ABSTRACT
This chapter analyses the claim that the causes of participation in some form of illegal activity are somehow different from the causes of the frequency of that illegal behavior once it occurs. Using data from the Bail Decisionmaking Study (BDS), Seattle Youth Study (SYS), and National Youth Survey (NYS), multivariate models of participation and frequency of offending were tested with probit and Tobit statistical models. The BDS sample of released offenders in Philadelphia seriously questions the validity of the criminal career claim that the causes of participation and frequency of offending are substantially different. Similar to the BDS analysis, demographic characteristics—age, race, and gender—are included to model the different mean levels of delinquency among the different groups. There is a great deal of similarity in both the SYS and NYS analyses comparing participation with frequency of delinquency, when the cutpoint is operationalized as one or more delinquent acts.
