The notion of specialization in offending has had a long history in criminology, and researchers have devoted considerable attention to it. Research questions have included: Do some offenders disproportionately commit some particular types of crime, such as violence, during their offending careers? Are patterns of specialization, if they do exist, concentrated in the juvenile years, or do offenders become more specialized at older ages (Cohen, 1986; Piquero et al., 1999). Or is it the case that offenders tend to engage in a diversity of criminal acts, be they violent, nonviolent, and/or drug related and that such diverse patterns of offending continue throughout one's criminal career?

Extant criminological theory also sees the specialization issue as important. For example, some theories, such as Gottfredson and Hirschi's general theory of crime, anticipate very little specialized offending in the career, while other theories, such as Moffitt's developmental taxonomy, hypothesize some level of specialization for one type of offender (i.e., adolescence-limited) but more generalist patterns for another type (i.e., life-course-persister). The issue of specialization is also important for policy reasons (Farrington et al., 1988). For example, if offending were more specialized, then knowledge about earlier types of offenses in one's career would help officials to predict later offense types and help criminal and/or juvenile justice/adult decision-making efforts.

Much research has been brought to bear on the issue of specialization/versatility. Three findings stand out from this line of research.