1. Introduction
Researchers have consistently found that a small proportion of an area's population accounts for a disproportionately large amount of urban gun violence.Reference Braga1 This finding, coupled with the ineffectiveness of traditional law enforcement approaches, drove the development of a range of policies aiming to reduce gun violence through the targeting of high-risk individuals and groups. Group Violence Intervention (GVI) is a popular strategy that shares this objective.
GVI combines a focused deterrence law enforcement approach with community mobilization and social services. Its main goals include (1) identifying individuals associated with street groups at high risk of gun violence, (2) preventing them from perpetrating or becoming victims of gun violence, and (3) directing them towards community support services.Reference Braga2
GVI strategies garnered national and international attention after Boston's Operation Ceasefire (1996) demonstrated great success in reducing gun violence. The effectiveness of the GVI model has been supported by empirical research.Reference Engel, Kennedy, Tilley and Sidebottom3 Since 2005, the National Network for Safe Communities (NNSC), an outgrowth of Boston's Ceasefire, has used the GVI model to address gun violence, overt drug markets, and other forms of violence in the United States and abroad.4 Yet, few studies have examined GVI implementation from the perspective of practitioners, making it difficult to pinpoint implementation challenges and potential ways of overcoming them.
The current study qualitatively examines Project Longevity, Connecticut's largest GVI initiative, to contribute to the limited literature on implementation of gun violence reduction strategies. Relying on interviews with 24 of Project Longevity law enforcement and non-law enforcement partners, we explore the establishment of interagency collaboration, which was viewed by study participants as the most pressing implementation challenge of Project Longevity.Reference Bass5 Our case study results offer important lessons to practitioners responsible for implementing GVI strategies.
2. Connecticut's Project Longevity
Group Violence Intervention takes seriously the notion that crime is highly concentrated among a specific and small group of offenders. Unlike traditional social intervention efforts, GVI blends a deterrence message focused on certain, swift, and severe punishment with social services provision (e.g., case management, educational opportunities, employment training and placement, crisis intervention, drug treatment). The deterrence message and opportunities to benefit from services are delivered to a small and specific group of offenders or individuals at high risk of becoming offenders or victims of violence.Reference Braga and Weisburd6
In response to an increase in gun offenses between 2003 and 2011, Connecticut state legislators approved funding for Project Longevity, a multi-city gun violence prevention strategy that closely follows the GVI model. Project Longevity was first launched in New Haven in 2012, and it was subsequently introduced in Bridgeport (2013) and Hartford (2014). Together, these three cities accounted for 71% of Connecticut's gun crime in 2014.7
Project Longevity mobilizes law enforcement, social service providers, and community representatives to deliver an anti-violence message to individuals associated with street groups who are likely to commit crimes together, and to engage in or be victimized by retaliatory violence. To identify these individuals, Project Longevity utilizes a two-part ‘Problem Analysis,’ composed of a ‘Group Audit’ and an ‘Incident Review.’ During the Group Audit, law enforcement practitioners map and share their knowledge of each street group's location, membership, and activities. During Incident Reviews, staff collect information about ‘group member involved’ (GMI) shootings, usually from police expertise, in an attempt to identify victims and perpetrators. Through this two-part information collection and sharing system, Longevity staff select individuals to engage with, either through group meetings, known as ‘Call-Ins,’ or through one-on-one meetings, known as ‘Custom-Notifications.’Reference Sierra-Arévalo, Papachristos, Bichler, Malm and Sierra-Arévalo8
‘Call-Ins,’ which involve bringing street-group members into a room with law enforcement agents, community representatives, and social service providers, typically take place two or three times a year in Connecticut. ‘Call-In’ participants are individuals on probation or parole whose attendance can be mandated by their community supervising officer. ‘Call-Ins’ are structured around a three-part message. First, criminal justice system agents (i.e., police and prosecutors) give a warning that those who choose to continue committing acts of violence will be met with swift and certain legal consequences. These warnings are followed by community representatives (usually formerly incarcerated individuals or individuals impacted by gun violence) who deliver a moral plea against gun violence, describing its impact on their families and communities. Lastly, Project Longevity's social services coordinator offers case management help and referrals to a range of support services that interested participants are encouraged to take advantage of.9
The second mode of engaging individuals at risk of gun violence is the ‘Custom Notification.’ ‘Custom Notifications’ involve one-on-one, in-person meetings that generally take place after a violent incident occurs, or when new knowledge collected by law enforcement intelligence points to a potential risk for violent crime. The notifications involve meeting with the identified person at their home or in their neighborhood, and they are done by a local program manager and/or a senior law enforcement agent.
Both ‘Call-in’ attendees and recipients of ‘Custom Notifications’ are put into contact with the program's social service coordinator. The coordinator supports individuals who either express certain needs or demonstrate the desire to avoid future involvement with gun violence by connecting them to community resources, raising their awareness of existing services, and guiding them toward opportunities that they can pursue on their own. Social service coordinators work collaboratively with parole and probation officers, local non-governmental organizations (NGOs), and social service providers. Available services generally include high school diploma or general education development classes, employment assistance, drug and alcohol abuse treatment, childcare assistance, mental health services, access to basic goods, and entrepreneurship assistance.
2.1 GVI Evaluations
Researchers who examined Project Longevity in New Haven, CT, found that three years after its implementation, GMI incidents were reduced by nearly five incidents per month.Reference Sierra-Arévalo10 A number of other GVI programs across the nation also have empirical support.Reference Braga, Braga, Weisburd and Papachristos11 For example, the Cincinnati Initiative to Reduce Violence (CIRV), which partnered political leaders, law enforcement, researchers, healthcare professionals, street advocates, and community and business officials in an attempt to reduce violence among at-risk gang members, demonstrated a statistically significant 61% reduction.Reference Engel, Tillyer and Corsaro12 Engel and colleagues (2013) found that both group-member involved homicides and violent firearm incidents significantly declined after the Cincinnati program's implementation for both 24- and 42-month post-intervention periods, demonstrating the potential long-term effects of such programming.13 The New Orleans' Group Violence Reduction Strategy, an intervention that sought to identify and change the behaviors of high-risk offenders through homicide incident reviews and gang audits, revealed a statistically significant 32% decrease in gang homicides.Reference Corsaro and Engel14 Similarly, the Indianapolis Violence Reduction Partnership, which involved a multi-agency working team, a research partnership, and problem-solving techniques aiming to lower gun assaults and homicides, experienced a 34% reduction in homicides in Indianapolis compared to six other Midwestern cities.Reference McGarrell15 In 2012, Braga and Weisburd conducted a meta-analysis of focus deterrence evaluations, and found that 10 of 11 noted a significant, medium-sized crime reduction impact.Reference Braga and Weisburd16 This conclusion was reinforced in their updated 2018 review of 24 quasi-experimental evaluations.Reference Braga17
Yet, evaluations, which tend to be exclusively quantitative, present well-known methodological challenges. For instance, GVI interventions are generally not designed as randomized controlled trials, and may overlap with other initiatives and/or changes in the environment that affect violence and crime rates (e.g., employment or poverty), making it difficult to isolate and evaluate the impact of specific policies. Moreover, evaluations depend on the continuous quality of data collection done by law enforcement, program administrators, and social service providers, who often experience data management updates and turnover of key personnel. An updated version of an evaluation of Project Longevity, for instance, would not be presently feasible given changes and inconsistencies in the GMI incident classification used by the New Haven Police Department.18
Additionally, given their focus on intervention outcomes, rather than the process of planning and implementing interventions, quantitative evaluations often leave out the voice of program administrators and do not explore challenges that appear to be critical to GVI implementation. A few progress and impact reports, however, have pointed to obstacles that can arise at various points of GVI implementation. For instance, the National Network for Safe Communities highlighted “irregular meetings and inconsistent participation by agency partners” as an implementation challenge of the Birmingham Violence Reduction Initiative in Birmingham, Alabama.19 In addition, a 2016 report produced by the London's Mayor Office for Policing and Crime identified “differences in interpretation of the core elements of the GVI model between some practioners” as key challenges for Shield, a pilot GVI program in the United Kingdom.Reference Davies20 Further, a study of Kansas City's No Violence Alliance (NoVA) noted various implementation challenges, including poor interagency collaboration and enforcement decisions being made at a variety of organizational levels.Reference Fox21 Reports such as these point to program implementation challenges, a topic we explore in this article based on narrative accounts of Project Longevity's administrators.
3. Methodology
Between October of 2019 and January of 2020, we interviewed 24 stakeholders of Project Longevity in New Haven, CT. In light of Project Longevity's success in reducing GMI incidents in New Haven, our goal was to qualitatively explore key practitioners' perceptions regarding Longevity's challenges, shortcomings, and conditions for success. Our semi-structured interview schedule included questions relating to implementation challenges and how they were overcome.22 The current study focuses on findings relating to this topic.
Our study focuses on New Haven given our access to local stakeholders and their overall perception that Longevity has been more strongly consolidated in this city than in Bridgeport or Hartford. Project Longevity's former state-wide coordinator helped us identify key interviewees. Twenty-four of the 28 identified individuals agreed to be interviewed. Participants include current and former police officers, state and federal prosecutors, parole and probation officers, Project Longevity staff, social service providers, and religious leadership.23 A third-party company transcribed audio recordings verbatim, and we coded the transcripts using MaxQDA, a qualitative data analysis software. Using a ‘flexible coding’ approach, we indexed transcript passages with codes derived from the interview schedule, and created additional subcodes as other themes emerged.Reference Deterding and Waters24
4. Results
Our interview participants were asked what the most pressing challenges were in implementing Project Longevity, as well as how they addressed them.25 The most frequently mentioned challenge was ‘poor interagency collaboration,’ followed by ‘cultural resistance/lack of buy-in,’ particularly by the police department. Notably, both challenges were spontaneously brought up by approximately 59% of participants without specific prompts. Because of word constraints, we focus in this article on interagency collaboration.26
For the purposes of our discussion, we broadly define interagency collaboration as coordinated efforts of various entities, possibly across different levels of government, sharing risks and responsibilities in the pursuit of shared objectives.Reference Himmelman27 The presumption carried by this definition is that organizations are more likely to efficiently achieve their goals by working together rather than separately through the reduction of duplicative efforts, more integrated services, and improved communication.Reference Fisher28
In discussing interagency collaboration, interviewees referred to two dynamics: (1) collaboration within and between law enforcement agencies (police, prosecutors, probation and parole) and (2) collaboration between law enforcement and non-law enforcement partners (Project Longevity staff, social service providers, and religious leadership).
4.1 Collaboration between Law Enforcement Partners
Before the program began, law enforcement agencies, and officers within these agencies, had largely operated in silos. Our interview participants spoke about Project Longevity's initial challenges relating to collaboration within and between law enforcement agencies. One high-ranking police officer commented on communication difficulties associated with this during the launching of Project Longevity:
[Police departments] … don't capitalize on the information that everybody, even within the department, has. Let alone other jurisdictions and other law enforcement agencies. So, like most departments, we operated in these silos, so to speak…. We shared only the information that we felt was necessary to share. But we pretty much kept our investigations private and to ourselves. We weren't communicating with each other the way that we do now…. We certainly weren't communicating with our partners. We always had a relationship with the federal authorities and with ATF, DEA. … but we really didn't have the type of relationship that we do now (103019).29
Another high-ranking police officer spoke about having positive, personal relationships with colleagues from other agencies, but noted that such relationships did not involve agency-wide open channels of communication that guaranteed “all the agencies at one table” (110519).
When asked about how initial challenges were overcome, interviewees emphasized the role of leadership in establishing top-down directives. A number of police officers described initial resistance within the department, especially by officers who understood the initiative as “another hug-a-thug program” (110519), distinct from what they typically consider to be the proper crime-control role of law enforcement.
Interviewees agreed that Project Longevity gained credibility within and across law enforcement agencies as shootings decreased and distinct agencies were able to establish trust and collective accountability among themselves. Participants noted that daily intelligence meetings, which were initially met with skepticism, ultimately became a pathway to interagency collaboration among New Haven partners. As described by a prosecutor:
[T]he challenge was just bringing everybody together, and sort of explaining the program's strategy and its philosophy. … [W]e started meeting maybe once a month with some of our federal counter-agencies represented, state probation, state parole, and NHPD… and we saw that those monthly meetings were valuable in just developing relationships, sharing information, putting a face to a name. … We started, then, doing them two times a week. And then, … [New Haven's Assistant Police Chief] said ‘Why don't we have these intel[ligence] sharing meetings every day?’ And we began to have them every day. … They've been fantastic intel[ligence] sharing and partnership forming meetings (111219).
The meetings also fostered trust and accountability between partner agencies. One probation officer explained:
So those intel[ligence] meetings have been huge … I think that was one of the hardest challenges initially, was trusting one another and learning to work together (090120).
This interviewee continued to explain that it “took time” to fully understand that different actors were “all working towards the same goals.” Relatedly, a prosecutor spoke about the intelligence meetings as an opportunity to create joint accountability among the different agencies (111219).
It is important to note that daily intelligence meetings were not part of the initial operational plan for the implementation of Project Longevity, nor were they part of the National Network for Safe Communities' guidelines. Rather, in New Haven, the meetings were a local response to the recognition of a problem that impacted stakeholders' trust and cooperation.
4.2 Collaboration between Law Enforcement and Non-Law Enforcement Partners
Interviews also revealed how Project Longevity's key players perceived initial frustrations with and tensions between law enforcement and community service providers. The two groups often demonstrated conflicting views about community interests and their agencies' roles and goals.
Though the GVI strategy is formally described as an intervention that provides a continuum of services, law enforcement interviewees tended to emphasize its deterrence effect, while service providers and religious leaders focused on the program's potential to foster behavior change through the provision of support services. Participants explored how these different views posed challenges to Longevity's implementation.
A civilian partner of Project Longevity spoke about the unsurprising character of such tensions, given that interagency collaboration requires overcoming cultural barriers without guidance or a formal mandate binding multiple agencies: “You have all these players. You have this consortium of players, but nothing really, you know, saying, ‘This design is mandated and you have some sort of stake in the game’”(102219).
Along these lines, another non-law enforcement partner commented on how establishing concerted efforts between multiple agencies requires time and trust:
[T]he process of aligning [the State's Attorney Office, the U.S. Attorney's Office, and the community partnership], which are all critically important to actually putting the intervention together, and I think most importantly, sustaining it, requires kind of a long period of both education, just getting everyone to the same place of understanding what we're actually proposing to do here and what their responsibility will be, and a position of sort of mutual trust and really believing that everyone else is committed at the same level (110819).
Skepticism about partners' alignment of views, values, and responsibilities was particularly evident in a comment offered by a non-law enforcement former partner:
I'm not sure everybody was on the same agenda. There were similarities in the kind of wellmeaningness of the program but from various different vantage points. Some people had pieces of the process or journey that others didn't have, which I think eventually created – for me as well – a suspicion and skepticism that I think could've been avoided from an open and full kind of understanding of where we are, where we're going… (111319).
Notably, tensions between non-law enforcement and law enforcement partners led to initial distrust, which was perceived as hindering collaboration. Both groups described these tensions. Law enforcement agents labeled the intervention's ‘deterrence message’ as its most effective mechanism, while non-law enforcement agents highlighted the importance of ‘support services and behavioral change.’ A former prosecutor described how “[i]t's very important that the public know that law enforcement will act, that the message that we give is real and not empty” (010119). In contrast, a non-law enforcement participant expressed a different understanding and “would have preferred to spend more energy around the transforming culture work than just going to your house saying. … [I]f you get arrested again, this is what's going to happen to you” (111319).
These tensions, acknowledged by most interviewees, were at least partially solved through the hiring of a local project manager. Interviewees noted that New Haven's program manager, a former police investigator with close ties to community organizations, brought in technical competence, as well as management skills that allowed for the balance of multiple and sometimes conflicting points of view. Ultimately, the project manager encouraged conflicting partners to focus on the construction of a joint agenda.
As described by a current police officer, the program manager “is like the face of Longevity for us and he basically helps balance us. He's intimately involved in what we're doing and identification [of street group members], but he also is the one that is out there, at all the community meetings” (102919). A retired highranking officer shared a similar perception:
When we hired [program manager] … I think him being a former law enforcement officer in New Haven, understanding the dynamics in New Haven, understanding the dynamics in this department, it made it very easy for us to have a relationship. … [T]hat's when the legitimacy of the program began changing [within the police department] (103019).
Importantly, non-law enforcement partners also recognized the program manager's skills. One such participant described the individual's “incredibl[e] passion about people” and “genuine car[ing] for the individuals that are involved in these gangs” (112519).
New Haven's program manager worked as a ‘boundary spanner,’ strengthening relationships between different organizations and helping Project Longevity gain the trust of community members.Reference Sullian and Skelcher30
5. Recommendations
Our law enforcement and non-law enforcement interviewees emphasized the critical role played by interagency collaboration in the implementation and establishment of Project Longevity. Based on our interviewees' experiences and other reports describing similar challenges, GVI programs might consider institutionalizing this feature.31 There are different ways of going about this. First, local, state, and federal grant providers might incentivize, or mandate, the hiring of a program manager. In Project Longevity, the program manager enhanced interagency collaboration by balancing Project Longevity's deterrence and social service support goals. If the use of such a manager was formally required, official roles and responsibilities could be outlined prior to program implementation, and manager progress could be tracked without sustainability concerns regarding the position being cut or downsized.
An alternative recommendation might involve formalizing interagency collaboration requirements in the legislative text authorizing GVI programs. Authorizing texts for these programs could require, for example, interagency partners to create memorandums of understanding detailing roles, co-responsibilities, and intervention milestones. This approach would institutionalize interagency collaboration during the front end through substantive commitments that spell out the balance between deterrence and social services before program implementation, rather than during the back end through the project manager. Although including a program manager role or formalizing programmatic goals cannot guarantee successful agency collaboration, more careful consideration of this issue as part of the GVI implementation process appears warranted.
Our exploratory study of perspectives of key stakeholders from Connecticut's Project Lonvegevity examines the challenges and successes that come with implementing a GVI strategy. Our results demonstrate that stakeholders perceive interagency collaboration to be crucial for their work. Yet, the ways in which implementing agents enact or formalize interagency collaboraton has not been fully explored in the GVI literature.
6. Conclusion
Our exploratory study of perspectives of key stakeholders from Connecticut's Project Lonvegevity examines the challenges and successes that come with implementing a GVI strategy. Our results demonstrate that stakeholders perceive interagency collaboration to be crucial for their work.
Our interviewees described collaboration challenges among different law enforcement agencies, as well as between law enforcement and non-law enforcement partners. Challenges involved law enforcement buy-in and the establishment of open communication channels for the exchange of information about GMI incidents and individuals involved in shootings. This challenge lingered until the creation of new, daily intelligence meetings which established formal communication channels, and fostered information sharing and mutual accountability.
When describing collaboration challenges between law enforcement and non-law enforcement partners, our interviewees noted cultural fragmentation and conflicts of value. This finding is consistent with the extant literature on interagency collaboration, which recognizes that it is not uncommon for organizations holding distinct missions to see the injunction to work together as a threat to their specialism, and interpret it as a negative assumption about the value of their own work.Reference Sedgewick and Hawdon32 In New Haven, this challenge was at least partially addressed by the hiring of a program manager who worked as a ‘boundary spanner’ and was capable of bridging the community-law enforcement divide. In the public management literature, boundary spanners (also known as ‘nurturing reticulists’) are skilled communicators capable of “talking the right language” and working across agency boundaries.Reference Sullian and Skelcher33 In New Haven, the program manager served as a reliable and trustworthy source for law enforcement and non-law enforcement participants alike, and was thus vitally important to strengthening interagency collaboration.
By examining stakeholders' perceptions' of challenges in Project Longevity's implementation, this study helps add detailed nuance and advice that may be useful to GVI researchers and policymakers elsewhere. Our exploratory study was, however, limited by a sample size that did not include community members beyond their representatives in the form of non-law enforcement implementers. Further, it relied on interviewees' retrospective accounts of successes and challenges. Self-reported data are subjective and not to be mistaken for mirrored reflections of people's experiences. Moreover, retrieved memories are known to be subject to modification.Reference Alshenqeeti and Bjork34
While the solutions to interagency collaboration challenges in New Haven evolved organically after trial and error, this might not always be the case. Our study suggests that gun violence prevention programs may benefit from greater attention to strategies, such as the formalization of collaboration. Ways of best formalizing collaboration, as we noted, could potentially include the use of legislative text or memorandums of understanding to strengthen agencies' efforts to work together.
Figure 1 Most Mentioned Implementation Challenges
Table 1. Summary of Interviewee Roles
APPENDIX
Quantitative Issues
When this study was first designed, our goal was to supplement our interview data with an updated quantitative evaluation of Project Longevity's outcomes similar to that conducted by Sierra-Arévalo et al. in 2017.Reference Braga1 Our initial plan, however, proved unfeasible given two data gaps we encountered. First, the New Haven Police Department (NHPD) did not have a system in place to keep accurate records on shootings prior to 2013, making a shooting trend analysis pre- and post-Project Longevity nearly impossible. We attempted to triangulate across internal police databases to create shooting trends, but inconsistencies between shooting logs and information inferred from incident reports made this method unsuccessful. Second, a precise assessment of the impact of Project Longevity would need to assess the decline in shootings classified specifically as Group Member Involved (GMI). But in New Haven, GMI classifications have not remained consistent over time. In conversations with a New Haven Assistant Police Chief who has been involved with Project Longevity since its inception, we learned that the point system currently used to identify individuals as members or associates of street groups was not implemented until 2015-2016.Reference Braga2 Unfortunately, to be used with some degree of confidence, the GMI classification would need to be audited line by line by the Assistant Chief based on his memory of the events, or by police staff, based on non-systematized notes and records. This time-consuming undertaking would exceed the data collection possibilities for this article.
Interview Schedule
What is your current job or profession?
How long have you been involved with this work?
Can you walk me through a typical day in your role?
Are you currently involved with Project Longevity?
IF YES:
What is your current role within Longevity?
IF NO:
Were you involved with Longevity in the past? What was your role then?
Were you involved in the initial implementation of project longevity? When and how did you become involved?
What do you think were the program's most difficult challenges during implementation?
How did you or others go about addressing those challenges?
In what cities [New Haven/Bridgeport/Hartford] have you been involved in Project Longevity?
Could you speak to the program's challenges in the different locations?
Has your role with Longevity changed over time? (How so?)
In your opinion, is Project Longevity a successful program?
IF YES:
What do you think are Longevity's greatest successes? Has your view on this changed over time?
What were some factors that helped Longevity to be successful?
IF NO:
What do you think have been Longevity's greatest failures? Has your view on this changed over time?
What were some factors that contributed to Longevity not being successful?
What are some of the project's most pressing challenges today?
Are these challenges different from challenges Longevity faced earlier on in its implementation?
With the knowledge you have today, is there something you would have done differently with Longevity?
In your opinion, who are the most important players in the Longevity program? Why?
Do you attend the program's call-ins?
If YES: How frequently and what role do you play in the meetings?
Are there resources within the program, to follow up with call-in participants or check whether they have benefited from any of the social programs?
Are there resources, not currently offered, that you believe should be offered to call-in participants?
Are you aware of resources within the program to address mental illness problems among call-in participants?
What type of support, if any, does Longevity receive from the National Network for Safe Communities?
To your knowledge, does Project Longevity interact with other gun violence reduction strategies in New Haven or elsewhere?
If YES: How do they interact with each other?
In your opinion, what is the root cause of the gun violence problem in New Haven? Is it different from the rest of the country?
What does Longevity need in order to be improved and continued?
If answer is “funding”: If funding was not an issue, what other challenges do you see to the program's expansion and improvement?
