Mitigating risk in restorative justice

• 1 Introduction

  A key task for facilitators taking on a new restorative justice case is to assess the risk to all potential participants before proceeding. Assessing risk is not a one-off procedure – it needs to continue throughout the restorative justice proceedings and be reviewed regularly. Strangely, however, although assessing risk is high on the agenda of any discussion regarding the introduction of restorative justice, like in Scotland, and features in the training of facilitators, there is a lack of published research findings on what are the key elements, the extent to which risks can be mitigated and how this relates to the core values of restorative justice.¹x A literature search on ‘restorative justice risk assessment’ produced only publications looking at the risk of reoffending if restorative justice was undertaken, not on the risk of undertaking a restorative justice process. One exception is a brief report by Swida (2020), on adjusting risk assessment in response to the particular risk demands of Covid-19. There was also some work on using restorative justice with those with disability (Bolitho, 2019). Considering risk seems to be more an art than a science. This is doubly surprising given the prominence of risk assessment in relation to offending and criminal justice in the last few decades (see, for example, Andrews, Bonta & Wormith, 2016; Beck, 2004). A potential danger is then that the two may be confused: risk assessment of offending is about whether a perpetrator is likely to reoffend or to commit a serious crime. Risk assessment in restorative justice, however, is not about the outcomes from a restorative justice process, such as whether victim²x The terminology used for potential participants in restorative justice currently varies between countries. The international instruments relating to restorative justice refer to ‘victims’ and ‘offenders’ in terms of key participants. In Scotland and some other countries, victims are called those who have been harmed, whereas offenders are called those who have harmed. In some contexts, particularly that of gender-related violence, victims are often called survivors. What participants may call themselves also varies. In this article, we use the international terminology of offenders and victims, because we are writing for an international audience, but we are very aware that each term has potential characteristics or stereotypes associated with it. needs are met or there is less reoffending, but about what risks might occur if the restorative justice process proceeds. It is the assessment of the potential process to be undertaken, together with whether any risks can be mitigated, rather than trying to predict outcomes.³x There is of course a large research literature on restorative justice outcomes.
  The importance of undertaking risk assessments is shown by their mention in relevant international instruments. So, for example, the Council of Europe, in its 2018 Recommendation on restorative justice in criminal matters, states ‘Facilitators must be afforded sufficient time and resources to undertake adequate levels of preparation, risk assessment and follow-up work with the parties.’ (pt. 29). The European Union’s (EU’s) 2012 Victim Directive says:

  Victims who have been identified as vulnerable to secondary and repeat victimisation, to intimidation and to retaliation should be offered appropriate measures to protect them during criminal proceedings. The exact nature of such measures should be determined through the individual assessment, taking into account the wish of the victim (para 58).

  The EU Directive not surprisingly concentrates on potential risks to the victim, but, as the Council of Europe indicates, risks of participation in restorative justice can relate to, or be generated by, any potential participant.
  The current research was inspired by recent developments in Scotland, where the Scottish Government has committed to ‘have restorative justice services widely available across Scotland’ (Scottish Government, 2018).⁴x Criminal law and criminal justice are devolved matters in the United Kingdom, so provisions in Scotland are different from those in England and Wales and in Northern Ireland. Though restorative justice has been used in criminal cases in Scotland for at least 30 years, both with young people and with adults who have offended, its provision has been patchy and has concentrated upon cases with young offenders, with restorative justice being delivered by local authority social work services⁵x In Scotland, local authority ‘Justice Social Work’ performs a role similar to probation services elsewhere. or contracted voluntary sector providers (Maglione, Buchan & Robertson, 2020). Implementing the Scottish Government’s plans therefore will require a significant step change in provision, with training and public awareness taking place across the country, in urban, rural and island areas. The authors realised that it would be helpful to pin down more thoroughly how risk assessment and mitigation of risk might best be accomplished for this new provision.
  There is no universally agreed definition of restorative justice in criminal cases, with restorative justice being an ‘umbrella term’ (Shapland, Robinson & Sorsby, 2011) covering a wide variety of restorative processes, including face-to-face and virtual meetings between victim and offender (sometimes with supporters present as well), indirect processes of communication through neutral facilitators (indirect or ‘shuttle’ mediation, or sending letters or artwork) and panels or circles. The Scottish Government’s (2017: 6) definition of restorative justice, adopted in this research, is:

  a process of independent, facilitated contact, which supports constructive dialogue between a victim and a person who has harmed (whether this be an adult, a child, a young person or a representative of a corporate or other body) arising from an offence or alleged offence.

  The definition does not bar any particular type of offence from the provision of restorative justice processes, so there is potential for restorative justice to be used in more complex cases, involving serious offences and participants who know each other or have been in a previous relationship, as well as more minor cases or where potential participants are strangers. However, the former type of case has hitherto been neglected in Scotland; most Scottish restorative justice provision before the publication of the Scottish Government Guidance (2017) was provided under diversionary youth justice arrangements and centred on young people’s minor or first-time offending (Maglione et al., 2020).
  Though there has been little information available from research on mitigation and risk in restorative justice (though see Shapland, 2009; Shapland et al, 2011; Swida, 2020),⁶x We searched the academic literature using the keywords [risk – restorative justice – mitigation]. Very few articles were found, and those which were primarily focused on risk of reoffending, harm minimisation for young people or the generic needs of victims, i.e. outcomes, rather than the risk of undertaking a restorative justice process. References to ‘mitigation’ were about mitigation in sentencing after restorative justice, not mitigating risks during restorative justice processes. This is congruent with comments critiquing the paucity of literature on exactly how restorative justice is done (Shapland, 2013). facilitators in many countries have developed their own practice over many years, including working on complex cases (see e.g. the mapping of some of these practices in Keenan & Zinsstag, 2022). Possibly related topics, however, have been considered in the literature, such as the tendency of referring agencies to be protective of ‘their’ potential participants, so not providing information about restorative justice to victims, for example, leading to patchy referrals (Banwell-Moore, 2022; Van Camp & Wemmers, 2016). These are not, however, about risk assessment for restorative justice per se. The aim of this research is therefore to tap into the expertise of facilitators as to how they do risk assessment and their experiences of using different elements of mitigation in the restorative justice process adopted. The research hence sought out experienced facilitators and asked them to describe and evaluate their own practice on risks and mitigation. There are different cultures of restorative justice in different countries, reflecting the different legal systems, the ways in which traditions of restorative justice have developed and preferred theoretical backgrounds (Zinsstag & Vanfraechem, 2012), so we wished to explore practice in a variety of countries; however, for practical and linguistic reasons, we confined our research to those working in European countries.
  The aim of the study was to draw upon the experience of facilitators working in a number of different countries as to how they did risk assessment and what mitigation elements or processes they had called upon where risks were identified. In the tradition of reflective practice used by professionals, facilitators were asked to reflect on their ‘success’ – where success is considered as the absence of harm occurring to the participants, and what participants themselves said to the facilitators in terms of benefits. We were very aware that risk assessment needs to be considered in relation to risks relevant to that activity and setting and that, as the Health and Safety Executive (2022) in Great Britain has advised, employers for example should look at their own workplace and talk to their own staff. Hence we must not assume that risk assessment instruments can be taken from one situation and used in another. We concentrated upon what facilitators themselves had found and used in their own environment, what general practices of risk assessment they routinely used, and what elements of risk related to restorative justice they had encountered and mitigatory techniques they had used. We hope that the results may inform the development of good practice in this area and how we should view risk and mitigation.

• 2 Method

  Given the paucity of previous research on risk assessment for restorative justice, this research needed to be exploratory and inductive, whereby the themes and results are suggested through the data acquired. It was not possible to set out theoretical postulates and derive hypotheses from them. The decision was therefore to use a qualitative method, in-depth interviews, with open-ended questions. The research method used was semi-structured interviews with some 30 experienced restorative justice facilitators from different European countries, with interviews conducted virtually, via Microsoft Teams or Google Meet. This allowed a wide geographic ‘reach’ with minimal additional time or resource constraints. There is no register of experienced facilitators across Europe, so the authors used contacts from their different previous research projects and conferences since the early 2000s, trying to approach facilitators from different providers who would give an overview of restorative justice provision in the relevant jurisdictions. This is essentially a convenience sample, almost a snowball sample (Arber, 1993), but the potential bias of having respondents too similar to each other was mitigated by having four authors/researchers acting as starting points for the sample, each of whom had conducted evaluations and research on restorative justice in different countries and with different providers. The sample was designed to include both those working with adults and young people as offenders.
  The topic guide for the interviews covered facilitators’ previous experience with different types of participants and offences and their career in relation to restorative justice; agencies referring to restorative justice; information checked before approaching potential participants; which potential participant they would contact first; risks considered at the beginning; types of case and issues that would make them stop and think; risks and potential mitigatory actions they have found associated with communication; sites; safety during meetings/communication; safety after restorative justice; large group meetings; whether they had any separate category of serious or complex cases; approach in relation to domestic abuse/intimate partner violence, sexual assault, young participants; follow-up measures; the usefulness of backup, mentoring, co-facilitation, and action learning sets.
  Ethical permission in relation to the research was granted by the Research Ethics Committee of the School of Law, University of Sheffield. Facilitators agreed that we would not identify either themselves/their restorative justice scheme or individual cases, but would give the jurisdiction in which the work was done. The interview covered both restorative justice used within the context of criminal justice in that jurisdiction and also restorative justice used as diversion from criminal justice prosecution or trial.
  Full transcripts were made of the results from all the interviews and the results analysed thematically through reading and rereading interview transcripts and marking them up manually. Transcripts were available to all the authors and the themes identified were discussed and refined together, with more than one author involved in setting out and writing each theme discovered.
  The authors, between them, did 28 interviews with a total of 30 interviewees (one interview was with a group of three interviewees from one restorative justice provider). Table 1 shows the numbers of interviews and interviewees from each jurisdiction and the abbreviation used to indicate each jurisdiction in quotes from the interviews. Given the context of assisting the development of restorative justice in Scotland, emphasis was placed on interviewing facilitators from the United Kingdom and particularly Scotland.

  Table 1 Experienced facilitators interviewed and number of interviews by jurisdiction

  Jurisdiction	No. of interviews (and interviewees)	Abbreviation
  Scotland	8 (8)	Scot
  England & Wales	5 (7)	E&W
  Northern Ireland	2 (2)	NI
  Belgium	2(2)	Belg
  Denmark	2 (2)	Denm
  Norway	2(2)	Nor
  Finland	2 (2)	Fin
  Republic of Ireland	2 (2)	RoI
  France	1 (1)	France
  Austria	1 (1)	Aus
  Estonia	1 (1)	Est

  We asked the facilitators whether they had worked with adult offenders, young offenders or both and also what kinds of restorative justice processes they typically used. Our respondents included thirteen interviews with facilitators who had worked with both adult and child offenders, nine with facilitators who had primarily worked with adult offenders and six with facilitators who had primarily worked with children or young people as offenders.⁷x Note that the age of criminal responsibility and the age at which offenders were typically tried in adult courts varied between jurisdictions, so the term ‘children’ or ‘young people’ refers to those who were above the age of criminal responsibility but under the age of offenders normally tried in adult courts for that jurisdiction, but there was no uniformity in using the term ‘young person’ or ‘child’ in chronological age across jurisdictions. It can be difficult to ascribe different restorative justice practices accurately in categories, because the distinction between mediation and conferencing, for example, depends both on who is typically invited to a meeting/attendees and cultural and historical traditions of restorative justice in that jurisdiction (Zinsstag & Vanfraechem, 2012), but interviewees had often had the experience of several types of restorative justice processes, with 22 describing themselves as doing conferences (face-to-face meetings with more people attending than the direct victim(s), direct offender(s) and facilitator(s)), eighteen mediation (face-to-face meetings with the key participants only), and nine other forms of restorative justice, including indirect mediation/shuttle mediation, panels and circles. The vast majority of interviewees were working for or had worked for a restorative justice scheme/provider, with some being self-employed. The type of scheme or provider varied between jurisdictions, but overall, twelve were, at the time of interview being employed, within third sector restorative justice providers, eight within the penal system (such as in prisons or probation/social work services for adults or young people), three within the police and five in other settings (within government, in a victim agency, in a national mediation service or self-employed). Most providers used volunteers as facilitators as well as paid staff (eighteen, with ten having only paid facilitators and two where we were not certain of the employment status of staff).
  Interviews normally took about an hour. We asked about interviewees’ previous restorative justice careers, what information they would obtain (and from whom) before approaching potential participants, how they did risk assessments, whether any particular risks would make them cease exploring restorative justice and at which point, and whether they used a risk assessment schedule or checklist. We then raised some particular risks and scenarios and asked about both risk assessment and which mitigatory elements or strategies they had tried, including difficulties in communication, the site of any meeting, safety during a meeting or communication, safety after restorative justice and large group meetings. We asked if they categorised some cases as ‘serious or complex’ and if so, what kinds of cases would be in that category, and what procedures would be used or how those cases would be managed. After asking about follow-up procedures for cases, we concluded by asking what they or their schemes did to manage more difficult cases, including use of backup facilitators, mentoring, co-facilitation and learning sessions/continuing professional development.⁸x The management of cases and facilitators is not covered in this article. See Shapland, Buchan, Kirkwood & Zinsstag (2022).
  The research deliberately looked to talk to experienced facilitators, so they had dealt with some very serious and difficult cases. We specifically asked about whether they had experience of particular offence types. Of our sample of 30 interviewees, fifteen said that they had facilitated one or more homicide cases, twenty said they had facilitated sexual assault cases and sixteen said they had facilitated cases where domestic abuse was a known element (almost all of these serious offences were with adult perpetrators) – although these features might not always be reflected in how criminal justice systems categorise and process cases. It is important to bear in mind in considering the results below that this level of experience is not the average – respondents often tended to be talking about matters which would only be taken on by facilitators with several years’ experience. Indeed, 23 of our interviewees had had the responsibility of supervising (and indeed training) other facilitators. This was very helpful in eliciting good practice in managing taking on more difficult cases within a scheme, as well as useful mitigatory techniques.

• 3 The nature of risk in restorative justice processes

  The relevant international instruments for restorative justice (see earlier section) prescribe that risk assessment for potential participants should be an individualised process. This was the key message from our experienced facilitator respondents as well. Given that restorative justice is all about communication between individuals, the potential risks derive from the nature of the offence and the social contexts of the participants and no case is the same as any other case. Moreover, the ways that potential participants evaluate risks also vary between those participants. Given that the main reason for considering risk and how it might be mitigated is not to do harm to the participants (or anyone else), it is important to proceed from the basis of how the participant himself or herself evaluates that risk and what they are concerned about. We know that the effects of an offence on a victim are very much an individualised matter (Shapland & Hall, 2007), with different victims experiencing similar types of offences differently. Risks and their likely effects were found to be a very similar phenomenon. This runs very much counter to the prevailing actuarial conceptions of risk popular in mainstream criminal justice from the 1990s onwards and exemplified in statistical tools and computer software such as OASys (in England and Wales) and LS/CMI (in Scotland).
  Our respondents would obtain initial information from their own and other agencies before approaching participants, in order to have some key information about the case. They would not though presume that particular factors would lead to the same level or nature of risk. Because referrals often come from criminal justice personnel, often information about the offender was more readily available than that on the victim:

  So our pick-up point will generally be police reports, in our case juvenile offence notifications, so that’ll be Police Scotland submitting the summary of the crimes that have taken place. We may have no prior relationship with the young person responsible, or as is more likely, they’ll be known to us and there’ll be emerging patterns of behaviour. Often they’ll have an allocated social worker, and the allocated social worker will have some additional insights into that young person’s engagement, empathy – or lack thereof – maturity and openness to participating in any kind of process. (Scot5)

  Well, having clarified the way the requests come to us, the information will come through a number of ways. The information that comes from the court is in a very standard form. It’s the standard form that basically outlines the fact that the court has adjourned the matter for a pre-sanction report, and it’s a tick-box to say that restorative justice to be considered as part of the assessment. We still practice a court duty system where probation officers are in court, although it’s very resource-intensive, and I suspect that may change in the coming years. But for the moment, if we have a probation officer sitting in court, they will often provide the context within which this arose. (RoI2)

  In some countries, information sharing between agencies (for example, between the police and restorative justice providers) has been problematic. This tends to be due to rules or fears from the agency, not the potential participant. Sometimes it is because some support or social work agencies feel protective towards their client, whether victim or offender, and worry that restorative justice would be too much pressure on them. Sometimes it is a concern that providing details may infringe data protection legislation, even if the person concerned has agreed. Certainly, this has been a problem both in Scotland (Maglione et al., 2020) and England and Wales, with a need for central guidance in England and Wales that providing information to restorative justice providers is similar to providing information to victim support or penal supervision agencies.
  After initial information has been sought from the referrer, facilitators would proactively seek out information from agencies working with the potential participant, whether before approaching the participant, or more likely after an initial pre-meeting with the participant. Partnership working with other agencies was seen as crucial, both to support participants through the restorative justice process and to cover details for the risk and mitigation assessment. There were areas of potential risk where facilitators were conscious they needed expert guidance and information. These were particularly around mental health and substance abuse:

  If there’s an organisation working with them, I would get permission to contact the agency. What I really would want to check with the agency is are they supportive. Because if they are in contact with an agent in the agency, if they are in contact with therapists, in contact with whoever, it is very important that they are supportive of whatever is going to happen. (Denm1)

  Both contact with the agency and any information sharing needs to be with the consent of the potential participant. Restorative justice, respondents stressed, must be empowering to the individual, not a professional-led conversation about them which does not involve them. Similarly, the other agency’s view of risk was usually found to be informative, but could never be decisive in terms of the restorative justice process. In criminal cases, the participant’s view of both the offence and risks can be different to those of both official criminal justice personnel and advice agencies and professionals. Participants’ views of restorative justice depended upon the goals they were seeking from the restorative justice process, which might be very different from those which other agencies may presume. So, for example, the participant may be seeking answers to questions (such as ‘why did you pick on me?’, or ‘what did you do with the property stolen?’), rather than apologies or compensation.

  I have had cases where the therapist said, you’re not ready, but the victim said, I think I will never be ready but I have to do this and I [facilitator] went on doing it. Because, for me, the most important is the person themselves, and then it’s that person that will decide, no matter what their lawyer or therapist advises them. I take that advice very seriously, I will discuss the advice with them, but if they then decide, I’ll go on with it. (Belg1)

  At initial meetings with a potential participant, it is important, our respondents said, to find out what the participant thinks might be a risk and what goals they have, and then to discuss how they might be dealt with through possible restorative processes. Restorative justice, by its nature, is a co-produced process, they stressed, not a top-down instructional one:

  But I think to me it’s really, really important that they define the aim…. And actually when I do a victim and offender mediation, that’s actually the first question I will ask them, okay, so what do you hope to get out of this meeting? (Denm2)

  It is the role of the facilitator to stop the process if the risks of continuing are, in the facilitator’s opinion, too great and cannot be mitigated or another process used. But otherwise it is the role of the facilitator, respondents thought, to facilitate safe restorative justice.
  Gathering information to assess risk and potential mitigatory strategies ran alongside talking with the potential participant about the process and preparing them for what it may involve. All restorative justice must be voluntarily undertaken, with the potential participant able to withdraw at any time. This means they must have sufficient information to be able to judge and make decisions as to what they themselves want from the process, what will happen, the emotions they may feel and potential outcomes. The role of the preparatory phase is to forestall and mitigate risk that might arise through communication or in a meeting.

  We’ll be looking at things around any additional needs, were there any mental health concerns, any physical health concerns or any extra support that either participant needs in order to have those conversations with us. And do they need somebody in as a supporter to help them manage, you know, those needs or can we manage them ourselves. (E&W2)

  Risk assessment in restorative justice is therefore individualised risk assessment, ideally conducted with the potential participant and working with any relevant support agencies (with the consent of the participant). All of this means it cannot be actuarial risk assessment, whereby risks are calculated on the basis of an individual’s membership of a group or in relation to a particular type of offence. It is possible to say that some types of offence (for example, which are particularly violent or involve previous relationships) or some elements of a person’s health or life (for example, communication difficulties, mental health vulnerability) will tend to make risks higher. But because it is about what that particular person sees as risk and their own social context, it is not possible to predict or standardise risk calculations through the judgment of an external professional, even a facilitator, in the same way as the likelihood of penal outcomes is measured.

  I think the importance is to take the lead from the participants. It’s what do they want and how can we make that happen in a way that’s safe for everybody. (Scot4)

  Some of our respondents used various types of checklists or forms to note down their risk assessment and mitigation findings in individual cases – two respondents from Scotland, four respondents from England and Wales, one respondent from Northern Ireland, one respondent from Finland and one respondent from Austria. None, however, used an actuarial risk assessment instrument. We noted that no respondent spoke about a formal, validated risk assessment, where a tool has been compared in previous research with what then happened during the restorative justice process. This kind of validation of risk measurement tools does not seem to exist. This is in line with the stress in assessment of the individual circumstances of the case and the skills of practitioners that we cite above.
  Interviewees said, however, that it is important that risk and mitigation assessments are recorded and that early risk and mitigation assessments are updated if circumstances change. This is a part of professional accountability and reflective practice and is vital if, for example, there has to be a change in facilitator.

• 4 Risks and mitigation in more complex cases

  The question then arises as to what kinds of circumstances would (and had) prompted respondents to conclude from the risk assessment that it was not possible to provide restorative justice safely and so they had stopped the process. This is linked to whether respondents had any category of cases, for example in terms of type of offence, which they regarded as particularly difficult or risky in relation to restorative justice. Of the countries in which our respondents had practiced, only England and Wales and Northern Ireland have such a category set out in guidance. For England and Wales, the category is that of ‘sensitive and complex cases’ in which domestic violence/abuse, sexual assaults and homicides fall by nature of the kind of offence, but in which other cases may be put because of their own particular circumstances (Restorative Justice Council, 2011). These would include especially vulnerable participants, manipulative behaviour by participants and there being several perpetrators or persons harmed. In Northern Ireland, in youth ‘indictable only cases’ (which include murder, manslaughter, rape and robbery), judges have the discretion whether or not to refer to restorative justice before sentence, whilst other types of offences must be so referred, all providing that the offender consents to the referral.
  Respondents from other countries certainly acknowledged cases with these features as complex, often requiring more preparatory work and potentially the use of a more experienced facilitator. However, they did not categorise them as qualitatively different from other restorative justice cases. Respondents from England and Wales would also not see the type of offence as determining necessarily a more complex case, though some providers had rules about whether they would accept offender self-referrals in cases of, for example, domestic abuse or other gender-based violence.
  Instead, respondents said that they would proceed on a case-by-case basis, considering possible harm to each potential participant, rather than assuming that certain kinds of offences always produced more likelihood of harm:

  Increasingly, I’m finding that they’re all serious and complex. I think, when I was young and naïve, there was a kind of hierarchy of, you know, down here you’ve got shop theft, someone’s handbag’s getting snatched, then you’ve got burglary, robbery, and you’d kind of build up to things that involve death, sex offences, domestic violence. And the more I work in restorative justice, the more I realise that, although that’s the way that it’s looked at in the criminal justice system, you can’t anticipate harm. (E&W1)

  There’s not a case that can’t be mediated or restorative justice process, but there are people who are not suitable for the process…. Domestic violence, sexual violence, always large groups, young people, maybe if somebody has mental health issues or some communicative disorder, you have to be careful. (Fin2)

  Key elements that would prompt facilitators to look more closely at potential harm which might be caused by restorative justice were very similar between facilitators from different providers and countries. They were prior relationships, dominance in the offence from one person, power differentials and major effects on the person harmed. These were not the same as the criminal justice distinctions between offences of different seriousness, though clearly there is a correlation between seriousness and harm caused. Cases where harm was potentially more likely would be given to more experienced facilitators, or to paid staff rather than volunteers. Though the restorative justice process might involve more preparation of participants or more consultation with other agencies specialising in supporting those who have been harmed in such cases, the process would not be significantly different in more complex cases or involve different values:

  I don’t know that it’s a different approach … I think it’s the level of intensity in terms of the checking and the re-checking, and the level of … I want to use the word supervision. (RoI2)

  Restorative justice measures in France depend on participants’ expectations (meeting several offenders and victims, meeting their offender/victim, being supported in a circle, etc.), but they do not depend on complexity or seriousness. It is in the preparation that we consider these aspects. (France1)

  Interestingly, those who had not done such complex cases were far more likely to be diffident about them, for example, saying that they required more specialised training (for a discussion on specialist training, see Keenan, 2018). As in previous research (Shapland et al, 2011), experience in facilitation had led to facilitators being prepared to take on (and be successful with) more difficult or complex cases. This included cases of historical abuse and domestic violence, where several facilitators said that victims had been rebuffed from restorative justice in the past as not being ‘suitable’:

  They go to people they think can help … And they’re told no, for your offence that’s impossible, you know, it’s often domestic violence or sexual crime … it tends to be framed as what is the risk of doing this. And what I’ve certainly come to the conclusion is, having talked to victims, and particularly victims who really did want to have a restorative process. What are the risks of not doing it, even if it is risky, do you know what I mean, to put it that way? (NI1)

  A key message for risk assessment in restorative justice is then that the risk to be assessed is not just the risk of proceeding with restorative justice, but the risk of not doing it, which then leaves potential participants with their questions and feelings not addressed.
  Are there then circumstances in which facilitators would not proceed with restorative justice, because it would be too risky? The answer is definitely yes, but that such cases are rare. Facilitators said that any such decisions should be discussed with participants already approached, and it was important not to presume what participants would want from the process:

  And you never know what kind of needs people have. And that’s things that we have to sort out. Can your needs be taken? Can [he or she] answer on your needs? And the goal is never that people are going to be the best friends forever. The goal is that people get the chance to say what they have to say, ask what they need to ask and may be able to go further in life. (Nor2)

  Circumstances where the facilitator would have to take a decision that restorative justice could not proceed related strongly to the values of restorative justice which appear in the international instruments (as discussed above) and which have been set out in both academic work and guidance (see, for example, Braithwaite, 2003; Scottish Government, 2017). They include that the key participants voluntarily wish to participate in restorative justice; the possibility of facilitative communication between the parties; that the offender has admitted responsibility for the offence;⁹x This is distinct from an admission of legal guilt. and that participants are not threatening each other or trying to coerce the other for another purpose. Threat and coercion negate the voluntariness of the process. Denial of responsibility (or that it is the wrong offender or victim) prevents communication, as does one key participant not being able to communicate at that time (because of mental health difficulties, language difficulties or substance misuse – though a restorative process may be possible at a later date). Essentially, we found that too risky a process is a process where the core values of restorative justice cannot be met. In other cases, as we shall see below, risks can often be mitigated so that it is possible to proceed, even if the type of process involved has to change.

• 5 Specific risks

  We now turn to consider specific risks which may be present in individual cases and the possibility of mitigating those risks. The universe of possible risks is very large; so, given the space constraints, we can only summarise potential mitigation here.¹⁰x A more complete discussion can be found in Shapland et al. (2022). We start by considering risks related to participants, before turning to risks concerned with different kinds of restorative justice processes (meetings, virtual communication and larger meetings). In relation to all of these risks, it is important to consider potential means of mitigation as well as the risk itself – and to involve the potential participant:

  We talk a lot about restorative justice, putting the participants at the heart of what we do. And then I think sometimes we fall into the trap of making those decisions without even telling them. So we will always try, I will always encourage our participants to say, if you’ve got a risk that needs managing, get your participants involved in that … And just being quite open and transparent with participants about does this work for you? (E&W2)

  5.1 Risks related to communication

  Communication between participants is at the heart of restorative justice, whether that be face-to-face communication or some form of indirect communication. Respondents said that it is often possible for facilitators to be creative to mitigate any communication difficulties, whether those be because of language difficulties, difficulties in expressing things in words, mental health problems or having to go through intermediaries (see also Bolitho, 2019, who comments in relation to an Australian sample with disability that there was often a need for adjustments (equivalent to our concept of mitigating risk) in relation to the preparation and process of restorative justice, particularly around acceptable behaviour and comprehensively preparing the participants on the process). The initial point is to check the extent of any problems, rather than assume they are major. If there are, for example, significant mental health issues, they may diminish over time or with treatment, so restorative justice processes may need to be paused rather than abandoned. Where there has been a prior relationship, it may be necessary to break or change past communication habits:

  Often in domestic violence … when we dive into it, it has been two monologues. And now the offer is a dialogue. So you have to break some patterns, how they have been communicating before. They have to communicate in another way when they are meeting in our meetings. (Nor2)

  It was common for our respondents, as experienced facilitators, to have used interpreters where participants spoke different languages or where English was not the first language. The facilitator as well as the participants need to be able to understand. Some providers were able to use their own staff as interpreters and it was useful if the interpreter was familiar with restorative justice processes, because they would then be used to the confidentiality necessary and the kinds of meetings, as well as to the potential for emotion in meetings. It was necessary to have a contract for interpreters which stressed confidentiality and the ground rules for when people would speak, as well as safety precautions. Particularly where restorative justice was taking place in small communities or in relation to offences where there had been considerable publicity, rules about not gossiping are key. If participants are nervous about others finding out what happened, another mitigatory measure is to assure them that they themselves are in charge of how much detail to give – this is not like a police interview:

  Also the agreements in these meetings that everyone can say how much they want and you … don’t have to be sharing your story, like, today, then you can say as much as you want, and that kind of thing. (Est1)

  Thinking about how communication can best take place needed to be undertaken not just in relation to face-to-face meetings, but also in terms of any information that is sent out to participants. Indeed some respondents had used images, art and crafts as instruments of explanation and communication between participants. Restorative justice is not only about speech. Prioritising non-verbal communication can be really useful if a participant is silent, or does not seem to want to talk. One facilitator had started using pre-printed cards, which could be pointed to in preparatory meetings and restorative meetings, with some young people:

  So you would get like a pile of cards pre-printed, and then you would give the young person the cards or you would put them out on the floor, and you would say, right, have a look over those questions. If there’s one that you want the answer to, pick it up and we’ll talk about it. So I’d be like, you don’t have to speak to me, you don’t have to … just pick up the question, I’ll read the question and I’ll give you the answer. (Scot8)

  It should be emphasised though that difficulties with speaking are not confined to young people:

  I think that the majority of my clients have communication difficulties, which are not all, but I’d say the majority because of all the things that I’ve said previously about adverse childhood experience, current or previous trauma, which makes it really hard to voice and articulate things, which is another reason that I find the arts making really helpful. (Scot2)

  Although questions of identity and power were not the focus of our study, Willis (2020) highlights that the power of participants to communicate in the restorative justice process may be very unevenly distributed between participants of different social backgrounds.
  Another situation which can involve difficulties in communication, particularly in the initial phase, is where it is necessary for the facilitator to go through an intermediary to get information to a potential participant. This was most likely to occur when the participant was involved in the criminal justice system, for example, an offender in prison or on probation.

  When you’re dealing with prisons, often what you can find yourself doing, is saying, ‘I need to speak to this person, and I want to ask them about this thing’, and you’ll have a middle party like a probation officer, or a prison officer, going, ‘we’ll pass that message on for you’. And you think, ‘oh, what have you passed on?’ … One of the ways that we have tried to mitigate that, and with various degrees of success is by being extra specially helpful and saying to those professionals, ‘what I can do is, do you want me to put all of this down in a letter for you?’ And then you can give them the letter and feedback what they say. (E&W1)

  Overall, communication difficulties were found to be common. Facilitators emphasised:

  • being alive to the possibility of communication difficulties, not just in obvious cases where participants speak different languages, but also in terms of the confidence the participants will have in expressing themselves

  • discussing potential difficulties with the participants before any meeting or starting indirect communication, so that they are aware of the situation

  • where there are particular mental health difficulties, or communication difficulties which can be alleviated using professional support, being in touch with the professional supporting that participant (with the participant’s agreement) before deciding what to do

  • being prepared to be creative in surmounting the difficulties by using different means of communication or ensuring interpreters or intermediaries do not confuse the message.

  5.2 Risks related to relationships

  We have considered above the question of whether there are specific risks where the key participants have had a previous relationship. Respondents indicated that this was always a flag for finding out more in preparation. It is important to extend this to all potential participants, not just the victim and offender. Indeed, respondents said the general rule was:

  It’s really important that no-one comes to a meeting that you don’t know, that you always meet everybody prior to that meeting … I’ve said, if you want that person to come then we need to postpone this restorative meeting because I haven’t got time to assess that person’s suitability for this meeting to take place. (E&W3)

  The key point was whether others might act inappropriately, particularly if they felt protective towards a key participant (whether they are relatives or support workers) and would not let them speak for themselves. Particularly, other professionals who often act as advocates should not speak instead of the victim or offender:

  You also have people that can advocate on behalf of people, social workers, et cetera. They’re allowed to come in, that’s fine, but they cannot speak for the person unless they’re asked to … So they can’t interject and just start speaking, they have to be prompted by that person to speak. (Scot7)

  Other participants can, however, be a benefit in terms of support, as well as potential risks. So the main thing, respondents said, was to deal with this in preparation, asking the key participants who they would like to be there to support them. If the key person did not have their own contact, then someone from a local agency specialising in that kind of offence might be helpful.

  5.3 Risks related to the participation of young or vulnerable people

  As we have seen, vulnerability in terms of restorative justice processes is not confined to young people and indeed adults are often vulnerable. However, where restorative justice is linked to criminal offences or to criminal justice processes, it can be that the restorative justice process done with young participants needs to follow the general rules that a parent or guardian or another responsible adult must be present (jurisdictions vary considerably as to the ages for which this is required) or, as in Denmark, that the legal guardian needs to give their permission before the restorative justice can start. Similarly, if restorative justice is taking place as diversion, but where the case is still subject to potential prosecution, some jurisdictions, for example, in Northern Ireland, prescribe that a lawyer may attend, to provide legal advice as necessary to the person responsible for the harm (though it is very rare that lawyers actually attend restorative justice meetings). These are safeguards against young people being dominated or adversely affected afterwards by what may be said in a restorative justice process. Though normally the facilitator has control over who is invited to attend a restorative meeting or take part in restorative communication, they do of course have to obey any relevant legislation.
  For the facilitators interviewed, the important element was that if such a person attended, they would be supportive to the young person/vulnerable person. Working with young people and with any other vulnerable person was not qualitatively different from working on other cases. Specific aspects to concentrate upon are modes of communication and understanding, where it is levels of emotional and intellectual maturity which are relevant, not necessarily age per se, and ensuring that the participant feels able to talk, without feeling overawed or dominated by the situation (see Daly, 2003; Campbell, Devlin, O’Mahony, Doak, Jackson, Corrigan & McEvoy, 2005). Another is to ensure that ground rules about, for example, not having mobile phones on or gossiping are clear and, where necessary, checked.
  Young people may participate not only as a direct victim or offender, but as having been affected by the offence, for example, as members of a burgled household or a witness. Facilitators recalled instances where they were in two minds as to whether a child should come to a restorative justice meeting:

  In [one case] the parents and a 12-year-old daughter in the house at the time, was burgled during the night. And the 12-year-old girl, who has Asperger’s, wanted to come along … Anyway, she was absolutely determined and the parents wanted her to come along…. She was brilliant. Because she asked the questions, the direct questions that other people were sort of hedging around and she just went straight in for it and she was amazing. And that was a real eye-opener for me that, you know, given the right support and the risk assessments and all the rest of it, that you can have children. (E&W4)

  This example also touches on the above-noted complexities around communication in restorative justice; neurodiversity and childhood may be stereotyped as barriers to communication, but this example presents them as positive assets for the restorative justice process.
  In general, facilitators had found that, providing there was sufficient preparation and risk assessments to enable age-appropriate communication, children could be accommodated in a restorative justice setting.

  5.4 Risks related to place

  Restorative justice face-to-face meetings clearly need to take place in a suitable site, but there are risks related to place for both preparatory meetings and virtual communication. The choice of place should reflect, respondents said, both practical matters, such as access, size, cost and safety, and the emotional needs of the participants, including the associations of the type of place and if they have encountered particular places previously.
  Ideally a place for a meeting should be neutral to all participants, comfortable, have refreshment and toilet facilities and, for a face-to-face conference, allow for the use of a breakout space. Providers had taken great care to furnish bespoke places for meetings when possible to provide a calming atmosphere. Facilitators also thought about how participants would enter the place and, particularly, how they would exit it, to reduce the possibilities of confrontation. Personnel needed to be assigned to escort participants, particularly in criminal justice facilities (such as prisons or police stations). Items within the space need to be assessed as to whether they pose particular risks (being able to be thrown, etc.), though no facilitator we interviewed had witnessed any actual aggression between the key participants, and only one described instances of physical aggression (directed at the facilitator). Apart from bespoke providers’ spaces, a safe local community space was important so that the space was not where one participant might attend for other purposes (for example, an office used for community sentences) or where participants worked or lived. It was also important for accessibility (including ensuring accessibility if there are physical disabilities), and transport might need to be provided. In meetings, considering where people would sit was also significant:

  So, you put the potentially aggressive person maybe in the back so that everybody else can get out of the room easily, you don’t leave any heavy glassware on the table, you use little plastic cups for your drink of water or whatever you serve, coffee. (Fin2)

  Safety related to place is relevant for the facilitator as well as participants. Facilitators can be lone workers; so, they need to take precautions for safety for lone workers (consulting systems about any ‘red flags’; leaving a record of where they will be; contacting a colleague after the meeting to let them know they are OK, etc.). This obviously applies to preliminary meetings and feedback meetings, as well as to any restorative justice face-to-face meeting.
  As well as being as safe as possible, spaces for restorative justice need to feel safe to participants. For this, one of the most important elements was having one or more additional spaces, as well as the main room, in which participants could prepare, debrief and discuss afterwards, and go to if it was necessary to take a break if the meeting became emotionally charged:

  Time off is good. Have a break, go out, separate rooms, let’s talk in separate rooms and then bring together again. (Denm1)

  Doing restorative justice in a prison setting is obviously more challenging in relation to the space and feeling safe. The strict security procedures can be very intimidating to those not familiar with them – and prisons are hardly neutral environments. Nonetheless, some mitigatory measures were described: such as using less ‘prison-like’ rooms (with anterooms); facilitating access through meeting participants in the car park; helping with transport (many prisons are in remote locations) and working with the victim during the pre-meeting process to familiarise them and visit the prison.

  5.5 Risks in meetings, including larger meetings

  Potential participants can have concerns about how other people may see them during a restorative justice meeting. Risks during a meeting are probably the most immediate risks that facilitators are aware of. Respondents, however, considered that most of the risks could be mitigated during the preparation phase, by discussing this with potential participants. Mitigation measures include having ground rules about voice, possible aggression in meetings, cases with possible power imbalances, where there may be gossip, and considering the safety of the facilitator as well as the participants. In these discussions, there can be a fine line between raising possible concerns in preparation and over-rehearsing participants, but it was considered far better to raise potential concerns and see whether they were shared by participants.
  Restorative justice requires all participants to have ‘voice’, i.e. a fair share of the time, so that they can put their questions and their point of view. Some respondents hence always used ‘ground rules’, including confidentiality and taking turns, along with breaks, tea and coffee, and use of breakout rooms, as well as reiterating that the meeting will stop if someone says they do not wish to continue.

  Sometimes, when we know that we’re going to have someone who is problematic, we talk very specifically about, ‘we think that this will happen, how do you want us to deal with that?’, and make that very open so that we are saying to … our difficult person, this is a problematic behaviour, you may not intend it to be a problematic behaviour, but it is, and if that happens, we will do a, b and c … One of my first conferences, and it was very good self-awareness on the part of the offender, it was a shop-lifting case, and the offender said, I’m nervous, I’m really nervous about this, but I also know that whenever I’m nervous I laugh, so please could you say that to my victim in advance, because I don’t want them to think that I’m taking the Mick [laughing at the person or situation], I’m not, I’m taking this very seriously. (E&W1)

  The likelihood of aggression, respondents said, is very slight (as it was, for example, in the large sample of conferences in Shapland et al., 2011), but people can get worked up and ‘kick off’. Here breakout rooms were often used, whether the break was suggested by the participant or the facilitator. Power imbalances were a particular worry, but facilitators said that, apart from where there was a long-term relationship, they only rarely caused difficulties in meetings:

  There’s always a power imbalance, and it doesn’t matter whether it’s petty crime, severe crime, there’s always a power imbalance. If there wasn’t a power imbalance we wouldn’t be there. Somebody took power over the other one at some point…. I will say that there’s more talk about power imbalance than actual power imbalance when you get to the meeting … With domestic violence I’m far more cautious because I know there are things that I don’t know, I know there are things that I don’t see; I know that that’s where the power imbalance is so invisible. Whereas with the sexual violence or the severe cases the power imbalance is far more balanced actually. (Denm1)

  Facilitating larger meetings (say ten or more participants) was a much rarer experience. There were no major differences with having many participants there, but it was important that everyone relevant was invited, and of course then they all have to be prepared, be met before the meeting, potentially be escorted for breaks and so forth. Healing circles or large meetings often required more facilitators. Yet, such meetings could be easier to facilitate – as long as the participants agree on what had happened:

  So ideally when you invite more people into the room, in many cases, it actually makes it easier to facilitate the process, because the people in the room will take care of the process quite often. … I think actually talking about what has happened can quite often be risky in a way … especially if you’re sitting with people who have a long history, then they normally have completely different versions of what has actually happened…. So, I quite often have a dialogue with the parties before the meeting to actually hear whether they think it’s important to talk about what has happened during the meeting or whether they would rather focus on something else? Especially in difficult neighbour disputes and sometimes also in cases where there has been a history of domestic violence or something like that. (Denm2)

  Facilitators also have to be aware of risks occurring after any meeting, with the key elements here being discouraging any gossip and if participants are worried about future victimisation. Discouraging gossip can be done by putting a confidentiality clause in outcome agreements and getting participants to agree to this verbally, whereas what participants might do if they encountered each other sometime afterwards could be discussed and agreed at the meeting, and this was particularly important in smaller or more rural communities. Respondents said it was vital to contact participants a day, a few days or at most a week after a meeting or the end of the process. For participants in custody, a shorter time was necessary. This was just to check in with them in relation to their well-being and any concerns about safety, as well as to gain their immediate impressions of the process. Contacting them again a few weeks later was also recommended, this time to reflect on how things had gone and offer feedback to the provider generally.
  Some providers always worked with two facilitators, but many did not, and relied more on the experience of certain of their facilitators: it seemed to be a cultural practice in terms of models of restorative justice. Equally, providing information to participants on the implementation of any outcome agreement was an important part of preventing any reoccurrence of difficulties.

  5.6 Risks for virtual processes

  Despite the experience of working through the Covid-19 pandemic, respondents had only occasionally used virtual communication in restorative justice rather than face-to-face preliminary meetings or conferences. Virtual meetings seemed mainly to have been used where one participant lived at a distance.
  The risks in relation to virtual meetings (video conferences) were primarily how facilitators could guarantee who else might be listening in at the participant’s end, because virtual meetings normally only show a segment of the room. The same problem occurs if it is a virtual meeting with a lawyer at a prison, etc. In more remote areas, a facilitator might be present with each participant, each located in a different town.

• 6 Discussion and conclusions: risk, mitigation, culture and values

  The most important message coming from experienced facilitators in relation to preparation for restorative justice is that risk assessment for restorative justice is about assessing risks of engaging in a restorative justice process, and the extent to which such risks could be minimised. It is not about evaluating potential participants’ ability to be treated, or the likelihood of committing further offences, or remorse, or the likelihood of receiving compensation, or any other outcome measure after the restorative justice process. Hence, criminal justice assessment instruments and actuarial measures are not relevant. This can be difficult for referrers and criminal justice personnel to appreciate, in what is currently a culture of criminal justice risk assessment.
  There are limitations in relation to the research. Due to the relative paucity of evaluations of the detailed processes of undertaking restorative justice, we are not able to judge or comment upon whether the methods being used by these experienced facilitators are the best way to measure risk or whether the risks identified were ‘real’ or substantial or whether the mitigatory techniques used succeeded completely or in the long term. We did, however, ask respondents to reflect on what they had found helpful and what seemed to lead to a safe process.
  Restorative justice is essentially a bottom-up process, where the communication is done by, and any outcome agreement created by, the participants. Risk assessment for that process is therefore about assessment in relation to each participant’s own goals in taking part and the risks they perceive, as judged by the facilitator. So communication with each participant is key and assessment needs to have a co-production element.
  We found that the main threats to safe participation in fact relate to the core values of restorative justice: communication and voice, the offender’s acceptance that they have caused harm, and wishing to participate in a safe, neutral, non-coercive process. Even in more difficult or complex cases, many elements of risk which might occur in a process respondents said could be mitigated, but, as we have discussed above, there are many different aspects to consider during preparation. Facilitators relating their experiences to us, whichever restorative justice process they were used to using and whichever offences they had worked with, saw no qualitative difference in the basics of the risk and mitigation assessment process they adopted. There were some cases too challenging to proceed with at the time they received them, but the problems did not necessarily reflect the type of offence or any categorisation based on one specific factor. Experiences of risk and mitigation were remarkably similar across jurisdictions – which implies that future research could profitably build on this study to explore aspects of risk and mitigation in more detail, using comparative methods.

References

• Andrews, D.A., Bonta, J. & Wormith, J.S. (2016). The recent past and near future of risk and/or need assessment. Crime & Delinquency, 52(1): 7-27. doi: 10.1177/0011128705281756.

• Arber, S. (1993). Designing samples. In N. Gilbert (ed.), Researching social life (pp. 68-92). London: Sage.

• Banwell-Moore, R (2022). Just an ‘optional extra’ in the ‘victim toolkit’?: the culture, mechanisms and approaches of criminal justice organisations delivering restorative justice in England and Wales. International Review of Victimology, 29(2), 217-235. doi: 10.1177/02697580221079993.

• Beck, U. (2004). Risk society: towards a new modernity. London: Sage.

• Bolitho, J (2019). Complex cases of restorative justice after serious crime: enabling spaces for those with disability. In T. Gavrielides (ed.), Routledge international handbook of restorative justice(pp. 159-176). London: Routledge.

• Braithwaite, J. (2003). Principles of restorative justice. In A. von Hirsch, J. Roberts, A. Bottoms, K. Roach & M. Schiff (eds.), Restorative justice and criminal justice: competing or reconcilable paradigms? (pp. 1-20). Oxford: Hart.

• Campbell, C., Devlin, R., O’Mahony, D., Doak, J., Jackson, J., Corrigan, T. & McEvoy, K. (2005). Evaluation of the Northern Ireland youth conference service. NIO Research and Statistical Service Report No. 12. Belfast: Northern Ireland Office. Retrieved from https://www.restorativejustice.org.uk/resources/evaluation-northern-ireland-youth-conference-service (last accessed 2 August 2022).

• Council of Europe (2018). Recommendation CM/Rec(2018)8 of the Committee of Ministers to member States concerning restorative justice in criminal matters. Strasbourg: Council of Europe. Retrieved from https://rm.coe.int/09000016808e35f3 (last accessed 2 August 2022).

• Daly, K. (2003). Mind the gap: restorative justice in theory and practice. In A. von Hirsch, J. Roberts, A.E. Bottoms, K. Roach & M. Schiff (eds.), Restorative justice and criminal justice: competing or reconcilable paradigms? (pp. 219-236). Oxford: Hart Publishing.

• European Union (2012). Directive 2012/29/EU of the European Parliament and of the Council of 25 October 2012 establishing minimum standards on the rights, support and protection of victims of crime, and replacing Council Framework Decision 2001/220/JHA. Brussels: EU. Retrieved from https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32012L0029 (last accessed 2 August 2022).

• Health and Safety Executive (2022). Managing risks and risk assessment at work. London: HSE. Retrieved from https://www.hse.gov.uk/simple-health-safety/risk/steps-needed-to-manage-risk.htm (last accessed 2 August 2022).

• Keenan, M. (2018). Training for restorative justice work in cases of sexual violence. The International Journal of Restorative Justice, 1(2), 291-302. doi: 10.5553/ijrj/258908912018001002007.

• Keenan, M. & Zinsstag, E. (2022). Sexual violence and restorative justice: addressing the justice gap. Oxford: Oxford University Press.

• Maglione, G., Buchan, J. & Robertson, L. (2020) The local provision of restorative justice in Scotland: an exploratory empirical study. European Journal on Criminal Policy and Research. DOI: https://doi.org/10.1007/s10610-020-09470-3

• Restorative Justice Council (2011). Best practice guidance for restorative justice. London: Restorative Justice Council. Retrieved from https://restorativejustice.org.uk/sites/default/files/resources/files/Best%20practice%20guidance%20for%20restorative%20practice%202011.pdf (last accessed 2 August 2022).

• Scottish Government (2017). Delivery of restorative justice in Scotland: guidance. Edinburgh: Scottish Government. Retrieved from https://www.gov.scot/publications/guidance-delivery-restorative-justice-scotland/pages/2/ (last accessed 2 August 2022)

• Scottish Government (2018). Delivering for today, investing for tomorrow: the Government’s programme for Scotland 2018-19. Edinburgh: Scottish Government. Retrieved from https://beta.gov.scot/publications/delivering-today-investing-tomorrow-governments-programme-scotland-2018-19/ (last accessed 2 August 2022).

• Shapland, J. (2009) Restorative justice conferencing in the context of community policing. Journal of Police Studies, 2009-2: 119-138.

• Shapland, J. (2013). Comparing conferencing and mediation: some evaluation results internationally. In E. Zinsstag and I. Vanfraechem (eds.) Conferencing and restorative justice: international practices and perspectives (pp. 47-62).Oxford: Oxford University Press.

• Shapland, J., Buchan, J., Kirkwood, S. & Zinsstag, E. (2022). Mitigation and risk in restorative justice. Report to the Scottish Government. Retrieved from https://www.rjforum.scot/files/mitigation-risk-report-final-version-to-scottish-government.pdf (last accessed 2 August 2022).

• Shapland, J., Robinson, G. & Sorsby, A. (2011). Restorative justice in practice. London: Routledge.

• Shapland, J. & Hall, M. (2007). What do we know about the effects of crime on victims? International Review of Victimology, 13: 175-218.

• Swida, A. (2020). A service response to the pandemic: Open circle and restorative justice during Covid-19. Australian Social Work, 73(4), 508-509. doi: 10.1080/0312407X.2020.1801123.

• Van Camp, T. & Wemmers, J.-A. (2016). Victims’ reflections on the protective and proactive approaches to the offer of restorative justice: the importance of information. Canadian Journal of Criminology and Criminal Justice, 58(3), 415-442. doi: 10.3138/cjccj.2015.E03.

• Willis, R. (2020). ‘Let’s talk about it’: why social class matters to restorative justice. Criminology & Criminal Justice, 20(2): 187-206. doi: 10.1177/1748895818804307.

• Zinsstag, E. & Vanfraechem, I. (2013). Conferencing and restorative justice: international practices and perspectives. Oxford: Oxford University Press.

Noten

• 1 A literature search on ‘restorative justice risk assessment’ produced only publications looking at the risk of reoffending if restorative justice was undertaken, not on the risk of undertaking a restorative justice process. One exception is a brief report by Swida (2020), on adjusting risk assessment in response to the particular risk demands of Covid-19. There was also some work on using restorative justice with those with disability (Bolitho, 2019).

• 2 The terminology used for potential participants in restorative justice currently varies between countries. The international instruments relating to restorative justice refer to ‘victims’ and ‘offenders’ in terms of key participants. In Scotland and some other countries, victims are called those who have been harmed, whereas offenders are called those who have harmed. In some contexts, particularly that of gender-related violence, victims are often called survivors. What participants may call themselves also varies. In this article, we use the international terminology of offenders and victims, because we are writing for an international audience, but we are very aware that each term has potential characteristics or stereotypes associated with it.

• 3 There is of course a large research literature on restorative justice outcomes.

• 4 Criminal law and criminal justice are devolved matters in the United Kingdom, so provisions in Scotland are different from those in England and Wales and in Northern Ireland.

• 5 In Scotland, local authority ‘Justice Social Work’ performs a role similar to probation services elsewhere.

• 6 We searched the academic literature using the keywords [risk – restorative justice – mitigation]. Very few articles were found, and those which were primarily focused on risk of reoffending, harm minimisation for young people or the generic needs of victims, i.e. outcomes, rather than the risk of undertaking a restorative justice process. References to ‘mitigation’ were about mitigation in sentencing after restorative justice, not mitigating risks during restorative justice processes. This is congruent with comments critiquing the paucity of literature on exactly how restorative justice is done (Shapland, 2013).

• 7 Note that the age of criminal responsibility and the age at which offenders were typically tried in adult courts varied between jurisdictions, so the term ‘children’ or ‘young people’ refers to those who were above the age of criminal responsibility but under the age of offenders normally tried in adult courts for that jurisdiction, but there was no uniformity in using the term ‘young person’ or ‘child’ in chronological age across jurisdictions.

• 8 The management of cases and facilitators is not covered in this article. See Shapland, Buchan, Kirkwood & Zinsstag (2022).

• 9 This is distinct from an admission of legal guilt.

• 10 A more complete discussion can be found in Shapland et al. (2022).