Friendship
The relationship research neglected for decades
Description
Relationship research for most of the twentieth century focused on romantic partnerships, marriage, and parent-child bonds. Friendship, despite being one of the most common and consequential forms of human connection, received relatively little scholarly attention until the past two decades. The neglect reflected a specific cultural hierarchy — romantic love was serious, family was primary, and friendship was treated as pleasant but secondary, a relationship you could afford to have only after the more important ones were working. This hierarchy turns out to be both historically specific and empirically wrong. Friendship is more consequential for adult wellbeing than most people realize, and the decline in friendship quality over recent decades is one of the underappreciated social patterns of modern life.
The research that has accumulated since the 2000s — work by Robert Putnam, Sherry Turkle, Lydia Denworth, Julianne Holt-Lunstad, and others — documents both the importance of friendship and its contemporary erosion. Adults have fewer close friends than previous generations did. Men in particular report striking declines in friendship networks, with the share of men reporting no close friends having risen substantially since the 1990s. The specific features of modern adult life — long work hours, geographic mobility, smartphone-mediated social time, suburban geographies — have systematically reduced the conditions under which deep friendships develop and persist.
What friendship actually does for wellbeing, why it has eroded, and what can be done about it are the questions that have increasingly preoccupied researchers and writers in this area. The answers are specific enough to be useful and counter-intuitive enough to matter. Friendship is not a soft topic; the research suggests it is as consequential for mortality, mental health, and life satisfaction as any other relational variable studied, and the current culture has been losing ground on it without fully noticing.
● The question we're asking: what does friendship actually provide, and why has it become harder to sustain?
● What we'll see: the research base, the specific mechanisms, the current erosion, and what it would take to reverse the trend.
Table of contents
01The research base
The scientific study of friendship began gaining momentum in the 1980s and accelerated in the 2000s. Julianne Holt-Lunstad's meta-analyses of social relationships and mortality, published starting in 2010, consistently showed that weak social connections predict mortality with effect sizes comparable to or exceeding those of smoking, obesity, and physical inactivity. The specific effect of friendship, distinct from family or romantic partnership, turns out to be substantial. Adults with strong friend networks live longer, recover better from illness, and report higher wellbeing than adults with comparable romantic and family relationships but weaker friendships.
The mechanisms are multiple. Close friends provide emotional support that reduces physiological stress responses. They provide practical help during specific life events. They provide social integration that reduces the isolation associated with depression and cognitive decline. They provide access to specific information, opportunities, and perspectives that strengthen decision-making. Each of these mechanisms has been documented separately in different research traditions, and together they explain why friendship effects are as large as they are. The relationships are not just pleasant; they produce measurable physiological and behavioral effects that aggregate into the mortality findings.
02The specific mechanisms
Close friendships operate through specific mechanisms that distinguish them from other relationships. Voluntary chosen connection is one — friendship is the one major adult relationship type that is neither biologically given (family) nor formally contractual (marriage, work). This voluntarism gives friendship a specific quality: both people are in the relationship because they actively choose to be, which produces a specific kind of affirmation that other relationships cannot fully provide. The friend who shows up when they do not have to is providing something the family member or spouse who has obligations cannot provide in the same way.
Shared activity is another mechanism. Close friendships typically develop through repeated shared activities — work projects, sports, shared homes, raising children at similar ages, attending the same institutions. The shared activity provides the specific context in which the emotional connection develops. Adults who change jobs, neighborhoods, or life stages often lose friendships because the shared-activity context that sustained them disappears. This is why adult friendship is harder than childhood friendship; the institutional contexts that brought people together (schools, playgrounds, neighborhoods) are less stable in adult life.
03The current erosion
American adults report substantially fewer close friends now than in previous decades. The General Social Survey and similar instruments show consistent declines: the share of Americans reporting no close friends has roughly tripled since the 1990s. The share of men reporting no close friends has risen most steeply. The average number of close friends has fallen across demographic groups, though the decline is not uniform. Educated urban adults generally have more friends than their rural and less-educated counterparts, but everyone has fewer than their counterparts did thirty years ago.
The reasons for the decline are multiple and interacting. Work hours have lengthened and work intensity has increased for professional adults, leaving less energy for friendship maintenance. Geographic mobility has increased, separating people from the institutions and neighborhoods where friendships developed. Suburban geographies have reduced the physical proximity that sustains casual repeated contact. Smartphones have filled the interstitial time that previously produced incidental social contact. The specific combination of these factors has made friendship maintenance measurably harder than it was a generation ago, even as the underlying human need for friendship has not changed.
04What it would take to reverse the trend
Reversing the friendship decline is partly individual and partly structural. At the individual level, the research suggests specific practices that improve friendship maintenance: regular unstructured time with specific friends, repeated activity together rather than only one-off events, willingness to be the initiator of contact rather than waiting for reciprocation, specific openness about emotional topics rather than keeping conversations transactional. Each of these is learnable, and the specific skill of making and keeping friends is one that can be improved through deliberate practice.
At the structural level, the conditions that support friendship are substantially about geography and time. Adults with long commutes, demanding jobs, and multiple caregiving obligations struggle to maintain friendships regardless of their intentions. Policies that reduce work hours, support family caregiving, and create more walkable communities with public third places (cafés, parks, community centers) indirectly support friendship formation by creating the conditions under which it happens. Finland, which has invested in public spaces and community infrastructure, reports higher friendship density than most developed countries, and some of the difference is plausibly attributable to these structural features.
05Conclusion
Friendship is one of the most consequential human relationships and one of the most neglected in research, cultural attention, and individual life-planning. The empirical evidence shows that close friendships produce substantial wellbeing effects, comparable to or exceeding those of other relational variables studied. The current erosion of friendship quality in modern societies — driven by specific features of contemporary work, geography, and technology — is a measurable loss that is only partially compensated by the social connections people do maintain. Taking friendship seriously, at both individual and structural levels, would substantially improve the wellbeing outcomes of the populations most affected by the erosion.