Assessment Findings

Results from the Philadelphia Community Violence Intervention (CVI) Needs Assessment

Franklin S. Moreno and Caterina G. Roman | October 2025

0 / 0

Introduction

This report presents the results from a needs assessment, conducted in early 2025, of community violence intervention (CVI) organizations in Philadelphia, Pennsylvania. A needs assessment is a critical tool for organizational and program capacity building, as it helps identify gaps in resources, training, and infrastructure and provides a roadmap for strategic improvements (Eschenfelder, 2010). Through an examination of strengths and challenges, organizations can align their programs with evidence-based practices, enhance service delivery, and build long-term sustainability. Needs assessments are not solely about identifying gaps in services; they also encompass a broader understanding of the capacity needs at both the organizational and programmatic levels.

The current assessment was specifically focused on the needs of organizations with firearm violence reduction programs that serve youth and adults through any type of case management function, broadly defined. We asked: “Does your agency have a community violence intervention (CVI) program specifically designed to mentor and/or case manage people involved in firearm violence (either as victim, witness, or perpetrator) or are at high-risk for involvement? By mentoring or case managing we mean a formal or informal program that meets with participants at least once per month. ‘Case manage’ also refers to the work of credible messengers and/or outreach staff, who meet participants somewhat regularly.” CVI programs can be defined in various ways, but we have chosen to utilize the definition closely aligned with the Vera Institute of Justice definition (Dholakia & Gilbert, 2021) and that of the U.S. Department of Justice, Office of Justice Programs (2022). CVI programs focus on reducing homicides and shootings by establishing relationships with people at the center of firearm violence in our communities. These programs support those at the highest risk of firearm violence (both perpetrators and victims of violence). We like this definition because it highlights the relational aspect in that CVI programs build pro-social relationships through their strategies, services, and supports and expose participants to new opportunities.

We also invited agencies without formal case management/mentoring components to respond to several key questions about their organizational needs and priorities. This was intentional because all CVI agencies, regardless of model, contribute valuable perspectives on implementation challenges, capacity needs, and priorities. However, we focused the majority of the survey on case management-based violence reduction programs because these programs have proliferated in recent years, are resource intensive, vary widely, and their effectiveness is not clearly understood. The potential for high impact among these case management–based programs is substantial given their high-dosage engagement and service model. Prior research on mentoring and juvenile justice interventions consistently shows that programs delivering frequent, sustained contact between staff and participants and emotional support produce stronger outcomes, including greater trust, skill development, and behavior change (e.g., DuBois et al., 2011; Tolan et al., 2014; Lipsey, 2009). Furthermore, CVI programs that utilize case management likely face significant challenges. Best practices for case management within violence reduction programs remain underdeveloped. Very few structured professional development opportunities exist to equip staff with enhanced skills in case management, conflict mediation, and long-term participant engagement. This creates variations in service delivery, as some programs may provide more robust forms of support than others, likely leading to inconsistent outcomes for participants. Although the needs assessment’s focus is on case management-oriented providers in Philadelphia, Pennsylvania, the survey tool and results may be informative to other stakeholders.

The findings from the assessment have been used to inform the development of the ARCvi online resource hub (ARCvi.org), which is maintained by faculty and staff from Temple University’s College of Liberal Arts with support from the Public Policy Lab. This resource hub has been developed to provide stakeholders with open access to a range of products, including reports, research briefs, webinars, and podcasts. ARCvi features materials specifically focused on CVI as well as broadly applicable resources designed to support the general capacities of community-based organizations in developing strong anti-violence programming, such as grant-writing best practices, and logic model templates. The initiative aims to facilitate both the creation of new products and the curation of existing resources from local and national sources (mostly with curated links to these sources) to enhance accessibility and utility.

Survey Development and Response Rate

The needs assessment survey, programmed and administered using Qualtrics, was distributed to CVI agencies that operate case management or mentoring programs (CMMs). A total of 41 responses were received, of which 37 represented unique agencies and programs eligible for analysis. We calculated the response rate to be 44%. Of these, 32 were case management or mentoring-based CVI programs (CMMs) and 5 were other types of CVI programs.1 Unless otherwise noted, findings presented in this brief focus on the 32 CMM programs. More information on survey development can be found in Appendix A.

Agency and Program Characteristics

Demographics of CVI Agencies (n=37)

Table 1 provides the organizational characteristics of the 37 responding agencies. Given the focus of the survey, it is not surprising that 86% have a case management or mentoring program (CMM) and 14% do not (non-CMM).

Table 1. Agency-wide Demographics
Agency Wide (n=37)	%
Case Management or Mentoring Program
No	13.5%
Yes	86.5
Agency Classification
Non-profit, focused on 1-2 areas	40.5
Non-profit, general human services agency	27.0
Hospital	10.8
College/university	5.4
Government agency	5.4
Community-based	5.4
Public/private/charter school	2.7
Grassroots cooperative with a fiscal manager	2.7
After school programming	2.7
Primary Focus of Agency
Violence reduction	29.7
General/multiple services	18.9
Education	10.8
Recreation and sports	8.1
Victim services	8.1
Advocacy	5.4
Medical or healthcare services	5.4
Other	10.8
We do not have a primary area	2.7
501(C)3 Status
No	13.5
Not yet, but it soon will be	2.7
Yes	83.8
Have More than 1 CMM Program (n=32)
No	50.0
Yes	50.0
Full-Time Staff across Entire Agency
1 to 2	16.2
3 to 10	35.1
11 to 50	27
51 to 100	8.1
101 to 500	2.7
5001 to 1000	2.7
Over 1000	8.1

Also shown in Table 1 is that the majority (83.8%) have 501(c)3 status and 5% are government agencies. Roughly 60% of agencies were either classified as a non-profit focused on a few service areas (35.1%) or a non-profit with a broad range of general human services (24.3%).

The primary program areas for most CMM agencies are violence reduction (29.7%), general/multiple services (18.9%), education (8.1%); and for non-CMM agencies, advocacy (2.7%), education (2.7%), disability, recovery, and restorative justice (2.7%), recreation and sports (2.7%), and victim services (2.7%). Size of agency with regard to staffing (entire agency) varied but leans towards small: with roughly half of agencies (51.3%) reporting having 10 or fewer full-time staff. Only 13.5% reported having over 100 full-time staff (2.7% have 101-500 FT staff; 2.7% have 501-1000, and 8.1% have over 1000 FT staff).

Among the 32 CMM agencies, half (50%) have only one CVI program and the other half have two or more. Those that have more than one CVI program typically are community-based programs (as opposed to hospital- or school-based).

Table 2 shows general program characteristics for the longest-running CMM programs for each responding agency. Because some agencies have more than one CVI program (which can complicate interpretation of responses), the survey asked agencies to answer questions about their “longest-running” program. Most were classified as community-based (65.6%), school-based (18.8%), and hospital-based (15.5%), followed by city-based, court/prison, and law-enforcement based (each at 3.1%, respectively).² Most CMM programs (71.9%) use an evidence-based model, and CMM programs that don’t (15.6%) and don’t know (12.5%) are almost evenly split. The length of time agencies have been operating their programs varied widely. One quarter (25%) were new or relatively new; roughly one-third (31.3%) had been operating for 3-5 years; 15.6% for 6-10 years. The primary sources of funding for the majority of the longest-running programs are the City of Philadelphia (65.6%) and PCCD/other state funding (50%). Table 2 has the full list.

Table 2. Program Characteristics (CMMs only)
Longest-Running CVI Program (n=32)	%
Program Classification*
Community-based	65.6%
School-based	18.8
Hospital-based	12.5
City-based	3.1
Court/Prison-based	3.1
Law enforcement-based	3.1
Use Evidence-Based Model
Yes	71.9
No	15.6
Don’t know	12.5
Geographic Area Served
Philadelphia region	18.8
City of Philadelphia	53.1
Only specific zip code(s)	15.6
Only specific neighborhood(s)	6.3
Other	6.3
Length of Time Operating Program
0 to 2 years	25.0
3 to 5 years	31.3
6 to 10 years	15.6
> 10 years	28.1
Primary Source of Program Funding*
City of Philadelphia	65.6
PCCD/other state funding	50.0
Donations	18.8
Federal grants	15.6
Private foundations	12.5
A mix of all the above	15.6
Other	9.4
Note: *Responses are not mutually exclusive. Percentages add to over 100%.

The City of Philadelphia is the primary source of funding for most CVI programs

Table 3 shows participant characteristics for the longest-running CMM CVI programs. The overwhelming majority (68.8%) do not limit their services to either juveniles (legal minors) or adults, but 15.6% of programs only serve juveniles and another 15.6% focus on adults only. The eligibility criteria vary widely for programs but across programs, eligibility (the categories are not mutually exclusive) often include age (68.8%) and assessed risk for violence (65.6%), with just over half including being a victim of firearm violence (53.1%) and half using geographic (50.0%) or school-based (50.0%) criteria; fewer hinge on court adjudication (34.4%) or hospital connections (31.3%), and only a small minority cite assaults/stabbings (9.4%) or any socioeconomic screens such as income or public benefits (each 3.1%).

Most agencies decide eligibility internally (56.3%); 28.1% follow a specific evidence-based model; and 9.4% are mandated by a court, funder, or government. Over half report they are not dependent on outside referrals (53.1%), 28.1% say “somewhat,” and 15.6% say “yes.” Graduation criteria (multi-select) most often include criterion that are education-focused—attending school or meeting education goals (53.1%)—though 37.5% report no formal graduation criteria.

Table 3. Program Participant Characteristics: Eligibility and Flow
Longest-Running CVI Program (n=32)	%
Participant Ages Served
Legal minors only (i.e., juveniles)	15.6%
Adults only	15.6
Both legal minors and adults	68.8
Eligibility Criteria*
Age	68.8
High risk for violence	65.6
Firearm violence victim	53.1
Geographic location	50.0
School-based program	50.0
Court adjudicated	34.4
Received hospital services	31.3
Victim of assault/stabbing	9.4
No specific criteria	9.4
District Attorney Office referrals	3.1
Income	3.1
Youth in danger of not graduating high school	3.1
Youth/parents receive public benefits	3.1
Determination of Program Eligibility
Own agency decided	56.3
Follow evidence-based model	28.1
Mandated by court/government agency/funder	9.4
Other	3.1
Missing	3.1
Dependent on Referrals from Other Agencies
No	53.1
Somewhat	28.1
Yes	15.6
Program Graduating Criteria*
Attending school or completed education goals	53.1
No formal criteria	37.5
Program based on months/years in program	21.9
Achieved ‘low’ to ‘no’ risk assessment	18.8
Has stable employment	15.6
Accomplished set goals	12.5
Improved mental health	6.3
Ideas, skill sets, and behaviors transformed	6.3
Murder case closed	3.1
Age	3.1
Stable housing	3.1
No legal involvement	3.1
Created lifelong mentorship	3.1
Note: *Responses are not mutually exclusive. Percentages add to over 100%.

Other markers include time in program (21.9%), achieving a low/no risk score (18.8%), and stable employment (15.6%), with smaller shares recognizing goal attainment (12.5%), improved mental health or transformed skills/behaviors (each 6.3%), and rare case-specific or stability indicators (e.g., murder case closed, age, stable housing, no legal involvement, lifelong mentorship; each 3.1%).3

Figure 1 shows the distribution of programs by size (the number of unduplicated participants served annually by the program itself, not across the whole agency). Roughly 41% indicated their longest-running program serves anywhere between 51-100 annually; only a small percentage of programs were very small (20-50 participants annually) or large (301 and over). We also asked agencies about the upper threshold of their capacity (not shown). Most programs report relatively modest annual capacity: over half can serve fewer than 100 participants (56.3%), and about three-quarters cap below 200 (75.1%). The most common band is 51–100 annually (28.1%), followed by 20–50 and 101–200 (each 18.8%). Only a small share can handle 300 or more in a year (9.4% combined), and just one program (3.1%) describes their capacity as “unlimited.” Overall, the distribution skews toward smaller to mid-sized programs.

The programs reported their average times for participants to remain in or complete the program (not shown): 40.6% reported that participants stay in the program for 3-6 months, one quarter (25%) reported that program stay is somewhere between 7 and 12 months, 15.6% reported just under 2 years and another 15.6% indicated that the average participant can remain in the program over two years. Only one program indicated that the average program duration was less than 3 months.

Participant Assessments

Figure 2 shows that most of the CMM programs (84.4%) conduct a needs and/or risk assessment for each participant at the beginning of the program; and of those who use some type of risk and/or needs assessment (n=27), 40.7% use a validated formal assessment tool, 37% do not, and 22.2% of respondents were not sure whether their assessment tool was considered validated.

Table 4. Program Staffing Characteristics
Longest-Running CVI Program (n=32)	%
Full-Time CVI Staff
None	3.1%
1 to 2	18.8
3 to 5	34.4
6 or more	43.8
Part-Time CVI Staff
None	34.4
1 to 2	31.3
3 to 5	15.6
6 or more	18.8
CVI Volunteers
None	50.0
1 to 2	18.8
3 to 5	9.4
6 or more	21.9
Full-Time Case Manager or Mentors
None	6.3
1 to 2	43.8
3 to 5	28.1
6 or more	21.9
Part-Time Case Manager or Mentors
None	50.0
1 to 2	25.0
3 to 5	15.6
6 or more	9.4
Train within 30 Days of Hire
No	6.3
Yes	90.6
Missing data	3.1
Who Conducts CMM Training*
Our own agency	84.4
Another Philadelphia-based agency	34.4
Global/national agency	15.6
Consultant	12.5
Other	6.2
Participant Caseload
1 to 2	15.6
3 to 5	0.0
6 to 10	21.9
11 to 15	37.5
16 or more	25.0
At Maximum Allowed Caseload
No	37.5
Yes	62.5
Note: *Responses are not mutually exclusive. Percentages add to over 100%.

Program Staffing

Table 4 shows program staff characteristics for the CMM programs. Overall, staffing patterns suggest that many of the longest-running CMM programs have established full-time teams but remain modest in scale. Nearly half (43.8%) report six or more full-time CVI staff, indicating a strong core workforce, while a similar proportion rely on only a few part-time staff or none at all, reflecting leaner operational models. We examined the configuration of staff across full-time, part-time and volunteer categories (not shown) and found that the most common configuration among agencies (10 agencies; 31.3%) is to have 6 or more full-time staff and no (0) part-time staff or very few part-time staff (i.e., 1-2). Smaller agencies—those with only 1–2 full-time staff—rarely have more than 2 part-time staff; those with 3–5 full-time staff are the most likely to have 6+ part-time (4 cases). Volunteers are less prevalent than part-time staff. It appears volunteer reliance increases with agency size, but half of agencies have no volunteers at all, indicating a primarily paid-staff model.

Staff stability appears relatively high, as over two-thirds (68.8%) of agencies did not experience turnover in the past year (not shown). Among the minority that did, most lost only two to four employees, suggesting moderate rather than chronic instability.

Training practices appear strong. Nearly all programs (90.6%) train new case managers or mentors within the first 30 days of hire, and the majority (84.4%) deliver this training internally, which demonstrates investment in program-specific knowledge. Philadelphia-based peer training also plays a meaningful role (34.4%), signaling a collaborative learning culture within the city’s CVI ecosystem. Programs that utilize global or national training (15.6%) are likely those who utilize models or strategies which have a national organization that is available to provide implementation training, such as Cure Violence Global or ROCA.

Caseload data show moderate participant-to-staff ratios overall. While over one-third (37.5%) of programs assign 11–15 participants per case manager, another quarter (25%) manage 16 or more, suggesting potential strain in some programs.

Most agencies (62.5%) formally cap caseloads, with limits typically between 11 and 19 participants, reflecting intentional efforts to balance engagement intensity with reach. Together, these indicators portray relatively professionalized CVI programs with attention to staff training, manageable caseloads, and generally stable workforces—key ingredients for sustaining long-term violence intervention efforts.

Core and Referral Services

We asked CMMs to review a list of twenty-three services and indicate whether the agency directly provides, refers out, or does not provide the service at all. (Note that respondents could choose both “directly provide” and “refer” out if they do both.) We report on the core services that are directly provided by 50% and more of the CMM programs (Figure 3). The services are case management or mentoring (87.5%), employment support (81.3%), social emotional skills (75%), and peer support groups (71.9%), followed by conflict mediation (68.8%), financial support (59.4%) and anger management (50.0%).

Figure 4 shows the services that are referred out by 50% and more of the CMM programs. Across the survey, several service categories stood out as those most commonly referred out rather than directly provided in-house. The most frequently referred services included housing relocation (71.9%), vocational training (68.8%), substance use support groups (68.8%), and legal assistance (62.5%). Over half also referred participants to education-related supports (59.4%), mental health counseling (56.3%), and medical or healthcare services (56.3%), while half referred to CBT or cognitive-behavioral therapy groups (50%).

CMM agencies directly deliver relationship-centered services and financial support while referring out resource-intensive or specialized supports.

This division reflects a strategic use of capacity—anchoring participants in mentorship and case management while leveraging broader community systems for housing, health, and legal aid.

These findings show that programs often rely on external partnerships for specialized or licensed services—especially those requiring clinical credentials, legal expertise, or substantial infrastructure (e.g., housing or medical care)—even while maintaining strong in-house delivery typical of case management, and employment-related supports.

By contrast, several services are not offered at all by a notable share of respondents (see Figure 5). Crisis response after a shooting or homicide (40.6%) and victim compensation (37.5%) are often missing from program portfolios—but this may be because programs are not necessarily serving victims of violence. Strikingly, over one-third (34.4%) reported not offering legal assistance in any form and none directly provide legal assistance (the latter not shown). Fatherhood and parenting programs also appear to represent a gap in current CVI service offerings, with 34.4% not offering these services directly or through referral. Parenting-skills and related supports can be especially important for both men and women in violence-impacted communities, helping them strengthen family relationships, develop positive identity roles. However, even within common service domains, some gaps remain: for example, around one-quarter do not directly provide or have referral partnerships for medical/health services (25%).

Overall, the pattern reflects a clear division between core violence intervention functions—such as mentoring and conflict mediation which are mostly provided directly—and specialized stabilization services like housing, mental health, and legal assistance, which are primarily accessed through referral. This mix suggests strong programmatic engagement capacity but a heavy dependence on broader community and institutional systems to address participants’ comprehensive needs.

Participant Needs

We did not want to infer participant needs based solely on the services agencies currently offer or lack. Instead, we asked CMM programs to identify and rank what they perceive to be their participants’ most pressing needs. Specifically, respondents reviewed a list of predetermined needs and services and ranked their top four priorities (1 through 4) responding to the prompt:

“Thinking about your participants, what do you perceive are the top four priorities related to participant needs and services for your longest-running CVI mentoring or case management program?”

The top-ranked participant needs (Figure 6) are case management and/or mentoring (59.4%), followed by employment support (43.8%), housing or relocation assistance (37.5%), and mental health services (34.4%) (see Figure 6 for the full list and note that crisis-intervention, although included in this survey question on top service needs, was not included in the survey items asking how program services were provided).

When asked whether participant needs differ for legal minors versus adults, 84.4% of CMM programs reported that they do—highlighting the importance of tailoring supports to different age groups (not shown).

We compared stated needs for participants against the services available in the programs and found:

Case management/mentoring and employment support stand out as both widely available (provided by over 80%) and top-ranked participant needs (59.4% and 43.8%). This alignment suggests that programs are investing heavily in the relational and economic foundations of change—consistent with CVI’s emphasis on trust-building and stability.
Participants’ priorities reflect a strong demand for mental and emotional well-being supports: crisis or pre-crisis intervention (31.3%), anger management (28.1%), social-emotional skills (28.1%), conflict mediation/resolution (25.0%), and cognitive behavioral-related therapies (21.9%). These areas are often directly provided by fewer than half of agencies, suggesting an opportunity to deepen cognitive and emotional skill building supports, as well as trauma-informed and therapeutic competencies within CVI teams. Service needs related to anger management, cognitive and social-emotional skill development are particularly notable as they sit at the center of many theories of change for community violence intervention programs. Each reflects a pathway through which CVI practitioners aim to reduce the likelihood of retaliatory or impulsive violence by helping participants regulate emotion, manage conflict, and rebuild a sense of control. In models such as Cure Violence, READI Chicago (P3 in Philadelphia), and other credible-messenger frameworks, emotional regulation is part of the process in interrupting cycles of harm and sustaining behavior change.
Housing/relocation assistance ranks as the third-highest participant need (37.5%) but is directly provided by only about one-third of agencies (31.3%); most (71.9%) refer out and 15.6% do not provide at all.

Staff Training Needs

CMM agencies were also asked to review eleven pre-determined training needs for staff and rank those needs 1 through 11. Figure 7 highlights the results for the top training needs ranked between 1-5. The top-ranked training needs include 59.4% of agencies indicating trauma-informed care, 56.3% mentioned finding/managing grants, 43.8% strategic marketing, 43.8% obtaining housing/relocation, and an equal share indicating training needs around engaging the hardest-to-reach participants (40.6%) and using risk and needs assessments (40.6%).

Although not shown in a graphic, we found that almost half (46.9%) of the CMM agencies reported needing bilingual case managers or mentors. For the CMM agencies needing bilingual staff (N=15), Spanish is the language most needed (86.7%), followed by French (13.3%), Creole (6.7%), and Russian (6.7%).

Partnerships and Collaborations

We asked CMMs which types of partners do you believe are most needed to support your CVI mentoring/case management program at the current time and to select up to four of twelve predetermined responses. The top three desired types (Figure 8) were CBOs who provide services we don’t (81.3%), city/state agencies (50%) and CBOs who do what our agency does (34.4%). Most agencies (81%) said they would most like to partner with CBOs that provide services they currently do not, reflecting a strong appetite for service complementarity and cross-organizational collaboration. Half (50%) identified city or state agencies as desirable partners, underscoring the need for stronger connections between community programs and public systems. About one-third indicated interest in partnerships with schools (34.4%), peer CBOs offering similar services (34.4%), and federal agencies (31.3%), suggesting that programs value both horizontal and vertical integration within the CVI ecosystem. Fewer agencies mentioned universities/research firms (25%), which may reflect existing ties or limited capacity to formalize such collaborations.

We then asked: what benefits do you hope to gain from partnerships/collaborations? Figure 9 shows the responses; the benefits named by three-quarters or more of programs were training/workshops (84.4%), networking opportunities (81.3%), and obtaining new funders (75%)—suggesting that agencies are eager to build skills, relationships, and financial sustainability.

Over half (53.1%) hoped that better access to data and general information on firearm violence would come from partnerships. Roughly four in ten saw value in partnerships to help their programs gain exposure through news media outlets (43.8%), to find volunteers (40.6%), or to share program data and successes (37.5%), reflecting interest in visibility and collective learning. About one-third (31.3%) cited research and evaluation support as a benefit of future partnerships, indicating that while building evidence remains important, most programs are currently prioritizing or are most interested in building collaborations that strengthen operational capacity—such as training, networking, and funding opportunities.

Program Challenges

We also wanted to obtain an understanding of key programmatic challenges without agencies having to choose from a pre-determined list. So we asked the open-ended question: Please list three of the most significant challenges your program faces. The answers are provided in Table 5 in alphabetical order; many have been recoded to collapse/group similar responses. The bold text indicates where more than 1 program listed that particular challenge (1st, 2nd or 3rd).

Table 5. Key Programmatic Challenges
1st Greatest Challenge Listed	2nd Greatest Challenge Listed	3rd Greatest Challenge Listed
Diversification of funding sources	Accessing/using social media	Aid (financial) for participants
Funding (11 programs)	City has limited resources to serve target population	Building partnerships with suburban counties
Funding to recruit and retain skilled staff (2 programs)	Collaborating with schools (2 programs)	Collaborating with schools
Hiring staff (2 programs)	Consistent referrals	Consistent communication w/partners
Housing services/finding housing (2 programs)	Engaging population through social media	Data collection
Lack of formal community partners	Finding sustainable employment for target population	Dedicated staff for social media outreach and marketing successes
Limited partners for transitional employment	Funding (6 programs)	Expanding to other neighborhoods
Participant attendance	Funding/ability to pay staff higher wages	Funding (2 programs)
Physical space	Housing relocation services/finding housing	Funding for expansion
Regular mentor participation	Keeping tabs on resources available	Hiring process
Slow reimbursable funding mechanisms	Lack of partners to employ youth	Intimate partner violence
Staff retention	Managing participant challenges	Involving more spiritual leaders
Staff support/wellness	Managing personnel	Limited incentives to reach target population
Staffing	Outdated equipment	Measuring impact and effectiveness
Sustaining funding	Participant engagement and retention	Non-responsive and overburdened community-based programs inhibiting possible collaboration
Understanding the popular culture	Partners knowledgeable about HVIP model	Participant outreach and engagement
Unresponsive larger ecosystem	Physical space	Physical space
	Providing job readiness skills and job placement	Recruiting/hiring staff
	Staff support/wellness (2 programs)	Security at events
	Staffing (2 programs)	Staff pay
	Transportation for youth	Staff training
		Staffing
		Supporting the travel needs of staff doing community-based work

Across all three columns in Table 5, two broad themes dominate: staffing-related challenges and funding limitations, with several cross-cutting issues around partnerships, participant engagement, and program infrastructure.

Funding concerns were by far the most prevalent, mentioned by at least 11 programs as their primary challenge and appearing repeatedly in the second and third priority columns (e.g., six programs citing it again in column 2 and two more in column 3). A number of other programs were more specific about challenges related to funding, such as slow reimbursement mechanisms, funding to attract top-quality staff and funding to expand program.

Staff-related issues were the second most common cluster of responses, cutting across hiring, retention, training, and wellness. Programs reported difficulty recruiting, hiring and retaining qualified staff and supporting staff well-being. Together, these findings indicate that even when funding is available, sustaining a stable, skilled, and supported workforce remains an acute challenge for many CVI programs.

Beyond staffing and funding, programs pointed to barriers in the broader CVI ecosystem, including a city that is resource-poor, particularly with regard to housing and employment opportunities, and limited collaboration with schools and employers. Not surprisingly, CVI programs operate in fragmented systems with uneven resource flow.

Evidence-based Models and Use of Data

Most responding CVI agencies demonstrate a strong orientation toward evidence-informed practice and data use (see Figure 10). Nearly three-quarters (71.9%) report following an evidence-based or evidence-informed model, and an equal share (71.9%) use a formal data management system to track participants or program activities. About two-thirds (62.5%) use government crime data to inform planning, strategy, or grant writing, while slightly fewer (56.3%) use social and economic data for those same purposes. These findings suggest that while most CVI agencies are integrating data and evidence into their operations, there is room to strengthen partnerships that expand access to real-time, neighborhood-level data and enhance cross-sector data sharing.

Conclusion

Within communities, nonprofits and government agencies working to fight community violence continue to compete for scarce resources, limiting the collaboration and coordination that could strengthen programming and service delivery (Marwell & Gullickson, 2013). The findings from this needs assessment make clear that many, if not most, of CVI programs face barriers that constrain their capacity to grow and sustain high-quality, data-informed programs and practices.

Agencies appear to be responsive to the complex needs of participants, yet resource limitations shape what they can directly deliver. Most programs provide a wide-range of relational supports—such as case management, mentoring, and employment readiness—while referring out critical but resource-intensive services like housing, mental health care, and legal assistance. These patterns point to both the strength of relational interventions and the fragility of the service ecosystem surrounding them. Gaps in behavioral health, housing, and family supports illustrate missed opportunities to address the broader social determinants of violence.

While most agencies demonstrate a strong commitment to evidence-based models and data-driven approaches and maintaining data management systems, some still lack reliable data systems or access to longitudinal firearm violence and socioeconomic data that would enable more precise targeting, planning, and outcome monitoring. Strengthening partnerships across community-based and government agencies, policymakers, news media, researchers, businesses and data-holding institutions may be one way to improve the ability of programs to provide a wide range of needed services, allocate resources effectively, perform high-quality outcome reporting and internal monitoring, and quickly showcase successes.

At the organizational level, the greatest challenges cited by respondents centered on funding instability and workforce capacity. Programs struggle to recruit, train, and retain skilled staff; slow reimbursement processes and limited funding flexibility and diversity add to these staff challenges. Although programs are resourceful and adaptive, we know from our research experiences that the current network for training, technical assistance, and organizational support in Philadelphia (and in other cities) typically is fragmented and difficult to navigate. Most technical assistance opportunities arise only in connection with specific grant awards, leaving many agencies—especially smaller or emerging ones—without access to consistent or comprehensive guidance. As a result, there are few accessible mechanisms to help programs strategize on how to secure new resources, strengthen operations, improve data use, or showcase successes outside of short-term grant cycles. This lack of accessible assistance makes it difficult to align capacity-building efforts across the CVI field and to ensure that programs at different stages of development can learn from one another and apply evidence-based practices effectively.

It is important to note that the findings from this needs assessment point not to deficits in will or vision, but to opportunities for strategic investment and coordinated infrastructure that can amplify the effectiveness of the city’s violence intervention ecosystem. With sustained funding, shared data access, and equitable partnerships among government, research, and community stakeholders, Philadelphia can move toward a more connected, infrastructure data-informed, and resilient CVI system capable of producing lasting reductions in violence and advancing community well-being.

References

Banyard, V., & Hamby, S. (2022). Strengths-based prevention: Reducing violence and other public health problems. American Psychological Association. https://doi.org/10.1037/0000267-000
Dholakia, N. & Gilbert, D. (2021). Community Violence Intervention Programs Explained. https://www.vera.org/community-violence-intervention-programs-explained (Accessed October 10, 2024).
DuBois, D. L., Holloway, B. E., Valentine, J. C., & Cooper, H. (2002). Effectiveness of mentoring programs for youth: A meta-analytic review. American Journal of Community Psychology, 30(2), 157–197. https://doi.org/10.1023/A:1014628810714
Eschenfelder, B. E. (2010). Using community-based needs assessments to strengthen nonprofit-government collaboration and service delivery. Journal of Health and Human Services Administration, 32(4), 405–446.
Lipsey, M. W. (2009). The primary factors that characterize effective interventions with juvenile offenders: A meta-analytic overview. Victims & Offenders: An International Journal of Evidence-based Research, Policy, and Practice, 4, 124–147. https://doi.org/10.1080/15564880802612573
Marwell, N. P., & Gullickson, A. (2013). Inequality in the spatial allocation of social services: Government contracts to nonprofit organizations in New York City. Social Service Review, 87(2), 319–353.
Sansfaçon, D., & Waller, I. (2018). Recent evolution of governmental crime prevention strategies and implications for evaluation and economic analysis. In Costs and benefits of preventing crime (pp. 225–247). Routledge.
Tolan, P. H., Henry, D. B., Schoeny, M. S., Lovegrove, P., & Nichols, E. (2014). Mentoring programs to affect delinquency and associated outcomes of youth at-risk: A comprehensive meta-analytic review. Journal of Experimental Criminology, 10(2), 179–206. https://doi.org/10.1007/s11292-013-9181-4
Office of Justice Programs. (2022). Community Violence Intervention. U.S. Department of Justice. https://www.ojp.gov/archive/topics/community-violence-intervention (Accessed October 10, 2024).

Appendix A. Methods

The needs assessment survey items were developed based on initial meetings with community and city-based CVI agencies in Philadelphia, as well as a review of the extant literature. To determine eligibility, respondents were asked whether their agency has a CVI program specifically designed to mentor and/or provide case management for individuals involved in firearm violence (as specified earlier in the introduction of this brief). The final survey protocol can be found in the appendix.

In late 2024 a unique survey link was generated for 73 agencies that were known to us as likely having a CMM CVI program. In addition, a QR code was created and distributed by a city-agency network for CVIs that were potentially not our initial list. Each agency with a unique link that had not completed the survey was contacted subsequently at least three times to encourage completion. The survey was live for four months. We calculated the response rate as 44%; 41 responses to the survey were recorded; 2 records were excluded as duplicates, and 2 were excluded for missing data—giving us 37 agencies reporting on 37 programs: 32 are CMMs and 5 are non-CMMs). Note that the majority of this summary report focuses on the 32 CMMs programs; but a few tables, including Table 1, show the results from all 37 responding agencies.