Escalation and Deceleration of Offending Behaviours From Adolescence to Early Adulthood
| DOI | 10.1375/acri.38.1.59 |
| Published date | 01 April 2005 |
| Author | Abigail A. Fagan,John Western |
| Date | 01 April 2005 |
Behaviours From Adolescence
to Early Adulthood
Abigail A. Fagan
University of Washington, USA
John Western
University of Queensland, Australia
Relatively little longitudinal research is available in Australia to describe
the age/crime relationship in much detail, particularly patterns of
offending occurring during the transition from adolescence to early adult-
hood. This paper addresses this issue using self-reported criminal involve-
ment from a school-based sample, a group of socially disadvantaged
individuals, and a group of officially identified offenders.The findings support
the widespread research that rates of offending peak during adolescence, at
which time offending is widespread, and that the criminal career is of
relatively short duration. However, the results also demonstrate that the
age/crime curve is not a unitary phenomenon.The type of offending behav-
iour being considered, the gender of the population, and the perpetrator’s
exposure to the criminal justice system contribute to the variability in the
curve. In this study, the prevalence and mean level of overall offending for
the total sample was higher during early adulthood than adolescence for
vehicle offences and drug-use, rates of theft were similar in both periods,
and vandalism and serious offending were lower. In addition, socially disad-
vantaged young people reported involvement in crime that peaked and
desisted earlier in the life course compared to the school-based sample,
and gender differences within these groups were also found. For the
school-based sample, offending for females began and desisted earlier than
for males, but within the at-risk group, the opposite was true. Implications
for crime-prevention programming are discussed.
The age/crime curve is one of the most generally accepted tenets of criminology
(Farrington, 1986; Hirschi & Gottfredson, 1983). Typically, the relationship
between age and criminal involvement is such that relatively few individuals display
serious problem behaviours during childhood; involvement in crime increases
Address for correspondence: Abigail A. Fagan, Social Development Research Group,
University of Washington, 9725 Third Avenue NE, Suite 401, Seattle, WA 98115-2060,
USA. E-mail: fagana@u.washington.edu
THE AUSTRALIAN AND NEW ZEALAND JOURNAL OF CRIMINOLOGY
59
VOLUME 38 NUMBER 1 2005 PP. 59–76
ABIGAIL A. FAGAN AND JOHN WESTERN
throughout adolescence when many, if not most, teenagers engage in some form of
criminal behaviour; and participation in offending declines from adolescence to
early adulthood. Despite the seeming inevitability that ‘what goes up must come
down’ for crime rates, detailed information regarding criminal involvement over the
life course is needed, particularly with respect to continuity and change in patterns
of offending (Catalano & Hawkins, 1996; Elder, 1998; Elliott, 1994; Loeber, 1996;
Piquero & Mazerolle, 2001; Sampson & Laub, 1990, 1993).
Review of the Literature
It is universally accepted that offending rates peak during adolescence, and that the
onset of serious and violent offending prior to ages 12 to 14 is relatively rare
(Blumstein & Cohen, 1987; Farrington, 2002, 2003; Moffitt, 1993). Developmental
criminologists posit that the many significant life experiences occurring during the
teenage years, including substantial changes in physical development, schooling,
roles and statuses, peer networks, and relationships with parents, leave adolescents
at considerable risk for offending. All of these transitions must be successfully
navigated, and failure to do so may increase the likelihood of involvement in crime.
For example, poor school performance and influence of deviant peers are both risk
factors for offending; conversely, protective factors such as strong ties to prosocial
individuals and institutions are demonstrated to reduce criminality (Farrington,
2002; Herrenkohl, Maguin, Hill, Hawkins, Abbott, & Catalano, 2000; Laub, Nagin,
& Sampson, 1998; Lipsey & Derzon, 1998; National Crime Prevention, 1999;
Sampson & Laub, 1990; US Department of Health and Human Services, 2001).
Even those who successfully cope with these experiences may engage in some
form of ‘adolescence-limited’ (Moffitt, 1993) criminal behaviour. Indeed, investiga-
tions based on self-reported behaviour demonstrate that the majority of teenagers
report committing one or more illegal activities during adolescence. For example, in
a New Zealand community sample (Krueger, Schmutte, Caspi, Moffitt, Campbell,
& Silva, 1994), 93% of males and 83% of females aged 18 reported engaging in at
least one delinquent activity in the past year. Likewise, Fagan (2001) found that,
among 11- to 17-year-olds in the National Youth Survey, conducted in the United
States with a representative sample of young people, 84% reported at least one
minor or serious offence during adolescence.
According to the age/crime curve, most individuals ‘age out’ of criminal behav-
iour, as rates of offending decline beginning in early adulthood. Although criminol-
ogists have long noted this trend, the specific ages at which desistance occurs and
the processes motivating the shift from crime have received somewhat limited
attention (Cernkovich & Giordano, 2001). Nonetheless, developmental criminolo-
gists speculate that disengagement from offending is contingent upon a successful
transition from adolescence into the greater responsibilities and stability of adult-
hood (Bushway, Thornberry, & Krohn, 2003; Farrington, 2002; Giordano,
Cernkovich, & Rudolph, 2002; Uggen, 2000; Warr, 1998). Sampson and Laub
(1990, 1993) assert that strong ties to prosocial institutions and individuals increase
the likelihood of disengagement, and there is evidence that, compared to those who
continue to break the law, those who desist are more likely to report supportive
marriages and full-time and/or satisfactory employment (Farrington, 2002; Sampson
60
THE AUSTRALIAN AND NEW ZEALAND JOURNAL OF CRIMINOLOGY
OFFENDING BEHAVIOURS FROM ADOLESCENCE TO EARLY ADULTHOOD
& Laub, 1990; Uggen, 2000; Warr, 1998). Exactly when these processes occur is
somewhat less clear, although there is some evidence that disengagement begins in
the early 20s, and that most individuals terminate serious violent offending by age
25 (Elliott, 1994; Farrington, 2003).
Hirschi and Gottfredson (1983) argue that the age/crime curve is invariable;
that is, that the rise and fall of criminal behaviour follows the same pattern for all
individuals, regardless of demographic characteristics, life experiences, or engage-
ment in different types of offences. Developmental criminologists offer a more
complex view, positing that there may be different pathways in and out of crime.
Moffitt (1993) distinguishes between the large number of adolescents who display
criminal behaviour only during the teenage years (adolescent-limited offenders),
and the small percentage of individuals who begin offending early in life and
continue well into adulthood (life course–persistent offenders). Other typologies
have also been offered. For example, Bushway and colleagues (2003) offered what
they considered a more dynamic model, which identified seven distinct trajectories
of offending, including Moffitt’s categories, as well as intermittent offenders, slow-
uptake chronic offenders and others. These findings point to the need for further
investigation of offending patterns, particularly regarding whether or not these
patterns can be universally applied to offenders.
This paper addresses these issues by examining changes in the prevalence and
incidence of offending from adolescence through early adulthood, using longitudi-
nal, self-reported data from young people in south-east Queensland, Australia. The
paper investigates whether or not different levels of offending are reported from
adolescence to early adulthood according to: (1) the type of crime committed; (2)
whether or not the respondent is from a normative, school-based population or an
‘at-risk’ population; and (3) whether or not gender differences in these relationships
are demonstrated.
Potential Variation in the Age/Crime Relationship
Several studies demonstrate that when individuals offend, they do not specialise in
certain types of crimes, but rather include a range of acts in their offending repertoire
(Farrington, 2003; Piquero, 2000). Thus, violent offenders also report acts of nonvio-
lence, and drug-users tend to commit acts in addition to using illegal substances. At
the same time, there is also evidence that offenders generally follow a sequence from
participation in less serious to more serious crimes, initiating minor offending at
young ages and severe, violent offences later in the life course (Elliott, 1994; Loeber,
1996; Kelley, Loeber, Keenan, & DeLamatre, 1997). Kelley and colleagues (1997)
report a progression in age of onset for males in Pittsburgh, Pennsylvania from
shoplifting (age 10) to property damage (age 12) to fighting and violence (age 13).
There is also debate as to whether or not offenders continue to commit illegal
behaviours already mastered, while also engaging in new types of criminal activities
(Farrington, 2003); or, if offenders move from one behaviour to another and disen-
gage from particular types of offences at different ages (Graham & Bowling, 1986;
Rutter, Giller, & Hagell, 1998). Because desistance is little studied, there have been
few attempts to illustrate whether or not certain crimes persist from adolescence to
young adulthood or are limited to the...
To continue reading
Request your trial