Abstract

This paper examines peer effects in adolescent cannabis use from several different reference groups, exploiting survey data that have many desirable properties and have not previously been used for this purpose. Treating the school grade as the reference group, and using both neighbourhood fixed effects and IV for identification, we find evidence of large, positive, and statistically significant peer effects. Treating nominated friends as the reference group, and using both school fixed effects and IV for identification, we again find evidence of large, positive, and generally statistically significant peer effects. Our preferred IV approach exploits information about friends of friends – ‘friends once removed’, who are not themselves friends – to instrument for friends’ cannabis use. Finally, we examine whether the cannabis use of schoolmates who are not nominated as friends – ‘non-friends’ – influences
own cannabis use. Once again using neighbourhood fixed effects and IV for identification, the evidence suggests zero impact. In our data, schoolmates who are not also friends have no influence on adolescent cannabis use.
Original languageEnglish
Publication statusPublished - 2012

Publication series

NameMelbourne Institute Working Paper Series
PublisherMelbourne Institute for Applied Economic and Social Research
No.12
Volume27
ISSN (Print)1328-4991
ISSN (Electronic)1447-5863

Fingerprint Dive into the research topics of 'Peer Effects in Adolescent Cannabis Use: It's the Friends, Stupid'. Together they form a unique fingerprint.

Cite this