A mentorship program is a society's admission that the family unit, by itself, cannot guarantee every child a path to adulthood. When the village shrinks — when extended kin disperse, when neighborhoods stop knowing each other, when single parents work two jobs — the developmental scaffolding that used to be automatic becomes scarce. Big Brothers Big Sisters of America, founded in 1904 when a New York court clerk named Ernest Coulter noticed how many boys were appearing before the bench without an adult who knew them, was the first formal attempt in the United States to manufacture that scaffolding at scale. The premise was simple and remains simple: pair a child who lacks a stable adult relationship with a screened, supervised volunteer who shows up, weekly, for years.
What the program actually delivers is not advice. It is presence. The 1995 Public/Private Ventures impact study, conducted by Joseph Tierney and Jean Grossman across eight BBBS affiliates with 959 ten-to-sixteen-year-olds randomly assigned to either an immediate match or an eighteen-month waitlist, found that mentored youth were 46 percent less likely to initiate drug use, 27 percent less likely to initiate alcohol use, and 52 percent less likely to skip school during the study window. These are not modest numbers. They were produced not by curricula or lectures but by an adult who took a kid bowling and asked, with some regularity, how things were going.
The collective implication is uncomfortable. If a stranger, meeting a child for two hours a week, can move outcomes that much, then the absence of such adults in many children's lives is not a private misfortune. It is a structural one. The Connect Law — that wellbeing is built in relationship, not delivered in services — names this directly. A society that produces large numbers of children with no consistent non-parental adult in their lives is a society that has rationed connection. Mentoring programs are the patch. They are also evidence of the underlying wound.
But the patch is not free of complication. Jean Rhodes, whose Center for Evidence-Based Mentoring at the University of Massachusetts Boston has produced more rigorous mentoring research than any other lab, has spent twenty years complicating the BBBS narrative. Match length matters enormously: relationships that end before twelve months can actively harm youth, producing measurable declines in self-worth and academic competence compared to never having been matched. The 2002 meta-analysis by David DuBois and colleagues, covering fifty-five evaluations, found a modest overall effect size of 0.18 — useful but not transformative — and the effects depended heavily on program quality. Screening, training, ongoing supervision, structured activities, and match closure protocols were not amenities; they were the difference between benefit and harm.
The Connect Law also predicts the failure modes. When mentoring is conceived as a service delivered to a deficient child by a generous adult, it tends to underperform. When it is conceived as a relationship between two people, one of whom happens to have more years and resources, it tends to work. The asymmetry is hard to design around. Volunteers who arrive expecting gratitude find themselves managing a child's testing behavior — late arrivals, missed sessions, sullen silence — that is itself the child checking whether this adult, like the others, will leave. Programs that prepare volunteers for this phase retain more matches. Programs that don't lose them at month three.
At the collective scale, mentoring sits inside a larger question about how a society distributes adult attention. In a neighborhood where most adults know most children, formal mentoring is redundant. In a neighborhood where adults and children pass each other as strangers, formal mentoring is a thin substitute for what used to be free. The Plan Law cautions against confusing the substitute for the original. A nation that funds BBBS generously while permitting the conditions — long commutes, declining civic membership, housing churn, criminalized parents — that make BBBS necessary is solving a symptom while feeding the disease.
The honest summary: BBBS works, modestly, when run well, for kids who get matched long enough. It is one of the few social interventions with random-assignment evidence of effect. It is also a confession. Every child in the program is a child for whom the ordinary architecture of connected adulthood failed. Scaling the program is good. Asking why it has to exist at this scale is better.