The 13-to-17 window: why TWIN's age range is calibrated to the research
The Seals findings are clear: gang participation responds to unemployment only at ages 16–17 — the first age of legal employment. That makes the window before formal employment age — 13 to 15 — the optimal intervention point. Here's why TWIN built its program around that exact range.
Most youth programs in NYC enroll anyone from 11 to 24. That's not a strategy. That's a fundraising compromise. The actual research on when interventions can still rewrite a trajectory points to a specific window — and TWIN is built around it on purpose.
What the Seals data showed
When researchers studied the relationship between teenage unemployment and gang participation, they found something specific and important: gang participation responds to unemployment only at ages 16–17. Younger boys' decisions about gang involvement are not driven by labor-market conditions. They're driven by something else — something developmental, social, observational.
Once you cross 18, the path locks in. By the time someone is in their early twenties, the relational, cognitive, and economic structures of street involvement have calcified. Interventions still work — READI Chicago proved that — but they cost more, take longer, and have lower base rates of success.
The optimal window
optimal intervention window — pre-employment age, identity malleable, habits forming
last on-ramp — first age of legal employment, the moment unemployment starts driving the decision
path locked in — interventions still work but cost orders of magnitude more
The window before formal employment age — 13 to 15 — is where the cheapest, highest-leverage prevention work can be done. It's the moment when career exposure can still reroute a developing identity. It's also the moment when most NYC programs are NOT focused. Most resources concentrate on either elementary kids (where the risk hasn't manifested yet) or 18-and-up adults (where the path has already calcified).
How TWIN is calibrated
TWIN's youth eligibility is 13–17. That's not arbitrary. The lower bound captures the start of the optimal window. The upper bound catches kids who are still in the last on-ramp — before the labor-market variable starts to dominate. Mentors, by contrast, are 19–28: old enough to have visibly built something, young enough that the mentees can still see themselves in them.
“Five years where the intervention can still rewrite a trajectory. TWIN catches the first three years of that window — when the cost of changing course is lowest.”
— TWIN Needs Assessment, Section 3
Why this matters for funders
Per-dollar effectiveness scales inversely with age in this model. A dollar spent on a 13-year-old produces dramatically more long-run change than the same dollar spent on a 21-year-old. This isn't because the 21-year-old is unworthy of the intervention. It's because the 13-year-old has more years of compounding ahead of them, fewer calcified patterns to undo, and a developmentally open window for identity formation. The math is the math.
