
Strict school phone bans lift maths and reading scores by about one to two extra weeks of learning a year, according to a major new Harvard study — the first national analysis to track individual students before and after the bans took effect.
Published on Tuesday, November 17, 2026, by Harvard's Center for Education Policy Research with researchers from NWEA and Dartmouth, it followed 3.3 million students in grades 3 to 10 at 4,600 schools across 49 states and Washington, D.C., during the 2021-22 through 2025-26 school years, measuring growth on NWEA MAP tests from fall to spring.
Schools that required students to lock their phones in pouches for the whole school day saw modest but steady gains in both subjects — about 0.01 standard deviations of annual growth. The gains appeared at every achievement level, began in the first year of a ban, and were still visible three years later.
The study, from Harvard's Center for Education Policy Research with NWEA and Dartmouth researchers, is the first national analysis to compare the same students' achievement growth before and after a strict school cellphone ban. Earlier research looked at single schools or districts; this one covers 3.3 million pupils and five school years of test data.
The researchers used NWEA MAP tests, which measure growth from fall to spring, so they could watch the same children's progress speed up after their school locked phones away.
The clearest gains came in schools that required students to lock phones in pouches — the Yondr system used by many districts — for the whole school day. The study also gives the first evidence on state-mandated bell-to-bell bans, in place in 22 states in 2025-26, with suggestive evidence they boosted achievement too.
The findings land against a long slide in American school results. Since 2013, 8th-grade maths and reading have fallen by roughly one grade equivalent by 2024, as students lost two to three weeks of learning a year while broadband smartphones spread.
Phone bans spread in response, first school by school and then by state law. Districts tried pouches, lockers and outright bans, but until now there was little national evidence on whether any of it actually moved test scores. The new study is the first to answer that question with data on millions of students rather than a handful of schools.
The headline number is 0.01 standard deviations of extra annual growth in both maths and reading. In classroom terms, the researchers translate that into roughly one to two additional weeks of learning per year — the kind of gain a teacher would notice over a full school year, even if no single test shows a dramatic jump.
The gains showed up at every achievement level, from struggling pupils to top performers, and they did not fade: they started in the first year and were still there three years later.
The authors are careful about the limits: a phone ban is not a cure for everything dragging scores down, and the gains are modest next to the post-2013 slide. But reversing about half of each year's learning loss with one policy change is, by education research standards, striking.
Co-author Professor Thomas Kane of the Harvard Graduate School of Education put the finding plainly: "Our answer is yes, phone bans do have a positive effect on growth. But it's not a panacea."
That caution is shared by the study itself, which notes the gains, while real, are modest. The findings are also contested: a separate National Bureau of Economic Research study in May found strict bans had little overall impact, setting up a debate more data must settle.
For now, the Harvard study is the largest and most detailed evidence available, and its design — tracking the same students before and after a ban — is the strongest yet. Supporters will cite it in statehouses; sceptics will point to the May study and demand replication.
Twenty-two states have already mandated bell-to-bell bans, and many more districts are weighing whether to follow. Until now, those decisions were made mostly on instinct and anecdote. This study gives policymakers the first national numbers to work with.
The results also matter beyond test scores. Phones in classrooms have become a proxy fight over attention, childhood and the role of technology in schools, as our report on the push for AI-free search in US schools shows. A policy that measurably improves learning, even modestly, changes that argument.
The debate now moves to replication: whether the gains hold as more states' bell-to-bell bans produce data, and whether the effect survives when bans are imposed by law rather than chosen by a school. School policy battles are already reaching the courts, as in our report on the Supreme Court preschool case, and phone rules may follow.
Districts face a practical choice too: pouches, lockers or a simple ban, each with different costs and enforcement headaches. The Harvard evidence favours the strictest version — phones locked away all day — but schools will weigh that against parent concerns about reaching their children during the day.
The bigger question is whether the gains compound: one to two weeks a year across a child's whole school career could add up to far more than any single year suggests — and that is what the next studies will try to measure.
What did the Harvard study find?
Strict school phone bans lift maths and reading scores by about 0.01 standard deviations a year — roughly one to two extra weeks of learning per year — for students in grades 3 to 10.
How many students and schools were studied?
It tracked 3.3 million students at 4,600 schools across 49 states and Washington, D.C., in the 2021-22 through 2025-26 school years.
How do the bans work?
The clearest gains came in schools requiring students to lock phones in pouches, such as Yondr pouches, for the whole school day. State-mandated bell-to-bell bans in 22 states in 2025-26 also showed suggestive benefits.
Do the gains last?
Yes. Gains started in the first year and were still visible three years later, at every achievement level.
Do all studies agree?
No. A National Bureau of Economic Research study in May 2026 found strict bans had little overall impact, so the findings are contested and more research is expected.