short time-to-crime intervals and, thus, were most likely to have been trafficked. (The time-to-crime interval is the time from the first retail sale to the time the gun is confiscated by the police.) The Boston Field Division of the Federal Bureau of Alcohol, Tobacco and Firearms (ATF) set up an inhouse tracking system that flagged guns whose traces showed a time-to-crime interval of 18 months or shorter. o Focused enforcement attention on traffickers of guns used by the city's most violent gangs. o Attempted to restore obliterated serial numbers of confiscated guns and subsequently investigate trafficking based on those restorations. o Supported these enforcement priorities through analysis of data generated by the Boston Police Department and ATF's comprehensive tracing of crime guns and by developing leads from the systematic debriefing of gang-affiliated arrestees or those involved in violent crime. The second strategic element, which became known as the "pulling levers" strategy, involved deterring the violent behavior (especially gun violence) of chronic gang offenders by-- o Targeting gangs engaged in violent behavior. o Reaching out directly to members of the targeted gangs. o Delivering an explicit message that violence would not be tolerated. o Backing up that message by "pulling every lever" legally available (i.e., applying appropriate sanctions from a varied menu of possible law enforcement actions) when violence occurred. Concurrently, the Streetworkers (a coalition of Boston social service workers), probation and parole officers, and, later, churches and other community groups offered gang members services and other types of assistance. Throughout the intervention process the Ceasefire message was delivered repeatedly: in formal meetings with gang members, through individual police and probation contacts with gang members, through meetings with inmates of secure juvenile facilities, and through gang outreach workers. The message was a promise to gang members that violent behavior (especially gun violence) would evoke an immediate and intense response. Although nonviolent crimes would be dealt with routinely within the criminal justice system, violence would receive the Working Group's focused enforcement actions. Street operations began in earnest in early 1996; the first comprehensive gang crackdown began in March and the Working Group's first meeting (or "forum") with gang members was held on May 15, 1996. A second major crackdown occurred in late August 1996, with other core Ceasefire activities-numerous forums, direct warnings to gangs, several lesser crackdowns, and gun trafficking investigations. The height of Ceasefire's operational activity, however, occurred during 1996 and 1997. Youth homicides in Boston decreased dramatically following the first gang forum in May 1996 and has remained low to the present. To determine if Operation Ceasefire was associated with this decline, the study team conducted a rigorous evaluation of the intervention's effects on youth violence in the city. Using carefully constructed, generalized linear models that controlled for trends and seasonal variations, they found that the implementation of Operation Ceasefire was associated with a 63-percent decrease in youth homicides per month, a 32-percent decrease in shots-fired calls for service per month, a 25-percent decrease in gun assaults per month, and a 44-percent decrease in the number of youth gun assaults per month in the highest risk district (Roxbury). Because these abrupt reductions in homicide and youth violence could have been caused or meaningfully influenced by other factors, the study team added control variables to the models that included-- o Changes in the employment rate. o Changes in Boston's youth population. o Changes in the citywide violent index crimes. o Changes in older homicide victimization. o Changes in street-level drug activity as measured by Boston Police Department arrest data. The addition of these control variables did not substantively change the findings: Operation Ceasefire remained associated with significant decreases in the monthly number of youth homicides and measures of nonfatal serious violence. The team also distinguished youth homicide trends in Boston from national and regional trends. Using Supplementary Homicide Report data, the monthly counts of the number of homicide victims ages 24 and under were analyzed for 29 major New England cities and 39 major U.S. cities. The generalized linear models suggested that Boston's significant youth homicide reduction associated with Operation Ceasefire was distinct--both larger and more abrupt--when compared with youth homicide trends in most U.S. and New England cities. Part I of this report provides a detailed description of the issues and processes involved in implementation of the Boston Gun Project's Operation Ceasefire. Part II discusses the design and findings of the evaluation study that assessed the impact of this strategic intervention. ------------------------------ Part I. Developing and Implementing Operation Ceasefire David M. Kennedy, Anthony A. Braga, Anne M. Piehl The seriousness and key dimensions of the relatively recent youth homicide problem in the United States are well known. Briefly, beginning in the mid-1980s and broadly coincident with the appearance of crack cocaine, youth homicide rates rose abruptly to historic highs. Homicide victimization rates tripled for young black males and doubled for young white males; juvenile handgun homicides increased 418 percent between 1984 and 1994.[1] For minority males in particular, homicides were concentrated in poor inner-city communities, where the combination of high rates of homicide and other violence, street drug activity, and preexisting social and economic difficulties combined to produce severe and perhaps unprecedented stresses.[2] Gang formation and gang activity also increased markedly.[3] Nearly all of the violence involved guns; the rate of homicide committed with knives and by other means stayed essentially flat, while the rate of youth gun crimes-gun homicide, gun assault, gun carrying, and the firing of shots-soared, particularly in troubled neighborhoods.[4] Youth Homicide in Boston Gun crime in Boston never reached the level attained in some other cities,[5] but when the use of crack cocaine became epidemic in roughly 1988, street violence became chronic in the primarily black neighborhoods of Roxbury, Dorchester, and Mattapan. In neighborhoods such as the one near the intersection of Normandy and Brunswick Streets in Roxbury-gang turf claimed by the extremely violent Intervale Posse-families avoided the front of their houses because of the frequency with which stray shots penetrated.[6] Arrests of primarily young, minority men for violent gun crimes and illegal possession of firearms skyrocketed.[7] In August 1988, in an event that seemed to crystallize the lethal anarchy that had taken over the city streets, young Tiffany Moore was killed in a street crossfire while sitting on a mailbox outside her Roxbury home. By 1990, the city seemed out of control. Boston's homicide rate, driven by street gun violence, increased by half that year, from 100 to 152. The police were overwhelmed. "We were responding to six, seven shootings every night," said Lt. Det. Gary French, former commander of the Boston Police Department's (BPD's) Youth Violence Strike Force. "You just ran from crime scene to crime scene." Although 1990 marked the peak, the problems persisted. Youth homicide rates remained at historically elevated levels[8] and the streets retained their violent new character. "I think there was a real question in people's minds about whether Boston would remain a viable city," said BPD commissioner Paul Evans. A district court judge, appalled by the shows of force gangs were making at trials, called for the National Guard to be deployed in the courthouses. In 1992, gang members invaded the Morningstar Baptist Church, where a slain gang rival was being memorialized, and attacked mourners with knives and guns. In 1994, after a shooting in Roxbury's Chez Vous skating rink, youths fleeing the scene shed so many handguns that investigating police officers had trouble figuring out which ones-if any-had been involved in the assault. For many of Boston's young people, the city had become a dangerous, complicated place. In a 1995 Centers for Disease Control and Prevention survey of 10 cities, some 15 percent of Boston's junior high school student sample said that they had avoided school in the past month because they were scared--the highest such response rate among the cities surveyed.[9] A young probationer interviewed as part of the Gun Project in 1995 said that he did his best to avoid his peers in an effort to prevent being dragged into dangerous conflicts. "I stay home or go over to my cousin's," he said. "It's too dangerous to go out or have a lot of friends." Another young probationer, more active on the streets, said that navigating their dangers required constant, demanding attention. "It's like a video game," he said. "You master one level, and they bump you up a level, and things get harder, and you keep on going until you just can't do it any more. That's what the streets have gotten to be like." Applying Problem-Oriented Policing Late in 1994, the authors approached Jim Jordan, BPD's Director of Policy Development, and Commissioner Evans to explore the possibility of working together to apply problem-oriented policing to youth homicide in Boston. Commissioner Evans, a career BPD officer recently named to head the department and committed to community and problem-oriented policing, was enthusiastic. In keeping with the general theory of problem solving, the idea was to take a close look at the youth homicide problem in Boston and, using a variety of methods, design an effective intervention (i.e., one that would have a substantial impact in the relatively near term). The hope was that this problem-solving process, if successful, would be broadly transferable to other jurisdictions. The further hope was that the intervention itself, while likely not transferable as a stand-alone "technology," would contain basic ideas or elements that could be adapted elsewhere in accordance with the nature of the youth violence problem in other cities and the resources and predilections of the agencies and communities involved. The authors' plan was to convene a Working Group of frontline practitioners from a variety of agencies and work with them to sort out what was happening on the streets, design a plausible intervention, work with the agencies involved to implement the intervention, and evaluate its impact. Assuming that an intervention would be designed and reach the implementation stage, the Working Group was intended to move into a core operational role, with the same frontline practitioners who had been involved in the strategic design taking central responsibility for carrying it out. While the authors envisioned an open-ended process--that is, they were agnostic about what shape the intervention should take, preferring that it emerge from the research, analysis, and strategy design of the Working Group-they recommended two areas in which both research and operational energies might be profitably focused. Illicit firearms trafficking One focus was illicit firearms trafficking.[10] In Boston, as elsewhere, most youth homicides were committed with firearms, especially handguns. Although youths cannot acquire handguns legally, they nonetheless had little difficulty in obtaining them. The authors surmised that at least some of these guns were being illegally trafficked to and by youths, either directly from retail outlets (by large-volume traffickers and small-volume "straw purchasers" diverting guns to youths from gun stores) or less directly through the illegal sale of stolen guns by burglars and fences and the illegal sale and bartering of guns among youths themselves. Surprisingly, both Federal and local authorities had ignored the issue of gun trafficking. "When we find a kid with a gun, we charge him with a gun crime," said Jordan at the time. "We don't even ask him where he got it."[11] The size and nature of this market, the relative importance of its different components, and the significance of illegal firearm pathways (large-volume trafficking versus small-scale traffickers versus firearms stolen by youths from private residences) were all unknown. However, because guns were clearly such an important part of the youth violence problem, it seemed worthwhile to determine whether direct law enforcement interventions might be employed against illicit firearms trafficking. Fear and decoupling of violence from crack The second focal point was fear. Although crack cocaine trafficking had almost surely sparked the wave of youth violence in Boston and elsewhere, there was evidence that the problem was now being driven by the fear of many young people that they were at risk of exposure to high levels of violence and were consequently getting and using guns, ostensibly for self protection.[12] Research by Joseph Sheley and James Wright, especially, showed that in communities with high levels of youth violence, even young people not heavily involved in drug and gang activity were getting and carrying guns in the name of self-defense.[13] The authors believed that the youth violence problem had thus become decoupled from the drug trafficking phenomenon and now might usefully be addressed in different terms. They hoped, in particular, that making the streets safer for such young people would diminish the demand for guns. How to do so was an open question. The authors wished to explore the idea of including youths in community and problem-oriented policing activities to find out, through consultation, what was driving them to obtain guns and to address particular issues (such as a threatening street gang or a patch of dangerous drug turf) using problem-oriented policing methods. The authors framed these two dimensions in gun market terms. Gun trafficking and other routes in the illicit acquisition of firearms represented the supply side of the problem. Fear and/or other factors that might be driving illicit gun acquisition and use represented the demand side. They remained open to the possibility that these ideas might be proved wrong in practice and that the problem-solving process might suggest other theories and methods; however, the authors felt reasonably sure that any workable intervention would have to address both the supply and demand sides of the equation. The Working Group and the State of the Art in Boston Commissioner Evans supported this basic approach and, early in 1995, with grant support from the National Institute of Justice, the authors began to convene the Working Group. Commissioner Evans asked Superintendent James Claiborne, then Commander of Field Services, to guide the authors with respect to BPD representation in the Group.[14] Claiborne advised the authors to work with officers of the Youth Violence Strike Force (YVSF), the department's gang unit.[15] The authors resisted. In their view, although illicit gun trafficking might be connected with gangs, gangs certainly did not represent, a priori, the best or even a good unit of analysis. And to the extent that the decoupling hypothesis was correct, one would want to address not only gangs, and perhaps not even particularly gangs, but a broader spectrum of youths. Superintendent Claiborne listened attentively to these arguments and rather firmly advised the authors to work with officers of YVSF. The authors reluctantly complied, and were soon grateful, for it was apparent almost immediately that Claiborne had been right. Early in 1995, the authors began meeting with officers of YVSF on their home ground, part of BPD's Special Operations headquarters on Warren Street in Roxbury. YVSF was at the core of a loose, largely informal, and extraordinarily dedicated network of frontline agency personnel and community activists that had created an extremely sophisticated and nuanced approach to Boston's youth violence problem. At the center of this network was Detective Sergeant Paul Joyce, then the commander of YVSF, an imposing, taciturn marathon runner with a shaved head. Widely regarded within BPD as one of the best police officers in the department, Joyce had been a member of the now-defunct Anti-Crime Unit (ACU) when youth violence first began to reach crisis proportions in Boston. An elite "flying squad" tactical unit deployed citywide in response to the needs of the moment, ACU-with at least the tacit support of department management-had responded to the chaos on the streets with a wholesale stop-and-frisk policy aimed at young black men in Roxbury, Dorchester, and Mattapan. The eventual and formidable backlash from the black community, the press, and the courts was a formative experience for Joyce and many others in ACU, which was disbanded; many of its members, however, would soon be assigned to the new YVSF. "We'd been out there trying to do this on our own, the only way we knew how, and it just hadn't worked," Joyce said. "It taught us that we couldn't do it alone and we couldn't do it without support from the community and other agencies. And that it couldn't be just policing, or just enforcement; there had to be prevention, too." During the early years of the 1990s, Joyce quietly and steadily teased into being a distinctive brand of community policing aimed at the city's youth violence problem. He reached out to and crafted partnerships with a variety of other local actors. Frontline practitioners from agencies such as probation, parole, the Department of Youth Services (DYS) (juvenile corrections in Massachusetts), the Federal Bureau of Alcohol, Tobacco and Firearms (ATF), and the Suffolk County District Attorney's Office became regular participants in YVSF planning and operations. So did less innately police-friendly groups such as Boston's Streetworkers. The Streetworkers were City of Boston social service workers who dealt with the most at-risk youths on the streets, trying to connect them with services, keep them out of trouble, and mediate disputes. They were not in the enforcement business; their loyalties were to the youths (they often looked and sounded like the kids with whom they associated) and, frequently, they mistrusted the police. As a result, they were sometimes regarded by the police as little better than gang members themselves. Joyce didn't care. "They have their own job to do, and their own way of doing it," he said. "There are things we can do together." Also drawn into YVSF's orbit was the Ten Point Coalition, a group of activist black clergy formed in the wake of the Morningstar Baptist Church incident and spearheaded by the Rev. Eugene Rivers, a minister who had been critical of the police department. Joyce and YVSF forged a strong working relationship with Rivers after an incident in which a street drug dealer with whom Rivers had been disputing shot at the minister's house. YVSF solved the case and induced the young man to surrender; his surrender included an apology to Rivers. All this was done through the personal, individual outreach of Joyce and the others involved, although BPD management was aware and supportive of what was happening. What emerged was an extensive network with a remarkable mix of capacities, well-informed about activities on the streets at any given moment and possessed of a guiding, bedrock pragmatism. If Tracy Litthcut, director of the Streetworkers, heard that two gangs were going to be fighting, he could quietly notify YVSF, which then could flood the area with officers to prevent an incident. If a YVSF officer realized in the course of such an operation that the gang member driving the dispute was under DYS community supervision, he or she could reach out to friends at DYS and arrange for the youth to be picked up and held in a DYS facility until matters calmed down. YVSF officers cracking down on a gang could introduce members to Litthcut and Rivers, who could broker social services or job referrals, or perhaps hold out the promise of a slot in the Summer of Opportunity, a job program YVSF had created in partnership with the John Hancock Insurance Company. Probation officers alerted by one of their charges that another kid was a danger on the streets-or in danger on the streets-could inform Joyce, who, if he thought it warranted, would arrest and remove people on the spot. The authors came to realize that practitioners in traditional enforcement roles, such as Joyce, as well as those in traditional prevention roles, such as Litthcut and Rivers, seemed to believe that they were involved with interesting, often intelligent kids who had been failed by the adults around them and were trying to cope with a dangerous environment not entirely of their own making. The practitioners believed the youths deserved protection and help, but also that they were, at times, extremely dangerous and therefore needed to be controlled. This mutual sensibility permitted a remarkable sharing of approaches, often carefully tailored to particular situations on the streets. Chief among the important formal programmatic innovations was Operation Night Light, a community probation program in which probation officers and YVSF officers jointly patrolled in Roxbury, Dorchester, and Mattapan in the evening to, for example, run curfew checks on probationers or carry out their area and enforcement restrictions. Created by Dorchester Court probation officers William Stewart and Richard Skinner and YVSF detectives Bobby Merner and Bobby Fratalia in 1992, Night Light has since been instituted statewide and received national recognition. "We never used to leave the office or talk to the police," said Stewart. "But in the early 1990s the probation office looked like a MASH unit and we were seeing these officers in the courthouse all the time, and we realized we were all dealing with the same kids. And one day they said, do you want to ride together?" Their first night out they happened on a shooting. The victim, on the ground with a gun shot wound in his chest, was a probationer, and milling around in the crowd were a dozen more probationers, all out in violation of their curfews. According to Stewart, "We got out of the car, and they said, 'What are you doing here?' And we said, 'No, what are you doing here?'" One of the probationers on the street, braced by Stewart and Skinner and facing a violation that could send him back to prison, provided information that helped solve the shooting. Operation Night Light was under way.[16] These new relationships and operations, for all of their innovation and vigor, did not have the intended effect on youth homicide. Although the killing remained below the historical peak of 1990, youth homicide (measured as victimization among those ages 24 and under) had remained essentially flat from 1991 on (and would remain so through mid-1996), at an average of about 44 victims a year (see exhibit 1-1). Paul Joyce welcomed the authors in the same spirit that he brought to all aspects of his job. He would work with anyone and try anything legitimate that held any hope of helping with the youth violence problem. He liked the idea of forming a Working Group and quickly brokered introductions to the core frontline practitioners with whom he had been working in YVSF: probation officers, DYS personnel, school police, and Streetworkers. The authors, with the support of the Boston Police Department, approached the U.S. Attorney's Office, the Office of the Suffolk County District Attorney, and the Boston Regional Office of ATF, each of which supported the project and named representatives to the Working Group. In January 1995, the Working Group started to convene meetings regularly--approximately every 2 weeks for several hours at a time--at YVSF's Roxbury headquarters.[17] The authors had been completely unprepared for the group members' extraordinary levels of initiative and accomplishment; this seemed, however, to raise the bar somewhat for the project as a whole. YVSF and its allies were already doing a great deal of what the authors had hypothesized might be helpful in addressing youth violence in Boston: They had formed powerful working partnerships, adopted a fundamentally preventive and problem-solving orientation, and made some important programmatic innovations. Nonetheless, the impact of these efforts on the streets was far less than could be wished. Youth violence was still commonplace and the threat of youth violence was constant. The group had no real sense of any obvious way to progress further. "I didn't think there was anything that could be done to stop the violence," James McGillivray, a Working Group participant and former Streetworker who ran a strict supervision program for probationers in Roxbury, said later. "I just didn't see how you could do it." The same sentiment was common in the interviews the authors conducted with probationers. Interviews included the question, "What could be done to stop the violence?" The answer, invariably, was "Nothing." At the same time, despite the very high quality of these practitioners and their work, the process had been less focused than might be hoped. Although the police, probation officers, Streetworkers, Ten Point Coalition, and others were toiling away, their efforts had no center; no person, or group of people, was responsible for understanding and acting against the problem of youth violence in Boston. The Working Group, it was hoped, could provide that locus. Proceedings of the Working Group The Working Group meetings at Warren Street passed through several phases. The first, which lasted roughly through the spring of 1995, involved a thorough sharing of agency activities and views about the causes of youth violence in the city. Essentially all members of the Working Group knew Paul Joyce, but few were acquainted with each other. In any case, this was the first forum in which everybody was asked to say more or less directly what they were doing and why. Key participants during this period included Paul Joyce and several other detectives from the Youth Violence Strike Force; Billy Stewart and Richard Skinner from probation; Hugh Curran from the Suffolk County prosecutor's office; Tim Feeley and later Ted Heinrich from the U.S. Attorney's Office; David Carlson, Lennie Ladd, and Phil Tortorella from ATF; Mike Hennessey from the City of Boston School Police; Tracy Litthcut and James McGillivray from the Streetworkers; and the authors. In the process, even members of fundamentally different orientation got used to dealing with each other. In one of the sharpest exchanges of this period, James McGillivray spoke at length about how the terror many of Boston's young people experienced on the streets had driven them to carry guns. Philip Tortorella, an ATF special agent who made no secret of his passion for taking gun traffickers and serious felons with guns off the streets, responded with disbelief, "Can that mean we tolerate kids carrying guns?" "No," McGillivray responded, "but it does mean that if we neglect the real pressures on the kids, we're missing something important." The positions Working Group members took were sometimes less predictable. Tim Feeley, from the U.S. Attorney's office, briefed the group on the case of Freddie Cardoza, who had recently been arrested by YVSF while in possession of a single round of handgun ammunition. Cardoza was regarded by many in Boston law enforcement as one of the city's worst gang members. Prompted by YVSF, ATF and the U.S. Attorney had worked together to charge him as an armed career criminal under a Federal statute aimed at those with prior convictions for three violent felonies, three predicate drug felonies, or a combination thereof. At the time of the discussion, Cardoza had been convicted and would soon be sentenced to almost 20 years in Federal prison (nearly 4 years were for an additional charge of transferring a handgun to a minor). Feeley voiced an unexpected concern, one he would raise regularly throughout the Working Group process. "Are we really buying anything with these prosecutions?" he asked. "Sure, we've got Cardoza off the street, which is fine, but the real purpose of these heavy Federal sanctions should be to create deterrence." He asked the group if other street kids knew about this incident. If not, he questioned whether it could possibly shape their behavior. The police, probation officers, and Streetworkers shared his suspicion that other street offenders did not, in fact, understand some of the sanctions to which they were exposing themselves. And Tracy Litthcut, the head of Streetworkers, voiced his own unanticipated position. "I wish you'd do more of these heavyweight prosecutions," he said, "You can't save everybody on the streets. Cardoza was dangerous, and if the kids saw more Cardozas, maybe they'd be more careful about doing things like carrying guns." Key Themes Throughout the process, the authors were looking for opportunities to frame new interventions. By late spring, three key themes had emerged. Gun trafficking From the beginning, the Working Group had agreed that addressing gun trafficking would be useful. In fact, YVSF had been working on Operation Scrap Iron, its first large-scale, anti-gun trafficking operation. Scrap Iron was a model of the way Paul Joyce and the larger network around YVSF had come to do business. Late in 1994, YVSF noticed a sudden and severe outbreak of gun violence in and around Dorchester's Wendover Street. At about the same time, a probationer who owed Dorchester court probation officer Richard Skinner a favor came to him to pay his debt. "There's a kid on Wendover Street," he told Skinner, "who's selling crazy guns." Skinner spoke to Joyce, who spoke to ATF. Over the next several months, YVSF and ATF built a case against a Cape Verdean named Jose Andrade, whose family lived on Wendover Street and who went to college in Mississippi. Andrade turned out to be using a network of straw purchasers to buy guns while he was at school, and he was bringing the guns back to Boston during school vacations to sell to the loose street crew with which he associated. YVSF arrested Andrade and (using means that remained mysterious to the authors for some time) induced his crew to hand over many of the guns; in some instances, they delivered bags of handguns to officers at the Warren Street headquarters. The outbreak of shooting stopped and Joyce and his colleagues became committed to the idea of doing something about gun trafficking. In their view--and in the view of most on the Working Group--the trafficking problem was one of guns being brought from the South. Gang kids, they thought, liked "new in the box," .380 or 9mm semiautomatic pistols that were safe because they had never been used in a crime before. Figuring out a solution to this problem became a Working Group priority. Fear among Boston youths Joyce and the rest of the Working Group also endorsed the notion that fear was spreading out from groups of core offenders to a larger youth population and must be addressed. YVSF was seeing a new phenomenon, Joyce said: youths with "a scholarship in one pocket and a gun in another." Probation officers and Streetworkers endorsed that view, and the authors thought prosecutors and ATF were willing to believe it. Litthcut and McGillivray were perhaps most eloquent on this subject, with McGillivray at one point treating the group to an abstruse but compelling disquisition on Locke and Hobbes: "Civil society has broken down on the streets. We are utterly failing to protect these kids and so they've reverted to a state of nature. Unless we can reimpose civil society, we're not going to make a lot of headway." Gangs and chronic offenders Thus far, the Working Group's views were tracking nicely with the gun trafficking and decoupling ideas with which the authors had entered the process. But they diverged dramatically when the Working Group members closest to the streets--Joyce and the others from YVSF, probation officers, and Streetworkers--insisted that youth violence in Boston was a problem committed by and against chronically offending gang members. This argument had two parts. One was that Boston's gangs were not the relatively organized and structured gangs of Los Angeles and Chicago--a contention that was undisputed among all Working Group members. YVSF informally called itself a gang unit; probation officers talked easily about gangs, as did the Streetworkers; prosecutors discussed prosecution of gang members. But the "gangs" in question were loose, informal, and usually relatively small. The practitioners' second point was that most of the violence was among members of these gangs. "Every time we lose a kid, we know him, and every time a kid kills another kid, we know the shooter," said Billy Stewart. Others on the Working Group echoed his comments, saying that a relatively small number of youths were at high risk for both killing and being killed. They were gang members chronically in trouble with the authorities and known by Working Group members, often personally, because of their participation in gang activities, frequent arrests, status as probationers and prisoners, and visibility both on and off the streets. Working Group members could not predict who would kill or be killed, but they believed they could say with some authority from which relatively small universe the youths would come. In particular, they said, chronic "beefs" between gangs could be counted on to generate casualties. According to police, probation officers, and Streetworkers, these vendettas were what primarily drove youth violence in the city. Sparked by some dispute or other--perhaps business, perhaps personal--they took on a life of their own and could persist for years, even after most or all of those involved had forgotten their origins. They could lie dormant for long periods and then flare up on reprovocation--because rival gang members ran into one another in school, on the street, or in clubs-or for no apparent reason at all. When the vendettas were active, Working Group participants said they had a very strong idea of who was at risk; YVSF members sometimes joked among themselves about taking life insurance out on certain gang members. Much of the energy of YVSF, probation officers, and the Streetworkers went to defusing or containing these disputes. One way of viewing this was to regard the youth violence problem in Boston as comprising two groups: a relatively small group of chronic gang offenders and a much larger group of other youths. Working Group members maintained that the larger, non-gang-involved group feared the smaller one but that chronic offenders in the smaller group also feared each other. Most of the violence occurred among members of the smaller group, who were constantly on watch for each other and, as a result, carried guns, used guns, and acted tough in various ways. The Working Group stated that fear was driving this dynamic and that the problem had become, at least in part, decoupled from crack cocaine. But the core of the problem was a gang problem. The authors resisted this diagnosis; they were drawn to the strong version of the decoupling hypothesis, which made fear and vulnerability among a wide spectrum of youths more central than a more particular dynamic among gang kids. Working Group members, however, were compelling on the subject. The matter of gangs thus became a major topic for exploration. Unraveling the Nature of Youth Violence Through Research One key ambition of the Boston Gun Project was to support the proceedings of the Working Group, wherever possible, with research. By late spring 1995, several key questions were identifiable: o What guns were being used in youth crime in Boston, where were they coming from, and how were youths acquiring them? o What was driving the violence and, in particular, what was the role of gangs and gang disputes? o If a gang problem existed, what was the nature of gangs, and of gang members, in Boston? Over the summer of 1995, research was able to shed light on all of these areas. Guns and gun tracing data Since 1991, BPD and the Boston Regional Office of ATF had been tracing all guns recovered by Boston police. The arrangement, unprecedented at the time, had been put in place by David Carlson, then an ATF special agent, and Robert Scobie, then a sergeant in charge of the BPD Ballistics Bureau. Working Group member and ATF special agent Lennie Ladd negotiated with ATF headquarters in Washington, D.C., to work out arrangements by which the authors were granted research access to this data set. The data set contained information on 1,550 firearms recovered in Boston from persons ages 21 and under.[18] Some information, such as gun type and caliber, was available for all of these firearms. After some additional laborious matching by the authors from BPD records, the name and birth date of the person from whom Boston police had recovered the gun were available for about 80 percent of these firearms. Further information, such as where and when the gun had first been sold at retail, was available for the roughly 60 percent of firearms that had been successfully traced by ATF's National Tracing Center.[19] Over the course of the project, additional information was obtained from both ATF and BPD. ATF provided the name and Federal Firearms License (FFL) number of the dealer who had first sold the gun at retail, as well as the name of the first purchaser; BPD determined whether the name of the person from whom the gun had been recovered was included in the YVSF gang database. The resulting data set turned out to be an extraordinary resource. Analysis confirmed the belief of Working Group members that young people in Boston had a taste for new guns; slightly more than 25 percent of traceable guns and some 33 percent of guns associated with gang members were less than 2 years old at the time of recovery. This fact defied the common belief in law enforcement and other circles that such guns were invariably old, either borrowed from their own homes or stolen in house burglaries. Further, more than half of the traceable guns were semiautomatic pistols, and these tended to be newer yet; this was even more true of guns associated with gang members, particularly the .380 and 9mm models identified by the Working Group as guns of choice on the street. Where popular models of semiautomatic pistols were concerned, the time from purchase to recovery by police was often very brief.[20] In all probability, these guns had been deliberately diverted from gun stores into young people's hands. Finally, nearly 20 percent of all guns recovered from youths had obliterated serial numbers--a marker of illicit firearms trafficking. Analysis disproved the belief, on the other hand, that all youth crime guns in Boston were coming from southern States. More than 33 percent of traceable guns had first been sold at retail in Massachusetts; the next largest source State was Georgia, at only 8 percent. Guns originating in all southern States combined equaled a volume slightly below that of Massachusetts. Contrary to the expectations of the entire law enforcement community, Boston had a large problem essentially in its own back yard.[21] These findings led to some quick and concrete action. Special Agent Phil Tortorella took steps in ATF's Boston office to flag for special attention every gun whose trace showed a time-to-crime interval of less than about 18 months on the premise that such guns were particularly likely to have been trafficked or to lead to traffickers. For the same reason, the BPD's Ballistics Unit instituted a policy of attempting the restoration of all obliterated serial numbers. Members of the Working Group, newly alerted to the importance of instate trafficking, began working with the Massachusetts Department of Public Safety to rationalize the maintenance of records of instate firearms transfers.[22] Youth homicide, chronic offenders, and gangs Criminal activity of offenders and victims. With the help of BPD's Homicide Bureau and Office of Research and Analysis, the authors constructed a list of 155 people ages 21 years and under who had been killed by gun or knife in Boston between 1990 and 1994. Mapping these slayings and gathering key demographic and incident information were simple but powerful steps. Almost all the killings fell within the neighborhoods of Roxbury, Dorchester, and Mattapan; most victims were young black men; and most had been killed with handguns. Even long-time members of local criminal justice agencies were taken aback by the resulting charts. "No more in Charlestown?" one prosecutor said almost plaintively on seeing the map; she was referring to a white ethnic part of Boston renowned as a tough neighborhood that had nonetheless escaped the youth violence epidemic afflicting the black community (see exhibit 1-2). Next the authors examined the criminal histories of the 155 victims and the 125 youthful offenders, identified by the homicide bureau as "cleared," who were associated with the incidents. The results were astonishing: Prior to their murders, 75 percent of victims had been arraigned for at least one offense in Massachusetts courts, 19 percent had been committed to an adult or youth correctional facility, 42 percent had been on probation at some time before their murder, and 14 percent were on probation at the time of their murder. Of the 125 offenders known to be associated with those homicides, 77 percent had been arraigned for at least one offense in Massachusetts courts, 26 percent had been committed to a facility, 54 percent had been on probation, and 26 percent were on probation at the time they committed the murder. For the 117 homicide victims with at least one arraignment, the average number of arraignments was 9.5, and 44 percent had 10 or more arraignments. For the 96 offenders with at least one arraignment, the average number of arraignments was 9.7, and 41 percent had 10 or more arraignments. For both victims and offenders, arraignments for property offenses, armed violent offenses, and disorder offenses outnumbered drug offenses. For offenders, unarmed violent offenses also outnumbered drug offenses. Even within this high-rate population, offending was skewed, with the worst 5 percent and worst 10 percent of the 125 offenders responsible for 20 percent and 36 percent of 1,009 total arraignments, respectively. The worst 5 percent and worst 10 percent of the 155 victims were responsible for 17 percent and 33 percent of 1,277 total arraignments, respectively. As the Working Group had said, this was indeed a highly criminal population. Gang involvement. Next, working with YVSF, probation, and Streetworker members of the Working Group, plus several other police and Streetworker participants whom they recommended, the authors mapped gangs and gang turf and estimated gang size. This process identified some 61 different crews with some 1,300 members (the map of gang turf coincided almost perfectly with the homicide map) (see exhibit 1-3). Gang members represented less than 1 percent of young people in the city and less than 3 percent of young people in high-risk neighborhoods. Another step in the mapping process produced a network map of gang "beefs" and alliances: who was feuding with whom and who allied with whom (see exhibit 1-4). Finally, working with the same group of practitioners, the authors systematically examined each of the 155 homicides and asked the group members if they knew what had happened in each instance and whether there had been a meaningful gang connection. This answer, too, was striking: Using conservative definitions and methods, at least 60 percent of the homicides were gang related. Most of these incidents were not in any proximate way about drug trafficking or other "business" interests; most were part of relatively longstanding feuds between gangs. None of these dimensions-the number of crews, their size, their relationships, or the connection of gangs and gang rivalries to homicide-could have been examined from formal records; the relevant information simply was not captured either within BPD or elsewhere. But the frontline practitioners in the Working Group had this knowledge, and obtaining it by qualitative methods was a straightforward if laborious and time-consuming task.[23] By late summer 1995, most of this work was complete. Boston clearly had a gang-related youth homicide problem. More specifically, it had a problem with 1,300 chronic offenders who, for the most part, were minority male gang members who hurt one another along identifiable "vectors" of gang rivalry. The practitioners on the Working Group had been correct. Bad Implications and Bad Times As these findings emerged over the summer of 1995, they were folded back into the Working Group process. Although it was exciting, and felt like progress, to arrive at such a specific, detailed, and in many ways unexpected picture of the youth homicide problem in Boston, this was a deeply discouraging time for the authors. For several reasons, the description that was developing seemed to diminish-though did not eliminate-the power of a supply-side, gun market-focused approach to youth homicide prevention. The gang members at the core of the city's youth violence problem were highly criminal and thus likely had fairly varied and robust means of access to firearms through drug connections and other illegal pathways. Many of them were involved to some extent in drug trafficking and other economic crimes in which guns were useful, which gave them a strong financial motivation to obtain guns. More importantly, they were at very high risk of getting hurt, which gave them a strong personal motivation to obtain guns for self defense. Although neither the authors nor the Working Group lost hope that a supply-side intervention would be possible and productive, it seemed clear that preventing this chronically offending population from acquiring firearms would be extremely difficult. That understanding placed more of a burden on the demand side: to reduce the fear these gang members felt, ideally by lowering their risk of violent victimization or by controlling their violent offending. This seemed a newly daunting task. Although the research had crystallized and refined Working Group practitioners' views of Boston's gang problem, it had not altered those views in any fundamental way. Their ongoing work had been shaped by that knowledge and addressed the problem as they (and now the authors) saw it. The current practice of YVSF and other agencies was of a very high quality (the authors' respect for the group's effort was only growing with longer exposure) and it was not at all clear what, if anything, could be added to the existing mix. The research also demonstrated that the authors' original frame of reference-juvenile gun violence-was wrong. The real problem was violence among chronic gang-involved offenders; it was mostly but not exclusively firearm violence, and it included not only, or even predominantly, juveniles but also offenders well into their 20s. The authors and the Working Group framed this new problem as gang violence among the 24-and-under population and began calling it, somewhat inaccurately, "youth violence." Nobody was inclined to give up. The Working Group had taken on a real energy by now; the members were caught up in the process, and many ideas were suggested and critiqued.[24] Some ideas appeared at first and even on review to have merit. There were discussions of improved information sharing, for example, so that BPD patrol officers in high-risk areas would know, as they currently did not, who was on probation or under DYS supervision. One YVSF detective, in particular, wanted illicit gun traffickers to be held jointly liable for any crimes subsequently committed with the guns they sold. The prosecutors on the Working Group warned that this would require an unusual and hard-to-prove fact pattern: It would be necessary to prove in court that the trafficker had had prior knowledge of the crime in question. They welcomed the idea of bringing such a case to trial, however, and vowed to look for an opportunity. The importance of Massachusetts as a source State for Boston crime guns led both ATF and the U.S. Attorney's Office to refocus some of their energies toward instate, as well as interstate, trafficking. New ideas about taking guns off the street were discussed. Gang members often hid "area" guns near their hangouts or drug-dealing sites; they had been observed stashing weapons on the street when making probation appearances because the courthouses had metal detectors at their doors. Could enough street surveillance and a coordinated call to high-risk probationers yield a one-time harvest of guns? The regular Working Group meetings also created a useful professional dynamic. Members typically used some time at each meeting to share information and intelligence, make plans, and generally keep one another up to date about the activities of their agencies and conditions on the street. Both Tim Feeley from the U.S. Attorney's Office and Hugh Curran from the Suffolk County District Attorney's Office rode along with YVSF officers, a mutually satisfying experience that led Curran's office to formally designate him to work with YVSF. The Streetworkers built working relationships with prosecutors and ATF, something they had lacked before. Nonetheless, the authors did not believe that the ideas generated either alone or in combination seemed likely to result in a substantial impact in the near term. In what was essentially a coincidence but served to underscore the stakes involved and heighten the authors' sense of urgency, the summer and fall of 1995 turned out to be deadly. More than a dozen young people were killed in the city within a few months, and numerous other shootings filled the gaps between homicides. Working Group meetings--and local newspapers--were filled with accounts of street violence. At this point, the authors had essentially moved in with the Working Group, spending day after day at Warren Street, riding with YVSF and probation officers, and holding one-on-one conversations with Working Group members to probe more deeply into what they did and could do, what they thought was going on, and what might be done about it. It was an intense and a deeply disturbing time, with no clear way forward.[25] Lessons of Wendover Street The answer, when it came, turned out to be embedded in the historical activities of YVSF. As part of the general push to understand the history of the gang problem in Boston and official responses to it, the authors had been working off and on with Paul Joyce and other members of YVSF, especially Detective Fred Waggett, to construct a history of gang enforcement practices in BPD. One thing that kept coming up in those conversations, as it often did in the Working Group, was YVSF's Wendover Street operation. Part of that operation was aimed at the gun trafficker and seemed fairly straightforward: YVSF, working with ATF, and using traditional investigative techniques plus gun traces, had gradually uncovered a multistate trafficking ring and shut it down. Another part, aimed at ending the outbreak of violence that had first drawn attention to Wendover Street, was less clear cut. Joyce, YVSF, and the Streetworkers (along with their partners in probation and elsewhere) had clearly done something remarkable. The problem had been all but completely suppressed, and the area had remained quiet for the 9 months that had since passed. YVSF officers reported Wendover Street gang members actually pleading with them to remain at the end of the operation, because the area was then so safe that they wanted it to stay that way. Above all else, the extraordinary picture existed of young gang members turning over their guns to the police-literally walking up to YVSF's Warren Street headquarters with paper bags full of guns and dropping them off. How had this happened? Focused attention. Pressed on this point, Paul Joyce--who was, in general, reluctant to talk, particularly about himself and his work--responded with a cryptic, "We were honest with them," or the equivalent: "We just told them the truth." What this meant, exactly, remained a mystery. Waggett and other YVSF officers were more forthcoming. YVSF had focused much of its attention on the area for a 3-month period. Part of its energy was devoted to developing the case against gun traffickers. Another part was devoted, in conjunction with its partners in other agencies, to disrupting the gang and discommoding its members in any legal way possible. YVSF officers enforced every law they could against gang members, shutting down street drug sales and making arrests for trespassing, public drinking, overt drug use, and disorder offenses. They ran gang members' names against BPD's extensive database of outstanding warrants, found many, and served them. A Registry of Motor Vehicles search revealed that gang members had no drivers' licenses; YVSF took their unregistered cars away. Probation officers focused Night Light visits and other street probation enforcement on gang members. One particularly adamantine gang member, who was effectively whipping his peers on to defy the police, turned out to be under DYS supervision; DYS placed him in a youth correctional facility in western Massachusetts where he had no friends. The Streetworkers then moved in, offering services and striving to broker a resolution to the Wendover Street gang's feud with its nearby Magnolia Street rivals. YVSF officers, probation officers, and Streetworkers reached out to gang members' parents, many of whom were desperate over their children's behavior and willing to help. An honest approach. This appeared to be a classic crackdown, albeit one of uncommon subtlety that incorporated an unusual breadth of agencies. But it had one unique component. Conversations with Fred Waggett finally revealed what Paul Joyce had meant by "honesty" and "truth." Throughout the operation, YVSF officers, probation officers, and Streetworkers had told gang members directly why the crackdown was occurring and what it would take to make it stop. "We're here because of the shooting," the authorities had said. "We're not going to leave until it stops. And until it does, nobody is going to so much as jaywalk, nor make any money, nor have any fun." Every week or so, Joyce, Waggett, or another YVSF officer would visit DYS in the western part of the State and talk to the gang member there. "Want to go home?" he was asked. "Ready to tell your friends to knock it off?" It took some time, but the message eventually got through to him and the rest of his crew. As a final condition, YVSF told gang members it would not leave until it had the illegal guns they were holding. "We knew by then who had what: makes, models, calibers," Waggett said. "And we pointed fingers and made them turn them over." At which point, with an explicit warning that renewed violence would bring a return of the pressure, the operation ended. "Let us see if we've got it," the authors said. "You focused on this gang because of the violence. You went in and told them why you were there and what it would take to get you to leave. You pulled every legal lever you could to impose costs on them, which turned out to be fairly easy because they were selling drugs, on probation, and otherwise chronically offending. You focused just on the gang and not on other kids in the same neighborhood. And they eventually complied, and you left." "That's right," Waggett said. "And we've done it a few other times, when other gangs have gone critical, and it always worked then, too." Designing Ceasefire This was, as far as the authors were concerned, a radically different type of demand reduction than the usual violence prevention training, dispute resolution, and similar facilitative approaches. It made powerful, strategic use of authority with a group of chronic offenders, who could probably be reached only by using approaches that incorporated authority. In an exceedingly clever fashion, it turned gang offenders' behavior against themselves, taking advantage of the vulnerabilities created by their chronic misbehavior and turning them into a violence prevention tool. It used many of the existing strengths of criminal justice agencies--such as conducting drug investigations, patrolling streets, and supervising probation--and marshaled them into a preventive regime intended not to take gang members off the street, but to protect them from themselves and other gang members. Through the Streetworkers, it incorporated social services, gang mediation, and other opportunities for those willing to take advantage of them. And, because of the explicit communication with gang members, the approach seemed to be fundamentally fair. "Here's how the game's going to be played," gang members were told. "After this, it's up to you." Was this the answer the Working Group had been searching for? Could the Wendover approach to violence prevention, henceforth used only rarely and in extraordinary circumstances, be made the routine response to gang violence in Boston? And if it were, did Working Group members believe it would have the strong preventive impact they had been seeking? During the summer and early fall of 1995, conversation in the Working Group focused increasingly on these possibilities. Modifying the strategy A positive feeling developed: If the Working Group could organize member agencies behind a Wendover approach, and if it were given the right institutional support, success might follow. After a fair amount of application, a basic strategy emerged. Despite wide support for marshaling agency resources behind a Wendover-style violence prevention strategy, a problem remained: The incidence of violence and the number of active gang feuds were greater than could be handled using the intensive Wendover approach. The Working Group surmised it could put in place, at best, three to four such focused operations at any given time, which would not always be enough. The group decided to put the new regime in place gradually, using the network of gang conflicts in the city as a template (see exhibit 1-4). If gangs A and B were fighting at the time the new strategy kicked off, they would get the Wendover treatment. Once they were calmed, they would get one message: We will desist, but the moment that violence starts up again, we will be back. Because gangs A and B were now most vulnerable to any other enemies they might have, and because the peace that had just been won was most vulnerable to being broken by such rivalries, those other enemies would be the Working Group's next target. The rival gangs would get a visit and a message: "See what just happened to gangs A and B? Well, they are off limits now, and if you provoke or engage with them, you will get the same treatment." Moving quickly enough to make an impact but slowly enough to be able to deal with any problems that emerged, the Working Group could make its way across the network until more or less all the gangs were effectively proscribed from hurting one another.[26] That was the hope. At the same time, work by ATF and BPD to crack down on gun traffickers, especially those supplying violent gangs, would serve to inhibit easy access to weaponry. The Wendover approach had worked because the actual intervention was carefully tailored to the gang in question. It offered a basic script, but that script would need to be applied differently to different gangs in different situations. Once the Working Group decided to subject a gang or gangs to the intervention, it would be necessary to share information, and perhaps gather more, and figure out what to do in this particular case. Working Group members were satisfied that they usually knew enough to mount a Wendover-type operation, however. In dry runs, picking gangs at random to see how such an operation would be structured, if necessary, an operation could always be designed that seemed sufficient to attract gang members' attention. Communicating Communicating clearly Direct communication with gang members, a key element in the logic of the overall strategy, appeared necessary at two levels. First, the Working Group wanted all gangs in the city to be aware that new rules existed regarding violence and to understand what they were and how to avoid coming to the Working Group's attention. Second, when focus on a particular gang became necessary, it wanted to make that gang understand that violence had drawn the official attention, and only a cessation of violence would lead to its easing. This communication could be carried out one on one by YVSF officers, probation officers, Streetworkers, and others. In addition, the Working Group decided to drive the message home by holding a series of semiformal meetings with selected gangs. Gangs would be brought into the Dorchester courthouse, either through the Streetworkers' good offices or by requiring probationers to attend, and given a clear message about how violence would be treated in the future. These "forums," as they came to be called, were intended to make a graphic show of the new reality in Boston: A variety of agencies were working together and sharing information, with an enormous range of capacities to deploy and with violence as their only target. At the same time, the Working Group wanted to make it very clear that it was not offering a "deal,"-that refraining from violence would not win gangs the freedom to deal drugs or commit other crimes. The energies that went into implementing the new strategy would represent a fraction of the total criminal justice activity in Boston; normal enforcement by police, probation officers, prosecutors, and others would continue as before. The new reality-it was hoped-was that violence would now bring a special response in addition to "baseline" enforcement activities. The statement that captured the Working Group's attitude and was used routinely in forums was, "This is a promise, not a deal." Nobody had any illusions about the power of the new idea. To implement it, sustain implementation, and succeed in any way would be a long, difficult struggle. No explicit discussion occurred as to when to expect results, but implicit in the planning was an expectation that at least 1 or 2 years, maybe more, would be required to change the streets in any meaningful way. At the same time, many in the group believed strongly enough in the idea that gang violence was begetting gang violence to at least entertain what the authors called the "firebreak" hypothesis: If violence could be quelled for a meaningful period, it might not naturally reemerge. The streets were violent now, so they stayed violent; if they were safe, might they not remain safe? Any such development, however, was several very difficult and tenuous steps into the future. At the moment, the consensus on the Working Group was that the new strategy was, at least, worth trying. Garnering support The original pledge to the agency heads who had detailed representatives in the Working Group was that any promising strategy that emerged would be shared with them and their support solicited. During the last months of 1995 and early 1996, the authors made a series of presentations to those agencies detailing the research findings and the intervention strategy that the Working Group had crafted. Although the basic diagnosis held some surprises, it seemed to reflect their views (Ralph Boyd, a long-time gang prosecutor in the U.S. Attorney's Office, reacted to a slide of the network of gang rivalries by leaping out of his chair, kneeling in front of the screen for a better look, and shouting, "Yes! That's it!"). They too thought the strategy worth pursuing. Sentiment was strong at both the policy and working levels to put the operation in place and generate some results before making a formal public announcement. Paul Joyce and the authors briefed key outside figures, however, including the presiding judges in both Dorchester and Roxbury district courts and the Ten Point Coalition's Rev. Eugene Rivers.[27] All appeared to find the strategy worth trying and, if implemented properly, unobjectionable. A sketchy, but formal, protocol was written by the Working Group to specify the steps to be taken during focused gang enforcement actions. BPD Commissioner Paul Evans, in particular, seemed to have faith in YVSF and the larger application of the Wendover approach. "[The gangs] will have to behave," he wrote soon after in a BPD strategy document on guns and youth violence. "They won't have any choice."[28] The Working Group hoped that he was right. Applying the Strategy In fact, the strategy--then called within the Working Group by various names, such as "Scrap Iron II" and "Wendover Plus"--was already under way.[29] Late in 1995, Paul Joyce and several other detectives in YVSF had launched an operation aimed principally at the Intervale Posse. The Intervale Posse Widely regarded as perhaps the worst crack-era gang in Boston, the Intervale Posse had essentially taken over its Roxbury neighborhood. Led by Sam "Sam Goody" Patrick and working primarily out of a large, wooded lot that it had equipped with couches, television sets (powered by extension cords run out of nearby houses that residents had abandoned because of gang activity), a large barbecue, and even firing ranges, the gang sold crack cocaine, invaded homes, and attacked young people who dared to venture onto its turf (a huge "Adidas tree" adorned with hundreds of pairs of shoes taken from interlopers was a principal feature on the lot). "Intervale was king of the hill in Boston," said Gary French. "[Its members had] been involved in a string of unsolved homicides in that area going back years. Everybody on the street knew Intervale, and nobody messed with them." Fed up with the gang's years of domination, Joyce aimed a Wendover-style operation at it and three other allied gangs nearby. He had intended to put in place the same gradual, meticulous blanket that had worked on Wendover Street: a heavy police presence backed up by probation officers, Streetworkers, DYS, community organizers, and constant communication with the gangs. In the end, it worked out somewhat differently. Early in 1996, Lt. Det. Gary French replaced Joyce as commander of YVSF (Joyce was promoted to BPD's homicide squad). French came, most recently, from assignments in narcotics and as a district patrol supervisor in Dorchester. He was in tune with YVSF's methods and quickly integrated himself into the Working Group. He agreed with and was willing to support the strategy being pursued. He also wanted to make some changes, particularly within his own unit. Paul Joyce and his team of detectives had essentially acted as a squad within a squad, an elite unit within the elite YVSF. French felt it essential to involve all of YVSF, which included both patrol personnel (who generally worked in plain clothes) and officers detailed from such other departments as the Massachusetts State Police. French began including other key members of the 40-strong YVSF in Working Group proceedings and took steps to market the Working Group's strategy with YVSF, including christening the strategy as "Operation Ceasefire." French also took over and refocused the Intervale operation with the Intervale Posse as its sole target. In view of Intervale's long, violent history and its resistance to pressure being applied--French knew Sam Patrick personally and was making repeated visits, telling him to calm things down, to no avail--he quietly reached out to the Boston office of the U.S. Drug Enforcement Administration (DEA). DEA was conducting a slow-moving investigation aimed at Intervale; at French's urging, that investigation was reinvigorated and adapted to incorporate YVSF personnel and other Working Group agencies. The Vamp Hill Kings While the Intervale operation was under way, several other gang violence problems erupted. Chief among them was a situation involving the Vamp Hill Kings. Another prominent, well-established Boston gang, with turf on Dorchester's Bowdoin Street, the Kings were going through an internal struggle in early 1996. Kings were openly feuding, carrying guns, shooting at each other, firing "expressive" shots up and down the street at night, and otherwise wreaking havoc in the neighborhood. Three King-on-King homicides would eventually result. French focused the Working Group's attention on the Vamp Hill Kings. The violence on Bowdoin Street became Operation Ceasefire's first real test. In March 1996, several planning sessions brought in not only core Working Group participants but also district (precinct) patrol and beat officers and prosecutors from the Safe Neighborhood Initiative--a joint operation between BPD, the Suffolk County District Attorney's Office, the Office of Massachusetts Attorney General Scott Harshbarger, and neighborhood groups focused on an area of Dorchester that included Bowdoin Street. Everybody involved knew the Kings and their key players, and together the various agencies were able to mount a comprehensive operation. Applying pressure. A critical decision involved the communications strategy. Should the Working Group warn the Kings before the operation, or after? A strong feeling existed at the working level that, at least this first time, the operation should come first to build credibility. A heavy BPD presence--YVSF officers and district and special unit officers--squeezed the street drug trade by close to 80 percent. Billy Stewart increased Night Light visits on the Kings' leaders and walked the streets during the day to make sure that probationers were behaving (in a twist he invented on the spot, he also visited the parents of Kings not on probation, warned them that their sons were in trouble, and urged them to keep them off the street). YVSF and district officers maintained a heavy uniformed presence, making disorder arrests and sometimes posting officers full time outside the houses of the main players in the Kings' two warring factions. Four Kings turned out to be under DYS supervision; DYS soon took them off the street. One King was determined to be a resident alien subject to deportation for his Massachusetts criminal record; French reached out to the U.S. Immigration and Naturalization Service and made it happen. A King stopped one evening by a YVSF officer was carrying a mask, gloves, and a semiautomatic pistol, which he drew on the officer before thinking twice and dropping the weapon. Normally that case would have been prosecuted by Massachusetts authorities; instead, alerted by the Working Group, the U.S. Attorney took the case. Judges, kept up to date by probation officers, imposed strict bail conditions on Kings arrested during the operation. ATF agents rode and walked the streets with BPD officers. Gary French even brought the Massachusetts Society for the Prevention of Cruelty to Animals in to remove a number of pit bulls the Kings had trained as fighting dogs. In their contacts with the Kings, YVSF and probation officers reminded them constantly that pressure was being applied because of the violence. And, throughout, Streetworkers delivered the same message and urged the gang to end the dispute and take advantage of the services the city and others could offer. Still the Kings resisted. In one particularly awful incident, probably the low point of the entire Boston Gun Project, a young King was murdered in his own house only minutes after Gary French, Billy Stewart, and ATF agents had warned his mother that their intelligence suggested she should do everything she could to keep him off the streets. Throughout the operation, the Working Group was never certain that it could make the strategy work. Members were in constant touch, in meetings at Warren Street and over the phone, to update each other on what seemed to be going on, reshaping the operation day by day and even hour by hour. Sending the message. By early May, however, Bowdoin Street was quiet. Whether this was due to the impact of the operation, the natural playing out of the dispute, or some combination thereof was impossible to determine. But quiet it was, and the Working Group moved on to the next step. Late on the afternoon of May 15, 1996, the Streetworkers brought about a dozen Kings, along with other Bowdoin Street community members, into the Dorchester courthouse for the Gun Project's first forum (DYS brought several Kings from their locked facilities, who sat off to the side in restraints). They were met there with an unprecedented sight: representatives of all the Working Group agencies, plus several others such as Safe Neighborhood Initiative prosecutor Marcie Jackson, sitting together at a table at the front of the room. They were flanked by posters spelling out what had just happened on Bowdoin Street (see exhibit 1-5) and collections of news clippings detailing, for example, the Federal prosecution of the Bowdoin Street gun carrier; these were also given to the audience as handouts. Billy Stewart was master of ceremonies, as he would generally be at subsequent forums. "Thanks for coming," he told the audience. "This isn't a sting, everybody's going to be home for dinner, we just wanted you to know a few things. And this is nothing personal, either; this is how we're going to be dealing with violence in the future, and you just happened to be first. So go home and tell your friends about what you hear today." The forum was dramatic. In essence, the Working Group's message to the Kings was that they and their activities were known, and although the group could not stop every instance of offending, violence would no longer be tolerated in Boston. The message also stressed that the crackdown was being conducted in large part to protect gang members, and it made its rationale clear and assigned responsibility for the future to individual gang members. Each representative took a few minutes to speak about the powers his agency possessed and how such powers would be exercised through Operation Ceasefire. Many gang members in the audience smiled and scoffed. They stopped when Ted Heinrich, as Assistant U.S. Attorney assigned to the group, spoke. "This kind of street crime used to be a local matter," he said: Not any more. [The] Attorney General cares more about youth violence than almost anything else. My boss works for [the Attorney General], so that's what he cares about more than anything else. Right now, the youth violence in Boston is happening in your neighborhood, which means that the U.S. Department of Justice cares about you. We can bring in the DEA, we can bring in the FBI, we can bring in the ATF; we can prosecute you federally, which means you go to Lompoc, not stateside, and there's no parole in the Federal system any more: You serve your time. We don't want to do that, and we won't if we don't have to, but it's violence that will get that kind of attention. The room became more silent when the panel turned to Freddie Cardoza, who was featured on his own poster and handout (see exhibit 1-6). "One bullet," said Gary French. "We are not putting up with this stuff any more." The talk was not only about sanctions. Tracy Litthcut, deliberately sitting in the audience rather than at the table to distinguish himself from the enforcement agencies, delivered an impassioned speech. "We know you're all caught up in something you can't control," he said. "We know it's dangerous out there. And we'll help, any way we can. If you need protection from your enemies, if you want a job, if your mom needs treatment, if you want back into school, tell us; here's my phone number." But Litthcut also made sure there were no mixed messages being delivered. "If you don't hear what's being said to you today," he told the Kings, "it's on your heads. Take what we're offering. I've been to over 100 funerals, and I'm not going to any more. The violence stops now." YVSF officers, probation officers, and Streetworkers reported that word of the meeting spread quickly on the streets, with gangs all over Boston talking about the forum and Freddie Cardoza. An unprecedented calm fell over Bowdoin Street, according to YVSF and Streetworkers, who were still spending a lot of time there. A key King gang member visited Billy Stewart and asked what kind of record Cardoza had had that deserved prosecution as an armed career criminal; Stewart simply held up a thick copy of Cardoza's sheet and matched it with the Kings' equally thick one. The Streetworkers, working to consolidate the situation, moved into Bowdoin Street with a summer jobs program aimed at both Kings and non-Kings. And, as it had on Wendover Street, the calm held. The Working Group continued its biweekly meetings during the summer of 1996. The sessions were straightforward and asked about the incidence of any gang violence or eruptions of shots fired, and whether anything should be done. Incidents continued to occur, but no further Bowdoin-scale operations took place. That summer, the Working Group discovered the power of the communication strategy, which the authors were beginning to think of as "retail deterrence." In several instances, YVSF and probation officers visited gangs that appeared to be headed for trouble. "We have our eyes on you," they would say. "We are the ones who brought you Bowdoin Street. If this violence does not stop, you are next." Tracy Litthcut and the Streetworkers would back up the message. "They are watching," they would say. "They know your names." In every instance, the trouble stopped. Tracy came back to the Working Group a few weeks after putting one gang on notice, shaking his head in wonder. "They believe," he said. "There's nothing happening." Intervale II The Intervale Posse did not get it. That Working Group operation had continued over the summer, both the overt side--heavy Bowdoin Street-style interagency pressure--and the covert aspect involving the DEA. The situation did not improve. On August 29, 1996, an early-morning sweep resulted in the arrest of 15 key members of the Intervale Posse on Federal drug charges and arrests of 8 more on State charges. Sam Patrick's first question to the arresting officers was telling. "State or Federal?" he asked. When the answer was Federal, he hung his head and made no reply. The sweep made headlines in Boston for weeks. The Working Group wanted to make sure that those on the streets understood exactly what had happened. In a flurry of forums at the courthouse, with young inmates being held in DYS lockup, and in the Roxbury middle school that many students from the Intervale area attended, plus in innumerable one-on-one conversations on the street and through a new flier, the Working Group told its story (see exhibit 1-7). "We warned them; they didn't listen," the Working Group said: The papers say this was a drug operation, and it was, but it really happened because of the violence, and it's violence that will draw anything similar in the future. We have BPD, we have DEA, we have ATF, we have probation, we have parole, we have DYS, we have the U.S. Attorney, we have the county District Attorney, we have everybody. If we focus on you, you can't win, so don't make us [act]. In what felt like a particularly useful step, the Working Group, with U.S. Attorney Donald Stern in attendance, met with all Streetworkers, some 40 strong, at Warren Street at the end of May and asked for their help putting the message on the streets. Head Streetworker Tracy Litthcut had developed his own relationships with Working Group members, and his own trust in them, but the same was not necessarily true of his rank-and-file associates. "Is Tracy naming names to you?" some demanded. "No," Gary French replied. "We frankly don't need him for that. We just want him, and you, if you're willing, to tell the kids on the streets how business is being done now." In one particularly tense exchange, an older Streetworker asked, "Is everybody who might sell drugs on the street facing these huge Federal penalties? My son, God forbid, might choose to do a little of that; is he going to be exposed?" "No," Stern replied. "This is about violence. Only the key players in the most violent groups have to worry, and they'll get fair warning, just like Intervale did." "Ah," said the Streetworker. "In that case, the crew I work with in Mattapan asked me to send you guys a message. The message is, we got it, we're not doing anything, leave us alone." Effectiveness of the Strategy The Intervale operation marked a turning point. Gang violence in Boston declined abruptly. "I almost took my beeper in to have it checked," Gary French said at the time. "It just stopped going off." To everybody's surprise, it was unnecessary to repeat the crackdowns or move out gradually along the gang network as originally planned. The Working Group continued to meet regularly and occasionally to reach out to gangs with warnings, but no further operations on either the Bowdoin Street or Intervale scale were necessary. During this period, Gary French expanded and refined what the Working Group viewed as its operational repertoire. The way the Operation Ceasefire group had imagined working, he pointed out, was too inflexible for the situations the streets kept putting forward. Some violent or potentially violent situations had to be addressed as soon as YVSF or Streetworkers heard of them; decisionmaking could not always wait for the Working Group to convene or even consult. At the same time, not all situations and not all gangs seemed to require or deserve a full-force Ceasefire intervention, particularly not the kind of focused Federal attention visited on the Intervale Posse. As French articulated it, Operation Ceasefire had four basic levels of intervention available to it: o Level One was a warning, through forums or other means, to a particular group or groups to stop the violence. o Level Two was near-term street enforcement focused on a group or groups, deliverable largely within the capacities of the YVSF and the police department, but perhaps with some help from other agencies: heavy police presence, drug market disruption, warrant service, misdemeanor enforcement, and similar actions, maybe with some help from probation or DYS. Both levels one and two could be launched by YVSF as it deemed fit, based on its own intelligence or on input from other agencies, the Streetworkers, or community sources. o Level Three was a large, interagency, heavily coordinated operation-- the Vamp Hill Kings intervention remained the model for this--that was readily apparent to the target group, with sanctions remaining primarily on the State side. o Level Four, which all hoped would rarely be necessary, was for those groups that were both violent and deemed essentially unsalvageable: undercover, gangwide investigations making heavy use of Federal sanctions and designed, as with the Intervale Posse, to permanently dismantle the group. In reality, nothing more than a level-one intervention was required in Boston for roughly 1 year, when a violent new gang, calling themselves the Buckshot Crew and operating in Dorchester's Codman Square area, was first warned in a forum and then, having ignored the warning, rounded up weeks later on street drug charges. A fascinating and encouraging aspect of that operation was that other gang members who approached DYS officers during the summer of 1997 brought the Buckshot Crew to the Working Group's attention. "These guys are out of control, throwing their weight around," they told DYS. "They're going to ruin it for everybody; you have to do something." It is impossible to say with certainty what caused the falloff in youth homicide in Boston or exactly what part Operation Ceasefire played. (See page 40, "Effectiveness of the Strategy," for complete discussion of this question.) Because Ceasefire was conceived as an intervention aimed at interrupting the overall dynamic of violence in which all Boston gangs and gang members were involved, the operation could not be set up as a controlled experiment, with certain gangs or neighborhoods excluded for purposes of comparison. Some things are clear, however. Youth homicide in the city declined abruptly following the first gang forum in May 1996, and this low level continued through 1998 and 1999 (see exhibit 1-8). The streets have essentially remained quiet, but the reason for this is unclear. One possibility is that the authorities in Boston delivered a deterrent message so powerful that it remains salient on the streets without repeated reinforcement. More likely, as most on the Working Group believe, is that one of the guiding ideas behind Operation Ceasefire was correct: Gang violence was generating gang violence, and in the absence of constant fear and provocation, the temperature on the streets has gone down. "Peace begets peace," said Emmet Folgert, head of the Dorchester Youth Collaborative, who works closely with at-risk youths. Jed Hresko, a Streetworker in Dorchester, supports that idea. "Things still happen on the street," he said, "but they don't spiral out of control any longer." It may be that the Working Group's largest aspiration--to create a "firebreak" across the cycle of youth violence--has actually been attained. Reflections on Problem-Oriented Policing The Boston Gun Project was framed as an attempt to bring the problem-solving model to bear on an important, large-scale issue. Problem solving is worth pursuing Perhaps the most fundamental lesson here is that the basic approach the project followed--serious, sustained, attention to an important problem, with ambitious goals--is worthwhile. This approach is an enormous departure from the reactive "we'll send another car" mode that was typical of policing until recently, and it may be relevant not only in matters of crime and criminal justice, but in other aspects of the public sector as well. One suspects that many difficult problems might appear less so if similarly addressed. However, it is possible, and important, to differentiate the production of a meaningful strategy from the actual efficacy of that strategy. Operation Ceasefire may or may not be principally responsible for the reduction in youth homicide in Boston, but most people would agree that it represents an innovative intervention strategy that was an advance over current practice and worth trying. Crime policy would benefit from having more such innovative strategies, and the problem-solving process is one way to produce them. A locus of responsibility must be created A closely related point is the importance of creating a locus of responsibility--be it a person, an agency, or an interagency group--for the problem in question. Boston was a hotbed of innovative practice aimed at preventing youth violence, and the people engaged in that work were extraordinarily dedicated, hard working, and serious. Nonetheless, nobody was in charge of either thinking about the problem or coordinating the various individuals and agencies actually doing the work. The Boston Gun Project Working Group assumed the former responsibility to a large extent and the latter to a lesser, but still important, extent. The authors are agnostic about how such responsibility should be established and structured--whether, for example, it should be vested in an individual, a lead agency, or, as here, an interagency group. Different problems and different settings no doubt require different arrangements. Having such a locus of responsibility, however, seems likely to be an important element in applying the problem-solving approach. Patience and uncertainty The issues likely to be addressed by sustained, large-scale, Boston Gun Project-style problem-solving exercises are probably difficult ones-- otherwise, lesser efforts would have been sufficient to deal with them. Such issues may not yield to ideas put forward on the first day or the first month of the process. The Boston Gun Project Working Group spent more than a year designing and fielding Operation Ceasefire. It could well have taken longer and still not produced anything very useful. Those who participate in and manage such exercises must be prepared for uncertainty. The power of frontline knowledge and perspective In the authors' view, the most important element in whatever success the Boston Gun Project may have enjoyed was the central role played by frontline practitioners. Their immersion in the problem shaped the project's research agenda and diagnosis; their knowledge was essential to the research itself; their sense of what was and was not possible or likely to be effective shaped the intervention design; their contacts and support in their respective agencies eased the authors' tasks in innumerable ways; their eventual support for the intervention bolstered its credibility and enhanced their agencies' willingness to risk implementation; and their operational activities were essential to the operation's actual implementation. Most important, of course, is that Operation Ceasefire emerged from work that they were already doing; it is extremely unlikely that the approach would have suggested itself otherwise. Working systematically with good frontline practitioners may well bear similar fruit where other problems are concerned. The power of an outside eye At the same time, the involvement of outsiders--the authors, in this case-- in the process was important. YVSF and its partners did not realize the strategic significance of what they had accomplished on Wendover Street, that in fact the kernel of a whole new approach to homicide prevention was represented there. Even if they had realized it, conveying the message to the top policy ranks in their own organizations would have been difficult. The authors, on the other hand, could see the significance of an existing practice, determine with the Working Group its potentially wider application, and articulate the developed strategy to practitioners' own agency leaders. Such outsiders need not be academics and probably need not even be from outside the agency or agencies involved. Having someone from outside the substantive and operational world of participating practitioners who is able to see practitioners' work from a different perspective and, where appropriate, validate it to agency managers, may be an important part of the larger problem-solving mix. The power of an appropriate description Youth homicide in Boston could have been accurately and variously described as a problem of-- o Gun violence. o Violence in poor minority communities. o Minority-male-on-minority-male violence. o Gang violence. o Drug-related violence. o Violence by young people with little respect for authority. o Violence by young people with poor values. All these descriptions were put on the table and seriously addressed in the Working Group process. But until the Working Group settled on yet another accurate description of the problem--1,300 more or less chronic offenders well known to the authorities, in 61 gangs, who largely hurt one another, principally through gang feuds--it was unable to make real progress in designing a solution. The point is not that this description was somehow more true than the others, but rather that it was both accurate and well matched to the capacities of the agencies represented on the Working Group and to their fundamental desire to make a substantial, near-term impact. None of the agencies, for example, had the capacity to deliver economic development in Boston; if they had, its impact on youth homicide would not have been felt for some years at best. One way to view the long Working Group process is as a continual worrying away at this question of an appropriate description and the question of what, if anything, participating agencies could do about the key elements of the description of the moment. When the key elements included, for example, poverty, absent fathers, or the availability of crack, little progress was forthcoming. Causal though those factors might have been, they offered the Working Group no foothold. The Working Group kept going until it came up with something it could work with. This process of arriving at an account of the problem that is relevant in policy and operational terms--not just in causal and historical terms--may be an important element in the problem-solving process. The power of research The research on guns, gangs, and homicide supported the Working Group process in important ways. None of this research was terribly sophisticated methodologically. But the ability to pin down key issues--where the guns were coming from, what guns gang members favored, who was killing and being killed, what role was played by gangs and gang conflict--kept the Working Group moving on solid ground, helped participating agencies understand the logic of the proposed intervention (and the relative illogic of at least some competing interventions), and helped justify the intervention to the public. Even the simplest research, when it illuminates elements of key policy significance, can be extremely important to the problem-solving process. The collapse of framework Finally, the authors cannot help but notice that both Operation Ceasefire and the best practices of YVSF and others in Boston from which it emerged do away almost entirely with the dualistic categories so favored by many students of youth violence and youth violence policy. Neither Operation Ceasefire nor the Wendover Street operation before it are enforcement operations in the usual sense of the term: They are not aimed at taking offenders off the streets, eliminating gangs, zero tolerance, or raising the probability that violent offenses will be met with stiff prison terms. Nor, however, are they composed of classic prevention elements. Fundamentally, they are not about the provision of facilitative services and they do not provide much in the way of additional resources, rely on changing offenders' character, or aim to change root cause elements of the environment. In essence, the Boston Gun Project was an exercise in deterrence and getting deterrence right. The power of the Boston Gun Project comes primarily from exercising authority in the service of prevention, a notion that in the current political and policy framework is unusual and, to many, even a contradiction. In important ways, the key insight in the Boston Gun Project remains the essential guiding belief of Working Group members and their various partners that gang members need, and indeed deserve, both help and control. These frontline practitioners--from Paul Joyce and Gary French to the Reverend Eugene Rivers--were, in the view of the authors, simply too close to the action to afford or be able to sustain more stylized notions. One wonders what other insights--impossible to imagine before being framed, painfully obvious afterward--may flow from the similarly sophisticated pragmatism of other problem-solving efforts. Conclusion The Boston Gun Project was an attempt to bring problem-solving policing to bear on a large, important, and apparently intractable issue. It was also an experiment in an unusual partnership between academics and practitioners. It appears to have been successful on both counts. The resulting Ceasefire intervention is innovative and seems to have played an important part in reducing youth homicide in Boston. The Working Group process at the heart of the Gun Project seems to have made good use of the complementary strengths of both scholars and frontline practitioners. The authors emphatically do not hold the Gun Project up as an ideal problem-solving exercise: There may be, and probably are, better ways to structure and carry out problem-solving efforts. But it appears to have been a good enough way to go about things and thus represents, at least, evidence that the basic notions behind problem solving are powerful enough to generate important contributions to preventing crime and helping troubled communities. Part II, Measuring the Impact of Operation Ceasefire, describes specifically the effects the project had on youth violence. Notes 1. James Alan Fox, 1996, Trends in Juvenile Violence: A Report to the United States Attorney General on Current and Future Rates of Juvenile Offending, Washington, DC: U.S. Department of Justice. 2. See Lois Fingerhut, D. Ingram, and J. Feldman, 1992, "Firearm and Nonfirearm Homicide Among Persons 15 Through 19 Years of Age," Journal of the American Medical Association 267:3048-58. See also Alex Kotlowitz, 1992, There Are No Children Here: The Story of Two Boys Growing Up in the Other America. New York: Anchor Books; and Terry Williams, 1989, The Cocaine Kids: The Inside Story of a Teenage Drug Ring, New York: Addison-Wesley. 3. G. David Curry, Richard A. Ball, and Scott H. Decker, 1996, "Estimating the National Scope of Gang Crime from Law Enforcement Data," in Gangs in America, 2d ed., edited by C. Ronald Huff, Thousand Oaks, CA: Sage Publications. 4. See Franklin Zimring, 1997, "Juvenile Violence in Policy Context," Valparaiso University Law Review 31:419-26. 5. In 1990, Boston's total homicide rate was 27 per 100,000 population. By comparison, Washington, D.C.'s 1990 total homicide rate was 78 per 100,000; New Orleans' rate was 61 per 100,000; and Baltimore's rate was 41 per 100,000. Federal Bureau of Investigation, 1991, Crime in the United States, 1990. Washington, DC: Federal Bureau of Investigation, U.S. Department of Justice. 6. Lt. Gary French, Commander, Youth Violence Strike Force, Boston Police Department; personal communication. 7. Weapons offense arrests of black males ages 24 and under more than doubled from approximately 160 in 1988 to more than 350 in 1991. Gun assault arrests of black males ages 24 and under increased from about 70 in 1988 to more than 120 in 1991. 8. Before the youth violence epidemic, yearly counts of Boston youth homicides averaged slightly less than 30 per year between 1981 and 1987. After the sudden rise between 1988 and 1990, Boston averaged about 44 youth homicides per year between 1991 and 1995. 9. See Centers for Disease Control and Prevention, 1996, Youth Risky Behavior Surveillance Survey, 1995. Atlanta, GA: Centers for Disease Control and Prevention. It is not possible to tell from the survey what the students were afraid of. Reason suggests, however, that violence and related street activity may have driven their fear to a considerable extent. 10. See David M. Kennedy, 1994, "Can We Keep Guns Away From Kids?" American Prospect 18:74-80. 11. Ibid. 12. Ibid. 13. Joseph F. Sheley and James D. Wright, 1993, Gun Acquisition and Possession in Selected Juvenile Samples. Washington, DC: National Institute of Justice, U.S. Department of Justice. 14. BPD has two main operation divisions: Field Services, which encompasses patrol and most uniformed operations; and Investigative Services, which encompasses detectives, the Major Crimes Unit, and most investigative operations. The Anti Gang Violence Unit/Youth Violence Strike Force, the part of BPD most heavily involved in the Gun Project and Operation Ceasefire, is unique within the department in being constituted of both detectives and patrol officers and reporting jointly to both divisions. 15. In early 1995, the Youth Violence Strike Force was an administrative unit within the larger Anti Gang Violence Unit; late in 1995, the two were merged and retained the Youth Violence Strike Force name. The authors worked with personnel from both units. For the reader's convenience, the Youth Violence Strike Force label is used throughout this report. 16. Ronald P. Corbett, Bernard L. Fitzgerald, and James T. Jordan, 1998, "Boston's Operation Night Light: An Emerging Model for Police Probation Partnerships," in Community Corrections: Probation, Parole, and Intermediate Sanctions, edited by Joan Petersilia. New York and London: Oxford University Press. 17. Key members in the Working Group have included Gary French, Robert Fratalia, Jim Jordan, Paul F. Joyce, Jr., and Fred Waggett of the Boston Police Department; Lennie Ladd and Phil Tortorella of the Federal Bureau of Alcohol, Tobacco and Firearms; Richard Skinner and William Stewart of the Department of Probation; Timothy Feeley, Marianne Hinkle, and Ted Heinrich of the Office of the U.S. Attorney; Lynn Brennan, John Burke, Hugh Curran, Elizabeth Keeley, and Mark Zanini of the Suffolk County District Attorney's Office; Chris Byner and Tracy Litthcut of the Boston Community Centers; Tim Zadai of the Department of Parole; James McGillivray of the Crime and Justice Foundation/Project Turnaround; Sonya Aleman, Jack Arnold, and Dan Tracy of the Department of Youth Services; Mike Hennessey of the Boston School Police; and the authors of this report. 18. Twenty-one was the somewhat arbitrary age used as a cutoff for "youth" in the authors' studies of youth homicide in Boston. 19. Of the roughly 40 percent of firearms ATF could not trace, approximately 20 percent could not be traced because the serial number on the gun was obliterated-serial numbers are usually obliterated to conceal illegal firearms trafficking. Another 10 percent were too old to trace-guns manufactured prior to 1968 cannot be traced because manufacturers and dealers were not yet required to keep records on transfers of guns. The remaining 10 percent could not be traced for a variety of reasons, but primarily because the request forms necessary to successfully trace guns were incorrectly filled out or contained incomplete or inaccurate information on the pedigree of the gun. Only 1 percent of these guns could not be traced because they were stolen from the retail dealer. For a complete discussion, see David M. Kennedy, Anne M. Piehl, and Anthony A. Braga, 1996, "Youth Violence in Boston: Gun Markets, Serious Youth Offenders, and a Use-Reduction Strategy," Law and Contemporary Problems 59:147-96. 20. For example, 90 percent of Lorcin pistols showed a time-to-crime of 2 years or less. See Kennedy, Piehl, and Braga, 1996. 21. See Kennedy, Piehl, and Braga, 1996. 22. Massachusetts law requires that records of all firearms transfers, including those of used guns, be recorded with the Department of Public Safety. At the time described, those records were being warehoused in volume on paper and, for all practical purposes, were not searchable or otherwise useful. 23. See David M. Kennedy, Anthony A. Braga, and Anne M. Piehl, 1997, "The (Un)Known Universe: Mapping Gangs and Gang Violence in Boston," in Crime Mapping and Crime Prevention, edited by David L. Weisburd and J. Thomas McEwen. New York: Criminal Justice Press. 24. By this point, the Working Group had adopted two informal rules that helped keep this process focused: "Everybody's entitled to his own opinions, but not to his own facts" (that is, if a matter is sufficiently central, such as the sources of gun trafficking or the gang involvement of homicide victims, an attempt should be made to pin it down with appropriate rigor) and "If we can't think of any way to influence it, let's move on." The Working Group was no more resistant than other discussion groups to fulminations about such things as family values, the fact that contemporary youth have no respect for authority, or the weak performance of various actors and institutions-both public and private. And although it proved essential to let those views air, the group became fairly efficient at bringing the conversation back to matters it could, at least in principle, influence. 25. Mark H. Moore has described the point person on such a project as "someone who should wake up every morning feeling sick to [his] stomach [that] it's not going to work." This captures the authors' experience well. 26. As will be related below, this plan was not necessary in the end. It may remain a useful idea, however, for other jurisdictions. 27. Formal public notice would come in June 1996, after the Bowdoin Street operation ended, described infra below, when policy-level figures and Kennedy described the Ceasefire strategy at a youth violence conference held by the Boston Bar Association. See also David M. Kennedy, "Stopping Youth Violence With More Than a Shot in the Dark," Boston Sunday Herald, June 30, 1996. 28. Paul F. Evans, 1996, "Illegal Firearms: An Epidemic We Can Cure," Boston: Boston Police Department. 29. The Working Group used several names for the strategy, including "Wendover Plus" and "Scrap Iron II," before finally settling on "Ceasefire" in the fall of 1996. For the reader's convenience, the Ceasefire label is used here throughout. ------------------------------ Part II. Measuring the Impact of Operation Ceasefire Anthony A. Braga, David M. Kennedy, Anne M. Piehl, Elin J. Waring Without the support of a formal evaluation, Boston's Operation Ceasefire has been hailed in the media as an unprecedented success.[1] The substantial reduction in the city's yearly youth homicide numbers certainly suggests that something noteworthy happened after Operation Ceasefire was implemented in mid-1996. Boston averaged 44 youth homicides per year between 1991 and 1995. In 1996, the number decreased to 26; in 1997, it decreased further to 15. Although these numbers demonstrate the sudden drop, they do not provide a rigorous assessment of Operation Ceasefire's contribution to the decrease. Consequently, the research team evaluating the operation's impact focused on four key questions: o Were significant reductions in youth homicide and other indicators of serious nonfatal gun violence associated with the implementation of Operation Ceasefire? o Did the timing of these reductions coincide with the implementation of Operation Ceasefire? o Were other factors responsible for Boston's reduction in youth homicide? o Was Boston's significant youth homicide reduction distinct relative to youth homicide trends in other major U.S. and New England cities? Study Design The design of this study, like that of most evaluations of crime prevention programs,[2] departs from the desirable, randomized controlled experimental approach.[3] Operation Ceasefire's strategy was aimed at all areas of the city with a serious youth violence problem. No control areas (or control gangs) were set aside within the city because: o The aim was to do something about serious youth violence wherever it presented itself in the city. o The target of the intervention was defined as the self-sustaining cycle of violence in which all gangs were caught up and to which all gangs contributed. o The communications strategy was explicitly intended to affect the behavior of gangs and individuals not directly subjected to enforcement attention.[4] Therefore, a comparison between areas and groups affected by the strategy to similar areas and groups not affected was not possible. Analysis of impacts within Boston associated with the Ceasefire intervention followed a basic one-group, time-series design.[5] In addition, a nonrandomized quasi-experiment was used to compare youth homicide trends in Boston with those in other large cities in the United States.[6] The Ceasefire intervention mostly targeted violence arising from gang dynamics, and earlier research suggested that most gang members in Boston were ages 24 and under.[7] Therefore, the key outcome variable in the assessment of the Ceasefire intervention's impact was the monthly number of homicide victims in this age group. The Boston Police Department's (BPD's) Office of Research and Analysis provided the homicide data used in these analyses. The youth homicide impact evaluation examined the monthly counts of youth homicides in Boston between January 1, 1991, and May 31, 1998. The preintervention period included the relatively stable, but still historically high, postepidemic years of 1991-1995 (see exhibit 2-1). In addition to preventing youth homicides, Operation Ceasefire was designed to reduce other forms of nonfatal, serious gun violence. Thus researchers also examined citywide data on monthly counts of "shots fired" citizen calls for service and official gun assault incident reports. These data were available for a slightly shorter time period than the homicide data set due to lags in BPD's data collection and preparation procedures. The nonfatal gun violence data were examined for the period January 1, 1991, through December 31, 1997. The computerized BPD incident data have what is, for study purposes, an important shortcoming: The records do not capture the age of the victim (this is, of course, also true for "shots fired" calls for service). To assess the effects of the intervention on gun assaults in specific age groups, information on the age of the victim was collected from hard copies of gun assault incident reports for the study period. Because the collection and coding of this information was time consuming, researchers chose to collect these data for only one high-activity police district. District B-2 covers most of Boston's Roxbury neighborhood and has a dense concentration of gangs: 29 of 61 identified gangs (47.5 percent) had turf in B-2.[8] Furthermore, of the 217 homicide victims ages 24 and under in Boston between 1991 and 1995, one-third were killed in B-2 (71 of 217, or 32.7 percent). Findings Reductions in youth homicides and nonfatal gun violence In these analyses, the date of the first direct communication with Boston gangs (May 15, 1996) was selected as the date Ceasefire was fully implemented; all elements of the strategy--the focus on gun trafficking, a special interagency response to gang violence, and the communication campaign--were in place as of that date. For convenience, June 1, 1996, was selected as the start of the "post" period. Exhibit 2-2 presents the monthly counts of youth homicides in Boston during the study period. The time series shows a 63-percent reduction in the mean monthly number of youth homicide victims from a pretest mean of 3.5 youth homicides per month to a posttest mean of 1.3 youth homicides per month. This simple analysis suggests that Operation Ceasefire was associated with a large reduction in Boston youth homicides. In any time series, however, intervention effects could be obscured by trends, seasonal variations, and random fluctuations.[9] Therefore, rigorous time-series models were used to analyze the data.[10] Since the underlying data were counts, Poisson regression generalized linear models were used to analyze the monthly counts of citywide youth homicide incidents, citywide shots-fired calls for service, citywide gun assault incidents, and youth gun assault incidents in District B-2.[11] Analyses suggest that the Ceasefire intervention was associated with statistically significant reductions in all time series, including: o A 63-percent decrease in the monthly number of youth homicides in Boston. o A 32-percent decrease in the monthly number of citywide shots-fired calls. o A 25-percent decrease in the monthly number of citywide all-age gun assault incidents. o A 44-percent decrease in the monthly number of District B-2 youth gun assault incidents. Timing questions Although the generalized linear model analyses revealed that implementation of Operation Ceasefire was associated with a significant reduction in youth homicide, the models do not establish whether the reduction started before or after the commencement of the program. In other words, the large decrease in youth homicides might have started several months earlier or later than the June 1996 commencement date. If this was the case, the association of Operation Ceasefire with observed youth homicide reductions would be weakened. An alternative approach to the standard time-series impact assessment methodology was to examine the entire time series for the point in time that experiences the maximal significant increase or decrease, if one exists.[12] To implement this test, the study team ran a model that checked for significant changes in the entire time series for each successive month.[13] These analyses suggested that the maximal significant decrease in the Boston youth homicide time series occurred in June 1996-about the same time Operation Ceasefire was fully implemented.[14] These results reinforce the observation that the implementation of the Boston program was associated with significant reductions in youth homicide. Examining rival causal factors The youth homicide and gun violence reductions associated with the Ceasefire intervention could have been caused or meaningfully influenced by other factors. Therefore, the study controlled for changes in-- o Boston's employment rate as measured by the Massachusetts Department of Employment and Training. o Boston's youth population (youths ages 14 to 24) as measured by the U.S. Bureau of the Census. o Citywide trends in violent Index crimes as measured by the Federal Bureau of Investigation Uniform Crime Reports. o Homicide victimization among older victims (ages 25 and older). o Youth involvement in street-level drug market activity as measured by BPD arrest data. Admittedly, these controls are far from ideal. For example, measuring changes in Boston's citywide youth population does not directly measure population changes among the target audience: gang-involved youth offenders. However, these variables represent the best available information on these alternate explanations for Boston's decrease in youth homicide. When these control variables were added to study models, the findings did not substantively change. The significant reductions in youth homicide, shots-fired calls for service, all-age gun assault incidents, and youth gun assault incidents in B-2 associated with Operation Ceasefire remained when the control variables were added to the Poisson regression time-series models.[15] Public health initiatives. Other initiatives have been associated with the noteworthy decline in youth violence in Boston. In a series of editorials and public statements, Dr. Deborah Prothrow-Stith and other public health practitioners made the case that public health initiatives were responsible for the fall in youth homicide in Boston. Analysis suggests that this is unlikely.[16] No evidence suggests that public health interventions had a direct effect on youth homicide in Boston, although some indirect, unquantifiable effects may be thus attributed. The period of these interventions, nearly two decades beginning in the early 1980s, covers the increase of youth homicide in the city, its peak in 1990, and the period 1991-1995 (in which youth homicide was lower but historically high) before the abrupt decrease began (see exhibit 2-1). Although public health interventions may have had some impact over time, nowhere do they show the effect on youth violence demonstrated by whatever, in fact, caused the abrupt decline in rates of youth homicide and gun assault in 1996. During the summer of 1996, Boston was home to a small (on a citywide scale) but meaningful group of chronic, gang-involved offenders who, as demonstrated by their violent behavior, had not been reached by years of public health or similar interventions. Only a few months later, their behavior had changed. No proponent of the view that public health interventions caused the overall decline in violence has suggested a mechanism by which this particular change might have been accomplished. Moreover, various components of the Boston public health interventions have been evaluated, in Boston and in similar settings elsewhere.[17] Nowhere have they shown strong impact on violent behavior and victimization. Independent group activities. Similarly, Operation Night Light, Boston's innovative probation/police partnership, and Boston's Ten Point Coalition have variously been credited with direct responsibility for Boston's dramatic reductions in youth violence.[18] Here, too, the strong claim seems unlikely. Both Operation Night Light and the activities of the Ten Point Coalition date from 1992. No diminution in homicide is evident between 1992 and mid-1996. Had either venture had strong solo impacts on violence, those impacts should have been evident. Both probation officers involved with Operation Night Light and, from late 1996 until the present, Ten Point clergy were tightly integrated into many Ceasefire interventions and made valuable contributions to the problem-solving process. However, directly attributing the sudden decline to either group independent of the larger problem-solving enterprise does not seem warranted. Anti-trafficking effects. Finally, the degree to which violence reduction in Boston should be attributed to the prevention of illegal firearms trafficking must be questioned. Trafficking was, of course, a principal original focus of Boston's Gun Project, and attention to trafficking was one of Operation Ceasefire's two fundamental planks. Study investigators believe that evaluating the particular contribution of supply-side interventions in Boston is essentially impossible. Anti-trafficking efforts were implemented at the same time as violence deterrence efforts, and both might be expected to influence, for example, gun carrying, gun use, and the mix of illegal guns found on the street. A stand-alone trafficking prevention intervention would not face these difficulties and could lead to definitive answers on the impact of supply-side interventions. Operation Ceasefire, however, was not a stand-alone trafficking prevention intervention. Here, as well, the distinctive characteristics of the decline in homicide and shootings in Boston offer the best insight into what might have happened. Two things are certain. First, supply-side efforts cannot be responsible for the abrupt reductions in gun-related violence during the summer and fall of 1996. Most Boston trafficking cases followed that reduction, rather than anticipated it. Second, anti-trafficking efforts in Boston did nothing to reduce the existing stockpile of illegally acquired and possessed firearms in Boston. The guns held by gang members in Boston in May of 1996 were, for the most part, still held by them several months later when the violence reached its new, lower level. The change that had occurred was not in the extent of gun ownership but in gun use. The principal impact, therefore, was almost certainly a demand-side, deterrence-based effect rather than a supply-side effect. It may well be that anti-trafficking efforts strengthened and prolonged that impact. Whether any such effects were large or small cannot be independently established in this case. Violence trends in Boston relative to violence trends in other cities Although the within-Boston analyses support the conclusion that a large reduction in youth homicide and gun violence was associated with Operation Ceasefire, it is necessary to distinguish Boston's experience from national and regional trends in youth homicide. Many major cities in the United States have enjoyed noteworthy reductions in homicide and nonfatal serious violence.[19] Violence reductions in other cities could be associated with a number of complex and tightly interwoven endogenous or exogenous factors, such as positive changes in the national economy, shifts in the age distribution of offending populations, or the stabilization of urban drug markets. Moreover, many cities, most notably New York,[20] have implemented crime prevention interventions that have been credited with substantial reductions in violence. Because many U.S. cities experienced varying decreases in homicide, the study team believed it was important to determine whether other cities experienced a similar sudden, large decrease in youth homicides. The following analyses provide insight into whether Boston's reduction in youth homicide was part of national youth homicide trends and whether the impact associated with the Ceasefire intervention was distinct in magnitude from other youth homicide reductions that occurred during the time series period. To examine these important issues, the study team obtained monthly counts of the number of homicide victims ages 24 and under for Boston, 29 major New England cities,[21] and 39 major U.S. cities[22] from Supplementary Homicide Report (SHR) data for the period from January 1991 through December 1997. To compare youth homicide trends in Boston relative to youth homicide trends in major U.S. and New England cities, the researchers built a model that maximized their ability to control for trends, seasonal variations, and random fluctuations in the time series of each city. They used a generalized linear Poisson regression model that predicted monthly youth homicide counts as a function of simple linear trends within each city time series, nonlinear trends within each city time series, monthly effects within each city time series, intervention effects within each city time series, and a simple autoregressive component for each city time series.[23] Using the June 1996 intervention date, these models revealed that four cities--Boston, Dallas, Jacksonville, and Virginia Beach--experienced significant changes in the monthly count of youth homicides coinciding with the implementation of the Operation Ceasefire program. Examination of the trends in youth homicides in the other cities with significant intervention coefficients also supports the distinctiveness of the Boston case. Virginia Beach, for example, had a significant increase in youth homicides in June 1996, although the yearly counts of youth homicides were stable between 1995 and 1997.[24] Declines in Dallas and Jacksonville both began months earlier than the decline in Boston. The study team is not aware of any connection between youth homicides in these four cities. Although based on exploratory analysis, the presence of these differences undermines the argument that the changes in Boston reflect trends in other major U.S. cities. Of course, other cities may have experienced a significant decrease in youth homicide before or after Boston experienced its decrease. Therefore, the study team conducted an exploratory analysis to identify significant youth homicide reductions in other months during the time series. This analysis looked at youth homicides in 39 major U.S. cities and 29 major New England cities, with a varying intervention point for each month in the time series.[25] Five of 39 cities--Dallas, Los Angeles, New York, Philadelphia, and Tucson--experienced a sudden, significant reduction in youth homicides at some point in the time series, but no substantive link can be made between youth homicide trends in the five cities and Boston because of differences in the yearly trends across cities. Philadelphia experienced significant reductions in monthly counts of youth homicides in December 1993 through February 1994--30 months before the implementation of Operation Ceasefire--followed by a steady increase in youth homicide between 1994 and 1997. Tucson experienced significant decreases in monthly youth homicide counts between November and December 1995, but this sudden decrease was followed by an increase in youth homicides in 1997. Dallas experienced a significant decrease in the monthly count of youth homicides between March and May 1996; although this significant reduction coincides with the implementation of Operation Ceasefire, youth homicide in Dallas declined almost linearly between 1991 and 1997. Los Angeles experienced a significant reduction in monthly counts of youth homicides during June and July 1993, and New York experienced sudden significant reductions in monthly counts of youth homicides during March and April, then August and September, 1994. As in Dallas, youth homicide trends in Los Angeles and New York declined sharply during the mid 1990s. Superficially, the steady declines in Dallas, Los Angeles, and New York seem different from the trajectory of youth homicide in Boston. Overall, the results from this analysis do not support the idea that changes in Boston either followed or trailed national changes or changes in other major cities. Careful within-city studies are necessary to unravel the youth homicide trends in the six cities. Although some cities may have experienced a similar decrease, these analyses suggest that Boston's significant youth homicide reduction associated with Operation Ceasefire was distinct when compared with youth homicide trends in most major U.S. and New England cities. Discussion The research presented here shows that the Boston Gun Project was a meaningful problem-oriented policing effort that brought practitioners and researchers together in new ways and led both to a fresh assessment of the youth violence problem in Boston and to operational activities that departed substantially from previous practice. The principal intervention, Operation Ceasefire, was likely responsible for a significant reduction in the city's rates of youth homicide and gun violence. At first glance, the effectiveness of Operation Ceasefire in preventing violence may seem unique to Boston. The intervention was constructed largely from the resources and capacities available in Boston at the time and deliberately tailored to the city's particular violence problem. Operational capacities of criminal justice agencies in other cities will differ and youth violence problems in other cities will have important distinguishing characteristics. However, the Working Group's problem-solving process and the "pulling levers" approach to deterring chronic offenders may be transferable to other jurisdictions. A number of cities--Baltimore, High Point and Winston-Salem (North Carolina), Indianapolis, Los Angeles, Lowell (Massachusetts), Memphis, Minneapolis, New Haven (Connecticut), Portland (Oregon), and Stockton (California)--have begun to experiment with these frameworks and have experienced some encouraging preliminary results.[26] The Boston Gun Project applied the basic principles of problem-oriented policing to a substantial public safety problem. Addressing this problem required the involvement of multiple agencies and the community, as well as substantial investments in analysis, coordination, and implementation. The experience of the Gun Project suggests that deploying criminal justice capacities to prevent crime can yield substantial benefits. The problem-solving orientation of the project means that the problem definition, the core participants, and the particulars of the intervention evolved over the course of the collaboration. Operation Ceasefire itself was highly customized to the goals of the collaboration, the particular nature of the youth violence problem in Boston, and the particular capacities available in Boston for incorporation into a strategic intervention. Therefore, Operation Ceasefire as such is unlikely to be a highly specifiable, transportable intervention strategy. However, certain process elements of the Boston Gun Project, such as the central role of the line-level working group and the use of both qualitative and quantitative research to "unpack" chosen problems, should be generally applicable to other problem-solving efforts. Using the Working Group's problem-solving approach, criminal justice practitioners in other jurisdictions can develop a set of intervention strategies that fit both the nuances of their youth violence problem and their operational capacities. Although the resulting package of interventions may not closely resemble the tactics used in Operation Ceasefire, the frameworks will be similar. The "pulling levers" deterrence strategy at the heart of Operation Ceasefire was designed to influence the behavior and environment of the chronically offending, gang-involved youths identified by Gun Project research as the core of the city's youth violence problem. The deterrence strategy may have something to offer other jurisdictions in which problem-oriented policing efforts are aimed at violence that is rooted in similar factors. The Operation Ceasefire intervention is, in its broadest sense, a deterrence strategy. Much of the literature evaluating deterrence focuses on the effects of changing the certainty, swiftness, and severity of punishment associated with certain criminal acts on the prevalence of such crimes.[27] In addition to increasing the certainty, severity, and swiftness of sanctions associated with youth violence, the Operation Ceasefire strategy sought to enhance deterrence by advertising the law enforcement strategy and personalizing its application. It was crucial that gang youths understood the new regime being imposed by the city. The "pulling levers" approach attempted to prevent gang violence by convincing gang members that violence and gun use would bring consequences, so that they would choose to change their behavior. A key element of the strategy was the delivery of a direct and explicit "retail deterrence" message to a relatively small target audience regarding the response that specific behaviors would provoke. Boston's law enforcement agencies made gang-related violence more costly to participants. Knowledge of what happened to others in the target population was intended to prevent further acts of gang violence in the city. Operation Ceasefire's Working Group understood that law enforcement agencies generally do not have the capacity to "eliminate" all gangs or powerfully respond to all gang offending in gang-troubled jurisdictions.28 Pledges to do so, although common, are simply not credible. The Working Group recognized that, for the strategy to be successful, a credible deterrence message must be delivered to Boston gangs. Because the Working Group could deploy, at best, only a few severe crackdowns at a time, the Ceasefire intervention targeted those gangs that were engaged in violent behavior rather than expending resources on those that were not. Through this focused application of deterrence principles, Operation Ceasefire suggests a new approach to controlling violent offenders. Notes 1. See, e.g., Fox Butterfield, 1996, "In Boston, Nothing is Something," New York Times, November 21; and Gordon Witkin, 1997, "Sixteen Silver Bullets: Smart Ideas to Fix the World," U.S. News and World Report, December 29: 67. 2. Paul Ekblom and Ken Pease, 1995, "Evaluating Crime Prevention," in Building a Safer Society. Crime and Justice, Vol. 19, edited by Michael Tonry and David Farrington, Chicago: University of Chicago Press. 3. See, e.g., Lawrence Sherman, 1997, "Thinking About Crime Prevention," in Preventing Crime: What Works, What Doesn't, What's Promising, edited by the University of Maryland Department of Criminology and Criminal Justice, Washington, DC: Office of Justice Programs, U.S. Department of Justice. 4. David M. Kennedy, Anne M. Piehl, and Anthony A. Braga, 1996, "Youth Violence in Boston: Gun Markets, Serious Youth Offenders, and a Use-Reduction Strategy," Law and Contemporary Problems 59:147-96. 5. Donald T. Campbell and Julian Stanley, 1966, Experimental and Quasi-Experimental Designs for Research, Chicago: Rand McNally; Thomas Cook and Donald Campbell, 1979, Quasi-Experimentation: Design and Analysis Issues for Field Settings, Boston: Houghton Mifflin. 6. Peter H. Rossi and Howard E. Freeman, 1993, Evaluation: A Systematic Approach, Fifth Edition, Newbury Park, CA: Sage Publications; Cook and Campbell, 1979. 7. Kennedy et al., 1996; See David M. Kennedy, Anthony A. Braga, and Anne M. Piehl, 1997, "The (Un)Known Universe: Mapping Gangs and Gang Violence in Boston," in Crime Mapping and Crime Prevention, edited by David L. Weisburd and J. Thomas McEwen. New York: Criminal Justice Press. 8. Kennedy et al., 1997. 9. David McDowall, Richard McCleary, Errol Meidinger, and Richard Hay, 1980, Interrupted Time Series Analysis, Sage University Series on Quantitative Applications in the Social Sciences, Newbury Park, CA: Sage Publications. 10. The general class of auto-regressive integrated moving average (ARIMA) models can be used to good effect in detecting these three sources of noise in a time series (McDowall et al., 1980). The study used ARIMA models to unravel the error structure of the preintervention time series for each outcome measure to guide researchers in accounting for these sources of noise in their generalized linear models. ARIMA models were not used to assess the impact of the intervention. All outcome measure time series exhibited varying seasonal effects; that is, all time series had either seasonal moving averages (a shock that is felt once each season and then disappears), seasonal autocorrelation (e.g., August 1991 figures correlated with August 1992, August 1993, and so on), or both. To account for these seasonal effects in their models, researchers included dummy variables for each month. None of the time-series data shows significant nonseasonal autocorrelation (i.e., monthly counts serially correlated); therefore, they did not estimate a nonseasonal autoregressive component in the models. The preintervention time series varied as to whether a trend was present. Youth homicides and youth gun assault incidents in District B-2 were relatively stable during the preintervention time series, while citywide shots-fired calls and citywide gun assault incidents in B-2 exhibited simple linear downward trends. To account for trends in the series, researchers included a simple linear trend variable in the model. 11. Generalized linear models, configured as Poisson regressions in log linear models, were used in the deeper analysis of impacts associated with Operation Ceasefire to analyze the time-series data; see Annette Dobson, 1990, An Introduction to Generalized Linear Models, New York: Chapman and Hall; Peter McCullagh and John Nelder, 1989, Generalized Linear Models, Second Edition, New York: Chapman and Hall. For the details of the models and a full discussion of the resulting parameter estimates, see Anthony A. Braga, David M. Kennedy, Elin J. Waring, and Anne M. Piehl, 2001, "Problem-Oriented Policing, Youth Violence, and Deterrence: An Evaluation of Boston's Operation Ceasefire," Journal of Research in Crime and Delinquency 38(3): 195-225. Inserting youth homicides as the dependent variable, the basic model was as follows: Monthly youth homicide count = Intercept + intervention + trend + month dummy variables + error. The SAS GENMOD procedure was used to analyze these data; see SAS Institute, 1993, SAS/STAT Software: The GENMOD Procedure, Release 6.09, Technical Report P-243, Cary, NC: SAS Institute, Inc. 12. For the details of applying tests of structural breaks to evaluate changes over time, see Donald Andrews, 1993, "Tests for Parameter Instability and Structural Change with an Unknown Change Point," Econometrika 61:821-56. 13. These models excluded months at the very beginning and the very end of the series to allow enough data to control for trends and seasonal variations in the time-series data. See Anne M. Piehl, Suzanne J. Cooper, Anthony A. Braga, and David M. Kennedy, 1999, "Testing for Structural Breaks in the Evaluation of Programs," NBER Working Paper No. 7226, Cambridge, MA: National Bureau of Economic Research. 14. Piehl et al., 1999. 15. For the details of these analyses, see Braga et al., 2001. 16. Deborah Prothrow-Stith and Howard Spivak, "Turning the Tide on Violence," op-ed, The Boston Globe, November 24, 1996; Deborah Prothrow-Stith and Howard Spivak, "With Youth Crime Down in Boston, What Do We Do For an Encore?" op-ed, The Boston Globe, March 2, 1997. 17. Marc Posner, 1994, "Research Raises Troubling Questions About Violence Prevention Programs," The Harvard Education Letter 10(3); Alice Hausman, Glenn Pierce, and LeBaron Briggs, 1996, "Evaluation of Comprehensive Violence Prevention Education: Effects on Student Behavior," Journal of Adolescent Health 19:104-10. 18. On the Ten Point Coalition, see John Leland with Claudia Kalb, 1998, "Savior of the Streets," Newsweek, U.S. Edition, June 1: 20. On Operation Night Light, see Fox Butterfield, 1996, "In Boston, Something is Nothing," New York Times, November 21. 19. See, e.g., Alfred Blumstein and Richard Rosenfeld, 1998, "Explaining Recent Trends in U.S. Homicide Rates," Journal of Criminal Law and Criminology 88: 1175-1216. 20. George L. Kelling and William J. Bratton, 1998, "Declining Crime Rates: Insiders' Views of the New York City Story," Journal of Criminal Law and Criminology 88:1217-32. 21. Researchers selected all New England cities with populations of more than 60,000. The 28 cities included Bridgeport, Danbury, Hartford, New Britain, New Haven, Norwalk, Stamford, and Waterbury, Connecticut; Brockton, Cambridge, Fall River, Framingham, Lawrence, Lowell, Lynn, New Bedford, Newton, Quincy, Somerville, Springfield, and Worcester, Massachusetts; Portland, Maine; Nashua and Manchester, New Hampshire; and Cranston, Pawtucket, Providence, and Warwick, Rhode Island. Although it has only 50,000 residents, Burlington, Vermont, was included in this pool because it was the only "major" city in Vermont. 22. The study team ranked the top 40 cities according to U.S. Census population estimates in 1990 and 1996. In this procedure they observed that Fresno, California, and Tulsa, Oklahoma, were not in the top 40 in 1990 but were in the top 40 in 1996. St. Louis, Missouri, and Oakland, California, were in the top 40 in 1990, but not in the top 40 in 1996. Rather than exclude either pair of cities, researchers decided to keep both pairs in the sample. Removing Boston from this group of populous cities left 41 cities. After a close examination of these data, two cities (Washington, D.C., and New Orleans) were excluded due to extensive missing data. This left 39 major U.S. cities in the comparison group. The final pool of major U.S. cities included Albuquerque; Atlanta; Austin; Baltimore; Charlotte, North Carolina; Chicago; Cleveland; Columbus, Ohio; Dallas; Fort Worth; Fresno, California; Honolulu; Houston; Indianapolis; Jacksonville, Florida; Kansas City, Missouri; Los Angeles; Long Beach, California; Memphis; Milwaukee; Nashville; New York; Oakland, California; Oklahoma City; Philadelphia; Phoenix; Portland, Oregon; San Antonio; San Diego; San Francisco; San Jose, California; Seattle; St. Louis; Tucson; Tulsa, Oklahoma; and Virginia Beach. See Braga et al., 2000. 23. After a number of analyses, researchers decided on the following model: Monthly count of youth homicide = intercept + trend + trend squared + month dummy variables + intervention + autoregressive (1) component + error. Trend controls for simple linear trends within each time series; trend squared controls for nonlinear trends within each time series; month dummy variables control for monthly seasonal effects within each time series; intervention estimates the effect of the intervention within each time series; and autoregressive (1) component estimates AR(1) serial lag-one correlation components for each time series. The SAS GENMOD procedure does not allow for the estimation of an autoregressive component in generalized linear models. However, the SAS GLIMMIX macro allows autoregressive components to be estimated in generalized linear mixed models. Mixed models are generally used by statisticians to estimate random effects in statistical models. However, they can also be used to estimate a variance component that is different from that assumed by generalized linear models. In their fixed-effects model, the GLIMMIX macro simply allowed the study team to estimate a variance component that includes an AR(1) coefficient in a generalized linear Poisson regression model. See Ramon C. Littell, George A. Milliken, Walter W. Stroup, and Russell Wolfinger, 1996, SAS System for Mixed Models, Cary, NC: SAS Institute, Inc. 24. The SHR data reported the following yearly counts for youth homicides in Virginia Beach: six youth homicides in 1995, seven youth homicides in 1996, and six youth homicides in 1997. The significant break in June 1996 was due to a 4-month period without youth homicides followed by a 6-month period that saw one youth homicide in each month. 25. The intervention point only could vary from month 12 to month 72, rather than the full time period of month 1 to month 84. A 12-month window was used to assure enough points to identify trends and autocorrelation in the time series. See Braga et al., 2001. 26. See, e.g., Veronica Coleman, Walter C. Holton, Kristine Olson, Stephen Robinson, and Judith Stewart, 1999, "Using Knowledge and Teamwork to Reduce Crime," NIJ Journal, October: 16-23; and David M. Kennedy and Anthony A. Braga, 1998, "Homicide in Minneapolis: Research for Problem Solving," Homicide Studies 2(3): 263-90. 27. See, e.g., Philip J. Cook, 1980, "Research in Criminal Deterrence: Laying the Groundwork for the Second Decade," in Crime and Justice: An Annual Review of Research, Vol. 2, edited by Norval Morris and Michael Tonry, Chicago: University of Chicago Press; Alfred Blumstein, Jacqueline Cohen, and Daniel Nagin (eds.), 1978, Deterrence and Incapacitation: Estimating the Effects of Criminal Sanctions on Crime Rates, Washington, DC: National Academy of Sciences; and Raymond Paternoster, 1987, "The Deterrent Effect of the Perceived Certainty and Severity of Punishment: A Review of the Evidence and Issues," Justice Quarterly 4: 173-217. 28. Malcolm Klein, 1993, "Attempting Gang Control by Suppression: The Misuse of Deterrence Principles," Studies on Crime and Crime Prevention 2:88-111.