NCJ Number
175731
Journal
Studies on Crime and Crime Prevention Volume: 6 Issue: 2 Dated: 1997 Pages: 169-199
Date Published
1997
Length
31 pages
Annotation
This study examined the relationship between age and crime for different offense types and assessed the implications of this relationship for the utility of California's "Three Strikes" law.
Abstract
Data for the analysis were drawn from the first four waves of the National Youth Survey (Elliott et al., 1985, 1989). This survey used a national probability sample of households in the continental United States and identified 2,360 eligible youths aged 11-17 at Wave 1. Of the eligible respondents, 1,725 participated in the study and were interviewed at Wave 1 in 1976. The respondents in the current analysis ranged from ages 11 to 17 at Wave 1 and from 14 to 20 years old at Wave 4. The study focused on reported delinquent participation during Waves 2 through 4 for those respondents who did not report delinquent activity during Wave 1. This approach maintained the appropriate temporal ordering between predictor variables and subsequent delinquent participation. Age 13 was selected for differentiating between adolescents who participated in delinquent activity at an early age compared to adolescents who participated in delinquency later. Using a composite measure for trivial acts of delinquency, the study compared age-specific logistic regression models to a model of complete invariance. The results support the null hypothesis: the effects of various predictors of delinquency participation for trivial acts of delinquency were apparently invariant across different age groups. Additional analyses compared age-specific models with a model of complete invariance for a composite measure for serious acts of delinquency. These results support rejection of the null hypothesis; that is, results suggest that the effects of certain predictors of serious delinquency vary across early and late participation groups. Overall, the results suggest that unitary theories that advocate age invariant explanations of delinquency are not necessarily appropriate for understanding serious delinquent conduct. The additional complexity of developmental models is apparently empirically justifiable. The results suggest that the understanding of variation in the predictors of delinquent participation may be a function of how researchers choose to define delinquency. Future research should continue to explore such variation across social groups, social settings, and varying definitions of crime and delinquency. 2 tables and 46 references