Australian Institute of Criminology

Skip to content

Restorative justice and deterring crime

Lawrence W Sherman and Heather Strang
ISSN 1328-3006 ; ISBN 0 7315 2803 4
Australian National University, Canberra
RISE Working Papers, no. 4
April 1997

The Canberra reintegrative shaming experiments indicate that offenders are more deterred from repeat offending after experiencing the restorative justice approach of diversionary conferencing than after court proceedings. Drink drivers predict more dire consequences especially with their families, if they are caught offending again after a conference than after court. Both young offenders and drink drivers are more likely than offenders sent to court to say they will obey the law in the future.

These preliminary findings of the evaluation by ANU researchers of the Canberra police conferencing program are especially important because some critics have called shaming conferences a "soft option". Interviews with the first 548 offenders in the study suggest that is not how the offenders see it, especially after experiencing over an hour under the spotlight of critical examination by family, friends, victims or community representatives sitting in a circle in a private room at a police station. Led by specially trained police officers, the conferences are emotionally intense discussions of what the offender did, whether the offender is sorry, and how the offender can repair the harm caused by the crime. For drink drivers, the average court case in the study takes 6 minutes, while the average conference takes 88 minutes. For young offenders, the average court case takes 13 minutes, while the average conference takes 71 minutes.

Whether these conferences are "hard" or "soft" is not nearly as important as whether they prevent repeat offending better than prosecution in court. That is the ultimate question of the long-term ANU evaluation of the program, requiring two years of follow-up monitoring of the offences committed by the offenders in the study. But the battle to prevent future crime is won or lost in the immediate aftermath of the court or conference, which is the time when ANU staff seek interviews from all the offenders. The first 111 young offenders and 437 drink drivers interviewed represent around three-quarters of all those eligible for interview. The young offenders were apprehended for a wide range of property and violent offences, from shoplifting to burglary to assault.

In order to draw conclusions about the comparative effects of court and conferences, the ANU evaluation used the most rigorous method available: the randomised clinical trial approach used in medicine, education and other fields. This approach virtually rules out other explanations for any differences between the two groups other than the treatment they received. Once a police officer has determined that a case would be suitable for either prosecution in court or diversion to a reintegrative shaming conference, the ANU research team advises the officer on which course to take. That advice is blind to the characteristics of the offender, and based solely on a mathematical formula that gives all eligible offenders an equal chance to go to court or conference.

The results of the interviews show what offenders dealt with in court think of their sentences, and how they view the prospects of future punishment and their own obedience to the law.

How Offenders Judge Their Sentences

The effect of criminal justice actions on repeat offending is arguably a result of two separate factors, called procedures and outcomes. Jeremy Bentham's classical approach to criminology placed the emphasis on outcomes, assuming that offenders are deterred by the fear of greater infliction of pain from the state in the form of anything from a fine to capital punishment. French sociologist Emile Durkheim modified that view by suggesting that "it is shame which doubles most punishments", and that the real power of formal punishment is to invoke informal disgrace among family and friends.

More recently, social scientists of law have argued that the fairness of the procedures may matter as much as or more than the actual punishment imposed. Listening to the offender's views, correcting errors of fact, and judging each offender by equal standards regardless of race, income, age or gender are all important elements affecting whether offenders respect the law. That in turn may affect their attitude to obeying it. In the report in this series entitled "Restorative Justice and Offenders' Respect for the Law", conferences were shown to be superior to court on all these elements of "procedural justice", which results in greater offender respect for the police, the justice system and the law itself.

The reason these findings of a deterrent effect are surprising is that they go beyond the theory of reintegrative shaming devised by ANU Professor John Braithwaite. It holds that offenders will be more likely to obey the law if shamed in a restorative way because that will make them want to be law-abiding. The theory does not suggest that shaming will make offenders more fearful of sanctions, thus coercing them into obeying the law. The theory places far more emphasis on the procedures of shaming than on the outcome, which is traditionally associated with the fear of punishment.

Procedures and outcomes join together in how offenders judge their sentences. In the case of drink driving, preliminary results are now available on how the outcomes differed between conferences and prosecution in court. Conferences are much less severe in terms of monetary fines and loss of drivers licences, but much more severe in terms of community service. The average "fine" (payable to a community charity) in a drink driving conference is $120, while the average fine imposed by the courts is $414. Whether these fines are ultimately paid is a separate question which is also being tracked. Because the police put special efforts into enforcing offenders' compliance with the outcomes agreed to in conferences, it is possible that the difference in paid fines will be less than the difference in fines imposed.

The average number of hours of community service imposed on drink drivers in conferences is 26 hours compared with two hours in court. Many criminologists claim that hours are fairer than dollars in imposing any sort of fine, because it levels out the effect of income differences. Whatever your income may be, a fine of one hour of your time has just as much proportionate effect on that income.

Court and conference are also different in their restrictions on driving privileges. Only in court can offenders have their driver's licences suspended or cancelled, although in conferences offenders sometimes agree voluntarily to stop driving at specified times or for a specified period. Whether formal loss of licence actually prevents drink driving again is another question being tracked and will be measured in both offender interviews and police records two years after the initial arrest of each offender.

This theory receives strong support in the ANU evaluation of drink driving cases. Offenders sent to conferences are more likely to say the outcome they received was fair (72 per cent) than offenders who were prosecuted in court (54 per cent). Drink drivers sent to court were also twice as likely to believe their sentences were too harsh (32 per cent) as offenders sent to shaming conferences (16 per cent). This is accompanied by drink drivers feeling bitter and angry far more often after court (26 per cent) than after conferences (7 per cent).

The Sword of Damocles

The three elements of classical deterrence theory are the certainty, severity and swiftness of punishment. The offender interview data on the first two of those elements suggest that conferences deter drink driving offenders more than prosecution in court. This is true even though all offenders fear future court more than future conferences. A conference creates a "sword of Damocles" effect in which prosecution hangs over the drink driver's head, whereas prosecution itself takes away some of the fear of future prosecution.

Getting Caught Again. Both young offenders and drink drivers who were sent to conferences are more likely than prosecuted offenders to think they will be caught if they offend again. The difference is not large: 29 per cent for drink drivers sent to conference versus 21 per cent for those sent to court; for young offenders it was 71 per cent of those sent to conference compared with 53 per cent of those sent to court. But it suggests how broadly the criminal justice system can shape perceptions, even for issues that have nothing to do with sentencing and everything to do with police operations on the ground.

What Getting Caught Would Mean. Both drink drivers and young offenders have equally severe perceptions of the consequences of a future prosecution. However, the results show that actual prosecution does no better than conferences at creating a deterrent fear of court, though conferences do much better than court at creating fear of conferences. It is important that conferences do not diminish fear of court, and that court remains an ever-present option for the next offence. But this factor may be less important for drink drivers than the role of family and employers.

Family, Friends, and Community: Fear of Shame. The clearest evidence of the "Sword of Damocles" effect of shaming conferences is found in the drink drivers' fear of being discovered as a repeat offender. Drink drivers are more likely to say that it would be "big problem" if their family and friends found out about a future repeat offence if they had been sent to a conference (48 per cent) than after they had been sent to court (30 per cent).

The reason for this difference lies in the AFP rules about drink driving conferences. Unless the offender can assemble at least five family, friends and supporters to attend the conference, the police almost always refuse even to schedule a conference date. If too few supporters show up for the conference, the session is cancelled and rescheduled. Ultimately, failure to get enough family and friends involved may mean that the case is sent on to court.

Once the offender has put family and friends to all the trouble of attending the conference, they generally become committed to helping the drink driver avoid repeat offending. The offender makes undertakings to the family and friends attending the conference not to commit another offence. Offenders sent to court, in contrast, can often hide the fact from their family. At the least, the family does not get involved in the criminal justice process to the same degree in court as they do in conference. Thus conferences may do a better job at mobilising family and friends as a deterrent to crime.

A similar pattern appears in the fear of drink drivers having their names published in the newspaper if they are caught again. Offenders who have been to court and thus had their name published already are less likely to see it as a "big problem" (26 per cent) than offenders who were sent to a conference (37 per cent), thereby avoiding both a criminal conviction and the community stigma of their name in the paper. This "sword of Damocles" means that the offenders still have something to lose if they re-offend. This does not necessarily mean that publishing drink drivers' names in the press is an effective crime deterrent. Whatever deterrent value it may have might come from the threat of publication, rather than actually carrying out the threat. The evidence for that interpretation comes from the strongest test of conferences we have carried out to date: their effect on offenders' prediction of future law-breaking.

Preventing The Next Crime

While we have yet to collect police records or interview data on how many crimes offenders actually commit in the years after being caught by police, the initial interviews offer an important insight about repeat offending. A common procedure in criminological research is to ask people if they are likely to obey the law, or commit crimes, in the future. Answers to that question in studies conducted in the U.S often show shocking admissions of intentions to re-offend. Such admissions are far less common in the ANU evaluation to date. But the preliminary results show a clear difference between court and conference cases.

Both drink drivers and young offenders are significantly more likely to say that they feel encouraged to obey the law after being sent to a conference than after being sent to court. The differences are not great, but they are not likely to be due to chance. Among drink drivers, the answers favoured conferenced offenders by 91 per cent to 82 per cent for court cases. Among young offenders, the difference was somewhat larger, favouring conferenced offenders by 92 per cent to 76 per cent for prosecuted cases.

These differences are small in absolute terms, but they are large relative to the amount of repeat offending in Canberra. Only 8 per cent of drink drivers are apprehended again within two years of their last offence, so a 9 per cent difference in intentions to offend could have a substantial effect on overall recidivism. Young offenders are apprehended again at a much higher rate - approximately 30 per cent for first offenders over the two years following their last offence. Thus a 16 per cent difference in intentions to offend, if correct, could cut juvenile recidivism in half.

Whether these offenders' deeds will live up to their words remains to be seen. Given the indications of both stronger deterrence mechanisms and perceptions by offenders of greater procedural fairness in conferences than in court, there is a clear need to complete the research based on official criminal records, so as to discover whether in fact offenders change their behaviour in the way they say they will.


Professor Lawrence W Sherman is an Adjunct Professor of Law at the Research School of Social Sciences, ANU and Scientific Director of the Reintegrative Shaming Experiments (RISE). Heather Strang is a Research Fellow in the Law Program at the Research School of Social Sciences, ANU, and Project Manager of RISE.