Punishment and restorative justice: an ethical comparison

Rev 27.6.20

 “BELTING SON “ACCEPTABLE” This was the headline in an English provincial newspaper, reporting a court case. The father was accused of hitting his three-year-old son with a leather belt after the boy damaged a television set and tried to blame it on his baby sister, although the father had repeatedly asked the boy to tell the truth. The boy had red marks and bruises on his arm, back and ribs. The judge indicated that he intended to pass a custodial sentence, but the accused failed to appear at court. His wife, a schoolteacher, said that her husband was “a good dad”, and had been brought up to believe that physical chastisement is an acceptable way of disciplining children. Although she had originally contacted the police, she said that she regretted doing so and tried to retract her statement, because she never expected that her husband would be arrested. This case illustrates several points: that once punishment is legitimated there is a danger that it will be overdone; that fear of punishment is liable to make people tell lies and deny what they did, like the little boy, or try to run away, like the father, instead of taking responsibility for their actions; and that the fear of overreaction by the state can make people, like the wife, think twice about reporting wrongdoing.

 This was clearly an example of the wrong sort of punishment. There are many more throughout history, and still to-day: amputations, solitary confinement, “supermaximum” security prisons … . Humankind has been horribly inventive in thinking up ways of inflicting pain. But is there a right sort of punishment?

 In comparing punishment and restorative justice, this chapter will consider definitions, and two kinds of justifications: instrumental and symbolic. It will show that punishment and restorative justice are based on different psychological principles, and different axioms. The question is asked how a “pure” restorative system would respond to certain problems, such as enforcement, and protection of the public; the concepts of “natural consequences” and “protective use of force” are proposed, but questions remain: how long should they last? Do restorative responses have to be “burdensome”, and if so do they cross the line separating them from punishment? An attempt is made to weigh the two philosophies in an ethical balance.


 Punishment is essentially the infliction of a painful or at least inconvenient measure on a person (including a “legal person” or organization: RJC, 2005) who has broken a rule. The intention is to show that the action was unacceptable, and to discourage or deter the offender, and others, from repeating itii. Some writers extend the definition to include coercive measures which are not intended to be painful, such as rehabilitation or reparation. They justify this on the grounds that the measure restricts a person’s liberty, and may be painful even when not intended to be, as with unreasonably long detention in a supposedly rehabilitative institution, or unreasonably heavy reparation. There is no objection to this definition so long as it is made clear that the word is being used in this way.

 But even if these measures are called “punishment”, they are clearly different kinds of punishment, intended to operate in different ways and should therefore be labelled differently, for example “rehabilitative punishment”, “reparative punishment”, and the somewhat tautologous “punitive punishment”. It seems simpler to call them rehabilitative or reparative “measures”, while recognising that their coercive aspects may unintentionally cause pain or inconvenience. We might use “interventions” for those that require the offender’s consent (but see discussion of consent, below). “Punishment” is then reserved for the deliberately punitive ones.

 Most definitions of restorative justice include the idea of reparation or healing of the harm done; dialogue between the victim and the offender in deciding how reparation can be made – sometimes called “deliberative” justice (for example by Walgrave 2005, 6); or both. Others add involvement of members of the community: as mediators, as organizers of NGOs providing a mediation or conferencing service, and/or as participants in the process itself: relatives and supporters of the victim and offender, and sometimes other people affected by the offence.

 In its purest form restorative justice replaces a court appearance, or takes place after completion of the criminal justice process, so that no one can feel under pressure to try to influence the decision to prosecute, the sentence, or the date of release from prison.. Punitive sanctions are normally imposed by a judge; so are rehabilitative ones, although some may require the offender’s consent, for example probation and community service orders when they were first introduced in England and Wales (but no longer, now that they have been re-named “community orders” with various “requirements” including “unpaid work” (Criminal Justice Act 2003 sec. 177)).

 Ideally, “Mediation in penal matters should only take place if the parties freely consent” (Council of Europe, 1999, Appendix, article 1). This consent is however limited if mediation takes place before the final decision of the criminal justice process: offenders inevitably hope for a lighter sentence or earlier release if they take part, even if they are told that this will not happen. Offenders may indeed accept the restorative justice process willingly, but this could only be entirely voluntary if they were told “You can take part in restorative justice, but if you do not, nothing will happen”; this would not be acceptable except for the most minor offences. Despite this, the fact that restorative justice offers offenders the choice whether to take part, and the opportunity to discuss reparation with the victim, does distinguish it from punishment imposed by a court.


 Punishment and restorative justice may be justified either by their effects or, as we shall see later, because they are believed to be a symbolically right response to wrongdoing. It is hard to justify ethically a society based on fear and the threat of harm, if another way of preserving social order is available. In addition, deterrence does not work when people think they can “get away with it”. What keeps most people from wrongdoing most of the time is primarily the desire to have self-respect and to be valued by those whose respect is important to them. They may also be influenced by the desire to keep the material possessions which often accompany their status. Thus, especially for acquisitive crimes, an appropriate strategy for prevention would be to ensure that, as far as possible, everyone can gain a reasonable amount of respect, status and possessions. Otherwise the only way to penalize those whose life conditions are already poor is to make them worse still. There is however a problem when the social status is regarded as arising from the material possessions, because this may tempt people to try to acquire more at other people’s expense. This applies at all levels, from the teenage robber who wants the latest mobile phone to the business executive who wants a bigger yacht and is willing to exploit people, or even commit ”white-collar crimes”, in order to get it.


 Since punishment (on the definition we are using) causes pain and often harm, it would seem that it can only be justified if it is more effective than any less harmful method of promoting the common good. In other words, only if it works, and certainly not if it makes matters worse: primum non nocere, the first thing is to do no harm.

 The wrongdoing of the powerful should not be (but too often is) overlooked. Regulatory or “white-collar” offences can result in financial hardship, injury and death. Methods for dealing with them could with advantage be applied to other crimes (see below, “natural consequences”) However, since “street crimes” generate most public debate, we will focus on them. A high proportion of convicted offenders come from a background deprived of basic needs for educationiii, affection and even safety; what they need is to be included, valued and respected for what they are, despite what they have done. Often they need help and support, ideally of course before they have done wrong; if it is not provided until afterwards, it may look as if they are being rewarded for bad behaviour. But if they are punished, they are further excluded, stigmatized and devalued: the opposite of what is likely to persuade them to behave more acceptably.

 Antony Duff argues for punishment as a way of communicating censure (Duff, 2005), although he concedes that present forms of punishment do not comply with the ideal of the communicative theory. But in practice its message is that the use of force is the way to control people. It communicates lack of respect and tells people what they should not do, not what they should do.

 Can punishment claim to be effective? A review of the psychological literature (Wright 1999, Chapter 2) shows that it inhibits unwanted behaviour only temporarily, and only if it is immediate, certain and not excessive – rare qualities in the criminal justice system. Also, an alternative form of behaviour must be available to the subject. Punishment can be counterproductive, by eliciting aggressive behaviour. People try to avoid it, but not necessarily by behaving well: offenders not only tell lies, try to blame others, or attack the police, but have even killed their victims in the hope of avoiding detection (ibid. pp. 41-2) – and the more severe the threatened punishment for one crime, the greater the incentive to try to escape it by committing another. It can produce what Braithwaite (2002, 106-7) calls “reactance”: people react to attempts to control them by acting in a contrary way, including aggressive behaviouriv. It can inflict disastrous “collateral damage” on the offender’s family. The most severe punishment (in civilized countries with no death penalty) is imprisonment, which is in a class by itself for producing unwanted side-effects: it separates people from potential good influences and prevents them from taking responsibility, besides being proverbially a “university of crime”. It is also commonly accepted that the younger people are when sent to prison, and the more often they have been imprisoned, the more likely they are to re-offend (Prison Reform Trust, 2006, 5).

 In her study of social and psychological influences, Fellegi (this volume) argues that retributive justice makes people more likely to stay at a “rule obeying” level, “neutralizing” (excusing) their behaviour to avoid punishment, rather than progress to moral values that take account of the needs of other people and society. It also stigmatizes them and makes it harder to form or preserve constructive social bonds.

 To compare restorative justice with punishment ethically, we must ask whether it achieves at least as much. The commonest question is, Does it deter offenders from re-offending? It has long been accepted that it is not the severity of punishment but the certainty of being caught that really deters (Wright 1982, 172-82). So a highly punishing system with ineffective authorities or with very low level reporting will not be effective in deterrence. As Cornwell (2006: 83) concludes, “The best that can be assumed about the operation of deterrence is that it might be a desirable by-product of punishment, but it remains one that we cannot reasonably rely on”. Advocates of restorative justice would say that in any case this is not its primary aim: so long as re-offending is not increased, it is justified by its other advantages, especially for victims. In any case, if it reduces re-offending, this is not the result of deterrence, fear of the consequences, but of persuasion, through empathy, showing what the offender can do, “earned redemption”. Moreover, restorative justice recognises – or ought to recognise – that although the quality of the process affects the outcome, offending or re-offending behaviour is mainly influenced by the person’s circumstances in early life before the restorative process, and the support they receive afterwards (Maxwell and Morris 2001). To over-simplify, whereas punishment says that offenders have done wrong, and rehabilitation says that they are disadvantaged or maladjusted and should receive help not blame, restorative justice recognises both individual free will and the effect of social and psychological pressures.

 There is, nonetheless, a growing amount of research indicating that when restorative principles are followed, re-offending can be reduced. The Restorative Justice Consortium (2006) has found that 34 of 46 studies showed positive results, 9 had mixed or neutral results, and only three, negative ones. Even allowing for methodological factors (most of them had comparison groups rather than random allocation), this is a favourable result. Sherman and Strang, in an extensive review of the evidence, conclude that “in many tests, offenders who receive restorative justice commit fewer repeat crimes than offenders ho do not”, especially in cases of violent crime, and that “diversion from prosecution to RJ substantially increases the odds of an offender being brought to justice” (Sherman and Strang, 2007, 88).

 In a culture conditioned to expect “toughness”, is restorative justice “soft”? The question implies that if it is, it is inferior to punishment. However, firstly, for many people, facing the person you have harmed is not a soft option (and for many, punishment is not a particularly tough one). Secondly, even if it is less tough, is that not justified as an encouragement to take part in a process which can benefit victims? Thirdly, as we have seen, it is doubtful whether the harm caused by tough punishment is outweighed, and therefore justified, by any beneficial results.

 Punishment has harmful side-effects; but could restorative justice also make matters worse? It can happen. Heather Strang found that 18 per cent of victims whose cases had been to conference said that the process made them angry; this is not a negligible figure, although when offenders had been to court, and hence mostly punished, nearly twice as many victims felt that way. Describing some of these cases, she concludes that most of the problems were due to poor practice, but in some cases it may be because the programme is set up in an offender-centred way (Strang 2002, 137-154). Some of the disappointment was due to the research design, in which some victims who wanted a conference did not get one because they were randomly assigned to the control group (ibid. p. 199). Kathleen Daly (2005) has also studied the victims’ perspective. In one study substantial numbers of victims were distressed, especially when the conference was not done well, but there was no comparison group to assess court experiences. Another study, of cases of sexual assault, did make the comparison: conferences were arranged more quickly than court hearings and led to more apologies, community service, and “measures that matter to victims” such as acknowledgement of the wrong. She adds however that she would hesitate to support conferencing for sex offenders unless a programme was available to help offenders to stop sexual abuse and harassment (ibid. p. 171-2, note 25). This is an example of how an offender’s co-operation with rehabilitative measures can be incorporated into the concept of reparation.

 There are also objections to restorative justice on the basis of its theory, not merely its implementation. Matthews (2006) is critical of the theory of reintegrative shaming, and of police-led conferencing; however, many, probably most, restorative justice programmes are not based on it, and are not run by police. Matthews does not make comparisons with the courts, which certainly impose non-reintegrative shaming; his thesis is mainly limited to recommending a “divorce” of restorative justice from reintegrative shaming, a conclusion with which many restorativists would not disagree. Other criticisms of restorative justice theory are dealt with by Morris (2002).

 As for rehabilitative measures, there is no space here for a detailed comparison but there is evidence that they can “work” (Bottoms et al. 2005). Thus, as there is no convincing evidence that punitive sanctions ar more effective than restorative interventions, punishment cannot claim ethical legitimacy on the grounds of being a more effective way of maintaining social order.

 Symbolism: a choice of axioms

 The basic axiom on which punishment is based is either simple retribution, “just deserts”, or the slightly more nuanced notion that if you gain some advantage over another person by unacceptable means, some harm should be inflicted on you to restore the status quo. The implication is that although this does no good, or even makes matters worse, it is still the right thing to do: “the guilty deserve to suffer” (Duff 2003: 48). The theory of deserts does not focus primarily on prevention and deterrence, but bases its argument on proportionate punishment for the act.

 Restorative justice is based on a different axiom: if you have caused harm, you should put it right as far as possible. It maintains that this has greater ethical validity than the criminal justice one. It also proposes a different procedure: the victim and the offender should have an opportunity to meet so that the incident can be seen as harm caused to another person, not just a breach of the law. Where there is no individual victim, the community may be regarded as the victim.v In fully restorative justice, the community is involved in the process, and community participation, it can be argued, is likely in the course of time to lead to community acceptance.

 One basic difference is that punishment is imposed, whereas reparation, in a restorative system, is agreed – or rather, there is an opportunity to agree on a measure, and only if that fails is it imposed. This is less than total voluntariness, but it can be argued that it is much less than total coercion, or at least coercion with an element of choice.

 Psychological principles

 The psychological basis for attempting to deter wrongdoing by punitive responses is a behavioural one: it is assumed that, like rats in a psychological laboratory, people who suffer pain after a certain action will stop doing it. If it “worked”, it could be considered justifiable, but the foregoing discussion does not support this.

 Restorative justice is based on a different psychological mechanism. In place of fear (of pain, rejection, being regarded as worthless) it puts hope (of actively earning reacceptance, and help towards it); instead of using victims, if at all, as a tool for securing a conviction, it offers them the opportunity to take part in the process of finding a constructive outcome. There is however a danger that it too can use victims as “a tool to prevent re-offending”, as Teresa Reynolds of Victim Support has pointed out (quoted by Johnstone 2002: 82).

 Punishment: the problem of quantum

 One problem with punishment, whether instrumental or symbolic, is that there is no logically defensible way of determining how much to inflict. (These questions are explored in more detail by Wright, 1999, Chapters 5, 6; Wright, 2003) If instrumental, how long is necessary to deter any individual? The more we have to lose in other ways, the less additional punishment is necessary to deter us. But then people who were already fortunate when they committed their crimes would be punished least, whereas a woman who was desperate to feed her family would risk even a severe punishment for doing what she believed to be right.

 If it is intended to be symbolic, how much punishment is needed to reflect our condemnation of, say, a burglar or an employer who ignores health and safety regulations? There is no way of quantifying the “wrongness” of an action, nor the amount of punishment that it should attract, except in relation to other punishments. Which is worse: a minor sexual assault or a major burglary? A short prison sentence, several years on probation, a fine, or some other measure? It is just as impossible to quantify these on symbolic grounds as on instrumental ones. Thus it appears that even if punishment were defensible, the amount of punishment has no logical basis. However, restorative justice has comparable problems, as will be seen below.

 Does restorative justice have benefits that punishment does not have?

 In several ways restorative justice claims to outdo criminal justice, and hence to be ethically superior. Courts and punishments will be considered together, since one normally leads to the other. It is important to compare like with like: restorative justice only takes cases where the accused has at least admitted involvement, so where the accused denies this, the victim’s experience of being cross-examined in the witness box has to be excluded from the comparison. However, if the restorative justice process accepts an admission of involvement, without insisting on a full admission of guilt, the accused can hope to avoid or mitigate punishment and is therefore more likely to make an admission and spare the victim her ordeal. Some victims of sexual offences may prefer to obtain such an admission, and possibly tell the offender the effects of his action, rather than use the criminal justice process with its notoriously high acquittal rate. As Daly (2005, 165-6) has pointed out, “The potential of restorative justice is that it opens up a window of opportunity for those who have offended to admit what they have done, without the potential risks associated with a court-imposed sentence”, whereas if a case went to court the chance of any sexual offence being proved was only 51 per cent.

 Victims have complained that the criminal justice process does not allow them to be heard. In some jurisdictions they are allowed to make victim impact statements, or even to speak to a court or parole board. But if this affects the punishment there are problems of fairness, because punishments would be even more inconsistent than they are already; and if it does not, they may wonder what is the purpose of it. Restorative justice, on the other hand, is concerned only with reparation, not punishment, and it seems appropriate that the victim should have a say. It also allows them to ask the offender questions, which the criminal justice process would exclude.

 In many cases, especially of violence, the victim and offender are known to each other, and the offence is the outcome of a dispute, often with faults on both sides. If one is responsible for having the other punished, their relationship is likely to be damaged or destroyed, leaving the offender unwilling or unable to make reparation; mediation, in contrast, offers the possibility of resolving the dispute.

 Potential for crime reduction

 The atmosphere of a mediation or conference session is intended not to be punitive, but is based on problem-solving; this encourages the offender to explain, and the victim to take part in the same spirit. It has also been claimed that a successful conference could then 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 but at 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, inadequate schools, lack of adequate recreational facilities for young people, members of ethnic minorities denied opportunities because of discrimination, and many more” (Wright 2005). Unlike the conventional system it encourages discussion of the offence and its background in a non-adversarial atmosphere, and can therefore point to types of situation where offending behaviour is more likely, so that preventive action can be taken – if there is the political will. “We should cease to look to severity of punishment to control crime …[but] should look primarily to measures of crime prevention outside the criminal justice system” (Cavadino et al., 1999, 51). One example comes from Zwelethemba and other townships in South Africa (Froestad and Shearing, 2006 and this volume; Roche 2003, 264-6). This is another feature that punishment does not offer.


 It comes down to this. There are two competing axioms: that lawbreakers should be punished, or that those who cause harm should repair it or make amendsvi. For some, the ethical choice is clear: a method which tends towards healing is preferable to one which inflicts further harm. For pragmatists, the decision between them should rest on their results. Here it is argued that the prevention of re-offending is by no means the only criterion for assessing the results, and restorative justice offers several advantages over criminal justice. There are however problematic areas which need to be considered when comparing restorative justice with punitive justice. One is proportionality: while judges do at least try to be consistent in applying a “tariff”, there would inevitably be wide discrepancies in the amounts of reparation asked by different victims. Restorativists reply that if one offender and his victim agree one form and amount of reparation, and a quite different settlement is reached in another similar case, each is right for the people concerned. Other questions include: what happens when the victim or the offender will not take part? How can reparation be enforced, and how would the public be protected? Should restorative justice make “burdensome” demands on the offender? In each case, does the restorative measure then become a punishment?

 It can be argued that in the case of serious offending the option of attempting to heal the conflict could intensify it and inflict more harm.  Van Stokkom (this volume) appears to accept the possibility that punishment can increase the risk of re-offending, but suggests that failure to punish serious offenders can stir up emotions, societal animosity and punitive mentalities, and perhaps do even more harm to society. For those individuals who demand punishment regardless of its effects this may be true; restorativists accept that actions must have consequences, but believe that those consequences should be constructive. They should benefit the victim and/or the community, and try to lead offenders to regret their actions if they do not already do so. Many restorativists would accept that if offenders do not make amends voluntarily, they could be compelled, and that if there is a serious risk of a further serious offence, they should be restrained or detained. But this, they would argue, is different from punishment, or at least a different kind of punishment.

 How would a “pure” restorative justice system respond?

 There are situations in which a fully restorative process cannot take place. Must we then resort to punishment? The victim may be unwilling to meet the offender or even to conduct an indirect dialogue, or the facilitator may decide, after speaking to them, that such a dialogue could be damaging or even dangerous to one or other. In these cases an offender who was willing to make amends to the community instead would be enabled to do so.

 To answer the following questions, two concepts distinct from punishment are proposed. The first is “natural consequences”. If someone is caught in an act of dishonesty, he will for a time not so readily be trusted; for a more serious offence, other people will naturally try to ensure that he will not be able to commit it again for some time, for example by disqualification from holding certain positions. Some traditional societies have no way of translating the word “guilty”; instead, people are asked if they take responsibility for what has occurred, and Aboriginal people emphasise not what has been done wrong, but on what needs to be done to put it right (Rudin 2005, 92) – which is also a central feature of restorative justice. The Navajo also speak of natural consequences rather than “right” and “wrong”: for example, if you do not tell the truth your fellows will not trust you and you will shame your relatives (Ross 1996, 107). If you have committed burglary or fraud, people will not readily employ you to paint people’s houses or handle money. In place of adversarial trials they aim at the restoration of harmony through co-operative problem-solving, and punitive prison sentences have been replaced by community healing programmes; “deterrence cannot be permitted to get in the way of healing” (ibid. 217, 216). Some natural consequences of wrongdoing are that others will want to change the behaviour of the wrongdoer, to eliminate any financial benefit, to put right any harm caused, and to encourage more considerate behaviour in future. These principles have been proposed for regulatory offences (Macrory, 2006) but could be no less applicable to “street crimes”. This is admittedly a difficult concept, and will need further work.

 If there is a serious risk of repeating a serious offence, the second concept may have to be applied: “protective use of force”. It is familiar to police officers: when a person cannot be controlled by other means, he can be restrained, but no more force than necessary must be used. It may result in restriction or deprivation of liberty (Wright 1982, 255-8, 262-3).

 What if offenders fail to perform reparation, or are totally unco-operative? For business regulation, which can also involve criminality, Braithwaite (2002: 31) proposes a “regulatory pyramid” in which, after persuasion and a warning letter, the next stages would be a civil penalty and a criminal penalty, followed by suspension and then revocation of licence. The civil penalty might be regarded as a “natural consequence”, but the criminal one seems out of place: if we use restorative measures because they are believed to be better than punitive ones, it seems illogical, when they don’t work, to use the punitive ones after all.

 If offenders refused to fulfil the agreement, first they would be warned and given an opportunity to re-negotiate it; after that it seems a natural consequence that their freedom could be restricted until they had done so. It can be argued that this is not punishment, because it stops as soon as they comply. But if they never do, the problem arises: for how long should their freedom be curtailed? This will be considered below.

 Although these measures would be for enforcement, not punishment, it has to be admitted that in this case the dividing line is very thin. It would also be a salutary reminder of the limits of state power: although the state can always inflict punishment, it cannot compel anyone to make reparation if they are determined not to.

 The offender may be assessed as so likely to commit a further serious offence that her liberty must be restricted. In that case, she might be disqualified from being a company director, placed under curfew, house arrest, or even detention, but without punitive intent. She could still make reparation – as some prisoners already do. Although these measures would have features in common with punishment, that would, once again, not be the intention; houses of detention would still take away freedom, but their ethos would be restorative rather than deliberately spartanvii.

 Restriction: another problem of quantum

 In all these cases where restriction of liberty is imposed, and especially where this is on grounds of public protection, the question arises: how long for? When we were looking at punishment, we argued that no logical basis for determining the amount of punishment can be justified, except in relation to other punishments. Is restorative justice open to the same criticism? Even in a mainly reparative system, the problem cannot be avoided. There are essentially only two ways of deciding. One is at the outset, on the basis of a “tariff” or a prediction of how long the person will remain a risk; but that can be affected by conditions in the institution and. especially after release. There is a risk that some individuals will have to be released despite being assessed as still being a danger to the public. Or the period can be indeterminate, and the individual can be assessed at intervals. There are strong objections to the latter option. There is no way of predicting “dangerousness” that is not wrong as often as it is right (Floud and Young, 1981). It was a feature of many totalitarian regimes which abused the criminal law, and could be oppressive even when ostensibly used with rehabilitative intentions. Indeterminate sentences ‘for public protection’ have however recently been introduced in England and Wales (Criminal Justice Act 2003, sec. 225). Their wide use has helped to increase the prison population to crisis levels, and they are being challenged by prisoners who allege that there are not enough rehabilitative programmes to enable them to qualify for release (Guardian, 23.6.2007, 13).

 It is hard to square this circle; one possible approximation is a semi-indeterminate sentence, with upper and lower limits, the exact release date being determined at regular assessments.

 Should there be lower or upper limits to the amount of reparation? A lower limit might be thought necessary because, even if a victim were so understanding and forgiving that they asked for little or no reparation, others in the community might want the offender to make substantial amends, possibly for non-restorative reasons such as retribution or deterrence. An upper limit to reparation might be needed because, although restorative justice does not attempt strict proportionality, it would be possible for an offender who was very contrite – or intimidated – to agree to a totally disproportionate amount of reparation. This raises the question, at what point does reparation become excessive, and who should decide?

 The final question is whether, as Duff (2005, 121) has suggested, although a restorative response should not be painful in the way that many prisons are, it should at least be “burdensome” to the offender. There are times when the boundary between restorative justice and punishment is not entirely clear. Walgrave (2002) argues that the intention of the punisher does make a difference. To take an analogy made by the late Dr Peter Scott, a prison psychologist, conditions on a small warship in a storm in the North Sea were almost unendurable; if they had been inflicted as punishment they would have been cruel, inhuman and degrading, but the men, even though they were conscripts, believed that they were serving their king and country and accepted the discomfort as part of their duty.

 Many people will have sympathy with the idea that restoration should entail at least some effort, if not pain, a $10 000 payment “weighed in favour of restorative objectives” (Roach, 2006, 185). But there is no mention of whether the victim was offered mediation or a conference, where he would have to face her and hear her feelings. If she did not want that, was making a payment enough? The restorative argument would be that when someone shows himself so insensitive to other people’s feelings and rights, they will naturally demand that he make a substantial effort to understand what he has done wrong, and make up for it, for example by attending a suitable course of gender awareness and performing community service. Does the fact of being ordered to make reparation turn it into punishment? It is only if we demand this because it is “burdensome” and writing a cheque is not burdensome enough, that it would be classified as punishment.

 This is comparable to the religious notion of repentance: In order to show that you feel sorrow for your action, you undertake a penance by confessing what you have done, promising to amend your life and to make restitution. It is recognised in the Austrian penal code as tätige Reue, active repentance, which can be a ground for reducing the sentence (Dünkel and Rössner, 1989, 157). Although the word penitence has the same root as “penal”, this definition does not include the infliction of pain for its own sake; but merely an appropriate act symbolizing contrition (Macquarrie and Childress, 1986, 462).


 If we take punishment to be the infliction of pain on a person who has behaved in an unacceptable way, a definition which, it is argued, does not include restorative justice, there could be two types of justification. One is instrumental: deterring that individual, and others, from similar acts. It is suggested that this could only be justified if it were more effective than measures which do not cause harm; but it is not – on the contrary, punishment often makes matters worse. The second type of justification is symbolic. It has traditionally assumed as an axiom that punishment is what ought to happen to wrongdoers. A different axiom is proposed: that people who cause harm ought to do as much as possible to put it right. It is argued that restorative justice achieves the main aims of punishment more effectively and with fewer side-effects, and other aims which criminal justice does not attempt. Since on this view punishment has no clear practical advantage, the ethical choice rests on the symbolic, declaratory functions; of the two, the one based on repairing the harm to the victim and enabling the offender to earn reacceptance seems preferable.

Originally published in Images of Restorative Justice Theory  by http://www.polizeiwissenschaft.de/, 2007.


 Braithwaite, J., Restorative justice and responsive regulation. New York, Oxford University Press, 2002.

Cavadino, M., Crow, I., and Dignan, J., Criminal justice 2000. Winchester, Waterside Press, 1999.

Cornwell, D. J., Criminal punishment and restorative justice. Winchester, Waterside Press, 2006.

Council of Europe, Recommendation No. R(99) 19 of the Committee of Ministers to member states concerning mediation in penal matters. Strasbourg, Council of Europe, 1999 (reprinted in I Aertsen et al., Rebuilding community connections: mediation and restorative justice in Europe, Strasbourg, Council of Europe, 2004.

Daly, K., “A tale of two studies: restorative justice from a victim’s perspective.” In Elliott, E., and Gordon, R.M. (eds.), New directions in restorative justice: issues, practice, evaluation. Cullompton, Willan, 2005.

Duff, R. A., “Restoration and retribution.” In von Hirsch, A., Roberts, J.V., Bottoms, A.E., Roach, K. and Schiff, M., eds. Restorative justice and criminal justice: competing or reconcilable paradigms? Oxford and Portland, Oregon, Hart Publishing, 2003.

Duff, R. A., “Punishment and the morality of law.” In Claes, E., Foqué, R. and Peters, T., eds. Punishment, restorative justice and the morality of law. Antwerp and Oxford, Intersentia, 2005.

Dünkel, F., and Rössner, D., “Law and practice of victim/offender agreements.” In Wright, M. and Galaway, B.,, (eds.) Mediation and criminal justice, victims, offenders and community. London, Sage, 1989.

Floud, J., and Young, W.. Dangerousness and criminal justice. London: Heinemann, 1981.

Froestad, J, and Shearing, C., “Conflict resolution in South Africa: a case study.” In Johnstone, G., and Van Ness, D.W. (eds.) Handbook of Restorative Justice. Cullompton, Willan Publishing, 2006.

Haney, C., Banks, C. and Zimbardo, P. (1973) ‘Interpersonal dynamics in a simulated prison’. International Journal of Criminology and Penology, vol. 1, 69-97.

Johnstone, G., Restorative justice: ideas, values, debates. Cullompton, Willan, 2002.

Macquarrie, J., and Childress, J. (eds.), A new dictionary of Christian ethics. London, SCM Press, 1967.

Macrory, R., Regulatory justice: making sanctions effective Final report.. London, Better Regulation Executive, 2006; available at http://www.cabinetoffice.gov.uk/regulation/penalties

Matthews, R., “Reintegrative shaming and restorative justice: reconciliation or divorce?” In Aertsen, I., Daems, T. and Robert, L. (eds.), Institutionalizing restorative justice. Cullompton, Willan, 2006.

Maxwell, G., and Morris, A., “Family group conferences and reoffending.” In Morris, A. and Maxwell, G. (eds.), Restorative justice for juveniles: conferencing, mediation and circles. Oxford, Hart Publishing, 2001.

Prison Reform Trust, Prison factfile. London: Prison Reform Trust, 2006. www.prisonreformtrust.org.uk

Restorative Justice Consortium, Corporate Manslaughter: The Government’s draft Bill for reform. London, Restorative Justice Consortium, 2005.

Restorative Justice Consortium, Restorative justice works! The positive effect of restorative processes on re-offending. London, RJC, 2006 (updated on website www.restorativejustice.org.uk )

Roach, K., “The institutionalization of restorative justice in Canada: effective reform or limited and limiting add-on?” In Aertsen, I., Daems, T. and Robert, L. (eds.), Institutionalizing restorative justice. Cullompton, Willan, 2006.

Roche, D., Accountability in restorative justice. Oxford, Oxford University Press, 2003.

Ross, R., Returning to the teachings: exploring Aboriginal justice. Toronto, Penguin, 1996.

Rudin, J., “Aboriginal justice and restorative justice.” In Elliott, E., and Gordon, R.M. (eds.), New directions in restorative justice: issues, practice, evaluation. Cullompton, Willan, 2005.

Sherman, L., and Strang, H., (2007) Restorative justice: the evidence. London, Smith Institute, 2007.

Strang, H., Repair or revenge: victims and restorative justice. Oxford, Clarendon Press, 2002.

Walgrave, .L, “Restorative justice and the law: socio-ethical and juridical foundations for a systemic approach.” In Walgrave, L. (ed.) Restorative justice and the law. Cullompton, Willan, 2002.

Walgrave, L., “Towards restoration as the mainstream in youth justice.” In Elliott, E., and Gordon, R.M. (eds.), New directions in restorative justice: issues, practice, evaluation. Cullompton, Willan, 2005.

Wright, M., Making good: prisons, punishment and beyond. London, Burnett Books, 1982. Now on www.restorativejustice.org

Wright, M., Restoring respect for justice. Winchester, Waterside Press, 1999.

Wright, M., “Is it time to question the concept of punishment?” In Walgrave, L. (ed.) Repositioning restorative justice. Cullompton, Willan, 2003.

Wright, M., “The paradigm of restorative justice”, in: Research and Practice section, VOMA Connections (Victim Offender Mediation Association, USA) 2002 (11), Summer. (also published in Russian: “Парадигма босстанобительного правосудия”. Вiдовне Правосуддя в Украïнi, Киïв, Украïнський Центр Порозумiння (1/2 2005).)

COST/First drafts Wright WG3 punt Dec 06C

i The Star (Sheffield), 11.5.2006 and e-mail from The Star, 30.5.2006

ii Van Stokkom (this volume) argues that state punishment has other intentions, such as deflecting demands for mob punishment; but community-based restorative justice programmes have shown that structured community responses are not necessarily punitive..

iii Presumably something must also have been lacking in the education of white-collar criminals.

iv Restorative justice also attempts to influence behaviour, but by persuasion rather than coercion.

v The issue of what “the community” is has of course generated much debate, which cannot be summarized here.

vi In either case, those considered to present a serious risk of committing further serious offences may have to be restrained by restriction or deprivation of liberty, but this is a different type of intervention which may or may not be combined with punishment or reparation.

vii There is always a danger that people with power over others in a closed institution will abuse it (Haney et al., 1973), and proper safeguards would be necessary.