Locking students’ phones in pouches for the school day sharply cuts phone use, but it does little for test scores. A nationwide study of US schools that adopted lockable pouches found average effects on achievement close to zero, a first-year rise in discipline problems, and a dip in student well-being that later turned positive.
The study is by Hunt Allcott, Thomas Dee and Matthew Gentzkow of Stanford University, E. Jason Baron of Duke University, Angela L. Duckworth of the University of Pennsylvania and Brian Jacob of the University of Michigan. It was issued in April 2026 as NBER Working Paper 35132 and later revised.
Tracking almost 5,000 pouch schools
Under the pouch approach, students seal their phones in a magnetic pouch on arrival and generally cannot open it until dismissal. Nearly 5,000 US schools used pouches by 2026. The researchers obtained adoption records from Yondr, the largest provider, and focused on middle and high schools that adopted in the 2023, 2024 or 2025 school years.
They compared those schools with similar schools that did not adopt, using a staggered difference-in-differences design that was pre-registered. Outcomes came from GPS device data for more than 40,000 schools, state test scores, attendance and discipline records, Panorama Education student surveys, and a national teacher survey fielded between April 2025 and February 2026.
Phone use falls, test scores barely move
The pouches worked on their own terms. GPS activity on campus during school hours fell by roughly 30% by the third year, which the authors treat as a conservative lower bound. Teachers reported that the share of students using phones in class for personal reasons dropped from 61% to 13%.
Academic gains did not follow. Over the first three years, the team could rule out average test score improvements larger than about 0.007 student-level standard deviations. High schools saw modest gains, notably in math, at about 0.024 standard deviations, or roughly 0.9 percentile points. Middle schools saw small declines.
Attendance was essentially unchanged, and there was no clear improvement in self-reported classroom attention or perceived online bullying.
A rough first year, then recovery
Disciplinary incidents rose in the year of adoption, equivalent to about a 15% increase in suspension rates, before fading. Student well-being fell by about 0.2 standard deviations in the first year and then rebounded, rising by 0.16 standard deviations by the second year after adoption.
The authors suggest the early spike in discipline may reflect enforcement of the new rules, or students swapping phones for other disruptive behaviour. Younger students may have less impulse control, and phone activity fell less in middle schools, which might explain the weaker results there. They call these interpretations speculative.
They also note limits. The data cover at most three years after adoption, the measures miss some possible effects, and stricter approaches such as campus-wide bans may work differently. The well-being pattern, they add, should be read cautiously because the short panel makes it hard to separate real change over time from differences between cohorts.
The high school math gain is similar in size to the 0.9 percentile point effect another study found for Florida’s phone restrictions.
Study Details:
- Title: The Effects of School Phone Bans: National Evidence from Lockable Pouches
- Authors: Hunt Allcott, E. Jason Baron, Thomas Dee, Angela L. Duckworth, Matthew Gentzkow, Brian Jacob
- Journal: NBER Working Paper No. 35132
- Publication Date: April 2026
- DOI: 10.3386/w35132
