The impact of timing and frequency of parental criminal behaviour and risk factors on offspring offending

Abstract

This paper explores mechanisms of intergenerational transmission of criminal behaviour by investigating specifically the timing and frequency of the parents' criminal behaviour while including risk factors for criminal behaviour. The results demonstrate a dose–response relationship: parents' number of criminal convictions is positively related to offspring's conviction rate. Furthermore, children whose parents had only been convicted before the child's birth have more convictions than those whose parents had never been convicted. Children whose parents had been convicted after the child's birth have more convictions than those whose parents had only been convicted before the child's birth, but this difference can be explained partly by the observation that the latter group had fewer risk factors for crime. When parental convictions at different ages were examined, children whose parents had been convicted between their 7th and 13th birthdays exhibit more criminal behaviour than children whose parents were convicted in other periods, but none of the differences were significant. There does not appear to be a sensitive period for the impact of parental criminal behaviour. The results demonstrate support for static as well as dynamic explanations of intergenerational transmission such as the transmission of a criminogenic environment and/or mediation through risk factors.

Keywords:

• intergenerational transmission
• mechanisms
• parental criminality
• developmental criminology
• risk factors

Acknowledgements

I am greatly indebted to David Farrington and Donald West for the data collection of the Cambridge Study in Delinquent Development. Furthermore, I would like to thank Catrien Bijleveld, David Farrington, and the IoC Writing Group as well as the anonymous PCL reviewers and the editor for their helpful comments on drafts of this paper. This work was supported by the Gates Cambridge Trust. Data collection for the Cambridge Study in Delinquent Development was funded by the UK Home Office.

Notes

1. Assortative mating is defined as ‘the tendency for people to form unions with similar others' (Moffitt et al., 2001, p. 185). People tend to affiliate with those who are similar to themselves, and antisocial people tend to marry and have children with other antisocial people. Children with two criminal parents have an increased risk to show antisocial behaviour, they experience a ‘double whammy’ effect (Moffitt et al., 2001, p. 195; West & Farrington, 1977). They inherit an antisocial phenotype twice and grow up in a criminogenic home environment. However, I argue that assortative mating is not a mechanism in itself, because other explanations such as genetic, social learning or risk factor mechanisms are needed to explain transmission in the first place. Assortative mating only posits that the intergenerational transmission is stronger with two than with one criminal parent.

2. Analyses with mutually exclusive groups led to similar conclusions. Furthermore, analyses where groups were combined (0–6 and 7–12, 7–12 and 13–18, 0–6 and 13–18) and compared with each other and with the groups studied in this article did not lead to significant results. The age cut-off points were chosen to create groups with a similar age interval. Furthermore, 0–6 years of age represents a pre-school age, 7–12 represents primary school, while 13–18 year olds are likely to be in secondary school, further education or working.

3. Log natural transformation of a count variable with a skewed distribution is recommended (Tabachnick & Fidell, 2007) and used widely in criminological research (e.g. Krahn, Hartnagel, & Gartrell, 1986; Messner, 1989; Neapolitan, 1994, 1995, 1997). The following formula was used for the log natural transformation: Log(x) = Ln(x+1).