1 Introduction

Reporting crime to the police normally results in considering a criminal justice trajectory, but, at the same time, this trajectory is regarded as an ultimum remedium, a last resort, when all other options have failed. Overuse of criminal law enforcement suits the interests of neither victims nor offenders. It is here that restorative justice, as an optimum remedium, comes in (Johnstone & Van Ness, 2007). Restorative justice can be depicted as an alternative to criminal justice, albeit it does not automatically signify a substitute. Rather, it seems integral to existing law enforcement and prosecution practices. The goal of restorative justice interventions is to connect victims and offenders (or suspects) in affirmative ways, within the scope of law enforcement and judicial response if necessary.
Under Article 51h of the Dutch Code of Penal Procedure (Wetboek van Strafvordering), the public prosecution office promotes the principle of the police informing both victims and suspects/offenders about the possibilities of restorative justice solutions, including mediation, at the earliest possible stage. However, actually making people aware of mediation and its advantages is still uncommon. Restorative justice remains a highly underdeveloped theme within policing (Clamp & Paterson, 2017), and empirical studies on the practices of conflict (i.e. victim-offender) mediation are relatively scarce (Choi & Gilbert, 2010; Hansen & Umbreit, 2018). Therefore, our article explores what happens when conflict mediation is put into practice in the ‘police phase’ (Wolthuis, Claessen, Slump & Van Hoek, 2019) of the Dutch criminal justice system.
We carried out a unique ‘action research’ (Coghlan & Brannick, 2014) in Amsterdam, the Netherlands, to both change and study police practice. This combination of conducting empirical research and taking action at the same time was necessary to create – previously non-existent – conflict mediations in a police context. A sworn-in professional and experienced non-police mediator facilitated the dialogues between conflicting parties during the research process. Within the action research framework, the objective of our article is to report about observations on conflict mediation in a police context by asking two questions: (a) what do suitable cases for conflict mediation in practice look like? and (b) how can these mediations be evaluated set against core principles of restorative justice?
This article is structured as follows: Section 2 gives a background to the use of restorative justice interventions in policing, offers a definition of such interventions and sketches a framework for analysis. In addition, Section 3 covers the methodology of our study, Section 4 illustrates empirical results and Section 5 offers reflections. The article ends with a conclusion and discussion.

2 Restorative justice in policing

2.1 Background

The birth of modern restorative policing initiatives can be traced to the Australian, indigenously inspired, Wagga Wagga model and the Bethlehem Pennsylvania Police Family Group Conferencing Project developed over the 1980s and 1990s (Clamp & Paterson, 2017; McCold & Wachtel, 1998; O’Connell, 1993). These models combine elements of family group conferencing with the concept of ‘reintegrative shaming’. Different from stigmatisation and exclusion, reintegrative shaming contains a disapproval of the offender’s deviant action, while not labelling the person as inherently evil. Relationships of respect and trust must stay intact. In addition, the police or another organisation convenes a ceremonial meeting involving the victim, offender and their supporters such as family members. All parties can express their feelings about the offence and its consequences, after which they try to develop a settlement or come to an agreement. The hypothesis behind family group conferencing is that it improves compliance with rules and laws and reduces re-offending (Makkai & Braithwaite, 1994). Since then, restorative policing has travelled in different directions, among which can be counted mediation, youth courts and alternative caution schemes throughout the world.
Conceptually, restorative policing, and more broadly, restorative approaches within policing, can be naturally placed at the crossroad of ‘community policing’ and ‘problem-oriented policing’ (Clamp & Paterson, 2017). Community policing refers to a decentralised force devolving authority and responsibility to small geographical units or neighbourhoods so as to establish contacts with citizens and develop relationships with professional partners (Skogan, 2006). The problem-oriented model emphasises the need to identify and solve the root causes of crime in the long term, instead of merely carrying out reactive police responses (Goldstein, 1979). Restorative policing shares similar ambitions in terms of preventing troublesome conduct, introducing resolutions that prevent conflict from escalating and reaching out to local civil society networks beyond the police.
A common thread through restorative justice studies is that interventions, one way or the other, are likely to decrease recidivism and increase people’s well-being (Hansen & Umbreit, 2018; Menkel-Meadow, 2007). Community police officers appear most likely to participate in restorative conflict resolutions and build relationships with victims, offenders and those who care about them (Glowatski, Jones & Carleton, 2017). These officers often do so quite informally and spontaneously (Van Steden, Miltenburg & Boutellier, 2014), but restorative interventions usually call for more structured and long-term efforts (Clamp & Paterson, 2017). Although such interventions hold promise, their practical translation into police work is something different. Despite reasonable scientific evidence that police can benefit from conflict mediation, family group conferences and peacemaking circles, restorative ways of thinking create a number of tensions among police forces around the world. Not everyone automatically shows ‘readiness’ (Mohammad & Gearhart, 2021) when it comes to alternative conflict resolution.
In an overview of international research, Vanfraechem evaluates restorative policing practices against the background of restorative justice principles and concludes that some issues arise ‘such as a lack of neutrality and the impossibility of guaranteeing confidentiality’ (2009: 39). As for the Netherlands, Van Stokkom and Gunther Moor (2009) argue that community police officers tend to stay away from institutionalised forms of alternative dispute settlements and prefer instead to refer citizens to external specialists such as neighbourhood mediators. Restorative policing, in the officers’ view, does not fit the force’s so-called ‘core tasks’ of combatting crime. Things might have changed a bit, but ‘restorative justice is still relatively unknown within the [Dutch] police and is only employed incidentally’ (Hoekstra, 2021: 93). Indeed, restorative justice assumes a conversion in the way police work is done. Many officers in the Netherlands and elsewhere have joined the force with the aspiration to be a ‘tough cop’ and a ‘problem solver’, not a ‘social worker’. They do not see policing as, in fact, a holistic function that also incorporates principles such as resolving problems and offering repair (Bazemore & Griffiths, 2003: 343). It will take long-term investment in culture change to ensure that officers adopt this more comprehensive outlook to their work.
Another caveat is that police officers and their managers continuously experience high work pressure, a shortage of personnel, budget cuts and financial austerity. Restorative policing may save finances in the long run, but also comes with start-up costs, which only add to an already severe workload. This, in turn, hampers experimentation with alternatives to criminal law measures, however promising they are (Gavin & MacVean, 2018). Furthermore, community involvement is essential for the practical implementation of restorative police interventions. Victims, offenders, supporters and volunteers must be willing to support and participate in dialogue, peacemaking and conflict resolution (Bazemore & Griffiths, 2003). This is easier said than done. Citizens do not automatically partake in local meetings and projects, especially when relationships with the police are unfavourable or when they are not structurally and systematically informed about the possibility and (dis)advantages of restorative justice.

2.2 A definition

In his research on two English forces in 2015, Marder (2020) shows that, on the one hand, the police are enthusiastic about restorative justice as a means to satisfy victims and possibly save budget in a time of austerity. On the other hand, he reports the risk of poor practice due to a lack of training in conflict resolution skills, conceptual ambiguity about what restorative justice actually means and ongoing uncertainty regarding the realisation of anticipated goals and gains. However, as touched upon earlier, community police officers should not necessarily deploy restorative methods themselves. It is also possible to hire a professional non-police mediator to carry out interventions. This brings us to the following broad definition of restorative interventions in the police phase of the criminal justice system:

Restorative interventions in policing, possibly with the help of external partners and based on a combination of the individual and collective responsibility of all stakeholders involved, focus on healing, repairing and restoring harm to victims and to communities that have been (un)intentionally created by crime, violations or other problems and conflicts.

This definition consists of several elements that deserve more explanation:

• In principle, restorative interventions in policing start with intake and service staff who manage reports of crimes and incidents at the front desk of a police station or on the phone. They are the main gateway both to the criminal law system and to restorative justice options if citizens want to file a crime report (Van Steden, Anholt & Boelens, 2023). Community police officers can informally mediate between victims and offenders (Van Steden et al., 2014), but it is also possible to refer them to an external partner who facilitates dialogue and deliberation.

• External partners are professionals who operate in the broad field of restorative justice (Wolthuis et al., 2019). Examples are independent conflict mediators or facilitators of restorative circles that bring together community members to settle harmful behaviour.

• Individual and collective responsibility means that in restorative policing all stakeholders involved participate voluntarily to deal with harm (EFRJ, 2021; Kirkwood, 2022; UNDOC, 2020). Individual responsibility applies to the harm done by offenders, but also to the willingness of victims and communities to participate in restorative processes. Collective responsibility, in turn, refers to the common interest of arriving at sustainable solutions that are satisfactory for everyone, including the police.

• Harm includes financial losses as well as social, emotional and psychological harm (UNDOC, 2020), manifest in past or in the future, inflicted on victims, the wider community and professionals such as police officers. Offenders themselves can also suffer from these kinds of damage. The concepts of harm and damage relate to legal principles, but have moral dimensions of, for example, hurt, guilt and shame too (Kirkwood, 2022), which must be determined on a case-by-case basis.

• Healing, repairing and restoring relate to attempts to cure and solve conflicts between victims and offenders (EFRJ, 2021; Kirkwood, 2022; UNDOC, 2020). This can take many forms: compensation for damage or loss, apologies, reconciliation and so on. It is not always (fully) possible to come to a negotiated solution, and this necessitates finding out in practice whether and how restorative justice can prevail and how restorative aspirations can be addressed in cases where a full repair is not possible.

• Victims are those who have been affected by a conflict, violation or crime and have suffered damage or loss as a result (UNDOC, 2020). Victimhood concerns individuals who are directly involved in a situation and their broader social environment of relatives: family, friends, neighbours and so on. Police officers, other professional stakeholders and witnesses can also become (emotional) victims of a traumatising event (McCold & Wachtel, 2003). ‘Moral injury’ (Antonelli, 2017) is a commonly used phrase to describe such a distressing aftermath.

• Communities are commonly thought of as caring networks involving parents, families, friends, relatives, neighbours, professionals and other acquaintances. They support victims and offenders (McCold & Wachtel, 2003). The criminal justice system deals with offenders by setting a standard for what is acceptable behaviour – and what is not. Restorative justice may come before, during or after a criminal procedure or sanction and sometimes contributes to the (difficult) task of reintegrating offenders into their communities.

• (Un)intentionally: within criminal law, suspects (and after conviction: offenders) have intentionally caused harm or damage as a result of their crimes. However, there are conceivable situations that do not fall under this narrow definition. Consider, for example, heavy traffic accidents that happen unintentionally (Lavanchy, 2022). Restorative approaches are also possible here.

An important question remains as to how restorative interventions in policing are implemented and how they play out in everyday practice. The next section provides a framework of analysis that assesses these interventions set against core principles of restorative justice.

2.3 Framework for analysis

While being aware of a larger set of restorative justice standards, values and principles (cf. EFRJ, 2021; Kirkwood, 2022; UNDOC, 2020), following from the definition set out above, practical interventions must contain at least three core principles that can be addressed when deployed in a police context. First, all stakeholders involved in a crime or conflict voluntarily enter into a dialogue. Second, these dialogues recognise the needs of victims and offenders and specifically hold offenders responsible for their behaviour. Finally, the ultimate goal of restorative justice interventions is to effectively respond to harm, trauma and shame and, as such, promote healing, repair and restoration. We will adopt this framework of analysis to exploratively evaluate practices of conflict mediation in policing in Section 5.

3 Methodology

3.1 Action research

Exploring restorative justice in a police context, we started a pilot study on conflict mediation. The ambition behind this was to arrive at a detailed understanding of the day-to-day practices of conflict mediation and learn from these practices to further the implementation of restorative justice routines in everyday policing. Funding for the project was granted by the Dutch Research Agenda (NWA) and allowed for piloting six exemplary cases, which will be described below. This made it possible to set up a research team and hire a mediator through our own network. At the start of our pilot study, which ran from 1 October 2019 to 30 September 2020, we decided to conduct a form of action research. This research orientation is not only capable of describing and explaining new ways of working, but also aspires to change and improve organisational realities. Put differently, our pilot study can be defined as ‘research in action rather than research about action’ (Coghlan & Brannick, 2014: 6; italics in the original). The goals were to provide the police with a new perspective for dealing with crime and disorder that hold the promise of mediation and conflict resolution where this would also serve the needs of citizens. We tried to reach these goals by creating a tight partnership between police staff, an externally paid mediator and ourselves as independent researchers.
Action research goes beyond describing, analysing and explaining what is happening. It is about setting up a natural experiment designed with the view to learning and introducing novel ways of working and thinking (Schuiling & Vermaak, 2017). Our intention was to further develop the police profession and generate knowledge that would contribute to the field of restorative justice studies. The experimental interventions were set up in two ways. First, the offering of mediations by police is not a standard practice in the Netherlands. Our project thus opened up a new working method for the police that gave citizens a unique opportunity to solve their problems without further criminal justice involvement. Second, the outcomes of mediations had real consequences for the conflicting parties. Mediations all ended with a settlement agreement in which mutual agreements were laid down for victims and offenders alike. The agreements were filed in official police records and have a legal status.

3.2 Three research phases

Our research endeavour was divided into three sequential phases. Together they form an iterative cycle that consisted of (a) planning the action, (b) taking or implementing the action and (c) evaluating the action (Coghlan & Brannick, 2014). In the planning phase, we, as researchers, interviewed a police manager, a community police officer and a police detective about what they thought of as exemplary cases for conflict mediation – that is, cases which the police were frequently struggling with, as criminal justice interventions often fail due to a lack of evidence, witnesses and severe offences. This led us to carve out three categories suitable for restorative justice interventions: neighbourhood disputes, domestic violence and other types of violence, and youth problems. Careful demarcation of roles and tasks within the criminal justice chain was crucial in the preparations for mediation to be successful. An agreement between the police and the public prosecution service, being formally in charge of criminal investigations, allowed the officers to play a pioneering role in selecting suitable cases at an early stage and to process these in the direction of alternative conflict resolution. As laid down in Article 51h of the Dutch Code of Penal Procedure, the police also have a statutory task to do so.
The aim of the pilot study was not to test the effects of restorative justice in policing in a representative sample of cases but let practices speak for themselves. Predetermined criteria for selecting our cases included the presence of known victims and offenders, the persistence of a conflict and/or the chance of escalation. Meeting these criteria, the police, in consultation with the researchers and the mediator, thoughtfully selected six exemplary cases from the daily flows of reported crimes and incidents that appeared promising for conflict mediation: three neighbourhood disputes (cases 1-3), a violent argument in the street (case 4), domestic violence (case 5) and non-consensual sharing of nude images among youth (case 6). These cases were reported to the police during our pilot study and involved seventeen participants. In the three neighbourhood disputes, conflicting parties were either couples or were accompanied by spouses. In every case selected, the police approached victims and offenders involved for a referral to the mediator who scheduled an appointment shortly afterwards at a local police station in Amsterdam.
In the action phase, the researchers observed the six conflict mediations without being participants. The professional mediator was fully responsible for facilitating the sessions. During these sessions, we used an observation scheme to document the case according to a prescribed pattern that followed the mediator’s guidelines: (a) contacting the victims and offenders, (b) interviewing them separately about their views on the conflict, (c) bringing them together for dialogue and deliberation (the conflict mediations), which resulted in signing a settlement agreement, and (d) providing aftercare (see more details below). We have manually analysed the observational data around three topics: the case content, the summary of mediation and the results or outcomes of these sessions.
A disadvantage of observational studies ‘is that it is an inherently subjective exercise, whereas research requires objectivity’ (Mack, Woodsong, MacQueen, Guest & Namey, 2005: 15). Given this weakness, we tried to filter out personal biases as much as possible. Observing the mediations, we wrote down the sequence of events in real time and transferred our findings to field notes. These field notes were cross-checked with and reviewed by the mediator. There was no intention to arrive at generalisable results about either the effects of mediation or actual police performance. Instead, during and after the action-taking and data collection phases, we reflected on empirical findings with the police in order to infer lessons for future practice.
The evaluative part of our project consisted of calling participants – victims and offenders alike – one month (or slightly more) after the mediation took place. In so doing, we provided aftercare, were informed about people’s well-being and asked about their personal experience with conflict mediation. One person was causing enduring problems for his neighbours. On police advice, we decided not to call him back. Two others indicated that they did not want to be called back for personal reasons.

3.3 Ethics

Before collecting our data, we completed an online self-check provided by the university. This led to discussions about the ethical framework of our research with police colleagues. The police carefully selected cases they deemed suitable for conflict mediations. From their perspective, such mediations were more likely to prevent conflicts rather than escalate them. We hired a sworn-in professional and experienced mediator to make sure that the mediations were conducted in a sound and ethical manner. Following official restorative dialogue standards (Hansen & Umbreit, 2018; Umbreit & Armour, 2011), the mediations followed a four-step process:

• The police, in consultation with the mediator, contacted the victims and offenders prior to the mediation to ensure that they were well-informed and suitable for the process.

• The mediator had separate meetings with the conflicting parties to gain a better understanding of the conflict, establish trust, explain the purpose of the sessions and manage expectations.

• The mediator facilitated structured communication between the parties, allowing them to talk and listen to each other, and helped craft settlement agreements.

• As researchers, we provided aftercare later to check on the well-being of the conflicting parties, gauge their experience with the mediations and assess any changes in their situation. The police also monitored the cases and contacted social support services if necessary.

The mediation sessions took place in local police stations to ensure the safety of all involved. Prior to the sessions, victims and offenders gave their signed consent to participate in our project. This ensured that we, as researchers, could quietly observe the mediations and take notes with the assurance of maintaining anonymity. We received no objections or refusals afterward.

4 Results

4.1 Neighbour dispute (case 1)

The conflict. The conflict concerned a neighbour dispute. Party 1 (a couple) had reported ongoing defamation and threats by Party 2. Party 1 had lived at the address for about ten years and both people had busy daytime and evening hour jobs. Party 2 lived alone, had gone through a divorce and was long-term unemployed. Party 2 was cross with Party 1 because, in his view, they were bullying him, for example by emptying an ashtray in his garden.
Party 1 denied this and insisted that Party 2 slammed doors at night, spat on their windows, called them racists, raised his middle finger and so on. The couple felt unsafe, were constantly alert and said they no longer enjoyed living in their apartment. According to Party 2, the quarrels and conflicts had been going on for years. Some quiet intervals aside, incidents had persisted and disturbed.
Prior to the mediation, Party 2 accused Party 1 (the couple) of stealing a postal package and complained about watching and intimidating him as he came and went. Party 1 had asked Neighbourhood Mediation (Beter Buren), a Dutch non-profit conflict resolution agency, to intervene, but was disappointed that Party 2 did not want to cooperate and talk. Party 2, at times, was verbally aggressive and said he ‘was about to hurt his neighbours’, but also felt confused about what he had done wrong and why the neighbours were taunting him.
The mediation. The mediator started with the allegation about the stolen mail. Much to Party 1’s annoyance, Party 2 had posted a note on their front door asking where his mail was. However, it appeared that the package had not been delivered yet. This example, Party 1 said, illustrated the confused and aggressive behaviour of Party 2. They had even called the police a couple of times to report the situation, but without any lasting result. During the conversation, Party 2 got angry a couple of times and wanted to walk away. He accused Party 1 of not wanting to talk. The mediator emphasised that there would be no winners or losers, all suffered from the conflict. Her aim was to look to the future, to put agreements on paper and, hopefully, to make a new start as neighbours.
The outcome. The mediator listed what the parties did not want from each other: spitting, emptying ashtrays, conflicts about mail, threats and insults. She concluded that they simply desired ‘peace’. Both parties agreed. The mediator wrote out the final agreement, which everyone signed.
The follow-up calls showed that Party 1 looked back positively on their own participation: the information provided by the police was clear, participation was completely voluntary, the mediator did her work ‘professionally’ and ‘impartially’ and she listened to everyone. Unfortunately, Party 1 reported that the problem had not come to an end. According to Party 1, this was because Party 2 used alcohol and/or drugs, was lonely and suffered from psychosis. Party 2 said he felt well-treated by the police and by the mediator. He cooperated voluntarily and was satisfied with the outcome of the mediation. He also indicated that the conflict had not been solved but instead blamed Party 1.

4.2 Neighbour dispute (case 2)

The conflict. This case was about a conflict between two neighbours, one living above the other. Party 1 was aged 23, and Party 2 was aged 17, therefore a minor. Party 1 came to the mediation with his mother. Mother 1 was not able to express herself in Dutch and needed an interpreter. Party 1 (her son) was a school leaver, had no job, and lived on benefits. Party 2 also came with his mother. Party 2 attended secondary school.
Party 1 said that the problems started when the neighbours called the police because of an argument he had with his mother. He said he ‘was crazy about her’, but also felt a lot of pressure and responsibility for administrative matters and tax matters because his mother did not speak Dutch. Party 1 admitted he and his mother sometimes screamed at each other but denied physically abusing her. However, Party 2 accused his neighbour of domestic violence, which outraged him. The hostile situation turned into a fight between the two youngsters on the stairs. Party 1 had filed a crime report against Party 2. Party 2, in turn, felt very much affected by the conflict between Party 1 and his mother and was tired of the conflict between him and his neighbour.
The mediation. The mediator stressed that there was a unique opportunity to make a clear agreement about future behaviour. She wondered why there had been no attempt to contact Neighbourhood Mediation, but Party 1 muttered that he did not want to talk to these volunteers. His mother’s interpreter said that she thought her son ‘was not easy’. Judging by his frown, her son did not appreciate her remark. He also complained loudly that the interpreter was not translating his mother correctly. The interpreter responded laconically that he understood his mother perfectly well. His mother urged the mother of Party 2 to take care of her son and find a viable solution. She felt embarrassed and threatened. The mothers indicated that they had no problems with each other. However, Party 1 – the son – should stop yelling and banging, then and in the future. The mediator managed to de-escalate the situation and, after a long discussion, established that the parties agreed to respect each other.
The outcome. The mediator read a long summary of the final agreement. Party 1 resisted a bit, but eventually agreed. Everyone put his or her signature at the bottom of the final behavioural contract. Party 1 then left in a hurry. His mother shook hands with everyone and even hugged Party 2 and his mother.
A few weeks later, after calling back, it turned out that the situation had not fundamentally changed. In Party 2’s view, the mother was totally dependent on her son (Party 1), he cherished his power asymmetry, denied nuisance problems and was unwilling to cooperate. This made a sustainable solution difficult. Party 2 described the mediator as ‘professional’ and ‘nice’; she ‘listened to everyone’ and ‘had done her best’. Since the problem had not been solved, Party 2 approached the housing association in the hope that the conflict would end. The housing association could issue official warnings and, in extreme cases, ensure that a tenant was evicted. It was decided not to call Party 1 back and stir up his emotion because he had not changed his misbehaviour and still troubled his neighbours.

4.3 Neighbourhood dispute (case 3)

The conflict. Party 1a/b (a couple) and Party 2a/b (a couple) had been living next to each other for five years. Party 1a/b rented their apartment; Party 2a/b had bought their house. Recently, there had been disagreements between them about noise, children and renovations. These disagreements culminated in a physical struggle between the two males: Party 1a grabbed Party 2a by the neck in response to verbal abuse. The spouses of both (Party 1b and Party 2b) tried to play a ‘de-escalating role’ and were also present at the mediation. Party 2a filed a crime report against Party 1a to make a clear statement.
The mediation. The male neighbours (Parties 1a and 2a) each blamed the other for causing the conflict. Interestingly, their wives (Parties 1b and 2b) fully agreed that continuing along these lines was annoying and unproductive. Party 2b proposed burying the hatchet that day. Party 1b responded that something needed to be done first. The mediator confirmed that moment as an avenue towards reconciliation. The quarrelling parties should not have held on to their anger. After a lengthy conversation, the mediator opted for a follow-up in four weeks. She added that the parties themselves were in charge of settling their dispute by expressing wishes, not accusations. Shortly after the session, Party 2a sent a text message to the mediator. He was grateful for her time and attention but felt unsatisfied about not hearing any apology from Party 1a. In his view, their roles in the conflict sometimes seemed reversed. He would therefore have little faith in the efficacy of another gathering. The mediator reflected that Party 2a did not ask for an apology.
After a month, a conversation between the wives (Parties 1b and 2b) took place. Party 1b wanted to discuss a number of matters. She thought that her husband’s health problems might have partly sparked his capricious behaviour. This was no excuse, but it was nevertheless relevant to know. Party 2b understood this. Their husbands were much alike in being stubborn, which led to irritations and worse. Party 2b hoped for a better mutual understanding and thought that her husband reacted disproportionately too. It was never good to stop talking and keep shouting.
The outcome. The mediator sensed that all neighbours wished to improve their communication and came to an agreement. From their telephone calls, it transpired that Party 2a/b were critical of Party 1a/b’s ‘influence’ on how the process went. The conflict may not have been completely solved, but important steps had been taken. In particular, the wives pointed at ‘appeasement’ taking place during the follow-up mediation. They were happy with the outcome.

4.4 Street fight (case 4)

The conflict. Party 1 and Party 2 had disagreements for years, mainly arising when they were walking dogs. Things went completely out of hand when Party 2 kicked Party 1 in the back, after which a friend X of Party 1 punched Party 2. Party 2 wanted Party 1 to better control his dogs. He also said his phone got damaged in the fight. However, the most important thing was to get to the bottom of the conflict. The disagreements had to end.
The mediation. Party 1 felt constantly provoked by Party 2. The mediator asked whether, according to Party 2, this was because both parties were ‘too nosy’ about each other’s lives. Party 1 confirmed this: ‘yes; just don’t bother me’. Party 2 clarified that the conflict had been going on for a long time. The mediator took both parties back to the incident reported to the police. Party 1 remembered a raging quarrel about the dogs, after which a fight occurred. Party 2 told a similar story. When the mediator asked what the parties hoped to achieve, elements of ‘peace,’ ‘ignoring each other’ and ‘keeping dogs apart’ came to the fore.
During the conversation, the parties continuously blamed each other. Party 2 accused Party 1 of running an illegal dog walking business. The mediator argued that even if Party 1 had such a business, Party 2 did not need to take the law into his own hands. Party 1 said that Party 2’s hurtful remarks and innuendos made her furious. The mediator: ‘You were so focused on each other that tensions kept on escalating. Turn your gaze elsewhere, turn to your own life!’ Both parties concurred with this insight.
The outcome. The parties agreed to leave each other alone and prevent further escalation. Our follow-up calls learned that Party 2 now avoided Party 1 and evaluated the mediation as ‘very positive’. Party 1 was also happy about participating in the mediation: the sting had been removed from the situation.

4.5 Domestic violence (case 5)

The conflict. This case concerned domestic violence between spouses. Party 2 (the wife) returned home two days prior to the mediation after being away from home for four months. She filed a crime report against her husband (Party 1) to ‘draw a line’. Her goal was not to start criminal proceedings but to change Party 1’s behaviour. The couple had an adolescent son who exhibited problematic behaviour. The husband (Party 1) was seriously ill.
Party 1 explained that the conflict centred on deep disagreement about how to handle their son. At some point, there was a physical confrontation between the two spouses. Party 2 had bruises and torn clothes, after which she left to stay with her sister. Party 1 said ‘it must not happen again’. He hoped for a restoration of his marriage beyond ‘peaceful coexistence’. ‘I am in a hurry because of my health’, Party 1 added. The spouses were in emotionally focused therapy.
The mediation. The mediator invited the parties to talk about the incident that had taken place four months ago. Party 2 outlined what had happened, that she was fed up and left. Party 1 admitted his bad conduct and understood what his wife went through after being assaulted by him. He really tried to listen to her but sometimes found it hard to understand what she was thinking. However, he longed to live in harmony. Party 2 wanted to be there for her family but needed a safe and peaceful place. The mediator concluded that the couple desired to give their relationship a go: ‘I think that is brave of you’. The mediator also mentioned the difficulties with their son. This softened the conflict because it involved a joint task and responsibility.
The outcome. Both parties expressed their intention to look for a solution should things get out of hand again. The mediator summarised everything and checked what Party 1 and Party 2 thought of the mediation. Party 2 was pleased to be able to talk about the conflict situation. Party 1 was too ill to participate in our interview on the phone. Party 2 said she agreed to the mediation because she was in need and hoped to rescue her marriage and family. The problems with her son and the daunting relationship with her husband had not improved much. Party 2 still felt glad that the mediation had contributed to some sort of solution: she was back home.

4.6 Non-consensual sharing of nude images (case 6)

The conflict. This case was about non-consensual sharing of nude images among youth. The suspect was a fifteen-year-old boy, who sent an intimate photo of himself to the victim, a thirteen-year-old girl. She returned a similar photo of herself to the boy, which he later on shared with a large network of contacts. Both parties, minors, were represented by their parents: the mother of the suspect (Party 1) and the father of the victim (Party 2). Neither the boy nor the girl was present at the mediation.
The text messages sent to the mediator by Party 2 (the father) were remarkably aggressive in tone. He talked about a ‘completely devastated 13-year-old girl’, wanted to see ‘money’ as a token of compensation, ‘couldn’t sleep for three days’, was ‘a ticking bomb’ and preferably did not want to ‘break the boy’s legs’. Party 2 seemed to suffer from a complex addiction background and had behavioural issues. Due to the risk of escalation, and because both parties intended to work things out, we initiated mediation. A community police officer had confiscated the suspect’s smartphone.
The mediation. Party 1 was the first to speak. The mother stated calmly, but visibly sadly, that her son’s despicable action was a one-off event. She also said that her son was pretty distressed and regretted the consequences of his thoughtless behaviour. Party 2 (the father) responded that his daughter was very upset. She cried constantly and had low self-esteem. Party 2 gradually became calmer in response to questions from the mediator.
Party 1 explained for a second time how badly her son felt but realised the girl was a victim. Party 2 underlined that they wanted to reach a mutual settlement last week, which did not succeed. He challenged Party 1 to donate an ‘acceptable amount of money’ to be spent on doing ‘something fun’ with his daughter so the case could be closed. His question lingered in the air for quite some time. In the end, the boy’s mother (Party 1) offered several hundreds of euros, stating: ‘This is the amount my son has saved up throughout his entire life’. Party 2 immediately accepted.
The outcome. After arranging the financial transaction on the spot, both parties agreed that their children didn’t need to talk to each other. They didn’t go to the same school and would not meet again. In response to our follow-up calls, Party 1 said that it had been ‘a stressful time for all involved’ and refrained from responding to our questions. After the police had erased the nude picture from her son’s phone, he got his device back. Party 2 was satisfied with the outcome of the mediation. The boy had received a ‘serious punishment’, his mother had been ‘anxious for a while’ and his own daughter had received a ‘fair compensation’. He said that the mediation had produced other positive outcomes too. His daughter now visited a coach, who helped her to be ‘stronger’ and ‘more resilient’.

5 Reflections

5.1 Voluntary entering into a dialogue

In all of our six case studies, people, whether in the capacity of victims or offenders (or a mix thereof), willingly entered into dialogue with each other. Participation in mediation was voluntary, which is a fundamental principle of restorative justice (EFRJ, 2021; Kirkwood, 2022; UNDOC, 2020). Yet, at times, using an element of police authority to convince people seemed to help. The invitation to come to a police station was not only necessary for reasons of safety and security; it had a symbolic function too. As Walker writes, ‘[t]he police … represent the state’s most visible guarantee that, however provisionally, authority will be restored wherever it may have broken down’ (1996: 57). Where, for example, voluntary mediators employed by Neighbourhood Mediation (Beter Buren) were not always able to pacify people causing a neighbourhood dispute (cases 1 and 2), police contact did have impact.
Case 6, concerning the non-consensual sharing of nude images, violates the principle of restorative justice, which states that the primary parties – that is, the boy as the offender and the girl as the victim – must be involved in dialogues and deliberations. However, the victim’s father was determined to protect his daughter from meeting the offender again. He also expressed their intention to press charges against the offender if the case could not be settled informally soon. Subsequently, the offender’s mother, the police and the mediator decided to set up a mediation process. The question of what the underaged victim wished for, a key element in deploying restorative justice in cases of sexual violence (McGlynn, Westermarland & Godden, 2012), remained unanswered.
Lastly, it might be worthwhile to involve people other than those directly involved in the conflict to create a fuller restorative circle (McCold & Wachtel, 1998). A wider group of social workers, family members, neighbours and representatives of housing associations, for example, are often stakeholders in ongoing quarrels. They can act as witnesses, draw up dossiers, have an interest in ending the conflict and can exercise peer pressure. Mediation sessions can then evolve into larger restorative conferences to foster dialogue and deliberation among a greater number of stakeholders.

5.2 Recognising the needs of victims and offenders

Reporting crime to the police does not necessarily represent a crime-reporting rationale or the wish to punish an offender. Instead, victims may be seeking help, wishing to ‘make a statement’ (case 3) or ‘draw a line’ (case 5). The mediator, in turn, attempts to meet these individual needs. As a complicating factor, in two cases involving neighbour disputes (case 1 and 3) and in the quarrel about the dogs (case 4), it was difficult to discern who was the victim and who was the offender; distinctions and needs became obscure. In the second neighbourhood dispute (case 2), it even appeared that it was the self-declared victim who caused the trouble himself. Hoekstra (2022) argues that police officers, specifically, are quite likely to offer restorative responses to such ‘messy victims’, who are seen as not completely genuine and somewhat culpable, because they fit the implicit judgement that situations are always two-sided and people just have to learn how to cope with life. However, the blurring of boundaries and images complicates the course and outcome of mediations if none of the participants takes full responsibility. Problems were indeed often not or not completely solved.
Yet, importantly, even with a disappointing outcome, victims and offenders (or, at least, as people framed themselves) felt satisfied with being given the chance to speak out. They also praised the mediator’s calmness, independence, neutrality and monitoring of equality. She prepared the meetings, listened to people, negotiated shared outcomes (the written statements) and provided follow-up contact if necessary (case 3). However, at times, the mediator did not strictly adhere to a neutral and unobtrusive style, but was quite active in evaluating people’s conduct as ‘too nosy’ (case 4) or ‘brave’ (case 5). This shows, as Choi and Gilbert also discovered in their empirical study on mediators’ practices in the United States, ‘variance in “real world” restorative justice’ (2010: 207). Mediators ‘translate’ abstract theories and principles their own way in fostering repair and reconciliation between victims and offenders.
Finally, elements of restorative justice seemed to connect with the procedural justice paradigm: being treated fairly and respectfully by public authorities positively affects their legitimacy in the eyes of citizens, regardless of unsatisfactory or ineffective results (Tyler, 2006). Yet, there was more at stake. Participants in our mediations appreciated being listened to and understood; they were able to tell their part of the story (in all cases). This appreciation goes beyond simple compliance with the justness of procedures followed. Being seen and heard are basic human needs. Thus, as Van Camp and Wemmers (2013) found earlier, underlying the procedural justice dimensions of ‘voice’ and ‘recognition’, conflict mediation encapsulates caring dimensions that meet distinctive moral and pro-social needs and purposes.

5.3 Promoting healing, repair and restoration

The mediator did try to promote healing and restoration in all observed cases by facilitating the signing of a formal agreement letter between conflicting parties, aimed at resolving the conflict and encouraging improved future behaviour. However, sometimes there was more to be done. In cases 1 and 2 (neighbourhood disputes), underlying factors such as social, economic and personal vulnerabilities played a major role in the conflict and seemed to reduce the chance that people would comply with the signed agreements. Much information about these factors came out in the course of our mediation and had not been present in the official police reports beforehand. Furthermore, as was explained in the methodological and ethical sections, the logic behind the piloted mediations was to interview the conflicting parties separately, bring them into dialogue with each other and arrive at a shared written agreement. Future mediation sessions may require more and longer preparation to increase the chance of positive outcome, but this comes with a dilemma: mediations in the police phase assume ‘on-the-spot’ treatment.
In addition, although words like ‘remorse’, ‘regret’ and ‘repentance’ easily come to mind when thinking about restorative justice (Van Stokkom, 2002), it was striking that none of the victims asked for an apology and offenders did not offer this either. Mediations mainly involved pragmatic problem-solving rather than emotional healing or restoration. Some participants, however, did speak in veiled terms such as ‘it must not happen again’ (case 5) or regretted that they did not ask for an apology (case 3). Despite this, all participants signed the final agreement promising better future conduct. Not everyone adhered fully to their promise, though. (Perceived) erratic behaviour among respondents seemed to detract from long-term solutions and peace between parties (cases 1 and 2).
The agreement of settling the conflict economically was controversial in the case of a non-consensual sharing of nude images (case 6). By consent of both parents representing their children, the mediation provided an alternative to further criminal investigation and conviction of the adolescent offender. Sending around explicit pictures may trigger a police and public prosecutor referral to HALT, the Dutch version of customised community service sanctions. It was most likely that the police saved time and costs in this case, and both parents deliberately approved the alternative settlement, but this case also demonstrated the ethical dilemmas of mediations in community policing. Other dilemmas arose concerning the ongoing toxic and abusive relationships of vulnerable females with their son (case 2) and husband (case 5). Safeguards were put in place, with the police carefully monitoring the cases and contacting social support services such as Safe at Home (Veilig Thuis).
Still, in conflicts where parties had adopted a ‘fighting attitude’ towards each other for an extended period, achieving transformation is difficult. It can indeed be argued that restorative solutions in a police setting are insufficient when it came to addressing domestic abuse (Westermarland, McGlynn & Humphreys, 2018). At the same time, if victims do not press charges against offenders, return home voluntarily and encounter difficulties in ending their relationships, the police are left empty-handed. A formal crime report is legally necessary to initiate further investigation and make arrests. Moreover, as Stanko (1995: 40) notes, ‘policing itself is not a panacea for domestic violence, a violence that finds support within wider societal structures that legitimise violence against women’. In these situations, the police require larger social care networks and anti-domestic violence programmes. More integrated policing, restorative and public health solutions are indispensable to effectively cope with cases of domestic abuse.

6 Conclusion and discussion

The starting point of our research project was the observation that many crime reports by citizens do not necessarily lead to adequate police action. This can be due, for example, to their low priority status and lack of resources. Citizens, in turn, feel disappointed; their problem has not been resolved. Moreover, criminal law should only be used as a last resort, an ultimum remedium. We therefore explored alternative police actions aimed at providing conflict mediation to victims and offenders (or a mix thereof), as an optimum remedium. If promised as an option, conflict mediations can contribute to a better community policing and improve public satisfaction with the police.
As our research demonstrates, suitable cases for police referrals to conflict mediation include neighbour disputes, fights, domestic violence and youth problems. Our research also shows that restorative justice, when put into practice, appears to be ambiguous. This finding does not imply that the police and the hired mediator deliberately ignored the normative standards and theoretical principles laid down in the scientific literature. Instead, it turned out to be quite challenging to entirely meet these high standards and principles in everyday community policing and mediation work (cf. Vanfraechem, 2009). Let us provide three reflections.
First, ideas of ‘voluntary’ and ‘inclusive’ participation presuppose equality among all stakeholders, but ‘power imbalances’ (Braithwaite, 2002) could not be avoided completely. Although the police never forced people to participate in the mediation sessions (far from that), their ‘symbolic dimension’ (Walker, 1996) may have played a role in bringing victims and offenders together when volunteer mediators failed to do so. This was most visible in two cases of neighbour disputes. Furthermore, in the case of non-consensual sharing of nude images, the two parents decided that their children should enter into dialogue. Otherwise, the mediation sessions would not have taken place, while, according to the police and the mediator, it was preferable to settle the dispute at an early stage to prevent further escalation.
Second, for mediators, it is important to maintain their ‘neutrality’ and ‘impartiality’. However, consistent with a previous study on conflict mediation (Choi & Gilbert, 2010), the mediator had developed her own style. She sometimes took on an active and evaluative role in the cases, which was not necessarily a problem. Her style appeared facilitative towards a negotiated outcome, and our respondents generally indicated positive experiences after the mediation sessions. Nevertheless, the revealed gaps between the guiding principles of restorative justice and the realities on the ground can create unintentional dissonance in meeting the needs of victims (Choi, Gilbert & Green, 2013), as well as those of offenders. In case 6 about the nude images, the two juvenile protagonists were not present at the police station.
Third, we did not witness any profound ‘moral emotions’ (Van Stokkom, 2002) of repentance and forgiveness during the mediation sessions. Following Jesilow and Parsons, restorative justice in a community policing context seemed to work best in a ‘give-and-take atmosphere’, where stakeholders showed interest in an open conversation and resisted seeing the world in terms of ‘us/them’ (2000: 178). Mediations revolved around pragmatic problem-solving rather than restoration and healing. In particular, the situations of the two female victims involved in toxic or abusive relationships with family members (husband, son) in cases 2 and 5 were not completely resolved, and in cases 1 and 2, some participants ostensibly displayed physical instability that hindered progress. Social work and support networks need to supplement policing and restorative justice to offer long-term solutions here.
This leads us to the barriers faced in implementing restorative justice and mediation within community policing and broader societal structures. While restorative practices in policing can thrive only after substantial investment in training, education, empowerment of frontline staff and organisational adaptations (Clamp & Paterson, 2017), the underlying question is whether offering alternative interventions as a means of supporting victims and offenders with tailor-made solutions that address their personal capacities – but go beyond the criminal justice scope – is genuinely feasible and executable. An overemphasis on top-down tactics is unlikely to be effective since restorative justice, in essence, must be understood as a transformative process that requires police engagement throughout the organisational hierarchy and within local communities (Clamp & Paterson, 2017; Stockdale, 2015). The observations and insights from six in-depth case studies in the Netherlands are, obviously, far too limited to provide a final answer.
That being said, a future research agenda for restorative justice should consist of large-scale qualitative and quantitative studies that describe the scale and scope of policing initiatives around the world. Such studies should analyse the outcomes of these initiatives in terms of victims’ and offenders’ needs, recidivism rates, public satisfaction with the police, improvement of police-community relations, public savings in the criminal justice chain and the willingness of police officers to embrace restorative justice solutions. This is not merely an ideological debate fuelled by restorative justice enthusiasts in the police force and academia. In the Netherlands, at least, there is a legal obligation as well. As touched upon above, Article 51h of the Dutch Code of Penal Procedure promotes police use of mediation (and other restorative justice practices) if victims and offenders voluntarily agree. Nonetheless, there is still a long way to travel.