An international approach: what is restorative justice?

Paper to Seminar in honour of Gunnar Marnell (27.2.1913 – 16.4.2002), ‘Mediation and restorative justice’ (Medling och reparativ rättvisa), Stockholm, 18 April 2002.

Revised version presented to Conference on Restorative Justice, held by the Ministry of Justice, Tel Aviv, Israel, May 2002. An edited version was published as ‘The paradigm of restorative justice’ in Research and Practice section, VOMA Connections (Victim Offender Mediation Association, USA) 2002 (11), Summer. (Published in Russian: ‘Парадигма босстанобительного правосудия’. Вiдовне Правосуддя в Украïнi, Киïв, Украïнський Центр Порозумiння (1/2 2005).)

It is a great honour to speak at a seminar held to honour Gunnar Marnell. You know better than I do his achievements in Sweden: how he founded the journal Tidskrift för Kriminalvård in 1946 and was its editor for many years. Under his editorship it gave special prominence to the treatment of offenders – in both senses of the word ‘treatment’: not only that they should receive therapy, such as group therapy, when they needed it, but that people, including prison officers, should show them respect. As regional director of prisons and probation he encouraged the practice of these ideas, notably at Gävle prison. He was one of the most progressive prison governors in this progressive country, and in my time at the Howard League for Penal Reform, where I used to work, I used to depend on him for information which might help to advance the cause of reform. For many years he was a member of the League and even after I left, we continued to correspond. He would write letters or telephone, with enquiries about bail hostels, day training centres, reparation, and many other points in which he was taking an interest, usually because he was going to a meeting and wanted facts with which to support a proposal to reform the system. He resisted the re-birth of the neo-classical notion of ‘just deserts’ which is based on punishment rather than treatment. Nor was his support for restorative justice uncritical; he resisted the idea of ‘shaming’ offenders, even if it claimed to be ‘reintegrative’; he described shaming to me as ‘a sort of vocabulary unknown to the lips of Socrates or of Christ.’ Långholmen prison, next to his house, has been converted into a hotel, but it contains a prison museum, and Gunnar would have approved of the caption of one of the exhibits:

Prisoners who endured a sentence in the stocks or pillory often suffered mental breakdowns afterwards or became so bitter towards society that they committed even more crimes. Punishment continued to result in the creation of an outlaw caste which in turn not only caused continued human suffering but cost society a substantial amount of money.

He was also a long-time supporter of Amnesty International. In 1986 he wrote to me that he had been only a ‘door opener’ for mediation, and was no longer involved; but two years later he helped to organize the Temadag on 27 May 1988 in Svea Hofrätt, and invited me to speak. It is therefore a particular pleasure to come back to Stockholm on a return visit fourteen years later. It is sad that he was not able to attend the seminar, but he knew before he died that it was to be held in his honour, and this gave him well-deserved pleasure.

At that time the main theme was ‘Alternativa brottspåföldjer i ett internationellt och skandinaviskt perspektiv’ (Alternative sanctions in an international and Scandinavian perspective); one of them was called victim/offender mediation, and was discussed at the conference. To-day the term ‘restorative justice’ is also widely used, and a European Forum for Victim/ Offender Mediation and Restorative Justice has been established. The basic idea has been to help the victim, but it has developed beyond that, as we shall see. It has been defined in various ways; the Restorative Justice Consortium in the UK says in its Manifesto that

Restorative justice seeks to balance the concerns of the victim and the community with the need to integrate the offender into society. It seeks to assist the recovery of the victim and enable all parties with a stake in the justice process to participate fruitfully in it.

Thus it also recognizes the other people who are affected by crime, such as the relatives of the victim and the offender; and it emphasizes that the process is no less important than the outcome; it can contribute towards healing the harm caused by the crime.

Restorative justice therefore challenges us to re-think our aims, our criteria for ‘success’. Although politicians, in Britain at least, tend to assume that the reconviction rate is the most important, advocates of R J often regard two of the primary aims as being satisfaction on the part of victims, and a feeling by offenders that they have been fairly dealt with, and research has shown generally good results on both these criteria. It is argued that if the performance on these measurements improves, restorative justice is justified, even if the reconviction rate is unchanged. Advocates of restorative justice are careful to point out that they do not regard restorative justice as a panacea; they hope, however, that it may be the first step in transforming our response to people who cause harm to others. The new vision has been expressed by Ezzat Fattah (1997: 265):

The primary aims of the criminal law should be to restore peace, heal the injury and redress the harm. Its main function should be to achieve conciliation, not retaliation; to settle human conflicts, not perpetuate them; and to bring the feuding parties closer together instead of setting them apart.

This paper is entitled ‘An international approach’; it is not necessary to describe how well (or how imperfectly) restorative justice is practised in various countries, because comparative studies and descriptions of the situation in individual countries have already been undertaken (Akester 2000, European Forum 2000, Miers 2001). What I should like to offer is a description of restorative justice as it could be, drawing on the experience of different countries in the process.

May I begin with an example? No case is typical, but this will show some of the restorative processes at work. Some details have been changed to preserve anonymity.

A teenager was caught grabbing a woman’s handbag in the street late at night. He and his victim were willing to meet, with a youth worker; his mother and the victim’s husband also came. He admitted the offence, and explained that he wanted money to take part in a big celebration; he apologized both to the victim and to his own mother for the distress he had caused them. He was doing well at school, and was afraid that this incident would harm his prospects. The victim said she was not concerned about the money, but she wanted him to do some community service for people less fortunate than himself, to make him realize how lucky he was.

It was arranged that he would work for a project which helped disabled children to learn to ride; he would avoid certain young people with whom he had been associating; and that his school would allow him to stay and complete his examinations.

This case, from one of the new Youth Offending Panels now being introduced in England, shows several things. Victims are not always primarily concerned about money. Community service can be relevant to an offender, and to the victim’s wishes, even if it is not related to the offence. It probably has more value if the offender can meet face-to-face with the people who benefit. Control over young people’s behaviour can be exercised by agreement, not necessarily by the order of a court. And the community has a responsibility to make the reparation possible: the riding project did so by allowing the young man to work with the disabled children, and the youth worker by arranging the placement.

In presenting an outline of this kind, it is simpler to think mainly of individual victims and offenders, and of disadvantaged offenders offending against ordinary ‘good citizens’; we should remind ourselves, however, that the poor often commit offences against their equally deprived neighbours, employers commit offences against their employees and the public by breaking laws on health, safety, price-fixing and the pollution, of earth, air and water, and so on; the dynamics of such offences may be different, but they could still be handled in a restorative way.

The contrast between the old and the new perspective has been summarized by the American Howard Zehr (1995), who has been the director of a victim/offender mediation service as well as a leading philosopher of restorative justice. Some of his points are:

The old and new aims

  • Wrong as violation of rules

  • Focus on infliction of pain

  • Justice based on rules, outcome

  • Focus on guilt and abstract principles

  • The state as victim


  • Wrong as violation of people, relationships, shalom 1

  • Focus on making right

  • Justice defined by substance, process

  • Focus on harm done

  • People, shalom as victims

Features of criminal justice

Traditionally criminal justice has tried to achieve two main aims, which are at least in tension with each other, if not incompatible: to deal with each individual case, and to send messages to the public. Each of these efforts is also composed of conflicting objectives (Wright 1999).

Individual cases In dealing with each case, one aim is individual deterrence: to deter the individual from offending again (and bear in mind that the word ‘deter’ has the same root as ‘to terrify’, to make people afraid to repeat the offence). This raises the question: Even if deterrence worked, which is open to considerable question, do we really want to live in a society where the reason why other people do not harm us is not respect, but fear? The effect on the total amount of crime is in any case limited: in Britain it is estimated that the proportion of crimes which reach the criminal justice system and result in a warning or a conviction is about 3 per cent. If this were substantially increased, the system would break under the burden. It is reported that in some parts of the United States the education budget is being cut in order to pay for prisons: a short-sighted policy if ever there was one.

The second is to rehabilitate: here the idea is to provide what many offenders lack, such as skills, self-esteem, accommodation and work, and to encourage them to take control of their lives and live without harming others. This can be described by the phrase ‘to persuade and enable’. This is very necessary, but it raises some questions: does it focus attention on the offender to the exclusion of the victim, and is it helping, or even coercing, the offender to conform to a basically unequal and unfair society?

More recently a further objective is often added: to require the offender to make reparation. In England, however, this idea has been confused with the idea of punishment: that is, punitive sanctions imposed on the offender which are intended to be unpleasant, and fail in their intention if the offender does not find them so. Restorative sanctions, as we shall see, work the other way round: they succeed if they are reparative, but if the offender does not find them unpleasant that does not mean that they have failed. It is therefore confusing to describe reparative sanctions as ‘punishment’, because they can succeed even if the offender finds pleasure in doing them: some offenders continue with their community service after completing the agreed number of hours of work (and an example of this will be given later).

Messages to the public In addition to the individual aims of the criminal justice system, it has the intention of sending messages to the public. The first of these is general deterrence: if you break the law, you will be punished. Three of the main objections to this are: that it only works to the extent that potential lawbreakers believe that they will be detected; that if a punitive sanction fails to deter, it has to be imposed, and is usually damaging to the offender and to his family if he has one; and that the state cannot be justified in inflicting deliberate harm on its citizens unless there is no other way of safeguarding its citizens (and, some would say, not even then). Advocates of restorative justice maintain that there is another way, as we shall see.

The other public aims of punitive sanctions are symbolic. The second is retribution, which conveys the idea that ‘justice has been done’; ‘justice’ in this sense is more-or-less synonymous with punishment and means that the offender who has caused harm should suffer harm in return. Restorative justice, in contrast, maintains that the harm should be balanced not by adding to it, but by constructive action to repair it.

The third public aim is often called ‘denunciation’: it is upholding social norms by indicating what behaviour is not tolerated; the worse the behaviour, the more severe the punishment. The equivalence is quite arbitrary (wright 1999), and varies between countries and even between judges. It too can be criticized on the grounds that it could be done in a less damaging way: it could be based not on the amount of pain inflicted, but on the reparation made.

Punitive sanctions have other disadvantages. They make the offender think of himself, not of his victim. They encourage him to deny, or minimize, the harm he has caused, whereas the victim wants the harm to be acknowledged. Where the victim was a witness to the crime, the defendant’s lawyer (in the Anglo-Saxon adversarial system at least) may subject the victim to a hostile cross-examination to try to create the impression that he or she is lying, or at least is not sure of the facts.

It is of course an oversimplification to present the criminal justice system as entirely punitive; some sentences are intended to be rehabilitative. Some prison regimes also attempt to include rehabilitative programmes in the regime; but this has obvious disadvantages: it has to compete with the harmful effects of imprisonment, and it depends on the support which the person receives in the community afterwards.

The prison system in some Western countries is in crisis. In England and Wales in August 2001 the prison population reached 67,000, 0r 125 per 100,000 of the population, and in 2002 it has risen to 70,000. In the United States it is 700 per 100,000, and 665 in Russia. Sweden, at 65, continues to set a relatively good example, but no one should be complacent: in the Faeroe Islands the rate is 20 per 100,000, representing 9 prisoners in a population of 43,000 – there are advantages in living in a small isolated community! (Walmsley 2002).

In short, people commonly over-estimate what the criminal justice system can achieve, and under-estimate its harmful side-effects (Wright 1999). We are looking for another way.

Features of restorative justice

In the classical model of criminal justice, the actors are the state and the accused. The state (symbolized in the United Kingdom by the Queen), through the Crown Prosecutor, acts on behalf of citizens in general. In restorative justice, there are three big differences. Firstly, the victim, if there is one, is brought into the process; secondly, the community rather than the state is actively involved so that the relationship is now a triangular one between victim, offender and community (and I will say more about that word ‘community’ later); and thirdly the process itself is seen as part of the restorative response, and not merely as a means of reaching an outcome.

For each of the three parties to the triangular relationship there are some Do’s and Don’ts.

  • Victims should have the opportunity to take part in the process of repairing the harm caused by the crime, but should not have a part in deciding punitive sanctions.

  • Offenders should be encouraged to acknowledge the harm they have caused, and to make up for it as far as they are able to do so, but they should not be held responsible for social pressures which influenced his or her behaviour (I have avoided the word ‘cause’).

  • The community should assist the reintegration of offenders, and should not stigmatize and exclude them; it should also offer support to victims

How does restorative justice work?

To see how restorative justice works, we can look again at the points of the triangle victim – offender – community; and then we can return to the point made at the beginning: if we are sceptical about the deterrent effect of punitive sanctions, what does restorative justice offer in their place as a strategy for trying to reduce crime and harmful behaviour?

1. For victims, we first have to remember that for the great majority their offenders are not known. We should not exaggerate: for many, being a victim is an annoyance rather than a trauma, but for those who have been seriously affected by the crime, there should be support. The Victim Support movement offers this to many victims: in the United Kingdom volunteers with basic training contact victims to offer support, and put them in touch with professional help if they appear to need it; in some countries the victim requests the service. For victims of violence, many countries offer compensation for medical costs, if these are not covered by the welfare state, and in some cases compensation for pain and suffering as well.

Where the offender is known, this may be because the offence resulted from a dispute, and there are many advantages in resolving it in a way which leaves both parties able to speak to each other, rather than to use a process which leaves one of them responsible for the fact that the other has received a criminal conviction. Often there are faults on both sides.

Where the crime was committed by a stranger, the British Crime Survey has shown that twice as many victims of burglary, mugging (street robbery) or violence by a stranger wanted a non-custodial sentence rather than prison for their offender, 41 per cent were willing to meet their offender, and a further 17 per cent were willing to receive reparative activity (Mattinson and Mirrlees-Black 2000: 42). In Northern Ireland, 75 per cent of a random sample would be prepared to attend a meeting to help decide what should happen to a young person who had stolen something from them (Amelin et al. 2000: 35).

The first victim-offender meetings used one-to-one mediation, with the help of a mediator. Since then other methods have developed. Gunnar Marnell told me that in 1997 Gun Hellsvik, the former Minister of Justice, visited New Zealand and was impressed by family group conferences (FGCs). As you will know, the Swedish Commission on Mediation (SOU 2000:105) looked at FGCs and considered that they could exert a positive influence on perpetrators, provided of course that they were well managed. The principle is similar to that of victim/offender mediation, but the extended family of the offender is encouraged to be present, as well as supporters for victims. Uinlike victim/offender mediation, a conference can take place in the absence of the victim. The optimistic assumption is that the more people there are present in the room, the greater the chance that some of them will have constructive ideas for the offender’s future, and may offer to help to implement them; the young person may go to live with a relative, and there have even been instances where the victim has offered the young offender a job. Until now the method has been used primarily for young offenders, but it is being extended to adults, and this is as it should be, because restorative justice is intended to benefit victims, and they should not lose that benefit merely because the offender happens to have passed his or her eighteenth birthday.

2.  Offenders, for their part, can offer amends or reparation. Agreeing to take part in mediation is already the beginning of reparation. Mediation shows them the consequences of their actions, and often when they understand that they are genuinely sorry for what they did. We need to think about the idea of reparation. In its most obvious sense, it means repairing the damage, or paying some compensation towards it. In some cases the offender can work directly for the victim; individual victims do not often want this, but corporate victims (‘legal persons’) such as a shop, a club or a school may find some useful work for the offender to do.

If the victim does not want reparation, the offender may do something for the community, because crime harms the community as well as the individual victim: other people are afraid that they will be the next victims, shops put up prices to cover the cost of theft and of closed circuit television, and so on. Thought needs to be given to this. Some work, such as picking up litter, can feel more like punishment than reparation. In England it has been suggested that the work should be related to the offence; this is reasonable, but it may be more effective if the work is related to the offender’s abilities. We need to think what we are trying to achieve. If we want the work to be boring and humiliating for the offender, that is primarily a punitive sanction, which does not show respect for the offender and is unlikely to make him respect other people. If the work is obviously valued, and especially if it involves meeting the people who benefit from it, it is more likely to build the offender’s self-respect and enable him or her to earn the respect of other people.

There is another way in which offenders can make reparation. A common reaction of victims, not only after crimes but after road crashes, medical tragedies and other disasters caused by human error, is that although nothing can undo the harm, they want to see action taken to make it less likely that others will suffer as they have done. They would like the offender not to create any more victims, and if he co-operates with any programme providing the skills, training or treatment that he needs, then they will regard that as a form of reparation.

3. The community can make it possible. There has always been a tendency to hold the individual offender responsible for the offence and for his success or failure in keeping away from crime. Of course offenders can make choices, but they are also subject to pressures, as we all are. The original idea of victim/offender mediation considered only the individual victim and offender, but as it develops, some community involvement appears to be desirable, and in some cases essential. To explain this, we had better look at that word ‘community’. I suggest that it should be defined in an inclusive way. It has (at least) five aspects.

Firstly, the victim and his or her family and friends, and the offender’s family, are all members of the community, and indeed the aim is to make sure that the offender him or herself is integrated as a full member of the community. The group conferencing version of restorative justice brings them into the process.

Secondly, in some places the mediators (facilitators) are trained lay volunteers (in some places paid a small fee for the work they do). Their involvement helps to spread understanding about the realities of crime and the process of restorative justice among their friends and colleagues. In other places mediation is carried out by professionals; this has some advantages, but it creates a barrier between the professionals and other people; and there is a tendency for professionals to be less flexible about working in evenings and weekends, the times which victims and others often find more convenient.

Thirdly, restorative justice should not be judged solely by what takes place in a meeting lasting for perhaps two hours. Offenders need work and accommodation, and employers and owners of property have to provide them Individuals can also contribute, as ‘mentors’ (volunteers who support offenders), and in Canada and England there is currently interest in ‘circles of support’, especially for sex offenders who have undergone treatment, in which a group of trained volunteers help – but also control – an ex-offender (Church Council 1996).

Fourthly, non-governmental organizations (NGOs), which are managed by members of the community (sometimes with the involvement of officials from social services or other statutory agencies) have several roles to play. In some places, they operate the mediation service to which cases are referred by police, prosecutors or courts. NGOs can also find socially useful work for offenders to do when their reparation takes the form of community service. They can also provide support for victims, especially those whose offenders are not known.

Lastly, I suggested earlier that one way in which offenders can make reparation is by co-operating with whatever education, training or therapy they may require. This is a two-way contract. If he is willing to co-operate, then the community has to make it possible; and in this case the community can means the NGOs which run drugs treatment programmes and so on. Or the local authority, the municipality, which is the agent of the community that should provide the necessary services for which people pay taxes.

4. Crime reduction: feedback After the victim – offender – community triangle, there is one more aspect which should be included in restorative justice. The ‘criminal justice system’ is often criticized for not being a system but a collection of agencies – police, prosecution, courts, probation, prisons and so on, which have differing aims and operate semi-independently. It lacks the essential feature of a system, which is feedback. In a system, the operation is monitored, failures are noted and information is fed back to the operator so that corrective action can be taken. The criminal justice system is designed to answer very specific questions, such as: did this person commit this act, was it a crime, how much punishment should be imposed? Much relformation that he has to, in case it leads to a heavier punishment.evant background information is excluded because it is not legally relevant under the laws of evidence, and as we have seen the offender has no inclination or incentive to give any more in

In restorative justice the dynamic is quite different. It only takes place when the offender has admitted his involvement in the causing of harm (and the great majority of offenders do admit their guilt). The atmosphere of a mediation or conference session is not adversarial, but is based on problem-solving: people are there because they are looking for ways of repairing the harm. The victim wants to ask questions which have no place in a court of law, such as ‘Why did you choose me?’; if he or she asks about the offender’s background, the offender can try to explain, and will not, as in the court room, sound as if he is only making excuses. A conference could be compared to a small ‘truth and reconciliation commission’, where the background can be explored. Restorative justice does not (or should not) aim only at persuading individual offenders not to re-offend, and for its effect on potential offenders it does not rely only on deterrence (apart from the fear of being caught) but on a more nuanced strategy of reducing social pressures towards crime. This means that the mediation service can build up a picture of factors which tend to lead to crime: not merely security factors such as easy-to-steal goods in supermarkets, but high unemployment, an inadequate school, lack of adequate recreational facilities for young people, members of ethnic minorities denied opportunities because of discrimination, and many more.

An example of how this can work comes from Germany.

Some children aged 11 to 15 were playing in the ruins of an old castle. It belonged to an insurance company, which was beginning to repair it, and in one room they found a stock of neon tubes, which made a big bang when they were broken. Some parts of the building were very dangerous, notably a shaft 30 metres deep. When the police visited the parents, some took a firm line with their children, but others pointed out that the local council provided no other play facilities.

A large meeting was held in the house of the local priest, with the facilitator of the mediation service. There was talk of the danger, especially as there were other ruins nearby; some of the parents had come close to being negligent. Then compensation was discussed. The foreman from the insurance firm showed evidence that the damage totalled DM 10 000; but said that the firm would forgo its claim to compensation in return for from 8 to 20 hours of work within two weeks, with the parents’ consent. The foreman would make sure that the work was not beyond their abilities. All this was confirmed in writing.

Finally the question of recreation facilities was raised. The Bürgermeister was brought into the discussion. It was agreed that some local sports clubs would be opened to young people for a reduced subscription, and security around two other ruins would be improved.

The children completed their work with enthusiasm, and did more than the agreed number of hours – they came back and asked if they could help again. The prosecutor discontinued the case. .

(Summarized from Hennig 2000)

Although this was not a very serious case, it shows several things about the restorative process. It can be used with a corporate ‘victim’ (a ‘legal person’) if the firm has an imaginative representative. Reparation can be effective without being punitive (the children enjoyed it). The community can be involved in the process, and may share some responsibility for allowing the offence to happen. Last but not least, a process based on problem-solving rather than punishment can encourage open discussion, from which the community can learn about pressures towards crime and can take preventive action.

If the process does not involve the whole community, as in this example, it should then be the responsibility of the mediation agency to feedback findings of this kind to the authorities responsible for social policy, so that remedial action can be taken. In England and Wales, under the Crime and Disorder Act 1998, the chief executive of every local authority, together with every chief constable of police, are given the responsibility of drawing up a crime strategy. This may be at different levels. So-called ‘situational crime prevention’ is quite straightforward: if there have been many thefts of unlocked cars, or of mobile phones, their owners can be encouraged to take greater care of their property, and in some cases manufacturers can be asked to make them more secure. Steering-column locks on cars are an obvious example. Security is a classic case of treating the symptoms, not the disease. The result is often displacement of the crime from one area to another, or, more frighteningly, to a higher level of violence: as cars become more secure, determined thieves steal them by forcing the driver to hand them over with the keys; as banks become more impregnable, thefts of cash in transit increase.

More difficult, but more effective in the long run, is ‘social crime prevention’, which is focused on those who are likely to be tempted or to be exposed to pressure to commit crimes. Restorative justice does not (or should not) aim only at persuading individual offenders not to re-offend. It does not rely, for its effect, only on deterrence (except the fear of being caught), but on a more nuanced strategy of reducing social pressures towards crime. If it found that many offenders are failing to attend a school which has a high truancy rate, or come from a part of the town where there are few jobs or recreational facilities, or belong to an ethnic group that is discriminated against. questions have to be asked about how to put right those deficiencies. This ought of course to be done as part of an equitable social policy to make life better for all, not merely to try to reduce crime.

Relationship of restorative justice to the criminal justice system

How would restorative justice be put into practice? Elements of the model which I will describe have all been put into practice in one country or another, but I know of no country which uses them all. Some are used only for juvenile offenders or for less serious crimes, and are therefore not fully restorative because they exclude victims of adults or of more serious offences, as was mentioned above. But I will describe what I believe to be a possible model, and you can consider whether it is indeed workable under existing law, whether it would be necessary to modify it to fit in with the law (or the proposed law Regeringens propositionen 2001/02:126, ‘Medling med anledning av brott’), or how the law could be changed. My idea would not be to see R J as an alternative method to be used alongside punishment and rehabilitation within a largely unchanged criminal justice system, but rather as a new paradigm, used on a small scale at present, but as experience grows, gradually being applied to a greater proportion of cases. Local community-based mediation services would provide mediation between neighbours, between victims and offenders, and in family matters, and could promote mediation in schools. There would still be an essential role for courts, however, as we shall see in a moment, and they would be encoluraged to fulfil it in as restorative a way as possible. .

For simplicity let us divide cases into five categories: those where the victim and offender know each other, petty offences, moderately serious ones and the most serious, and those which will have to be dealt with by courts.

1. Victim and offender acquainted When the offence is the result of a dispute, and especially when both sides have put themselves in the wrong, either of them could go to a community mediation service, operating as an NGO; without involving the criminal justice system at all. I deally each community would have a mediation service for all kinds of interpersonal conflicts: in families, schools, workplaces, between neighbours and so on.

2. Petty offences could be dealt with by a warning from the police or the prosecutor, in countries where that is possible; the warning could be a ‘restorative’ one, which means that instead of emphasising that the offender has broken the law, and will be in trouble if he offends again, it points out that he has harmed someone.

3.  Moderately seriousoffences could still be dealt with by the police or the prosecutor, who could divert them out of the system by referrring them to the local mediation service.. In some jurisdictions this will require permission from a judge; in England and Wales there is no express power for prosecutors to divert cases, although it can be argued that the Code for Crown Prosecutors does not rule it out (Crown Prosecution Service 2000, article 6 (h); Wright 1994, 1999). ‘Seriousness’ for this purpose should be defined not by an abstract legal category but in terms of the harm and trauma suffered by the victim, which justify the use of the resources required to organize a mediation or conference session. (It has been found that arranging a conference can take an average of 12 to 15 hours of staff time.)

4.  The most serious offences would be brought before a court; if the offender acknowledged his or her involvement in causing the harm, the court would ask the mediation service to arrange a conference in place of the pre-sentence report which is used a the moment. The agreement of the conference would be brought back to the court for ratification. This is what happens in the New Zealand juvenile system, and it is reported that in about 80 per cent of cases the court endorses the agreement. In some cases the court will feel it necessary to alter an agreement, for example if an offender has agreed to an unreasonable amount of reparation; but the reasons should be carefully explained to the parties, so that they do not feel that their ‘empowerment’ was not genuine.

5. A restorative role for courts What about the other 20 per cent? We have been talking about agreements, but if there is no agreement the court will have to decide on a suitable restorative measure. Although offenders sometimes agree to a curfew or to avoid certain areas in their town, they are unlikely to agree to a substantial curtailment of their liberty; so what is to be done if coercion is needed to ensure that they keep their agreements, or to protect the public? Again, the court will have to decide.

Secondly, when there is a serious risk that the offender will commit another serious offence, other members of the community will have to be protected and the court will have to go beyond the agreement and impose controls. Where possible these should involve restriction of liberty, for example by a ban on practising certain professions, or driving a car; only in the most serious cases there would be deprivation of liberty, detention. This would be only for the protection of the public, not punishment for its own sake; the offender should be enabled to make reparation as far as possible, and to receive any treatment that will satisfy the authorities that he will no longer be a danger if he is released, so that he can be released in the shortest time consistent with safety. This will of course entail difficult decisions, which will not always be right; but that is true of any system.

Thirdly, courts will continue, as now, to determine guilt where the offender pleads ‘Not Guilty’. And lastly, just as in the present process they uphold the procedural rights of the offender (‘due process’), they will take on the role (as the juvenile courts in New Zealand have done) of ensuring that mediation and conferencing are carried out in the spirit of restorative justice. In New Zealand lawyers are also available, not to represent offenders but to advise them.

The future: problems and aspirations

There will be tensions in any system, and that is true of restorative justice, just as it is of the criminal justice system. In some cases restorative justice simply uses a different philosophical base, in others two or more legitimate requirements pull in different directions.

One example of the former is proportionality. This victim and offender agree this amount of reparation; that victim and offender, after a similar crime, make a quite different agreement. Is this unfair? In restorative justice, what is agreed by both parties is fair, provided that it is not grossly excessive. Even if consistency in sentencing offenders were possible, which is very doubtful (Wright 1999 chapter 6), the involvement of victims makes it impossible to adhere to any form of strict proportionality. The aims of restorative justice are different, and this has to be recognized.

What happens if an agreement is not kept? The evidence suggests that people are more likely to do what they have agreed to, such as paying compensation, than what is imposed on them by a court. In England there is an additional incentive: if they complete the reparation their conviction is ‘spent’, that is, they do not normally have to mention it for example when applying for jobs. If they have difficulty in keeping the agreement, for example because it is too onerous or their circumstances have changed, it can be re-negotiated. They might have to sell a treasured possession such as a motor-bike or stereo equipment in order to pay compensation, or the victim could agree to a reduction of the amount of compensation. The matter will be treated more like a civil debt than a criminal sanction. In short, Justice may have to put away her sword and her blindfold, and use her ingenuity to devise a new set of scales with three pans, to balance the needs and wishes of the victim, for acknowledgement and reparation; of the community, for healing and reintegration; and of the offender, for re-acceptance by the community and the opportunity to lead a fulfilling life without harming others. ‘The pen is mightier than the sword’: so in place of a sword she would hold a pen, with which to write down victims’ and offenders’ agreements.

The restriction or deprivation of liberty poses a problem to which, I believe, there can only ever be a compromise solution: how long should it last? If it is for a fixed period, the duration is arbitrary and may be too long for some, but may release others who will re-offend; if release depends on the assessed risk of re-offending, someone has to predict the future, and predicitons are likely to be wrong as often as they are right. It is important to remember, however, that behaviour depends not only on individuals but on their relationships and the conditions into which they are released: once again, the role of the community is vital.

The traditional system is strong and well-established, and there is a danger that it will absorb restorative justice and brain-wash it into conventional punitive values. There could be pressure to make reparation proportional to the offence rather than the agreement of the parties; to make it punitive by choosing tasks because they are unpleasant rather than because they are constructive. This is not to suggest that unpleasant forms of community service such as picking rubbish from a polluted river should not be chosen; but the emphasis should be on the usefulness, the pride of achievement, not on the unpleasantness, and the task should be presented as one in which they can take pride, not (as Gunnar Marnell would have agreed) one of which they are ashamed. If the work involves personal contact with the people who benefit, so much the better.

In an imperfect world, even restorative justice could perpetuate injustice. If a person who has little or nothing steals something, and is required to pay back, the original inequality remains. If he or she has given way to pressures which few of us would have been able to withstand, the task of making things right should fall on all of us, not only the victim and the offender. This is shown by another case history, about a woman in New York. Her mother (the daughter of a bootlegger) was an alcoholic, had been severely abused by her husband, and taught her daughters to get money for sex. (I am only selecting a few of the adverse factors.) The woman herself was addicted to crack and cocaine, was jailed for fighting, and had her children selling drugs while young. One of her daughters, also addicted, neglected her daughters, but joined Narcotics Anonymous. Her daughter had a baby at sixteen by a crack addict who beat her. Five generations had little chance to break out of the cycle of poverty and rejection, and enter the conventional world of which they had little knowledge. The authors conclude that to tackle ‘[c]onditions that favour unemployment, educational failure, family dissolution, crime and drug abuse’ would involve much more than mere provision of services: be it improved education, policing or job training and availability. A transformation … will also necessitate a cultural change either by revolution or evolution.’ It also involves hearing the voices of those concerned, who are living under very hostile conditions (Dunlap et al. 2002).


Too much is often expected of criminal justice, and it focuses primarily on the offender. Restorative justice includes not only the victim, but also the community which, like the offender, has responsibilities as well as rights. It is like a triangle, victim – offender – community, but with an extra element: feedback from the restorative process to crime reduction strategy and social policy. It can be used at different levels: disputes which may be resolved through mediation without using the criminal law; petty offences, through a warning; moderately serious offences could be diverted before coming to trial; the most serious offences would be brought before a court, but mediation or a conference could still be used in order to determine the reparation and the offender’s action plan; and lastly the courts would have a further role in regard to the fulfilment of agreements, restriction or deprivation of liberty, cases where the accused denied guilt, and overseeing the process to ensure that it was as restorative as possible. It is important to recognize that a community’s response to crime does not focus only on the offender, or on the victim, but on the contribution of the whole community, in the way in which it responds to the harm and seeks ways of reducing it in future – even if that means changes to the structure of society. We need to look for ways in which our society can be made better for everyone to live in, and restorative justice is one of the ways through which we can become aware of the improvements that are needed.

Thanks to Kate Akester and Margarita Zernova for valuable comments on the draft of this paper.


Akester, Kate (2000) Restoring youth justice: new directions in domestic and international law and practice. Justice, 59 Carter Lane, London EC4V 5AQ

Amelin, K, M Willis, C Blair and D Donnelly (2000) Attitudes to crime, crime reduction and community safety in Northern Ireland. Criminal Justice Review Group, Statistics and Research Branch, Northern Ireland Office, Belfast.

Church Council on Justice and Corrections (1996) Satisfying justice: a compendium of initiatives, programs and legislative measures. CCJC, 507 Bank Street, Ottawa, Ontario K2P 1Z5.

Crown Prosecution Service (2000) The code for crown prosecutors. 4th ed. London: CPS.

Dunlap, Eloise, Andrew Golub, Bruce D Johnson and Damaris Wesley (2002) ‘Intergenerational transmission of conduct norms for drugs, sexual exploitation and violence: a case study.’ British Journal of Criminology, 42 (1), 1-20.

European Forum for Victim/Offender Mediation and Restorative Justice, ed (2000) Victim/Offender mediation in Europe: making restorative justice work. Leuven, Belgium: Leuven University Press.

Hennig, Silvia (2000) ‘Täter-Opfer-Ausgleich und Mediation: Methoden und Beispiele – Zweifel und ketzerische Gedanken.’ [Victim/offender mediation: methods and examples – doubts and heretical thoughts.] In: Günter Gutsche and Dieter Rössner, eds. Täter-Opfer-Ausgleich: Beiträge zur Theorie, Empirie und Praxis. [Victim/offender mediation: contributions to theory, research and practice.] Mönchengladbach: Forum Verlag Godesberg.

Mattinson, Joanna, and Catriona Mirrlees-Black (2000) Attitudes to crime and criminal justice: findings from the 1998 British Crime survey. (Home Office Research Study 200). London: Home Office.

Miers, David (2001) An international review of criminal justice. (Crime Reduction Series 10.) London: Home Office (as above).

Wright, Martin (1994) ‘New prosecution code may open way for mediation referrals.’ Mediation (Mediation UK), 10 (4) 8-9. See also ‘After mediation and reparation, prosecution may no longer be in the public interest.’ Mediation, 1999 15 (4), 13.

Wright, Martin (1999) Restoring respect for justice. Winchester: Waterside Press.

Stockholm 023

1 Shalom is the biblical word used by Howard Zehr to describe a community which is peaceful, orderly, and free. It does not mean that there will be no conflicts, but conflicts and crimes will be dealt with in a way that respects the rights of everyone – especially children.