Credible Messenger Mentoring to Promote the Health of Youth Involved in the Juvenile Legal System: A Narrative Review

Abstract

This narrative review examines the literature on credible messenger mentoring (CMM) as an intervention to promote the health and well-being of youth involved in the juvenile legal system. In the CMM model, individuals with shared life experiences (e.g., from the same neighborhoods or marginalized communities, with former gang or incarceration history) serve as mentors, leveraging their own history of transformation to engage youth involved in the juvenile legal system and promote individual and community change. CMM is an increasingly popular approach for working with youth involved in the juvenile legal system, yet the state of the research on this intervention is unclear. This article provides a narrative review of existing research on CMM to understand what is known, and unknown, about the intervention. Results find an emerging, but incomplete body of evidence supporting the impact of CMM on youth involved in the juvenile legal system and adult mentors. Qualitative and observational findings provide stronger support for the model, while quantitative findings provide more mixed and incomplete evidence, indicating that CMM may be a promising life course health intervention, yet needs more empirical study. Findings from this review suggest the value of integrating community-generated evidence in the evaluation of health interventions. Future research can inform contemporary interest in the CMM approach and guide implementation and measurement standards for optimizing intervention delivery with youth involved in the juvenile legal system.

Credible messenger mentoring (CMM, defined as mentoring by individuals with lived experience) is an emerging strategy to promote the healthy development of young people involved in the juvenile legal system (JLS), their families, and communities. Youth involved in the JLS face a number of health disparities. Youth are more likely to enter the JLS with greater health needs, and JLS involvement itself—especially detention and incarceration—contributes to long-term health consequences persisting into adulthood, including higher morbidity and mortality rates.^1–3 Health disparities are further compounded by the over-representation of young people of color, LGBTQ+ youth, and youth from impoverished communities in the JLS.⁴ Moreover, most interventions for this population focus on mitigating risks and preventing recidivism, rather than promoting health or prosocial development. There is an urgent need for policy and practice innovations that foster health equity and promote long-term healthy development and well-being for youth who are already involved or are at risk of becoming involved in the JLS.

Purpose of this Review

CMM is a burgeoning health promotion strategy that aligns with the growing recognition of violence and carceral system involvement as strong determinants of health.^5,6 Multiple technical reports have described the CMM model in depth, provided examples of programs, and advocated for its expansion (see Table 1). However, there remains a need for thorough synthesis of empirical research on this increasingly popular intervention. This narrative review seeks to address the gap in understanding the scope and quality of research on the impact of CMM interventions, especially in terms of how it may promote health equity and long-term healthy development and well-being. To that end, we survey the CMM literature by analyzing the state of evidence on this model. We first describe the core components of the CMM model. Next, we describe our narrative search methods and provide findings and analysis on the state of empirical evidence on CMM, followed by considerations for CMM through a life course health framework. We conclude with implications and future directions for research to further understand the impact of CMM on healthy development.

Table 1:

Additional technical reports with examples of CMM programs

Ready4Work 38,39

DYRS Credible Messenger Initiative 19

Youth Advocate Programs 19

Project Safe Neighborhoods 8

reVision 8

READI 8,19

Roca, inc. 19

Community Connections for Youth 19

Southeast Credible Messengers 11

Maine Credible Messenger Movement 11

NuEntry Opportunity 11

Personal reflection from an Arches mentor 29

Other cities and county-level CMM initiatives (Described in: Credible Messenger Justice Center¹⁰ ): City of Milwaukee; Racine County, WI; King County, WA; Onandaga County, NY; Harris County, TX; Middlesex County, NJ

The Credible Messenger Mentoring Model

The credible messenger concept was first developed in 1979 by a group of men in prison, led by Black Panther Eddie Ellis, when they envisioned a radical social movement of people who were currently and previously incarcerated, applying their lived experience to empower and lift up marginalized communities.^7,8 In the years since, grassroots credible messenger approaches have supported communities in multiple ways, including serving as first responders to shootings, de-escalating gang conflicts, conducting street outreach, and hosting community events.^6,9 Credible messengers also increasingly work in a mentoring capacity with youth in the JLS in many regions of the United States, garnering national media coverage and recognition.^10–15 Credible messenger mentoring (CMM) is practiced in the community with young people identified as high risk for involvement in violence or the JLS, inside juvenile facilities, in alternatives to incarceration programs, and during the re-entry process.¹⁶

At its core, CMM connects youth involved in the JLS with trusted adult mentors who have relatable life experiences and can show them a different path forward.¹¹ Credible messengers are uniquely skilled and positioned to engage young people in the JLS who may be disengaged from traditional services and hard to reach through more traditional social service approaches. Credible messengers can relate to the challenges that youth face, and have successfully navigated the types of changes that they are encouraging youth to make, increasing their credibility in the eyes of youth. As one report describes: “The problem is not the message; the young person simply has a hard time identifying with the messenger” (p. 3).¹⁷ Mentors’ credibility, based on their lived experience, helps build a foundation of trust and strong relationships to support young people’s growth and change. In addition to the benefits for youth, CMM is also designed to promote intergenerational and community health and well-being. The CMM model views mentors as assets to be supported and developed as leaders, and provides them with opportunities for gainful employment and to make meaningful change in their communities where they otherwise often face structural barriers related to their own system-involvement.¹⁸

Credible Messenger Mentoring Defined

Our review of the literature indicates that there is not one universal definition of CMM; rather, it is a mindset and philosophy constituted of certain core elements, with room to adapt to individuals and local communities.⁸ CMM programs generally aim to reduce violence, crime, and contact with the carceral system, and to transform young people’s thinking and behavior in ways that help shift their life trajectories.^8,11 However, CMM programs vary considerably, such as in the frequency and duration of services, compensation for mentees, and the extent of collaboration with probation and law enforcement. Additionally, the format of service provision can differ to include individual and/or group mentoring, therapeutic and mental health programming such as cognitive behavioral therapy, and referrals to community-based services and opportunities. Regardless of the specific approach, CMM is grounded in relationship-building, and a deep commitment to intensively and holistically supporting youth and communities impacted by the JLS and carceral system.⁹

Similarly, there is no universal definition of what makes someone a “credible messenger.” A participatory study conducted by the Urban Institute⁸ of a diverse sample of CMM programs and providers found that credible messengers define their role fluidly. Respondents understood lived experience broadly, where an “intimate knowledge, connection, and embeddedness” in the communities they work with fosters their credibility (p. 25). While many credible messengers have been incarcerated or involved in gangs, CMM recognizes a range of direct and indirect experiences of marginalization and involvement in the carceral system. The results of the Urban Institute participatory study⁸ offer the following guidelines for defining a credible messenger:

• Lived experience: having direct, or indirect (e.g., a family member or close friend who is or was incarcerated) experience with the carceral system. This experience may also include living in heavily criminalized communities or involvement in other systems such as foster care.

• Transformation: experienced a major life change, such as desisting from crime or leaving a gang, and now acting as a leader in their community to support others in making similar transitions.

• Community credibility: considered credible in the eyes of youth and community members they work with (rather than by an external stakeholder or system).

• Mentoring and relational skills: skilled in communication, listening, rapport-building, empathy, and mentoring, with a deep commitment to lifting up marginalized youth and communities.

We initially approached the present review with the intention of differentiating credible messengers and CMM programs from other similar models (e.g., peer mentors, peer navigators, youth coaches, transformative mentoring, community health advocates, returned citizens, agents of change, lived experience experts). However, our review found that these terms are often used interchangeably. Furthermore, credible messengers in the Urban Institute report suggest that the specific language used is not as important as the spirit of the work itself.⁶ Thus, for the purpose of this review, we utilize the term credible messenger mentoring (or CMM) for consistency, but define it inclusively based on the core criteria of lived experience, transformation, community credibility, and mentoring and relational skills outlined above.

Method

We conducted an extensive literature search to assess the state of empirical knowledge on CMM. Search terms included: “credible messenger mentoring”, “credible messenger re-entry”, “credible messenger juvenile”, “credible messenger youth”, “credible messenger probation”, “credible messenger aftercare”, and “transformative mentoring juvenile.” We also searched the CMM synonyms listed above, but did not find additional studies meeting the aforementioned criteria defining CMM. The search period included literature from 2010 to 2023, retrieved from the following databases: Google Scholar; ProQuest Social Service Abstracts; National Criminal Justice Reference Service; and Crime Solutions Program Profiles. Additionally, we reviewed citation lists of the studies obtained, as well as technical reports advocating for CMM programs and providing local examples,^{8,11,13,17,19} and the research library of the Credible Messenger Justice Center, a national organization focused on credible messenger programs.

We included studies for consideration if they focused on youth and young adults as the client population, described work aligned with core criteria of credible messenger mentoring outlined above, and served youth formerly or currently involved in the JLS, or who were designated high-risk for JLS involvement. Studies that examined credible messengers as a community-based violence intervention approach, but without an explicit mentoring component were not included. Reports that focused on program implementation, capacity-building, or cost-savings of CMM, without findings related to their impact on youth or adults were also excluded. We included quantitative, qualitative, and mixed-methods studies, but excluded reports that referenced findings or impacts but lacked primary data. Some studies also included other data or analyses from the organization, such as cost-benefit analyses; however, we focused on findings related to CMM’s impact on youth and mentors. Additionally, we excluded articles that referenced CMM as model programs in technical reports, but where no evaluation had been conducted or it was not possible to determine or access the original evaluation (see Table 1).

We conducted a narrative review rather than a systematic review due to the emerging nature of research on CMM, the varying types of methods and research designs examined, and the importance of incorporating community-generated evidence and grey literature. Unlike a formal systematic review, there are fewer established standards for evaluating evidence in a narrative review. We referenced existing narrative reviews conducted on youth justice programs,^20,21 and best practice standards for evaluating grey literature²² to develop the following review criteria:

• Quantitative evidence:

  • Strong: use of comparison group/experimental or quasi-experimental design; detailed description of research procedures for recruitment, sampling, and bias; statistical significance level set at p < .05; use of validated measures; controlling/addressing for relevant confounds
  • Moderate: statistical test of pre/post change, without comparison or control group; statistical significance set at p <.1; incomplete description of methods, sampling, or instruments.
  • Weaker: did not use statistical analyses or tests of association, minimal or no detail on research procedures or methods.
• Qualitative and observational evidence:

  • Strong: themes reflected across several participants; rigorous qualitative methods described, participant quotations provided to illustrate themes, researcher positionality defined
  • Moderate: limited detail regarding method; paraphrasing of participant narratives; few direct quotations; less robust analysis
  • Weaker: minimal or no detail on method or analysis procedures; no direct quotations

Search Results and Studies

Our search yielded studies offering two broad categories of findings on CMM: Type 1) quantitative evidence of CMM’s impact on youth and/or mentors; and Type 2) qualitative or observational evidence of CMM’s impact on youth and/or mentors. Table 2 summarizes the empirical evaluations included in this review. Table 3 offers a detailed overview of methodology and findings for each study.

Table 2:

Summary of empirical studies evaluating CMM

CMM Program Studied	Citation	Type 1: Quantitative Impact Measures	Type 2: Qualitative/Observational Impact Measures
AIM	Cramer et al., 201828	Moderate evidence	Moderate evidence
Arches	Lynch et al., 201823	Strong evidence	Moderate evidence
Credible Messenger Institute	Lopez-Humphreys & Teater, 202024	Moderate evidence	N/A
	Lopez-Humphreys & Teater, 201931	N/A	Strong evidence
Peacemaker Fellowship	Corburn & Fukutome-Lopez, 202026	Weaker evidence	N/A
	Wolf et al., 201525	Weaker evidence	Moderate evidence
Healthy, Wealthy & Wise	Remington, 202030	N/A	Strong evidence
	Gonzalez et al., 202027	Weaker evidence	Moderate evidence
No model specified	Fader et al., 202240	N/A	Strong evidence

Table 3:

Analysis of Empirical Studies of Credible Messenger Mentoring

Study	Sample	Design	Outcomes	Results	Evaluator	Evidence Assessment
Advocate, Intervene, Mentor (AIM): program for youth assessed as high risk, who are sentenced to probation and would otherwise be placed in a residential facility, are facing a violation of probation, or were rearrested for certain felonies. Youth work individually with credible messenger mentors 7–30 hours per week for 6–9 months. Mentors help youth create and work on goal plans, link youth to services, and partner with probation officers to provide case management and referrals. AIM also involves regular meetings with youth, families, probation, and mentors.
Study 1: 28 Location: NYC	• n=229 youth at 5 sites • Ages 13–18, m=15 • 77% male • 81% African American • 16.2% Hispanic/Latino Interviews: • 20 youth • 9 parents • 17 mentor • 7 staff • 34 stake-holders	• Analysis of program data from launch (2012) to time of evaluation (2016) • Within group change (statistical test not described) • Interviews • Focus groups • Case file reviews • Court records • Historical comparison group of youth released from placement	• Youth, family, staff and court stakeholder perceptions of program • Program completion • Goal attainment • Re-arrest • Re-conviction (general, felony) at 6, 12 months post enrollment • Out of home placement • Demographic differences	Quantitative Results • 2–10 hours/week with mentors • 61% program completion • Most progressed or achieved goals • 67% not resentenced to placement (80% excluding technical violations.) • Recidivism varied by court, charge, and timepoint: 0–25.8% • Significant results:  ∘ Family court adjudication: general at 6, 12 months, felony at 12 months (p < 0.1)  ∘ Criminal court felony re-conviction 12 months (p < .05) Qualitative Results: • Shared experiences enabled connections • Youth comfortable opening up with mentors • Parents found mentors supportive and responsive • Concerns with capacity, mentor matching, program length, working with probation, and engaging younger youth	Independent evaluator at the Urban Institute Peer review: No	Quantitative: Moderate Findings that youth are on a promising track at the end of the program, with progress towards goals, significant differences in some measures of recidivism at varying levels, and more than half of youth not resentenced to placement. The report provides detailed description of data procedures, and conducts tests of association for some measures. The authors did attempt to form a historical comparison group, however there were major differences between samples that likely confound results. Findings do not provide a basis for causal inference. Qualitative: Moderate Interview findings provide insight into some positive impacts of the fellowship, as perceived by youth, family, mentors, and other stakeholders. Some direct quotes used to illustrate themes, moderate detail provided on methods and analysis. Results provide a more varied picture of participants’ views of CMM’s impact.
Arches: 6–12-month program for youth sentenced to probation in the community, with high risk of re-offending. Program components: individualized support, and group sessions held twice per week facilitated by credible messenger mentors, where youth complete a cognitive behavioral therapy journaling intervention. Mentors work in partnership with probation officers to provide case management and referrals. Youth receive stipends for participation.
Study 1: 23 Location: NYC	• n=943 youth, 8 sites (273 in Arches, 670 probation as usual) • Observed 61 youth, 24 mentors • Demographics (Arches youth) • Ages 16–24, m=19 • 86.5% male • 75.6% African-American, 17.9% Hispanic/Latino • 73.5% less than high school degree • 20.8% employed at intake • 22.9% assessed high risk at intake	• Quasi-experimental design • Groups formed by zip code, propensity score matching • Paired t-tests • 8 sites randomly selected. • For Arches programs only:  ∘ Focus groups  ∘ interviews  ∘ Observation  ∘ Pre/post surveys	• Rates of re-arrest and re-conviction overall, and felony at 12, 24 months post-program. • Demographic differences in recidivism • Changes in youth behavior • Youth, mentor, stakeholder perceptions of program • Youth engagement • Mentor facilitation	Quantitative results: • Retention varied by site 0–100% (most ~30–60%) and attendance ~25–50% • Recidivism varied by court, charge, and timepoint: 2–74% • Significant differences between Arches and comparison group youth in some indicators, but only for youth under 17.  • Felony re-conviction (youth under 17) at 12 (p < 0.01), 24 (p <0.001) months  • Some differences in arrest (felony and non-felony), and reconviction (non-felony) at 12 and 24 months for older youth (p < 0.05)  • Change in positive development and psychosocial skills (p < 0.01) • Qualitative results: • Youth felt supported by and close to mentors • Journaling curriculum not engaging • Mentors need more support • Program too short • Program capacity limited • Group engagement and model fidelity varied	Independent evaluator at the Urban Institute Peer-review: No	Quantitative: Strong: Findings that Arches was effective in reducing some indicators of recidivism compared to probation as usual, especially for younger youth. Arches youth also improved in indicators of positive youth development and psychosocial skills. The study uses rigorous methods and a strong design, with validated measures and comparison groups. Tests of association were significant below the threshold of p<.05. Results provide a strong basis for causal inference. Qualitative: Moderate Findings demonstrate some positive impacts of the program, as perceived by youth, mentors, other stakeholders, and in observations. Direct quotes are used to illustrate themes, and moderate detail is provided on methods and analysis. Results provide a more varied picture of participants’ views of CMM’s impact.
Credible Messenger Institute: 6-week intensive training program for credible messenger mentors. Training focuses on facilitating cognitive-behavioral therapy, mentoring skills, and interpersonal communication. Training also aims to support mentors in their own healing, such as providing tools for self-care, self-awareness, reflecting on trauma and how it may influence mentoring, revisiting own stories of struggle and resilience, and practicing emotion regulation.
Study 2: 24 Location: NYC	• n=35 adult mentors • Ages 21–61+, m=40 • 77% Black/African American • 20% Hispanic/Latino/a/x • 71% Male • Time incarcerated: < 1–36 years, m= 15.4 • Time since release: 1–12 years, m=5.2	Pilot study evaluating a 6-week CMM training. Pre-post t-test of change after training.	• Self-esteem • Self-awareness • Professional boundaries • Self-narratives • Hope • Experience with training	• Statistically significant increases in self-esteem, hope, self-awareness, and self-knowledge after training (p <0.1). • No significant increases in professional boundaries or self-narratives. • Mentors reported strong satisfaction with the training, and intentions to apply what was learned in practice.	University based academics Peer-review: Yes	Quantitative: Moderate Findings that training is associated with increases in some psychosocial indicators for mentors. Detailed description of methods and analysis, used validated measures, and conducted tests of significance, at the (p<.1) threshold. However, the study did not use comparison groups or calculate effect sizes. Findings do not demonstrate causality. Results also focus on CMM training, and may not be generalizable to the mentoring context.
Study 1: 31 Location: NYC	• n=11 adult mentors • Ages 29–63 • All formerly incarcerated, 2–25+ years • Time since release: 2–11 years • 64% Black/African American • 27% Hispanic/Latino/a/x • 82% male	Focus groups conducted 1 year after completing the institute training	• Mentors hopes for training • Reasons for participating • Experience with training • Main takeaways • Impact of training on practice	• Training was meaningful for mentors, and created a safe space to be vulnerable, build community, receive/give mutual mentorship, continue healing, practice self-care and self-forgiveness, recognize and affirm their own value. • Mentors felt seen/validated by peers • Training promoted mentor growth (such as seeing self as more than their past)	University-based academics, conducted member-checked of findings with participants Peer-review: Yes	Qualitative: Strong Compelling findings on mentors’ positive experiences of joining the CMM community and receiving training. To a lesser extent, provides insights into the ways benefits for mentors passed down to youth they work with. Methodology is detailed and rigorous, researcher defines their positionality and uses a member checking process, and rich example quotes are provided to support findings.
Peacemaker Fellowship: 18-month program where cohorts of fellows work with teams of credible messenger mentors. Fellows are individuals identified as active firearm offenders in the community but do not currently have prosecutable cases. Participation is not mandated, fellows are encouraged to join. Program components include daily mentor contact, intergenerational mentoring groups, developing life maps (individualized goal plans), case management, internships, travel and excursions, and linkage to services, and mentors work with fellows’ support networks. Fellows receive stipends for participating.
Study 1: 26 Location: Sacramento, CA	• n=50 fellows • Ages 14–25 m=23 • 98% male • 96% African American • 65% previously incarcerated or arrested • 12% offered social services prior to enrolling • 84% unemployed • 84% victim of gun violence	Percentage of fellows meeting each outcome during and after program completion. Uses internal program data.	• Program completion • Life map completion • Program participation • New arrests (general and gun-related) • Gun victimization • Paid work/internship/ • Service referrals • Mortality • Mentor contacts	• 64% program completion • 90% no new gun charges • 44% no new arrests • 2% (1 fellow) shot or killed by firearms during the program • 100% received social services • 25% AVG goal completion • 26% began paid internship • 38% obtained work • AVG services provided (per fellow):  • 3.7 referrals, 7.7 hours of referral advocacy  • 31 mentor contacts  • 50 hours face-to-face mentor contact	Independent evaluator at UC Berkeley Institute of Urban and Regional Development Peer-review: No	Quantitative: weaker Findings that the program helps connect many fellows to services and opportunities, make progress on their goals, and that many fellows are on a promising track after completing the program. However, the report provides minimal information on methods, and did not conduct statistical analyses or tests of association, or use a comparison group; findings do not demonstrate causality.
Study 2: 25Location: Richmond, CA	• n=68 fellows, 14 with interviews • Age: 14–25 • 97% African American • 45% are fathers • 21% were prior victims of gun violence	Process evaluation of organization from launch (2010) to evaluation (2015), select findings on the fellowship.• Interviews • Program data	• Perceptions of program • Experience in program • Exposure to violence • Arrest rates • Access/referral to services • Mortality • Educational and vocational achievements	Quantitative Results • Services completed:  ∘ 100% life maps  ∘ 83% life skills training  ∘ 77% anger management  ∘ 77% financial management  ∘ 61% employment  ∘ 14–46% accessed other social services: • 84% no new gun injury • 79% no new gun arrests • 94% (64/68) are alive • 20% received GED • 10% enrolled in college/vocational program 50% became employed Qualitative results: • Fellows felt understood and supported by mentors, program felt like a family • Program improved world view, showed possibilities for change, helped learn skills to feel safer in community	Independent evaluator at the National Council on Crime and Delinquency Peer-review: No	Quantitative: weaker Findings that the program helps connect many fellows to services and opportunities, make progress on their goals, and many fellows are on a promising track after the program. Some description of methods and procedures, but no statistical analyses or tests of association, or use of a comparison group; findings do not demonstrate causality. Qualitative: moderate Interview findings provide rich insight into the positive impact of the fellowship for participants. Limited detail is provided on interview methods or analysis, but direct quotes are used to illustrate themes.
Healthy, Wealthy & Wise (HWW) intensive, 14-week mentoring program led by credible messenger mentors for youth and young adults considered high risk for juvenile system involvement, or may be formerly involved. Weekly sessions held for 1.5–2 hours, and curriculum addresses topics such as decision-making, identity, overcoming pain and trauma, life skills and financial literacy. The program also draws on aspects of cognitive behavioral therapy, and uses journaling techniques to promote self-reflection, goal setting, and changes to thinking. Participants receive stipends for participating.
Study 1: 30 Location: Oakland, CA	• n=5 youth (3/5 male, 4/5 Black, 1/5 Latinx) • Ages 18–35 • All assessed as high risk (involved or affiliated with gangs, JLS, formerly incarcerated, victim/perpetrator of gun violence) • n=15 staff (12/15 male, 11/15 Black, 2/15 Latinx) • ages 37–54	Larger organizational case study, with select findings on HWW program impact • Interviews • Observation • Document analysis	• Process of transformative mentoring • Mentor perceptions of their roles • Youth perceptions of mentors • Mentor’s process of engaging youth	• Mentors support youth in navigating hard decisions, staying accountable to goals, modeling positive change, advocating/connecting to resources, changing mindset. • Mentors develop deep understanding of clients, draw on credibility to build relationships, provide individualized support. • Mentors foster belonging, engage families, and strengthen youth support networks • Mentors/staff provide meaningful support to each other.	Doctoral dissertation Peer reviewed: no	Strong qualitative/observational evidence: Findings provide compelling insights into the positive experience and impact of CMM for youth, mentors, and their larger community. Methodology is detailed and rigorous, researcher defines their positionality and process in depth with rich example quotes, and supporting observations.
Study 2: 27 Location: Oakland CA	• n= 37 youth • 67% African American • 12% Hispanic/Latino* • 72% Male* • 44% ages 19–24* • 100% history/immediate risk of gun violence • 97% close relation harmed by gun violence • 82% on probation or parole for violent incident • 39% shot or seriously injured due to turf/group violence • *Demographics reflect full organization	Process and outcome evaluation of full organization, select findings relevant to HWW. • Interviews • Document review • Program data • Survey data collection • Participant outcomes from 12 months before program start to 12 months after program end.	• Arrests for any offense, gun offense, and violent offense • Probation sentence • Conviction for any offense • Violent victimization • Resilience	Quantitative resultsDecrease in arrests (from 16 to 5 participants) • 49% program completion. Some re-joined or completed multiple times • Decrease gun offenses (11 to 0 participants), violent offenses (8 to 0 participants), placement on probation or convicted of a crime (8 to 0 participants) • Varied rates of violent victimization for youth completing different parts of the program youth • Minimal change on resilience surveys Qualitative results: • Participants felt program changed mindset, built life skills, fostered a sense of pride. • Participants felt mentors could relate to them.	Independent evaluator at Mathematica Peer review: no	Quantitative: weaker Findings that mentees are on a promising track at the end of the program, with declines in indicators of recidivism and offending behavior, and varied rates of violent victimization. However, the study does not utilize comparison groups nor conduct tests of statistical significance, has a relatively small sample size of youth participating in the target program, and is missing some detail on methods and analysis procedures; findings do not demonstrate causality. Qualitative: Moderate: Interview findings provide rich insight into the positive impact of the program for participants. Limited detail is provided on interview methods or analysis, but direct quotes are used to illustrate themes.
Impact on mentors (non-program specific)
Study 1 40 Location: Philadelphia suburb	• n=45, results focused on the 28 men of color • 20/28 reported some form of youth mentorship role • 24/28 had a criminal history or record, or self-reported past criminal activity	Interview study on experiences of racialization, criminalization, and finding redemption, with select findings relevant to CMM.	• How men of color navigate racial-criminal stigma, masculinity and redemption	• Role offers a chance to teach others from mistakes. • Mentoring promoted upward economic and social mobility • Mentoring helped develop positive self- narratives and change life trajectories. • Mentoring helped build community standing and leave a positive legacy	University-based researchers Peer-review: Yes	Qualitative: Strong Findings provide a rich example of the benefits of mentoring for men who have been racialized and criminalized. Study provides a detailed description of rigorous methodology and analysis, with numerous quotes to illustrate findings. This study was the least explicit in language and focus on CMM, but the work described was consistent with the criteria outlined earlier.

Overall, the search yielded nine studies evaluating the impact of CMM on youth, mentors, or communities. Independent evaluators conducted most of these studies and examined specific adaptations of CMM: Peacemaker Fellowship, Arches, AIM, Healthy, Wealthy & Wise (HWW), and the Credible Messenger Institute (see Table 3 for a description of each adaptation). One study examined the impact of participating in CMM more generally, outside of a specific program. Each adaptation of CMM had one or two distinct studies assessing some aspect of its impacts. The studies evaluated CMM in New York City, in three different cities in California, and in Philadelphia, predominantly in neighborhoods with high rates of crime and poverty.

The programs examined served youth and young adults ages 13 to 25 (with one program working with young adults up to age 35) who were mandated to complete the program as part of probation or as an alternative to incarceration, as well as those who were not formally involved in the JLS or criminal legal system at the time of participation but were previously involved, or were considered high risk for system involvement. Mentors in the studies had currently or previously worked with youth, or were in training to do so. In studies that reported mentor characteristics, mentors were ages 21–63, were formerly incarcerated or had a history of criminal activity, and had been out of prison for varying lengths of time. Both mentors and mentees predominantly identified as people of color.

Research designs varied across the research reports. Most of the studies followed participants throughout their involvement in a specific program (and sometimes after), and used mixed methods (e.g., some combination of interviews, focus groups, observations, document analysis, and/or analysis of law enforcement records or internal program data). Most studies used convenience samples of participants who were connected to some type of CMM program during the study period, although some studies included data from prior program participants. Only one study²³ used a quasi-experimental design, where participants’ zip codes were used to form the treatment and comparison groups. Sample sizes varied, with some studies focused on in-depth reflections from a few participants, and other studies including hundreds of youth.

Findings

Quantitative evidence on the impact of CMM (Type 1).

The most commonly assessed outcomes were measures of recidivism, reporting indicators of youth participants’ delinquent or criminal activity, or future JLS involvement at varying points in time from when they started or ended the program. Other common outcomes assessed included progress on individualized goal plans, program participation and dosage, and linkages to other services and opportunities (e.g., employment, counseling, educational programs). Some studies also assessed changes in psychosocial indicators (e.g., self-esteem, social support, emotion-regulation) for youth or mentors. Participation in CMM programs was associated with varying degrees of improvement in psychosocial skills and achievement of individual goals, though findings were largely correlational. Only one study collected quantitative data on mentors,²⁴ finding small increases in some psychosocial indicators following completion of a CMM training program.

The strongest quantitative support came from the one study with a comparison group.²³ This study found statistically significant effects for reduced recidivism, where the rate of felony reconvictions was over two times higher in the group receiving probation as usual compared to the group participating in a CMM program within 12 (p < 0.01) and 24 (p < 0.001) months of completing the program, especially for younger participants (under age 17 at the start of the program). However, other indicators of recidivism in this study, including felony and non-felony arrest, and non-felony reconviction at 12 and 24 months were not significantly different between the groups. This study also found statistically significant improvements from program start to end in psychosocial indicators of positive youth development (p < 0.01), and cognitive-behavioral skills targeted by a journaling curriculum (p < 0.01). However, the study did not include effect sizes, which limits ability to interpret the scale of psychosocial changes in practice.

There were a limited number of quantitative findings speaking to the impact of CMM on health outcomes. A few studies tracked participants’ experiences of violent victimization, gun violence/injury, and death, with varying results and weaker evidence. Studies of the Peacemaker Fellowship found that 84% of youth participants at the Richmond, California site did not report new gun-related injuries during their time in the program.²⁵ At the Sacramento site, one of the 64 fellows participating died due to gun violence during the 18-month program.²⁶ Additionally, an evaluation of HWW found that the rate of violent victimization amongst participants before and after the program dropped from 35% to 10% amongst youth who received mentoring only, however for youth receiving the full curriculum the rate of violent victimization actually increased slightly, from 5% to 8%.²⁷

Qualitative and observational evidence on the impact of CMM (Type 2).

Qualitative studies provided stronger evidence speaking to the impact of CMM for mentees and mentors. Studies varied in the extent of detail provided regarding methods and analytic approaches, but almost all studies provided rich quotes that clearly demonstrated participants’ positive experiences and views of the program.

Across essentially all qualitative and observational findings, mentees described or displayed close, supportive relationships with their mentors. For example, mentees described feeling like they could relate to their mentors, and like mentors understood their experiences. Multiple studies found that forming these strong relationships and sense of community in CMM programs created a supportive and trusting environment for young people’s growth, which is in line with the program’s theory of change. One participant is quoted describing, “It’s not just a class, it’s a support group, it’s grown to a family…a group of people who help you to be a better version of yourself. This is what HWW does” (p.18).²⁷ Studies asking youth about their experience in the program indicated that to varying degrees, mentors helped shift their world view, show them a different kind of lifestyle, and learn key skills to change behavior. Youth felt supported in making these changes. For instance, one Peacemaker Fellow was quoted as saying: “I’ve seen the path I was on. [The CMM Program] pulled me from a lot of things. They saved my life. They are committed to me even when I am not” (p.15).²⁵ Similarly, a participant from another program described how the “goal is to get [participants] focused; [mentors] don’t want us to take the same path that they went through. Every time, the conversation is to [put] yourself in a different situation, to change your mindset” (p. 16).²³ There were also some challenges reflected in qualitative studies, such as engagement and mentor matching. For example, staff in AIM described challenges engaging younger youth or not having enough time to build relationships,²⁸ and many participants in Arches expressed concerns with the fit of the curriculum.²³

Regarding mentors’ experiences in CMM programs, multiple studies found that mentors were successful in connecting with youth and supporting them in navigating structural barriers. Mentors shared that CMM offered them an opportunity to make amends with their complicated past by “paying it forward” to youth facing similar circumstances, to continue their own personal growth and healing, and to improve their economic and social mobility. For example, a mentor from HWW stated, “I wanted to find a way that I could give back to my community, ‘cause I had took so much from it” (p.110).³⁰ Program observations, such as those conducted for HWW, also describe the positive impact of belonging in the CMM community, and providing as well as receiving support from other mentors. These aspects of CMM helped mentors grow in their own role and capacity to support youth.³⁰ Mentors reported that feeling validated by their peers in the safe space of the CMM community helped increasingly recognize their own worth, hope, and self-esteem. For instance, one Arches mentor shared of the experience of mentoring and participating in training: “It made me feel like I’ve got more, more to give these young men and women other than my story” (p. 149).³¹ These findings demonstrate mentors’ views of the programs benefits for their own well-being.

Discussion

This narrative review investigated the range of evidence concerning the impact of CMM for youth involved in the JLS. We found weaker to moderate quantitative evidence, and moderate to strong qualitative and observational evidence on the impact of CMM on youth and mentors. Regarding the quantitative findings, most results found that participation was associated with some degree of reduced recidivism, increased connection to services and opportunities, and growth in some kinds of psychosocial indicators. However, the influence of CMM on health or violent victimization was not clear, and there is very limited quantitative evidence examining the program’s impact on mentors. Moreover, only one study²³ used a research design with an appropriate comparison group, and many. Without the use of comparison groups, it is not clear from the current state of the evidence that CMM program participation itself (rather than some other factor) actually causes these positive changes, or how outcomes compare to other models or programs offering general mentoring or other interventions. Additionally, many studies discussed challenges with attrition, and it may be that those who completed CMM programs are more motivated to make major life changes, so the results may reflect characteristics of the participants, rather than the effects of the program itself. Finally, a number of quantitative results were presented as standalone percentages without corresponding tests of statistical significance, which limits the ability to determine whether findings were due to chance. In sum, there is some emerging evidence on quantitative outcomes of CMM for youth involved in the JLS, but at present there is not sufficient research to make causal claims about the impacts of CMM.

The qualitative and observational evidence on the model is more compelling, offering rich insight into the transformative effects of CMM. Qualitative findings support the model’s theory of change, including the value that community members attribute to lived experience for building trust, supporting youth through navigating complex structural barriers, fostering hope, and modeling a path for meaningful change. The qualitative evidence collected from youth reiterates this impact, with descriptions of how program participation changed their thinking, behavior, outlook, and for some, their larger life trajectory. Likewise, mentors shared how their experience as credible messenger mentors promoted their own growth, healing, and sense of purpose. It should be noted that some studies were more robust than others, and a few did point to some challenges and limitations of current implementations, especially related to engaging youth, program capacity, and expanding support for mentors. Though studies provided varying levels of detail regarding methods, analysis, and researcher reflexivity, all qualitative evidence included direct quotes to illustrate claims through participants’ own words. There were also multiple studies using rigorous qualitative and observational methods to substantiate these findings. In sum, our review found moderate to strong qualitative and observational evidence demonstrating that those who have been involved in the CMM model consider it an effective approach for promoting well-being in their communities.

The breadth of settings, adaptations, and methods in the current body of research also offers strengths and limitations for understanding the CMM evidence base. First, most studies evaluated youth and adults who identified as people of color and with past or current involvement in the carceral system, or who were considered at high-risk for carceral involvement, which provides a basis for claims that the model is effective with the population it is designed to serve. However, it should be noted that no evaluations were conducted in custodial facilities, and that the large majority of participants identified as male, which limits the ability to understand the impact of CMM in some JLS settings and with female-identified and gender expansive youth and mentors.

Similarly, the range of study methods has strengths and limitations. On the one hand, the range of approaches used provides a nuanced and holistic understanding of the program’s impact. The positive findings across regions, service settings, and with different program adaptations also point to its diverse potential. On the other hand, with such divergent methods and contexts, it is difficult to compare findings and assess the larger effects of CMM in a generalizable way. For instance, it is not clear how the model’s impact differs if mentoring is provided in a group or individual setting, if mentees are compensated, or if cognitive behavioral therapy is included. As another example, the curriculum and structure of adaptations differ enough that we cannot say how they might translate in another region or JLS setting. Overall, this review found that there is some evidence to support the model’s flexibility in a range of contexts and populations, but more limited evidence demonstrating its replicability or transferability to other parts of the country, in some of the highest risk settings, and with other system-impacted populations such as LGBTQ+ and female youth.

One final point is important in describing the state of the research on CMM. Our review found that most studies were conducted in close collaboration with CMM providers, published as technical reports through think tanks or independent evaluators, and with only a few studies published in peer-reviewed journals. The CMM evidence base was generated in closer connection to communities than is typical of intervention studies, which aligns well with the model’s philosophy of community-determined credibility. Such community-driven and validated evidence offers a unique contribution to understanding the impact of interventions, especially when taking seriously the goal of health equity.³²

At the same time, in recognizing the community desire to expand the model, it is important to address the gap in the CMM evidence base in the context of a funding and service landscape that may discredit non-traditional research evidence. Whether out of convention, elitism, or concerns about methodological rigor, major funders and the academy often ignore or under-value such community-driven and validated evidence and instead prioritize studies that use more traditional randomized controlled designs and are published in peer-reviewed journals.³³ However, there are a number of unique ethical and logistical challenges to conducting research with credible messengers,³⁴ which may contribute to the seemingly unconventional CMM evidence base, and pose challenges to continued research on CMM. We do not advocate that researchers force CMM to fit into traditional evaluation paradigms, especially given that CMM’s informal, fluid, and collective approach seems to be part of what makes it so impactful. The promising findings described in this review reiterate the importance of lifting up and prioritizing community-generated research for its strengths alongside other types of evidence. We encourage researchers and funders to engage in creative thinking and innovate evaluation methods to continue expanding the CMM evidence base.

Life Course Health Development as a Potential Theoretical Frame for Credible Messenger Mentoring

The results of this review point to developmental science, and in particular the life course health development (LCHD) framework, as a helpful direction for guiding future research on assessing the impact of CMM. LCHD views health as a developmental process unfolding over the lifespan and across generations, and influenced by multiple biological, environmental, social, and psychological factors.³⁵ Life course health development interventions integrate this complex understanding of health to move beyond the prevention and treatment of illness, and instead shift health trajectories towards longevity and optimal well-being.³⁶ We mapped the results of our narrative review onto the twelve principles of LCHD interventions (see Table 4), and found emerging support for CMM as a life course intervention. The evidence on CMM appears especially promising for the LCHD intervention principles of strategic timing, health equity, and collaborative co-design. For example, in regards to strategic timing, CMM challenges conventional definitions of sensitive periods by demonstrating possibilities for transformative change beyond the bounds of age or developmental stage. There also seems to be some evidence supporting CMM as aligned with other LCHD intervention principles (such as being developmentally and longitudinally focused, multi-level, health optimization focused, and addressing emerging health capabilities), but more research is needed to understand how the model promotes these dimensions of health. Given the strong conceptual fit and emerging empirical support between CMM and LCHD, we recommend future research draw on the LCHD framework and intervention principles to inform research designs and the processes and outcomes measured, such as health concepts.

Table 4:

Assessing CMM as a Life Course Intervention

Life course criteria	CMM Evidence Base
1. Developmentally focused	• Strong evidence that CMM promotes positive development in mentors, and core developmental processes in youth by connecting them with caring, trusted adults and opportunities for growth. • Varied evidence regarding the development appropriateness of curriculums.
2. Longitudinally focused	• Strong evidence for the model’s impact in promoting mentor’s resilience and well-being long-term. • Limited evidence for the model’s long-term impact for youth beyond recidivism.
3. Strategically timed	• Some evidence that CMM is targeted to key turning points (e.g. individuals identified as high-risk, or re-entering the community) but no evidence on CMM in facilities, which may be a critical transition. • CMM redefines “turning points” with strong evidence of transformative shifts for those who are often considered “past change,” and at varying points in individual’s trajectories.
4. Multi-level and holistic	• Strong evidence that CMM helps youth and mentors navigate complex environments, meeting a wide range of needs, and matching social context. • Limited evidence on whether CMM specifically is more holistic compared to other types of interventions.
5. Strengths based	• Strong evidence that CMM leverages mentor and community strengths. • Limited evidence that programs matched individualized needs and strengths.
6. Health Optimization Focused	• Strong evidence that CMM helps optimize positive life trajectories for mentors. • Some evidence that CMM promotes health and well-being for youth, but most indicators focused on problem-oriented outcomes.
7. Health Equity Focused	• Strong evidence that CMM serves and is led by those who often face the greatest health disparities, including people of color, people with histories of carceral system involvement, and people experiencing poverty and community violence.
8. Family centered	• Strong evidence that CMM fosters a sense of family, community, and belonging amongst mentors in particular, and youth in some programs. • Some evidence that CMM engages young people’s larger support networks.
9. Anti-Racist	• Some evidence for CMM explicitly addressing racism in health trajectories. • Strong evidence for CMM promoting meaningful change in communities of color. • Model explicitly seeks to lift up people of color and people with experiences of structural racism and institutional violence as leaders.
10. Horizontally, Vertically, And Longitudinally Integrated	• Some evidence for effective collaborations between CMM and other stakeholders to promote multi-dimensional change across institutions and contexts.
11. Collaboratively Codesigned	• Strong evidence for community leadership and support for CMM.
12. Addresses Emerging Health Development Capabilities	• Strong evidence for CMM promoting mentors’ capabilities and supporting their continued development and transitions through adulthood. • Some evidence for CMM promoting changes in youth thinking and behavior, achieving goals, and connecting to opportunities that may promote their later success and adaptability.

Future directions

The results of this narrative review reinforce the potential positive life course effects of CMM, as well as the importance of further expanding the evidence base to clarify how the model promotes health and well-being of youth impacted by the JLS and their communities. In addition to applying LCHD to inform research, there are a number of directions that can further strengthen the CMM evidence base. First, future research using comparison or control groups will help demonstrate how the impact of CMM differs from other mentoring models or programs, and investigate causal associations between CMM and intended outcomes. Quasi-experimental designs may be promising, in addition to historical (but more closely matched) comparison groups or other approaches such as time series methods. Another pressing direction is to expand assessment in other geographic areas and JLS settings, especially custodial facilities, where health risks are particularly exacerbated and young people may be the most disengaged and hard to reach. Finally, we encourage future research to incorporate a wider range of outcome measures beyond recidivism, especially markers of health and positive development. The recidivism measures used in most of the existing studies on CMM generally do not capture the structural, environmental, and relational influences that shape youth behavior, nor do they demonstrate the complex processes of changing thinking, behavior, and ultimately life trajectories.³⁷ CMM is specifically designed to promote this kind of transformative, long-term change and to support youth in navigating structural barriers to health and well-being. More holistic measures of CMM’s effects over time will likely better demonstrate the models’ impact. Ultimately, these directions can help to strengthen the empirical grounding of CMM to match the growing interest and expansion of the model.

Limitations

This review has a number of limitations. First, this review is narrative, not systematic. Due to the small and varied nature of the evidence base, we did not believe that a formal systematic review would be feasible until a more robust body of research is developed. However, a systematic review uses more rigorous and established criteria and would likely be valuable once more studies on CMM are available. It is also important to acknowledge that this review was conducted by academic researchers. While all of the authors have varying degrees of prior or current practice experience in the juvenile and adult legal systems and with CMM organizations, we are not ourselves credible messengers. Though we aimed to prioritize the full body of evidence on CMM, and hope that our review can help advance CMM, our analysis and recommendations are influenced by our position in the academy, and may not resonate with the CMM community.

Conclusion

This narrative review summarizes an emerging body of research on credible messenger mentoring, and underscores the need to continue understanding how this intervention may promote health and wellbeing. CMM appears to be a promising approach to promote life course health development, led by community members from the ground up; however, more research is needed to examine health equity and pro-social development. As CMM continues to expand, participatory research will be particularly important to engage youth, mentors, and other youth justice stakeholders in expanding what is known about the approach. There is much potential for CMM to grow as an evidence-based intervention for youth and adults impacted by the carceral system, both as mentees and mentors. Further community-engaged research can advance CMM’s growth and extend its usage into other communities to advance health equity for young people involved in the JLS.