Advancing Interorganizational Crime and Violence Reduction Goals Through a Relational Change Intervention

Abstract

Effective interorganizational collaboration is deemed essential to comprehensive crime and violence initiatives but rarely is it empirically assessed. An 18-month intervention to improve collaboration Comprehensive Gang Model (CGM) locations was used in this study to examine the impact on increasing community capacity to address gangs and violence and reducing gang and violence in the community. Relational coordination theory grounded the collaboration intervention. Results from the quasi-experimental design showed significant, increased collaboration and reduction in violent crime in one intervention site. Other crime reduction efforts were explored as counterfactuals. Matched comparison sites saw no change in the ability to work together or violence reduction. Study implications are that relational coordination interventions may facilitate the goal of working better together, but parallel evaluations for each of the five core CGM strategies are needed to understand the independent effects of each strategy on gang and violence reduction goals.

Keywords

criminal justice policy crime reduction Comprehensive Gang Model interorganizational collaboration

Comprehensive crime and violence reduction approaches bring diverse entities together, such as criminal justice agencies, social service providers, government units, school personnel, and community stakeholders, to tackle problems of common interest. Scholars have noted that comprehensive criminal justice initiatives are complex, difficult to evaluate, and when evaluated, success is not often demonstrated (Klein & Maxson, 2006; Papachristos, 2011). Yet governments as well as public and private funders are increasingly moving away from single-agency fixes and programs and shifting toward collaborative approaches to tackle complex social issues (Skogan, 2011). Compounding funding realities is research showing that in communities most affected by violence, residents are unlikely to support crime and violence reduction initiatives unless problems are addressed holistically (Chaskin, 1999).

While comprehensive efforts to reduce crime and violence in communities align with current funding structures and community interests, there must be an understanding of how diverse entities best coordinate and collaborate to reach those shared goals. Comprehensive initiative policy graphs, logic models, and evaluations of individual components can help clarify processes and the mechanisms through which goals are met (e.g., Gravel et al., 2013; Sampson et al., 2013), but interorganizational collaboration is a strategy that undergirds all comprehensive initiatives and needs research attention. To date, much of the focus of large-scale crime and violence reduction initiatives has been to evaluate specific components, like individual programs or solitary program strategies (e.g., suppression), rather than to directly examine the interorganizational aspects of collaborations themselves.

Coordinating activities among diverse stakeholders are often mandated through policies and funding streams but doing so is complicated and does not necessarily translate to impactful, sustainable relational connections among entities. Research demonstrates that collaboration in interorganizational collaborations (ICs) is complex and dynamic (Bryson et al., 2015). If collaboration among organizations is successful, the result should be improvement in outputs and outcomes with the assumption that the sum of the collaborative effort is greater than any one individual entity. Insufficient research exists, however, on crime and violence reduction ICs and how to evaluate their collaboration as a contribution toward shared goals. The Comprehensive Gang Model (CGM) is an ideal crime and violence reduction IC through which to examine the collaborative impact. The CGM is a multisectoral initiative that seeks to reduce gangs and violence through five coordinated strategies. Organizational change and development is explicitly identified as a core strategy for success, and IC is a critical component of that strategy. This study pointedly evaluates the outcomes of a quasi-experimental organizational change intervention focused on collaborative improvement in a CGM initiative. We examine the effect of the intervention on increasing community capacity for the gang, crime, and violence reduction as well as whether increased capacity has any measurable effect on crime reduction.

ICs in Criminal Justice

Criminal justice ICs have been in existence since the widely known Chicago Area Project and similar initiatives in Boston and New York (Marris & Rein, 1967; Miller, 1962; Shaw & McKay, 1942) addressing a range of issues such as community violence, drug abuse, and child maltreatment. Crime and violence reduction ICs have proliferated in recent decades with the belief that these issues are too complex for any one agency to solve alone (McGarrell et al., 2013; Rosenbaum, 2002; Sparge & Curry, 1993). Recent ICs that comprehensively address crime and violence are promoted through various U.S. federal policy initiatives including the CGM, and its variants such as Gang Reduction Programs as well as through Project Safe Neighborhoods, Group Violence Interventions, Weed & Seed, and Strategic Approaches to Community Safety Initiative. In these initiatives, collaboration between criminal justice organizations and other entities, including community organizations, are key elements of a comprehensive approach. While crime and violence reduction impacts of these initiatives have been examined through rigorous evaluations and systematic reviews (e.g., Braga et al., 2018; McGarrell et al., 2009; Roehl et al., 2008), their collaborative dimensions have not been assessed directly. A major obstacle to evaluating collaborative aspects of these ICs is the fact that there is a high degree of difficulty in linking collaborative activities to outcomes (Roehl et al., 2008; Rosenbaum & Schuck, 2012; Skogan, 2011). Unfortunately, lack of attention to the collaborative has inhibited our full understanding of how collaborations impact initiative goals.

Research in the criminal justice field has shown that communication and coordination across organizations are fundamental to successful collaborative functioning and to meeting shared goals (Rengifo et al., 2017). Effective coordination and communication in ICs go beyond the mashing together of individual organizational resources and programs. It requires interorganizational communication about goals and methods to leverage collective resources so that each organization can employ its own complementary skill sets and resources to the shared problem to produce positive change (Rosenbaum, 2002). Research in the organizational science field has demonstrated that communication and coordination are linked, with effective coordination resulting from communication practices that focus on collaboration inclusivity, cooperation, and shared goal identification (Koschmann et al., 2012). Meanwhile, criminal justice research has demonstrated that communication and coordination are enhanced when organizations have a history of collaboration, have dedicated resources to collectively address the problem, and when relevant organizational leaders and policymakers support the collaboration (Braga & Winship, 2006; Fox et al., 2015; McGarrell et al., 2013; Skogan et al., 2009).

Evaluating communication and coordination as measures of collaborative functioning is a difficult task in any large-scale initiative (Daley, 2009), including those that address crime and violence (Gebo et al., 2010). Missing from the literature on collective efforts toward crime and violence reduction is focused research and evaluation of what works regarding ICs as well as the impact of the collaboration itself on shared goals (McGarrell et al., 2013; Rosenbaum & Schuck, 2012). This study begins to fill that gap by evaluating crime and violence reduction goals from a quasi-experiment that utilized relational coordination (RC), an organizational science promising practice, to improve collaborative functioning.

The CGM

The CGM is ideally suited to examine the collaborative function in crime and justice reduction ICs because a key ingredient of the organizational change and development component of the model requires multisectoral IC. The CGM is not a single program but a dynamic program structure through which communities can organize their gang and youth violence reduction approaches consisting of five strategies that, when combined, are expected to reduce gangs and violence. Those strategies are organizational change and development, community mobilization, social intervention, opportunities provision, and suppression.

Organizational change and development is at the heart of CGM IC requiring agencies, such as police, schools, and community organizations, to alter their policies and practices to work together more effectively to reduce gangs and violence. This means sharing information, creating or altering roles and functions, and allocating resources to do so. Community mobilization refers to building support and sustainability through community involvement as equal initiative partners, engaging in dialogue and decision-making. Notably, because this requires organizations to change to allow community members into the fold, community mobilization is a by-product of true organizational change (Gebo et al., 2010). Social intervention refers to prevention activities that aim to reduce the chances of youth becoming involved in gangs, such as mentoring programs and education about gangs and peer pressure. Opportunities provision refers to services provided to gang-involved individuals and those most at-risk of gang membership. This includes education and training programs directed at gang members and those with criminal records as well as street outreach by former gang members to those currently involved. Finally, suppression refers to arrest and prosecution, targeting high violence locations as well as gang members and others who are committing the most violence.

The model was developed and piloted in the Little Village area of Chicago by Irvin Spergel. Initial results showed effectiveness at reducing some gang and youth crime (Spergel & Grossman, 1997). Prior to full evaluative results, the model was quickly adopted by the U.S. Office of Juvenile Justice and Delinquency Prevention, and five replication sites across the United States were funded. Evaluations of replication sites were not all positive (Spergel et al., 2006). One of the key problems in sites that did not see positive results was the lack of organizational change and development (Spergel et al., 2006). Unfortunately, the organizational change and development strategy was never operationally defined. In their review of these CGM studies, Klein and Maxson (2006) noted that although youth violence was significantly reduced in some sites, gang violence failed to be reduced in most sites. Among other things, they argued that the comprehensiveness of the initiative tried to do too many things (prevention, intervention, and suppression) without coordination and consistent, direct oversight. They further pushed for clear articulation of theoretical propositions and rigorous evaluative reviews of gang interventions prior to any further model implementation. Alternatively, in his review, Howell (2012) identified the successes and challenges in these CGM iterations and showed that, overall, the CGM initiative was more beneficial than not.

The U.S. government has since funded other projects that require the utilization of core CGM strategies in programs, such as Gang Free Schools and Communities, the Gang Reduction Program, and SafeFutures. Some state and local efforts have done the same (Gebo & Bond, 2012). The model is known by several other names including the Spergel Model, the OJJDP Comprehensive Model, and the Comprehensive Gang Model. Although there have been some evaluations of these initiatives (e.g., Cahill & Hayslip, 2010; COSMOS Corporation, 2007), none has been as extensive as the initial evaluations conducted by Spergel and his team, and none have focused on the organizational change and development and concomitant IC aspect of the initiative.

RC Theory and Practice

Examinations of comprehensive criminal justice initiatives reveal the need to understand relational dimensions essential to effective IC (McGarrell et al., 2013). RC is a robust organizational theory and practice with the potential to overcome that knowledge gap through strengthening relationships between individuals working toward collaborative outcomes and evaluating the results of those interventions. RC theory suggests that strong communication and coordination between individuals and entities working together can help facilitate the achievement of desired outcomes. It is based on the premise that coordination is a relational process (Faraj & Sproull, 2000) and that positive outcomes occur through frequent, high-quality communication supported by partner relationships grounded in shared goals, shared knowledge, and mutual respect (Gittell, 2002). RC stresses the need to create organized, cooperative relationships through focused attention on improving communication and coordination across entities in support of collective problem-solving (e.g., Wolff, 2001; Woolcock & Narayan, 2000). Relational and collaborative strength is accomplished by building mutual respect through understanding each other’s roles and responsibilities, refining shared goals, and building shared knowledge. These changes, in turn, increase communication, coordination, and needed organizational changes in pursuit of shared goal achievement. Grounded in network assessment methodologies (Bolton et al., 2021), and designed to capture relationships and interactions between collaborating participants, a validated survey instrument measures seven dimensions of RC (Gittell et al., 2000; Valentine et al., 2015). Those dimensions are frequent, timely, accurate, and problem-solving communication; and trust, mutual respect, and shared goals among partners (Gittell, 2002).

RC prescribes certain evidence-based practices to increase the strength of relationships as indicated by RC dimensions. A relational model of change (RMOC) translates these practices by grouping them into relational, structural, and work process interventions (Gittell, 2016). Relational interventions support improved understanding through creating mutually respectful opportunities for dialogue as well as partner coaching and reflection. Structural interventions enhance communication and coordination through formal practices and policies integrated into the everyday work of participants. Examples include new or revised job descriptions, regular collaboration meetings, shared protocols, and shared information systems. Finally, work process interventions support relational and structural changes and include continuous reflection of the current collaborative practice as well as identification of necessary steps to move to the desired collaborative practice. Taken together, these practices are considered a relational model of change aimed at improving functioning to meet goals.

A systematic review of RC research demonstrates that when RC is high, goals are achieved (Bolton et al., 2021). For example, high RC has led to more efficiency and productivity among collaborative partners and a higher likelihood of adopting more effective collaborative practices in healthcare contexts (Cramm & Nieboer, 2012; Gittell, 2002; Hartgerink et al., 2012; Lundstrøm et al., 2014; Noel et al., 2013). High RC also was associated with more parent involvement in an education context (Douglass & Gittell, 2012). These results—efficient use of resources, increased client engagement, and adoption of effective practices—are also key outcomes in criminal justice settings. RC has been used in two criminal justice collaborations. While outcomes were not assessed in either study, results reveal that RC may fit well in interorganizational criminal justice contexts. In one study, Bond and Gittell (2010) uncovered weak communication and coordination among key reentry agencies demonstrating the need to create better linkages across those organizations to ensure individual reintegration success. Hajjar and colleagues (2020) also used RC to assess interorganizational relationships to guide youth safety coalition members in strengthening their collaboration in support of collective goals. Given the positive and promising evidence of using RC across varied contexts, including criminal justice, it is fitting to apply RC to the CGM.

The current research builds on a previous study examining the implementation of an RC intervention in the CGM context. Results from the first article demonstrated that RC was a promising strategy to increase IC among diverse entities implementing the CGM (Gebo & Bond, 2020). The focus of this current study is on understanding process differences among sites implementing the intervention as well as on evaluating the intervention impact on shared goals. Aligned with CGM goals, two hypotheses assess RC intervention impact.

Hypothesis 1 (H1): Did RC lead to increased community capacity to address gang and youth violence problems? If RC led to increased community capacity, then we explore the possibility of whether the RC intervention was powerful enough to influence violence reduction.

Hypothesis 2 (H2): Did increased community capacity reduce gang and youth violence problems? Our findings will contribute to an understanding of the utility of RC and the effectiveness of the CGM ICs net of other practices and programs with implications for other crime and violence reduction ICs.

Method

A matched quasi-experiment was conducted with four cities implementing the CGM in a Northeast state. This type of design is appropriate and desirable when a randomized controlled trial is not feasible (Shadish et al., 2001). Using a mixed-method approach, researchers analyzed within and between-group differences in cities to understand if bolstering the organizational change and development strategy shown to be problematic in prior CGM evaluations influenced communities’ abilities to address gangs and youth violence as well as if any changes translated to appreciable crime and violence reduction benefits.

Study Site Descriptions

Massachusetts provided a unique opportunity to conduct a natural experiment of a CGM intervention. Since 2006, Massachusetts has awarded more than 75 million dollars in funding to communities to reduce gang and youth violence. As part of the funding terms, communities must articulate how their violence reduction strategies fall into the five strategic areas of the CGM. Consistent with OJJDP guidelines on the implementation of the CGM (National Gang Center, 2010), each community is mandated to have a partner committee consisting of entities providing frontline activities and services and a steering committee consisting of individuals who have influence and decision-making power in the community. Many communities also have a local action research partner who assists with those tasks. The current study capitalizes on this natural experiment.

There were two intervention sites (Intervention City A and Intervention City B) and two comparison sites (Comparison City A and Comparison City B).¹ Generally, sites were well matched relative to population, ethnicity, household income, poverty, and crime (see Table 1). There were three significant differences. Comparison City B had a lower percentage of Caucasian residents than the intervention city but was closely matched otherwise and neighbor each other geographically. Intervention City A had a significantly higher volume of violent crime than its comparison city. Likewise, the preintervention median monthly violent crime arrests committed by gang members was seven in Intervention City A compared with five in Comparison City A. These differences are considered in statistical analyses. The median preintervention monthly gang member violent crime arrests in both Intervention and Comparison Cities B were one. Finally, a review of resource allocation revealed that state funding allotted to each of these cities for gang prevention was proportionally consistent for the 4-year study period (E. Fontaine, personal communication, September 4, 2021).

Table 1.

Demographic Characteristics of Study Cities.^a

Site	Population	Ethnicity (single ethnicity)	Median HH income	Families in poverty %	Violent crime volume^b
Intervention City A	180,000	53% White 12% Black 21% Hispanic	$45,000	17%	1,700
Comparison City A	110,000	58% White 7% Black 17% Hispanic	$50,000	16%	1,200
Intervention City B	90,000	83% White 4% Black 7% Hispanic	$30,000	19%	1,100
Comparison City B	95,000	68% White 6% Black 17% Hispanic	$356,000	20%	1,100

U.S. Census Bureau 2013 rounded estimates.

FBI UCR 2010 rounded actual.

The lead agency for the CGM initiative was the police department for the two comparison cities. The City Manager’s office served as co-lead with the police department in Intervention City A, and the City Human Services Office served as the lead in Intervention City B. Lead agencies managed the administrative aspects of crime and violence prevention and intervention efforts as part of the collaboration. Core CGM partner agencies in each site consisted of the police department, street outreach, schools, social service agencies, employment agencies, and courts, including juvenile probation. In three of the four sites, juvenile corrections representatives also participated. Other entities, such as city government representatives and grassroots organizations, sometimes attended meetings but were not part of the core mix.

Site coordinators employed by the lead agency managed the grant, coordinated partner and steering committee meetings, and acted as boundary spanners to better connect various organizations together in CGM efforts. None of the site coordinators were dedicated full time to their city’s CGM strategies; rather the position was either partially funded or an add-on duty to existing full-time jobs. Site coordinators could make suggestions about funding but did not have the authority to grant funding themselves. Those decisions were typically delegated to steering committees made up of higher level decision makers in the city, such as city officials, state and local government authorities, and other community leaders. Sites used the CGM as a framework for organizing their collective work. The study intervention was introduced as sites continued their respective CGM activities. Site coordinators were already holding approximately monthly meetings to coordinate CGM efforts. Site activities included coordinated action between police and social services for identified youth, bringing awareness of social services to neighborhoods through different outreach and canvassing campaigns, and advocating for more funding to support their prevention and intervention efforts. Grant proposals showed no site providing funding for the organizational change and development strategy of the CGM, the key intervention for this project.

RC Intervention Description

The 18-month intervention focused on the translation of RC principles to practice. The RMOC grounded the intervention. This involved introducing relational, structural, and work process practices to bolster intervention cities’ capacities to collaborate toward reducing gang and youth violence. The intervention began with a 2-day RC workshop with approximately 30 participants from both intervention sites representing workgroups involved in the implementation of the CGM. Workgroups represented police, outreach, city government, corrections, district attorney’s office, employment, social services, and mental health services. Site meetings and coaching calls occurred approximately monthly at each intervention site. In addition, researchers also facilitated meetings approximately monthly with CGM site coordinators and partner subcommittees about collaboration challenges and priorities. Finally, researchers introduced evidence-based and best practices research in support of collaborative change as part of the RC intervention. Figure 1 displays the project’s theory of change and illustrates how the intervention activities functioned in relation to each other.

Figure 1.

Relational coordination—Comprehensive Gang Model theory of change.

The intervention emphasized working together to achieve CGM goals. Through monthly meetings and coaching calls, the intervention addressed the relational aspects of implementation, seeking to build a sense of interdependency through the identification of shared goals, clarity around roles and responsibilities, and interorganizational information sharing. Researchers used humble inquiry to probe answers to questions such as “What barriers exist to facilitate information sharing?” This was complemented by coaching on how to facilitate productive, interorganizational meetings to ensure information sharing and accountability. The intervention focused heavily on constructing an ideal violence prevention system that would connect organizations as part of a work process intervention. CGM site coordinators served as boundary spanners who managed communication and collaboration across diverse sectors and entities.

Researchers introduced a number of structural tools including shared protocols (e.g., memorandums of agreement), information sharing agreements, and shared accountability practices (e.g., transparency of performance metrics) to embed the relational and work process efforts into the initiative. RC intervention tool use, or dosage, was administered based on an action research approach that prioritized need and site desire to work on specific aspects of their organizational change and development approach to violence reduction. Dosage was monitored throughout the study (see Appendix) as a process indicator (Linning & Eck, 2018).

The current study seeks to build on our initial study by furthering an understanding of process differences in pursuit of CGM goals in intervention sites and to assess the impact of the intervention on CGM goals. Summary implementation findings of the initial study showed the RC intervention in City A prioritized structural and relational improvements in support of CGM implementation (Gebo & Bond, 2020). They constructed a system of meetings, information sharing, and accountability discussions that formed the center of a citywide plan focused on broad participation and organized action. The RC intervention in City B was similar in the sense that they focused on relational intervention efforts through meetings and information sharing, but these relational intervention efforts were reliant on one person (the site coordinator) and their efforts to change work processes and craft structures remained informal and superficial. The current study examines how these inputs impact the CGM community capacity-building outcome as well as their influence on crime and violence reduction.

Data Collection and Analytic Strategy

H1 was examined through the administration of a validated RC survey tool (Gittell, 2000). The RC survey quantitatively measured the strength of communication and coordination across four survey rounds (baseline, after 6 months of intervention, after the full 18 months of intervention, and 1 year postintervention). Questions assessed the seven dimensions of RC. Sample questions included, “How frequently do people in each of these groups communicate with you about gang and youth violence?”; “Do people in each of these groups share your goals regarding gang and youth violence?”; and “Do people in each of these groups respect the work you do with regard to gang and youth violence?”

Survey respondents were participants in their city’s CGM initiative. Again, those included police departments, street outreach, schools, social service agencies, employment agencies, and courts, including juvenile probation. The individual most connected to youth and gang violence in their respective organization was asked to respond to the survey because RC focuses on roles, not individuals, as central to sustained coordination and communication. The survey was administered through the web-based software Qualtrics, and researchers ensured accurate contacts prior to each survey round. While response rates varied slightly across survey rounds, they averaged 55.4% in Intervention City A (between 9 and 14 respondents), 52.0% in Comparison City A (between 10 and 22 respondents), 58.9% in Intervention City B (between 14 and 17 respondents), and 57.5% in Comparison City B (between 9 and 12 respondents). These response rates are considered good for web-based surveys (Kaplowitz et al., 2004). No discernable differences regarding gender or role in their city’s CGM initiative were detected in an examination of nonresponders.

To further explore changes in the capacity to work better together, researchers conducted appreciative inquiry interviews with key informants in intervention sites (City A = 6, City B = 7). Appreciative inquiries assess the changes and the potential changes that were or could be made to achieve goals (Coghlan et al., 2003). Researchers selected participants based on their centrality to gang and youth violence prevention and intervention work in intervention cities. Selections were confirmed by site coordinators. Participants included those who worked in city government, social services, street outreach, schools, and police departments. Every individual asked agreed to participate. Every respondent had been to at least one, but mostly all, of their city’s monthly partner violence prevention meetings during the intervention period. Phone interviews were conducted immediately following the conclusion of the 18-month intervention. Interviews lasted up to an hour. Transcribed and handwritten interview notes provided rich qualitative data to help contextualize and expound RC survey results. Observational notes taken by researchers at monthly partner meetings (City A = 12, City B = 11) were added to the qualitative data. Researchers used a hybrid inductive-deductive approach to qualitative data analysis, examining notes for RC dimension themes as well as a grounded theory method to identify themes addressing outcomes of the intervention from the perspective of these key informants (Glaser & Strauss, 1967). Two researchers line coded each interview for dimensions and themes and then came back together to discuss findings. An iterative process was used to refine themes and to identify linkages to the seven dimensions of RC.

To answer H2, monthly crime data from the National Incident-Based Reporting System (NIBRS) of the four cities were used as were gang calls for service, monthly arrests of gang members, whether gang-motivated or not, and shots fired data gathered from police departments in each site. All sites used the state’s gang database criteria to identify gang members. In total, there were 52 months of preintervention crime data (January 2014–February 2016), 18 months of intervention crime data (March 2016–September 2017), and 15 months of postintervention crime data (October 2017–December 2018). Comparisons within and between intervention sites were examined through difference-in-differences regression models. Throughout the intervention, researchers monitored site CGM activities by talking with site coordinators, observing at monthly partner meetings, and reading through site quarterly reports submitted as part of state funding requirements. These steps served as additional guards against threats to internal validity due to CGM strategy changes. No major CGM shifts were identified during the course of the experiment. While state funding was also proportionally consistent for each site over the experimental period, it is possible that there were changes in activities not captured by the data collection process.

Results

H1: Providing RC support will increase community capacity to address gang and youth violence.

RC survey scores were based on a 1 to 5 scale where 1 is considered “weak” and 5 is considered “strong.” Scores of 4 or above were considered very good based on prior RC research (Gittell, 2012).² Bivariate summary results from baseline through 1-year postintervention show that there was a continual, positive improvement in Intervention City A (see Table 2). City B, in contrast, had initial positive results and then returned to baseline measures by 1 year postintervention. Neither of the comparison cities showed a change over time. Positive results for City A are consistent with prior process results that showed a better implementation of RC in City A (Gebo & Bond, 2020).

Table 2.

RC Survey Global Results.

RC index sores	Baseline	6 months into intervention	18 months of intervention	1-year postintervention
Intervention City A	3.1	3.4	3.8****	4.0***
Comparison City A	3.8	3.7	3.6	3.6
Intervention City B	3.4	3.8****	3.7	3.4
Comparison City B	3.4	3.3	3.3	3.4

Note. RC = Relational coordination.

***

p < .01. ****p < .001.

The strength of RC across workgroups was measured by the average RC score for each workgroup based on all other workgroup responses over the four rounds of surveys. Results revealed major differences over time (see Table 3). In both intervention cities, RC was weak to moderate at baseline. Intervention City A’s interorganizational relationships generally improved during the course of the initiative and were sustained, especially with the government, the district attorney’s office, probation, school, and research partners. This was not the case in Intervention City B where RC generally remained stagnant. Comparison sites saw no appreciable differences during the study period. Observational notes from coaching calls and partner meetings in City A early in the intervention identified the need to engage the DA’s office, schools, and city leadership in the initiative. Site coordinators and other partners who had connections to these entities actively reached out to those entities to engage them in the process and invited them to participate in the partner or steering committee meetings. They were generally successful in gaining some participation from them. Observational notes from City B for the course of the intervention revealed that relationships with the government and schools were weak. Other relationships needing improvement mentioned by City B participants during the course of the study included probation, faith-based, health services and corrections. Although ideas were circulated about how to bring in these entities to the initiative, no concrete steps were taken to do so.

Table 3.

Intervention City Relational Coordination Scores Across Workgroups.^a

Workgroup	Baseline		6 months into intervention		18 Months of intervention		1-year postintervention
Intervention City	A	B	A	B	A	B	A	B
Criminal justice
Police	3.5	3.6	4.0	4.2	4.3	3.9	4.4	4.0
District Attorney	3.5	3.8	3.7	4.0	4.2	3.7	4.3	3.6
Probation	3.1	3.2	3.4	3.3	4.0	3.5	4.0	3.3
Parole	2.2	3.2	2.6	3.6	3.0	3.3	2.9	3.3
Corrections	2.4	2.9	2.5	3.5	3.2	3.5	3.4	3.3
Services
Social Services	2.9	3.3	2.9	3.8	3.5	3.7	3.5	3.1
Health Services	2.6	3.0	2.8	3.6	3.5	3.6	3.5	3.4
Government	3.0	3.0	3.6	3.6	4.0	3.4	4.5	3.1
School	3.2	3.1	3.6	3.5	4.0	3.5	4.1	3.1
Faith Community	2.8	2.1	3.3	2.8	3.4	2.8	3.3	2.2
Research Partner	3.2	2.6	3.7	3.6	3.9	3.4	4.5	3.0

Probation and parole are part of the court system in this state. Corrections include sheriff’s department and the Division of Youth Services.

Appreciative inquiry interviews provided context to the RC Survey and explored how the intervention inputs related to outcomes. Appreciative inquiry results showed that City A institutionalized collaborative organizational change elements related to increasing capacity to work together into their city’s youth violence prevention plan. They began to formalize meetings that, as three of the six appreciative inquiry informants stated, helped them become more productive and deliberate, eliminating the redundancy in discussions and addressing race and ethnic disparities in the system, especially regarding gangs and violence. Two key informants noted that this is what allowed them to have open conversations about disproportionate minority contact (DMC). One key informant stated, “I think we’re at a place now within the planning group where we’re like, this is the conversation we need to be having now [about disparities].” Another key informant stated that institutionalizing meetings and communication structures helped gang and youth violence partners “. . .share some information [about youth] and be able to communicate more frequently” across different agencies and programs. They also created an organizational chart to identify how entities could best communicate and coordinate on violence issues and DMC.

These changes were illuminated by another interviewee who stated that the intervention helped them focus

. . .on systems, organizational change, where, you know, it’s not just implement programs, and let’s find new grant money to implement this, implement that. I think there’s been a more consistent messaging around—we need to find system issues that we can deal with and address. But then also making sure that we stay true to the assessment. . .which is the disparity in numbers in terms of the Latino and Black community.

While participants also identified various challenges, including engaging youth, communities, and one reluctant city leader in the process, key informants noted positive shifts in relationships and in how the meetings were conducted, the content of meetings, and the institutionalization of intervention practices into the citywide plan. Interviewees’ comments were consistent with the positive RC results over time showing better interorganizational communication and coordination on gang and youth violence issues.

All appreciative inquiry respondents in City B noted that there were excellent working relationships among program providers and that all gang and youth violence initiatives flowed through one person in the city. As one responded stated, “A lot is driven by X [person], but he can’t be the only person doing the outreach because I think it doesn’t have the impact.” The need for formalizing relationships and structures as a way to communicate and sustain work after someone leaves a position or is out for an extended period was identified by two respondents, noting that the coordinator was out for several months and initiative progress stalled. One respondent raised awareness of the need for improved structure and relational aspects of the collaboration as a result of the initiative stalling. They stated,

And you don’t want to be so formal [about relationships] that you don’t have flexibility, but I think, you know, just thinking about the future and how to ensure certain things happen in certain ways can be very valuable, especially if people do change in organizations.

Creating formal communication structures to help with the gang, crime, and violence reduction efforts was identified by another respondent in a concrete way. This respondent said that since the intervention, city partners had an email chain and regular meetings about gangs and high-risk youth; yet barriers remained in how to work better together. Four of the six identified the need for government leaders to support their work. One of those respondents identified the difficulty of that given the city’s “very dysfunctional civic infrastructure” of elections every two years. This was reinforced by another respondent who noted, “the need for more integration in terms of [city] leadership.” Respondents felt that city leadership support could increase commitment from other groups, such as the community, necessary to optimal CGM implementation, but partners did not actively reach out to city leadership.

Yet another respondent stated that the intervention was helpful for “shaking it up” to “reignite” the work that needed to be done to effectively address gang and youth violence. This respondent discussed the need for better coordination among task forces and issues to eliminate overlaps among them. This was echoed by another respondent who said that there now was a concerted effort to do so, or “considering it [gang and youth violence prevention] more holistically” since the intervention took place so that those most at-risk for gangs and violence were identified and received services. Overall, there was some movement toward building capacity to work better together in Intervention City B, but by the end of the intervention period, those gains were not fully realized. The lack of change in RC scores in Intervention City B reflects these challenges.

H2: Increased community capacity will result in decreased gang and youth violence.

H2 explores the possibility that the intervention focusing on one aspect of the CGM caused gang and youth violence reduction. Given evidence from H1 showing changes beginning to occur in Intervention City A during the first 6 months, intervention and postintervention data points were combined and used in models to examine pre- and during/postintervention. This also helped increase statistical power to detect effects. Intervention and Comparison Cities A were examined in one model and Intervention and Comparison Cities B were assessed in another model because qualitative results from H1 showed different processes were happening in these intervention cities.

The average marginal effect over time was examined through regression count models of NIBRS data. Count models were used to address overdispersion in the dependent variables. The first month of the intervention was coded as 0 with incremental increases of 1 to capture monthly change over time. For all models, except for nonviolent gang arrests in Intervention and Comparison Cities B, the likelihood ratio test for the negative binomial versus Poisson model indicated that the negative binomial model fit significantly better (p < .01). A dummy variable for seasonal effects was included to account for higher crime rates during the summer months. Preintervention slopes were allowed to vary in all models, and the parallel trend assumption held for all models. Only the differences-in-differences (DID) table is displayed for simplicity (Table 4).

Table 4.

Difference-in-Differences Across Cities.

DID Intervention & Comparison Cities A
	β (SE)	Z	95%CI
Violent Crime	−.98 (.57)	−1.62*	[−2.05, −.20]
Nonviolent Crime	−1.36 (1.13)	−1.21	[−3.57, .84]
Shots Fired	.14 (.13)	1.09	[−.11, .39]
Violent Gang Arrests	−.16 (.16)	−0.98	[−.47, .16]
Violent Crime in Intervention & Comparison Cities A
Difference-in-differences	1.38 (.63)	2.17**	[13, 2.62]
DID Intervention & Comparison Cities B
Violent Crime	.65 (.51)	1.27	[−3.57, 1.66]
Nonviolent Crime	−1.07 (1.27)	−0.84	[−3.56, 1.43]
Shots Fired	−0.08 (.11)	−0.73	[−.28, .13]
Violent Gang Arrests	.24 (.12)	2.00*	[.01, .48]

Note. DID = differences-in-differences.

p < .10. **p < .05.

DID estimates the actual effect of treatment by looking at the intervention site before versus after and the comparison site before versus after, measuring the magnitude of differences in those sites. The average marginal effect was examined to determine if the change pre and during/post was significantly different, while DID assessed if those differences were substantial between sites. These analyses account for the fact that crime rates were dropping in all sites during the experimental period. Outcomes were violent crime, nonviolent crime, and shots fired, conducting analyses for full crime data as well as analyses for “youth” aged 12 to 29.

Unfortunately, any intervention effect on reducing gangs or gang violence exclusively could not be determined with available data because gang calls for service were either not recorded or reliably recorded in three of the four police departments. Gang arrests were analyzed, but because there is no clear argument that gang arrests should decrease throughout the study period, it is difficult to contextualize. It could be, for example, that there was an increase in arrests of the most active and serious offenders to reduce crime, or it could be that a decrease in arrest was the result of decreasing gang activity. Researcher discussions with partner committees and police department personnel did not reveal clear indications of how these data should be characterized. Thus, gang arrests should be viewed as an indicator of significant changes in gang and/or police behavior over this time period rather than as an outcome variable.

Crime rates trends were similar in Intervention and Comparison A cities preintervention, and while violent crime decreased in both cities during the intervention, violent crime decreased significantly more in Intervention City A. There were no significant differences between Intervention and Comparison Cities A in nonviolent crime, shots fired, or gang arrests. Because violent crime was significant in Cities A, a test of whether the average marginal effect for violent crime was lower in Intervention City A during and postintervention as compared with preintervention was conducted. Results revealed that this was indeed the case. The monthly violent crime rate in Intervention City A averaged 39 violent crimes per 100,000 residents preintervention, while during and postintervention, the rate dropped to 30 violent crime incidents per 100,000 residents. This is arguably a substantial drop, and more investigation into other plausible causes for this drop must be investigated before a causal claim can be made. A contribution analysis will help to ascertain attribution. Violent crime trends in both Intervention and Comparison Cities B show decreases over time. The DID test showed that there were no significant violent crime, non-violent, and shots fired differences between cities postintervention. While there were significant decreases in the number of arrests for violent gang arrests in Intervention City B during and postintervention, the estimate for the average marginal effect was essentially 0 or no difference.³

Contribution Analysis

A contribution analysis helps to determine if an intervention made an important contribution to the observed results accounting for counterfactuals. Contribution analyses are best used when research studies take place in natural settings that do not allow for researcher manipulation. A 6-step contribution analysis process was used to determine how much of a contribution this intervention had on identified outcomes (Mayne, 2012). The process included identifying a theory of change, articulating how the intervention leads to change, and examining counterfactual information for change. The contribution question for this study was, Did the RC intervention impact the community’s ability to work together, crime rates, shots fired, and/or gang arrests?

Rival factors were partially addressed through a pre–post quasi-experiment research design. Results showed that the RC intervention precipitated changes in City A. To assess a counterfactual that something other than the RC intervention produced these results, researchers conducted an environmental scan, examining other initiatives aimed at gang and youth violence as well as daily news reports in each city during the study period. Examination of cities’ quarterly and yearly grant reports from two state initiatives and federal youth violence funding did not identify organizational change, increased communication and coordination, or better collaboration as purposes or reported outcomes nor did discussions with lead coordinators for those grants.

A statewide evaluation two years into the CGM initiative funding in 2009 showed that an early outcome of funding was an increase in collaboration among stakeholders who had not previously worked together (McDevitt & Wolff, 2012). There was no mention of a focus on organizational change and development or an outcome of better collaboration in support of CGM in later statewide partner reports. Given that the current initiative assessed improved collaboration in support of organizational change as measured by the RC survey eight years after this report, it is likely that survey results are reflective of RC intervention changes in support of H1.

Evidence in support of H2 is murkier. A statewide evaluation report showed cities that received state funding for a person-based violence reduction strategy during our study period had a greater drop in violent crime relative to comparison cities (Campie et al., 2017). The drop in community violence was attributed to individual-level intervention effects of working with youth, including gang members, identified by law enforcement as the most likely to commit violence. The analysis, however, failed to account for the parallel trends assumption, and crime rates were already declining, with regression to the mean and additional factor given that cities were selected based on having the highest crime rates. That said, there is the possibility of the individual-level initiative influenced violent crime rates in Intervention City A, and it is counterfactual to any causal claim for H2.

Discussion

The RC intervention in this study was designed to improve collaborative capacity to address gang and youth violence. That increased capacity was hypothesized to result in a decrease in violence. To date, this is one of the first studies to subject the CGM to this type of inquiry. Pre–post and matched comparison results showed mixed findings. Significant improvement in collaborative capacity and violence reduction was observed in one intervention site but not the other. No improvements were observed in comparison sites. Qualitative analyses revealed that there were relational improvements, work processes, and structural adjustments in City A in support of organizational change and CGM goals. RC improved over time and was sustained one-year postintervention. Taken together, it appears that additive gains in collaborative capacity to address gang and youth violence can be possible with attention to the organizational dimensions of IC, although such positive changes may not be powerful enough to affect a change in crime reduction.

Although gains are small, results must be viewed in context. Because all cities in this study use the CGM framework, there was no ability to control for an absence of treatment. The treatment-as-usual effect was likely to be large given that the state funding mandate of a CGM approach. Furthermore, many studies centered on large population bases often have the statistical power needed to quantifiably argue effects, but in small- and medium-sized cities like those in this study, interventions are full coverage and intervention effects are more difficult to detect. While medium and small cities are likely to have lower base rates of gangs, violence, and crime, excluding them from systematic observation and rigorous evaluation ignores the problems they face. In these cases where statistical comparisons are more difficult, triangulating the results through mixed-method approaches is critical, and here we find evidence of positive effects. Further unpacking other causal mechanisms is a task of research beyond the scope of this study.

Our intervention focused on one strategy: organizational change and development. Specifically, we examined the capacity to work together. Our observations, however, reinforce the need to evaluate all five CGM strategies and their independent and collective effects on gang and youth violence reduction. Under the CGM structural umbrella, there must be programs, policies, and collaboration arrangements that have demonstrated effectiveness in all core strategies although the CGM does not proscribe. Evidence-based and evidence-informed practices and programs should be part of the CGM as well as all collaborative criminal justice initiatives (Rosenbaum & Schuck, 2012; Taxman et al., 2009). Ensuring reliable and valid gang data is also particularly important to initiatives that seek to reduce gangs and gang violence (Bond & Gebo, 2012; Klein & Maxson, 2006; McGarrell et al., 2013). Overall, equal attention must be paid to policy, programs, and the organizational dimensions of ICs as the mechanisms through which strategies are carried out. Ideally, an experimental design that includes best-practice programs and policies as well as RC organizational interventions would be tested with like locations to best understand the overall effects of the CGM on reducing the gang and youth violence.

There are several limitations of this research. Although a true experimental design could not be constructed to eliminate rival factors, a quasi-experiment was introduced in a natural setting and internal validity was relatively strong given that this was the first time cities had deliberately focused on collaboration in support of GCM-related organizational change and development. Unlike a validated risk-need-responsivity instrument for individuals to help guide interventions and dosage (e.g., Dyck et al., 2018), our intervention was a novel approach to the problem of IC with no validated instrument that identifies the dosage level and mix of RC tools that best aligns with location need. Over time, with systematic use and documentation of RC strategies, such tools can be developed to better understand the impact of an RC intervention. In addition, the collection and interpretation of gang data remain problematic and differ across cities. To accurately assess gang-related changes, cities must systematically collect data on gangs, gang members, and gang violence with uniform definitions for each (see Klein & Maxson, 2006). While otherwise matched well, lead agency differed among some sites and that may have affected results beyond what we could assess and control for through our intervention, research design, and contribution analysis. Finally, external validity was limited to the extent that these Massachusetts cities may be unlike other cities elsewhere.

These results are most applicable to the four study locations and the larger statewide CGM initiative, but they have lessons for other CGM initiatives and the broader study of crime and violence reduction ICs that include diverse, multisectoral partners. Study implications point to the need to reconsider how comprehensive initiatives are conceptualized and implemented to achieve goals. Research and evaluation on collaboration and organizational change in crime and violence reduction ICs are complicated by the fact that they are intersectoral concepts not readily suited to isolating effects. Yet evaluating the effects of collaboratives themselves is a challenge that should be undertaken to better understand the net collaboration effect on shared goals and the ways in which collaboration can support goal attainment. How organizations collaborate and the significance of relational roles should not change relative to the focus of an initiative, gang or not. As interest in organizational dimensions of criminal justice continues to grow, scholars and practitioners can learn from other fields in which theory and methods for intervening and assessing organizational dimensions are better developed. The black box of these interorganizational collaborative initiatives needs comprehensive scrutiny. Finally, given the likelihood of similar concurrent initiatives in like locations, assessing counterfactuals, such as through contribution analyses, is essential unless a true random assignment can be achieved.

The key aspects of the study—utilization of RC as a method for bolstering collaborative organizational change and development in a dynamic environment—have implications far beyond these local contexts, especially as collaborative ventures struggle with inter- and intraorganizational problems (Daley, 2009). Importantly, addressing collaborative aspects of organizational change and development is compatible with the modern criminal justice focus on evidence-based interventions (e.g., Braga et al., 2018; Roehl et al., 2008). At this point, efforts are largely piecemeal when we consider the enormity and complexity of large-scale interorganizational initiatives. Results from studies on organizational practices could help transport the findings of how to better engage in organizational change to other criminal justice applications that utilize evidence-based and informed programs and practices. How those may translate to crime and violence reduction goals also must be assessed.

This study also speaks to other evidence beyond the programmatic “what works” paradigm (Clear, 2010). Working within and across organizations toward shared goals is a critical area of study that has been neglected in the quest to understand what works to reduce crime, violence, and gangs. We need to expand our knowledge on how organizational change intersects with evidence-based practices and programs toward the achievement of desired outcomes, particularly when validated tools and evidence in disciplines outside the traditional criminal justice field can inform that approach. In the current environment of moving away from creating agencies and programs to other ways of organizing and collaborating, we must take seriously how diverse entities working toward reducing gangs, crime, and violence can work well together.

Footnotes

Appendix

Data Availability

Data for this project are available at

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was supported by Award No. 2015-R2-CX-0013 awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. The opinions, findings, and conclusions or recommendations expressed in this presentation are those of the authors and do not necessarily reflect those of the Department of Justice.

ORCID iD

Erika Gebo

Notes

Author Biographies

Erika Gebo, PhD, is a Professor in the Department of Sociology and Criminal Justice at Suffolk University in Boston, MA. Her areas of research are in youth and gang violence, interorganizational criminal justice collaborations, and policy and program evaluation. Her interdisciplinary work has been published in the fields criminology, sociology, political science, public health, and public administration.

Brenda J. Bond is Professor of Public Administration at Suffolk University. Dr. Bond’s areas of expertise are organizational change and improvement, and collaborative approaches to community safety. She studies the structures, processes, and mechanisms that support the adoption and implementation of innovation, evidence-based and promising practices.

References

Bolton

Logan

C. K.

Gittell

J. H.

(2021). Revisiting relational coordination: A systematic review. Journal of Applied Behavioral Science, 57(3), 290–322. https://doi.org/10.1177/0021886321991597

Bond

B. J.

Gebo

. (2012). Learning from community responses to gangs. In Gebo

Bond

B. J.

(Eds.), Looking beyond suppression: Community strategies to reduce gang violence (pp. 177–186). Lexington.

Bond

B. J.

Gittell

J. H.

(2010). Cross-agency coordination of reentry: Testing collaboration outcomes. Journal of Criminal Justice, 38, 118–129. https://doi.org/10.1016/j.jcrimjus.2010.02.003

Braga

A. A.

Weisburd

Turchan

(2018). Focused deterrence strategies and crime control: An updated systematic review and meta-analysis of the empirical evidence. Criminology & Public Policy, 17, 205–250. https://doi-org.ezproxysuf.flo.org/10.1111/1745-9133.12353

Braga

A. A.

Winship

(2006). Partnership, accountability, and innovation: Clarifying Boston’s experience with pulling levers. In Weisburd

D. L.

Braga

A. A.

(Eds.), Police innovation (pp. 171–187). Cambridge University Press.

Bryson

J. M.

Crosby

B. C.

Stone

M. M.

(2015). Designing and implementing cross-sector collaborations: Needed and challenging. Public Administration Review, 75, 647–663. https://doi.org/10.1111/puar.12432

Cahill

Hayslip

(2010). Findings from the evaluation of OJJDP’s Gang Reduction Program. Office of Juvenile Justice and Delinquency Prevention, U.S. Department of Justice. https://www.ncjrs.gov/pdffiles1/ojjdp/230106.pdf

Campie

Petrosino

Fronius

Read

(2017). Community-based violence prevention study of the Safe and Successful Youth Initiative: An intervention to prevent urban gun violence. American Institute for Research. https://www.air.org/sites/default/files/downloads/report/Intervention-to-Prevent-Urban-Gun-Violence-SSYI-April-2017.pdf

Chaskin

R. J.

(1999). Defining community capacity: A framework and implications from a comprehensive community initiative. The Chapin Hall Center for Children, The University of Chicago. https://www.lisc.org/media/filer_public/27/0f/270f0aa3-a66d-4f57-ad8c-f733584dca64/08102018_resource_defining_community_capacity.pdf

10.

Clear

T. R.

(2010). Policy and evidence: The challenge to the American Society of Criminology—2009 Presidential address to the American Society of Criminology. Criminology, 48, 1–25. https://doi.org/10.1111/j.1745-9125.2010.00178.x

11.

Coghlan

A. T.

Preskill

Catsambas

T. T.

(2003). An overview of appreciative inquiry in evaluation. New Directions for Evaluation, 100, 5–22. https://doi.org/10.1002/ev.96

12.

COSMOS Corporation. (2007). National evaluation of the Gang-free schools initiative [Unpublished draft report submitted to the Office of Juvenile Justice and Delinquency Prevention].

13.

Cramm

J. M.

Nieboer

A. P.

(2012). Relational coordination promotes quality of chronic care delivery in Dutch disease management programs. Health Care Management Review, 37, 301–309. https://doi.org/10.1097/HMR.0b013e3182355ea4

14.

Daley

(2009). Interdisciplinary problems and agency boundaries: Exploring effective cross-agency collaboration. Journal of Public Administration Research and Theory, 19, 477–493. https://doi.org/10.1093/jopart/mun020

15.

Douglass

Gittell

J. H.

(2012). Transforming professionalism: Relational bureaucracy and parent-teacher partnerships in child care settings. Journal of Early Childhood Research, 10, 267–281. https://doi.org/10.1177/1476718X12442067

16.

Dyck

H. L.

Campbell

M. A.

Wershler

J. L.

(2018). Real-world use of the risk-need responsivity model and the level of service/case management inventory with community-supervised offenders. Law and Human Behavior, 42, 258–268.

17.

Faraj

Sproull

(2000). Coordinating expertise in software development teams. Management Science, 46, 1554–1568. https://doi.org/10.1287/mnsc.46.12.1554.12072

18.

Fox

A. M.

Novak

K. J.

Yaghoub

M. B.

(2015). Measuring the impact of Kansas City’s No Violence Alliance. Department of Criminal Justice and Criminology, University of Missouri Kansas City. http://kansascitynova.org/images/presentations/KCNoVAImpactReport-Aug15.pdf

19.

Gebo

Bond

B. J

. (2020). Improving interorganizational collaborations: An application in a violence reduction context. The Social Science Journal. Advanced online publication. https://doi.org/10.1016/j.soscij.2019.09.008

20.

Gebo

Bond

B. J.

(Eds.). (2012). Looking beyond suppression: Community strategies to reduce gang violence. Lexington Books.

21.

Gebo

Boyes-Watson

Pinto-Wilson

(2010). Reconceptualizing organizational change in the Comprehensive Gang Model. Journal of Criminal Justice, 38, 166–173. https://doi.org/10.1016/j.jcrimjus.2010.02.007

22.

Gittell

J. H.

(2000). Organizing work to support relational coordination. International Journal of Human Resource Management, 11, 517–539. https://doi.org/10.1080/095851900339747

23.

Gittell

J. H.

(2002). Relationships between service providers and their impact on customers. Journal of Science Research, 4, 299–311. https://doi.org/10.1177%2F1094670502004004007

24.

Gittell

J. H.

(2012). Relational coordination: Guidelines for theory, measurement, and analysis. Relational Coordination Analytics. https://doi.org/10.12998/wjcc.v2.i12.757

25.

Gittell

J. H.

(2016). Transforming relationships for high performance: The power of relational coordination. Stanford University Press.

26.

Gittell

J. H.

Fairfield

Bierbaum

Head

Jackson

Kelly

Laskin

Lipson

Siliski

Thornhill

Zuckerman

(2000). Impact of relational coordination on quality of care, post-operative pain, functioning and length of stay. Medical Care, 38, 807–819. https://doi.org/10.1097/00005650-200008000-00005

27.

Glaser

B. G.

Strauss

A. L.

(1967). The discovery of grounded theory: Strategies for qualitative research. Aldine Publishing.

28.

Gravel

Bouchard

Descormiers

Wong

J. S.

Morselli

(2013). Keeping promises: A systematic review and a new classification of gang control strategies. Journal of Criminal Justice, 41, 228–242. https://doi.org/10.1016/j.jcrimjus.2013.05.005

29.

Hajjar

Cook

B. S.

Domlyn

Ray

K. A.

Laird

Wandersman

(2020). Readiness and relationships are crucial for coalitions and collaboratives: Concepts and evaluation tools. New Directions for Evaluation, 165, 103–122. https://doi-org.ezproxysuf.flo.org/10.1002/ev.20399

30.

Hartgerink

J. M.

Cramm

J. M.

Bakker

J. E. M.

van Eijsden

R. A. M.

Mackenbach

J. P.

Nieboer

A. P.

(2012). The importance of relational coordination for integrated care delivery to older patients in the hospital. Journal of Nursing Management, 22, 248–256. https://doi.org/10.1111/j.1365-2834.2012.01481.x

31.

Howell

J. C

. (2012). Gangs in America’s communities. Sage.

32.

Kaplowitz

M. D.

Hadlock

T. D.

Levine

(2004). A comparison of web and mail survey response rates. Public Opinion Quarterly, 68, 94–101. https://doi.org/10.1093/poq/nfh006

33.

Klein

M. W.

Maxson

C. L.

(2006). Street gang patterns and policies. Oxford University Press.

34.

Koschmann

M. A.

Kuhn

T. R.

Pfarrer

M. D.

(2012). A communicative framework of value in cross-sector partnerships. Academy of Management Review, 37, 332–354. https://dx-doi-org.web.bisu.edu.cn/10.5465/amr.2010.0314

35.

Linning

S. J.

Eck

J. E.

(2018). Weak intervention backfire and criminal hormesis: Why some otherwise effective crime prevention interventions can fail at low doses. The British Journal of Criminology, 58, 309–331. https://doi.org/10.1093/bjc/azx019

36.

Lundstrøm

S. L.

Edwards

Knudsen

T. B.

Larsen

P. V.

Reventlow

Søndergaard

(2014). Relational coordination and organisational social capital association with characteristics of general practice. International Journal of Family Medicine, 2014, Article 618435. https://doi.org/10.1155/2014/618435

37.

Marris

Rein

(1967). Dilemmas of social reform: Poverty and community action in the United States. Atherton Press.

38.

Mayne

(2012). Contribution analysis: Coming of age? Evaluation, 18, 270–280. https://doi.org/10.1177/1356389012451663

39.

McDevitt

Wolff

(2012). Impact of the Shannon CSI: Challenges and observations from a multisite evaluation. In Gebo

Bond

B. J.

(Eds.), Looking beyond suppression: Community strategies to reduce gang violence (pp. 153–176). Lexington Books.

40.

McGarrell

E. F.

Corsaro

Melde

Hipple

N. K.

Bynum

Cobbina

(2013). Attempting to reduce firearm violence through a Comprehensive Anti-Gang Initiative (CAGI): An evaluation of process and impact. Journal of Criminal Justice, 41, 33–43. https://doi.org/10.1016/j.jcrimjus.2012.11.001

41.

McGarrell

E. F.

Hipple

N. K.

Corsaro

Bynum

T. S.

Perez

Zimmermann

C. A.

Garmo

(2009). Project safe neighborhoods—A national program to reduce gun crime: Final project report (Document No. 226686). National Institute of Justice, U.S. Department of Justice. http://www.ncjrs.gov/pdffiles1/nij/grants/226686.pdf

42.

Miller

W. B.

(1962). The impact of a “total-community” delinquency control project. Social Problems, 10, 168–191. https://doi.org/10.2307/799049

43.

National Gang Center. (2010). Best practices to address community gang problems: OJJDP’s Comprehensive Gang Model. Office of Juvenile Justice and Delinquency Prevention. https://www.ncjrs.gov/pdffiles1/ojjdp/231200.pdf

44.

Noel

Lanham

Palmer

Leykhum

Parchman

(2013). The importance of relational coordination and reciprocal learning for chronic illness care in primary care teams. Health Care Management Review, 38, 20–28. https://dx-doi-org-s.web.bisu.edu.cn/10.1097%2FHMR.0b013e3182497262

45.

Papachristos

A. V.

(2011). Too big to fail: The science and politics of violence prevention. Criminology & Public Policy, 10, 1053–1061. https://doi.org/10.1111/j.17459133.2011.00774.x

46.

Rengifo

A. F.

Stemen

Amidon

(2017). When policy comes to town: Discourses and dilemmas of implementation of a statewide reentry policy in Kansas. Criminology, 55, 603–630. https://doi.org/10.1111/1745-9125.12146

47.

Roehl

Rosenbaum

D. P.

Costello

S. K.

Coldren

J. R.

Jr Schuck

A. M.

Kunard

Forde

D. R.

(2008). Paving the way for project safe neighborhoods: SACSI in 10 US cities. https://www.ncjrs.gov/pdffiles1/nij/216298.pdf

48.

Rosenbaum

D. P.

(2002). Evaluating multi-agency anti-crime partnerships: Theory, design, and measurement issues. Crime Prevention Studies, 14, 171–225. https://popcenter.asu.edu/sites/default/files/library/crimeprevention/volume_14/06-Rosenbaum.pdf

49.

Rosenbaum

D. P.

Schuck

(2012). Comprehensive community partnerships for preventing crime. In Farrington

D. P.

Welsh

B. C.

(Eds.), The Oxford handbook of crime prevention (pp. 226–246). Oxford University Press.

50.

Sampson

R. J.

Winship

Knight

(2013). Translating causal claims. Criminology & Public Policy, 12, 587–616. https://doi.org/10.1111/1745-9133.12028

51.

Shadish

W. R.

Cook

T. D.

Campbell

D. T.

(2001). Experimental and quasi-experimental designs for generalized causal inference. Houghton Mifflin Harcourt.

52.

Shaw

C. R.

McKay

H. D.

(1942). Juvenile delinquency and urban areas. University of Chicago Press.

53.

Skogan

W. G.

(2011). Community-based partnerships and crime prevention. Criminology & Public Policy, 10, 987–990. https://doi:10.111/j.1745-9133.2011.00782.x

54.

Skogan

W. G.

Hartnett

S. M.

Bump

DuBois

Hollon

Morris

(2009). Evaluation of Chicago Ceasefire. National Institute of Justice. https://www.ncjrs.gov/pdffiles1/nij/grants/227181.pdf

55.

Spergel

I. A.

Curry

G. D.

(1993). The National Youth Gang Survey: A research and development process. In Goldstein

Huff

C. R.

(Eds.), Gang intervention handbook (pp. 359–400). Research Press.

56.

Spergel

I. A.

Grossman

S. F.

(1997). The Little Village Project: A community approach to the gang problem. Social Work, 42, 456–470. https://doi.org/10.1093/sw/42.5.456

57.

Spergel

I. A.

M. K.

Sosa

R. V.

(2006). The Comprehensive Community-Wide Gang Program Model: Success and failure. In Short

J. F.

Hughes

L. A.

(Eds.), Studying youth gangs (pp. 203–224). Altamira.

58.

Taxman

F. S.

Henderson

C. E.

Belenko

(2009). Organizational context, systems change and adopting treatment delivery systems in the criminal justice system. Drug Dependence, 103, S1–S6. https://doi.org/10.1016/j.drugalcdep.2009.03.003

59.

Valentine

M. A.

Nembhard

I. M.

Edmondson

A. C.

(2015). Measuring teamwork in health care settings: A review of survey instruments. Medical Care, 53, e16–e30. https://www-jstor-org-443.web.bisu.edu.cn/stable/26417947

60.

Wolff

(2001). Community coalition building—Contemporary practice and research. American Journal of Community Psychology, 29, 165–172. https://doi.org/10.1023/A:1010314326787

61.

Woolcock

Narayan

(2000). Social capital: Implications for development theory, research and policy. The World Bank Research Observer, 15, 225–249. https://doi.org/10.1093/wbro/15.2.225