The influence of friends on teenage offending: How long does it last?
| Author | David J. Smith,Russell Ecob |
| Published date | 01 January 2013 |
| Date | 01 January 2013 |
| DOI | http://doi.org/10.1177/1477370812456345 |
| Subject Matter | Articles |
European Journal of Criminology
10(1) 40 –58
© The Author(s) 2013
Reprints and permission: sagepub.
co.uk/journalsPermissions.nav
DOI: 10.1177/1477370812456345
euc.sagepub.com
The influence of friends on
teenage offending: How long
does it last?
David J. Smith
University of Edinburgh, UK
Russell Ecob
Ecob Consulting, UK
Abstract
This study focuses on the question whether peers have a lasting effect on offending behaviour. In a
longitudinal study of 4300 teenagers in Edinburgh, Scotland, respondents nominated three friends
at age 13–14 most of whom were identified as other members of the cohort. The growth curve
of own offending was modelled over a period of three years from age 13–14. By allowing for the
effect of own prior offending on the friendship network, the model specifies the influence running
from friends to self. The level of offending of the three friends nominated at age 13–14 has a clear
immediate effect on own offending, which steadily diminishes over time but remains statistically
significant for a period of four to five years in a ‘slim’ model of broad offending and for two to
three years in a ‘fatter’ model that adjusts for a wide range of explanatory variables. Many of these
other variables have a significant influence on the growth curve of offending, but friends’ offending
is among the most important variables explaining own offending. It is argued that social learning
about the techniques, pleasures and justifications of offending may be important in explaining the
persisting influence of friends, whereas situational factors and peer pressure may be important in
explaining their immediate, short-term influence.
Keywords
Friends, peer pressure, social learning, social networks, teenage offending
Introduction
A large body of research supports the theory that friends have an important – even a
crucial – influence on adolescent offending. This fits with the fact that most adolescent
Corresponding author:
David J. Smith, School of Law, University of Edinburgh, Old College, South Bridge, Edinburgh, EH8 9YL, UK.
Email: d.j.smith1@blueyonder.co.uk
456345EUC10110.1177/1477370812456345European Journal of CriminologySmith and Ecob
2013
Article
Smith and Ecob 41
offending involves groups rather than individuals acting alone, and with the basic
assumption that social behaviours and attitudes emerge out of interactions with other
people.
Yet the importance of peer influence is less firmly established than at first appears. In
two respects, the evidence is weak. First, the great majority of studies, until recently,
have relied on the respondents’ second-hand reports about their friends’ offending, a
method that exaggerates the similarity between own and friends’ behaviour. Second,
most studies have focused on immediate effects, typically over periods of up to one year.
This leaves open the question of whether the influence of teenage friends can be a turn-
ing point in the life course.
The last few years have seen the emergence of a new generation of studies that over-
come the first of these problems by using direct measures of friends’ offending (Haynie,
2001, 2002; Weerman, 2011), yet there is still a lack of reliable information about the
longer-term consequences of associating with offenders in the teenage years. Hence, the
present study aims to show how far the immediate influence of friends persists. We ana-
lyse data from the first six annual sweeps of the Edinburgh Study of Youth Transitions
and Crime, a longitudinal study of a single cohort of 4300 young people in the City of
Edinburgh, Scotland, who were aged 11–12 at sweep 1. At sweep 3 (age 13–14) cohort
members were asked to name their three best friends from the same school year, and in
the vast majority of cases the named friends can be identified as members of the cohort.
Hence a direct measure of self-reported offending (along with other information) is
available for the named friends. We estimate latent growth trajectories of offending over
a four-year period from sweep 3 to sweep 6, and we analyse the effects of friends’ offend-
ing at sweep 3 on these trajectories, after controlling for a range of other variables.
Key findings from earlier research
Research on the influence of friends has to tackle two outstanding problems. The first is
finding an uncontaminated measure of friends’ offending: where an indirect measure of
friends’ offending is used, so that respondents rate their own and their friends’ offending,
the measures are not independent. Often having only vague information about what their
friends do, young people are likely to normalize their own behaviour by projecting it
onto their friends. Second, there is a need to distinguish between assortment (a tendency
for teenagers to cluster together into groups of similar individuals) and influence (a ten-
dency for teenagers to cause a change in each others’ behaviour).
The second problem is complex, since assortment and influence are linked elements
of a dynamic process. A fundamental feature of social relations is that similar individuals
tend to group together because of the way they choose associates (McPherson et al.,
2001), although social and institutional structures also play an important role – prisoners,
for example, can associate only with other prisoners or with guards. Only if assortment
produces more and less deviant groups can associates influence individuals to be more or
less deviant. As spelt out by Thornberry (1987), any reasonable theory of peer influence
must involve reciprocal causal loops involving both influence and choice of associates,
but a central task for analysis is to determine the extent of peer influence over and above
similarity between associates arising purely from assortment.
To continue reading
Request your trial