Designed to target adult offenders, the three schemes under evaluation offered a range of services:

• CONNECT provided mediation (both direct and indirect) and conferencing.
• Justice Research Consortium (JRC) offered conferencing only.
• REMEDI offered both direct and indirect mediation.

For the reconviction analysis, the researchers developed a matched control group using information from criminal justice databases for the CONNECT and REMEDI schemes. JRC developed a control group by randomly assigning offenders to either conference or court after consent to participate in the restorative justice conference had been given.

The study considered the:

• frequency of convictions over the two-year period
• seriousness of convictions
• cost of offending

The researchers used the England and Wales standard definition of re-offending: re-conviction is “the extent to which an offender has been reconvicted (or received another official disposal, such as a caution, reprimand or final warning) during a period of two years for an offence committed since the sentence for the first offence (p. i).”

To determine the cost-effectiveness of restorative justice, the researchers compared the cost of offending (taking into consideration seriousness and frequency of re-offending, the cost to victim and the cost of processing the offences in the criminal justice system) to all the costs of operating the programmes (employees' wages, training, and administration).

Key findings included:

• When data from all three schemes are combined, the offenders going through the restorative justice schemes committed significantly fewer new offences than the control group.
• In terms of likelihood of reconviction there was not a statistically significant result although the restorative justice cases tended to go in the ‘right direction.’
• There was not a significant difference in severity of reconviction between the restorative justice and control group.
• The JRC groups showed a lower cost of conviction in comparison with the control group. However, the REMEDI and CONNECT groups did not show a significant difference.
• The individual trials presented too small a sample size to be expected to show statistically different changes on their own. The only exception was one of the JRC locations (Northumbria) that produced significantly fewer reconvictions than the control group.
• There was no evidence that restorative justice worked better from any one demographic group.
• For adult offenders, certain elements of the conference experience correlate with decreased re-offending. These include:
• • the conference made them realise the harm done
  • the offender wanted to meet the victim
  • the offender actively participated in the conference
  • the offender perceived the conference as useful
• “Value for money was calculated by looking at the cost saving (or benefit) by subtracting the cost of convicted offending in the two years after the restorative justice from the two years prior to the restorative justice, and then comparing restorative justice and control groups. On this measure, JRC produced a net benefit in terms of reconviction, whiles CONNECT and REMEDI produced a net cost. (p.vi)”

The full report is available online.

July 2008