How long do criminal careers last? Is their duration short (a few years), medium (several years), or long (many years)? Do some identifiable offenders have longer careers than other offenders? Are careers in one crime type longer than those in another crime type? Can estimates of career length be obtained given the truncation observed in many longitudinal data sets? What about residual career length, or the expected time remaining in an offender's career at the time of a particular intervention (such as incarceration)? Can we develop estimates of an offender's expected time left in a career?

Career length questions are central to both theory and policy, but the policy questions are particularly salient. For example, if offenders have short careers, then criminal justice policies such as three-strikes laws, with their lengthy incarceration terms, may waste scarce resources. This will be the case because offenders' careers are likely to be over well before the end of their lengthy incarceration. Thus, knowledge of career lengths, particularly residual career lengths, can provide important information for policy guidance. Additionally, information on the correlates associated with both short and long careers will be of use in attempts to modify criminal careers. For example, a strong relationship between legitimate employment and termination of criminal careers may suggest greater attention to employment facilitation as a useful policy intervention (Blumstein et al., 1986:85).

Research has examined the career length issue in one of three ways.