The male friendship recession.

In 1990, 27% of American men reported having no close friends. By 2021, that figure was 15% — which sounds like an improvement until you read the question correctly. The Survey Center on American Life data is measuring the proportion of men with zero close friends, not the size of close friendship networks. And when you look at the full distribution, the 2021 picture is not better than 1990. The average American man now has fewer close friends than at any point in the survey’s history. The 15% figure is lower than the 1990 figure because some men who previously reported zero friends now report one. The network has not grown. Its floor has risen while its middle has collapsed.

The direction of travel is consistent across multiple surveys, across demographic subgroups, and across the fifteen years of data that followed 1990. This is not a data artifact.

The Survey Center data

The American Enterprise Institute’s Survey Center on American Life published its comprehensive analysis of American friendship in 2021, drawing on data collected since the late 1980s. The male friendship findings are among the most striking in the report.

In 1990, 55% of American men reported having six or more close friends. By 2021, that figure had dropped to 27%. The proportion of men with zero close friends rose from 3% in 1990 to 15% in 2021. The gender divergence is significant: women show a similar but less pronounced decline in close friendship numbers over the same period.

The Pew Research Center’s work on friendship and social connection corroborates the directional finding. Multiple survey instruments using different methodologies have produced consistent results: men’s friendship networks have contracted substantially over three decades, and men are now significantly less likely than women to report having a close friend outside their immediate family.

One measurement complication: the definition of “close friend” in survey research is not standardized. Studies using stricter definitions (people you would turn to in a personal crisis) find smaller friendship network sizes than studies using looser definitions (people you consider friends). The Survey Center data uses a middle definition — people you consider close friends, distinct from family members and casual acquaintances. Even with this middle-ground definition, the contraction is substantial.

The structural explanations

The standard error in interpreting data like this is to reach immediately for psychological explanation — men are emotionally avoidant, men don’t invest in relationships, men are socialized to suppress vulnerability. These explanations contain some truth. They are insufficient.

The data shows a decline over a specific thirty-year period, across demographic groups, in a pattern that tracks structural changes in American life. Psychological characteristics don’t shift that dramatically across a thirty-year window without structural drivers. The structural drivers are identifiable.

Geographic mobility: the past three decades have seen substantial increases in long-distance migration for employment. People move away from the cities and communities where they built friendships during adolescence and early adulthood. Adult friendships — particularly male friendships, which research consistently shows are more context-dependent and less portable than female friendships — do not survive geographic relocation at the rates people expect. The friendship network stays behind. The new city provides professional colleagues, not close friends.

The post-college cliff: the institutional settings that produce friendship without deliberate effort are overwhelmingly concentrated in the first twenty-two years of life. School, dormitory life, team sports, shared housing — these create the conditions Rawlins (1992) identifies as necessary for friendship formation: proximity, repeated unplanned interaction, and a setting that encourages disclosure. Those conditions largely disappear in the mid-twenties. Adult friendship requires intentional cultivation in a way that adolescent friendship does not. Most people — most men in particular — are not socialized to cultivate friendship intentionally.

Work as identity: as careers have become more demanding and longer in hours, work has substituted for social life in ways that appear to provide connection while actually preventing it. Work friendships exist within a conditional context — they are contingent on continued employment, they are bounded by professional norms, they are often suspended when employment changes. They are not the same as close friendships formed outside of transactional contexts. The substitution feels adequate until it isn’t.

Oldenburg’s (1989) concept of the third place — the informal public space, neither home nor work, where community forms — is relevant here. Bars, barbershops, community centers, religious institutions, bowling leagues: these are the spaces where repeated low-stakes social contact accumulates into familiarity and, sometimes, friendship. The decline of third places in American life — documented by Putnam in Bowling Alone (2000) and by subsequent researchers — is a structural condition, not an individual choice.

The health dimension

Holt-Lunstad et al.’s 2010 meta-analysis in PLOS Medicine examined the relationship between social relationships and mortality across 148 studies. The finding: individuals with adequate social relationships had a 50% greater likelihood of survival compared to those with poor social relationships. The effect size is comparable to that of smoking cessation and exceeds the effect of obesity, exercise, and alcohol consumption as mortality predictors.

The Surgeon General’s 2023 advisory on loneliness and isolation positioned the decline of social connection as a public health crisis. The advisory does not frame this as primarily a male problem, but the male friendship data it cites is among the most striking. Men who lose close friendships in middle age rarely rebuild equivalent networks.

The research does not distinguish cleanly between friendship and other forms of social connection — the health effects of connection are more general than the friendship literature alone. But friendship, specifically, has properties that other forms of connection do not: mutual investment, voluntary maintenance, disclosure without professional or family obligation. The loss of those properties in men’s lives has health consequences that aggregate data on social connection captures imperfectly.

The self-improvement industry’s response

The books, podcasts, men’s retreats, and online communities that have emerged in response to the male friendship crisis are a cultural phenomenon worth noting. Mark Manson’s work addresses the difficulty of male connection explicitly — the framing in The Subtle Art of Not Giving a F*ck and in subsequent writing identifies social disconnection as a problem while positioning individual attitude change as the primary lever. Jordan Peterson’s work similarly addresses male meaning-making and community, with a heavier emphasis on individual responsibility and hierarchical ordering.

Both frameworks reach audiences of millions. Both provide something useful to men who feel isolated: a vocabulary for the problem, and a framework for individual response. The limitation of both is the same limitation that affects most individual-level responses to structural problems: the structural problem does not resolve at the individual level.

Putnam’s critique in Bowling Alone is relevant: the decline in social capital is not primarily a product of individual choices that individuals can reverse by choosing differently. It is a product of structural changes — in housing, transportation, work, technology, and urban design — that require structural responses. The man who joins a men’s group, reads Peterson, and works on his individual psychology may have a better experience of his own isolation. The thirty-year trend in male friendship networks will not be reversed by that.

What actually works

Rawlins (1992) on friendship across the life course identifies the conditions that sustain adult friendship: proximity, repeated unplanned interaction, and a setting that encourages disclosure. The data suggests these conditions are now rare in adult male life.

What emerges from the research on adult friendship formation and maintenance: the contexts that most reliably produce close adult friendships share structural features with the institutional settings that produced adolescent friendship — regular, repeated contact in a setting that is not primarily transactional, where disclosure is normalized. Recreational sports leagues, religious communities, hobby groups, neighborhood associations, and military service are among the contexts that appear in the friendship research as sites where adult friendship forms. None of these require significant individual psychological change. All of them require structural conditions — proximity, regularity, a shared non-transactional purpose — that most adult men do not currently have access to and cannot produce through individual decision alone.

This is not a hopeful conclusion in the conventional self-improvement sense. It is an accurate one.