I sat in a quiet Italian classroom and watched a boy tap an app between questions. You could see his attention fracture, a small thread pulled away from the test. That moment is the seed of a finding that quietly rattles education policy across continents.
A student taps Instagram during a worksheet; the study then tracks that class for four years.
I read the new Nature paper that followed more than 5,000 Italian teenagers from late primary school into high school. The team matched students who had the same 5th-grade scores and then split them by when they first opened a social account. The headline: kids who joined social media in 6th grade (ages 11–12) scored worse on standardized math and Italian exams in 10th grade (ages 15–16)—about half a school year behind peers who signed up later.
Does early social media use harm grades?
Short answer: the study shows a strong correlation, and the pattern intensifies over time. If you opened an account in 6th grade, your probability of repeating a grade was higher and your scores fell further by grade 10—even though most classmates had joined by then. The authors suggest two plausible pathways: excessive time on platforms and attentional disruption that compounds like a slow-drip faucet, eroding study time and focus.
A teacher notices more off-task phone checks in 7th grade; the numbers reflect that observation.
The effect shrinks if kids wait until 8th grade to sign up and is “smaller but still meaningful” for 7th-grade adopters. That gradient—biggest harm for the earliest adopters—argues that timing matters. I trust the method here: picking students equal on prior achievement strengthens the case that early sign-up predicts later lag, not just that struggling kids join earlier.
At what age should children be allowed on social media?
Governments are answering that question in real time. Australia and the UK have policies targeting under-16s, France recently banned social media for under-15s, and the EU has debated an under-13 limit. The researchers say delaying access until children can manage distractions may help, but they also warn against overclaiming cause-and-effect from a single educational outcome.
A parent reads headlines about lawsuits while scrolling a lawyer’s filing; the legal fallout mirrors the academic debate.
The legal pressure is heavy. States and plaintiffs have sued Meta, ByteDance and Google, citing design features like infinite scroll and autoplay that they argue hook young users. Meta has said it might face damages of $1.4 trillion (€1.29 trillion) in one set of claims—an amount that would dwarf most corporate penalties and change the business calculus for platforms if awarded.
Can platform design cause addiction in teens?
Design matters. Instagram, TikTok and YouTube optimize for engagement; autoplay and algorithmic feeds reward short, repeatable attention bursts. The study cannot definitively name whether harmful content or sheer time spent is the main driver of poorer grades, but it does tie earlier account creation to more persistent academic decline. That puts product design squarely in the conversation.
A classroom in grade 10 shows mixed results across subjects; English scores buck the trend.
Not every subject suffered the same way. The drop in math and Italian did not appear for English—possibly because much online content is in English and social platforms provide informal language exposure. That nuance matters: social media isn’t all downside, and effects can be subject-specific.
What this means for you: If you’re a parent, teacher or policymaker, the study gives a readable, methodical signal: earlier access is linked to worse academic outcomes later on, and the harm compounds. If you build or regulate platforms, you now have an education impact to weigh alongside mental-health and safety concerns.
I will watch how regulators respond. France and Australia have already acted; courts and state attorneys general are testing the limits of platform responsibility. Platforms like Meta, ByteDance and Google are central players in that fight, and their product choices—feeds, autoplay, infinite scroll—are at the heart of the debate.
The study closes with a cautionary tone: treat the findings with care because the topic is politically charged and complex. I’ll add a different caution: don’t let a single metric—test scores—be the only lens we use. We should combine academic data with mental-health studies, behavioral metrics from platforms, and careful policy experiments.
If governments tighten age limits, the next set of research will tell us whether delaying sign-up reduces the cognitive drag the authors observed. And if platforms change features or offer better parental controls, we’ll need rigorous evaluation of those changes, too.
Are we prepared to decide, as a society, whether the convenience of endless feeds is worth a measurable hit to children’s schooling?