TB: School Is Way Worse for Kids Than Social Media (Stark-Elster)
Core Thesis
Eli Stark-Elster argues that the moral panic about teen social media has the comparison backwards: by the available data, school is the larger driver of youth mental-health decline. Governments are racing to legislate the lesser harm while the larger one runs uninspected.
Key Takeaways
The natural experiment: COVID
- Youth suicide rates fell during COVID-era school closures and spiked when in-person school resumed. Whatever else closures did, on this particular metric they were a positive shock.
- Emergency psychiatric visits show the same seasonal signal: peaks during school terms, troughs during holidays.
School-induced stress is rising fast
- Self-reported teen school-stress rose from 43% in 2009 to 70% in 2024 — nearly a doubling in 15 years.
- About half of stressed teens report skipping class because of anxiety. Only 28% report distress from peer pressure to fit in socially — a much smaller number than the social-media narrative implies.
Timing argument against the social-media story
- The acute mental-health decline began in 2012, which is also the year of broad Common Core implementation. The standard story attributes 2012 to smartphone-induced harm, but the school-policy timing is at least as good a fit and rarely tested.
The proposed interventions
- Later school start times (the strongest empirical case).
- Reduced homework loads, especially past the threshold where studies show non-linear deterioration.
- Expanded recess, lunch, and unstructured time.
- All three are evidence-supported, low-cost, and politically tractable — yet absent from the youth-policy debate, which is consumed by social-media bans.
Mental Models
- Inversion — instead of "social media is harming kids; restrict it," try "what is producing the seasonal mental-health signal we already measure?"
- Survivorship Bias — the social-media story survives because adults find it intuitive; the school-stress story dies in committee
- Incentives Matter — politicians can ban social media without disturbing institutions; school reform requires confronting unions, parents, and standardized-testing regimes
- Over-Determination by Design — the COVID natural experiment, the seasonal psychiatric-visit data, and the 2012 timing all point the same way; the case is over-determined
See also
- Selfish Reasons to Have More Kids (Caplan) — Caplan on overstated parenting costs; this piece on a specific overstated harm
- The Case Against Education (Caplan) — same direction of argument: schooling produces less than it claims and costs more than is acknowledged
- Don't Give Up on Kids Because of Climate Change (Alexander) — same family of "panic is misdirected" arguments
- Lifelong Learning — what schooling should be doing; Stark-Elster on what it is actually doing
Source
Eli Stark-Elster, Unpublishable Papers Substack, 2026 Original article