Research & Data
The goldfish stat was invented. The screen addiction is real
The OECD's Digital Well-being Report, drawing on a 14-country study from 2025, finds that digital engagement has complex effects on wellbeing — and that individuals spending over five hours a day on screens for personal use show markedly higher odds of poor mental health outcomes. Lifestyle factors (sleep, finances, physical activity) remain significant predictors, but screen time is not simply noise.
The most-cited statistic in a decade of screen time panic — that humans now have an 8-second attention span, shorter than a goldfish — was sourced from a 2015 Microsoft Canada report citing an organisation called Statistic Brain. They fabricated the figure. There is no peer-reviewed study behind it. Goldfish, incidentally, can sustain attention considerably longer.
The OECD study is more careful and useful. Surveying over 14,000 individuals across 14 countries, it finds that the relationship between screen time and wellbeing is genuinely dose-dependent: moderate use (one to three hours daily) is not clearly harmful; more than five hours is. It also finds that what you do on a screen matters more than how long you do it — social connection and purposeful use behave differently than passive scrolling. Not all screen time is equal, which is the one thing the goldfish conversation was incapable of capturing.
But the OECD data doesn't let the optimists off either. Five hours of personal screen time is not an unusual day. And the category of "screen" now encompasses not just social media and streaming but AI assistants, which have their own particular capacity to occupy attention indefinitely and frictionlessly. Meanwhile, the Reverse Flynn Effect — a decline in measured IQ scores in several developed countries began reversing around 2010, which maps uncomfortably onto the spread of smartphones. The causal link remains contested as IQ tests themselves are being debated as valid cross-generational measures.
In response to all this, 114 education systems — 58% of countries worldwide, up from 24% in 2023 — have now introduced restrictions on phones in schools, according to UNESCO's Global Education Monitoring Report (March 2026). Australia went furthest, banning under-16s from holding social media accounts entirely, with fines up to $50M for platforms that fail to enforce it. These are not small regulatory nudges; they are structural interventions premised on a view of addictive design that the goldfish debate never quite got to. The science is still assembling its verdict. The policy moved without waiting.
