The United States has never had universal national service. It has had a draft, which it used during four major wars and the Cold War standoff, and it has had voluntary service programs, which it has expanded and contracted according to political winds. The countries that do have universal national service — Israel, South Korea, Norway, Switzerland, and others — produce, as a structural byproduct, something the United States produces only sporadically and accidentally: a class of lifelong friendships that cut across the social boundaries that civilian life maintains.
The mechanism is not complicated. Take a diverse national population, require service in conditions of shared challenge and genuine interdependence, remove the pre-existing social markers that sort people in civilian life, and assign identities based on new criteria — unit, role, mission — and you produce the conditions under which people from radically different backgrounds form bonds that are more durable and more cross-stratifying than almost anything civilian social life generates. This is what armies have always done as a side effect of their primary purpose, and it is what universal national service programs do deliberately.
The friendship research is relevant here in ways that policy discussions of national service rarely incorporate. George Vaillant's Harvard Study of Adult Development, which tracked men for over eighty years, identified the quality of social relationships as the single most powerful predictor of physical and psychological health in later life — more predictive than wealth, intelligence, class of origin, or the absence of trauma. Sebastian Junger's analysis of military cohesion and its aftermath established that combat veterans' frequently reported grief for their service years was not grief for violence but for the intensity of human connection that combat conditions produced, a level of mutual investment and trust that civilian life rarely replicated. The implication — that the social conditions of intense shared service produce the highest-quality friendships, and that those friendships have lifetime health and wellbeing consequences — should be central to the policy argument for national service, and is almost entirely absent from it.
The current policy case for AmeriCorps, or for universal service proposals, rests primarily on direct outputs: the tutoring hours, the conservation work, the disaster response. These outputs are real and worth measuring. But they are also replaceable: the same work could, in principle, be done by expanded government employment or contracted service delivery. What is not replaceable, and what is not being measured or argued for, is the friendship infrastructure that organized national service uniquely produces.
The policy stakes are significant. The United States is experiencing what the Surgeon General's 2023 advisory called an epidemic of loneliness and isolation: elevated rates of reported loneliness across all age groups, declining numbers of close friends per person, declining civic participation, and declining inter-group trust across racial, economic, and political lines. These trends have documented public health consequences — loneliness is as damaging to mortality as smoking — and documented civic consequences — the weakening of the bridging social capital that democratic institutions require. A national service program that enrolled even a fraction of each birth cohort in a year of organized shared service would, the evidence suggests, produce lifelong friendship networks that cross the social boundaries those cohorts would otherwise maintain for life.
That argument — for national service as friendship and social capital infrastructure, not merely as public work delivery — is available and not being made. The political economy of national service discourse is oriented around direct outputs because outputs are countable and countability is what the current accountability infrastructure rewards. The friendships that last forty years, the trust that crosses party lines, the neighbor who knows the other neighbor because they served together — these do not appear in any program evaluation. They are nonetheless the most important things that national service can produce, and treating them as incidental byproducts rather than primary purposes has shaped the scale and design of programs in ways that have cost the country substantially.