Is screen time actually bad for you, or is that boomer talk

Let's grade the discourse.

There are two failure modes in the public conversation about screen time. The first is panic: phones are destroying a generation, every minute is poison, the children are doomed. The second is dismissal: this is moral panic, every generation has its boogeyman, television was supposed to ruin us too. Both positions are loud. Both positions are wrong in the same way. They treat "screen time" as a single variable when the underlying research describes several different effects with several different confidence levels.

I have read the literature. Let me give you the calibrated version.

High confidence: sleep

Sleep is the most settled finding in this entire field. Screens before bed delay sleep onset and reduce sleep quality. The mechanisms are independently documented and they stack:

• Blue light suppresses melatonin production, shifting the circadian signal that tells your body to sleep.
• Bright stimulation extends physiological arousal past the point where the body would otherwise wind down.
• Content keeps the brain in active processing mode. A scrolling feed at 11:47pm is not passive; it is a series of micro-decisions.

You can argue about the size of each effect. You cannot reasonably argue the effect is zero. If you change one input from your current routine, change this one. Phone out of the bedroom, charged in another room. The intervention is free, the evidence is robust, and the downstream effects on mood, attention, and weight are well documented.

High confidence: heavy social media use correlates with worse mood, especially in teen girls

The signal here is consistent across multiple study designs. Heavy users of image-based social platforms report higher rates of anxiety, depression, and body image concerns than light users. The effect is largest in adolescent girls.

Jonathan Haidt's The Anxious Generation (2024) is the most aggressive statement of this case. His causal claim is contested. The correlational pattern he documents is not. The researchers who disagree with Haidt mostly disagree about how much of the teen mental health decline is attributable to phones, not whether the relationship exists.

Calibrate: heavy use, especially of image-based platforms, especially in adolescent girls, is the population where the signal is strongest. This is a real finding. Treat it as a real finding.

Medium confidence: attention effects exist, smaller than the headlines suggest

The popular claim is that phones have shredded the human attention span. The lab data is more modest. Effect sizes for attention degradation in controlled tasks land around Cohen's d of 0.15 to 0.25. That is detectable, replicable, and small. It is not nothing. It is also not the apocalypse implied by the trade press.

What the data does support: heavy multitasking is associated with worse performance on attention-switching tasks. What the data does not yet clearly support: that this generalises to broad cognitive decline, or that the effect is permanent.

If your subjective experience is that your attention has gotten worse, your subjective experience is probably correct. The published effect sizes suggest the cause is real but smaller than your feed says. Both can be true.

Low confidence: physical brain changes

Studies showing fMRI differences in heavy phone users exist. They are small, the samples are small, and the causal direction is genuinely unclear. Heavy users may have different brain activity patterns because they use phones heavily, or they may use phones heavily because of pre-existing patterns. The brain also changes constantly in response to every input it receives. Reading a book changes your brain. Driving changes your brain. The phrase "phones change your brain" is true and almost meaningless without an effect size and a comparison.

Down-weight any headline citing brain scans. The methodology rarely supports the certainty of the claim.

Calibration by age bracket

The "screen time is bad" question has different answers for different age groups.

• Under 8: limit. Real evidence on language and developmental outcomes. The intervention is straightforward; this group is not the hard case.
• Teens 13 to 18: most concerning bracket, especially for image-based social media in girls. The strongest evidence base for restriction lives here.
• Adults: smaller effects, contingent on use pattern. A 35-year-old who uses their phone two hours a day for messaging and maps is in a different category than a 35-year-old who uses it six hours a day on short-form video. "Screen time" is not the variable; which screen time is.
• Older adults: limited research. Existing data suggests screen-based social connection is net positive. Loneliness is the dominant risk factor in this group, not engagement.

Recommendations by audience

• If you are panicking about your kid's phone: cool it slightly. The evidence is real but the doom framing is not in the data. Restrict image-based social, not all screens.
• If you are dismissing other people's concern: cool it as well. Sleep and adolescent girls are real findings. "Television was fine" is not a counter-argument.
• If you are an adult who feels their phone is too much: you do not need academic backing to act on that signal. Subjective experience is data. Move.

Interventions that hold up regardless of age

The research keeps refining. The interventions don't really change. These work for almost any age bracket, almost any use pattern:

• Phone in a different room at night.
• Cap social platforms at 60 to 90 minutes per day.
• Notifications off by default; turn on only the few that earn it.
• Pre-commitment if soft tools fail. ScreenFine, Beeminder, a friend who will check on you. Same mechanism, different wrapper.

End-state

The research will keep refining. The behaviour change is independent of the refinement. End of report.