Screen time debates
The Screen Time Panic Is Built on Shakier Science Than You Think
The most cited statistic linking smartphones to teen depression was later found to explain about as much variance in wellbeing as eating potatoes.
The Idea
For the past decade, a consensus has solidified in parenting culture, policy circles, and op-ed pages: screens are making us — and especially young people — miserable. The evidence always seemed intuitive. More time on Instagram, less time sleeping. More scrolling, less talking. Anxiety rates rising in lockstep with smartphone adoption. What's not to believe? Quite a lot, it turns out. The psychologist Amy Orben and statistician Andrew Przybylski ran what's called a specification curve analysis on the most widely cited datasets — essentially testing thousands of different ways to slice the same data rather than picking the single framing that supports your hypothesis. Their conclusion: the association between screen use and wellbeing is real, but so small as to be practically meaningless. Wearing glasses and eating potatoes showed similar-sized correlations with adolescent mental health. The problem isn't that the researchers are wrong — it's that the field has been doing what's called 'researcher degrees of freedom': choosing which variables to include, which age groups to focus on, and which outcomes to measure in ways that quietly inflate alarming findings. Meanwhile, the studies generating headlines tend to use self-reported screen time, which turns out to be wildly inaccurate even compared to what people's own phones record. None of this means screens are harmless. It means the science is genuinely messier than the confident warnings suggest — and that the moral panic may be outrunning the data by some distance.
In the World
The figure who arguably started the modern screen time panic is Jean Twenge, a San Diego State University psychologist whose 2017 Atlantic essay 'Have Smartphones Destroyed a Generation?' went viral in the way that only a perfectly timed piece of cultural diagnosis can. Twenge pointed to the sharp rise in teenage depression and loneliness that seemed to begin around 2012 — the year, she noted, when smartphone ownership crossed 50% among American teens. The timing looked damning. But when Orben and Przybylski got access to the same datasets Twenge was using — notably the Monitoring the Future study and the UK Millennium Cohort Study — they found that the dramatic findings dissolved under scrutiny. The specific correlations Twenge had highlighted were real but cherry-picked from a vast sea of possible variables. When you accounted for all of them, screens looked far less sinister. Twenge pushed back, arguing that even small effects matter at a population scale. That debate is ongoing and legitimate. But what happened next is instructive: several major tech platforms cited the 'screen time is harmful' consensus to justify new parental control features — tools that generated positive press coverage and regulatory goodwill without requiring the platforms to change anything structural about how their products are actually designed to capture attention. The panic, in other words, was easy to exploit even if the underlying science remained genuinely unsettled.
Why It Matters
The screen time debate affects decisions you probably make every week — whether to put your phone in another room, how firmly to set limits for children, whether to feel guilty about the hour you spent reading threads before bed. If the science underpinning those decisions is weaker than the headlines suggest, that is worth knowing. More broadly, this is a case study in how statistical findings get transformed into cultural certainties. A correlation becomes a trend piece becomes a bestselling book becomes bipartisan policy consensus — each step adding confidence that the original data never actually warranted. The sociologist Stanley Cohen called this a 'moral panic': a condition, episode, or group of people that becomes defined as a threat to societal values, with experts offering diagnoses and solutions that feel proportionate but may not be tethered to the underlying reality. Being a more careful consumer of tech-and-wellbeing research doesn't mean being complacent about genuinely harmful design patterns. It means being able to distinguish between what we actually know and what we've simply agreed to believe — which is a different and more useful kind of digital literacy than counting screen hours.
A Question to Ponder
If the evidence that screens harm wellbeing is much weaker than you assumed, what belief about technology — or about yourself as a technology user — might you need to revisit?
Get a new one of these every morning.
Start learning with Thinkable