It is widely understood that males are generally much more antisocial than females (Cook and Laub, 1998; Eme and Kavanaugh, 1995; Giordano and Cernkovich, 1997; Rutter, Giller, and Hagell, 1998; Steffensmeier and Allan, 1996). However, the implications of this sex difference for understanding the fundamental causes of antisocial behaviour have been virtually unexplored. We think that studying sex differences across the first decades of life offers an untapped resource for uncovering the causes of antisocial behaviour. Consider two fundamental facts about the distribution of antisocial behaviour across sex and age. It shows a male preponderance, and it shows a large increase in prevalence during adolescence. Other problem behaviours increase a lot during adolescence as does antisocial behaviour, but they show a female preponderance: depression and eating disorders, for example. Generally, social and psychological explanations have been put forward to explain the female preponderance, adolescent rise, and high prevalence of these emotional problems (Bebbington, 1996; Emslie, Hunt, and MacIntyre 1999). In contrast, some problem behaviours show a strong male preponderance as does antisocial behaviour, but they do not increase at all in adolescence, for example, attention-deficit hyperactivity, language delay, reading retardation, and autism. Generally, neuro-biological explanations have been put forward to explain the male preponderance, stability across age, and low prevalence of hyperactivity, dyslexia, and autism (Earls, 1987; Eme, 1979; Ounstead and Taylor, 1972).

Antisocial behaviour seems to be the sole anomaly in this otherwise orderly scheme.