CoChat

The Educational Impacts of School Phone Bans: Evidence from Brazil

Open access
Lichand, Guilherme, Moreno-Louzada, Luca, da Costa, Thiago, Gentzkow, Matthew · 2026 · National Bureau of Economic Research

The DOI matches its Crossref record; 30 of the 30 references it cites were found in a registry or an index; 3 of the 21 sources its claims cite were checked for support.

Abstract

Concerns about negative impacts of student phone use have led to calls around the world for tighter restrictions on phones in schools. This paper evaluates the impact of a 2023 policy that banned non-pedagogical uses of phones within schools in Rio de Janeiro, Brazil. To isolate the causal effects of the policy, we contrast middle schools that already had strict rules on phone use prior to the policy ("control schools'') to similar schools that did not have strict rules ("treatment schools''), before and after the ban. While restrictions were imperfectly implemented both before and after the ban, we show that in-school phone use fell substantially in treatment schools relative to control. We then show that test scores, which were trending similarly in the two groups prior to the ban, improved by 0.06 s.d. in treatment schools relative to control. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Overview

CoChat Summary

TL;DR

This paper evaluates Rio de Janeiro’s 2023–2024 school phone ban using a difference-in-differences design that contrasts schools with no or partial pre-ban phone rules (“treatment”) against schools with strict pre-ban rules (“control”). It finds that the school-wide ban improved standardized test scores by 0.06 s.d. in treatment schools relative to control, with calibrated total effects as large as about 0.2 s.d.

Key findings

  • In-school phone use fell substantially: the share of principals reporting that most or all students used phones during class fell by 23.9 p.p. in control schools and by an additional 10.3 p.p. in treatment schools (matched sample), closing about two-thirds of the baseline gap; post-ban shares were 15% in control and 19% in treatment schools.
  • The school-wide ban improved the summary math-and-Portuguese test-score measure by 0.061 s.d. in treatment schools relative to control (significant at 5%); math improved by 0.077 s.d. and Portuguese by 0.045 s.d. The earlier classroom-only ban had small and statistically insignificant effects.
  • Effects were concentrated among lower-achieving students and in grades 7 and 8; event-study estimates were flat before the ban and rose after the school-wide phase.
  • A calibration using the first-stage decline in reported phone use implies a total treatment-school effect of about 0.20 s.d. relative to a no-ban counterfactual, though the authors present this as illustrative.

Methods

  • Difference-in-differences design using quarterly standardized math and Portuguese test scores from Rio de Janeiro’s municipal school district, 2022–2024, for middle-school grades 6–9. Treatment is schools with partial or no pre-ban phone rules; control is schools with strict pre-ban rules.
  • Main sample: 196,757 students across 325 schools and 1,116,689 student-quarter observations; matched sample: 97,380 students across 154 schools and 549,969 observations. Matching uses 2022 school-census covariates including infrastructure, student race shares, Bolsa Família coverage, region, and school management complexity.
  • Data include administrative records, a May 2025 principal survey (327 middle-school responses, 90% response rate) on pre-ban rules and perceived phone use, and qualitative interviews with five principals. Robustness checks include event studies, alternative treatment definitions, matching specifications, additional controls, alternative clustering, balanced panels, and attrition/mobility checks.

Limitations

  • The paper states that the design does not identify the effect of a phone ban relative to no ban; it identifies a relative contrast across schools with different baseline exposure. Mapping this to treatment effects generally requires assuming homogeneous treatment effects, so estimates could be over- or under-stated.
  • The authors note that treatment and control schools differ in baseline observable characteristics, that concurrent policies could in principle have heterogeneous effects, and that estimates are less precise when inference allows arbitrary serial correlation within schools.
  • The paper states that phone-use evidence comes from principals’ retrospective categorical reports, which are coarse, noisy, and potentially affected by social-desirability bias; the calibration also relies on a coarse measure and assumes a linear, homogeneous relationship between phone-use reductions and test scores. The text makes evident that phone use also declined in control schools, so the reduced-form estimates likely understate total gains.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.2139/ssrn.6824971

Cites

0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

3 Valid · 0 Not valid · 18 Not checked · 200 Own results and interpretations

101 more claims

Create a free account to see all 221 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 41
  • It may state two separate claims as one. 64
  • The quote only partly states this claim. 38

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

The publisher hasn’t deposited this paper’s reference list.

Research works printed but not linked: 2

  1. Allcott, Hunt, Matthew Gentzkow, Benjamin Wittenbrink, Juan Carlos Cisneros, Adriana Crespo-Tenorio, Drew Dimmery, Deen Freelon, Sandra Gonz´ alez-Bail´ on, Andrew M Guess, Young Mie Kim, et al. Forthcoming. 'The Effect of Deactivating Facebook and Instagram on Users' Emotional State.' American Economic Journal: Economic Policy .
  2. B¨ ottger, Tobias, and Klaus Zierer. 2024. 'To ban or not to ban? A rapid review on the impact of smartphone bans in schools on social well-being and academic performance.' Education Sciences , 14(8): 906.

Cited by

Metadata

Authors
Lichand, G., Moreno-Louzada, L., da Costa, T., & Gentzkow, M.
Venue
National Bureau of Economic Research
Publisher
Elsevier BV
Published
2026
OpenAlex
W7162300084
Cited by
1
References
0
Citations by year
2026: 1
Type
report
Language
en
Open access
Locations
  • National Bureau of Economic Research (publishedVersion)
  • SSRN Electronic Journal (acceptedVersion)
Funders
Centro de Excelencia en Geotermia de Los Andes
Topics
  • Social Sciences › Education › School Choice and Performance
  • Social Sciences › Sociology and Political Science › Impact of Technology on Adolescents
  • Social Sciences › Education › Child Development and Digital Technology
Keywords
education policy evaluation, academic performance, test scores, junior high school, school regulations, student distraction, Rio de Janeiro

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-03, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-03 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.