Nobody wants to quit social media first

Published 6 October 2026 · Updated 6 October 2026

You have probably thought about deleting Instagram. Maybe you did, for a week. Then you missed a party, a joke, a photo everyone else had seen, and you were back.

That is not weakness. Economists have a name for it: a collective trap.¹

One question changed the answer

In 2023, four economists asked US college students what it would take to switch off Instagram for four weeks while everyone else carried on. On average: $47.¹

Then they changed one thing. What if everyone taking part at your university switched off at the same time, you included? Now almost half of the Instagram users would pay to make it happen. Among TikTok users, 60% would.¹

Same students, same apps. The only difference was leaving alone.

Asked straight out, 57% of the Instagram users said they would rather live in a world without Instagram. For navigation apps, asked as a comparison, it was 5%.¹

Why everyone stays

The students who wished the app did not exist were asked why they kept using it. Three in four of the Instagram users named fear of missing out. One wrote: “I feel like if I stop using it, I will be completely out of the loop.”¹

That is the trap. While everyone else stays, leaving costs you. So everyone stays, including people who would rather the app were gone. The researchers stress that this is not a failure of willpower: a trap like this can catch perfectly rational people.¹

One caveat. This is one study of US college students, mostly women, in the summer of 2023, and the group switch-off it asked about was never actually run.¹

So don’t leave alone

If the cost of leaving is missing your people, take your people with you. You do not need your whole university. You need the five or ten people you would actually miss, and one other place to hear from them.

What you gain in mood is less certain: studies of people taking a break from social media find small or unclear effects.^{2, 3} What you keep is the people.

Try this

Write down those names this week. Ask them to try one other place together for a month, a group letter or a shared album, and see who writes back.

See the group letters. We make one of them, DearAll, and it is marked as ours.

Sources

Leonardo Bursztyn, Benjamin Handel, Rafael Jiménez-Durán and Christopher Roth, "When Product Markets Become Collective Traps: The Case of Social Media", American Economic Review 115(12), December 2025, 4105-4136, doi:10.1257/aer.20231468. Incentivised online experiments with US college students aged 18 to 30, recruited through the College Pulse panel in July to September 2023, during the summer break; the main samples were 371 TikTok users and 235 Instagram users from 365 universities, most of them women, and only students willing to join a deactivation study took part. Answers decided real payments, but each student had only a small chance of being the one whose answer counted, and the group switch-off needed two-thirds of a university to sign up, so it was never carried out. The published text is paywalled; figures are quoted from the authors' final version dated 15 November 2024 (home.uchicago.edu/bursztyn/CollectiveTraps_Nov2024.pdf), which thanks the AER editors and referees, and match NBER Working Paper 31771 (revised July 2024). https://doi.org/10.1257/aer.20231468
Hunt Allcott, Luca Braghieri, Sarah Eichmeyer and Matthew Gentzkow, "The Welfare Effects of Social Media", American Economic Review 110(3), 2020, 629-676, doi:10.1257/aer.20190658. A randomised experiment in which US Facebook users were paid to deactivate their accounts for the four weeks before the 2018 midterm elections (1,661 people in the main sample; only those who would leave for under $102 were randomised). A well-being index rose by 0.09 of a standard deviation, a small effect. Figures quoted from NBER Working Paper 25514 (revised November 2019), pp. 2, 4 and 23. https://doi.org/10.1257/aer.20190658
Laura Lemahieu, Yannick Vander Zwalmen, Marthe Mennes, Ernst H. W. Koster, Mariek M. P. Vanden Abeele and Karolien Poels, "The effects of social media abstinence on affective well-being and life satisfaction: a systematic review and meta-analysis", Scientific Reports 15, 2025, 7581, doi:10.1038/s41598-025-90984-3 (open access; preregistered). Ten studies with 4,674 participants: no significant effect of complete breaks on positive mood, negative mood or life satisfaction, and longer breaks did no better; the authors note important differences in how the studies were run. https://doi.org/10.1038/s41598-025-90984-3