Key points are not available for this paper at this time.
The extent and nature of contextual effects on juvenile offending are frequent subjects of current research, mainly in the USA. After a short literature review, this paper presents empirical results of a new study which hints at the existence of neighbourhood contextual effects on serious offending by adolescents. The study is based on three types of cross-sectional data on 61 neighbourhoods in two German cities and a rural area: a self-report survey of students aged about 13 to 16, a separate survey of residents in the survey neighbourhoods, and census and administrative data on the same neighbourhoods. Multilevel analysis is applied to identify and explain the neighbourhood-level variance of selfreported serious juvenile offending. Hypotheses from both main traditions of theoretical reasoning about contextual effects on juvenile delinquency - subcultural and disorganization theories - are supported by the empirical findings. The spatial concentration of adolescents with attitudes typical of delinquent subcultures increases the likelihood of serious offending net of relevant individual predictors, whereas the social capital of neighbourhoods (as measured by the independent survey of residents) reduces it. Methodological problems of applying multilevel analysis to delinquency research are discussed.
Dietrich Oberwittler (Thu,) studied this question.