Self-Reported Offending, Victimization, and the British Crime Survey
Mayhew, Pat, Elliott, David, Violence and Victims
Both the 1982 and 1984 British Crime Surveys contained questions on self-reported offending. In 1984 the questions were changed in order to try and improve admission rates. This article looks at the effect of this change. Admission rates in 1984 are shown to be approximately double those in 1982. Also examined are differences between offenders and non-offenders, what offenses 1984 respondents saw other people as engaging in most often, and who saw other people as least law abiding. The relationship between self-reported offending and victimization in both this and Gottfredson's (1984) analysis of BCS data is taken up.
The main aim of the British Crime Survey (BCS), conducted in 1982,1984, and 1988, has been to provide information on the nature and extent of crime, and on patterns of risk. In each sweep, one person aged 16 or over was interviewed in about 11,000 households in England and Wales.^sup 1^ In common with other national victimization surveys, the BCS has had substantial impact on improving understanding of crime measurement, particularly as regards the dark figure, of crime and the pitfalls of relying on police statistics as an indicator of crime trends (Mayhew & Hough, 1985). Like other victimization surveys too, the BCS has been important in assessing the level and nature of the victimization risks that different groups face.
However, the BCS has differed from many other national surveys in the range of additional crime-related topics it has covered- these varying from sweep to sweep. This has contributed to a better understanding of, for example, fear of crime (Box, Hale, & Andrews, 1988; Maxfield, 1988); the effects of victimization (Maguire & Corbett, 1986); the nature of interactions between the police and public (Southgate & Ekblom, 1984); attitudes toward sentencing (Hough & Moxon, 1985), and crime seriousness (Pease, 1988). In particular, though, the BCS (in the 1982 sweep especially) included direct measures of "lifestyle" and risk factors which have generated analysis of, for instance, elderly victimization, controlling for exposure (Clarke, Ekblom, Hough, & Mayhew, 1985); lifestyle and household victimization (Maxfield, 1987) and lifestyle and predatory victimization (e.g., Sampson & Wooldredge, 1987). The BCS has also been unusual in incorporating measures of self-reported offending.
The article does two main things. First, it compares the results from the 1982 and 1984 surveys, given that in 1984 a new question design was introduced to try and improve what seemed to be rather low rates of admission in the 1982 survey. Secondly, some findings are presented about the characteristics of those who admitted offenses, and about respondents' perceptions of how far other people keep within the law.
Self-reported offending questions were included in the BCS not only to try and establish lower-limit estimates of offending among an adult population, but also to investigate the association between being a victim of crime and a perpetrator of it.^sup 2^ This association has been given indirect support insofar as victims and offenders have been shown to come from groups with the same demographic and social characteristics, signifying youth, poverty, ethnic minority concentration, and urbanization (e.g., Gottfredson, 1981). It has been evidenced more directly in studies showing that those who report more transgressions of the law also report most victimizations (Pagan, Piper, & Cheng, 1987; Lang, Baker, & Ball, 1969; Sparks, Genn, & Dodd, 1977; Singer, 1981; van Dijk & Steinmetz, 1983; Wolfgang, 1958). Singer, for instance, found victimization experience to be the best predictor of commission of a serious assault. Pagan et al. showed that inner-city students who had been victimized were consistently more likely to have engaged in delinquent behavior, though they found the direction of the relationship to be uncertain and argued instead that the social processes which cause someone to offend and become a victim may well be different. …