Peer-focused prison reentry programs: Which peer characteristics matter most?

Over the past two decades, recidivism rates have remained relatively stable, leading practitioners to explore innovative reentry solutions. One reentry model, based on the concept of peer mentorship, has received renewed attention. Unfortunately, little is known about which peer characteristics make mentors most effective in a prison setting. This study uses participant observation and semi-structured interviews, embedded with survey questions, to understand which “peer” characteristics prison staff, peer mentors, and mentees perceive as the most important. Analysis of survey data also suggests that a history of incarceration is perceived as the most important characteristic for peer mentors in a reentry context. Additionally, the qualitative analysis reveals that mentors need to be perceived as credible to be effective role models for reentry. This credibility was almost exclusively linked to a lived experience of incarceration. Peer mentorship remains a viable option for improving reentry outcomes, but hiring the appropriate, credible peers is essential for effective implementation.


Introduction
An American trend toward mass incarceration creates a new problem of mass reentry. Every year approximately 600,000 people are released from prison and expected to successfully reintegrate (Carson, 2020). In the United States, people returning to the community often 1 Gonzaga University, USA receive a list of obligations when they check in with probation and parole, but little support or guidance regarding how to meet them (Pleggenkuhle, 2018). For those who lack adequate support, the future is bleak. Over 80% of people released from state prisons in the United States will be rearrested within 9 years (Alper et al., 2018). Adequate support upon release is vital to successful reintegration and desistance (Brown and Ross, 2010), but who is best suited to provide the support?
Organizations tasked with rehabilitation in various contexts have demonstrated that people with similar lived experiences may be best positioned to help individuals set aside dangerous or unhealthy lifestyles (Davidson et al., 2006;Lopez-Aguado, 2013). Mead and MacNeil (2006) explain that peer mentors offer empathy and validation, in addition to offering practical advice and strategies for overcoming shared challenges. Research of violence prevention programs reveals that people involved in gangs respond positively to advice from former gang members who opted for a different life (Cheng, 2017;Lopez-Aguado, 2013). Additionally, the use of peer mentorship in substance abuse programs is equally promising (Bassuk et al., 2016;Tracy and Wallace, 2016). However, using mentors with histories of justice involvement as reentry service providers is somewhat controversial, as traditional criminal justice policy generally prohibits people returning to society from associating with individuals who have a criminal record (Buck, 2020;Davidson and Rowe, 2008;Hinde and White, 2019).
Despite these concerns, the model of peer mentorship is now being tested in criminal justice settings, but research into the efficacy of this model is far from conclusive, largely because there is little consistency between interventions. A lack of consistency in peer-delivered services is not exclusive to prison settings. Bassuk and colleagues (2016) conducted a systematic review of the literature on the efficacy of peer support in addiction and recovery settings and concluded that while the evidence suggested positive benefits for those in recovery, more concrete and generalizable conclusions were difficult because interventions lacked consistent definitions regarding who was considered a peer, what the peer mentor's role should be, and what the specific goals the interventions should achieve. In the area of mental health, Davidson and colleagues (2006) also highlight the importance of fully understanding peer support in mental health settings, because it directly relates to the efficacy of any peer-delivered services. The authors contend that the concept of peerness is complex and not widely investigated, which leads to inconsistency in intervention development, peer roles, and goals. Similarly, little is understood about how peer mentors can be best utilized in prison reentry programs, as the definition of a peer and the ultimate goals of the intervention are inconsistent, as in other settings. Buck (2020) provided the most comprehensive overview of how peer mentorship is best utilized in criminal justice settings; however, the investigation was limited to community-based interventions in the United Kingdom.
Hence, the current study seeks to fill in some of these gaps using an exploratory sequential mixed methods design to investigate the use of peer mentorship in several reentry programs at two prisons in the Northeastern United States. Each prison has multiple programs, with various formulations of the peer mentor model, including variable definitions of a peer mentor. This research aims to investigate which characteristics mentees, mentors, and prison staff believe are most important in a peer mentor, and why those characteristics are most important. This article first presents a brief overview of what is currently known about the benefits and definitions of peer mentors in criminal justice settings. Next, the study site and research design are outlined. Then the survey results and qualitative findings are presented. The article closes with a discussion regarding the implications of the integrated research findings.

Peer mentors in carceral settings
The promise of peer mentorship Despite a lack of consistency in the peer-mentorship model, research suggests that peer-focused reentry services appear to be beneficial in a variety of ways (Barrenger et al., 2017;Brown and Ross, 2010;Buck, 2020;Hinde and White, 2019;LeBel, 2007;Matthews et al., 2020;Marlow et al., 2015). First, research suggests that peers in criminal justice settings derive a variety of emotional and psychological benefits, similar to other settings (Buck, 2020). For instance, Marlow and colleagues (2015) found that peer mentors helped mentees evaluate their choices more effectively and provided guidance about appropriate interpersonal skills and ways of coping, ultimately giving the mentees confidence they could achieve stability. Brown and Ross (2010) interviewed mentors and mentees involved in a peer support program for women returning to the community in New Zealand. Like Marlow and colleagues (2015), the authors found several psychological and emotional benefits, including increased confidence and peace of mind. Matthews and colleagues (2020) investigated an employment program for people returning to the community and found that peer mentorship from formerly incarcerated staff not only improved self-confidence, but promoted self-empowerment. Similarly, Koschmann and Peterson (2013) reported that self-empowerment and self-sufficiency were common benefits of the peer mentorship model.
In addition to the emotional and psychological benefits derived from having a peer mentor, research suggests that peer support also provides a role model, regardless of the setting (Barrenger et al., 2017;Buck, 2020;LeBel et al., 2015;Marlow et al., 2015;Matthews et al., 2020;Nixon, 2019;Portillo et al., 2017). Barrenger and colleagues (2017) reported that mentees viewed peer specialists as role models, which allowed mentors to help clients transition out of stigmatized identities, while also providing support and connection to service providers. Peer providers used their own personal experience and peer training, in tandem, to best meet the unique needs of clients. Buck (2020) noted that peer mentors relied heavily on their lived experience of justice involvement to model how mentees could manage the stigma and systemic barriers associated with a criminal history. According to Mead and MacNeil (2006), a peer mentor's personal experience with the negative aspects of a specific lived experience was central to the peer relationship. Mentors with a history of incarceration who have been denied employment or housing can provide advice regarding how they managed that negative experience, but they can also share resources that ultimately helped them secure housing and obtain a job, despite the barriers.
Resource connection is one area where role models are particularly effective (Buck, 2020). Marlow and colleagues (2015) found that peer mentors helped mentees returning to the community connect to education, employment, transportation, legal, and housing resources, all while modeling effective coping strategies and interpersonal skills. Research suggests that people returning to society frequently have difficulty asking for help, but mentees interviewed by Matthews and colleagues (2020) reported finding it easier to ask for help from a peer mentor who had successfully transitioned from prison to the community. Because their mentors had successfully navigated the reentry experience, they were knowledgeable about which resources were reliable.
In addition to being role models, or likely because of it, peers can be legitimizers for an organization or resource (Cheng, 2017;Matthews et al., 2020;Portillo et al., 2017). Portillo and colleagues (2017) evaluated a program that provided formerly incarcerated individuals diagnosed with a mental health disorder with a peer who had a similar dual lived experience. The authors found that clients were more willing to trust the organization and engage with services because the people they saw as role models represented the organization. The mere presence of a trusted peer legitimized the organization and the resources it provided. Similarly, Cheng (2017) found that street outreach workers (SOWs) at Bullet-Free Bridgeport had to build relationships in order to achieve any outcomes, and legitimacy was a key component in this process. The lived experience of the SOWs enabled them to establish legitimacy, which allowed them to successfully connect to clients who had never before accessed services.
While many of the benefits of programs are difficult to quantify, a few studies report quantitatively measurable outcomes (Goldstein et al., 2009;Sells et al., 2020;Vigilante et al., 1999). Goldstein and colleagues (2009) found that women who participated in a peer-focused reentry program experienced dramatically lower rates of recidivism, when compared to the state average. Similarly, Vigilante and colleagues (1999) found that women who participated in the Women's HIV Prison Prevention Program had lower recidivism rates, when compared to a control group. Finally, Sells and colleagues (2020) used a randomized controlled trial to compare recidivism rates between individuals who received peer support in addition to traditional reentry services and those who only received traditional reentry services. The results showed that those who were assigned to the peer support condition were significantly less likely to recidivate than those in the control condition. However, the authors noted that peer services needed to focus on ''early intervention, relationship quality, criminal desistance, social navigation, and gainful citizenship,'' suggesting that the specific details of a program can significantly affect outcomes (Sells et al., 2020: 1).
Ultimately, the benefits of effective peer mentorship programs are myriad, ranging from psychological benefits to more tangible, quantifiable benefits. However, the success of peer mentorship is not a given. Much like any other intervention, success depends on understanding the essential elements of peer support, including the definition of a peer, which provides the foundation for what his or her role as a peer mentor should be, in order for the program to have the anticipated effect.

Defining peers
Because peer support is relatively new in criminal justice settings, there is no set definition on ''what makes a peer a peer'' (Clark et al., 2016). Most scholars agree that lived experience is an important part of the peer relationship (Buck, 2020: Davidson andRowe, 2008;Mead and Mac-Neil, 2006;Reingle Gonzalez et al., 2019), however, exactly which components of a person's lived experience matter is still widely undefined. For example, does the fact that someone has a lived experience of overcoming a serious mental illness make them a peer for an incarcerated person with a serious mental illness who is dealing with the dual stigmas of having a diagnosis and a criminal history? Some programs categorize people who have overcome mental health challenges or are in recovery from addiction as forensic peers, even though the peers have no lived experience of incarceration or reentry (Bellamy et al., 2019;Chapman et al., 2018;Reingle Gonzalez et al., 2019). Mead and MacNeil (2006) argue that the important component of lived experience is not actually the shared experience of the condition, but the shared experience of the negative treatment you encounter because of that condition. Hence, peer-focused organizations must determine if having a serious mental illness or a history of addiction and recovery equates to the stigma associated with having a criminal record. Davidson and Rowe (2008) extend this thought. The authors describe people with serious mental illnesses who return to the community from incarceration as having a dual stigma. They define peers as having a lived experience with both stigmas. Davidson and colleagues (2006: 8) quote D. Banks McKenzie in their explanation, saying peers ''fully understand each other's language, thoughts, feelings, sorrows, signs, grips, and passwords.'' Although Buck (2020) reported that the ''ex-offender'' identity was a core component of peerdelivered services, according to both mentors and mentees, this finding was supplemental, not the result of a targeted or comparative investigation.
Recently, Clark and colleagues (2016) investigated the importance of various characteristics specific to peers in forensic settings by asking veterans in a jail diversion program to rank 13 different characteristics. The authors found that gender, age, and race were ranked as the least important characteristics, while service in the military was the most important peer characteristic. A history of incarceration was somewhere in the middle. While this finding is useful, the sample population was very specific and likely not generalizable to people who never served in the military, it highlights the need for more research. A review of the current literature revealed the absence of a study investigating ''peerness'' among the general population of individuals living in prison. Perspectives of prison staff regarding who should be considered a peer were also noticeably absent from the literature. As the use of peer mentorship expands into prison settings, further exploration of the most essential components and correlated benefits of the peer connection in prison-based reentry programs is crucial.

The current study
Given the lack of scholarship regarding what characteristics qualify someone to provide ''peer'' services in criminal justice settings, this study used an exploratory sequential mixed methods approach, guided by two main research questions. First, what are the most important characteristics of a peer, in prison reentry settings, according to relevant stakeholders: mentees, mentors, and prison staff? Second, what benefits do mentees, mentors, and prison staff believe people returning from prison derive from the peer mentorship experience? These two questions are interrelated because certain peer characteristics are presumably more likely to lead to desired benefits than others. An initial round of mentee interviews was coded, and six core characteristics, linked to ''peer'' status, emerged. Once the most salient characteristics were identified, future interviews were embedded with survey questions asking mentees, prison staff, and peer mentors to rank the importance each of the six peer characteristics on a Likert-type scale from Not Important to Very Important.

Research design The study site
This research project is part of a larger study investigating four peer mentorship programs in two prisons in the Northeastern United States. The research was approved by the researcher's University Institutional Review Board (IRB), in January 2019, and a research board with the state's Department of Corrections, in March 2019. An advocate for incarcerated people was included in the review by the IRB, to ensure the ethical treatment and well-being of this vulnerable population. The two prisons were selected because they are classified as reentry prisons and are the only prisons in the state with multiple peer mentorship programs. Most of the state's prisons lack any sort of peer mentorship reentry programming. Two of the peer mentorship programs exclusively utilize mentors who were formerly incarcerated. The other two program uses both incarcerated and formerly incarcerated mentors. Both prisons offer extensive traditional reentry programming to individuals nearing release. Each peer mentorship program includes a multi-session classroom component, as well as one-on-one mentoring, during incarceration and after release. During the classroom component and individual mentoring sessions, peer mentors offer guidance regarding community resources, financial planning, employment, self-care, managing stigma, and practical strategies for overcoming various barriers and reintegrating in their communities. After release, mentees receive individual assistance obtaining employment, housing, health care, and transportation, among other things.

Data and sample
Data sources included ethnographic participant observations and one-on-one semi-structured interviews, embedded with survey questions regarding the importance of several peer characteristics. Observations and interviews began in May of 2019 and continued until March 2020. Approximately 370 h of participant observation occurred in the field. The researcher sat in and participated in the classroom component of each program and observed mentoring sessions, as well as various prison activities while classes or interviews were not taking place. In one prison, the researcher was permitted to bring in a laptop to type interview and field notes, when possible, and not disruptive. When a laptop was not permitted or appropriate, the researcher took detailed notes in a notepad throughout observations and interviews. During frequent downtime, the researcher would fill in notes and include any details that could not be jotted down or typed during observations and interviews. In most cases, before the next interview, the researcher was able to create full sentences for each sentence fragment written down during the interviews. According to established ethnographic practice, at the end of the day, the researcher would transcribe the notes as soon as possible after leaving the research site, to provide the most accurate account possible (Emerson, 2001;Emerson et al., 1995).
Participants were recruited during observations at the prisons, at various times when individuals could be approached without disrupting regular activities. All prison staff with knowledge of the peer mentorship programs were recruited, as well as all of the peer mentors and mentees participating in the prison component of the mentorship programs. All mentors who provided in-prison mentoring agreed to be interviewed. All but two of the prison staff with in-depth knowledge of the prison mentorship programs agreed to be interviewed. Only four mentees did not consent to an interview.
Once participants agreed to be interviewed, they were scheduled to meet with the researcher in a private space, so confidentiality could be assured. Initial interviews lasted approximately 1 h. Follow-up interviews lasted between 15 min and 1 h. The first round of mentee interviews was used to create the survey items later embedded into the second round of interviews, but not all of these participants were available for a second interview. Some of the men living in prison had already been released and a couple of the prison staff had been relocated. Summary statistics for the sample can be viewed in Table 1.

Analysis
Since the study relied on a mixed method research design, both quantitative and qualitative analysis were used. The quantitative data comprised demographic and background characteristics, in addition to the survey results. The data was analyzed using Stata Version 16. The initial phase of interviews with the mentees (n ¼ 51) was coded to reveal salient characteristics. Six characteristics-mentioned by at least ten percent of those interviewed during the initial phase-were identified as the most common. Once the survey items were created, they were added into the interview script. The researcher asked participants to rank the importance of the six characteristics, varying the order the characteristics were presented to participants, to reduce the chance of bias.
Participants were asked to rank the importance of each characteristic on a scale from 1 to 4, where 1 indicated that the characteristic was Not at all important to the peer status and 4 indicated the characteristic was Very Important. The six peer characteristics participants ranked were as follows: (1) a history of incarceration and successful reentry, (2) a history of overcoming trauma, (3) a history of addiction and recovery, (4) the age of the mentor, (5) the race of the mentor, and (6) the gender of the mentor. The first three characteristics are experience-based and the last three pertain to demographic information. The decision to use a general term like trauma was intentional, to encompass a variety of disturbing or distressing experiences, as the mentees often were not specific about what experiences they had survived. The researcher did not provide clarification or ask follow-up questions during the survey portion of the interview, to reduce the chance of bias.
All field notes and interview transcripts were uploaded to the qualitative software program Atlas.ti. In order to create the survey items, the researcher relied on open coding, followed by selective coding (Glaser, 1978). For the final analysis, all interviews were analyzed using a semigrounded theoretical framework, as outlined by Corbin (1990, 1998). As such, a hybrid approach of inductive and deductive analysis was utilized. The author adhered to the three phases of coding suggested by Corbin (1990, 1998): open coding, axial coding, and selective coding. During the open coding phase, the data was first analyzed line-by-line, coding specific words, phrases, concepts, and categories. Next, the author used axial coding to establish relationships and connections between the concepts and categories. During this phase, coded categories were also broken down into new categories, when appropriate. In the final stage of coding, a core category was selected, and all other categories were viewed in relation to that category. Throughout the process, analytic memos were created to document and track the entirety of the analytic process. The author met with another senior researcher, at various stages of analysis to review codes, talk through complex categories, and crosscheck emergent themes.

Results and findings Quantitative results
Frequency analysis, including mean and standard deviation, was used to quantify the survey responses (peer characteristics), by two groups: (1) mentees and (2) prison staff and peer mentors. The survey results for mentees can be seen in Figure 1. The prison staff and peer mentor results are in Figure 2. Regardless of the sample group, the participants indicated that a history of incarceration and a history of overcoming trauma were the most important characteristics for a peer mentor to have; 73% of the staff and peer mentors ranked a history of incarceration as Very Important, whereas 68% of mentees did. A few of the mentees ranked every characteristic a 1, maintaining that someone could be a good mentor regardless of any of the characteristics. A history of trauma was ranked as Very Important by 60% of staff and mentors but only 50% of mentees.
Despite building a rapport with the men, very few mentees discussed a need to heal from trauma, possibly because doing so could be considered a weakness and survival in prison requires the men to hide all signs of weakness (Ricciardelli et al., 2015). The race of the peer mentor was the least important peer characteristic, regardless of the sample group. For mentees who were incarcerated, the mean importance of a history of incarceration was 3.47, and the mean ranking for race of the mentor was 1.53. For mentors and prison staff, the mean importance of a history of incarceration was 3.67 and the importance of race of the mentor was 2.00. Quantitative results indicate that experience-based characteristics, especially a lived experience of incarceration, are more important than any demographic characteristics of the mentor. Additionally, the researcher estimated an ordinary least squares regression to determine whether the importance a participant assigned to a characteristic was correlated with the mentee's background and demographic information. There are not enough prison staff and peer mentors to accurately estimate a regression model, so only the mentee data were regressed. There were six dependent variables: the importance of (1) race of the mentor, (2) gender of the mentor, (3) age of the mentor, (4) a history of incarceration, (5) a history of trauma, and (6) a history of addiction and recovery. There were four independent variables: (1) age of the mentee, (2) race of the mentee, (3) mentee's total time incarcerated, and (4) mentee's age at first incarceration. Race was dummy coded, where 1 indicated the participant was non-White. Regression models for all six dependent variables resulted in null findings. There were no statistically significant differences based on any of the independent variables. The results of all six regression models can be seen in Table 2.

Qualitative findings
In order to understand why various stakeholders-mentees, mentors, and prison staff-believed certain characteristics were most important in a peer mentor, understanding what individuals hoped to gain from the program was pivotal, as that would provide insight into why some characteristics were more valued than others. Hence, the analysis focused on what interviewees described as the program's ultimate goal and why the peer mentorship program could help accomplish that goal. The similarity between the various stakeholder perspectives was apparent, and a clear narrative began to emerge. Three interconnected findings underpin that narrative. First, interviewees hoped peer mentors would serve as role models for reentry, someone the mentees could rely on for advice to help them successfully reintegrate into the community. Second, participants were clear that in order for a person to be a role model-someone the mentees would actually listen to-the individual needed credibility, which was closely connected to relatability. Finally, the interviews helped illuminate why the survey results revealed a history of incarceration as most important overall: it provided a level of credibility, that none of the other single characteristics could. The mentor's lived experience of incarceration made their advice believable and inspired trust. Because the survey questions were embedded in the interview, interviewees frequently explained why they ranked a particular characteristic as Not Important or Very Important. To protect the anonymity of the participants, the data was de-identified. The names used to identify quotes are not the participant's real names. The numbers in parentheses after each quote represent the numerical code assigned to each participant to provide anonymity.

Role models
When the men were talking about their ultimate goal-why they wanted to participate in the peer mentorship program-their fundamental goals were fairly uniform: they wanted lead successful lives and never to come back to prison. When asked why he decided to participate in the program, Carl said, ''it caught my attention because the key thing was helping me try to stay out. I grasped onto that. I wanted to do that . . . '' (27). Additionally, the mentees liked that they shared that goal with their peer mentors. Leroy explained: The main goals of trying to stay out are the same. We deal with the same problems. Because you . . . it's like having another set of eyes and you'll have similar goals. He could advise me to save and such. He could also warn me about stuff. (46) To better understand their primary goal, mentees were also asked how they defined successful reentry. One mentee, James, elaborated, ''me judging myself as successful would obviously be to stay out of prison and be able to become someone who could help others and not come back to jail.'' (51). James indicated that in order for him to stay out of prison, he would have to make changes to become someone who stayed out of prison. Many of the men talked about this concept of change, from a dual perspective. They wanted others to see that they had made a change, and they wanted to see it in themselves. The idea of being a new, changed person, or being perceived as one by others appeared frequently in the data, which aligns with prior scholarship that suggests an identity change is a key component of desistance. (Giordano et al., 2002;Maruna, 2001). Overall, the analysis revealed that mentees wanted to make changes that would enable them to accomplish their primary goal: staying out of prison and leading a successful life. Many of the mentees indicated it was easier to believe they could make changes and avoid coming back to prison when they had evidence it was possible. Angel discussed how his peer mentor would help him stay out: I find Martin motivating. He did real time and he's successful. Cause I know I can do it, if he did it. I haven't had a drink since 2015. What Martin did, I can see me doing the same thing. (58) Several of the mentees mentioned the importance of hearing specific success stories. In one class, the peer facilitator told the men a former mentee buying a [U.S.] $7,000 house at an auction, renovating it, and renting rooms. The mentee repeated the process multiple times. During interviews, mentees often mentioned this example. Pedro explained: I think what I like to hear most are success stories. The guy that got out, saved his money and is doing things. It's new for him . . . I probably could do it myself, however I haven't kept my freedom long enough to be able to show you: this has been me, for this amount of years. (14) The mentees explained that the way they had approached reentry in the past had not worked; they always came back to prison. Hearing specific examples about others-many of whom they remembered-getting out and staying out provided a possible roadmap for success.
The ability to call and ask for their mentor for advice was one of the most valuable components of the peer relationship. Eddy said, the ''gentlemen walked the same streets, made the same mistakes, so they can give you advice on how to get out of that lifestyle'' (56). The advice the mentees sought was not always about how to achieve their mentors' material success, but also how to achieve the cognitive change they saw in the mentors. Harris noted, ''I think it will be helpful for me to pay attention to what were their thoughts, and how did they think positive, so I can start changing my thoughts too'' (30). Similar to community-based peer mentorship programs discussed by Buck (2020), the mentees often rely on mentors to help make critical tangible and cognitive decisions.
Analysis revealed that mentees often looked to mentors for advice regarding one of their biggest obstacles: disconnecting from a violent lifestyle. Mentees frequently mentioned their safety as a primary concern after release. They explained that carrying a gun was mandatory, because of their past lifestyle choices. When asked about his biggest worry, Harold said, ''dying, that's the only thing I'm worried about. A lot of people have to go back to their old neighborhood, and you adjust back to your old beliefs, so you either end up dead or in jail'' (54). The validity of these concerns was evidenced by the fact that a mentee was shot and killed in a shootout at a local supermarket less than a month after he went home. The news affected the entire unit, as most of the men were about to be released, many to the same neighborhood.
Mentee's fears regarding safety were poorly understood by staff, even the most well-liked staff. They did not seem able to understand the dilemma. After the news of the death broke, a prison staff member asked the researcher, ''why was he carrying a gun?'' In contrast, Teron, the peer facilitator of one of the groups who had a substantial street reputation, talked to his mentees about how he balanced his safety, his former associates, and his commitment to never going back to prison, after serving 26 years. The mentees asked him how he protected himself from people he had previously been ''at war'' with on the street. Teron explained that he never even touched a gun because it turned him into another person, gave him false confidence. He described the gun as a tease, so he advised the men to never carry one. The mentees asked him questions about his thought process, how he interacted with former associates on the street, and how he weighed the pros and cons associated with his decision. This exchange was the most engaged conversation the researcher witnessed. The mentees appeared to believe the facilitator understood their dilemma and could provide advice about a real concern pertaining to their reentry.
When the men were talking about their ultimate goal-staying out and going on to lead successful lives-they almost always talked about how important it was for mentors to be credible and relatable. For the mentees, prison staff were not viewed as credible sources of advice for certain topics like whether one should carry a gun.

Credibility and relatability: Intertwined
Prison staff, mentors, and mentees talked about the important role credibility played in whether or not someone could be an effective role model for successful reentry. The prison staff frequently mentioned their lack of credibility when providing reentry advice. Martha explained: Having someone like me, who's only seen it from one side, having me tell you something, it's different. It's easier to take to heart when it comes from someone you identify with. There's lots to be gained from using people like that. (S3) The correlation between relatability and credibility was often discussed by peer mentors as well. Charles noted, ''it's about credibility. Credibility covers everything. If they can't relate to the speaker, it's useless'' (42). One of the mentees mirrored this sentiment when explaining why he valued his mentor. Jerome said, ''He [Chris] can relate. He understands. It's empathy, he's not just shooting from the hip. He knows what he's talking about'' (51). Buck (2020) reported the value mentors and mentees placed on similar lived experience, but the participants in the current study suggest that the perspective is shared by prison staff, as well.
While nearly unanimous in importance, the concept of credibility is not simple; it is multifaceted. According to the interviewees, credibility is derived from the ability to understand the various inner pains caused by the experience of incarceration, much like D. Banks McKenzie explained (Davidson et al., 2006). One of the mentees, Carl, said: Because he [Chris] knows the pain. It don't just come with stress . . . the minute you tell me you're incarcerated, and if you tell me a few things that a person on the outside wouldn't know, it clicks. He knows what's going on, how we get treated, how some COs take their job and overdo it, how being in the hole feels; he knows how the food is. (27) While several of the mentees talked about the pain caused by the carceral environment itself, the inner pain, and the wounds left behind, came from other sources as well. For many of the mentees, it was important their peer mentors knew the pain that comes from loss associated with incarceration. Andre said, ''Yeah, cause real recognize real. If you haven't walked in here, you don't know. I have a child. It hurts to be in here'' (59). Missing out on their children's lives was a loss frequently discussed by the men and one of the biggest changes they wanted to make. Ultimately, mentoring was more than just providing advice about how to find housing and employment; it was providing advice about how to do those things while recovering from the deep and varied wounds of incarceration (Sykes, 1958).

The importance of a history of incarceration
The importance of peer mentors having a lived experience of incarceration was discussed consistently throughout the entire interviews, regardless of whether interview questions pertained to why people participated, which component helped most, or during the survey portion. Conversely, the other characteristics appeared only sporadically, usually when directly asked. While other characteristics were certainly desired in an ideal peer mentor, none of the other attributes conferred a similar credibility that inspired faith in one's ability to provide advice regarding how to manage the varied and specific challenges associated with staying out of prison and leading a successful life.
The interviewees consistently explained that a history of incarceration was the most important experience because that was one thing they all shared. Not every individual in jail struggles with addiction or trauma, so many of the mentees talked about the limited credibility of peer mentors who only had a lived experience of one or the other. One of the mentors without a history of incarceration was well-liked, but his credibility, and thus his ability to be an effective role model for reentry was limited, in the mentee's mind. Jeremiah explained: He wasn't a type of person where he was brought up to get caught in something. In his mindset, he's very educated, and he's a business-minded person. When I talk to him, it's more of me telling him how I feel, so he can understand. I wouldn't take certain advice from him, like about being incarcerated or how to get housing. I wouldn't take everything he says, but some of it. (31) Even those who had a history of addiction or trauma recognized the limited applicability of that experience. For instance, Ron struggled with addiction before coming to prison and he was nervous about how he would manage it after release, but when ranking the importance of characteristics, he elaborated: ''I can go to an NA meeting and talk about my addiction all day, but they don't know anything about getting out'' (43). For Ron, his primary goal was staying out of prison and leading a successful life. While someone with a history of addiction could help him manage his recovery, the person was not a credible source of advice for never coming back to prison or overcoming barriers associated with a criminal past.
Prison staff and mentors without a history of incarceration added additional insight to the issue. One of the peer mentors who had never been to prison talked about how his experiences, or lack thereof, affected his ability to be an effective mentor. Brandon said: I guess it's kind of like, would you go to an auto mechanic who's never driven a car. Somebody who can understand a bit. Sometimes it's hard to explain, but I sometimes feel like I'm doing a bit of disservice because I haven't been incarcerated. I have been through some shit. Not that that in any way equals what they've been through, but it does give me a little bit of an understanding. (M4) Prison staff also often noted their lack of a lived experience as limiting their ability to offer credible reentry support. One of the prison staff explained: I teach and sometimes I can identify with violent guys. But, I don't know what inmate life is like. Someone like him [Chris] can satisfy both things at once: success and formerly incarcerated. I can only demonstrate how someone doesn't get in trouble. (S7) According to all of the stakeholders, the lack of direct experience limited one's ability to understand the negative aspects of their experience and thus be credible role models. Similar to Mead and MacNeil (2006), the peer mentor quoted above believed it was important for a peer mentor to be able to personally identify with the negative treatment one receives as a result of being incarcerated or having a criminal record. These feelings confirm what Davidson and Rowe (2008) argued: there is no substitute for real-world experience.
Analysis of the interview data also helped illuminate why the demographic characteristics of age, gender, and race were perceived as a less important determinant of peer status. Mixed feelings were common. In terms of race, one of the reentry staff is a young Black male, demographically very similar to the men who are incarcerated. Mansfield is extremely well-liked by the men. The mentees described him as well-intended, but they still viewed him as an outsider. Charles explained, ''people want to get us all together and try to do something. I was upstairs with Mansfield and he had a worksheet. It wasn't pertaining to us. If you haven't been in here, you don't know about it'' (30). During direct explanations about the importance of race, some mentees indicated it was of some degree of importance, but it did not confer credibility about reentry advice. For instance, of the two mentees who ranked race as Very Important, one indicated in his explanation that it was not decisive. Marion explained, ''it could be very important, unless they have sense. Like you're white, but you have sense. That's not always true'' (57). Gender caused a similar divide. Some mentees preferred a female mentor, while others preferred a male. Simon's explanation about the importance of gender was common. He said, ''I can identify with a female . . . We've both been to jail and had no resources coming out'' (26). For Simon, a shared experience of struggles associated with reentry was more important than whether someone was male or female. Similar feelings were expressed about age; mentees could see advantages and disadvantages to older and younger mentors. Overall, most of the mentees agreed that demographic characteristics were not as important as lived experiences of incarceration, trauma, or addiction. In their view, experiences were what made the peer mentors relatable. One mentor summarized, ''race, age, gender, they're only important until you get to know someone, hear what they've gone through. Then, the superficial qualities don't matter'' (M2).
Not only did a lack of credibility limit the effectiveness of a peer mentor in terms of being a role model for reentry, but without a formerly incarcerated peer component, a few of the mentees would not even participate in a program. Like Portillo and colleagues (2017), peer mentors legitimized resources. Some of the interviewees reported that the peer program they were in was the first prison reentry program they had ever signed up for. Stephen said: I heard about it from when he first came because honestly I don't do much of their programs here, they're bs . . . With him [Chris], I feel a level of confidence that he'll help us on the outside, if we can't do it on our own . . . he has insight, he can relate and apparently he didn't let it hinder his progress and let it build walls around him. (46) For the first time since coming to prison, Stephen said he was open to take advantage of a prison provided reentry resource. Other mentees expressed similar feelings. When asked why he participated, Isaiah said: Yeah, I only go if it's people that have been incarcerated. They have the same look and the same feel as people incarcerated now. So, they know where to go and how to get around obstacles. I don't want to take advice from someone who's never been in jail. (24) Ultimately, participation in the programs was sometimes contingent on whether the mentees believed the information and strategies presented were credible. Many of them used the facilitator's background, or lack thereof, as an indicator of legitimacy, to determine whether they participated.
Overall, the analysis revealed that the needs of people returning to the community after incarceration were specific. Incarceration causes specific pains, self-doubt, and stigma. These unique experiences correlate to unique challenges that are difficult to understand if you have never personally experienced and overcome these negative, interwoven challenges. The shared experiences of incarceration-the interconnected pains and challenges-connect the mentees to their peer mentors. These bonds allow the mentees to believe that they can successfully reenter society, like their mentor did.

Conclusion
The widespread use of peer mentorship programs in carceral settings was almost unheard of at the turn of the century, but peer-based reentry programming has become more common in prisons and community organizations around the world, despite the fact that much is still unknown about this potentially beneficial avenue for rehabilitation and reentry (Buck, 2020;Fletcher and Batty, 2012;LeBel et al., 2015;Reingle Gonzalez et al., 2019). A lack of understanding regarding the most fundamental element of the model-the peer mentor-could derail this promising intervention. Thompkins (2010) argues that reentry service providers often see people returning to the community as ''cash cows.'' Many of these organizations promise to provide useful services, but often fall short because of a lack of standards and accountability in the industry. During the year the researcher was inside the prison, traditional reentry providers in the prison started adding ''peer mentoring'' components to their services. One of the prison staff joked that everyone seemed to be a peer. In order to reap the benefits of peer support, we must solidify what characteristics qualify someone to be a peer mentor and what benefits those characteristics produce.
The current study employed an exploratory sequential mixed-methods design to expand our understanding of the central component of peer-focused prison reentry programs and the main benefit of these programs. The integrated findings add to the current literature in three important ways. First, qualitative analysis of participant interviews revealed the main benefit of the mentorship program: peer mentors serve as role models who offer specific guidance to help mentees successfully reintegrate back into society. Second, the qualitative analysis showed that credibility was a pivotal factor in determining whether peer mentors were perceived as effective role models for reentry. If a peer was not deemed credible, he or she would likely not be viewed as a source of inspiration and advice regarding how to navigate the varied and complex barriers people returning to society face. Finally, an integration of quantitative and qualitative analysis revealed that mentees, mentors, and prison staff ranked a history of incarceration as the most important ''peer'' characteristic because it confers credibility. The lived experience of incarceration was the most relevant and applicable experience for the mentee's principal goal: never going back to prison. The quantitative results revealed the most important peer characteristics, but the qualitative data helped explain why each characteristic was, or was not, perceived as an important determinate of ''peer'' status.
Although the peer mentor is arguably the most essential component in determining the effectiveness of a peer-focused reentry program, only one prior study, Clark and colleagues (2016), empirically investigated the most salient ''peer'' characteristics. While the findings were informative, the sample only included justice-involved veterans, which makes the findings difficult to generalize to the larger prison population. The present research expands our understanding of ''peerness'' because the sample included individuals that are more likely to resemble the general prison population, as they were recruited from the general population in two state prisons. Although the results identify a history of incarceration is the most important characteristic, there is ''no one size fits all'' mentor. Some individuals did indicate they would prefer to talk to a male or a female, someone older or younger, or someone of a particular race. Some of the men had experiences that made it easier to trust women than men, and vice versa. Matching mentees with effective mentors is not as straightforward as some practitioners make it out to be. It takes more than compassion and similar demographics to be deemed a credible source of support, inspiration, and advice. Every formerly incarcerated mentor knows what it is feels like to be locked in a cage, to feel powerless, to be degraded, and to feel loss. Additionally, every formerly incarcerated mentor has experienced the stigma and discrimination associated with the presence of a criminal history. They know how to overcome the systemic barriers while managing the private and unique pains of incarceration. While mentees may benefit from mentors who are demographically similar, analysis suggests those characteristics alone are not enough to qualify the individual to be an effective, role model for the purposes of reentry.
Finally, most prior research investigates the use of peer mentorship in community-based reentry programs (Barrenger et al., 2017;Buck, 2020;LeBel et al., 2015;Matthews et al., 2020;Portillo et al., 2017), so the study findings are particularly useful, as very little is known about how the peer support model will translate into prison-based reentry programs. The inclusion of prison staff perspectives is uniquely valuable as prior research identifies a reluctance on the part of traditional criminal justice agencies and staff to embrace peer support workers (Buck, 2020;Chapman et al., 2018;Portillo et al., 2017). Both the qualitative and quantitative analysis revealed that prison staff support and value the inclusion of peer mentors with a lived experience of incarceration. Prior research on peer mentorship in criminal and noncriminal justice settings documents the substantial implementation issues linked to traditional, non-peer staff being unwilling to value workers who have the same stigmatized identity as their clients (Buck, 2020;Chapman et al., 2018;Portillo et al., 2017), so these findings are encouraging. However, the interviews with staff did reveal frustration with administrative barriers that often prevented formerly incarcerated mentors from entering the facility to meet with their mentors. These barriers sometimes meant that there were not enough mentors to meet the demand, so mentees had to wait to meet their mentors until after they had been released. The programs are designed to cultivate a bond while the mentee is incarcerated, so that bond can provide stability through the often turbulent period people must navigate following their release.

Limitations and implications
Although this study adds to the current literature in important ways, it is not without limitation. First, the participants were only able to discuss how peer mentorship affected males. While research suggests that men and women benefit from peer mentorship (Buck, 2020;LeBel, 2007;Vigilante et al., 1999), further research is needed to discover any gender-specific differences that might emerge. Second, while the present study did collect data from two prisons, both are located in one Northeastern state in the United States, so results may not be generalizable to diverse prison populations. However, concerns about generalizability might be overstated as various studies with diverse populations reveal similar benefits (Brown and Ross, 2010;Goldstein et al., 2009;Marlow et al., 2015;Portillo et al., 2017) and emphasis of the most important peer characteristics (Buck, 2020;Matthews et al., 2020). Finally, the findings only reveal perceptions of the most important peer characteristics. Longitudinal research comparing the effectiveness of mentors without and without a history of incarceration is needed.
Despite these limitations, the findings have several important implications. First, the survey results and interview analysis indicate that a history of incarceration is the most salient peer characteristic. Although Buck (2020) indicated that the identity of ''ex-offender'' was a core component of peer mentors in community-based interventions in the United Kingdom, the inclusion of survey questions in the present study helps quantify and compare the perceived importance of different peer characteristics according to a variety of stakeholders. Therefore, prisons should only utilize peer-focused service providers that require peers to have a history of incarceration, at a minimum. Additionally, prisons, and other criminal justice agencies, should reduce existing barriers that prevent individuals with a criminal background from providing peer services. At the very least, prisons should provide clear guidelines about the background clearance process, to aid in the expansion of peer mentor services. Finally, all criminal justice agenciesprisons and community corrections-should attempt to include peer-delivered services. Not only did the mentees and mentors advocate for this change, but the prison staff also placed a high value on peer-delivered services. Ultimately, past research and the findings of this study suggest that peers who have successfully navigated the tumultuous journey of reentry are better able to assist, advise, and mentor individuals tasked with doing the same, than individuals who lack the same lived experience.

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

Funding
The author(s) disclosed receipt of the following financial support for the research and/or authorship of this article: This study was funded by 'Neil and Ann Kerwin Doctoral Fellowship'.