The American Perspectives Survey, conducted by Daniel Cox at the Survey Center on American Life and published through the American Enterprise Institute, is the most comprehensive dedicated study of friendship in the contemporary United States. Unlike omnibus surveys that include a friendship question, the APS devoted full survey instruments across multiple waves to the topology, quality, and trajectory of American friendship. Its findings are not anecdotal. They are nationally representative, methodologically documented, and — where they can be tested against other surveys — replicated.
The headline numbers are damaging. Between 1990 and 2021, the share of Americans with ten or more close friends fell by more than half. The share with no close friends more than tripled. Among men, the trend is severe enough to constitute a distinct social phenomenon: 15 percent of American men reported having no close friends in 2021, a figure five times higher than the 3 percent reported in 1990. The median American man now reports having three close friends, down from a historical norm of closer to six.
But the headline numbers are the beginning, not the end, of what the APS reveals. The survey examined not just quantity but quality — how often people discuss personal problems with friends, whether they feel they can rely on friends in a crisis, whether they are satisfied with the state of their social lives. On every quality dimension, the trend lines move in the same direction as the quantity data. Americans are not maintaining a stable number of close friends while the depth of those friendships erodes. Both are declining simultaneously.
The survey also revealed the role of structural context in friendship formation. Adults who are embedded in organized religion, civic associations, or stable workplaces report more close friends and higher friendship satisfaction than those without such institutional anchors. The mechanism is straightforward: institutions create repeated, low-stakes contact with the same people over time — the preconditions for friendship formation. As Americans have disaffiliated from institutions at historic rates, the structural scaffolding for friendship has weakened proportionally.
The data disaggregates in ways that matter for understanding causation. Married Americans report more close friends than unmarried Americans — but the advantage is declining. Americans with college degrees report more non-kin close friends than non-graduates. Rural Americans report fewer close friends than urban ones. Young adults, who once formed friendships almost automatically through school and early employment, now report lower friendship satisfaction than older cohorts.
The APS data is not a lament for a golden age that never existed. It is a measurement of a real change in the social fabric — a change that is distributed unequally, driven by identifiable structural forces, and measurable in its consequences for mental health, civic participation, and the social trust that democratic institutions require.