NDARC Technical Report No. 264 (2007)


Aim: The aim of this study was to assess whether there was evidence of cost-savings due to crimes averted while individuals were engaged in methadone maintenance as a treatment for heroin use. Analyses were undertaken over a four year period for a population-based sample of NSW methadone clients.

Methods: Linked administrative data bases (methadone treatment records, court appearance records, and imprisonment records) were used to estimate the costs of crime on- and off-treatment, the costs of gaol and the costs of treatment for a sample of 10,925 NSW methadone clients.

Certain crimes were excluded due to uncertainty around the completeness of the data (e.g. homicide); lack of cost data (e.g. illicit drug use crimes, dangerous or negligent acts endangering a person, abduction and related offences); and potential for double counting costs (illicit drug use). Following data linkage, costs were applied to treatment days, charge and gaol data in order to estimate the economic costs and potential cost savings. Costs were attached to days in methadone treatment in and out of gaol (Warren et al., In press, Mattick et al., 2001); days in gaol (Department of Corrective Services, 2005); and to the various types of crime (Mayhew, 2003). All costs are reported in 2005 Australian dollars.

There are two sets of analyses in this paper. The first is descriptive and presents the total number of treatment days and crimes committed on- and off-methadone, and the costs of treatment, gaol and crimes for the sample over the four year study period. A second analysis was conducted using regression analysis to explore whether methadone treatment impacts upon the costs of crime and gaol.

Days and charges

  • The sample spent more free time on-methadone (53%) than off-methadone (47%).
  • The large preponderance of time was spent out of gaol (92% of the total days).
  • There was considerable variation across the sample with regards to the amount of time spent on-methadone with some individuals spending only one day in treatment during the four year study period while others spent the entire 1,461 days enrolled in treatment.
  • With respect to crime, 46% of all crime was committed while on treatment and 54% while off treatment; which translates into 0.0026 crimes per free day on-methadone and 0.0034 crimes per free day off-methadone.



  • The largest component of total costs was the cost of gaol, which accounted for 65% of the total despite only accounting for 8% of the days.
  • Methadone treatment costs, both in and out of gaol, accounted for 26% of total costs, and the crime costs included accounted for 9% of total costs.
  • Crimes committed on-methadone accounted for 43% and crimes committed off-methadone account for 57% of total crime costs.


Regression analysis
The results of the regression analysis suggest that the mean cost of treatment, crime and incarceration was $17,274 per episode. The coefficient for days in treatment was negatively related to cost of crime and gaol and highly significant, thus indicating that for each additional day in treatment the costs of crime and gaol decreases by $15. Gender was also significant: the cost per female was, all else constant, $5,129 less than the cost per male.

Discussion: The present study found a reduction in the cost of crime associated with enrolment in methadone consistent with the broader literature. When comparing only the treatment and crime costs, it was apparent that the investment in methadone treatment was only partially offset by savings from averted crime. However, the results from regression analysis, which examined the relationship between time in methadone treatment and costs of crime and gaol, found that every day an individual was enrolled in methadone treatment paid for itself in terms of a decrease in gaol and crime costs.

Citation: Shanahan, M., Hetherington, K., Mattick, R. P. and Weatherburn, D. (2007), Estimating the cost-savings of reduced crime while in methadone treatment, Sydney: National Drug and Alcohol Research Centre.


Date Commenced
13 Feb 2007
Resource Type
Technical Reports