Background: This article is dedicated to David Farrington who was a giant in criminology and, in particular, a pioneer in studying developmental pathways of delinquent and antisocial behaviour. Numerous studies followed his work. Systematic reviews of his and others' research described between two and seven (mainly 3-5) trajectories. The variation is due to the age of individuals, kind and seriousness of problem behaviour, data sources, assessment methods and cultural context. Reviews stated a lack of research on very early starting problem behaviour, broad developmental outcomes, data from multiple informants and (beyond description) on risk and protective factors or potential causes of the different trajectories.
Aims: The present study addresses these issues in a prospective longitudinal design and test of the concept of antisocial potential (AP) in Farrington's ICAP theory.
Methods: Data on more than 600 children and their families were gathered in a prospective longitudinal design over 10 years in Germany. To avoid potentially negative effects of criminal justice interventions, the study concentrates on child development from ages 4-5 to 6-7, 8-9, 10-12 and 13-14 years. Child externalising behaviour problems were assessed using the social behaviour questionnaire by kindergarten educators, mothers, school teachers and youth self-reports. Developmental trajectories were analysed by general growth curve modelling (GGCM) across five time points. The prediction and explanation of different pathways included family factors (SES, stressful life events, aggressive and inconsistent parenting) and child characteristics (intelligence, resting heart rate, disruptive behaviour, temperament and social adaptability). In accordance with dose-response relationships, we also tested accumulated factors in the Cracow Risk/Needs Instrument.
Results: The GGCM analysis revealed five developmental trajectories: high-chronics (2.4%), high-reducers (7.9%), medium-reducers (22.4%), late-starters medium (8.7%) and low-chronics (continuously unproblematic youngsters; 58.6%). Although the group with high externalising problems across all time points was rather small due to the affluent context of the region, there were significant social and individual differences between this and the other groups that fitted to ICAP theory. Furthermore, the study revealed differences between those youngsters that desisted from behaviour problems or started later. The predictive validity of accumulated factors in the Cracow Risk/Needs Instrument was very good for the comparison of the groups with persistently high versus no serious behaviour problems.
Conclusions: Our results showed that different pathways of aggressive, delinquent, impulsive and other externalising behaviour already commence in early childhood. Behavioural stability (high- vs. low-chronic problems) was well predicted by child and family characteristics, but there were also plausible findings on trajectories of behavioural change. Overall, the findings underline the need for early developmental prevention.
© 2025 The Author(s). Criminal Behaviour and Mental Health published by John Wiley & Sons Ltd.