Interpreting Age, Period and Cohort Effects in Juvenile Delinquency: Predicting County Level Offending

Anthony Fabio, University of Pittsburgh
Rolf Loeber, University of Pittsburgh
David P. Farrington, University of Cambridge

The Pittsburgh Youth Study is a longitudinal study of three cohorts of Pittsburgh children followed from 1987 to 2001. The database tracks changes in juvenile crime trends as well as risk and protective factors. These data provide a valuable opportunity to develop a predictive model for offending based on a large amount of individual data of separate cohorst collected over many years. We plan to use these data to develop a predictive model of juvenile offending utilizing as predictors risk factors publicly available data at the county level and data specific to the PYS. This approach allows us to test whether Pittsburgh juveniles' offending behavior is best explained by trends in the prevalence of risk factors (period effects), changes in children's behavioral responses to risk factors (cohort effects), or both. we can also test the utility of this model in predicting county level offending. This will provide important information to aid in public policy decisions on whether currently available data is sufficient to predict offending. We can also improve the power to predict turning points in offending by improving estimates the lags between the emergence of risk factors and precursor behaviors as well as offending behavior.

(Return to Program Resources)

Updated 05/20/2006