A child’s peers are a vital determinant of their performance at school. While it has been well documented that being exposed to disruptive peers can have adverse effects on educational outcomes, little is known about the impact of being exposed to a peer with a criminal background. In this paper, leveraging a rich administrative dataset that links crime and education records for nearly the entire universe of school students in England, I study the extent to which criminality spreads within schools from juvenile criminal peers. Exploiting the plausibly random timing of school transfers of students with a criminal record, I find that exposure to a criminal peer increases the peer’s probability of starting a criminal career by 12.9% within four months of transfer, and the effect is persistent for three years. These peer effects account for about 20 million pounds, or 5% of the total social cost of youth crime in England. Given the substantial heterogeneity in treatment effects across school characteristics, reassigning transfer offenders to a different school could reduce crime by 36% under reasonable assumptions. Furthermore, exposure to a peer with an offense record has an adverse effect on educational attainment, with approximately one quarter of the observed decline attributable to direct exposure effects, and the remaining three quarters mediated through engaging in criminal behavior.
Presented by:
Hikaru Kawarazaki (IFS, UCL)
Date & time:
November 25, 2026 12:30 pm - November 25, 2026 1:30 pm
Venue:
SSRC416 (2N2.4.16)
External seminars home