The Limits of Mentorship

Published January 13, 2022

National Affairs - Winter 2022 issue

As America suburbanized in the decades following World War II, neighborhoods — the places where family and civic life is lived on a day-to-day basis — gradually became more socio-economically stratified. Between 1980 and 2010, the percentage of low-income households located in a majority low-income census tract rose from 23% to 28%, while the percentage of upper-income households located in a majority high-income census tract doubled, from 9% to 18%. As a result, children in today’s America are less likely than children of earlier generations to grow up around families from different income brackets.

Youths raised in communities where poverty is concentrated not only want for material resources; they are also surrounded by broken families, addiction, a lack of opportunity, and a paucity of positive role models. Many well-meaning organizations seek to redress this imbalance through formal mentoring programs, wherein low-income or otherwise at-risk youth are paired with adults to help them foster positive habits, build valuable skills, and achieve personal goals.

In theory, formal mentoring offers a way to generate the kind of cross-class “bridging” social capital, to borrow Robert Putnam’s term, that builds relationships between individuals from different economic as well as social, racial, ethnic, and religious groups. And indeed, many mentees have had their lives changed through these programs, as have their adult mentors.

But while the desirability of helping youth achieve their full potential is self-evident, even the best of intentions do not guarantee positive outcomes. The available evidence should engender a sense of caution with regard to relying too heavily on these programs to improve the lives of low-income children. Although a relationship with a mentor may be beneficial, meaningfully supporting at-risk children requires a strategy that focuses on strengthening the institutions in which mentoring relationships develop organically.


In a 2016 report on the growing opportunity gap between rich and poor Americans, Putnam, along with Patrick Sharkey of New York University and Margery Turner of the Urban Institute, explained how community investment (or a lack thereof) can influence a child’s life:

Youths from more-advantaged families and communities are typically surrounded by mentors and role models, and when they mess up (as most do) they have access to a range of “airbags” in their families and community to cushion any blow to their success. When young people from areas of concentrated disadvantage make bad decisions, they are far less likely to have access to the same airbags in their families or communities and are far more likely to experience harsh punishment in the schools or on the streets.

Some of this disparity is due to poverty itself — after all, an affluent family is more likely than a poor one to be able to afford a tutor for a child who is failing out of school, a therapist for an adolescent who develops an addiction, or a private attorney to help a young person who finds himself in trouble with the law. But much of it is also due to the social networks in which families are embedded. As Putnam points out in his book Our Kids, high-income parents tend to have a wealth of both “strong” and “weak” ties to members of their communities. The former include connections with close friends and neighbors, while the latter consist of connections to broad, diverse networks of individuals within the community, and even outside of it. These ties offer the children of affluent families emotional, social, and financial support, as well as opportunities to access employment and higher levels of education. And if a child from a well-off family strays from the straight and narrow — as young people are wont to do — a variety of community cushions can help protect him from the harshest consequences of his actions.

By contrast, social networks in our poorest communities tend to be both overlapping and insular. According to Putnam, “lower-class parents’ social ties are disproportionately concentrated within their own extended family (and perhaps a high school friend and a neighbor or two), who, because of their own location in the social hierarchy, are unlikely to expand the reach of the parents.” Families in disadvantaged neighborhoods also tend to be cut off, voluntarily or otherwise, from resources outside the community. In his pivotal book The Truly Disadvantaged, Harvard University’s William Julius Wilson stressed the social isolation of poor urban neighborhoods as a key driver of poverty:

[T]he residents of highly concentrated poverty neighborhoods in the inner city today not only infrequently interact with those individuals or families who have a stable work history and have had little involvement with welfare or public assistance, they also seldom have sustained contact with friends or relatives in the more stable areas of the city or in the suburbs.

When combined with the fact that, as Wilson points out, “a person’s patterns and norms of behavior tend to be shaped by those with which he or she has had the most frequent or sustained contact and interaction,” it follows that the behavioral patterns and norms children from underprivileged families develop are likely to reflect those of other individuals living in unstable situations. While Wilson was writing about poor blacks living in urban neighborhoods in the 1980s, one could easily extend his analysis to residents of rural towns left reeling by the departure of their economic base in more recent years.

Unsurprisingly, given our socio-economically stratified times, family income plays a major role in determining which individuals children interact with. It also seems to play a role in young people’s tendency to form relationships with adult mentors. Estimates suggest that roughly two-thirds of American children have a close relationship with a non-parental adult, but those mentoring relationships are not evenly distributed across the population: Youth from low-income families, those whose parents are receiving public assistance, and those who are living in neighborhoods with greater levels of poverty, as well as youth who are male, black, or Hispanic, are all disproportionately less likely to report having an influential relationship with a mentor. More specifically, “research has shown that young blacks living in disadvantaged inner-cities have relatively few adults in their communities that they can depend on for guidance.”

Mentoring programs are intended to counteract these circumstances by offering children from lower-income backgrounds opportunities to engage with positive role models and expand their social networks. But for policymakers, the empirical question of whether these programs are effective remains key.


Formal mentoring programs consist of structured relationships between youth and adults. They may be community based — typically oriented around non-academic activities — or school based, often held during or after school but separately from academic tutoring. Most structured mentoring programs aim to supplement a child’s existing relationships with parents, peers, and community members, and use a set period of relationship building to instill young people with the habits and skills that can set them on an upward socio-economic trajectory.

Formal mentoring in the United States stretches back over a century. The largest youth-mentoring organization, Big Brothers Big Sisters of America (BBBSA), got its start in 1904, when a New York Children’s Court clerk recruited 39 volunteers to each befriend a boy who had entered the juvenile-justice system. A dozen years later, the model had spread to 96 cities nationwide.

Hundreds of formal mentoring programs have followed in BBBSA’s footsteps since the early 20th century, with a particularly large surge in the creation of such programs occurring over the last three decades. A 1999 survey conducted by sociologists Cynthia Sipe and Anne Roder identified 2,320 local organizations offering formal mentorship programs; of the 722 programs responding to the survey, nearly 40% had been founded within the preceding five years. Three years later, the number of such programs had more than doubled, amounting to an estimated 5,000.

What accounts for this surge? The trend has roots in the late 1980s and 1990s, when scholars began discovering positive effects associated with formal mentoring programs like BBBSA. In 1998, researchers Joseph Tierney and Jean Baldwin Grossman published a landmark study finding that youths randomly assigned to a mentoring program showed reduced incidence of interpersonal violence, drug and alcohol use, and absenteeism than those who had not been involved in a program.

Around the same time, popular books like Marc Freedman’s Kindness of Strangers and Jean Rhodes’s Stand by Me stoked broader interest in mentoring among the public. Non-profits like the Points of Light Foundation, United Way, America’s Promise Alliance, and MENTOR began experimenting with models that enabled both public and private investment in formal mentoring programs. These developments coincided with a bipartisan enthusiasm for volunteerism among public officials, manifested in events such as the 1997 Presidents’ Summit wherein President Bill Clinton joined his living predecessors at Independence Hall to encourage the creation of 2 million mentoring relationships. By 2015, survey data showed roughly 2.6 million American adults — or about 1% of the adult population — had volunteered as a mentor for at least one academic or calendar year.

There is a modest history of federal efforts to encourage mentorship. Between 1994 and 2003, Congress appropriated a total of $104 million to the Department of Justice’s Office of Juvenile Justice and Delinquency Prevention (OJJDP) for a program intended to reduce the number of juvenile offenders through relationships with adult mentors. In 2003, a White House Task Force on Disadvantaged Youth called for better coordination of all federally sponsored mentoring programs and activities, which led to the creation of the Federal Mentoring Council (active from 2006 to 2008). All told, since the early 2000s, the federal government has spent over $2.5 billion on formal mentoring programs.

Of course, as Rhodes and her co-authors Elizabeth Raposa and Nathan Dietz have noted, “[f]orging an ongoing relationship with a child, particularly one who may be struggling with the effects of poverty or trauma, requires a more substantial investment of time and self than more typical episodic volunteer activities.” And in fact, subsequent analyses have found only “small overall positive effects of formal mentors on the psychological, emotional, behavioral, and educational functioning of participating youth.” A 2019 meta-analysis of 70 youth-mentoring studies estimated a positive mean effect size of one-fifth of a standard deviation (a standard deviation is the typical amount by which any given youth differs from the average).

To be clear, these effect sizes are meaningful. But they are also modest. A cost-benefit calculation by Rhodes and others found “monetized benefits of program participation to exceed total costs by only a narrow margin.” These limited effects are roughly in line with those discovered in past generations of mentoring research, suggesting that current interventions are not improving outcomes even when best practices are implemented.

While research continues into promising new models, the most straightforward reading of the evidence should prompt some humility toward the belief that formal mentoring programs dramatically change outcomes for participants. While no one disagrees that the right mentor in the right program at the right time can have a transformative effect on a child’s life, for lawmakers, the key question should be whether they can reliably create such relationships through public policy. And in the case of formal mentoring, the evidence suggests they might be better off pursuing other strategies.


Informal mentoring consists of the kind of relationships that form naturally between young people and the adults in their lives. Typical informal mentors include teachers, coaches, neighbors, and religious leaders. By definition, such relationships cannot be engineered by policy initiatives. Gaining a more systematic understanding of how they develop, however, may suggest ways in which policymakers can establish and nurture environments that tend to foster them.

Any rigorous evaluation of informal mentoring is hindered by the lack of random assignment, as any number of characteristics may influence a child’s likelihood of forming a relationship with an adult that is correlated with the child’s long-term success. But a 2018 meta-analysis of 24 studies found larger effect sizes for natural mentorship relationships with adults in the “helping professions” (teachers, guidance counselors, priests, rabbis, etc.) than in others, suggesting that caring adults within the social institutions that children are involved in — such as schools and churches — can have a meaningful impact on the lives of vulnerable youth.

Indeed, as Putnam asserts, “[m]entoring works best as the by-product of a connection that rests on some shared interest, like tennis or skateboarding or fishing.” Rhodes concurs: “In mentoring, a close relationship is vitally important. But when it is construed as the only active ingredient, it creates unrealistic pressure on everyday relationships — zapping them of the closeness that can come when there are shared goals and a sense of purpose.”

Extracurricular activities like sports, dance troupes, clubs, and volunteer organizations offer natural opportunities for young people to come into contact with potential mentors in an environment of shared goals and interests. Unfortunately, access to such activities increasingly correlates with income. Putnam and Reuben Finighan found that the “participation gap in all extracurriculars between poor and rich kids grew from less than 10 percent to more than 20 percent” between 1972 and 2004, and that it was driven primarily by falling participation among youth of the lowest socio-economic status.

“Because of financial constraints,” write professors Pamela Bennett, Amy Lutz, and Lakshmi Jayaram, “working-class families rely on social institutions [like schools] for affordable participation opportunities.” Yet as Putnam notes, thanks to educational budget cuts and an emphasis on high-stakes testing, “school boards everywhere [are deciding] that extracurricular activities and soft skills are ‘frills'” they can ill afford. As a result, children from financially disadvantaged families are losing access to free or low-cost extracurricular opportunities.

The proliferation of pay-to-play policies — those that require parents to pay fees or make other monetary contributions before their child can engage in an extracurricular activity — have only added to the problem. Putnam estimates the average total cost of such activities to be about $400 per student, per activity for each semester, with the predictable result being that children from low-income families can’t afford to participate. That more-affluent parents have more resources to invest in their children should not come as a surprise, but some research suggests that the share of income they spend on their kids relative to what less well-off parents spend on theirs has grown disproportionately over time.

Aside from traditional extracurricular activities, research also indicates that participation in religious groups may have a beneficial effect on adolescent outcomes. Researchers Jennifer Glanville, David Sikkink, and Edwin Hernández find that religious activities promote overlapping intergenerational relationships that earlier generations of theorists have identified as crucial for building social networks. Their findings also suggest that greater involvement with religious organizations predicts greater participation in other extracurricular activities, which leads to additional opportunities for young people to find mentors.

Religious groups not only provide an activity for children to participate in, they encourage well-meaning adults to involve themselves in the lives of youth. Yet trends in religious disassociation have hit families whose children could benefit most from natural mentors in religious settings particularly hard. As recently as 1982, high-school seniors from the lowest socio-economic quartile were only slightly less likely to participate in church activities or religious youth groups than those from higher quartiles. But over the following two decades, as Putnam and his co-authors report, “church attendance [fell] much more rapidly among children from working class families than among children from affluent families.” “This divergence,” they add, “may contribute to the growing opportunity gap” between children from advantaged and disadvantaged families.

As poor and working-class families become less likely to participate in community institutions like churches, their children have less exposure to the ancillary benefits that belonging to those social organizations can provide. No one would recommend joining a church so that a child has access to a better summer internship, but no one would dispute that such a benefit may accrue from a church membership, either. These sorts of gains are precisely what Pierre Bourdieu and other early theorists were trying to measure when they came up with the term “social capital.”

The social capital available to a child raised in a stable family that is embedded in a rich web of community institutions can be immense. But given the rise in socio-economic segregation, children from poorer families not only have reduced access to such institutions, the benefits they can gain from their participation in them will be commensurably lower. After all, participating in a youth group or sports league will be less beneficial for children if they are not building relationships with counselors or coaches who can model healthy social scripts.

The available research we have suggests that lessons of what it means to be a successful adult are best taught not through formal programs isolated from broader institutional settings, but as part of larger projects involving shared interests and a joint purpose. Yet in recent years, children in our nation’s highly disadvantaged families have lost access to the kinds of environments and institutions where mentoring relationships can develop organically. To ensure that youth from all socio-economic backgrounds have the opportunity to become involved in an activity or institution where they are likely to encounter mentors, policymakers should strive to make these institutions as strong and as widely accessible as possible.


Given the evidence indicating that formal mentoring programs do not result in more meaningful or long-term beneficial relationships than do informal ones, our public-policy strategy should shift accordingly. Instead of focusing on stand-alone mentorship programs, elected officials at all levels of government should prioritize approaches that enable communities to build and maintain institutions that bring young people into contact with potential adult mentors more naturally. They should also work to help institutions incorporate mentorship techniques into existing activities and programming.

To start, lawmakers can take steps to give economically disadvantaged children greater opportunity to participate in extracurricular activities. Though family structure is a contributing factor, financial burdens are often the driving force behind why children from low-income families are unable to participate in such programs.

Some states have already taken steps to expand access to extracurricular activities. Arizona, for instance, offers a tax credit of up to $400 for fees or donations to public schools to fund extracurricular programs. States and school districts with sufficient resources might explore the feasibility of reducing or eliminating pay-to-play policies altogether.

If one of our goals is to help disadvantaged children encounter a wider range of social networks and resources than they come across in their schools, we should also consider policy prescriptions beyond those that assume school-based extracurricular activities are always the appropriate answer. Fordham Institute president Michael Petrilli has suggested policymakers establish “enrichment savings accounts” that give parents “the equivalent of a debit card to be used for sports, art, summer camps, Girl Scouts, and all the rest.” Such an approach could be piloted by states or localities, as well as philanthropic organizations.

Another option lawmakers may want to consider is broadening the definition of allowable expenses under 529 savings plans. At the moment, only tuition at an eligible school is considered a qualified expense at the elementary- and secondary-school levels. Expanding the definition to include expenses related to extracurricular activities could enable more children from disadvantaged families to participate in them.

Aside from extracurricular programs, some of the most natural vehicles for cultivating informal mentoring relationships are religious institutions. This is not only because they enable young people to form mentoring relationships with the pastors, priests, and rabbis who minister to the congregation, but because of the relationships that develop among families that attend religious institutions. Survey data suggest that adults who participate as mentors tend to be more religious than those who don’t, while an older analysis found that 43% of adults who engaged in mentoring-type activities did so through religious organizations. America’s long-standing tradition of the separation of church and state should urge caution about directly subsidizing religious programs aimed at bolstering mentoring relationships. But faith-based groups should not be left out of ongoing efforts to build community-wide webs of support that help foster mentoring relationships among adults and vulnerable youth.

One underappreciated vehicle for encouraging faith-based organizations to undertake more mentoring opportunities is the 21st Century Community Learning Centers (21st CCLC) program. The grants distributed through this program — totaling $1.2 billion in 2019 — support community learning centers that offer academic and extracurricular activities outside of school hours. From the outset, faith-based organizations have been deemed eligible to participate in the program, but as of 2016, only 1.2% of 9,592 centers receiving grants were faith-based organizations, while 82% were public schools. If federal lawmakers are interested in helping at-risk and disadvantaged youth find mentors outside the traditional school system, they could direct the Department of Education to issue guidance to state educational agencies — the entities responsible for subcontracting 21st CCLC funds — encouraging them to give priority to applications featuring partnerships with faith- or community-based groups. More aggressive measures could be taken at a later stage if the percentage of funding directed to public schools remained disproportionately high.

States and localities should also experiment with creative ways to inject mentor-like components and opportunities into existing youth programs. In Boston, City Connects’s model helps construct networks of support around students and their families. The program assigns counselors or social workers to coordinate with teachers and students in developing customized plans to help students achieve academic and personal success while identifying community resources that can help them achieve their goals. City Connects has been evaluated as generating a moderately positive return on investment and better academic outcomes among participants.

While formal mentoring programs operating in isolation may not be the answer to connecting youth with mentors, that doesn’t mean they can’t be part of the solution. Schools, sports teams, religious institutions, and others might find it useful to partner with local mentoring organizations to incorporate mentor education and training into their activities. Teachers, coaches, pastors, counselors, and other adults who interact with young people should all be considered potential participants.

In fact, many organizations — including current OJJDP grantees — have found success with this approach. The U.S. Soccer Foundation’s Soccer for Success program, to take just one example, trains coaches in programs like Boys & Girls Clubs, YMCAs, Police Athletic Leagues, and other youth soccer leagues to serve as mentors. These coach-mentors interact with youths both on and off the field, developing ties grounded in the practices unique to soccer while serving as positive role models more generally. Over 190 organizations run 1,500 Soccer for Success sites across the country, and since the program started in 2009, it has served roughly 100,000 participants.

Protecting children from abuse is a mandatory part of any youth program, but the costs and complexity of meeting overlapping federal, state, and local screening regulations can create unnecessary hurdles for cash-strapped non-profit organizations, particularly those looking to scale up. To address this issue, lawmakers at the federal level might consider taking steps to streamline the web of background checks and screening practices with which youth organizations must contend.

Policymakers at the state level, meanwhile, should be aware that the traditional, volunteer-based approach to mentoring is being supplemented — and in some cases, replaced by — mentors who receive stipends, particularly in programs that require training or a time commitment, like Soccer for Success. A handful of states that have increased their minimum wages have not commensurately increased their funding for youth programs, which are forced to make up the difference for paid mentors out of their own pockets. Granting youth programs waivers from binding wage requirements could make it easier for them to continue to recognize mentors’ time and investment without incurring heavy financial burdens.


These ideas are deliberately designed to be both modest in scope and politically achievable. We could dream bigger — a broader rethinking of our neighborhoods and patterns of development, for instance, could naturally lend itself to more socio-economically integrated communities. A cultural “great awakening” could make it more likely that families of all walks of life participate in religious communities. A nationwide recommitment to the institutions of civil society could lead to a reforging of the kind of associational life that social conservatives prize.

Such changes would certainly leave fewer children at risk of falling through the cracks. But grand transformations of this sort require changes of heart and mind that are beyond the ability of public policy to bring about directly and in short order. In the meantime, policymakers can and should take steps to address existing challenges without waiting for cultural shifts to make broader transformations possible.

There is a reason why mentoring has traditionally received bipartisan support: We all want to believe that the right relationship can help a young person reach his full potential — and many mentoring relationships do. But effective programs that are narrowly targeted are difficult to scale, while easily scalable programs can lose the individualized attention that is key to success. Policymakers should not comfort themselves by thinking that a weekly meeting with a formal mentor is enough to reduce rates of poverty; we need communities, not just individuals, to invest in at-risk youth.

Patrick T. Brown is a fellow at the Ethics and Public Policy Center.

Patrick T. Brown is a fellow at the Ethics and Public Policy Center, where his work with the Life and Family Initiative focuses on developing a robust pro-family economic agenda and supporting families as the cornerstone of a healthy and flourishing society.

Most Read

This field is for validation purposes and should be left unchanged.

Sign up to receive EPPC's biweekly e-newsletter of selected publications, news, and events.


Your support impacts the debate on critical issues of public policy.

Donate today