The Phones Went in the Pouch. The Attention Stayed Out.
Nearly five thousand American schools locked the phones away, and it worked — the phones were gone. What the largest study yet found in the space they left is the part worth an adult's notice.
Put the phone in a drawer on Monday morning and see what the hour does. That is the whole experiment, and nearly everyone who has run it reports the same faintly embarrassing result. The phone was gone and the hour was not noticeably better. You looked out of the window. You reorganised a folder. You went to find a biscuit.
American schools have now run that experiment at a scale no individual could. By this year nearly five thousand of them were sealing pupils' phones in lockable fabric pouches at the gate and releasing them at the final bell. In a working paper dated August 2026, six economists — Hunt Allcott, Jason Baron, Thomas Dee, Angela Duckworth, Matthew Gentzkow and Brian Jacob — compared schools that adopted the pouches with similar schools that did not, drawing on test scores, attendance and discipline records, student surveys, phone location data and a national survey of teachers.
The pouches worked. Before them, teachers reported 61 per cent of students using phones in class for personal reasons. Afterwards, 13 per cent. As a way of getting a phone out of a room, it is hard to do much better.
And the attention did not come back. Average effects on test scores were close to zero. Attendance did not move. The students' own reports of how well they attended in class showed no clear improvement.
Removing a distraction opens a gap. It does not say what goes in the gap, and nothing in the history of attention suggests the answer is automatically the lesson.
Where it went instead
The authors are careful about why, and offer possibilities rather than a verdict. Attention may have moved to the person at the next desk. It may have moved to the laptop, which nobody sealed in anything. Some of the year went on the rule itself: in the year of adoption, suspension rates rose by roughly 15 per cent, which is what happens when a new restriction arrives and people test it.
The more useful finding is about time, and it is the one that never survives the headline. Students' reported wellbeing fell in the year the pouches arrived. Then it recovered, and by the second year after adoption the estimate had turned positive. The rise in discipline faded over the same stretch. On those measures the first year looked worse than doing nothing, and the third looked better.
Consider what that does to the drawer experiment. Almost every adult attempt to take the phone away is judged within a week, and usually within a day. That is an evaluation carried out entirely inside the dip — the stretch in which the old exit is shut, the replacement habits do not exist yet, and the only measurable change is that you are more irritable. Naturally it reports failure. It was read at the one moment that could report nothing else.
The cautions belong here and they are not small. This is a study of teenagers in classrooms, not adults at desks, and a school day has a structure most working days lack. The wellbeing pattern is, in the authors' own word, suggestive: the data run to three years at most, and they say plainly that they cannot fully separate a genuine recovery from differences between the schools that adopted early and those that adopted late. The attention measure was drifting differently in the two groups before any pouch arrived, which limits what can be concluded from it. And the null on scores is an average — high schools saw a modest gain in maths, middle schools a smaller loss. Nobody should read this as proof that taking phones away does nothing. It reads better as evidence that taking them away is the first half of something.
The second half
Decide what the hour is for before you clear it. Not focus — a named thing, small enough to begin: the chapter, the one difficult email, the drawing. The schools removed the phone and left the lesson as it was, and the lesson turned out to be competing with rather more than a phone. A cleared hour with nothing assigned to it gets filled by whatever is nearest, and you already know what is nearest.
Then set the review date a long way out and write it down. A month at the least. The first fortnight is supposed to be worse; that is not a sign the thing is failing but what the thing is like at that stage, in a school of a thousand or a study of one. Do not ask whether it is working while you are still annoyed.
And keep the question honest when the date arrives. Not whether the phone stayed in the drawer — the pouches managed that from the first morning. Whether anything grew in the space.