A new paper released by the National Bureau of Economic Research (NBER) authored by a team of researchers (Allcott et al) examines the effect of locking students’ smartphones up for the day. NBC news is reporting it as the “largest study ever of school cellphone bans.” The study looks at more than 4500 schools over three years.
So what were the results? Not as clear-cut as you may guess. Apart from the first year, students report greater well-being. Academically the picture is more complex. Overall, there was little to no statistically significant impact on test results. There were some positive impacts for particular groups (high school math test scores improved, for instance). On the flip side, slightly negative impacts were observed for middle schoolers.
Furthermore, the study finds some negative effects on classroom attention. psychologist Jonathan Haidt observes, however, teachers reported higher satisfaction. Insofar as teachers’ satisfaction is derived from students paying attention or improving academically, this suggests the results may be more positive than some of the findings indicate.
The mixed results of the study seems to have muted some of the optimism around the screen-theory of the universal decline in test scores across the country.
While this new study is the largest, it is not the only paper we have. Other research by Figlio and Ozek finds Florida’s 2023 ban on smartphones led to student score improvements in the second year. Interestingly, the authors suggest these improvements could be due to a lower number of unexcused absences, perhaps implying that students are less likely to skip school when they can’t communicate with their friends who are in school under a phone ban.
A Norwegian study by Abrahamsson shows unambiguously positive effects of a ban. She finds lower rates of bullying and academic improvement, especially from girls, as a result of the ban. On the flip side, a 2025 Lancet study found no evidence of mental wellbeing improvement as a result of phone restrictions.
Smaller scale experiments have also been run on this question. In 2022, The Wall Street Journal ran an article with a provocative title: “This School Took Away Smartphones. The Kids Don’t Mind.” The article details how a small boarding school replaced their students’ smartphones with the intentionally “dumb” Light Phone. The most recent reporting indicates the change has stuck, and the students are happy. Although Buxton (a small boarding school in Massachusetts) may not be representative of typical schools, it follows an interesting pattern found in many other studies. That is, after an initial period of being upset, students report being subjectively happier after the bans.
This consistent finding is perhaps indicative of teenagers facing a prisoner’s dilemma, wherein if all their peers coordinate to put their smartphones away, everyone is better off. Each person, though, has an incentive to individually use a smart phone, and this leads to everyone adopting it, making the whole group worse off than if no one did.
Despite mixed results, there does seem to be some reason to be positive about the impact of bans. Much of the theory behind why smartphone bans will succeed is based on the idea that access to smartphones has an impact on children’s attention spans. By highschool, most kids have had access to some kind of smartphone or tablet for years already, so some lag on impact would be expected. The fact that several results are finding positive impacts after a couple of years seem suggestive of more positive impacts down the line.
The School as Laboratory
So should smartphones be banned? My own inclination is to keep smartphones away from my children. Like the administrators at Buxton, my kids will get browser-free, app-free Light Phones. The early evidence (and my general intuition) is that dopamine-centric algorithms aren’t good for the happiness or growth of children.
But that personal judgment is different from advocating a national policy. Whether public schools should implement bans is a different question, and it’s an empirical one that can be answered through further studies like the NBER paper. But the current public school system is poorly positioned even to discover an answer, much less how to implement it. If smartphone bans make a difference, we’d expect to see parents “vote with their feet,” moving to districts that ban smartphones. Schools that use — or exclude — technology effectively would gain students, and schools that tried to accommodate them would lose students until they shaped up.
But the institutional incentives of our current system do not allow for such competition and choice. In the world of for-profit, a business which fails to satisfy customers loses money and ultimately goes under. But government school administrators maintain their budgets even when their policies — and indeed, the education offered — are ineffective. Research compiled by Brookings suggests that school spending and learning outcomes are, at best, weakly related.
Ideally, our school system should be a kind of laboratory where competing ideas for what makes the best education are allowed to play out to see which idea is the most successful. But modern government schools do not have to participate in this laboratory, they often lose in head-to-head comparisons with private schools, homeschooling, and even the government schools of years past. Government school supporters often explain away private and homeschool success with selection bias, but this cannot account for the reason why families with successful outcomes are the type of people to select out, nor can it explain the falling standards of public schools over time.
Given this, government schools have weak incentives to ban phones even if that’s the best outcome for students and parents. Students’ educational failure imposes no costs, and the friction caused by bans might.
The data will take decades to really understand. But if my children were being educated in a public school, I’d be begging their district to run the smartphone ban experiment.
