Intergenerational transmission of crime: An international, empirical assessment

Full text

This article was downloaded by: Vrije Universiteit Amsterdam (VU Amsterdam)
On: 03 Aug 2018
Access details: subscription number 13856
Publisher:Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place, London SW1P 1WG, UK
The Routledge International Handbook of Life-Course Criminology
Arjan Blokland, Victor van der Geest
Intergenerational transmission of crime
Publication details
https://www.routledgehandbooks.com/doi/10.4324/9781315747996.ch16
Steve van de Weijer, Megan Bears Augustyn, Sytske Besemer
Published online on: 06 Mar 2017
How to cite :- Steve van de Weijer, Megan Bears Augustyn, Sytske Besemer. 06 Mar
2017 ,Intergenerational transmission of crime from: The Routledge International Handbook of Life-
Course Criminology Routledge.
Accessed on: 03 Aug 2018
https://www.routledgehandbooks.com/doi/10.4324/9781315747996.ch16
PLEASE SCROLL DOWN FOR DOCUMENT
Full terms and conditions of use: https://www.routledgehandbooks.com/legal-notices/terms.
This Document PDF may be used for research, teaching and private study purposes. Any substantial or systematic reproductions,
re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents will be complete or
accurate or up to date. The publisher shall not be liable for an loss, actions, claims, proceedings, demand or costs or damages
whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
279
The intergenerational transmission of antisocial and criminal behavior has been the topic of scien-
tific research for a long time. However, previous studies on this topic usually focus on a single coun-
try and do not take into account the diversity of criminal careers of offenders. This research uses
three intergenerational data sets from different countries to empirically explore the link between a
parent’s and child’s criminal behavior. Using the Transfive data from the Netherlands, the Cambridge
Study in Delinquent Development (CSDD) from England, and the Rochester Intergenerational Study
(RIGS) from the United States, this work empirically demonstrates that a link between parental
offending and child offending exists. More so, this work also demonstrates that the relationship
between parental crime and child crime is conditioned by the timing of the parental offending.
Introduction
The popular idiom ‘the apple does not fall far from the tree’ illustrates the common assumption
in society and popular press that children and parents share similarities. Those similarities can be
reflected in physical appearances but also in certain behaviors, like antisocial and criminal behavior.
However, the empirical foundation on which the assumption that crime runs in the family rests
has been quite weak until recent decades.
Some of the first scientific studies that focused on the topic of antisocial and criminal families
were Dugdale’s (1877) and Goddard’s (1912) classic studies on the Jukes and the Kallikak family,
respectively. These studies were important in first identifying intergenerational continuity in
antisocial behavior but their research designs are nowadays considered as weak and outdated
(Christianson, 2003; Elks, 2005; Karp et al., 1995). Subsequent studies on the intergenerational
transmission of crime were scarce during the first half of the twentieth century. Many criminol-
ogists were reluctant to conduct research suggesting potential biosocial causes of crime because
biological theories were linked to sexism, racism, and immoral eugenic policies (Wright &
Miller, 1998). Another reason for this scarcity is the demanding data requirements for conduct-
ing an adequate multigenerational study. Prospective longitudinal data on the criminal behavior
of consecutive generations is needed over an extensive period, ideally multiple generations,
making these multigenerational studies a costly and a longwinded effort.
16
Intergenerational
transmission of crime
An international, empirical assessment
Steve van de Weijer , Megan Bears Augustyn and Sytske Besemer

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
280
During recent decades, however, several multigenerational studies with a prospective longi-
tudinal study design and population studies were started around the world. These studies enable
scientists to properly study the intergenerational transmission of offending. Among others, the
CSDD, the Rochester Youth Development Study (RYDS), and the Transfive study provided empirical
evidence for intergenerational transmission of crime in England (e.g. Farrington et al., 1996;
Rowe & Farrington, 1997), the United States (e.g. Thornberry et al., 2003; Thornberry, 2009),
and the Netherlands (e.g. Bijleveld & Wijkman, 2009; Van de Weijer et al., 2014). While extant
research with these data sets contributed to our understanding of the intergenerational trans-
mission of crime, they have some limitations with respect to the generalizeability of the inter-
generational transmission of crime. Each of these studies, for instance, only examines a sample
from a single country and the relevance of the intergenerational transmission of crime across
countries is seldom made (for an exception see e.g. Besemer, 2012a). Therefore, the first aim
of this chapter is to verify the intergenerational transmission of offending across multiple
countries—England, the Netherlands, and the United States—by investigating it in a consistent
manner.
We further expand upon prior research to examine how different aspects of the criminal
career may condition the intergenerational transmission of crime. In many studies on intergen-
erational transmission of offending, the data on criminal behavior is reduced to a dichotomous
variable indicating whether or not an individual has ever committed a crime (e.g. Bijleveld &
Wijkman, 2009; Farrington, Coid & Murray, 2009). An advantage of this approach is that easily
interpretable odds ratios can be computed that indicate the relative risk of committing a crime
for children of criminal parents compared to children without criminal parents. A disadvantage
of this approach, however, is that it ignores the diversity of criminal careers within the population
of offenders and how these elements may affect the offending behavior of the next generation.
Two potential dimensions are the timing of criminal behavior across the life course and the
number of crimes committed by each individual. Given that the causes and consequences of
crime vary across the life course (Thornberry, 2005), it is also reasonable to suspect that the
consequences of the frequency and timing of offending may differentially affect the next generation.
Although some studies have examined the influence of the timing and frequency of (parental)
offending on the intergenerational transmission of crime (e.g. Besemer, 2014; Besemer &
Farrington, 2012; Van de Rakt et al., 2009; Van de Weijer et al., 2014), this is often done using
very different methods precluding comparisons across studies or countries. Therefore, the
second aim of this chapter is to examine the influence of the timing and frequency of parental
offending on the intergenerational transmission of offending in a consistent manner across the
three countries. Finally, we will expand upon prior research and examine the intergenerational
transmission of offending across two different periods of the child’s life course—adolescence
and young adulthood—to further explicate how the timing and frequency of parental offending
may affect the timing and frequency of the child’s offending as well.
Theory
The intergenerational transmission of offending can be expected based on several criminological
theories which offer different possible explanations for this transmission, such as (social) learning
theories, labeling theories, and biological or biosocial theories. Among these is Thornberry’s
intergenerational extension to his and Krohn’s Interactional Theory of Offending (Thornberry,
1987, 2005; Thornberry & Krohn, 2001, 2005) which provides one of the most comprehensive
and inclusive explanations for the intergenerational transmission of offending. Farrington
(2011) also acknowledges these various theories and proposed six different mechanisms that

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
281
can account for the intergenerational transmission of offending. We now discuss each of these
theoretical perspectives.
Thornberry’s interactional theory of intergenerational transmission
The main tenets of Thornberry’s (2005) Interactional Theory are that offending can start at
any point during the life course and that (offending) behavior is intertwined with other domains
of life. Extending from these main points, the causal influences of crime and consequences of
crime are different at various developmental stages. For example, parents have more influence
on behavior in childhood whereas peers play a stronger role during adolescence. The inter-
generational extension to this theory posits that people who were involved in delinquent
behavior during adolescence will face problems making successful transitions into adult roles
including parenthood. This in turn will impact their children’s development. While Thornberry
(2005) acknowledges a direct effect of antisocial behavior from one generation to the next in
the form of social learning and genetics, he predominantly explains intergenerational transmission
through the mediating mechanisms of near-term consequences of offending and longer-term
consequences of offending that affect family processes such as “family conflict, hostility, and
especially by the quality of parenting” (p. 183). In other words, the main focus of this theory of
intergenerational transmission of antisocial behavior is the mediating processes that promote
criminal behavior in the next generation. For instance, persons involved in criminal behavior are
more likely to have children at a young age, experience structural adversity, continue to engage
in antisocial behavior including substance use, experience more stress, and have weak prosocial
bonds in adulthood. All of these factors, in turn, lead to difficulties in childrearing such as a lack
of supervision and monitoring of children, erratic parenting, and low levels of affect between
parent and child, each of which is a risk factor for offending in children and adolescents. Taken
together, Thornberry (2005) proposes that parental involvement in crime exerts a negative effect
on subsequent development, successful transitions into adult roles, and parenting styles, which,
in turn, lead to ineffective parenting and an increased risk that children will engage in criminal
behavior.
Farrington’s mechanisms of intergenerational transmission
Acknowledging the body of theoretical and empirical work that examines the intergenerational
transmission of offending, Farrington (2011) identified six possible mechanisms that can account
for the relationship between parental criminal behavior and a child’s criminal behavior. The first
two mechanisms that Farrington (2011) describes involve risk factors for criminal behavior,
such as disrupted families, teenage parenting, high school dropout, unemployment, and living in
deprived neighborhoods. Many of these risk factors place the intergenerational transmission of
crime within a broader cycle of intergenerational deprivation. This explanation suggests that
criminal behavior is not directly transmitted between generations. Instead, it suggests that inter-
generational continuity is a result of dual exposure to risk factors for offending among parents
and children, increasing the likelihood that both parent and child engage in criminal behavior.
It is also possible that many of these risk factors act as a mediator in the intergenerational
transmission of crime. According to this explanation, criminal parents expose their children to
risk factors for criminal behavior, which leads to an increased risk for their children to become
criminal as well. The third mechanism identified by Farrington (2011) is assortative mating. It
points to the finding that offenders are likely to marry or cohabit and have children with offenders
of the opposite sex (e.g. Boutwell et al., 2012; Krueger et al., 1998). As a consequence, children

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
282
with two criminal parents have an increased risk of becoming criminal themselves. However, it
is debatable as to whether this is an independent mechanism or if intergenerational transmission
is stronger with two criminal parents (Besemer, 2012a). Farrington’s (2011) fourth mechanism
is related to social learning theories and suggests that younger generations imitate and learn crim-
inal behavior from older generations. In his fifth identified mechanism, he posits that genetics
may play a role in the intergenerational transmission of offending. This explanation suggests that
a genetic predisposition for criminal behavior is transmitted from parents to their offspring.
Finally, Farrington’s sixth mechanism suggests a possible official bias toward known criminal
families (see also Besemer et al., 2013). According to this explanation, law enforcement agencies
monitor criminal families more intensively and consequently children of criminal parents are
more likely to be arrested and convicted for crimes than their peers from non-criminal families.
Support for each of these six mechanisms has been found in the empirical literature on the
intergenerational transmission of crime (e.g. Besemer et al., 2013; Farington et al., 2001;
Giordano, 2010; Rhee & Waldman, 2002; West & Farrington, 1977). Each of these different
studies, however, have their own strengths and weaknesses making it hard to compare the results
and evaluate the relative importance of each of the theorized mechanisms in explaining inter-
generational transmission of crime (Farrington, 2011).
Previous research on intergenerational transmission
Criminal and antisocial families have been the topic of empirical research for a long time.
Dugdale’s (1877) and Goddard’s (1912) classic studies on the Jukes family and the Kallikak family,
respectively, were among the first large-scale quantitative studies to argue that crime is hereditary.
They will be discussed in brief before reviewing some contemporary multigenerational studies
and their noteworthy results regarding the intergenerational transmission of offending. Finally, this
section will discuss previous research with respect to the influence of timing and frequency of
parental crime on the intergenerational transmission of crime and their limitations.
The Jukes family and the Kallikak family
As an inspector of the New York Prison Association, Richard Dugdale discovered that six persons
being held at the Ulster County jail were blood relations in some degree: the Jukes family. In
order to discover the roots of their criminal behavior, Dugdale traced their ancestors back seven
generations and collected data on 709 persons. Dugdale showed that of 709 Jukes, 140 were
criminals including 60 habitual thieves, 50 were prostitutes, and 180 were a recipient of financial
relief. In addition, seven persons were murdered and 440 persons were infected with a sexually
transmitted disease by members of the Jukes family.
Some decades later, Henry Goddard (1912) published his study on five generations of the
Kallikak family. At the beginning of the American Revolutionary War, Martin Kallikak Sr.
joined one of the military companies and had a short-term romance with ‘the nameless
feeble-minded barmaid’. When the war was over, Martin Kallikak Sr. married a respectable girl
and all children and descendants born from this line became respectable citizens such as doctors,
lawyers, judges, educators, traders, and landholders. The feeble-minded barmaid, however, was
also impregnated by Martin Kallikak Sr. and among their 480 descendants, Goddard (1912)
found “paupers, criminals, prostitutes, drunkards, and examples of all forms of social pest with
which modern society is burdened” (p.116).
Today, the classic studies of Dugdale and Goddard are heavily criticized. For example, the
Jukes family was actually a composite of 42 families rather than a single clan (Christianson,

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
283
2003). Moreover, Goddard was accused of retouching photographs of the Kallikaks in order to
make darker eyes and more menacing faces (Elks, 2005). Karp and colleagues (1995) also
suggested that prenatal exposure to alcohol could be responsible for the intergenerational
antisocial behavior of the Kallikaks rather than heredity.
Prospective multigenerational studies
During the last decades several multigenerational studies emerged using a prospective, longitu-
dinal study design to investigate the intergenerational transmission of crime more adequately.
For instance, 411 working-class boys born around 1953 from the inner-city area of South
London were followed prospectively from age 8 onwards in the CSDD. This study contains
self-reported measurements of delinquency for these boys, as well as reports of convictions for
the boys, their full biological siblings, their parents, and their children. With this data, Farrington
and colleagues (1996) showed that crime is concentrated within families: 6 percent of the
families were responsible for half of all convictions among the sample. Moreover, convictions of
one family member were strongly related to convictions for other family members indicating
intergenerational transmission of crime. Moreover, this transmission of offending between
same-sex relationships was generally found to be stronger than for opposite-sex relationships
(Farrington et al., 1996; Rowe & Farrington, 1997). Subsequent research by Besemer (2012b)
showed that fathers who specialized in violent offending were more likely to have violent
children. Research with the CSDD also demonstrated that a convicted parent increased the risk
of convictions for offspring even after controlling for the child’s self-reported offending (Besemer
et al., 2013). Building upon the initial CSDD data, Farrington, Coid, and Murray (2009) added
the children of the 411 original men to the dataset and found that there was an intergenerational
transmission of crime from these original men to their children. However, there was little
evidence to suggest an intergenerational transmission from grandparents to grandchildren
(Farrington, Coid & Murray, 2009).
In 1988, RYDS oversampled seventh and eighth grade students from public schools in
Rochester, New York, USA who were at a high risk for criminal behavior. Prospective, longitu-
dinal data was collected from this sample through the average age of 33 and from at least one
primary caregiver of the respondent through early adulthood. These measures include self-
reports of offending behavior and official arrest records of the respondents. As an extension of
RYDS, the primary investigators identified the first born children of the original respondents
who participated in RYDS beginning in 1999. Currently, the first born child, aged 2 years or
older, of 210 females and 463 males who were part of RYDS make up the RIGS. The RIGS
data include yearly self-report interviews from the biological parent who was part of RYDS,
yearly self-report interviews from their first born child beginning at the age of 8, and yearly
self-report data from the other primary caregiver. Analyses using RYDS and RIGS demonstrate
that parental offending leads to early antisocial and externalizing behavior in offspring, both
directly (Thornberry et al., 2003; see also Dong & Krohn, 2014) and indirectly, mediated by
parents’ depressive symptoms (Thornberry et al., 2009), parenting stress, and parenting behaviors
(Thornberry, 2009).
Using a more qualitative approach to study the intergenerational transmission of crime,
Giordano (2010) followed 127 girls from a state institution for delinquent girls in Ohio over
their life course in the Ohio Life-Course Study (OLS). The 158 biological children of these girls
were interviewed as well, during adult follow-up interviews of the initial respondents. Empirical
support for intergenerational transmission of offending was provided in these qualitative
in-depth interviews (ibid.). Comparisons of self-reported delinquency between the offspring of

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
284
the OLS girls and the offspring of a control group showed that the former group engaged in
significantly more delinquent behavior than the latter group, implying some intergenerational
transmission of behavior (ibid.).
In the Netherlands, the Criminal Career and Life-course Study (CCLS) is also used to study
intergenerational transmission of crime. This sample includes 3,015 men, who were prosecuted
in the Netherlands in 1977, and their 6,921 children. As a control group, 485 men who were never
convicted and their 1,066 children are used. For both groups, official conviction data was available
from age 12 onwards. Van de Rakt and colleagues (2009) found support for intergenerational
transmission of crime from parents to children.
Finally, the Transfive study ’s starting point is a sample of 198 boys who were placed in a Dutch
reform school in 1911. Official conviction data has been collected for these boys, their parents,
children, grandchildren, great-grandchildren, and all partners. Bijleveld and Wijkman (2009)
showed that crime was transmitted between the five consecutive generations and that this
transmission was generally stronger for more serious offenses. Results from this study also
showed that intergenerational transmission of crime was stronger for violent crimes than for
non-violent crimes (Van de Weijer et al., 2014). Moreover, the intergenerational transmission of
violence was shown to be mediated by parental divorce (Van de Weijer et al., 2015b) and by
the son’s heart rate level (Van de Weijer, 2014). Besemer (2012a), however, found a group of
fathers who were specialized in violent offending in a subsample of the Transfive study—similar
to the CSDD—but their children were not significantly more often violent offenders compared
to children of other offending fathers. Empirical evidence for transmission of sex offending
between siblings and from father to sons was found in the Transfive study as well (Van de
Weijer et al., 2015a).
Another method to study the intergenerational transmission of crime is the use of data about
the criminal behavior of an entire population of a city or country. Besjes and van Gaalen (2008)
studied all individuals in the Netherlands who were between the ages of 18 and 22 in 2005 and
found that having a parent who was suspected of a crime led to an increased risk for offspring
to be a suspected offender as well. Junger and colleagues (2013) gathered arrest data from the
members of all families with a child born in 2006 in a medium-sized city in the Netherlands
and showed a strong concentration of arrests within families; Junger et al. (2013) also found that
arrests were transmitted from grandparents to parents and from mothers to children. Evidence
for intergenerational transmission of (violent) crime was also found in studies that focused on
the entire Swedish population (e.g. Frisell et al., 2011; Kendler et al., 2014). Finally, Hjalmarsson
and Linquist (2012) studied all individuals born in 1953 and who lived in the Stockholm
metropolitan area in November 1963. They showed that children were at increased risk for
having a criminal conviction if their father had at least one conviction.
In addition to the studies described above, many more studies demonstrated evidence
supporting the intergenerational transmission of crime. However, many of these works used
limited research designs or measures. For instance, some of these works relied on a single family
member’s reports of the offending behavior of multiple relatives (e.g. Farrington et al., 2001;
Shlafer, 2010). With respect to differential measurement, some studies examined the effects of
parental imprisonments rather than parental offending on the criminal behavior of the next
generation (e.g. Murray et al., 2014; Skardhamar, 2009). Finally, other research that attempted to
evaluate the intergenerational nature of crime included a generation that was too young to
offend. Consequently, only externalizing behaviors were predicted (e.g. Capaldi et al., 2003;
Dong & Krohn, 2014; Kim et al., 2009). Nevertheless, these studies still contribute to the
literature on intergenerational transmission of crime and our understanding of the mechanisms
behind it.

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
285
Timing and frequency of parental crime
Some of the previously mentioned multigenerational studies took into account different dimen-
sions of the parent’s criminal career and examined how the timing and frequency of parental crime
affect offspring offending. Several studies used dummy variables to indicate whether or not a parent
(usually the father) offended during a certain time period. For instance, analyses from the Transfive
study showed that parental crime before the birth of the child did not lead to an increased risk of
offspring offending while parental crime after the child’s birth did (Bijleveld & Wijkman, 2009).
Van de Weijer and colleagues (2014) found similar results for violent offending and showed that
paternal violence during the offspring’s childhood (0–12 years) and adolescence (12–18 years)
increased the risk of offspring violence. The same study did not find a relationship between pater-
nal violence and offspring violence during the child’s adulthood (18 years or older). Using data
from the CSDD, Besemer (2014) found that children whose parents were only convicted before
the birth of the child had more convictions than children of non-criminal parents. When the par-
ents were (also) convicted during the child’s life, the child’s number of convictions increased (ibid.).
However, Besemer (2014) did not find significant differences in the relationship between parental
crime occurring at different age periods in the child’s life (i.e. 0–6 years, 7–12 years, 13–18 years).
Extant research also analyzed the influence of the timing of parental crime on offspring
offending using longitudinal panel designs in which the dependent variable (i.e. offspring offending)
is measured in each year of the child’s life. Using data from the CCLS, Van de Rakt and colleagues
(2010) performed logistic multilevel regression models and showed that children were at
increased risk of being convicted in the years in which their father was convicted. Interestingly,
this effect decreased over time, but the decay was weaker with each new conviction of the father.
Other research takes into account the frequency of parental offending using either a continuous
variable indicating the number of parental crimes or a categorical variable indicating the number
of parental crimes (e.g. 0 convictions, 1 conviction, 2–5 convictions, more than 5 convictions).
Generally, both approaches showed that the more crimes the father committed, the higher the
number of crimes committed by the child (Besemer, 2014; Van de Rakt et al, 2009; 2010; Van de
Rakt, 2011). Van de Rakt and colleagues (2009) found a similar relationship between the number
of convictions of mothers and their children.
Other research used semi-parametric group based trajectory models (see Nagin, 1999, 2005)
to investigate the influence of the timing and frequency of parental crime on the criminal behavior
of one’s child. Criminal trajectories of fathers from the CCLS were shown to be associated with
those of their children: fathers in groups with high offending rates (i.e. Medium Rate-Desisters
and High Rate-Persisters) more often had children who belonged to groups with high offending
rates (i.e. Early Desisters and Chronics) as well (Van de Rakt et al., 2008). In contrast, Besemer
(2012a) found less resemblance between fathers’ criminal trajectories and their offspring’s
criminal trajectories in the CSDD and the Transfive study. Although children of sporadic and
chronic offenders were convicted more often than children of non-offenders, the chronic
offenders did not have significantly more chronic offending children than sporadic offenders.
Similarly, Besemer et al. (2016) did not find significant relations between parent and child
offending trajectories in three generations in Sweden.
Methods
Sample: CSDD
The CSDD is a prospective longitudinal study that has followed 411 males from London, born
around 1953. At the time they were first contacted in 1961–2, these males were all living in a

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
286
working-class inner-city area of South London. The sample was chosen by taking all of the boys
who were then aged 8–9 and on the registers of six state primary schools within a 1-mile radius
of a research office that had been established. Hence, the most common year of birth for these
males was 1953. In nearly all cases (94 percent), their family breadwinner in 1961–2 (usually the
father) had a working-class occupation (skilled, semi-skilled, or unskilled manual worker). Most
of the boys were white and of British origin. Donald J. West originally directed the study and
David P. Farrington, who has worked on it since 1969, has directed it since 1982. The males have
been studied at frequent intervals between the ages of 8 and 50. Information on criminal
convictions and self-reported delinquency was collected over the course of these years.
Additionally, police records of offending of the parents and siblings of these 411 males have been
collected. For more information and major findings see West (1969, 1982), West and Farrington
(1973, 1977), Farrington and West (1990), Farrington (1995, 2003), Farrington et al. (2006), and
Farrington, Coid, and West (2009). For the research carried out in this chapter, the full biological
siblings of the original 411 men were included in the analyses.
Sample: RYDS and RIGS
RYDS is a multiwave panel study analyzing the development of delinquency and drug use
through adulthood among a high-risk sample of adolescents attending the public school system
of Rochester, New York (N = 1,000). The target population was seventh- and eighth-grade
students. Because the base rates for serious delinquency and drug use are relatively low (Elliott
et al., 1989; Wolfgang et al., 1987), RYDS oversampled youth who were at a high risk for serious
delinquency and drug use across two dimensions. First, males were oversampled (75 percent
versus 25 percent) because they are more likely than females to be chronic offenders and engage
in serious forms of delinquency (Blumstein et al., 1986). Second, students from areas of the city
where large populations of adult offenders lived were oversampled on the premise that youth
residing in these areas are at a greater risk for offending. To identify high arrest rate areas, each
census tract in Rochester was assigned a resident arrest rate reflecting the proportion of the total
population living in that tract that was arrested by the Rochester police in 1986. Subjects were
oversampled proportionate to the rate of offenders living in a tract.
Data collection began in 1988 when members of the sample were either in seventh or eighth
grade (average age 14). A total of 14 waves of data were collected during three phases of data
collection. The first phase of data collection (Phase 1) covered adolescence, from 14 to 18 years
of age. During Phase 1, each respondent was interviewed nine times (waves 1–9) and their
parents were interviewed eight times at six month intervals. Phase 2 began after a 2.5 year gap
in the data collection. The respondents and their parents were interviewed at three annual intervals
(average age 21–23). Phase 3, waves 13 and 14, consists of respondent interviews at age 29 and
31 years of age. For the duration of the study, additional data on arrests and police contacts were
also collected from the Rochester Police Department, the New York State Division of Criminal
Justice Services (DCJS), and the County Wide Registration for juveniles for Monroe County,
New York. At the time of this analysis though, only official data on police contacts and arrests
was available through 1997 (the end of Phase 2).
RIGS began in 1999, and is the intergenerational component of RYDS. The initial goals of
RIGS were to describe intergenerational continuity and discontinuity in antisocial behavior
across the generations and to identify the mediating processes that explain the intergenerational
transmission of antisocial behavior. Currently, RIGS is a three-generation prospective panel
study consisting of a subsample of the original RYDS participants (G2s), the first born biological
child of the original subjects of RYDS (G3s), and the primary caregiver of the original RYDS

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
287
respondents (G1s; data originally available in RYDS). The focal subjects for the intergenerational
study are the oldest biological child of each of the original RYDS subjects. Each RYDS respondent
and child are eligible to enter the study when the child turns 2 years old (340 G3s in year 1 of
data collection). Subsequently, annual interviews are conducted with the original RYDS respondent
(G2), another primary caregiver for the child (OCG—not always a parent), and all children aged
8 and older (G3). Self-report information from the parent and other primary caregiver was
collected through the G3 age of 17. The collection of self-report data from the child is ongoing.
In the current study only data from the G2s and G3s are analyzed. Furthermore, the G3s who
were not 18 at the last time of data collection were excluded from the sample. This yields a
sample size of 295 G2 and G3 pairs for analysis.
Sample: Transfive study
The Transfive study contains data on individuals from five consecutive generations. The second
generation (G2) consists of 198 boys who were placed in a Dutch Catholic reform school
between 1911 and 1914. These boys constituted a high-risk sample in terms of offending since
they were placed in this institution due to concerns about their character and problem behavior
or because their parents were not able to take proper care of them. Their parents (G1), children
(G3), grandchildren (G4) and great-grandchildren (G5) were traced in Dutch genealogical and
municipal records, with a retrieval rate of 100 percent. Emigrated sample members were
considered as lost to follow-up and their descendants were not traced. More detailed information
on the Transfive study can be found in Bijleveld and colleagues (2007) and Van de Weijer (2014).
In the current study only the data on the G3, G4, and G5 were taken into account. On
average, the G3, G4 and G5 were born in 1932, 1960, and 1986, respectively. For the analyses
focusing on offspring offending in adulthood (between 18 and 40 years old), data on the G4 and
their biological parents (G3) was used. Only those G4 who reached the age of 40 at the moment
of data collection (December 2007) were included in the analyses in order that offending was
measured up to age 40 for all participants. For the analyses focusing on offspring adolescence
offending (up to age 18), data on the G5 and their biological parents (G4) was used. The G5 who
had not reached the age of 18 at the moment of data collection were excluded from the analyses.
Measurements
The dependent variables in this study are a binary variable indicating whether or not a child
offended and a count variable indicating the number of times the child committed a crime. For
all three studies these dependent variables were measured up to the eighteenth birthday. For the
CSDD and Transfive study these dependent variables were measured between the eighteenth
and fortieth birthdays as well. The independent variables in this study are variables indicating
parental offending and parental offending during four time periods: prior to the child’s birth,
during the child’s early childhood (0–5 years), during the child’s late childhood (6–11 years) and
during the child’s adolescence (12–17 years). These variables have been measured as a binary and
a count variable as well. The child’s gender and the parent’s age at the child’s birth (and the parent’s
gender in RYDS) were included as control variables in all analyses. The next section specifically
describes how these variables are measured in each of the three studies.
Measurement: CSDD
For the CSDD , convictions were searched in the central Criminal Record Office in London
(see Farrington et al ., 1996). The date when the offense was committed was used to time the

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
288
delinquent incidents. If no commission date was known, the conviction date was used. Offenses
were defined as acts leading to convictions, and only one offense per day was counted. This
rule was adopted so that each separate behavioral act could yield only one offense; if all
offenses had been counted, the number of offences would have been greater than the number
of criminal behavioral acts, resulting in an overestimation of criminal behavioral acts
(Farrington et al., 2006).
Measurement: RYDS
To construct the official arrest histories of G2s and G3s in the RYDS and RIGS data, two
different data sources from RYDS and RIGS are used. Official police data on police contacts
and arrests from the RYDS data were used to construct arrest histories through the end of Phase
2 (average age 23). Information on the age at each arrest and the type of arrest was available.
Furthermore, any arrest that was related to a minor traffic offense (e.g. speeding) was excluded
from the sample for continuity across the RYDS and RIGS data sets as well as the other data sets
included in this research. Although data regarding official police contacts and arrests were
collected for G2s for the duration of RYDS, data regarding arrests after 1997 are not available
for analysis. Therefore, the official police arrest histories for G2s are supplemented with self-
report data on arrests for the duration of RYDS and the entirety of RIGS. Beginning in Phase
3 of RYDS, respondents used life history calendars to identify the number of arrests that
occurred in each month since the date of the wave 12 interview. Additionally, when G2s entered
RIGS, each yearly interview contained questions regarding whether or not the subject was
arrested for anything other than a minor traffic violation since the date of the last interview and,
if so, how many times they were arrested. This allowed for the construction of arrest histories for
G2s through the time when the G2s’ first born child (G3) was 17 years old. Additional efforts
were made to ensure that there were not any gaps or overlaps in the official arrest records,
self-reported arrests from the Phase 3 life history calendar data, and the RIGS yearly interviews.
All information regarding the arrest histories of G3s come from self-report data in RIGS. In
the age 12 interview, respondents (G3s) were queried regarding whether or not the subject was
arrested previously, and if so how many times and at what age(s). Subsequent yearly interviews
included questions regarding whether or not the subject was arrested since the date of the last
interview (at age 16 the wording changed to exclude arrests for minor traffic violations) and, if
so, how many times they were arrested.
With the available information, it was possible to construct the full arrest histories for the G2s
and G3s by age. In other words, two variables were created for each subject at each age—
whether or not the subject was arrested at each age and how many times the subject was arrested
at each age. This information was then used to create the variables used for the analyses presented
in this research.
Measurement: Transfive study
Data on the criminal behavior of participants—both children and parents—were obtained from
the archives of the Dutch Criminal Records Documentation Service, in December 2007. All
registrations which resulted in a conviction were considered as offenses. The date of offenses was
used to determine parental offending in each time period. If the offense date was unknown it
was estimated as one year prior to the conviction date, since that is the average time period
between conviction dates and known offense dates. If the conviction date was also unknown, the
offense date was timed as July 1 of the year of registration.

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
289
Information on the gender, dates of birth, and dates of death of the participants was obtained
from the Dutch population registration data (GBA). These records did not contain any information
on gender for those born after 1994. In virtually all cases, however, it was possible to determine
the gender from the given name. Based on these records the control variables—child’s gender,
and parent’s age at birth of the child—were constructed.
Analyses
Logistic regression analyses were used to examine the intergenerational transmission of offending
with the binary variables, and negative binomial regression analyses were used for the count
variables given the presence of overdispersion for each of the outcomes. The first two analyses
examine the effects of having an offending parent(s) and the frequency of parental offenses on
offspring offending. In the third and fourth analyses the timing of the parental offense(s) are
taken into account by examining the separate effects of parental offending prior to the child’s
birth, early childhood (0–5 years), late childhood (6–11 years) and adolescence (12–17 years).
These analyses were conducted for offspring offending up to age 17 for all three studies.
Additional analyses for offspring offending between ages 18 and 40 were conducted for the
CSDD and Transfive study.
Since both the CSDD and the Transfive study include multiple children from the same
parents, the assumption of independence among observations is violated. Therefore, generalized
estimation equations (GEE-models) were used to estimate the logistic and negative binomial
regression models in order to control for this intra-cluster correlation. These models ignore the
correlations between repeated measures (i.e. the children) within the same family and correct
the standard errors of regression coefficients by estimating robust standard errors. For RYDS/RIGS,
such GEE-models were not necessary since its sample only includes one child for each parent.
Results
In Table 16.1 , the descriptives of all variables used in the analyses are shown. The prevalence of
offending
1 children through age 18 is 21 percent, 18 percent and 12 percent among the participants
of RYDS/RIGS, the CSDD, and the Transfive study, respectively. Moreover, 25 percent of the
participants of the CSDD committed a crime before their fortieth birthday, while this percentage
is 32 percent for the participants of the Transfive study.
In RYDS/RIGS (65 percent) and the Transfive study (52 percent by age 18 and 55 percent
by age 40) more than half of the children had a parent who offended at least once. Only
37 percent of the children in the CSDD had (an) offending parent(s). The number of parental
offenses is the highest among the Transfive study, with each set of these parents committing, on
average, 3.75 offenses between them. The timing of parental offending is quite similar for the
CSDD and Transfive study: parental offending is most prevalent prior to the child’s birth and
parents offend at a relatively similar rate during the three time periods during the youth of the
child. On the other hand, the parents in RYDS/RIGS commit most crimes during their child’s
early childhood.
Table 16.2 shows the results of the logistic regression analyses in which parental offending
predicts offspring offending, up to age 18 and from 18 to 40. The results for the child’s offending
up to age 18 are very similar across studies. All three studies found a significant intergenerational
transmission of offending. The odds ratios of 2.80 (Transfive study), 3.16 (CSDD) and 3.42
(RYDS/RIGS) indicate that children with (an) offending parent(s) are more likely to commit a
crime before they reach the age of 18, compared to those with (a) non-offending parent(s).

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
290
The results for offspring offending from age 18 to 40 show that parental offending is also
related to their children’s offending behavior during adulthood as significant odds ratios were
found for both the CSDD (2.92) and the Transfive study (1.77).
Table 16.3 shows the results of the negative binomial regression analyses in which the number
of offenses committed by the children are regressed on the parents’ offending frequency and
control variables. Again, the relationship between parental offending and child offending are
Table 16.1 Descriptives of all variables used in the analyses
Binary variables CSDD RYDS Transfive study
G3–G4 G4–G5
Dependent variables
Offending child up to age 18 18.12% 21.01% — 11.98%
Offending child from
18 to 40
25.13% — 32.33% —
Independent variables
Any offending parent 37.04% 65.08% 52.27% 54.90%
Timing of parental offending
Prior to birth child 27.29% 31.53% 35.45% 36.84%
Age child 0 to 5 14.30% 48.14% 14.80% 21.42%
Age child 6 to 11 11.48% 26.44% 15.41% 22.23%
Age child 12 to 17 11.62% 17.63% 13.70% 22.32%
Control variables
Gender parent (1 = male)
* — 56.27% — —
Gender child (1 = male) 63.75% 51.53% 51.96% 52.09%
Continuous and count
variables
CSDD RYDS Transfive
study
G3–G4 G4–G5
Dependent variables M SD M SD M SD M SD
# offenses child up to
age 18
0.46 1.51 0.62 1.98 — — 0.36 1.72
# offenses child from 18 to
40
0.87 2.46 — — 1.86 7.32 — —
Independent variables
# offenses parents 1.38 2.77 — — 1.49 2.67 3.75 11.66
# offenses parent
* — — 3.36 4.64 — — — —
Timing of parental offending
# offenses prior to birth child 0.72 1.65 0.97 2.09 0.76 1.65 1.44 5.52
# offenses age child 0 to 5 0.22 0.70 1.59 2.57 0.24 0.74 0.82 3.79
# offenses age child
6 to 11
0.21 0.79 0.52 1.10 0.25 0.70 0.74 2.56
# offenses age child
12 to 17
0.22 0.81 0.27 0.69 0.24 0.84 0.76 3.06
Control variables
Age of father at birth child 27.66 6.21 — — 30.14 5.90 27.54 4.50
Age of mother at birth child 31.52 7.33 — — 26.85 5.19 24.88 4.00
Age of parent at birth child — — 18.90 2.18 — — — —
Number of children (N) 1,385 295 993 1,102
Note: *Only in RYDS data.

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
291
significant in all three studies. For offspring offending up to age 18, incidence rate ratios (IRR)
of 1.19 (CSDD), 1.11 (RYDS/RIGS) and 1.04 (Transfive study)
2 were found.
The results for the frequency of offspring offending from age 18 to 40 also show significant
intergenerational transmission of offending. The significant estimates indicate that each additional
parental offense increased the number of offenses committed by their children. For the CSDD
an IRR of 1.21 was found, while the IRR was 1.12 for the Transfive study.
Tables 16.4 and 16.5 take into account the timing of the parental offense(s). Table 16.4 shows
the results of the logistic regression analyses. The results for offspring offending up to age 18 vary
across the three studies. For children in the CSDD, having (a) parent(s) offend in each of the four
time periods significantly increases the odds of committing a crime during adolescence. On the
other hand, the increased risk of offending is only significant if a child has a parent who offended
during early childhood (0–5 years) in RYDS/RIGS. In the Transfive study, children are at
increased risk of being convicted during adolescence if their parent(s) commit(s) a crime prior
Table 16.2 Logistic regression analyses predicting child offending
Child offending up to age 18 Child offending from 18 to 40
CSDD RYDS Transfive study CSDD Transfive study
Variables OR OR OR OR OR
Constant 0.04
*** 0.04
* 0.05
*** 0.07
*** 0.13
***
Parental offending 3.16
*** 3.42
** 2.80
*** 2.92
*** 1.77
***
Control variables
Gender child (male=1) 6.30
*** 3.53
*** 5.42
*** 4.71
*** 6.49
***
Gender parent (male=1) — 1.07 — —
Mother’s age at birth 0.98 — 0.96 0.99 1.00
Father’s age at birth 1.02 — 1.01 1.01 1.00
Parent’s age at birth — 1.01 — —
Number of children (N) 1,385 295 1,102 1,385 993
Note: ***p < 0.001; **p < 0.01; *p < 0.05 (two-sided); OR = odds ratio.
Table 16.3 Negative binomial regression analyses predicting number of offenses child
Child offending through age 17 Child offending through age 40
CSDD RYDS Transfive study CSDD Transfive study
Variables IRR IRR IRR IRR IRR
Constant 0.05
*** 1.01 0.31
* 0.17
*** 0.42
***
Number of parental offenses 1.19
*** 1.11
* 1.04
*** 1.21
*** 1.12
***
Control variables
Gender child (male = 1) 7.85
*** 2.41
** 8.50
*** 5.37
*** 4.18
***
Gender parent (male = 1) — 1.36 — — —
Mother’s age at birth 0.98 — 0.90
*** 0.98 1.02
Father’s age at birth 1.02
* — 1.03 1.01 0.99
Parent’s age at birth — 0.91 — — —
Number of children (N) 1,385 295 1,102 1,385 993
Note: ***p < 0.001; **p < 0.01; *p < 0.05 (two-sided); IRR = incidence rate ratio.

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
292
to their child’s birth, when the child is 0–5 years old and when the child is 6–11 years old. The
significant odds ratios range from 1.69 to 1.90 across the studies.
Table 16.4 shows the results for offspring offending between age 18 and 40 for the CSDD
and the Transfive study. In both studies, only parental offending prior to the child’s birth and
during the child’s adolescence (12–17 years) increase the child’s risk of becoming an offender
during adulthood.
Table 16.5 shows the results of the negative binomial regression analyses in which the child’s
offending frequency is predicted by the parents’ offending frequency during the four time
periods and control variables. The results again differ across the three studies, possibly as a
Table 16.4 Logistic regression analyses predicting child offending by parental offending at different
periods
Child offending up to age 18 Child offending from 18 to 40
CSDD RYDS Transfive study CSDD Transfive study
Variables OR OR OR OR OR
Constant 0.04
*** 0.07
* 0.03
*** 0.06
*** 0.11
***
Offenses prior to birth (yes = 1) 1.68
** 1.28 1.86
** 2.03
*** 1.48
*
Offenses when child is 0-5 (yes = 1) 1.80
** 1.88
* 1.65
* 1.35 1.07
Offenses when child is 6-11 (yes = 1) 1.68
* 1.43 1.63
* 1.42 1.34
Offenses when child is 12-17 (yes = 1) 1.90
** 1.34 1.31 2.23
*** 1.99
**
Gender child (male = 1) 6.96
*** 3.42
*** 5.75
*** 5.21
*** 6.69
***
Gender parent (male = 1) — 1.09 — — —
Mother’s age at birth 0.97 — 0.97 0.99 —
Father’s age at birth 1.02 — 1.02 1.01 1.00
Parent’s age at birth — 1.00 — — 1.00
Number of children (N) 1,385 295 1,102 1,385 993
Note: ***p < 0.001; **p < 0.01; *p < 0.05 (two-sided); OR = odds ratio.
Table 16.5 Negative binomial regression analyses predicting child offending by parental offending at
different periods
Child offending through age 17 Child offending through age 40
CSDD RYDS Transfive study CSDD Transfive study
Variables IRR IRR IRR IRR IRR
Constant 0.06
*** 1.84 0.33
* 0.16
*** 0.33
***
Offenses prior to birth (yes = 1) 1.12
*** 1.19
* 1.05
*** 1.15
*** 1.96
Offenses when child is 0–5 (yes = 1) 1.09 1.08 1.03
* 1.25
*** 1.07
Offenses when child is 6–11 (yes = 1) 1.35
*** 1.11 1.06
** 1.28
*** 1.43
***
Offenses when child is 12–17 (yes = 1) 1.27
*** 0.81 1.04
* 1.30
*** 1.26
***
Gender child (male = 1) 7.85
*** 2.56
** 8.33
*** 5.42
*** 4.18
***
Gender parent (male = 1) — 1.38 — — —
Mother’s age at birth 0.98 — 0.90
*** 0.98 1.02
Father’s age at birth 1.02
* — 1.03 1.01 1.00
Parent’s age at birth — 0.89 — — —
Number of children (N) 1,385 295 1,102 1,385 993
Note: ***p < 0.001; **p < 0.01; *p < 0.05 (two-sided); IRR = incidence rate ratio.

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
293
consequence of the differences in measurement. In the Transfive study, each additional parental
offense in each time period increases the number of child offenses up to the age of 18. Each
additional parental offense committed prior to the child’s birth, when the child is between 6 and
11 years old or when the child is between 12 and 17 years old significantly increases the number
of offenses for CSDD-children up to age 18 only. In RYDS/RIGS, only the number of parental
crimes prior to the child’s birth significantly increases the child’s number of offenses up to age 18.
Table 16.5 also provides the results for offspring offending between age 18 and 40 for the
CSDD and Transfive samples. For the children in the CSDD-study, the number of parental
offenses during all four time periods significantly increases the number of offspring offenses
between ages 18 and 40. However, only the number of offenses committed by one’s parent
between 6 and 11 years old and between 12 and 17 years old significantly increase the number
of child offenses between the ages of 18 and 40 for children in the Transfive study.
Looking at the significance of other variables that were not central to the intergenerational
analysis, results demonstrate that men are more likely to offend than women, during both ado-
lescence and adulthood. Furthermore, the parent(s)’ age and gender in general are not related to
offspring offending.
Conclusion
In this study, the intergenerational transmission of offending was examined using three studies,
from England (CSDD), the United States (RYDS), and the Netherlands (Transfive study).
Moreover, the influence of the timing and frequency of parental offenses on intergenerational
transmission was examined, as well as whether or not the effects of parental offending on
offspring offending are different during adolescence and adulthood.
First, this research is in line with several criminological theories that predict intergenerational
transmission of offending, and numerous studies. The results showed a significant degree of
intergenerational transmission of offending in all three studies. The odds ratios for offspring
offending up to age 18 indicated that children of offenders are 2.80 to 3.42 times more likely to
offend during adolescence compared with children of non-offenders. The results for offspring
offending between age 18 and 40 were also significant but showed that the degree of intergen-
erational transmission of offending was relatively low in the Transfive study. This finding is not
remarkable since the transmission from the third generation (G3) to the fourth generation (G4)
of the Transfive study was estimated in this analysis. Previous results of the Transfive study already
showed that the degree of transmission from G3–G4 was lower than from G4–G5 (see e.g.
Bijleveld & Wijkman, 2009; Van de Weijer et al., 2014), possibly due to cohort or period effects.
3
Second, this work also examined the relationship between the frequency of parental offending
and the frequency of offspring offending using negative binomial regression models. Among the
three samples, there was evidence that the frequency of parental offending influenced the
frequency of child offending, in both adolescence and adulthood. Thus, the offending behavior
of children is not only related to whether or not a parent is criminal but also to the frequency
of parental offending. Importantly, non-offending parents and their children are included in this
second analysis which may drive some of the results. Ideally, additional analyses should investigate
the relationship between the frequency of parental offending and the frequency of offspring
offending only among parents who offended to further verify this relationship. Unfortunately
this was not possible since excluding non-offending parents and their children would result in
low sample sizes and consequently low statistical power in these data sets.
Third, this study investigated the influence of the timing of parental offenses on offspring
offending by examining the separate effects of parental crimes prior to the child’s birth, parental

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
294
crimes during the child’s early childhood (0–5 years), parental crimes during the child’s late
childhood (6–11 years) and parental crimes during the child’s adolescence (12–17 years). The
results of these analyses were not consistent across data sets as the significance of some effects
differed within and across studies. The relationship between the binary indicators of parental
offending during the four periods and offspring offending were significant for the CSDD and
Transfive study. This was not the case in the RYDS/RIGS data. These differences are possibly
due to different measurement for offending (arrests versus convictions; self-report versus official
data) or the relatively low sample size.
Given these results, it is important to acknowledge that some of these findings vary from
other investigations as well. The finding that parental crimes before the birth of the child are
related to the child’s criminal behavior is in contrast to some previous studies that found no
effect of parental crimes committed before the child was born (e.g. Bijleveld & Wijkman, 2009;
Van de Weijer et al., 2014). The latter studies, however, compared parents who only offended
prior to their child’s birth with non-offending parents. Parents who offended during the youth
of the child were, thus, excluded from the analyses. The current study, on the other hand, includes
all parents regardless of if and when they offended, as does Besemer (2014). Thus, the divergent
findings may be a result of sample differences.
As with any research, the current study has its strengths and limitations. A strength of this
study is that it is among the first studies to verify the intergenerational offending across three
countries: England, the United Sates, and the Netherlands. All three data sets used prospective
longitudinal data from a large sample of individuals from consecutive generations. Another
strength of the current study is that it takes into account the diversity within the offender
population with respect to the criminal career. Individuals are not simply divided into groups of
offenders and non-offenders but the frequency and timing of offenses were taken into account
to examine how they condition the intergenerational transmission of crime.
This study, however, is also limited in some ways. First of all, the samples from each of the
three countries consist of individuals at high risk for criminal development. As a consequence,
the results of the current study might not be generalizable to the entire population of each
country. Another limitation is the fact that the measurement of crime differs somewhat across
countries. In the CSDD and Transfive study, offending from each generation was measured by
official data, while in RYDS/RIGS, the offending behavior of the youngest generation was only
measured by self-report data. Furthermore, the CSDD and Transfive data use convictions to
measure offending whereas the RYDS data uses arrests. While these differences limit comparisons
across data sets, they should not prevent conclusions regarding the importance of the intergen-
erational transmission of crime and the conditioning effect of different elements of the criminal
career of a parent on the offending of offspring cross-nationally.
In sum, the results of this study show that crime is transmitted across generations within three
different countries. In addition, it shows that offspring offending is not only related to whether or
not a parent offends but also to the timing and frequency of parental offenses. Intergenerational
transmission of offending is thus an international phenomena, at least in Western countries. As
the current study and most previous studies on this topic examine individuals from Western
countries—England, the United States, the Netherlands, and Scandinavian countries in particular—
it would be interesting for future studies to focus on other, non-Western, countries. Moreover,
it would be interesting if more studies would make international comparisons when studying
the intergenerational transmission of crime. If such international comparisons are made, it is
important to use consistent measures of crime across studies in order to enable reliable comparisons
of results. For future research it would also be desirable to focus more on the mechanisms
underlying the intergenerational transmission rather than establishing that crime is transmitted

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
295
across generations. Examining possible explanations and how these interact with each other
would be important for policy as it would offer insights in how the cycle of crime can be
interrupted. In this respect it would also be an important topic for further research to study
possible protective factors that prevent offspring of criminal parents from becoming offenders
and factors that increase the odds of intergenerational transmission taking place. Preventive
measures can then be aimed at those who are most likely to continue in a cycle of crime.
Notes
1 As described in the previous section, different measurements for offending are used in the three studies
(i.e. arrests and convictions; self-reports; and official data). Whenever ‘offending’ or ‘committing crimes’
is mentioned in this section it refers to the measurement of the specific study.
2 The relatively low incidence rate ratio of the Transfive study seems to be the consequence of some
outliers. Some children have parents who together committed up to 139 offenses. When the number of
parental offenses is set to a maximum of 25 offenses, the incidence rate ratio for the Transfive increases
to 1.12 and is, thus, quite similar to the incidence rate ratios of the other studies.
3 For a discussion of these cohort or period effects see Van de Weijer (2014).
References
Besemer , S. ( 2012a ). Intergenerational transmission of criminal and violent behaviour (Doctoral dissertation).
Leiden, The Netherlands : Sidestone Press .
Besemer , S. ( 2012b ). Specialized versus versatile intergenerational transmission of violence: A new approach
to studying intergenerational transmission from violent versus non-violent fathers: Latent class analysis .
Journal of Quantitative Criminology , 28 ( 2 ), 245–63 .
Besemer , S. ( 2014 ). The impact of timing and frequency of parental criminal behaviour and risk factors on
offspring offending . Psychology, Crime & Law , 20 ( 1 ), 78–99 .
Besemer , S. & Farrington , D. P. ( 2012 ). Intergenerational transmission of criminal behaviour: Conviction
trajectories of fathers and their children . European Journal of Criminology , 9 ( 2 ), 120–41 .
Besemer , S. , Farrington , D. P. & Bijleveld , C. C. J. H. ( 2013 ). Official bias in intergenerational transmission
of criminal behaviour . British Journal of Criminology , 53 ( 3 ), 438–55
.
Besemer, S., Axelsson, J., & Sarnecki, J. (2016). Intergenerational transmission of trajectories of offending
over three generations. Journal of Developmental and Life-Course Criminology, 2 (4), 417–41.
Besjes , G. & Van Gaalen , R. ( 2008 ). Jong geleerd, fout gedaan . Bevolkingstrends , 2 , 23–31 .
Bijleveld , C. C. J. H. & Wijkman , M. ( 2009 ). Intergenerational continuity in convictions: A five-generation
study . Criminal Behaviour and Mental Health , 19 ( 2 ), 142–55 .
Bijleveld , C. C. J. H. , Wijkman , M. & Stuifbergen , J. A. M. ( 2007 ). 198 Boefjes? Leiden, The Netherlands : NSCR .
Blumstein , A. , Cohen , J. , Roth , J. A. & Visher , C. A. (Eds.) ( 1986 ). Criminal careers and ‘career criminals’ .
Washington, DC : National Academy Press
Boutwell , B. B. , Beaver , K. M. & Barnes , J. C. ( 2012 ). More alike than different assortative mating and
antisocial propensity in adulthood . Criminal Justice and Behavior , 39 ( 9 ), 1240–54 .
Capaldi , D. M. , Pears , K. C. , Patterson , G. R. & Owen , L. D. ( 2003 ). Continuity of parenting practices across
generations in an at-risk sample: A prospective comparison of direct and mediated associations . Journal
of Abnormal Child Psychology , 31 ( 2 ), 127–42 .
Christianson , S. ( 2003 , February 8). Bad seed or bad science? The story of the notorious Jukes family .
New York Times .
Dong , B. & Krohn , M. D. (2014). Exploring intergenerational discontinuity in problem behavior: Bad
parents with good children . Youth Violence and Juvenile Justice , 13 ( 2 ), 99–122 .
Dugdale , R. L. ( 1877 ).
The Jukes: A study in crime, pauperism, and heredity . New York : Putnam .
Elks , M. A. ( 2005 ). Visual indictment: A contextual analysis of the Kallikak family photographs . Journal
Information , 43 ( 4 ), 268–80 .
Elliott , D. S. , Huizinga , D. & Menard , S. ( 1989 ). Multiple problem youth . New York : Spr inger Science & Business Media .
Farrington , D. P. ( 1995 ). The development of offending and antisocial behaviour from childhood: Key
findings from the Cambridge Study in Delinquent Development . Journal of Child Psychology and Psychiatry ,
36 ( 6 ), 929–64 .

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
S. van de Weijer, M.B. Augustyn and S. Besemer
296
Farrington , D. P. ( 2003 ). Key results from the first forty years of the Cambridge Study in Delinquent
Development . In T. P. Thornberry & M. D. Krohn (Eds.), Taking stock of delinquency: An overview of findings
from contemporary longitudinal studies (pp. 137–83 ). New York : Kluwer .
Farrington , D. P. ( 2011 ). Families and crime . In J. Q. Wilson & J. Petersilia (Eds.), Crime and public policy
(pp. 130–57 ). New York : Oxford University Press .
Farrington , D. P. & West , D. J. ( 1990 ). The Cambridge Study in Delinquent Development: A long-term
follow-up of 411 London males . In H.-J. Kerner & G. Kaiser (Eds.), Kriminalität: Personlichkeit,
Lebensgeschichte und Verhalten [Criminality: Personality, life history and criminal behaviour] (p. 115–38 ). Berlin :
Springer-Verlag .
Farrington , D. P. , Barnes , G. C. & Lambert , S. ( 1996 ). The concentration of offending in families . Legal and
Criminological Psychology , 1 ( 1 ), 47–63 .
Farrington , D. P. , Jolliffe , D. , Loeber , R. , Stouthamer-Loeber , M. & Kalb , L. M. ( 2001 ). The concentration
of offenders in families, and family criminality in the prediction of boys’ delinquency . Journal of
Adolescence , 24 ( 5 ), 579–96 .
Farrington , D. P. , Coid , J. W. , Harnett , L. , Jolliffe , D. , Soteriou , N. , Turner , R. , West , D. J. ( 2006 ). Criminal
careers up to age 50 and life success up to age 48: New findings from the Cambridge Study in Delinquent
Development . London : Home Office (Home Office Research Study No. 299 ).
Farrington , D. P. , Coid , J. W. & Murray , J. ( 2009 ). Family factors in the intergenerational transmission of
offending . Criminal Behaviour and Mental Health , 19 ( 2 ), 109–24 .
Farrington , D. P. , Coid , J. W. & West , D. J. ( 2009 ). The development of offending from age 8 to age 50:
Recent findings from the Cambridge Study in Delinquent Development . Monatsschrift für Kriminologie
und Strafrechtsreform [Journal of Criminology and Penal Reform] , 92 ( 2 ), 160–73 .
Frisell , T. , Lichtenstein , P. & Långström , N. ( 2011 ). Violent crime runs in families: A total population study
of 12.5 million individuals . Psychological medicine , 41 ( 1 ), 97–105 .
Giordano , P. C . ( 2010 ). Legacies of crime: A follow-up of the children of highly delinquent girls and boys . Cambridge,
UK : Cambridge University Press .
Goddard , H. H. ( 1912 ). The Kallikak family: A study in the heredity of feeble-mindedness . New York : Macmillan .
Hjalmarsson , R. & Lindquist , M. J. ( 2012 ). Like godfather, like son: Exploring the intergenerational nature
of crime . Journal of Human Resources , 47 ( 2 ), 550–82 .
Junger , M. , Greene , J. , Schipper , R. , Hesper , F. & Estourgie , V. ( 2013 ). Parental criminality, family violence
and intergenerational transmission of crime within a birth cohort . European Journal on Criminal Policy
and Research , 19 ( 2 ), 1–17 .
Karp , R. J. , Qazi , Q. H. , Moller , K. A. , Angelo , W. A. & Davis , J. M. ( 1995 ). Fetal alcohol syndrome at the
turn of the 20th century: An unexpected explanation of the Kallikak family . Archives of Pediatrics &
Adolescent Medicine , 149 ( 1 ), 45–8 .
Kendler , K. S. , Ohlsson , H. , Morris , N. A. , Sundquist , J. & Sundquist , K. ( 2014 ). A Swedish population-based
study of the mechanisms of parent–offspring transmission of criminal behavior . Psychological Medicine ,
45 , 1–10 .
Kim , H. K. , Capaldi , D. M. , Pears , K. C. , Kerr , D. C. & Owen , L. D. ( 2009 ). Intergenerational transmission
of internalising and externalising behaviours across three generations: Gender-specific pathways .
Criminal Behaviour and Mental Health , 19 ( 2 ), 125–41 .
Krueger , R. F. , Moffitt , T. E. , Caspi , A. , Bleske , A. & Silva , P. A. ( 1998 ). Assortative mating for antisocial
behavior: Developmental and methodological implications . Behavior Genetics , 28 ( 3 ), 173–86 .
Murray , J. , Bijleveld , C. C. , Farrington , D. P. & Loeber , R. ( 2014 ). Effects of parental incarceration on children:
Cross-national comparative studies . Washington DC : American Psychological Association .
Nagin , D. S. ( 1999 ). Analyzing developmental trajectories: A semiparametric, group-based approach .
Psychological Methods , 4 ( 2 ), 139 .
Nagin , D. S. ( 2005 ). Group-based modeling of development . Cambridge, MA: Harvard University Press .
Rhee , S. H. & Waldman , I. D. ( 2002 ). Genetic and environmental influences on antisocial behavior:
A meta-analysis of twin and adoption studies . Psychological bulletin , 128 ( 3 ), 490–529 .
Rowe , D. C . & Farrington , D. P. ( 1997 ). The familial transmission of criminal convictions . Criminology ,
35 ( 1 ), 177–202 .
Shlafer , R. J. ( 2010 ). Intergenerational transmission of criminal behavior: Understanding the effects of family criminality,
interparental violence, maltreatment, and hostile parenting (Doctoral dissertation). University of Minnesota.
Skardhamar , T. ( 2009 ). Family dissolution and children’s criminal careers . European Journal of Criminology ,
6 ( 3 ), 203–23 .
Thornberry , T. P. ( 1987 ). Toward an interactional theory of delinquency . Criminology , 25 ( 4 ), 863–91 .

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Intergenerational transmission of crime
297
Thornberry , T. P. ( 2005 ). Explaining multiple patterns of offending across the life course and across
generations . Annals of the American Academy of Political and Social Science , 602 ( 1 ), 156–95 .
Thornberry , T. P. ( 2009 ). The apple doesn’t fall far from the tree (or does it?): Intergenerational patterns of
antisocial behavior—the American Society of Criminology 2008 Sutherland Address . Criminology ,
47 ( 2 ), 297–325 .
Thornberry , T. P. & Krohn , M. D. ( 2001 ). The development of delinquency. An interactional perspective . In
S. O. White (Ed.), Handbook of youth and justice (pp. 289–305 ). New York : Kluwer Academic – Plenum
Publishers .
Thornberry , T. P. & Krohn , M. D. ( 2005 ). Applying interactional theory to the explanation of continuity and
change in antisocial behavior . In D. P. Farrington (Ed.), Integrated developmental and life-course theories of
offending (pp. 183–209 ). London : Transaction Publishers .
Thornberry , T. P.
, Freeman-Gallant , A. , Lizotte , A. J. , Krohn , M. D. & Smith , C. A. ( 2003 ). Linked lives: The
intergenerational transmission of antisocial behavior . Journal of Abnormal Child Psychology , 31 ( 2 ), 171–84 .
Thornberry , T. P. , Freeman-Gallant , A. & Loveg rove , P. J . ( 2009 ). Intergenerational linkages in antisocial
behaviour . Criminal Behaviour and Mental Health , 19 ( 2 ), 80–93 .
Van de Rakt , M. G. A. ( 2011 ). Two generations of crime: The intergenerational transmission of criminal convictions
over the life course (Doctoral dissertation). Nijmegen : Ipskamp Drukkers .
Van de Rakt , M. , Nieuwbeerta , P. & De Graaf , N. D. ( 2008 ). Like father, like son: The relationships between
conviction trajectories of fathers and their sons and daughters . British Journal of Criminology , 48 ( 4 ),
538–56 .
Van de Rakt , M. , Nieuwbeerta , P. & Apel , R. ( 2009 ). Association of criminal convictions between family
members: Effects of siblings, fathers and mothers . Criminal Behaviour and Mental Health , 19 ( 2 ), 94–108 .
Van de Rakt , M. , Ruiter , S. , De Graaf , N. D. & Nieuwbeerta , P. ( 2010 ). When does the apple fall from the
tree? Static versus dynamic theories predicting intergenerational transmission of convictions . Journal of
Quantitative Criminology , 26 ( 3 ), 371–89 .
Van de Weijer , S. G. A. ( 2014 ). The intergenerational transmission of violent offending (Doctoral dissertation).
Alblasserdam : Haveka .
Van de Weijer , S. G. A. , Bijleveld , C. C. J. H. & Blokland , A. A. J. ( 2014 ). The intergenerational transmission
of violent offending . Journal of Family Violence , 29 ( 2 ), 109–18 .
Van de Weijer , S. G. A. , Besemer , S. , Blokland, A. A. J. & Bijleveld, C. C. J. H. (2015a ). The concentration of
sex offenses within British and Dutch families . In A. A. J. Blokland & P. Lussier (Eds.), Sex offenders:
A criminal career approach (pp. 321–48). Oxford : Wiley-Blackwell .
Van de Weijer , S. G. A. , Thornberry , T. P., Bijleveld , C. C. J. H. & Blokland , A. A. J. ( 2015b ). The effects of
parental divorce on the intergenerational transmission of crime . Societies , 5(1), 89–108.
West , D. J. ( 1969 ). Present conduct and future delinquency . London : Heinemann .
West , D. J. ( 1982 ). Delinquency: Its roots, careers and prospects . London : Heinemann .
West , D. J. & Farrington , D. P. ( 1973 ). Who becomes delinquent? London : Heinemann
.
West , D. J. & Farrington , D. P. ( 1977 ). The delinquent way of life . London : Heinemann .
Wolfgang , M. E. , Thornberry , T. P. & Figlio , R. M. ( 1987 ). From boy to man, from delinquency to crime . University
of Chicago Press .
Wright , R. A. & Miller , J. M. ( 1998 ). Taboo until today? The coverage of biological arguments in criminology
textbooks, 1961 to 1970 and 1987 to 1996 . Journal of Criminal Justice , 26 ( 1 ), 1–19 .

Downloaded By: Vrije Universiteit Amsterdam (VU Amsterdam) At: 06:21 03 Aug 2018; For: 9781315747996, chapter16, 10.4324/9781315747996.ch16
Taylor & Francis
Taylor & Francis Group
http://taylorandfrancis.com