The Friendship Recession: What the Data Actually Show About Shrinking Social Circles

TEST_YOURSELF · How well do you know this science?

1. 01 In the 2021 American Perspectives Survey, what share of Americans reported having no close friends at all?

   3% — essentially unchanged since 199012% — four times the 1990 figure30% — nearly a third of the country50% — half of all adults

   The Survey Center on American Life's May 2021 survey found 12% of Americans reported no close friends, versus 3% in 1990 — one of the clearest markers of what its authors called the friendship recession. source ↗

2. 02 According to Jeffrey Hall's 2018 research, roughly how many hours of shared time does it take to turn an acquaintance into a close friend?

   About 10 hoursAbout 50 hoursMore than 200 hoursTime doesn't matter — only shared values do

   Hall found a dosage curve: roughly 50 hours to move from acquaintance to casual friend, about 90 to friend, and more than 200 to close friend — with time spent hanging out and in meaningful conversation counting far more than hours merely shared at work. source ↗

3. 03 What did McPherson, Smith-Lovin and Brashears report when they compared Americans' core discussion networks in 1985 and 2004?

   Networks grew as the internet added new confidantsThe mean number of confidants fell by about a third, from 2.94 to 2.08Networks stayed the same size but shifted to coworkersOnly rural Americans lost confidants

   Comparing the 1985 and 2004 General Social Surveys, they reported the mean number of confidants dropped from 2.94 to 2.08, the modal answer went from three confidants to zero, and reported isolation appeared to nearly triple. source ↗

4. 04 What was Claude Fischer's 2009 critique of the famous 'shrinking networks' finding?

   That the decline was even worse than reportedThat the dramatic isolation increase was implausible and likely a survey artifactThat confidants don't matter for well-beingThat the 1985 survey never actually happened

   In American Sociological Review, Fischer argued the reported near-tripling of isolation was implausible given anomalies in the GSS data and contrary results elsewhere, and was likely an artifact of survey procedure — the authors replied, and the debate became a standing caution about measuring isolation. source ↗

5. 05 According to American Time Use Survey data cited by the US Surgeon General, how did Americans' daily time with friends change from 2003 to 2020?

   It rose slightly as remote work freed up eveningsIt stayed flat at about an hour a dayIt fell from about 60 minutes to about 20 minutes a dayIt fell only among people over 65

   Time-use analyses of the 2003–2020 American Time Use Survey, highlighted in the Surgeon General's 2023 loneliness advisory, show daily time with friends dropped from roughly 60 minutes to about 20 — a structural, society-wide decline in friendship time, not an individual one. source ↗