The recidivism of prisoners released for the first time: reconsidering the effectiveness question

Abstract

This article describes the recidivism of 16,381 prisoners released from Western Australian Prisons between June 30 1975 and June 30 1987. Probabilities of ultimately returning to prison were estimated at 0.45, 0.76, 0.36 and 0.69 for male non-Aborigines, male Aborigines, female non-Aborigines and female Aborigines, respectively. A declining trend in recidivism over the period 1975–1980 was found to have levelled off from 1981–1985. These trends were related to important alterations in the definitions of the law, particularly to the decriminalisation of drunkenness and the introduction of mandatory prison for repeat offences of drunk driving. Recidivism also varied significantly with other demographic and institutional factors. Recidivism in this study was defined as a return to WA prisons for any offence, but varying definitions by deletion of fine defaulters made only minor differences to the estimated proportions returning.

The probabilities of sex offender re‐arrest

“…Although a 'race' descriptor is a poor guide to ethnicity, previous research has shown the importance of distinguishing between aborigines and nonaborigines 6 in studies of recidivism for Australian populations (Broadhurst and Maller, 1990;Broadhurst and Loh, 1995). Ta ble 2 shows very significant differences in p and lambda between aborigines and non-aborigines, with aborigines experiencing more rapid and much higher probabilities of re-arrest, irrespective of the definition of recidivism applied.…”

The Weight of the Sword of Damocles: A Reconviction Analysis of Suspended Sentences in Tasmania

“…There is some debate, however, over the value of comparing reconvictions of different sentencing dispositions as a measure of effectiveness. Broadhurst and Maller (1990), for example, considered it unclear whether differences in recidivism rates between custodial and noncustodial penalties are due to selection factors or the utility of the noncustodial program. Goldblatt (1998) in turn concluded that such studies have consistently indicated that any apparent differences between reoffending rates are largely the result of other factors, such as age and prior record, rather than the sentence itself.…”

Testing for Sufficient Follow-Up and Outliers in Survival Data

“…Here we consider two criminological examples of failure time data with an immune proportion (probably) present. They are drawn from a large data set described by Broadhurst and Mailer (1990) and con sist of two groups of prisoners convicted of serious sexual offenses, one with no prior records (No-Prior group) and the other with prior convictions (Prior group) and released, fol lowing imprisonment for the serious offense, after June 30, 1975. "Failure" for these individuals is "recidivism" (a return to prison for any offense), and we observe their times to return, possibly censored by the necessity to cease observation at a predetermined time (June 30, 1987).…”

“…The data are listed in Appendix B. Broadhurst and Mailer (1990) and Mailer (1993) suggested that the data are fairly well described by an improper WeibuU failure distribution with a proportion 1 -p of immune in dividuals who will not return to prison. Analysis of this data according to Mailer (1993) shows significant differences be tween Prior and No-Prior groups in probability of recidivism and in rate of return to prison, but not in respect to the shape parameters of the fitted WeibuU distributions.…”