The theory of reintegrative shaming was developed not only to generate new predictions and new policy implications about crime, but also to explain as adequately as possible what are the well-established strong relationships between crime and other variables. Below are a list of what I believed to be, in developing the theory, the strongest and most consistently supported associations in empirical criminology, bearing in mind that we are concerned with a general theory and not with propositions relevant only to specific types of crime. Any credible theory would at least have to be consistent with these findings, and preferably would offer an explanation for most of them.

Crime is committed disproportionately by males.

In most, if not all, societies men constitute over 90 per cent of adult prison populations. Arrest and court data also consistently show in all countries (e.g. Simon and Sharma, 1979) a massive overrepresentation of men in criminal statistics, while self-report measures tend to show much more modest gender differences in offending rates (e.g. Braithwaite, 1977: 26; Smith and Visher, 1980; Canter, 1982; Warner, 1982; Ouston, 1984; McGarrell and Flanagan, 1985:340–1; Morash, 1986; Riley, 1986). Smith and Visher's (1980) review of forty-four studies suggests that the more serious the type of offense, the greater the gender differences in rates of offending.