In 1990, when the General Social Survey asked American men how many close friends they had, 3 percent said none. In 2021, the Survey Center on American Life's follow-up survey asked the same question, and 15 percent said none. Twenty-one percent — one in five American men — said the same when the survey ran again in 2023. [1] Among men under 30, the number rises to 15 percent reporting zero close friends, the highest rate ever recorded for the cohort. Gallup's 2025 data, released in early 2026, found that 25 percent of American men between 15 and 34 — one in four — report feeling lonely daily.

The health economics are well-established. Brigham Young University psychologist Julianne Holt-Lunstad's 2015 meta-analysis of 148 studies found that chronic loneliness carries mortality risk equivalent to smoking fifteen cigarettes a day. [1] Loneliness raises cardiovascular risk, dementia risk, and the risk of depression. For men specifically, the U.S. Surgeon General's 2023 advisory — the first federal document to name loneliness an epidemic — noted that social isolation is one of the strongest predictors of suicide. Nearly four in five deaths by suicide in the United States in 2022 were men.

The Pew Research Center's January 2025 report adds nuance, not contradiction. Pew found that men and women report feeling lonely at similar rates — about one in six say "all or most of the time" — but that men turn to their networks less often for emotional support. [2] Thirty percent of men say they share personal feelings with a close friend. For women the number is 48 percent. Forty-one percent of women say they have received emotional support from a friend in the past year; for men the number is lower by double digits. Men and women have close friends at similar rates; they use them differently.

What explains the shift from 3 percent to 15 percent in thirty-five years is, in most of the empirical literature, structural. A University of Kansas study on friendship formation found that it takes about fifty hours of repeated interaction to move from acquaintance to casual friend and about two hundred hours to reach close friendship. [3] The structures that historically produced those hours — school through the early twenties, church, the local bar, union halls, neighborhood associations, the civic infrastructure Robert Putnam mapped in Bowling Alone — have eroded across most of the West. Working from home after 2020 removed one of the last reliable sources of low-stakes daily contact. The rise of texting and social media replaced voice contact with the kind of asynchronous communication that, per the research, does not produce the bonding reciprocity that sustained friendship produces. Men, more than women, seem to rely on the structures that eroded; women more often build intentional, deliberate replacement structures that men do not.

The number has political valences, but the mechanism is not political. Men who identify as conservative and men who identify as liberal report friendlessness at similar rates. Rural men and urban men report friendlessness at similar rates. The consistent variables, per the WashU cross-national study released in March, are age (younger is lonelier), marriage status (unmarried is lonelier), and the eight-country comparison's structural finding that cultures with stronger extended-family networks — Korea, Turkey, Brazil — have lower loneliness rates than cultures that prize individual mobility. [4] The United States is an individual-mobility culture whose young men have lost the local structures that previously did the mobility work. The result is measurable.

What the paper reads in this, paired with the 53.1 fertility number running alongside it today, is a single system with two outputs. Young men who do not have friends are less likely to form stable romantic relationships. Young people who do not form stable romantic relationships are less likely to have children. The 23 percent drop in American fertility since 2007 and the 5x increase in friendless men since 1990 track the same curve. It is not that loneliness causes low fertility, or vice versa. It is that both are produced by the same structural reshaping of American young-adult life — the decoupling from the neighborhood, the job that is not located in a workplace, the screen that is the primary point of contact, and the compressed timelines that now separate the four traditional markers of adult life (leaving home, marriage, first job, first child) by a decade each.

The treatment, per Holt-Lunstad's later work, is not complicated. At least one reliable close relationship lowers mortality risk to match the non-lonely baseline. The mechanisms are specific: a person you can talk to when something hard happens, a person who will answer a phone call, a person who knows where you are most nights. The Pew data suggests men have slightly fewer of these than women but not zero. The 15 percent who say they have zero is the population the Surgeon General's advisory was written for. The paper's demographic-winter thread, which promotes today from embryonic to live with this piece, tracks what that 15 percent becomes in twenty years if the loneliness figures keep moving in the direction they have been moving. What they become is the demographic story. Loneliness is the first number. 53.1 is the second. The third has not been written yet.

-- NORA WHITFIELD, Chicago