Americans Spend Just 35 Minutes a Day With Other People. The Decline Started in 2003, Not 2020.

By Courtney Turrin · September 10, 2026 · 8 min read read

Category: Health & Wellness

Americans spend less time socializing face-to-face now than in 2003. This decline predates the pandemic and is linked to loneliness. The cure? Correcting a maladaptive cognitive bias.

This is part of a four-part September series on suicide awareness and prevention, running through National Suicide Prevention Awareness Month. It discusses depression and suicidal ideation directly. If you are struggling, dial 988 to reach the Suicide and Crisis Lifeline. It's a three-digit number, available by call or text, 24 hours a day, for free and confidential support.

Face-to-face socializing in America has been declining for more than two decades, well before the pandemic came into play. Research shows that the link between loneliness and suicidal thoughts is real, but weaker than the "loneliness epidemic" headlines suggest. The intervention that actually works isn't more socializing; it's correcting a cognitive bias.

In 2025, 30% of Americans engaged in face-to-face socializing on an average day. In 2015, that figure was 38%. Among people who did socialize, the average time spent doing it fell from 41 minutes a day to 35 over the same 10-year period. This comes from the Bureau of Labor Statistics' American Time Use Survey (ATUS), which has tracked how Americans spend their days, minute by minute, since 2003 (Figure 1).

[datawidget key="atus-socializing-decline"]

The instinct is to blame the global pandemic that locked us all down in 2020. The data shows that this trend goes back much further than that.

The Decline in Face-to-Face Socializing Predates the Pandemic

Enghin Atalay, an economist at the Federal Reserve Bank of Philadelphia, tracked the proportion of Americans' free time spent with people from outside their own household from 2003 through 2020. That figure fell from 21.9% in 2003 to 17.3% in 2019, a significant decline that happened before COVID-19 existed. The pandemic then cut it further, to 12.3% in 2020, but the floor had already been dropping for 16 years before that (Figure 2).

[datawidget key="time-with-other-households"]

Sociologist Eric Klinenberg, director of NYU's Institute for Public Knowledge, referenced the same survey in a conversation with The Ringer, noting that face-to-face socializing among American teenagers in particular fell by roughly half between 2003 and 2022, compared with a 30% decline among adults. Something changed in the texture of American life well before anyone had heard of a novel coronavirus, and it changed fastest for teens.

What that something was is contested. Smartphone adoption, the built environment, the shift to two-earner households, and the decline of civic and religious institutions all overlap with the timeline. The ATUS survey itself doesn't isolate a cause. It only documents that Americans have spent less and less time with other people over the past 20 years.

Loneliness Is an Evolutionary Warning System, and It's Working Correctly

The late psychologist John Cacioppo spent much of his career arguing that loneliness should be thought of the way hunger is: as a meaningful signal that evolved because it was adaptive. Social species that felt bad when isolated were more likely to seek the group, and the group supported survival and improved the likelihood of passing along one’s genes. Loneliness, in this framing, is the aversive feeling doing exactly what it evolved to do. It just doesn't distinguish well between "I am in physical danger without my tribe" and "I skipped Sunday dinner with my friend group for three weeks running."

Thwarted Belongingness Predicts Suicidal Thoughts, but the Belief that One is a Burden Has an Even Greater Effect

Thomas Joiner's interpersonal theory of suicide names two specific cognitive states as central to suicidal ideation: perceived burdensomeness (the belief that one's existence is a cost to others) and thwarted belongingness (the belief that one does not belong, is not connected, is on the outside). The theory predicts that suicidal desire is highest when both are present and both feel permanent.

A 2017 meta-analysis by Carol Chu and colleagues, published in Psychological Bulletin, tested this across 122 samples. The interaction between thwarted belongingness and perceived burdensomeness was significantly associated with suicidal ideation.

What gets lost in translation to the popular press is the size of the effect and an asymmetry that complicates it. The average effect size for the interaction was weak, under 0.20 in several analyses of the same dataset. When the two components are compared separately, perceived burdensomeness carries more predictive weight than thwarted belongingness, the loneliness half of the pair.

So "loneliness causes suicidal thoughts" is an oversimplification. A more accurate framing is: thwarted belongingness is one measurable contributor to suicidal ideation, its effect is modest on its own, and it does most of its work in combination with the belief that one is a burden.

Solitude and Loneliness Are Not the Same Measurement

The ATUS measures minutes spent with other people who are not part of your household. It cannot measure whether the time someone spends alone was chosen or imposed, restorative or corrosive. One person can log hours alone by preference and feel entirely connected. Another person can be surrounded by coworkers, roommates, or family all day and feel lonely.

Atalay flags this directly, theorizing that more time alone predicts "a deterioration in social ties" as a population-level pattern, not a diagnosis of any individual's inner life. We can measure the minutes someone spends alone, but what those minutes mean to the person living them is a separate question.

The Loneliness Intervention That Works Isn't the One Most Advice Assumes

Given all of the above, the obvious prescription is "spend more time with people." The research on what actually reduces loneliness doesn't fully support that as the best intervention.

Christopher Masi and colleagues ran the first meta-analysis of loneliness-reduction interventions in 2011, sorting fifty years of prior studies into four strategies: improving social skills, enhancing social support, increasing opportunities for social contact, and addressing maladaptive social cognition (the distorted, threat-oriented way a lonely person tends to interpret other people's behavior toward them). Across randomized controlled trials, the average effect size across all intervention types was weak, about 0.2. The fourth category, correcting the cognitive bias rather than adding social contact, produced an average effect size of roughly 0.6, about three times larger (Figure 3).

[datawidget key="loneliness-intervention-effects"]

The researchers were careful about the limits of their findings. Only four randomized trials tested the social-cognition approach directly, which is a small enough number that the authors called for independent replication. Their paper has been cited heavily in the fourteen years since, and later reviews have generally reaffirmed the pattern, but the caveat is real: this is the best current evidence, not a closed case.

The underlying logic connects back to Cacioppo's signal-system framing. If loneliness evolved as a threat-detection system, then chronic loneliness may function like a hair-trigger alarm: it starts reading neutral or even friendly social cues as rejection, which makes a lonely person withdraw further. This in turn produces more of the isolation the alarm was meant to protect against. Under that model, adding more social contact without addressing the misfiring alarm is a bit like adding more smoke detectors to a house where the existing one goes off every time someone makes toast. More detectors don’t fix an alarm that’s miscalibrated.

The State Is Not the Trait

Thirty-five minutes a day of face-to-face interaction describes a population-level shift towards increasing isolation in American social life since 2003. It says much less about what any individual is experiencing.

For those who are seeking more and higher quality social connection, the evidence points away from the standard prescription. Filling a calendar with more appointments addresses the wrong layer for a meaningful share of people. Correcting the threat-oriented lens chronic loneliness creates, the tendency to read a friend's slow reply or a coworker's short answer as rejection, was three times more effective than other interventions. 

Neuroscientist Stephanie Cacioppo, John Cacioppo's wife and longtime research partner, describes how to use this technique in her group's clinical review of these interventions: treat a negative read on someone else's behavior as an unverified guess rather than a settled fact. A friend's slow reply becomes one data point, not a verdict, until it's weighed against equally plausible explanations: a demanding week, a dead phone, or a mood that has nothing to do with you.

The skill being trained is catching the jump from ambiguous signal to certain rejection while it's happening, then deliberately generating the neutral explanation before deciding which one the evidence actually supports. It is a small habit that’s easier to implement than the generic “just put yourself out there” advice, and it also has a stronger track record.

If you or someone you know needs support, dial 988 to reach the Suicide and Crisis Lifeline by call or text, 24 hours a day. In the Richmond region, the Richmond Behavioral Health Authority also operates a crisis line at 804-819-4100.

Related: Listen and watch

Erin Gallagher turned personal loss into the Jay Walkers Challenge, a Richmond effort built around getting people together in person. Her conversation with Channel RVA is below.

[embed url="https://www.youtube.com/watch?v=iN5r1Aob7OA"]