ENDE

Attention & Time

How much time do people spend with friends?

In 2003, Americans averaged about 60 minutes a day with friends. By 2020 that had fallen to 20 minutes, a two-thirds drop that predates COVID by more than a decade. Partial recovery continued through 2025, bringing the average back to around 35 minutes a day, still well below the pre-decline baseline. Time spent alone rose by the equivalent of 24 hours a month over that period. In 2023, the US Surgeon General formally declared a loneliness epidemic, citing health risks comparable to smoking 15 cigarettes a day.

Roughly 1.46 minutes every hour.

35 minper person/day
-42%change since 2003
≡ 15cigarettes/day (health risk)

Source: BLS American Time Use Survey 2020; US Surgeon General Advisory on Loneliness 2023.
Last checked: 18 October 2026

This is not a soft social issue - declining friend time is a measurable health and mortality risk

In 2023, US Surgeon General Vivek Murthy issued a formal advisory on loneliness and isolation, comparing its mortality risk to smoking 15 cigarettes a day. That comparison comes from a 2010 meta-analysis by Julianne Holt-Lunstad covering 308,849 people, which found that social isolation was associated with a 29% increase in mortality risk. Loneliness is not an emotion that passes. It is a chronic physiological stressor that raises cortisol, drives up inflammation markers, and compromises immune function.

The time-use data makes this concrete. American adults now average just 35 minutes a day with friends, down from around 60 minutes in 2003. For people under 35, the drop is steeper still. The hours that left in-person contact have not been replaced by equally satisfying digital connection: text-based and passive social media use consistently correlates with higher loneliness, not lower. Seeing a friend's life on Instagram is not the same as being with them.

Harvard's Adult Development Study followed hundreds of people for over 80 years. The single strongest predictor of late-life happiness and cognitive health was not wealth, career success, or physical fitness. The quality and warmth of close relationships predicted late-life outcomes most consistently, and that result has held across every cohort the study tracked.

Time with friends over time (minutes/day)

BLS time-use data shows Americans lost two-thirds of their daily friend time from 2003 to 2020, declining from 60 to 20 minutes per day. Recovery continues slowly: 35 minutes/day in 2025, up from 30 in 2024. The trend predates COVID and correlates with the rise of smartphone social media.

2003
60 min
~2012
45 min
2019
35 min
2020
20 min
2024
30 min
2025
35 min
0.00173552692003201220192020202420252028ESTIMATED604535203035~25
YearMinutes/day with friendsContext
200360 minPre-smartphone; in-person social baseline
~201245 minSmartphone era begins
201935 minDecline accelerates; social media dominant
202020 minCOVID-19 + pre-existing trend
202430 minPost-COVID partial recovery; still well below 2003
202535 minGradual recovery from COVID low; 30% participate daily (down from 38% in 2015)
2028 (forecast)25 minStructural loneliness trends continue; screen substitution persists

Key statistics: the loneliness epidemic

Time Americans spend with friends fell from ~60 minutes/day (2003) to ~20 minutes/day (2020), a two-thirds reduction

Social isolation increased by 24 hours per month between 2003 and 2020 (BLS ATUS)

The US Surgeon General declared a loneliness epidemic in 2023, calling it equivalent in health impact to 15 cigarettes/day

Time spent alone increased from 43.5% of leisure time (2003) to 49.7% (2022)

The decline in friend time is steepest among young adults (18-34), men, and lower-income groups

BLS ATUS 2025: Americans spent 35 minutes/day socializing and communicating, up from 30 min in 2024 but still below 41 min in 2015

Time with friends vs. time on social media, today

The starkest illustration of the numbers: Americans now spend more than four times as much of their day on social media as in face-to-face contact with friends.

Person-years with friends today
18minutes so far today6,300minutes this year
35 min/day, face-to-face friend time
vs.
Person-years on social media today
80minutes so far today28,620minutes this year
159 min/day, digital social platforms

The friendship recession: why modern life is crowding out connection

The great withdrawal

The ATUS data goes back to 2003, and the story it tells started well before COVID. From 2003 to 2019, Americans were steadily cutting time with friends, non-household family, and neighbors, while total leisure time stayed roughly flat. The hours were not disappearing into work. They were shifting from social leisure to solitary leisure, and increasingly to screens. COVID pushed the trend sharply lower in 2020, and the 2025 rebound has not closed the gap.

The paradox of digital connection

The sharpest decline in face-to-face friend time happened between 2010 and 2020, the same decade smartphones became near-universal. That timing is suggestive, though it does not settle causation. Urbanization, longer commutes, and career pressure were all shifting simultaneously. Jonathan Haidt's 2024 book "The Anxious Generation" makes the case for a direct link between social media and the withdrawal from in-person life, particularly among teenagers. The ATUS data fits that argument. It does not prove it.

The research behind the friendship recession and loneliness epidemic

YearFindingValueSource
2003BLS ATUS 2003: Americans spend ~60 minutes/day socially with friends; time with non-household family ~40 min/day60 minutes/day with friends (2003)U.S. Bureau of Labor Statistics
2012BLS ATUS 2012: friend time declining; early smartphone era; approx. 45 min/day with friends45 minutes/day with friends (2012 est.)U.S. Bureau of Labor Statistics
2019BLS ATUS 2019: ~35 min/day with friends; time alone 48.7% of leisure35 minutes/day with friends (2019)U.S. Bureau of Labor Statistics
2020BLS ATUS 2020: friend time collapses to ~20 min/day; COVID-19 effect; social isolation +24 hours/month vs 200320 minutes/day with friends (2020)U.S. Bureau of Labor Statistics
2022Post-COVID partial recovery; time with people outside household 14.3% (vs 21.9% in 2003)partial recovery post-COVIDU.S. Bureau of Labor Statistics
2025BLS ATUS 2025: 35 min/day socializing and communicating, gradual recovery continues; 30% participate daily (down from 38% in 2015)35 minutes/day socializing (2025)U.S. Bureau of Labor Statistics

From Putnam's "Bowling Alone" to the Surgeon General's advisory: how we documented the loneliness crisis

  1. 2003BLS ATUS begins; Americans spend ~60 min/day with friends
  2. 2010iPhone mass adoption; social media begins scaling; friend time begins accelerated decline
  3. 2019~35 min/day with friends, about 25 minutes less than in 2003
  4. 2020COVID: friend time collapses to ~20 min/day; social isolation peaks
  5. 2023US Surgeon General declares loneliness epidemic; equivalent health risk to 15 cigarettes/day
  6. 2025BLS ATUS 2025: friend time rises to 35 min/day, continued recovery from COVID low; 30% participate daily vs 38% in 2015

In perspective

The average American now spends barely more time with friends each day than one half-hour sitcom slot lasts, down from a full hour in 2003.

At 35 minutes a day with friends and roughly 159 minutes on social media, Americans give digital platforms about four and a half times more of their day than the people they know in person.

Frequently asked questions

How much time does the average American spend with friends per day?
According to BLS ATUS data, Americans averaged about 60 minutes a day with friends in 2003, falling to around 20 minutes by 2020. Partial recovery through 2025 brings the current average to about 35 minutes a day. The decline started well before COVID: the data shows a steady downward trend from 2003 onward.
Is the decline in friend time related to increased screen time?
The evidence points to correlation, not settled causation. Face-to-face friend time started falling before smartphones were widespread, which makes a simple substitution story hard to sustain. The acceleration after 2010 does line up with rising social media use, but longer commutes, urban sprawl, and changing work patterns were also shifting in the same direction. Both things can be partly true.
What are the health effects of spending less time with friends?
The US Surgeon General's 2023 Advisory on Loneliness put social isolation's mortality risk on par with smoking 15 cigarettes a day. That figure comes from a 2010 meta-analysis by Julianne Holt-Lunstad covering more than 300,000 people. The associated conditions include higher rates of depression, anxiety, cardiovascular disease, and faster cognitive decline in older adults.

Cite this rate

AnythingCounter derived rate, last checked October 2026: about 35 minutes per day.

Method: Based on U.S. Bureau of Labor Statistics 2025, shown as a daily rate.

Suggested citation: AnythingCounter, "Time spent with a friend (per person)," October 2026, https://anythingcounter.com/time-with-friends-statistics

Primary source: U.S. Bureau of Labor Statistics.

How this number is calculated, and how much to trust it

The current live counter uses the 2025 BLS ATUS estimate of ~35 minutes/day with friends, neighbours, and acquaintances – continued recovery from the 2020 low of 20 min/day. This is a person-level measure, not a global rate. The counter shows accumulation for a single person since midnight: seconds elapsed × 35/1,440. The historical decline from 60 min (2003) to 20 min (2020) is drawn directly from BLS ATUS annual data; the 2025 value represents continued recovery, though still well below the 2003 baseline.

Time-use figures come from the US Bureau of Labor Statistics American Time Use Survey (ATUS), which has surveyed a nationally representative sample of Americans annually since 2003. The ATUS covers both weekdays and weekends across age groups and is the primary dataset used here. International context draws on the OECD Time Use Database and data from the UK Office for National Statistics. The 2025 figure is the most recent full-year ATUS release; the counter uses it as the baseline until updated data is published.

Sources: see below. More detail on our methodology page.