Confidants per person — global data

The General Social Survey's confidant module, introduced in 1985, asks respondents to name the people with whom they have discussed "important matters" in the past six months. It then asks follow-up questions about each named person: their relationship to the respondent, how long they have known them, and whether they consider them a close friend. The module was designed to operationalize the sociological concept of the "strong tie" — a relationship sufficiently intimate to bear the transmission of sensitive personal information.

The module's strength is its specificity. By anchoring the question to actual disclosure behavior ("discussed important matters") rather than relational label ("who are your close friends"), it cuts through the definitional noise that plagues friendship count data. The confidant is not self-defined by the respondent — it is identified by what the respondent actually does, or reports doing, with the person. This makes the measure more cross-culturally consistent than friendship count data, though not fully immune to cultural variation in what "important matters" means.

The module's limitation is its reliance on recall and willingness to name. Some people have confidants they would not name to a survey researcher. Some name confidants they discuss relatively little with. The self-report problem is not eliminable, but the GSS data's longitudinal consistency suggests the measure is tracking something real across the decades it has been administered.

The McPherson paper's 2006 publication crystallized what the GSS data had been showing since 1985: between that year and 2004, the American confidant network underwent a structural collapse. Mean network size fell from 2.94 to 2.08. The percentage of Americans reporting zero confidants — no one with whom they discussed important matters — rose from roughly 10 percent to nearly 25 percent. The percentage reporting a network of one rose similarly.

The collapse was not evenly distributed. It concentrated among men, among less-educated Americans, among those who had recently relocated, and among those who had experienced significant life disruptions like divorce or job loss. The confidant network, the data showed, was a social asset that depleted under stress and did not automatically replenish. People who went through a major life disruption and emerged on the other side often emerged with fewer confidants than they had before, not more.

The paper also documented a shift in the composition of confidant networks. The percentage of respondents naming only their spouse as their confidant increased substantially over the period. The percentage naming a friend — a non-family member — declined. The confidant network became more family-centered and less friend-centered, narrowing the type of disclosure available to most Americans and increasing relational dependency on intimate partnerships.

The GSS confidant module was not readministered in consistent form after 2004, making the precise post-2004 trajectory difficult to track from that source. Other surveys have attempted to fill the gap. The Survey Center on American Life's 2021 AEI friendship report found that 12 percent of Americans reported having no close friends, up from 3 percent in 1990. It also found that 15 percent of men reported no close friends. These figures are not identical to the GSS confidant measure but are consistent with a continued trajectory of decline.

The Cigna loneliness surveys, conducted in 2018 and 2020, found that more than half of Americans reported sometimes or always feeling lonely, left out, or without meaningful social connection. The 2020 wave, conducted in the first months of the pandemic, showed a significant increase over 2018 — but the 2018 baseline was already alarming. The pandemic accelerated an existing trajectory rather than creating a new one.

Post-pandemic data from various sources suggests partial recovery in reported social connection, but the baseline to which the recovery is returning was already depleted relative to mid-20th century norms. The question of whether pandemic-related declines will prove permanent is not yet fully answerable.

The most consistent finding in confidant data across surveys and time periods is gender divergence. Women report more confidants than men in nearly every dataset examined. The divergence is not small: in the 2021 AEI survey, women were more than twice as likely as men to report having six or more close friends. Men were more than three times as likely as women to report having no close friends.

The divergence has widened over time. In the 1990 AEI data, 3 percent of men and 1 percent of women reported no close friends. By 2021, those figures were 15 percent for men and 10 percent for women. Men's confidant networks have deteriorated faster than women's, and from a lower baseline.

The mechanisms are multiple. Men are less likely to use friends as confidants, more likely to route emotional disclosure through their romantic partners, and more likely to organize friendship around shared activity rather than explicit emotional exchange. When the activity context disappears — the sports league ends, the job changes, the group chat goes quiet — men frequently lack the relational infrastructure to maintain the friendship absent its contextual scaffolding. Women are more likely to maintain friendships across contextual disruption through explicit relational investment.

Confidant networks have a lifecycle that differs significantly by life stage. Young adults typically report relatively high confidant counts — the concentration of peer relationships in educational institutions creates conditions favorable to close friendship formation. The late twenties and thirties tend to show decline, as career investment, partnership formation, and early parenthood consume the time and attention that friendship requires. Middle age is the nadir for confidant availability in most longitudinal data.

Late middle age and older adulthood show a more complex pattern. Retirement can produce either increase or decrease in confidants depending on whether it releases time for friendship investment or strips away the workplace social context that was providing it. Men who retire without close friends outside work frequently report sharp declines in social contact; women who retire often report increases in friendship investment.

The very old show the most severe confidant deficits, compounding health vulnerability at exactly the point of greatest need. Research on social isolation among the elderly consistently identifies confidant absence as one of the strongest predictors of health decline, cognitive deterioration, and mortality. The confidant deficit is not evenly distributed across the lifespan, but its worst consequences are concentrated in the years when the body can least afford it.

European data on confidants is less systematic than the American GSS data, but the available evidence suggests patterns that partially parallel and partially diverge from the American case. The Eurobarometer's social isolation module, administered irregularly across EU member states, finds high rates of reported loneliness in Southern and Eastern European countries despite cultural stereotypes of Mediterranean warmth — suggesting that the aggregate friendship culture and the distribution of confidant availability within it can diverge substantially.

United Kingdom data from the Office for National Statistics loneliness survey and the Campaign to End Loneliness research shows that approximately 9 percent of UK adults report feeling "always" or "often" lonely, with concentration among young adults (16–24) and the elderly (75+). The UK government's appointment of a Minister for Loneliness in 2018 — itself a measure of the scale of the problem — generated significant comparative data collection that documented confidant deficits across demographic groups.

German data, consistent with the German friendship culture discussed in the previous article, shows lower reported confidant counts than the UK or the United States but also lower reported loneliness — consistent with the hypothesis that deeper but fewer relationships can provide equivalent or superior wellbeing outcomes to larger but shallower networks.

Asian data on confidants is the least systematically collected of any major region. Japan, where the phenomenon of hikikomori (social withdrawal) and kodawari (dying alone) have received sustained public attention, has generated the most detailed national data. The NHK's 2020 survey estimated that over 600,000 adults in Japan had been hikikomori for more than six months, with a substantial overlap with people reporting no confidants.

South Korean data shows high rates of reported loneliness among both young adults (amplified by extreme academic competition) and the elderly (amplified by rapid urbanization that separated families). Chinese data is limited by survey access constraints and the complexity of defining confidant relationships in a culture where the concept of guanxi (relationship networks) structures social life differently from Western friendship norms.

The general finding from Asian data, where available, is that the relationship between institutional participation and confidant availability is stronger than in Western societies — the confidant is more likely to have been met through a structured institutional context (school cohort, work unit, neighborhood association) and less likely to have been deliberately sought as an individual relational investment.

The GSS data's most sociologically significant finding may be the growth of the spouse-only confidant pattern. In 1985, approximately 5 percent of Americans named only their spouse as their confidant. By 2004, that figure had grown substantially, particularly among men. By the 2021 AEI survey, married men reported significantly more close friends than unmarried men — suggesting that the social connection infrastructure increasingly routes through marriage for men in ways it does not for women.

This pattern has distributional consequences. Men who are married have access to a confidant; men who are not married and have no close friends have access to none. The growth of the unmarried male population — particularly in middle age — combined with the decline of male friendship means a growing segment of the male population navigates significant life events and personal crises with no one to talk to.

The spouse-as-sole-confidant pattern also creates relational fragility. When the marriage ends — through divorce, separation, or death — the man who has organized his entire disclosure life around his partner loses his confidant network in the same moment he loses his partnership. The emotional and health consequences of spousal bereavement for men are substantially worse than for women, and the confidant structure helps explain why: women typically retain a close friendship network outside the marriage; men often do not.

The health research on confidant availability is unambiguous in its direction and substantial in its effect sizes. Holt-Lunstad's meta-analyses — the most comprehensive synthesis of social connection and mortality data — find that social isolation increases premature mortality risk by approximately 26–32 percent, with effect sizes comparable to smoking fifteen cigarettes a day and larger than the effects of obesity or physical inactivity. Zero confidants is not a social comfort matter. It is a public health emergency distributed in individual lives.

The mechanisms connecting confidant absence to health outcomes include behavioral (no one to monitor health behavior, encourage treatment seeking, or accompany to medical appointments), physiological (social isolation activates chronic stress responses that damage immune and cardiovascular function), and psychological (absence of emotional processing of stress events leads to rumination, depression, and anxiety). The confidant relationship is therefore health-protective through multiple pathways simultaneously, making its absence multiply harmful.

Research on childhood and adolescent confidants is a distinct literature with its own findings, but its relevance to adult confidant data is significant. Children who have confidant relationships in childhood — typically with one or two close friends rather than large peer groups — develop relational skills and attachment patterns that facilitate confidant formation in adulthood. Children who are socially isolated in childhood are substantially more likely to be isolated in adulthood, not simply because isolation compounds but because the skills that close friendship develops are not acquired.

The adolescent confidant relationship is particularly significant. Research by Niobe Way and others has documented that the quality of adolescent friendship — including the ability to confide, to receive and give emotional support, and to maintain closeness under social pressure — predicts adult relational outcomes decades later. The current concern about adolescent social media use is therefore not only about present wellbeing but about the skill and relational infrastructure formation that the adolescent period normally provides.

Confidant data in the United States shows significant race and class stratification that aggregate national figures obscure. Black Americans report confidant networks that are smaller in absolute size than white Americans' but are more likely to include non-family members — friends, church community members, and neighbors — suggesting a different architectural pattern rather than simple deficit. Hispanic Americans show higher confidant counts than both, consistent with extended family and community norms.

Class stratification runs in complex directions. Working-class Americans report smaller confidant networks in most surveys, but the relationships within them are more likely to be geographically proximate and practically oriented — the neighbor who could be called in a crisis. Professional-class Americans report larger networks but are more geographically dispersed and show more dependency on deliberate scheduling for maintenance. The practical availability of the confidant in a genuine emergency may not correlate simply with reported count.

The confidant decline is structural and will not be reversed by individual effort alone, though individual effort matters. The structural interventions that research suggests would have the largest effect are: rebuilding the institutional contexts (civic associations, religious participation, neighborhood organizations) that historically provided the preconditions for confidant formation; redesigning built environments to provide third places and walkable public space that enables repeated casual social contact; and reducing the residential and labor mobility that disrupts established networks before they can reach confidant depth.

None of these are fast or politically straightforward. The short-term palliatives — social prescribing programs, loneliness helplines, organized social events for isolated individuals — address symptoms without addressing causes. They are worth doing, but they cannot substitute for the structural conditions that make confidant formation possible in the first place. The confidant count decline is a measure of the depletion of those conditions over several decades. Reversing it requires rebuilding them — a generation-scale project, not a program.

---

1. McPherson, Miller, Lynn Smith-Lovin, and Matthew E. Brashears. "Social Isolation in America: Changes in Core Discussion Networks over Two Decades." American Sociological Review 71, no. 3 (2006): 353–375.

2. Cox, Daniel A. "The State of American Friendship: Change, Challenges, and Loss." Survey Center on American Life, American Enterprise Institute, June 2021.

3. Holt-Lunstad, Julianne, Timothy B. Smith, and J. Bradley Layton. "Social Relationships and Mortality Risk: A Meta-Analytic Review." PLOS Medicine 7, no. 7 (2010): e1000316.

4. Holt-Lunstad, Julianne, Timothy B. Smith, Mark Baker, Tyler Harris, and David Stephenson. "Loneliness and Social Isolation as Risk Factors for Mortality: A Meta-Analytic Review." Perspectives on Psychological Science 10, no. 2 (2015): 227–237.

5. Cacioppo, John T., and William Patrick. Loneliness: Human Nature and the Need for Social Connection. New York: W. W. Norton & Company, 2008.

6. Cigna Corporation. "Loneliness and the Workplace: 2020 U.S. Report." Bloomfield, CT: Cigna, 2020.

7. Office for National Statistics. "Loneliness — What Characteristics and Circumstances Are Associated with Feeling Lonely?" Newport: ONS, 2018.

8. Murthy, Vivek H. Together: The Healing Power of Human Connection in a Sometimes Lonely World. New York: Harper Wave, 2020.

9. Way, Niobe. Deep Secrets: Boys' Friendships and the Crisis of Connection. Cambridge, MA: Harvard University Press, 2011.

10. U.S. Department of Health and Human Services. "Our Epidemic of Loneliness and Isolation: The U.S. Surgeon General's Advisory on the Healing Effects of Social Connection and Community." Washington, DC: Office of the Surgeon General, 2023.

11. Fischer, Claude S. To Dwell Among Friends: Personal Networks in Town and City. Chicago: University of Chicago Press, 1982.

12. Christakis, Nicholas A., and James H. Fowler. Connected: The Surprising Power of Our Social Networks and How They Shape Our Lives. New York: Little, Brown and Company, 2009.