The Cure for Crisis

Executive Summary

While gun violence in New York City has declined since its peak in the 1990s, shootings persist and are heavily focused in a small handful of neighborhoods including the South Bronx, Brownsville, East New York and Harlem. Much of the recent rise in gun violence is rooted in the COVID-19 pandemic, with shootings and homicides spiking in 2020. Despite a gradual decline since the height of the pandemic, gun violence “hotspots” —a small geographic area of high crime where the police focus resources— continue to tragically claim hundreds of lives each year.

To tackle the crisis of gun violence, New York City has increasingly embraced community-based and preventative strategies to address gun violence. The Crisis Management System (CMS) is the cornerstone of New York City’s Community Violence Intervention (CVI) efforts. Formalized in New York City in 2012 under the leadership of then-City Council Member Jumaane Williams, CMS, rooted in the Cure Violence model, employs “violence interrupters”—trusted community members with firsthand experience in violence prevention—to mediate conflicts, prevent retaliation, and address the root causes of violence through a public health lens, treating gun violence as a contagious disease that spreads through social dynamics and environmental factors. Today, CMS includes over 20 CVI organizations that address gun violence through de-escalation, conflict mediation, and the provision of social services. CVI programs like Save Our Streets (SOS) in Crown Heights, Man Up! Inc. in Brooklyn, and Street Corner Resources in Harlem have demonstrated measurable success in reducing shootings and fostering community resilience.

Over the past decade, the City has expanded CMS significantly, both in funding and geographic reach. The City budget for CMS has grown from $4.8 million in its first year to nearly $100 million annually, marking a twentyfold increase in funding. Over that period, the program’s footprint has expanded from just a handful of neighborhoods to 29 (out of 77) police precincts, serving nearly every New York City neighborhood with notable rates of gun violence. Despite these investments, shootings persist in these neighborhoods—with reports that just 4% of city blocks account for the majority of shootings as of September 2024, many of which have active CVI programs in place.[1] Understanding how CVI efforts impact these high-risk areas is critical to ensuring resources are being deployed effectively and identifying where additional investment or strategic adjustments are needed.

Key Findings

Building on prior research demonstrating the efficacy of CVI programs in NYC, this report presents the most up-to-date analysis using 2024 data, reflecting neighborhoods where CMS was recently expanded. To understand the historical context for gun violence in New York City, this report includes a spatial and temporal analysis of all 23,298 shooting incidents across New York City over the past 19 years, examining trends across all 77 police precincts and all 41 CVI service areas—a novel, dual-layered approach that goes beyond previous studies. By analyzing both localized CVI service areas and broader precinct-wide trends, this study provides a more comprehensive and nuanced understanding of CMS effectiveness while identifying critical gaps that should be addressed to further optimize outcomes.

This analysis also identified both areas persistently plagued by shootings (those among the most violent locations in the city for most of the last 19 years), as well as the emerging hotspots amid the pandemic-era spike in shootings. The analysis looked at how both the long-term and near-term violent locations intersect with CVI service areas to make recommendations for improvements and expansions to CVI service.

To comprehensively assess the scope, effectiveness, and challenges of CVI programs, this analysis is informed by site visits and interviews with CMS providers operating across Brooklyn, the Bronx, Manhattan, and Queens. Additionally, this report includes an analysis of vendor payment data from the Department of Youth and Community Development’s (DYCD) Office of Neighborhood Safety—where the CMS program is currently housed—to assess the financial and operational challenges facing CMS providers. This analysis examined approximately 1,400 payment requests from 112 contracts for CVI services between Fiscal Years 2016 and 2025, totaling $192 million. Finally, this report examined national best practices in CVI programs from cities such as Chicago, Baltimore, Oakland, Miami-Dade, Baton Rouge, and Pasadena. This report identifies evidence-based strategies that have successfully reduced shootings and improved program effectiveness.

Gun Violence & CMS Program Impact

Even as overall rates of gun violence have ebbed and flowed over the last three decades, shootings have consistently been clustered in specific hotspot areas such as the South Bronx, Brownsville, East New York, and Harlem. Data from 2006 onward—the first-year localized shootings data are available—shows that more than half of all shootings each year occur in fewer than 200 of the City’s 2,325 Census Tracts (i.e., compact areas with populations of approximately 4,000 residents designated by the Census Bureau). Further, fewer than 75 Tracts account for a quarter of shootings annually.

• Based on this report’s analysis, the CMS programs contributed to 1,567 fewer shootings citywide between 2012 and 2024. At the precinct level, this analysis found an average reduction of 7.4 shootings per year attributable to active CVI programs. This is a 21% reduction, from an estimated 35 shootings per precinct per year if CVI programs had not been active, to the actual 28 shootings per precinct per year observed in the typical precinct and year in which CVI was active.
• CVI programs are currently operating in the vast majority of neighborhoods with persistent rates of gun violence, with the exception of Harlem (Precinct 28), Inwood (Precinct 34), West Harlem (Precinct 26), and Washington Heights (Precinct 30), and Longwood, Bronx (Precinct 41) where there are currently no CVI programs in place.
• Despite CVI’s demonstrated impact in reducing shootings, several neighborhoods with CVI programs are also among the areas with persistent concentrations of shootings. These include Brownsville, Northern Harlem, East Harlem, and the South Bronx and Central Bronx, where shootings remain frequent despite CMS presence.

Management & Leadership Deficiencies

Despite being termed a “system,” CMS often functions more like a collection of separate groups with minimal integration. Critics have pointed to a lack of effective coordination and management from City Hall, especially following the transfer of oversight to the DYCD, which lacks specific expertise in violence prevention.[2] This fragmentation has led to inconsistent support for individual programs, undermining the overall effectiveness of CMS.

• The City Lacks a Standardized Approach to Data Collection: While some metrics such as de-escalations and retaliatory shootings are collected and reported by CMS providers, other critical evaluation measures—such as, the number of conflicts successfully mediated over time, participant retention in services, and reductions in recidivism for high-risk individuals—are not consistently monitored or included in CMS-wide reports. The absence of longitudinal data—such as tracking participants over time across CMS to determine the sustained impact of interventions—further limits the ability to measure success beyond immediate violence reductions.
• CVI Organizations Lack Access to Essential City Data: CVI organizations rely on real-time, localized data to effectively identify and respond to emerging violence trends. However, CMS organizations currently do not have direct access to critical citywide data sources, such as real-time New York City Police Department (NYPD) shooting data, crime trends, or public health indicators related to violence. This lack of access limits CVI groups’ ability to anticipate violence, deploy resources effectively, and tailor interventions to high-risk individuals and communities. Without access to this data, CVI groups are forced to operate with incomplete information, relying on community intelligence rather than Citywide data that could enhance intervention accuracy and effectiveness.
• The City Lacks a Comprehensive Evaluation Framework for CMS Programs: CMS lacks a standardized evaluation framework, limiting its ability to refine strategies and allocate resources effectively. While CMS has demonstrated success in reducing gun violence, the absence of regular, data-driven assessments prevents a full understanding of which interventions are most effective. Without systematic assessments, NYC cannot optimize its violence prevention efforts or ensure that resources are directed where they are most needed.

Payment Delays

Structural inefficiencies in the City’s procurement-to-payment pipeline continue to impede timely payments, limiting CMS providers’ ability to operate at full capacity and sustain violence prevention efforts. Interviews for this report with CVI programs revealed that significant payment delays disrupt payroll, threaten service continuity, and undermine the credibility of violence intervention programs at critical moments when rapid response is needed.

• This report’s analysis of transaction timelines reveals the average wait time for CMS service providers to receive payments has increased from 130 days in 2016, to 184 days in 2022, and to 255 days—over eight and a half months—in 2024. This represents a 5% increase in payment processing times since 2016, and a 36% increase since 2022.

National Best Practices

• Real-Time Data Dashboards: Several U.S. cities, including Chicago and Oakland, have integrated real-time data-sharing dashboards into their CVI programs, leading to a 30% reduction in gun violence on average where real-time data is utilized.[3] These dashboards help violence interrupters track incidents, monitor socio-economic factors contributing to violence, and deploy resources more effectively.
• ShotSpotter Data Integration: While the efficacy of law enforcement’s use of ShotSpotter data has been thrown into question,[1] Miami-Dade, Baton Rouge, and Pasadena counties have integrated ShotSpotter data into their CVI efforts. This integration, which involves sharing data with CVI groups that are typically kept confidential within police departments, has led to significant reductions in gun-related crimes (35% in Miami-Dade, 28% in Baton Rouge, and 25% in Pasadena) at a far lower cost to taxpayers than police-only responses.[4] [5]
• Ongoing Training: Baltimore’s Safe Streets program demonstrates that comprehensive, ongoing training for violence interrupters significantly enhances violence reduction efforts, contributing to a 56% reduction in homicides and a 34% decline in non-fatal shootings in targeted neighborhoods.[6]
• Program Evaluation: Chicago, Baltimore, and Oakland have successfully integrated routine evaluations into their CVI programs, to refine strategies, maximize resources, and adapt CVI interventions to the evolving needs of high-risk communities. These evaluations have led directly to data-driven improvements and better resource allocation. Oakland’s National Institute for Criminal Justice Reform (NICJR) uses real-time monitoring to adjust intervention strategies dynamically.

Recommendations

Improve the City’s Management & Coordination of CMS

Despite being called a “system,” CMS functions more as a collection of separate groups with minimal coordination. The transfer of oversight to DYCD, which lacks expertise in violence prevention, has contributed to fragmented management and a lack of strategic oversight.

1. Create a Dedicated CVI Oversight and Reform Team: City Hall should establish a dedicated team to oversee CMS reforms, improve coordination, and ensure streamlined funding and programmatic support. This team should include violence prevention experts, data analysts, and program managers to enhance CMS oversight and effectiveness.
2. Standardize Data Collection & Reporting Across CVI Programs: Implement a uniform framework requiring all CVI organizations to track key performance indicators, including outreach efforts, intervention outcomes, long-term participant engagement, and demographic data. This should include clear written guidelines on what data is collected, how it is recorded, and how it is shared to create a truly integrated violence intervention system.
3. Enhance Technical Assistance, Data Training, and Capacity-Building for CVI Organizations: Expand CVI training beyond the initial 40 hours by drawing inspiration models like Chicago’s Community Violence Intervention Leadership Academy (CVILA) five-month education program for emerging CVI leaders. This extended training should provide CVIs with skills in real-time data analysis, intervention strategies, and resource allocation to improve program effectiveness.
4. Improve Coordination Across CVI Organizations: The NYC Mayor’s Office of Criminal Justice (MOCJ) should facilitate regular monthly coordination meetings among CVI organizations, ensuring information-sharing, collaborative strategy development, and unified messaging across CMS providers.
5. Deploy a Data-Driven Community Engagement Strategy through Monthly “NeighborhoodStat” Meetings: The NeighborhoodStat model has been a successful collaborative framework for addressing public safety and quality-of-life issues in NYCHA developments.[7] NYC should expand this model beyond public housing and apply it to all precincts with active CVI programs, creating a structured forum for collaboration between violence interrupters, community organizations, NYPD precinct commanders, and local residents. These monthly precinct-based meetings should review shooting incidents, CVI intervention metrics, and service provision outcomes, allowing all stakeholders to track progress, evaluate trends, and refine violence prevention strategies. This data-driven community engagement approach will ensure that CVI programs adapt in real-time to emerging violence patterns, while reinforcing the city’s commitment to collaborative public safety solutions.

Implement a Real-Time CVI Dashboard in NYC

NYC lacks a centralized, real-time data-sharing dashboard for coordinating violence intervention efforts. Cities like Chicago and Oakland have successfully implemented such systems to track violence trends, allocate resources, and improve response times. The City should fund the development of a centralized, two-way data-sharing system for CVI organizations, allowing them to access and contribute real-time information that improves intervention effectiveness. A non-governmental CBO should manage and operate the system, to build trust and relationships between non-profit and City actors while ensuring data privacy of CVI program participants.

1. Integrate Real-time Data from City Agencies & CVI Providers: Establish a Unified CVI Dashboard providing CVI organizations access to shooting data, crime trends, public health data (e.g., substance abuse and mental health incidents), socioeconomic data (e.g., unemployment rates and housing instability), and available community resources (e.g., job training programs and support services). This will help CVI organizations better identify high-risk areas, tailor interventions, and connect individuals to critical services. CVI organizations should upload standardized data on outreach efforts, conflict mediations, intervention strategies, engagement with high-risk individuals, and follow-up interactions. This will enable the City to better measure program impact, track long-term violence reduction trends, and allocate resources more effectively. CVI-provided data should be securely managed, de-identified where necessary, and protected to ensure confidentiality while still allowing for meaningful analysis and policy development.
2. Leverage ShotSpotter’s “Data for Good” Program to Strengthen CVI: In February 2025, the NYC Comptroller’s Office declined to register the renewal of the NYPD’s contract with ShotSpotter, citing impermissible procedural flaws under New York City’s procurement laws. Following the NYPD’s correction of those flaws, the Office of the Comptroller registered a 3-year $21,853,900 contract pursuant to the Office’s Charter mandated duties. While the NYPD corrected the procedural procurement flaws, concerns around the number of hours spent by NYPD chasing unconfirmed shootings remain. Related recommendations provided by the Comptroller’s audit team have yet to be implemented.[8] To ensure these data are used effectively, the Comptroller’s Office additionally recommends that the City expand the use of Shotspotter data to support CVI efforts in line with best practices implemented in Miami-Dade, Baton Rouge, and Pasadena, rather than being used solely for policing. Integration of ShotSpotter data into the City’s CMS has the potential to not only improve CVI efficacy but reduce the costs of police-only response strategies that are extremely resource intensive and only confirmed shootings 13% of the time.
3. Establish a Data-Sharing Agreement between the Dashboard Manager, NYPD and CVI Organizations that Protects Data Privacy: Implement a written agreement ensuring a two-way exchange of anonymized, de-identified data between CVIs and NYPD. The CBO managing the Unified CVI Dashboard should have access to real-time gun violence data from NYPD, which it can securely share with CVI organizations to improve intervention efforts. Similarly, CVI organizations should only share anonymized, de-identified, aggregated data with the City, ensuring law enforcement receives high-level insights without access to personally identifiable information. Strict privacy safeguards should be in place to prevent misuse and maintain community trust.

Take a Data-Driven Approach to Expanding Capacity and Efficiency of CVI

To maximize the impact of CVI programs, New York City should adopt a comprehensive data-driven approach by enhancing data collection and analysis skills and conducting regular evaluations to track program effectiveness, refine intervention strategies, and allocate resources efficiently.

1. Conduct Regular Longitudinal Studies of CVI Programs: MOCJ should fund independent evaluations every 2-3 years to assess which CVI strategies yield the greatest reductions in gun violence, highlight areas needing improvement, and inform future resource allocation.
2. Establish a Rapid-Response Improvement Mechanism for CVI Programs: Create a formal, structured process within a designated working group to regularly identify and implement real-time improvements based on insights from monthly coordination meetings, a reimagined NeighborhoodStat, interviews with CVI practitioners, and real-time dashboard data. This process should be managed by City Hall and ensure that policy adjustments, resource shifts, and operational improvements occur quickly and consistently, to accompany changes made based on long-term evaluations.
3. Expand CMS into High-Risk Areas Without Coverage & Enhance Resources in CMS Coverage Areas with Persistent Violence: Based on this analysis, the City should prioritize CMS expansion into Harlem, Inwood, Washington Heights and Longwood, which currently lack CVI programs despite being current hotspots for violence. NYC should also consider increasing funding and staffing in Brownsville, Northern Harlem, East Harlem, and the South and Central Bronx, where shootings remain frequent despite CMS presence. This expansion should occur in coordination with CVI providers to ensure the resources are well-spent.

Address Persistent Funding and Payment Delays for CVI Providers

NYC should eliminate chronic delays in CMS provider payments and shift to an advance-payment model to ensure financial stability.

1. Fulfill the longstanding promise to launch ContractStat: Transparency is an essential step to improved efficiency; the administration should provide an update on its progress to implementation. The City should launch ContractStat, a public tracking system for contract delays, ensuring transparency and accountability in funding timelines.
2. Establish Deadlines for Paying Vendors on Time: Set a formal 90-day maximum reimbursement timeframe for City contractors, including CVI providers, and require interest penalties for delayed payments, as recommended by the Human Services Council.

Gun Violence in New York City

In the early 1990s, New York City saw an unprecedented wave of gun violence, with more than 5,000 shootings annually and homicides peaking at 2,245 in 1990. Over the following three decades, gun violence steadily declined. By 2019, shootings had dropped to 777 and homicides to 319, reflecting a dramatic reduction in gun violence.

However, the COVID-19 pandemic reversed this trend. Shootings nearly doubled to 1,531 in 2020, marking a 97% increase from the previous year, while homicides rose to 468. Despite some recovery in subsequent years, with shootings declining to 1,200 and homicides to 386 in 2024, these figures remain above pre-pandemic levels, indicating a new baseline for gun violence in the city. As of 2024, shootings remain approximately 54% higher than 2019 levels, and homicides are up by approximately 21%.

Notably, 2019’s homicide figure was among the lowest in over 60 years, with only 2017 and 2018 reporting slightly lower numbers at 292 and 295, respectively. Because 2019 was an exceptionally low year for homicides, the subsequent increase in 2020 appears even more dramatic by comparison. Furthermore, while homicides rose in 2020, they remained well below the levels seen in earlier decades. In fact, even with 2020’s uptick from the previous year, when homicides reached 468, the total remained below 500—something that, prior to 2011, had not happened since 1961. While recent years have seen an increase from the record lows of the late 2010s, today’s homicide levels remain a fraction of what New York City experienced in the early 1990s, when homicides regularly exceeded 2,000 per year.

Even as overall rates of gun violence have ebbed and flowed, shootings in New York City have long been geographically concentrated in a small number of neighborhoods. While the total number of shootings has fluctuated over the past three decades, these incidents have consistently been clustered in specific hotspot areas such as the South Bronx, Brownsville, East New York, and Harlem. Data from 2006 onward—the first-year localized shootings data are available—shows that more than half of all shootings each year occur in fewer than 200 of the city’s 2,325 Census Tracts (i.e., compact areas with populations of approximately 4,000 residents designated by the Census Bureau). Further, fewer than 75 Tracts account for a quarter of shootings annually. The locations of concentrated shootings are largely stable over time, with fewer than 15 Tracts newly emerging in this high-violence category in any given year.

A geographic analysis of shooting data illustrates that the darkest-shaded areas on the map—representing the most persistently violent Census Tracts—have remained largely unchanged over the years. Figure 1 below shows the Tracts by the number of years in which the Tract is part of the concentration of Tracts covering half of shootings in each year. The areas shaded darkest on the map have the most persistent violence: nearly every year since 2006, these locations have been among the few most violent Tracts in the city. Notably, some of these Census Tracts, such as those in Brownsville and the South Bronx, have consistently ranked among the top 10% most violent areas year after year. The persistence of these patterns suggests that targeted interventions within these areas have the potential to yield significant reductions in gun violence.

Figure 1: Locations with persistent incidence of violence

Overview of NYC’s Strategies to Address Gun Violence

New York City has addressed gun violence through a combination of law enforcement, prosecutorial strategies, gun control laws, and community-based initiatives. In the 1990s and early 2000s, the NYPD implemented data-driven approaches such as CompStat, focusing on identifying and increasing police presence in gun violence hotspots. Prosecutors also introduced specialized strategies, such as gun courts designed to expedite firearm-related cases, and initiatives aimed at dismantling illegal gun trafficking networks.

In recent years, District Attorneys Alvin Bragg and Eric Gonzalez have implemented targeted initiatives to combat gun violence in New York City, enhancing investigative capabilities, addressing the proliferation of untraceable firearms, and funding community-based programs. Additionally, both DAs have emphasized treating non-fatal shootings with the same level of investigative resources and attention as homicides, recognizing that the intent behind non-fatal shootings is as serious as in homicides.

In 2022, New York City Mayor Eric Adams established the Gun Violence Prevention Task Force, co-chaired by First Deputy Mayor Sheena Wright and Man Up! Inc. Founder A.T. Mitchell. The task force developed “A Blueprint for Community Safety,” outlining more than $485 million in investments to create safer, more resilient communities.[9] The plan expanded law enforcement initiatives, such as increased use of warrant squads to conduct gun-related arrests and expedited prosecution of gun cases, strengthening coordination between the NYPD and District Attorneys, alongside investments in housing support, job creation, and neighborhood revitalization efforts in the six precincts with the highest rates of gun violence.

A pivotal shift in NYC’s violence prevention strategies came in 2010 with the introduction of the Cure Violence model, which complemented existing law enforcements strategies with an evidenced-back community-based approach.

Cure Violence & New York City’s Crisis Management System

Originally developed by Dr. Gary Slutkin in Chicago in 2000, Cure Violence is a well-established public health approach used in the United States and internationally to treat gun violence as a contagious phenomenon that spreads through social dynamics, much like a virus. In this model, reducing harm requires both supporting individuals already affected and implementing preventative measures for those not yet impacted. This model deploys “violence interrupters”—trusted community members with firsthand experience of violence—who work within specific geographic areas to monitor social dynamics, identify brewing conflicts, and intervene by leveraging their credibility with residents to mediate disputes, de-escalate tensions, and discourage retaliation through direct conversation. In NYC, the SOS program in Crown Heights was among the first to adopt this model, laying the groundwork for similar initiatives throughout the City. Today, Cure Violence is one of the many strategies employed by CVI organizations, which refers broadly to organizations which use evidence-based solutions for violence prevention that tackle underlying social determinants of violence, often by providing complementary services, such as job training, mental health services, and educational support.[10] Key components of the Cure Violence model include:

• Data-Driven Interventions: Providers use crime data from law enforcement and community sources to identify hotspots—areas with high incidents of gun violence or retaliation risk. Real-time data, such as police blotters and emergency response logs, help pinpoint locations needing immediate attention.
• Community Intelligence Gathering: Violence interrupters maintain deep connections in the community to gather insight into brewing conflicts, gang tensions, or potential retaliation, enabling them to act proactively.
• Rapid Conflict Mediation: When alerted to a conflict, interrupters deploy quickly to mediate disputes and de-escalate tensions. Structured protocols—such as separating parties, clarifying misunderstandings, and offering alternatives to violence—are used to prevent escalation.
• Proactive Community Presence: Interrupters patrol high-risk areas to maintain a visible presence, deter violence, and engage individuals before disputes arise.
• Resource Connection: By partnering with local organizations, interrupters connect high-risk individuals to services like job training, mental health support, housing assistance, and legal aid, addressing root causes of violence.
• Trust-Building: Interrupters build long-term relationships within the community, fostering cooperation and sustained behavior change.
• Rigorous Training and Monitoring: Interrupters undergo training in conflict resolution, trauma-informed care, and de-escalation techniques. Regular evaluations refine their strategies and help manage the emotional challenges of their work.^[11]

NYC formalized its violence intervention efforts with the launch of the CMS in 2012 under the leadership of then-City Council Member Jumaane Williams. CMS was created to expand and coordinate violence interruption efforts citywide under MOCJ, centralizing CVI programs within a structured framework intended to strategically deploy resources in neighborhoods and amplify the effectiveness of individual programs.[12] Today, CMS includes over 20 CVIs that address gun violence through de-escalation, conflict mediation, and the provision of social services.

Figure 2: Operating areas of CVI providers

City Funding for CMS Providers

Beginning as a package of appropriations from the City Council in Fiscal Year 2013 under the moniker of “Anti-Gun Violence,” CVIs coalesced first under the CMS umbrella, based on the City’s public health framework Cure Violence. While CMS was initially exclusively funded by the City Council, the Council’s contributions to anti-gun violence were eventually dwarfed by the money invested by first Mayor de Blasio during his administration and later by Mayor Adams. The City’s overall spending on CMS increased from $5 million in 2013, to almost $100 million annually to sustain the CVI programs as seen in Figure 3 and 4. These funds sustain programming that includes hospital-based counseling for shooting victims, mediation and conflict resolution skill building in schools, access to legal services and job training.

Figure 3: Anti-Gun Violence Programming – Budgeted Spending

In the fiscal year 2018 and 2020 budgets, CVI programs saw slight declines in City funds, but on the whole, funding for CMS is twenty times greater than it was at the program’s inception.[13] While the majority of this funding was routed through the MOCJ as a centralized coordinating agency, funds supporting these anti-violence initiatives have since been allocated across nine City agencies, including the DYCD, the Department of Health and Mental Hygiene (DOHMH), and Health and Hospitals (H+H).

Figure 4: Council & Mayoral Anti-Gun Violence Spending by Agency FY 2013 – 2025 ($ millions)

Analysis of Gun Violence & CMS Service Areas

To better understand how the City’s current violence hotspots intersect with CMS providers, this report analyzed shootings data from 2020 onward to locate emerging patterns of violence amid and following the sharp increases in shootings during the COVID-19 pandemic. By overlaying concentrations of shootings with CVI service areas and the precincts in which CVI operate, this analysis locates neighborhoods or areas with concentrated shootings that still lack CVI service, and identifies areas where shootings persist despite CVI presence.

Figure 5’s color scales indicate the relative concentration of shootings, with the most intense concentrations shown in orange-red. CVI service areas are shown in darker blue and precincts which contain CVI areas are outlined in lighter blue.

Figure 5: Concentrated shootings and CVI service areas

These data show that there are over 20 CVIs, in 41 unique service areas, in 29 out of 77 precincts. This analysis reveals several areas where gun violence persists at relatively high rates, despite the presence of CVI programs: Precincts 23 & 25 (East Harlem), 32 (Northern Harlem), 40 (South Bronx), 42 (Morrisania), 44 (Concourse, Concourse Village), 46 (Fordham Heights), 48 (Belmont, Tremont), and 73 (Brownsville). Most of these areas have persistently suffered high levels of concentrated violence: several Census Tracts within these neighborhoods have been among the few Tracts which cover half the city’s annual shootings in more than 10 of the past 18 years.

Harlem and Inwood have the most severe concentration of shootings in areas where there are no CVI programs in operation. Additionally, West Harlem, Washington Heights, and Longwood, the Bronx are other notable hotspots with concentrated shootings since 2020, but which lack coverage from CVI programs (Figure 5).

Analysis of CMS Efficacy

To better understand the impact of CVI programs on gun violence in New York City, this report analyzes their efficacy. This analysis estimates the effect of CVI programs in reducing the number of shootings both within their immediate service areas (specific locations in which CVI organizations are contracted to work) and at the precinct level, the standard geographic unit for evaluating CVI programs. Precinct-level analysis accounts for any positive “spillover” effects beyond CVI’s specific service areas, reflecting the operational reality that CVI programs influence violence dynamics beyond their immediate service areas. This also controls for policing activities or policies that vary between precincts, such as differences in deployment strategies, resources, or leadership styles. This report’s interviews with CVI providers revealed that they regularly communicate and coordinate with precinct commanders—captains, deputy inspectors, or inspectors who lead precinct operations—to address emerging violence trends and resolve conflicts. This close collaboration enables CVI organizations to extend their impact across entire precincts, amplifying the effectiveness of their targeted interventions.

This analysis finds that areas where CVI organizations are active have fewer shootings than would be expected in these locations without CVI organizations active, with an average reduction of 7.4 shootings per precinct per year in areas where CVI programs were active (Figure 6). This is a 21% reduction from the estimated 35 shootings per year in the typical precinct where CVI was implemented in a year it was active, had CVI not been present. This statistically significant result underscores the effectiveness of CVI programs in reducing gun violence. In service areas where CVI programs are operating, this research reveals an average reduction of 0.5 shootings per year per site—an additional endorsement of CVI’s efficacy, but one that does not have statistical significance in part because of the small and highly varied number of baseline shootings in these geographically small service areas. Across 41 CVI service areas, this represents a total reduction of 21.5 shootings per year within service areas citywide. This is an 11.6% reduction from the 4.5 shootings predicted in the typical service area in a year when CVI is active, had CVI not been in place.

Based on the model finding that precincts where CVI is active see a reduction of 7.4 shootings per year (with a 95% confidence interval between -2.6 and -12.2), this analysis predicts the number of shootings that would have occurred each year without CVI by multiplying this average effect by the number of precincts with CVI programs active each year. Figure 6 shows the cumulative effect by year, with the actual number of shootings in blue and the modeled number of shootings without CVI in orange (with the 95% confidence interval shaded in lighter orange).

Figure 6: Actual and predicted shootings per year

From 2012, when the first CVI program was funded, through 2024, there were 14,368 total shootings across New York City. This model suggests a total reduction of 1,567 shootings compared to the number that would have occurred without CVI programs active (with a confidence interval from -580 to -2,555). This is an 9.8% estimated reduction in shootings over these years.

Figure 7: Total actual and predicted shootings (2012-2024)

The data show that precincts with active CVI programs experience statistically significant reductions in shootings, demonstrating that these initiatives work. At the same time, gun violence continues to occur in areas where CVI is active, highlighting the need to further examine how and why these programs are effective, where their impact is strongest, and what additional support or adjustments could enhance their impact. Understanding these dynamics is critical to strengthening CVI efforts, filling existing gaps, and ensuring that the communities most affected by gun violence receive the most effective interventions possible.

National Best Practices

A survey of national best practices in CVI programs across major U.S. cities, including Chicago, Baltimore, Oakland, Miami-Dade, Baton Rouge, and Pasadena reveals several evidence-based strategies that have the potential to demonstrably reduce shootings and improve program effectiveness in New York City. Key findings include the use of real-time data dashboards for intervention coordination, integration of ShotSpotter data to enhance rapid response, structured and ongoing training programs for violence interrupters, and data-driven evaluation frameworks to refine strategies and improve long-term outcomes. These approaches have contributed to double-digit percentage reductions in shootings and homicides in multiple jurisdictions, reinforcing the importance of data, training, and evaluation in effective violence prevention efforts.

Real-Time Data Dashboards

Several U.S. cities have successfully integrated real-time data-sharing dashboards into their CVI programs, enhancing their ability to reduce violence through targeted interventions. Research strongly supports the effectiveness of real-time CVI dashboards that integrate data on CVI program activities and engagement alongside robust government data on crime, demographics, and engagement with City services. These dashboards can be expanded to integrate emerging technologies that track violence trends in real time and inform intervention strategies accordingly. Cities that have implemented data-driven CVI dashboards have also contracted with community-based organizations specializing in crime analytics to provide direct support to violence interrupters, ensuring they can effectively utilize complex datasets to refine intervention strategies. A 2020 RAND Corporation study analyzing data-informed CVI programs across multiple U.S. cities found that such programs reduced shootings by an average of 30% in cities where real-time data systems were utilized. Additionally, the Health Alliance for Violence Intervention (HAVI), which advocates for integrating real-time dashboards into CVI programs, found that cities using these data systems respond more quickly to emerging violence trends, enabling violence interrupters and service providers to intervene before conflicts escalate. This approach contributed to a 20-25% reduction in shootings over time. [14]

Chicago Create Real Economic Destiny (CRED), supported by funding from the U.S. Department of Justice’s (DOJ) Bureau of Justice Assistance, offers technical support and helps community organizations use real-time gun violence data to tailor intervention strategies.[15] The Chicago CRED dashboard collects data on shootings, helps CVI programs track its engagement with program participants and intervention outcomes, and provides analytics to assess the effectiveness of CVI strategies in high-risk areas. Chicago CRED provides CVI organizations with comprehensive technical assistance, including regular consultations and training for organizations to strengthen their operational capacity. Expanding on this model, cities could further enhance technical support by contracting with community-based organizations that specialize in analytic crime reduction strategies, equipping CVI groups with advanced tools to monitor real-time violence trends and develop targeted intervention strategies. As a non-governmental community-based organization, Chicago CRED anonymizes and deidentifies the data to rigorously protect the privacy of program participants.  Chicago has also implemented data-sharing initiatives where law enforcement collaborates with community members, combining healthcare and law enforcement data to develop hyperlocal, neighborhood-based programming for communities and individuals most at risk.^[16]

Similarly, Oakland, California, has implemented a real-time CVI dashboard developed by the National Institute for Criminal Justice Reform (NICJR). This dashboard tracks incidents of violence, monitors high-risk individuals, and allows violence interrupters to respond to gunfire incidents in real time. It also enables CVI organizations to monitor socio-economic factors that contribute to violence, such as poverty rates or access to mental health services. By combining real-time data with ongoing technical support, the Oakland model ensures that CVI organizations can adapt their strategies as conditions change, making their interventions more responsive and effective.[17]

Shotspotter Data Integration

ShotSpotter is a gunshot detection system that claims to use acoustic sensors to detect the sound of gunfire throughout neighborhoods. Once gunfire is detected, ShotSpotter aims to identify the location of a gunfire incident and notify law enforcement within minutes. The data is intended to help law enforcement identify the time and place of shootings, although its accuracy and cost-effectiveness have been questioned. In New York City, ShotSpotter is exclusively used by the NYPD as part of its crime-fighting toolkit. In June 2024, the NYC Comptroller’s Office found that the NYPD’s use of the technology resulted in only 13% confirmed shootings, leading to thousands of hours of potentially wasted police time.[18] Furthermore, in February 2025, the Comptroller’s Office declined to register the renewal of the NYPD’s contract with ShotSpotter, citing impermissible procedural flaws under New York City’s procurement laws. Following the NYPD’s correction of those flaws, the Office of the Comptroller registered a 3-year $21,853,900 contract pursuant to the Office’s Charter mandated duties. While the NYPD corrected the procedural procurement flaws, concerns around the number of hours spent by NYPD chasing unconfirmed shootings remain. NYPD records showed that officers spent an average of 20 minutes investigating alerts deemed unfounded and 32 minutes investigating alerts that went unconfirmed. NYPD officers spent approximately 427 hours investigating alerts that did not result in confirmed shootings, equating to nearly 36 twelve-hour shifts if only one officer responded to each alert. Related recommendations provided by the Comptroller’s audit team have yet to be implemented.[19]

In other cities, through its “Data for Good” program, ShotSpotter has shared data with non-law enforcement organizations, including CVI programs. These organizations can then use real-time gunfire alerts to respond rapidly to shootings, mediate conflicts, and deploy social services. Miami-Dade County, Florida, for example, has incorporated ShotSpotter data into its community violence intervention strategies, enabling violence interrupters to determine where gunfire incidents occur most frequently. This integration has reportedly coincided with a 35% reduction in gun-related crimes in targeted areas, as organizations use real-time alerts to provide immediate trauma response, vocational training, and mediation services. Similarly, in Baton Rouge, Louisiana, faith-based organizations and violence intervention groups use ShotSpotter data to engage with communities experiencing high levels of gun violence, and the data integration was linked to a 28% decrease in reported shooting incidents within its first year of implementation.[20] Pasadena, California, has similarly used ShotSpotter’s Data for Good program to allow community organizations to access real-time gunfire data, contributing to a 25% reduction in reported gunfire-related incidents in targeted areas. [21]

Ongoing Training

Cities across the U.S. have developed a comprehensive and ongoing training frameworks to better equip violence interrupters. Baltimore’s Safe Streets program provides a training program, incorporating role-playing activities, scenario-based simulations, and practical exposure to real-world violence intervention efforts. This structured training ensures that violence interrupters are well-prepared to handle high-stress situations and respond effectively to violence hotspots. Beyond the initial training, Safe Streets prioritizes continuous professional development, requiring workers to undergo regular refresher courses, crisis simulations, and trauma-informed care workshops. The program also integrates ongoing mentorship programs where experienced interrupters train and guide new recruits, reinforcing national best practices and real-time crisis response skills. Baltimore’s model demonstrates the importance of continuous training in ensuring that violence interrupters remain adaptable, skilled, and prepared to respond to ever-changing community dynamics.[22]

Program Evaluation

Other cities have successfully integrated regular evaluations and data-driven assessments into their CVI frameworks to ensure continuous improvement and optimal resource allocation. In Chicago, CVI initiatives are regularly evaluated through partnerships with research institutions such as the University of Chicago Crime Lab. These evaluations often occur every two to three years, assessing violence trends, program effectiveness, and areas for improvement. The findings inform funding decisions and allow policymakers to scale successful initiatives while addressing inefficiencies.

Baltimore’s Safe Streets program similarly conducts longitudinal studies to track the long-term effectiveness of its violence intervention programs. A 2015 study by the Johns Hopkins Bloomberg School of Public Health on Baltimore’s Safe Streets program found that neighborhoods served by the program experienced a 56% reduction in homicides and a 34% reduction in non-fatal shootings. The City of Baltimore continues to incorporate similar studies into policy decisions to ensure sustained impact.[23]

Finally, Oakland’s National Institute for Criminal Justice Reform integrates continuous data monitoring and evaluation into its CVI efforts, ensuring that intervention strategies are adjusted in real time based on evolving violence patterns. These studies have led to adaptive interventions and better alignment of funding with community needs. The study attributed much of this success to the rigorous training and professional development that violence interrupters received. [24]

Key Challenges

To better understand the challenges New York City Crisis Management System providers face, this report’s analysis included site visits and interviews with CMS providers operating in Brooklyn, Bronx, Manhattan, and Queens. Additionally, the analysis included vendor payment data from 921 contacts to assess how delays impact service delivery.

This analysis reveals that despite being termed a “system,” the Crisis Management System functions more like a loose network of separate organizations with minimal coordination from City leadership. A major critique of the program is the lack of centralized oversight from City Hall, particularly following the transfer of CMS management to the DYCD, which lacks specialized expertise in violence prevention. This fragmentation has resulted in operational disparities and limited strategic cohesion, ultimately weakening the program’s overall effectiveness.

The City Lacks a Standardized Approach to Data Collection

One of the most pressing consequences of the City’s lack of coordination is the absence of a comprehensive, standardized data collection framework for CMS programs. While CVI organizations track their outreach and engagement efforts, there is no comprehensive citywide standard governing how data is recorded, analyzed, or shared. As a result, the information reported to the City is highly fragmented, making it difficult to aggregate meaningful data and assess the overall impact of violence intervention efforts.

This lack of comprehensive standardization extends to the core metrics used by CVI organizations. Each organization independently tracks interactions and interventions within its target area, but the specific data points collected vary significantly. While some performance metrics—such as the number of de-escalations, retaliatory shootings, and community events—are tracked at a city level, other critical metrics such as the number of conflicts successfully mediated, long-term participant engagement, and reductions in recidivism among high-risk individuals are not consistently measured across all CMS programs. The lack of uniform data collection methods limits the City’s ability to track violence trends, assess real-time intervention impact, and ensure strategic resource allocation.

CVI Organizations Lack Access to Essential City Data

CVI organizations rely on real-time, localized data to effectively identify and respond to emerging violence trends. However, CMS organizations do not have direct access to critical citywide data sources, such as real-time NYPD shooting data, crime trends, or public health indicators related to violence. This lack of access limits CVI groups’ ability to anticipate violence, deploy resources effectively, and tailor interventions to high-risk individuals and communities. Public health data, such as substance abuse trends, mental health crises, and hospital admissions for violence-related injuries, in particular, could allow CVI organizations to connect at-risk individuals to appropriate services and address underlying causes of violence before they escalate. Similarly, socioeconomic data, including unemployment rates and housing instability, could help CMS providers better target their outreach and intervention efforts to address systemic risk factors.

Beyond crime and socioeconomic data, CMS programs would benefit from access to citywide information on community resources, such as job training programs, mental health services, and social services. These programs play a critical role in violence prevention by ensuring that individuals in crisis are connected to the support they need. However, CVI organizations lack citywide real-time access into this data, making it difficult to direct individuals to the most appropriate resources.

Without access to this data, CVI groups are forced to operate with incomplete information, relying solely on community intelligence rather than citywide data that could enhance intervention accuracy and effectiveness. Other cities have implemented formalized data-sharing agreements between CVI programs and City agencies, allowing violence interrupters to work with the most up-to-date information. NYC has yet to take this step, leaving CMS providers at a disadvantage in addressing violence proactively and strategically.

The City Lacks a Comprehensive Evaluation Framework for CMS Programs

While CMS does collect some performance data, there is no standardized evaluation system to measure the effectiveness of violence intervention strategies across the entire system. Existing assessments do not systematically track key long-term indicators, such as sustained participant engagement or reductions in recidivism, making it difficult to determine the lasting impact of CMS efforts.

Without a structured citywide evaluation process, it is difficult to determine which intervention models are most effective, what improvements are needed, and how CMS efforts contribute to long-term violence reduction. The lack of consistent, citywide evaluations means that policymakers must rely on fragmented and anecdotal reports rather than comprehensive, empirical assessments of program effectiveness. The absence of longitudinal studies that track participants over time to determine the sustained impact of interventions, further limits the ability to measure success beyond immediate violence reductions.

This report’s analysis of gun violence trends from 2012 onward highlights the need for more regular, data-driven evaluations to track the impact of CMS interventions over time. While some neighborhoods have seen sustained reductions in shootings, others—particularly those with persistent or resurgent violence—show gaps in program coverage, requiring targeted funding increases and refined strategies. Without a structured approach to evaluation, it is difficult to determine which CMS interventions are most effective and where resources should be concentrated to maximize impact.

Gaps in Training for CVI Programs

CMS provides foundational training for violence interrupters and credible messengers, focusing on essential skills such as conflict mediation, trauma-informed care, and de-escalation techniques. However, the current training model lacks structured ongoing professional development, leaving violence interrupters without the advanced skills and continuous reinforcement needed to adapt to evolving violence dynamics.

While NYC’s training includes an initial 40-hour curriculum, it does not require refresher courses, crisis simulations, or mentorship programs—all of which have been proven effective in strengthening violence interrupters’ ability to handle real-world conflict scenarios. This lack of continuous training limits their ability to respond effectively to high-pressure situations, mediate conflicts, and prevent retaliations in the most at-risk communities.

Delayed Vendor Payments

Persistent bureaucratic inefficiencies hinder CMS providers from operating at full capacity. Like most other nonprofits contracted with the City, violence interrupters in Harlem reported weeks-long delays in receiving funding and other necessary support, limiting their ability to mediate conflicts in real time, disrupting payroll, and threatening service continuity.[25] These delays not only undermine the program’s credibility but also place communities at greater risk during periods of escalating violence.

To better understand this challenge, this report reviewed vendor payment data provided by the Department of Youth and Community Development’s Office of Neighborhood Safety, identifying 24 vendors that have historically or currently performed CVI work. Using the City’s Financial Management System, the analysis found approximately 1,400 vendor payment requests for CVI services between Fiscal Years 2016 and 2025 totaling $192 million (Figure 8). The analysis found that the average wait time for CVI service providers to receive payment has been steadily increasing. In 2016, vendors waited an average of 130 days to receive a fully processed payment; by 2022, that figure had risen to 184 days. In 2024, providers waited an average of 255 days—eight and a half months—for a cleared check (Figure 9). This represents a 96.5% increase in payment processing times since 2016, and a 36% increase since 2022.

Figure 8: Total Amount Paid to CVI Organizations – FY 2016 – 2025

Figure 9: Average Days from Completed Work to Payment: FY 2016 – 2025

Source: Financial Management System

These late payments represent failures at various points in the payment process. The time between when an agency accepts an invoice and money is transmitted has consistently been approximately 60 days for each of the observed years. However, there are numerous conditions that must be met to properly request payment – vendors must submit an approved annual budget and detailed invoices for services rendered, and agency staff must review documents for accuracy. Often, this is an iterative process that results in back-and-forth rejections and resubmissions before all necessary paperwork is in order. On average, this process takes approximately 112 days for CMS groups to navigate.

Submitting invoices for more than 30 days of work also adds to the complexity of processing the payment; for smaller CVI vendors, approximately 20% of invoices transmitted are for 3 or more months of service compared to 10% for fiscal conduits. This underscores the structural challenges that smaller CMS providers face in sustaining their work. Addressing these systemic delays is critical to ensuring that frontline organizations have the resources they need—when they need them—to continue their impactful work.

CVI organizations in New York City face significant procurement challenges that impede their effectiveness. Complex administrative procedures and stringent compliance requirements often lead to delays in funding and resource allocation, hindering these organizations’ ability to respond promptly to conflicts. For instance, the City’s procurement process mandates that vendors submit approved annual budgets and detailed invoices, which agency staff must meticulously review. This iterative process, involving back-and-forth communications to rectify any discrepancies, averages approximately 112 days for CVI groups to navigate, as noted in this report’s analysis.

Additionally, some CVI organizations have faced scrutiny over financial management and operational practices. In 2019, the New York City Department of Investigation (DOI) conducted a probe into Man Up Inc., a Brooklyn-based violence interrupter organization, uncovering instances of financial impropriety. In response to these findings, Man Up entered a five-year Corrective Action Plan (CAP) with the MOCJ in August 2019. The CAP mandated several organizational reforms, including the development of a comprehensive fiscal manual detailing policies for purchasing, expenses, and cash management; the establishment of a revised conflict of interest policy compliant with city and state laws; the creation of a corporate governance plan outlining the responsibilities of the board and executive staff; and the implementation of a computerized accounting system to enhance financial transparency. To assist in these reforms, Man Up hired NCheng LLP as a fiscal agent in early 2020. By November 2022, MOCJ reported that Man Up had successfully fulfilled the CAP’s requirements, leading to improved financial management and organizational structure.[26] The organization’s Founder and Executive Director, Andre T. Mitchell, described the CAP as a “blessing in disguise,” acknowledging that it facilitated significant improvements in the organization’s operations.[27] Despite these past challenges, Man Up has continued its efforts in violence prevention. In June 2022, Mayor Eric Adams appointed Andre T. Mitchell as co-chair of the City’s anti-gun violence task force, recognizing his ongoing commitment to addressing gun violence in New York City.[28]

Recommendations

Improve the City’s Management & Coordination of CMS

Despite being called a “system,” CMS functions more as a collection of separate groups with minimal coordination. The transfer of oversight to DYCD, which lacks expertise in violence prevention, has contributed to fragmented management and a lack of strategic oversight.

1. Create a Dedicated CVI Oversight and Reform Team: City Hall should establish a dedicated team to oversee CMS reforms, improve coordination, and ensure streamlined funding and programmatic support. This team should include violence prevention experts, data analysts, as well as program and contract managers to enhance CMS oversight and effectiveness. Given the persistent procurement and financial management challenges that CVI organizations face, this team should also work to reduce bureaucratic delays in vendor payments, improve fiscal oversight, and provide technical assistance to smaller CVI organizations struggling with procurement compliance. This work could be done in partnership with, or as part of, the administration’s recently established Nonprofit Advisory Council which is expected to implement policy recommendations that address inefficiencies in the City’s contracting process. Strengthening fiscal accountability while expediting funding disbursements will help ensure that CVI organizations can operate at full capacity without unnecessary disruptions.
2. Standardize Data Collection & Reporting Across CVI Programs: City Hall should implement a uniform data collection framework that requires all CVI organizations to track and report key performance indicators, including outreach efforts, intervention outcomes, long-term participant engagement, and demographic data. This should include clear written guidelines on what data is collected, how it is recorded, and how it is shared to create a truly integrated violence intervention system.
3. Enhance Technical Assistance, Data Training, and Capacity-Building for CVI Organizations: NYC should invest in technical support, structured training, and capacity-building to ensure CVI organizations have the skills and resources needed to use real-time data effectively. This includes training in data collection, analysis, and interpretation, equipping violence interrupters with the ability to track violence trends, measure program impact, deploy new technologies to track emerging trends in the community, and adjust intervention strategies accordingly. Community-based organizations with expertise in analytic crime reduction strategies could be contracted out to provide this support to CVI groups directly. By providing in-depth training in data collection, analysis, and interpretation, violence interrupters will be better equipped to monitor violence trends, assess program impact, and refine intervention strategies accordingly. This effort should also extend beyond the initial 40-hour CVILA, which offers a five-month education program for emerging CVI leaders.
4. Improve Coordination Across CVI Organizations: MOCJ should facilitate regular coordination monthly meetings across all participating CVI organizations, ensuring information-sharing, collaborative strategy development, and unified messaging across CMS providers. These meetings should also be used to identify emerging operational challenges among CVI providers to address issues like payment delays and contract uncertainty before they impact CVI providers’ day-to-day work.
5. Deploy a Data-Driven Community Engagement Strategy through Monthly NeighborhoodStat Meetings: The NeighborhoodStat model has been a successful collaborative framework for addressing public safety and quality-of-life issues in NYCHA developments. NYC should expand this model beyond public housing and apply it to all precincts with active CVI programs, creating a structured forum for collaboration between violence interrupters, community organizations, NYPD precinct commanders, and local residents. These monthly precinct-based meetings should review shooting incidents, CVI intervention metrics, and service provision outcomes, allowing all stakeholders to track progress, evaluate trends, and refine violence prevention strategies. This data-driven community engagement approach will ensure that CVI programs adapt in real-time to emerging violence patterns while reinforcing the City’s commitment to collaborative public safety solutions.

Implement a Real-time CVI Dashboard in NYC

NYC lacks a centralized, real-time data-sharing dashboard for coordinating violence intervention efforts. Cities like Chicago and Oakland have successfully implemented such systems to track violence trends, allocate resources, and improve response times. The City should fund the development of a centralized, two-way data-sharing system for CVI organizations, allowing them to access and contribute real-time information that improves intervention effectiveness. A non-governmental Community-Based Organization (CBO) should manage and operate the system, to build trust and relationships between nonprofit and City actors while ensuring data privacy of CVI program participants.

6. Integrate Real-time Data from City Agencies & CVI Providers: The City should provide CVI organizations with real-time access to shooting data, crime trends, public health data (e.g., substance abuse and mental health incidents), socioeconomic data (e.g., unemployment rates and housing instability), and available community resources (e.g., job training programs and support services) through the Unified CVI Dashboard. This information will allow CVI groups to better identify high-risk areas, tailor interventions, and connect individuals to critical services. CVI organizations should use the Unified CVI Dashboard to upload data on outreach efforts, conflict mediations, intervention strategies, engagement with high-risk individuals, and follow-up interactions. This standardized reporting will enable the City to better measure program impact, track long-term violence reduction trends, and allocate resources more effectively. CVI program’s-provided data should be securely managed, de-identified where necessary, and protected to ensure confidentiality while still allowing for meaningful analysis and policy development.
7. Leverage ShotSpotter’s “Data for Good” Program to Strengthen CVI: A June 2024 audit by the NYC Comptroller’s Office found that only 13% of ShotSpotter alerts resulted in confirmed shootings, leading to thousands of hours of potentially wasted police time. Furthermore, in February 2025, the Comptroller’s Office declined to register the renewal of the NYPD’s contract with ShotSpotter, citing impermissible procedural flaws under New York City’s procurement laws. Following the NYPD’s correction of those flaws, the Office of the Comptroller registered 3-year $21,853,900 contract pursuant to the Office’s Charter mandated duties. While the NYPD corrected the procedural procurement flaws, concerns around the efficacy of the technology and the number of hours spent by NYPD chasing unconfirmed shootings remain. Related recommendations provided by the Comptroller’s audit team have yet to be implemented. If the City continues to use Shotspotter for law enforcement purposes, then its data should be deployed to support CVI initiatives rather than remaining solely within law enforcement. Cities like Miami-Dade, Baton Rouge, and Pasadena have demonstrated that integrating ShotSpotter data into violence prevention efforts leads to measurable reductions in shootings. In Miami-Dade, this approach coincided with a 35% drop in gun-related crimes in targeted areas by allowing violence interrupters to immediately deploy trauma response teams, mediation services, and outreach programs in response to shootings.[29] NYC should adopt a similar model to ensure that existing ShotSpotter data is leveraged for violence prevention rather than solely for policing purposes. Integration of ShotSpotter data into CMS has the potential to not only improve CVI efficacy but reduce the costs of police-only response strategies that are extremely resource intensive over time.
8. Establish a Data-Sharing Agreement between the Dashboard Manager, NYPD and CVI Organizations that protects data privacy: A formal, written agreement should be implemented to ensure a two-way exchange of anonymized, de-identified data between NYPD and CVI organizations while maintaining strict privacy safeguards. The CBO managing the Unified CVI Dashboard should have access to real-time gun violence data from NYPD, which it can securely share with CVI organizations to improve intervention efforts. Similarly, CVI organizations should only share anonymized, deidentified, aggregated data with the City, ensuring that law enforcement receives high-level insights without access to personally identifiable information. Chicago CRED provides a strong model for this approach, as it is an independent CBO that manages gun violence data separately from law enforcement to prevent misuse and maintain community trust, ensuring data security while enabling effective violence prevention. The City should adopt a similar framework to maximize the effectiveness of CVI efforts while maintaining the confidentiality of individuals involved in violence intervention programs.

Take a Data-Driven Approach to Expanding Capacity and Efficiency of CVI

To maximize the impact of CVI programs, New York City should adopt a comprehensive data-driven approach by enhancing data collection and analysis skills and conducting regular evaluations to track program effectiveness, refine intervention strategies, and allocate resources efficiently.

9. Conduct Regular Longitudinal Studies of CVI Programs: The Mayor’s Office of Criminal Justice (MOCJ) should fund independent, third-party evaluations every 2-3 years to assess the effectiveness of CVI interventions. These studies should identify which strategies yield the greatest reductions in gun violence, highlight areas needing improvement, and inform future resource allocation. The University of Chicago Crime Lab’s evaluations of CVI programs in Chicago provide a strong model for NYC to follow, ensuring data-driven policy adjustments.
10. Establish a Rapid-Response Improvement Mechanism for CVI Programs: The City should create a formal, structured process to regularly identify and implement improvements based on insights from monthly coordination meetings, a reimagined NeighborhoodStat, interviews with CVI practitioners, and real-time data from the Unified CVI Dashboard. This process should be managed by City Hall and ensure that policy adjustments, resource shifts, and operational improvements occur quickly and consistently, to accompany changes made based on long-term evaluations.
11. Expand CMS into High-Risk Areas Without Coverage & Enhance Resources in CMS Coverage Areas with Persistent Violence: Based on this analysis, the City should prioritize CMS expansion in Washington Heights and Longwood, which currently lack CVI organizations despite being current hotspots for violence. NYC should also consider increasing funding and staffing in Brownsville, Northern Harlem, East Harlem, and the South and Central Bronx, where shootings remain frequent despite CMS presence. This expansion should occur in coordination with CVI providers to ensure the resources are well-spent.

Address Persistent Funding and Payment Delays for CVI Providers

NYC should eliminate chronic delays in CMS provider payments and shift to an advance-payment model to ensure financial stability.

12. Fulfill the longstanding promise to launch ContractStat: Transparency is an essential step to improved efficiency; the administration’s long-unmet commitment to launch ContractStat, a uniform public standard for tracking dysfunction and delays in the contracting process, is an impediment to a better contracting process. A functioning ContractStat would put meaningful data in the hands of key leaders meeting regularly to address processes that stand in the way of timely payments. The administration should provide an update on its progress to implementation.
13. Establish Deadlines for Paying Vendors on Time: Currently, the only statutory deadline in the contracting process is a mandatory maximum 30 days for the NYC Comptroller’s Office to review and register City contracts. Meanwhile, nonprofits like CVIs must carry out time-sensitive work regardless of whether their contract has been approved, or paperwork is timely reviewed. In countless cases, nonprofits report taking out costly loans to pay employees and bills while they await delayed payment from the City. NYC should adopt a formal written policy setting a maximum 90-day reimbursement timeframe for City contractors including CVI providers to prevent delays in critical funding. In line with recommendations made by the Human Services Council, the City should establish clear, published milestones for timely review of budgets, invoices, and requests for payment that require the City to pay interest penalties for exceeding deadlines.

Conclusion

Gun violence in New York City remains highly concentrated within a small percentage of city blocks, with persistent violence in neighborhoods like the South Bronx, East Harlem, Brownsville, and East New York. This report’s analysis confirms that CVI works, as shootings decline where these programs are in place. However, despite CVI’s demonstrated effectiveness, violence persists in these same areas, making it critical to refine and expand CVI efforts. To ensure sustained reductions in gun violence, NYC should fully support, strategically expand, and better manage its CVI programs.

The Crisis Management System (CMS) has proven its ability to reduce shootings, but a lack of leadership from City Hall and an absence of data-driven strategies to improve programs, alongside gaps in coverage, funding inconsistencies, and bureaucratic inefficiencies continue to limit its impact. This report’s findings make clear that CVI programs are a vital part of NYC’s public safety strategy, but these programs need to be properly resourced, systematically evaluated, and continuously refined. The reforms laid out in this report will support New York City in effectively addressing the systemic causes of violence, ensuring long-term safety and stability in the communities most affected by gun violence.

Methodology

CMS Efficacy Analysis

Data

The outcome this report’s analysis models is the count of shootings per precinct per year. These shooting incidents are tracked by NYPD and available on NYC OpenData. CVI programs are specifically focused on reducing shootings, so this outcome is a good measure of a program’s efficacy.

The dates each CVI program first became active were drawn from two sources. City Council analysts studied the effectiveness of CVI in 2023 and published with their work a list of CVI providers and their start dates provided by the Mayor’s Office of Criminal Justice in 2022.[30] As part of this report’s investigation, the Department of Youth and Community Development provided an updated list of currently-active groups, and the start dates of providers that have started since 2023. DYCD additionally provided geographic files for the boundaries of the service areas for each CVI provider.

Units of analysis

We evaluated the effect of CVI organizations at the police precinct level and within the specific areas which groups are contracted to serve. Analysis at the precinct level is meaningful because this analysis determined that CVI activities and providers’ relationships with local police cause some effect across the broader precinct area. Moreover, analyzing at the precinct level allows this study to control for particular policing activities or strategies that vary across precincts.

We use calendar years as the temporal unit of analysis. This aggregation smooths over the observed seasonality in shooting events (which peak in summer months) and creates fairly stable counts. In reality, CVI programs started operation throughout the year; this report’s analysis count precinct or service area as having active CVI if a group was active for any portion of the year.

Method

This analysis uses a difference-in-differences analysis to compare the number of shootings in precincts where CVI programs were implemented (“treated” in conventional difference-in-differences terminology) against the estimated number of shootings in those same areas in a counterfactual scenario in which CVI programs were not implemented.

This design assumes that the areas “treated” with CVI programs would have followed the same trend as the “untreated” areas in the counterfactual scenario in which the CVI treatment was not initiated. This allows estimating the difference between this counterfactual expectation and observed reality to estimate the reduction that could be attributable to CVI programs.

A strength of the difference-in-differences methodology is that is does not rely on directly comparing the level of the outcome between treated and untreated areas. CVI is targeted for areas with high levels of violence, and this report’s analysis observe much higher numbers of shootings within the precincts that are treated than those that are not. But the difference-in-differences method only depends on the treated and untreated areas following the same trends, relative to their own baseline levels.

A key requirement for the model is that treated and untreated precincts follow parallel trends before intervention, allowing a credible assumption that they would have continued following parallel trends without intervention. This analysis tested this parallel trend assumption by visually comparing mean shootings per precinct across treated and untreated precincts, normalizing both to the first treatment year.

Figure 1: Pre-treatment parallel trends

Typical implementation of difference-in-differences models depends on a single year of policy intervention, to categorize all years as either before or after this intervention. Because of the staggered implementation of CVI, it was not possible to make this simple before/after distinction across all precinct-years at once. Instead, this analysis applied Callaway and Sant’Anna’s novel approach for computing difference-in-difference estimates across all start years and outcome years and then aggregating a single average treatment effect.³ Because there are many cohorts (precincts grouped by their first year of implementation) with few precincts each, this report’s analysis applied the simple weighted average to compute the average treatment effect. This averaging necessarily gives more weight to the locations where treatment was started earlier for which there are more years of outcome effects. This potential bias could also be considered to be weighting long-term effects.

We modeled the number of shootings per area per year as a function of whether the area was treated with CVI and whether the year was after implementation of CVI:

[Shootings] ~ [Treated] + [Post_implementation] + [Treated&Post_implementation]

We fit models on two geographic scales:

The first model used specific service areas where CVI organizations are contracted to operate as the treatment geographies and included precincts without CVI organizations and the remainder of the precincts outside CVI boundaries as untreated areas. Shootings were spatially joined to these geographies, then totaled by area and year. This yields a panel of 2,242 area-years, across 41 treated areas and 77 untreated areas.

The second model used precincts as geographies, categorizing precincts as treated if CVI organizations were active anywhere in the precinct. This model uses a panel of 1,463 precinct-years.

Results

The model based on specific CVI service areas shows an average reduction of 0.52 shootings per area per year, compared to the number estimated if CVI organizations were not active. Across 41 service areas, this represents a total reduction of 21.5 shootings per year just withing these areas. However, there is a high degree of uncertainty with this model and the reduction is not statistically significant. CVI service areas are geographically small, and shootings remain a rare occurrence, so the median number of shootings per year in a treated CVI area is just 4. This small baseline number of shootings makes it difficult to precisely estimate the treatment effect of CVI.

The second model, measuring the effect precinct-wide, shows a reduction of 7.4 shootings per precinct per year compared to the estimate if CVI were not active. This estimate also has somewhat wide uncertainty, with the 95% confidence interval for the reduction spanning from -2.6 to -12.2. That there is some reduction (i.e. that the observed number of shootings is fewer than the counterfactual estimate) is statistically significant at alpha=0.95.

Building on prior research

We build upon prior research on the efficacy of CVI programs in New York City. This report’s new research corroborates earlier studies which showed CVI was effective at reducing violence.

In a 2017 study, Sheyla Delgado, et al., of the John Jay College Research and Evaluation Center compared counts of shootings and hospitalizations from shootings in two precincts where CVI had been recently started with two comparator precincts which the authors identified as demographically similar.[31] Using an interrupted time series analysis, the authors found reductions in shootings in both precincts where CVI. The South Bronx precinct a statistically significant break from the prior trend, while the decline in the East New York precinct was not large enough to be statistically significant. Both areas where CVI was initiated show larger declines in gun violence that the comparator areas.

This methodology depended on selecting comparator precincts, a necessarily subjective process this analysis aimed to avoid and one more difficult now that CVI has been implemented in so many of the areas that could have served as control areas.

Analysts for the New York City Council produced a 2023 study evaluating the change in shootings after CVI programs were implemented across precincts^.[32] The event-study design compared the number of shootings in the year CVI began, 1, 2, and 3 years after with the count before program start. Overall, this study found an 17% decline in shootings where CVI was implemented, with a 15% reduction in the first year and 16% and 14% reductions in following years.

This report’s analysis advances the field by introducing service areas into the analysis alongside precinct-level impacts and using the most recently available data from 2024. Using a difference-in-differences methodology, this analysis estimated the effect of CVI programs by comparing trends in shootings in precincts with and without active programs. This approach avoids the need for subjective selection of comparator areas and accounts for broader trends in violence. Testing the parallel trend assumption, the analysis confirmed that treated and untreated precincts followed similar trajectories before CVI implementation, ensuring the credibility of these findings.

Historical Financial Analysis

This report’s analysis reconstructed funding for anti-gun violence programming using Detail Initiative Funding Reports (DIFR) produced by the Office of Management and Budget for each quarterly financial plan from November 2014 through November 2024, recording the assigned budget code and unit of appropriation (UA) for each initiative. For prior calendar years where DIFRs were unavailable, the office relied on public Financial Plan Reconciliation documents. The records included any initiative attributed to “Anti-Gun Violence,” “Cure Violence,” or “Crisis Management System.” All initiatives with funding allocated for all years of the published financial plan were assumed to be recurring in perpetuity.

The report’s analysis matched the names of Community Violence Intervention organizations provided by the Office of Neighborhood Safety to their Vendor IDs in the City’s Financial Management System (FMS), and queried FMS for all Payment Request records for known vendors within budget codes and UAs known to have anti-gun violence spending. These results were further filtered for only those that entirely matched known combinations of fiscal year, agency, UA, and budget code. To create calculations on time-to-pay, this report’s analysis used the date fields “Commodity Service To/From Date,” “Payment Request Record Date”, “Disbursement Check Issue Date”, and “Disbursement Check Cleared Date” as relevant milestones as recorded in FMS.

Historically Concentrated Shootings Analysis

After observing that shootings are generally concentrated within a few high-violence areas, this analysis tested the stability of these locations over time. Shootings were spatially joined to Census Tracts (relatively compact areas with populations of approximately 4,000 residents, designated by the Census Bureau), then shootings were summed by Tract by year. For each year, this analysis sorted Tracts by the number of shootings then identified the set of high-violence tracts which together covered half of all shootings in that year. The small number of Tracts in this high shootings set shows how spatially concentrated shootings are. The analysis then tallied how many of the past 19 years each Tract has been among that set of high-shootings Tracts. This number accounts for how persistently a location has suffered from a high number of shootings. Finally, this analysis totaled how many new Tracts were included in the high-shootings set of Tracts each year, finding that there were few one-off locations included, and most were frequently observed to be in the high-shootings group.

Emerging Hotspots Without CVI Present Analysis

This report’s analysis mapped all shootings from 2020 through the end of 2024 to capture recent trends amid and following the COVID-era uptick in shootings. These were visualized as a heatmap, using a kernel bandwidth of 1000 feet to link nearby shootings into hotspot clusters. This map was overlaid with service locations of CVI groups, and with the boundaries of precincts either served or not served by CVI. Analyzing this layers map showed locations with concentrating shootings that lacked CVI (precincts 26, 28, 30, 34, and 41), as well as several areas where shootings were concentrated that were already withing CVI service areas, or precincts that had CVIs. The latter set of areas were compared to the historic concentrations of shootings, revealing that many sites which have concentrated shootings despite the presence of CVI have been locations of concentrated violence for much or even most of the past 19 years.

Acknowledgements

This report was primarily authored by Justyn Richardson, Senior Policy Researcher for Criminal Justice and Public Safety, with research and writing support provided by Annie Levers, Deputy Comptroller of Policy; Dan Levine, Senior Data Analyst and Products Manager; Robert Callahan, Director of Policy Analysis; Khalil Sulker, Strategic Organizer for Community Safety & Restorative Justice, and Nedyah Alexis, Policy Analyst for Mental & Behavioral Health. Report design was completed by Archer Hutchinson, Creative Director.

The Office of the Comptroller thanks the individuals and community violence intervention organizations that took time to meet with our office to share their experiences and insights on violence interruption efforts in New York City. We are also grateful to Senior Research Director Gregory Stoddard and Research Director Terence Chau from the University of Chicago’s Crime Lab for their valuable consultation in refining the methodology used in this analysis. Most importantly, we recognize and deeply appreciate the people working on the streets every day to prevent gun violence—violence interrupters, outreach workers, and community leaders—who put in tireless effort to make our communities safer.

Endnotes

[1] A June 2024 audit by the NYC Comptroller’s Office found that only 13% of ShotSpotter alerts resulted in confirmed shootings, leading to thousands of hours of potentially wasted police time.

[1] Kriegstein, B. (2024, July 25). Hot spots: NYPD data shows most shootings occur on the same blocks year after year. Gothamist. https://gothamist.com/news/hot-spots-nypd-data-shows-most-shootings-occur-on-the-same-blocks-year-after-year

[2] John Jay College Research and Evaluation Center. (2023, November 10). Designing safety: Public safety initiatives managed by the New York City Department of Youth and Community Development. John Jay College of Criminal Justice. https://johnjayrec.nyc/2023/11/10/designingsafety/

[3] RAND Corporation. (2020). The impact of data-informed community violence intervention programs on gun violence in U.S. cities. RAND Corporation. https://www.rand.org

[4] New York City Comptroller. (2024). Audit report on the New York City Police Department’s oversight of its agreement with ShotSpotter Inc. for the gunshot detection and location system. NYC Comptroller’s Office. https://comptroller.nyc.gov/reports/audit-report-on-the-new-york-city-police-departments-oversight-of-its-agreement-with-shotspotter-inc-for-the-gunshot-detection-and-location-system

[5] New York City Comptroller’s Office. (2024, June 20). Audit report on the New York City Police Department’s oversight of its agreement with ShotSpotter Inc. for the Gunshot Detection and Location System (Report No. FP23-074A). New York City Comptroller. https://comptroller.nyc.gov/wp-content/uploads/documents/FP23-074A.pdf

[6] Webster, D. W., Whitehill, J. M., Vernick, J. S., & Curriero, F. C. (2013). Effects of Baltimore’s Safe Streets program on gun violence: A replication of Chicago’s CeaseFire program. Journal of Urban Health, 90(1), 27-40. https://doi.org/10.1007/s11524-012-9731-5

[7] City of New York. (n.d.). NeighborhoodStat (NStat). NYC.gov. https://www.nyc.gov/site/safestbigcity/keep-nyc-safe/neighborhoodstat-nstat.page

[8] New York City Comptroller. (2024, June 20). Audit report on the New York City Police Department’s oversight of its agreement with ShotSpotter, Inc. for the gunshot detection and location system. Retrieved from https://comptroller.nyc.gov/reports/audit-report-on-the-new-york-city-police-departments-oversight-of-its-agreement-with-shotspotter-inc-for-the-gunshot-detection-and-location-system/

[9] New York City Office of the Mayor. (2023, July 31). Mayor Adams’ Gun Violence Prevention Task Force releases “A Blueprint for Community Safety”. NYC.gov. https://www.nyc.gov/office-of-the-mayor/news/555-23/mayor-adams-gun-violence-prevention-task-force-release-a-blueprint-community-safety-#/0

[10] Henderson, H., & Brown, D. (2022, March 31). Preventing gun violence takes more than police. Brookings Institution. https://www.brookings.edu/articles/preventing-gun-violence-takes-more-than-police/

[11] Office of the Mayor. (2023, April 5). Mayor Adams to fund crisis management system at record level of $86 million in next fiscal year. NYC Mayor’s Office. https://www.nyc.gov/office-of-the-mayor/news/462-23/mayor-adams-fund-crisis-management-system-record-level-86-million-next-fiscal-year

[12] City of New York. (n.d.). Crisis management system. PeaceNYC. https://www.nyc.gov/site/peacenyc/interventions/crisis-management.page

[13] New York City Council. (2021). Mayor’s Office of Criminal Justice and Legal Aid Society budget note. Retrieved from https://council.nyc.gov/budget/wp-content/uploads/sites/54/2021/05/MOCJ-LAS-Budget-Note.pdf

[14] CVI Ecosystem. (n.d.). CVI Ecosystem: Advancing violence intervention strategies. https://www.cviecosystem.org

[15] Bureau of Justice Assistance. (n.d.). Bureau of Justice Assistance (BJA). U.S. Department of Justice. https://www.bja.ojp.gov

[16] Yale School of Public Health. (2023). Advancing data and evidence to prevent gun violence. Yale University. https://ysph.yale.edu/news-article/advancing-data-and-evidence-to-prevent-gun-violence/

[17] National Institute for Criminal Justice Reform (NICJR). (n.d.). Gun violence data & analysis. NICJR. https://nicjr.org/gun-violence-data-analysis/

[18] New York City Comptroller. (2024, June 20). Audit report on the New York City Police Department’s oversight of its agreement with ShotSpotter, Inc. for the gunshot detection and location system. Retrieved from https://comptroller.nyc.gov/reports/audit-report-on-the-new-york-city-police-departments-oversight-of-its-agreement-with-shotspotter-inc-for-the-gunshot-detection-and-location-system/

[19] New York City Comptroller. (2024, June 20). Audit report on the New York City Police Department’s oversight of its agreement with ShotSpotter, Inc. for the gunshot detection and location system. Retrieved from https://comptroller.nyc.gov/reports/audit-report-on-the-new-york-city-police-departments-oversight-of-its-agreement-with-shotspotter-inc-for-the-gunshot-detection-and-location-system/

[20] SoundThinking. (n.d.). Miami-Dade County deploys ShotSpotter technology to reduce gun-related crimes. Retrieved from https://www.soundthinking.com

[21] SoundThinking. (n.d.). Pasadena’s use of ShotSpotter technology reduces gunfire-related incidents by 25% in targeted areas. Retrieved from https://www.soundthinking.com

[22] Baltimore Magazine. (2018, January 16). Don’t Take it Outside. https://www.baltimoremagazine.com/section/gamechangers/safe-streets-baltimore-ending-city-violence/

[23] Webster, D. W., Whitehill, J. M., Vernick, J. S., & Parker, E. M. (2013). Effects of Baltimore’s Safe Streets program on gun violence: A replication of the Chicago CeaseFire program. Journal of Urban Health, 90(1), 27–40. https://doi.org/10.1007/s11524-012-9731-5

[24] CVI Ecosystem. (n.d.). CVI Ecosystem: Advancing violence intervention strategies. https://www.cviecosystem.org

[25] Wilson, A. (2021, November 1). There are only two city-funded violence prevention sites tackling Manhattan’s shootings. The Trace. https://www.thetrace.org/2021/11/nyc-gun-violence-prevention-shootings-save-harlem/

[26] Gonen, Y. (2019, September 10). Probers flag anti-violence group cited after mass shooting. The City. https://www.thecity.nyc/2019/09/10/city-probers-flagged-anti-violence-group-cited-after-brownsville-shooting/

[27] Stark-Miller, E. (2022, November 2). Violence interrupter group turns things around with ‘corrective action plan’ after probe, Adams admin says. amNewYork. https://www.amny.com/politics/violence-interrupter-group-corrective-action-plan/

[28] Adams, E., & Mitchell, A. T. (2022, June 13). Transcript: Mayor Eric Adams, Gun Violence Prevention Task Force Co-Chair A.T. Mitchell appear live on Fox5’s “Good Day New York”. The Official Website of the City of New York. https://www.nyc.gov/office-of-the-mayor/news/389-22/transcript-mayor-eric-adams-gun-violence-prevention-task-force-co-chair-a-t-mitchell-appear-live

[29] SoundThinking. (n.d.). Miami-Dade County deploys ShotSpotter technology to reduce gun-related crimes. Retrieved from https://www.soundthinking.com

[30] New York City Council. (n.d.). Cure Violence Programs in NYC. Retrieved from https://council.nyc.gov/data/cure/

[31] Delgado, S. A., Alsabahi, L., Wolff, K. T., Alexander, N. M., Cobar, P. A., & Butts, J. A. (2017). The effects of Cure Violence in the South Bronx and East New York, Brooklyn. Research and Evaluation Center, John Jay College of Criminal Justice, City University of New York. Retrieved from https://academicworks.cuny.edu/jj_pubs/417/

[32] Nuñez, M., Avram, R., Koepcke, E. J., Martinez, R., & Moussawi, A. (n.d.). Cure Violence Programs in NYC. New York City Council Data Team. Retrieved from https://council.nyc.gov/data/cure/