This chapter aims to compare prospective (proximate) onset ages with retrospective (distant) and official onset ages for different offense types. The key question addressed here is how much retrospective and official onset ages agree with prospective ages, which are assumed to be the most accurate.

Age of onset, the age at which offending behavior begins, is a central concept in criminal career research. Some studies have explored the predictors of onset (Farrington and Hawkins, 1991; Farrington et al., 1990; Nagin and Farrington, 1992a), whereas others have focused on the impact of early onset on later features of criminal careers (see Chapter 5). Age of onset is one of the best predictors of the length and intensity of the criminal career (Blumstein et al., 1985; Farrington, 1973; Farrington and Hawkins, 1991; Farrington et al., 1990, 1998, 2003; Le Blanc and Fréchette, 1989; Loeber and Le Blanc, 1990). Research has also shown that a later onset can have a considerable favorable impact on the later criminal career (Farrington et al., 1990).

Farrington (1989) showed that the age-crime curve differed from one offense type to another. This suggests that the age of onset may vary from one offense category to another, although this aspect of criminal career research has been somewhat neglected (see Jolliffe et al., 2003; Le Blanc and Féchette, 1989).