Australian Institute of Criminology

Skip to content

The right kind of shame for crime prevention

Lawrence W Sherman and Heather Strang
ISSN 1328-3006 ; ISBN 0 7315 2800 X
Australian National University, Canberra
RISE Working Papers, no. 1
April 1997

As American judges move steadily towards greater humiliation and stigma as punishments for convicted offenders, the Australian Federal Police in Canberra are showing that shame does not require humiliation. In hundreds of drink driving and young offender cases over the past two years, the AFP have adopted alternative restorative justice techniques recommended by ANU Professor John Braithwaite.

Professor Braithwaite has defined two different kinds of shame. One kind is stigmatic shaming, which disintegrates the moral bonds between the offender and the community. The other is reintegrative shaming, which strengthens the moral bonds between the offender and the community.

Stigmatic shaming is what American judges employ when they make an offender post a sign on his property saying "a violent felon lives here", or a bumper sticker on his car saying "I am a drunk driver". Stigmatic shaming is designed to set the offender apart as an outcast for the rest of the offender's life. By labelling him or her as someone who cannot be trusted to obey the law, stigmatic shaming says the offender is expected to commit more crimes.

Professor Braithwaite's alternative to stigmatic humiliation is to condemn the crime, not the criminal. It gives offenders the opportunity to re-join their community as law-abiding citizens. In order to earn that right to a fresh start, offenders must express remorse for their past conduct, apologise to any victims and repair the harm caused by the crime.

Canberra police have adopted Professor Braithwaite's principles by diverting confessed offenders from court to a more intense, personal (and lengthy) alternative known as Diversionary Conferencing. In these conferences, which are convened by a police officer, offenders, their family and friends, and their victims or a community representative all actively participate. Conferences focus on the crime rather than the criminal, drawing out the bad consequences of the crime and planning how best to make up for them. If the offender agrees to the group's proposals for restoration, the police then monitor the offender's compliance in carrying out that plan. If offenders do not keep their promises, then these cases can be referred for prosecution, but nothing said during the conference can be used in a court.

Professor Braithwaite predicts that this kind of shaming and restoration will be more successful at preventing repeat offending than current methods, where most courts ignore shame most of the time, maintaining the dignity of legal process without unpacking the emotional forces resulting from the crime. These two approaches have never before been compared for their effectiveness at crime prevention.

The Canberra Reintegrative Shaming Experiments

For almost two years, staff at the ANU Research School for Social Sciences have been studying shame in Canberra's criminal justice system. What they have found is encouraging on two fronts. First, with one small exception, Canberra does not employ the "wrong" kind of shame - the humiliation that is growing so rapidly in the US. Magistrates are observed to be unfailingly polite to offenders, and interviews with offenders show they are well aware of the respect accorded them in court. Publication of the names of all convicted drink drivers in The Canberra Times is a modest form of public humiliation, but the stigma may be short-lived.

In contrast to the "no-shame" (or limited shame) approach of the Canberra courts, the ANU research team has found clear evidence that the police have succeeded in delivering the "right shame" approach in its program of diversionary conferences. These conferences are by no means simple to manage, and could easily have become an exercise in stigmatising offenders with a humiliating experience. But ANU observations of hundreds of conferences show that police officers who lead them have generally succeeded in preventing any participants from condemning offenders as bad people. While the conferences have been far more emotionally intense than court, most of the anger and shame have been aimed at the offenders' acts and not their character. Canberra police are succeeding in making offenders feel ashamed of what they have done without making them into shameful people.

Under funding from the Commonwealth Department of Health and Family Services and the Criminology Research Council, the ANU has gathered two kinds of data on shaming. One kind comes from observations of what happens when drink driving or young offender cases go to court, or when they are handled by a police-led conference. These observations measure the amount and kind of shame expressed during the proceedings, as well as other factors such as the emotional intensity of the participants.

The other kind of data collected so far is interviews with offenders and victims involved in the experimental cases. These interviews measure how the offenders felt during and after their case being handled through court or conference.

Still to be collected are the data from the offenders' criminal records, which will show how many offences they were charged with in the two years after they went to court or conference. This information, to which ANU has tightly controlled access under a formal agreement with the AFP and approved by the Privacy Commissioner, is absolutely vital to settling the debate over shaming - or at least the right kind of shame - and its effects on crime.

Using the same controlled research methods employed in medical trials, the Canberra Reintegrative Shaming Experiments (RISE) compare two groups of offenders who are similar in virtually all respects except one: whether their cases were handled in court or in a conference. These comparisons are made possible by Canberra police deciding which method to use based upon an ANU recommendation in each case. The recommendation 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 ANU research team has recently compiled preliminary results from the first 548 offenders interviewed for these experiments (111 young offenders and 437 drink drivers, representing around three-quarters of all those eligible for interview). The young offenders include juveniles apprehended for property offences ranging from shoplifting to car theft, and offenders up to the age of 29 involved in violent crimes (excluding sexual assault and domestic violence). The team has also completed observations of drink drivers appearing at 270 court cases and 237 conferences, and of young offenders appearing at 70 court cases and 63 conferences.

All the results reported here are statistically significant at the five per cent level, which means there is less than a five per cent probability that the finding is due to chance.

Shame without humiliation

Conferences are a challenge to the specially trained police officers who serve as conference "facilitators". They must be careful not to dominate the proceedings, and they are required to keep the discussion focused on what the offender did and how the harm can be repaired. This is sometimes difficult to achieve because of the anger that victims and their supporters may feel and their temptation to humiliate the offender.

The risk of humiliation is also greater because of the length of the conference process compared to court. For drink drivers, the average court case in the study takes 6 minutes, while the average conference takes nearly an hour and a half (88 minutes). For young offenders, the average court case in the study takes 13 minutes, while the average conference takes 71 minutes. Clearly the conferences put the offender under the spotlight of critical examination far longer than court proceedings. That represents both a danger and an opportunity.

The danger is that the conference may humiliate or stigmatise the offender more than court. The opportunity is to allow shame to follow from a discussion of all the harm the offender caused, with a forgiving acceptance of the offender's remorse.

What the Observers Saw. Rating scales used by the research team who observed both court and conference proceedings show conferences succeed on most dimensions of reintegrative shaming. Conferences are much more likely than court to express clear disapproval of both the type of offence and the specific crime the offender committed. The observers found no difference, however, in the amount of stigmatic shaming expressed in court and conference. This supports the conclusion that the Canberra experiments are comparing the "right shame" in conferences to "no shame" - rather than "the wrong shame" - in court.

Conferences are superior to court in evoking offender expressions of remorse. Both drink drivers and young offenders are rated as more apologetic in conferences than in court. Young offenders are twice as likely to offer apologies for their crimes in conferences as in court.

Conferences are also superior in evoking forgiveness in response to offenders' remorse. Both drink driving and young offender conferences were much more likely to elicit forgiveness than court appearances.

Canberra's police have also succeeded in creating the emotional intensity that Professor Braithwaite recommends for effective shaming. The research team rated offenders as more "emotionally engaged" in conferences than court appearances. Offenders were also more likely to cry in conferences, although their tears are rare events in both settings.

How the Offenders Felt. An essential element in reintegrative shaming is offenders feeling ashamed of what they have done. Both drink drivers and young offenders interviewed after conferences by the research team were more likely to say they felt ashamed of the crime they committed than those interviewed after court proceedings (though in the case of the young offenders this was statistically significant only at the 10 per cent level). Drink drivers sent to conferences were also more likely to say they felt ashamed of themselves, as distinct from what they had done. Drink drivers (but not young offenders) sent to court were more likely to say they felt humiliated.

Clearing the slate

Reintegrative shaming aims to prevent crime by allowing offenders to put their crimes behind them. Before this can happen it is essential that they repay society and their victims the costs their crimes have incurred, material and emotional. This should allow them to look forward to the regard of the law-abiding people about whom they care most, rather than joining a group of fellow-outcasts from society who have been stigmatised by the criminal justice process. The evidence shows that the Canberra police is succeeding in engendering this sense in the minds of offenders sent to conferences to a greater extent than may be possible in the court process.

ANU interviews of offenders asked whether they felt they had repaid their debt to their victims and to society - almost 80 per cent of the conference participants said yes compared to around 40 per cent of offenders sent to court. Both drink drivers and young offenders were much more likely to say that a conference had allowed them to "make up" for what they had done. Drink drivers, but not young offenders, were also twice as likely to say that a conference had "allowed them to clear their conscience".

Getting shaming right

The preliminary evidence strongly supports the conclusion that Canberra police have created the right kind of shaming. The delicate process of enlisting widespread community participation in a dialogue about an offender's crime could have resulted in either the wrong kind of shaming, or no shaming at all. If the police facilitators had permitted humiliation of the offenders during the conferences, they could have created "outcasts" from society and so possibly increased the risk of repeat offending. An even greater risk with drink driving cases is that offenders may bring supporters who are reluctant to condemn the crime of their friend or relative, and who then offer moral support for repeat offending. The additional presence of volunteer community representatives recruited by police to participate in the drink driving conferences provides the counterweight in such circumstances.

The operational challenge of reintegrative shaming thus requires a delicate balance between too much and too little condemnation and the evidence so far suggests police have met that challenge. Whether they have succeeded in reducing the risk of repeat offending remains the ultimate question for ANU researchers.


Lawrence W Sherman of the University of Maryland 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.