Skip to content. | Skip to navigation

Personal tools


Restorative Practices and Reoffending

Recently, a short article in The Report (1) questioned Canada's use and support of restorative justice programmes. The article quoted a claim in the May issue of Canadian Lawyer that after five years of use there was no proof that restorative justice programmes work. However, recent research has demonstrated that restorative justice programs do in fact reduce recidivism.
The chart below highlights seven such studies. Five involved comparison groups of offenders going through normal court processes, and revealed a decrease in recidivism for offenders who go through restorative programs. Two studies identified elements in conferencing that appear to have an impact on reoffending. Both studies found that when those elements are met, conferencing lowers recidivism when compared to anticipated recidivism using established predictors of offending behavior.







Luke, Garth and Bronwyn Lind. 2002. B69-Reducing Juvenile Crime: Conferencing Versus Court.”

New South Wales; Australia

This study compares reoffending rates of young people who went to a conference with reoffending rates of young people who attended court. The time period of follow-up is 27-39 months.  The study sample consisted of 590 young people who went to conference in 1998; 5,516 young people who appeared in court the year before conferencing became an option (1997); and 3,830 who appeared to court in 1998.

The results indicated that conferencing produces a moderate reduction of up to 15 to 20 percent in reoffending across different offence types and regardless of the gender, criminal history, age and aboriginality of offenders.

Hayes, Hennessey and Kathleen Daly. 2001. “Youth Justice Conferencing and Reoffending.”

South Australia

This study focuses on a sample of 89 conferences. It draws from conference observations and official police data to explore the relative importance of conference dynamics and offender characteristics in predicting future offending. The reoffending data comes from an 8-12 month follow-up period. There was no comparison with a control group in non-restorative programs.


The post-conference results showed that :  60% of sample had no official contact with police

17% had one contact

23% had two or more contacts 

Daly and Hayes identified the following conference conditions as having the greatest impact on reoffending:

-- Remorse shown by the offender (reoffending is 1/3 as likely)

-- Consensual decision-making (reoffending is ¼ as likely).

Latimer, Jeff, Craig Dowden, and Danielle Muise. 2001. “The Effectiveness of Restorative Justice Practices: A Meta-Analysis.

Mainly North America

This meta-analysis took a sample of 35 studies that looked into recidivisms rates, victim and offender satisfaction, and restitution completion. Each studies used control groups to measure the outcomes.

The 32 studies that covered recidivism showed a mean decrease of 7%.

Maxwell, Gabrielle, and Allison Morris. 2001. “Family Group Conferences and Reoffending.”

New Zealand

Maxwell and Morris conducted a 6.5 year follow-up of 108 offenders who had gone to conference. The study attempts to identify the characteristics of conferences that are more likely to be associated with less reoffending. 

Using an evidence-based approach, the researchers  attempt to identify the characteristics of people who do not reoffend. From past  research that identifies  circumstances that lead to offending and reoffending, the researchers developed a model to predict reoffending based on

-- Early life experiences (deficits in the family's circumstances and the child's environment) 

-- Early negative experiences (experiencing bullying, violence, and abuse).  

This model for understanding reoffending was then used to determine if factors related to conferencing impacted on future behavior.

The researchers identified 5 reconviction categories and self-reporting of offenses to measure recidivism.

--Persistent reconvicted-characterized by the frequency and volume of their offending in criminal matters

-- Improving reconvicted- had offended persistently for a time but had not been reconvicted in the 12 months prior to the interviews

-- Occasional reconvicted- had appeared in court more than once but had committed less than 5 offenses

-- Once only reconvicted- had appeared in court only once

-- Not reconvicted.

The study revealed the following percentages for each reconviction group:

29%  not reconvicted

14% reconvicted only once

21% occasional reconvicted

8% Improving Reconvicted

28% Persistent Reconvicted

The key finding was that family group conferencing can contribute to lessening the chance of reoffending even when other important factors such as adverse early experiences, other events which may be more related to chance, and subsequent life events are taken into account.

Sherman, et. al. 2000. Recidivism Patterns in the Canberra Reintegrative Shaming Experiment.

Canberra Australia

The study uses random assignment for treatment (court or conferencing) to investigate the effect of diversionary conferencing on different offense types. The majority of the cases studied had a one-year follow-up period.

The study looked at 3 offense types:

-- violent offences (with a sample size of 110 offenders)

-- drink-driving (with a sample size of 900 offenders)

-- juvenile property crimes (with a sample size of 117 offenders).

The study found that, when compared to court, the effect of diversionary conferences is

-- a 38% decrease for young violent offenders

-- a 6% increase for drunk driving

-- No difference in property offenses or shoplifting.

Griffiths, Mark. 1999. The Implementation of Juvenile Justice in Victoria.

Victoria Australia

Sample size of 71

12 month follow-up period

Matched probation group

37% of the Control group received further sentencing orders

21% of the Conferencing group received further sentencing.

Bonta, James. Jennifer Rooney, Suzanne Wallace-Capretta. 1998. “Restorative Justice: An Evaluation of the Restorative Resolution Project.”

 Winnipeg, Canada

The program targeted offenders who were likely to go to prison (with at least a 9 month sentence). Once an offender was accepted into the program, the Restorative Resolutions (RR)  staff started working on a plan which included attempts to contact the victim and the community to help in creating a plan.

The study identified matched groups of offenders who either went to prison or were given probation. 

Since many of the study participants were still in the program, recidivism was defined as new crimes or as breaking the terms of probation.

At two years, there was a significant difference between the RR participants and  (11.5% vs. 33.3%).

When compared to the two groups of probationers, the RR participants showed significantly lower recidivism (14.1% vs. 56.3%).


1. "Maybe 'punishment' works better than 'healing'". The Report. June 24, 2002. p.37

By Lynette Parker

July 2002


Document Actions

Restorative Justice Online - Featured Video

Restorative Justice Library Search

Search 11427 publications on restorative justice