Abstract
The continuing gap between the percentages of black and white people entering the criminal justice system (particularly in prisons), observed both in the USA and the UK, raises an old but very important question of whether the over-representation is the result of proportionately more black than white people committing offences ('differential involvement') or whether it is the result of racial bias in the administration of criminal justice ('racial discrimination'). The data provided in this study lends some support to the latter view and suggests that the over-representation and differential treatment by racial origin of the defendants at each stage of the criminal justice process cannot be explained by the differential involvement hypothesis alone. The research involved a multivariate analysis of all male referrals and outcomes during a five year period from the inception of the Criminal Justice Act 1982, in a London Metropolitan Police District's local juvenile criminal justice system which comprises the largest numbers of ethnic minorities in the United Kingdom. The study found that it was not possible to attribute outcomes on the basis of significantly more young people from one of the three racial groups of defendants being associated, for example with committing certain types of offences when compared to the other two groups. In other words, there was no significant interaction found between race and each of the control variables of interest, such as the type of the primary offences, the type of offences taken into consideration at the same hearing as the primary offences, the total number of offences dealt with at the same hearing and the number of arrests or court appearances in any model predicting court sentences as against police diversionary measures. However, evidence is provided which suggests that black young persons were significantly more likely than white young persons to have been acquitted in the courts because of insufficient evidence from the police rather than to have received police diversionary measures, notwithstanding the influence of the control variables of interest considered in combination with each other. A similar result was also found in respect of custodial sentences, where black young persons were again significantly more likely to have received custodial sentences compared to their white counterparts, the latter proving significantly more likely to have had no further action taken against them by the police or to have been cautioned. There were differences between the probabilities of acquittals and of custodial sentences between black and Asian young persons in favour of black young persons, but these were not significant. Likewise, the difference in the probabilities between Asian and white young persons in the custodial sentences analysis was not significant, although it was in favour of Asian young persons.