Intergenerational continuity of crime: A comparison between children of discordant siblings.

Background: Various studies have shown that fathers and children show similarities in criminal behaviour, but little is known about the nature of this relationship. By using a family-based research design, this study controls for familial confounders and gives a better estimate of the extent to which paternal crime has a direct effect on offspring offending.

Aims: To test the extent of any relationship between paternal offending during the childhood of offspring and adolescent offending by those offspring and to examine the effect of potential confounders of this relationship.

Methods: Data were from records held by Statistics Netherlands for 1,155,771 individuals born in the Netherlands between 1996 and 2001. Police data were used to measure paternal offending during the childhood (age 0-11) of this cohort and their adolescent offending (age 12-18). Logistic regression analyses were used to estimate the bivariate relationship between paternal and offspring offending, as well as this relationship after controlling for various demographic and socio-economic variables. Conditional logistic regression analyses were used to compare children of discordant brothers (N = 9232). By comparing within families rather than between unrelated individuals, all unmeasured familial factors that are shared between these cousins were controlled for.

Results: Offending during adolescence was about three times as likely among offspring whose fathers had offended while they were 11 years old or younger than among adolescents with no such paternal problem (OR: 3.21, CI 3.17-3.26). This relationship was attenuated after controlling for measured confounders (OR: 1.78, CI 1.75-1.81) and for unmeasured familial confounders (OR: 1.47, CI 1.36-1.59), but remained significant.

Conclusions: Paternal offending has an association with offspring offending, but this is small after controlling for measured socio-economic and unmeasured familial confounders. Previous studies that did not control for unmeasured familial confounders seem likely to have overestimated the effect of paternal crime on their offspring's offending. This has implications for interventions for the offspring. If confined to mitigating the negative consequences of paternal offending, they are likely to have limited effectiveness.