Statistical models for adult offender populations

Criminology Research Council grant ; (23/90)

A large population of all persons arrested by police in Western Australia for the first time between 1 April 1984 and 30 June 30 1993 were followed up to determine if they had ever been re-arrested. Probabilities of re-arrest were estimated at 0.52 for male non-Aboriginal people, 0.36 for female non-Aboriginal people, 0.88 for male Aboriginal people and 0.85 for female Aboriginal people. Significant variations in the probability of re-arrest and/or the time to re-arrest for different age groups, the number of times arrested, occupational status, offence group, place of birth and bail status were observed. Co-variate analysis of non-Aboriginal people found with the offence of "driving under the influence" (DUI) was also undertaken to test the significance of difference in probabilities of re-arrest for sub-groups. Probabilities of re-arrest were: for males 0.47 for any offence and 0.31 for a repeat DUI; and for females 0.34 for any offence and 0.20 for a repeat DUI. The results are discussed in the light of estimates of re-imprisonment and the utility of offender risk assessment.