Straight answers
Frequently asked questions
Short, direct answers to what people actually ask about social media harm — each one backed by the same evidence documented across this site. Where the honest answer is "it's complicated" or "nobody knows," it says so.
The basics
Is Facebook actually bad for you?
On average, yes — and the strongest evidence comes from randomized experiments, not opinion.
When economists paid 2,743 people to deactivate Facebook for four weeks, their subjective well-being improved by an amount equal to 25–40% of the effect of clinical depression treatment, their political polarization decreased, and they voluntarily kept using Facebook less after the study ended.[evidence] The effects across this literature are small to moderate on average and largest for heavy users — but the direction is consistent across labs, countries, durations, and methods. See the full evidence on leaving →
Is Instagram bad for teenagers?
Meta's own researchers concluded it was, and the company shipped it anyway.
A March 2020 internal slide, prepared by Meta's researchers for Meta's executives, stated: "We make body image issues worse for one in three teen girls." Another found that "teens blame Instagram for increases in the rate of anxiety and depression" — unprompted and consistently across groups. Randomized deactivation experiments published in 2025 found the largest benefits for women under 25, the same group Meta's own research had flagged five years earlier.[evidence] Read the Instagram case file →
Is TikTok addictive?
By TikTok's own internal math, a new user is addicted in about 35 minutes.
Documents unsealed in Kentucky's lawsuit in October 2024 show TikTok internally estimated addiction at roughly 260 videos — about 35 minutes. Its own research linked compulsive use to "loss of analytical skills, memory formation, contextual thinking, conversational depth, empathy, and increased anxiety." An internal document about its teen screen-time tool stated the quiet part plainly: "Our goal is not to reduce the time spent."[evidence] Read the TikTok case file →
Why is my feed full of things that make me angry?
Because anger holds attention, and attention is what these companies sell.
When Facebook introduced emoji reactions, its ranking algorithm weighted the "angry" reaction five times more heavily than a like — systematically pushing inflammatory content into more feeds. Across 27 studies covering 4.8 million posts, each additional moral-emotional word in a post increased sharing by about 13%. An internal 2021 Facebook presentation put it directly: "Our algorithms exploit the human brain's attraction to divisiveness."[evidence] See how the machine works →
How much of social media is bots and fake accounts?
Nobody knows exactly — and the platforms hold the data needed to check.
What is documented: automated traffic made up 51% of all web traffic in 2024, the first time bots passed humans in a decade of measurement. Facebook removed about 3.5 billion fake accounts in 2025 alone, and in several years has banned more fake accounts than it had monthly active users. Meta's last public estimate put fakes at roughly 5% of monthly active users — and it has since stopped reporting Facebook's monthly active users at all.[evidence] Read the bots case file →
Did the companies know they were causing harm?
Yes — according to their own documents, repeatedly.
Meta's researchers documented the teen body-image findings in 2020. That same year Meta ran a deactivation study code-named Project Mercury, which found that quitting Facebook and Instagram reduced depression, anxiety, loneliness, and social comparison — then halted the project and never published it. In April 2020, Mark Zuckerberg personally vetoed a proposed ban on plastic-surgery-style camera filters, dismissing staff and outside-expert concerns as "paternalistic." Meta disputes the characterization of these documents, calling them "cherry-picked quotes and misinformed opinions."[evidence]
Getting out
Does quitting social media actually help?
Yes, measurably — and faster than most people expect.
A randomized trial at the University of Bath found significant improvements in well-being and significant reductions in depression and anxiety after just one week off. A University of Pennsylvania trial found that capping social media at about 30 minutes a day cut loneliness and depression within three weeks, with the most depressed participants improving the most. Meta-analyses put the average effect at g = 0.17–0.25 — small on average, larger for heavy users.[evidence] See what you get back →
How do I delete my Facebook account?
Download your data first, then request deletion — and know you have about 30 days to change your mind.
Every platform will hand you a full archive of your photos, posts, messages, and contacts; take it before you do anything else. Facebook and Instagram both apply a roughly 30-day grace window in which logging back in cancels the deletion. Deleting Facebook also removes Messenger. Follow the step-by-step guide, with every official link →
Is deactivating better than deleting?
For most people, yes — deactivation is the test drive.
Deactivating hides your profile but restores everything the moment you log back in, and it lets you keep using Messenger so nobody loses the ability to reach you. Deletion is permanent. The research suggests the test drive tends to stick on its own: in the largest randomized deactivation experiment, participants voluntarily kept using Facebook less after the paid month ended.[evidence]
What should I use instead of Facebook?
Replace each job separately instead of looking for one replacement.
For messaging, Signal is the strongest pick — nonprofit-run, end-to-end encrypted by default, no ads or data mining. For news and interests, an RSS reader (Feedly, Inoreader, NetNewsWire) gives you a chronological feed you control. For communities, Discord or Mastodon; for events, Partiful or a plain group text; for family photos, shared albums in Google Photos, iCloud, or Ente. See all the alternatives by job →
How much time does social media actually take?
About 2 hours 15 minutes a day — roughly 51 waking days a year.
That is the reported average, and self-reports typically undercount actual use by 30–50%, so the real figure is likely higher. At that rate the habit costs about 821 hours annually. People who deactivated Facebook in a randomized experiment got roughly an hour a day back immediately, redistributed to in-person socializing, physical activity, and sleep.[evidence]
Everyone I know is on Facebook. How do I leave without losing touch?
Move the ten to twenty people you actually talk to, not all 500 friends.
Text your closest people individually with a small, concrete ask: "I'm using Facebook less — would you message me on Signal? Takes two minutes to install." Keep Messenger during a bridge period, checked once a day, while real conversations move. You are not asking anyone to make a statement or delete anything — just to message you somewhere else. More on the network-effect problem →
The honest objections
Isn't this just a moral panic, like TV or video games?
It's a fair challenge — and the honest answer is that the average effects are real but modest.
Randomized studies find effect sizes around g = 0.17–0.25. Statisticians call that "small," though for scale it is two to three times the effect of taking aspirin to prevent heart attacks. Two things separate this from earlier panics. First, the causal evidence comes from randomized experiments and natural experiments, not correlation alone. Second, the companies' own internal research reached the same conclusions before outside researchers did — and in Meta's case, the company halted and never published the study that found quitting reduced depression and anxiety.[evidence]
Is social media the cause of the teen mental health crisis?
The research does not support "the sole cause" — it supports "a significant, causal contributor."
Rates of major depression among U.S. teen girls rose roughly 145% between 2010 and 2020, and the timing matches smartphone and social media adoption. But timing alone is not proof. The strongest causal evidence is a natural experiment exploiting Facebook's staggered 2004–06 college rollout, published in the American Economic Review, which found the platform's arrival caused worse student mental health. Serious researchers disagree about magnitude; almost none argue the effect is zero.[evidence]
Who runs this site, and is it neutral?
It is an independent, non-commercial evidence project — and it is not neutral, it is sourced.
This site is not affiliated with Meta Platforms, Inc., ByteDance Ltd., or any product or tool it mentions. It sells nothing and takes no advertising. It has an argument — that these platforms are engineered to manipulate attention and emotion — and it makes that argument entirely from primary documents, court records, regulatory findings, and peer-reviewed research. Where evidence is weak, contested, or cuts against the argument, the pages say so in "honest caveat" notes. Every claim, with its source →