Department of Justice is now the Department of Communities and Justice.  Find out more >

CJB141

Author Rohan Lulham and Clare Ringland
Published August 2010
Report Type Crime and Justice Bulletin No. 141
Subject Offenders; Recidivism / Re-offending; Statistical methods and modelling
Keywords criminal careers, offence seriousness, trajectory modelling

Download this publication

Summary

Aim

To investigate whether the crimes committed by offenders early in their criminal careers change in severity over consecutive offence episodes.

Method

Offence seriousness was measured across conviction episodes for a cohort of people born in NSW in 1994. We examined the relationship between offence seriousness and conviction episodes using two techniques. The first analysis involved stratifying offenders by total number of episodes and then using non-parametric tests to compare offence seriousness between and across episodes. The second analysis involved using group based trajectory modelling to investigate if there were groups of offenders who had different trajectories of offence seriousness over their first three conviction episodes.

Results

Across all offenders, non-parametric tests showed no consistent relationship between offence seriousness and conviction episode. In contrast, group based trajectory modelling provided evidence for four offence seriousness trajectory groups:

(1) an escalating group (16.1% of sample),
(2) a low stable group (32.7% of sample),
(3) a high stable group (26.0% of sample), and
(4) a de-escalating group (25.2% of sample).

Conclusion

Group based trajectory modelling provided evidence that the relationship between offence seriousness and conviction episodes varies for sub-groups of offenders. Potential implications for understanding and informing the prevention of serious crime are highlighted. Limitations of this study and challenges for future research are discussed.

Download this publication