The Police as Gatekeepers of Restorative Justice – Perspectives of Professionals, Perspectives on Professionals

Methodology

I look into different areas, which could help shine light on the access to restorative processes in Denmark and Norway.⁶ I combine qualitative and quantitative data in an eclectic approach to answer the abovementioned research questions. The data collection for this article evidently does not reflect the entirety of either the Danish VOM programme or the Norwegian Mediation Service. Furthermore, there will often be considerable differences between what informants think they do or say they do and what can be observed by researchers. This includes the interviewing and observation of police officers.⁷ Hence, caution is to be exercised when basing conclusions on interviews and observations.

I have attempted to avoid misrepresentation by triangulating data as well as presenting contrasting findings or points of views when present.⁸

The data from Denmark are focused on professionals affiliated with the police-based national VOM-programme (Konfliktråd); either as employees of the police or as employees in municipalities having cooperated with the police on VOM cases. In the Norwegian data, my focus is on professionals involved in the screening for and execution of the two restorative reactions to youth crime. The Norwegian professionals are employed either in the Norwegian Mediation Service (Konfliktråd), which is responsible for the two restorative reactions or in cooperating institutions involved in the screening and execution.

The collected data are:

Qualitative data from Denmark:

Qualitative data Norway:

Quantitative data:

Other data on both services:

Data collection and access

The qualitative data collection took place from September 2015 to the end of 2018. All interviews and observations were conducted in Danish or Norwegian. Direct interview quotations appearing in this article were translated into English by the author. For clarification purposes some quotations have been edited.

Data access was obtained partly through the national secretariats for the Danish VOM programme and the Norwegian Mediation Service, and partly by direct responses by VOM coordinators (Denmark) and youth coordinators (Norway). The informants where to some extent selected as a result of non-random snowball sampling in that the first informants I contacted would sometimes suggest that I contact others, which I would then do. This could – among other – mean that possibly on average my informants are somehow ‘cherry picked’ and as such maybe more engaged than the average of employees in their position. Also, in both Denmark and Norway some of my requests remained unanswered, entailing a certain element of self-selection in the sample as well as a limitation in the geographical representation. Overall, in the case of the Danish VOM programme, my informants represent current or former employees from 5 out of 12 Danish police districts. The Norwegian informants are all from the Oslo region.

The balancing of interview data and observations in my samples from Denmark and Norway respectively is highly related to the organisational frameworks and screening approaches of the two organisations. As elaborated later in the article, the screening of potential cases in the Danish VOM programme is largely one which takes place within the mind of individual police officers and VOM coordinators, and it is thus a process, which is difficult to observe. Also, when it comes to the conscious or unconscious initial screening of VOM cases carried out by police officers during interrogations of parties, the large variations in practice suggested in my data set would call for a very large set of observations as well as a wide access within the Danish police, which can be difficult to get.⁹ In the Norwegian service, it is possible to observe groups of professionals as they discuss the screening of cases for two restorative youth sanctions.¹⁰

Informant protection

For anonymity purposes informants are not referred to by their names but only their generic titles. Specific comments are not attributed to specific anonymised informants. Exceptions are the national secretary from Denmark, the leader of the Norwegian Mediation Service, Oslo Akershus department, and the national coordinator for the Norwegian Mediation Service (now retired). There is only one person in each of these positions, making anonymization impossible. Even if their names are not mentioned in the article, for this reason none of these informants were promised any anonymization.

Pre-screening by police colleagues

Judging from my data, the majority of cases where parties are not informed about the possibility of VOM participation can most likely be ascribed to lack of awareness caused by the general pressure on police resources, the overall framework of the police, and the relatively low prioritized position of the VOM programme within this framework (Rasmussen, 2020a, 2020b). However, most of my informants in the Danish VOM programme vocalise a concern that at times some colleagues do tend to actively pre-screen cases based on what the VOM coordinators would describe as inappropriate criteria.

A VOM coordinator illustrates the potential dilemma of inappropriate pre-screening as follows:

VOMC: I find the colleagues in our district to be predominantly positive towards VOM. […] Yet, there are times when a colleague tells me about a case. And then I go ‘oh yes, and this is when you remembered to inform about VOM, right?’. And then he says ‘you know what? I actually thought about it because VOM is a good thing’ – and this is not a lie, what he is saying. […] But police officers – we are these autonomous people who are used to that there are constantly things, which are for us to decide. We drive around in the streets and to some extent we decide who gets a ticket today. Because we cannot give them all a ticket. So, you have to select someone who extraordinarily earned it. And you could say that there are different parameters for selecting. So, we are used to having our own attitude towards things. And therefore, he says ‘VOM is a good thing. But in this specific case the offender is such a jerk. I don’t think it will make any sense at all. It would just be wasting their time or even make things worse for the victim’. Or they can have all sorts of ideas about it. I try to challenge them and say ‘well, this is YOUR view. You do not know enough about this [to make that estimate]’. They do not realise that sometimes this is all it takes [for VOM to be a success]. That the victim just needs to SEE that the offender is a jerk – to stay in that terminology.

I just feel like the colleagues sometimes make an individual assessment. […] And I think that sometimes they are not aware of their own criteria in this regard. […] And then there are lots of prejudices too. ’Will he do it again? – Goddammit, he will do it again, that one! He is not going to stop this shit just because he enters a meeting with someone.’ That prejudice likely exits out there, right? […] ’And is it fair then, to put the victim in the same room as someone who maybe just SAYS that he will not do it again? Because I know he will do it again. Because I can see in his records that he has done it so many times before.’ And then we are back to ’well that decision is not for you to make’.

[…] This is a big challenge in the police. And how to put that nicely in writing – I don’t know. Haha.

I find that the above quotation provides an interesting illustration of a casting situation: how the combination of the police typologisation – including the previously described moral estimate of who is right – and the feeling of responsibility towards the victims, might lead to inappropriate pre-screening. It can result in a de facto limitation of VOM access for some types of offenders who are unappealing to the police. As offenders are ideally to be asked about VOM participation before victims, such pre-screening is also likely to limit VOM access for their victims.

One could note that this might be interpreted as primarily related to lack of training of the police officers in how to appropriately screen cases for VOM. My data indicate that inadequate VOM knowledge and training of police personnel is indeed a factor, and that the options of systematically addressing this are limited in a police organisation under pressure (Rasmussen, 2020a, 2020b). But as noted in Rasmussen 2020a, it appears to not just be an issue of lack of training or lack of time to train but also an issue of the clashing of perceptions of what is to be viewed as real police work or core police tasks. In this regard, in the case of the Danish police, my data indicate that the VOM programme is not seen as belonging to this category, either by police officers in general, police managements, or the political decision-makers. As described, this influences the options of training police officers in how to engage with the VOM programme screening as well as the options of enforcing accountability on this point (Rasmussen, 2020a, 2020b).

In the following, I will explore how mechanisms of typologisation might play into VOM pre-screening.

The primary typologies observed by Holmberg (1999) were customers (people with a criminal record), clientele (people who – based on their socio-economic appearance are considered potential offenders), and (good) citizens (Mr. and Mrs. Denmark) (the rest). The classification of a good citizen might change for the worse if their actions challenge the category. According to Holmberg’s typology, such turn of event can lead to placing citizens in the category of assholes. Assholes can be surprising or expected, with especially women and children falling into the first category. The emerging of a surprising asshole can be an event, which takes some time for police officers to process (Holmberg, 1999).

Notably, according to Görtz (2015), Holmberg (1999), and Sollund (2007), ethnic minorities serve as a special category because – disregarding their remaining socio-economic appearance – they are more likely than the ethnic majority population to be perceived as clientele than other citizens are.

In Goffman’s understanding, the interactions involving offenders and police will direct both parties towards certain roles. The presence of the police role is more likely to set the scene for a dramatic interaction than for instance a comical one. In this way, police presence invites certain other roles to enter the play while rendering other roles unlikely: Just as police officers at work will not likely come across as jazz enthusiasts, family men/women, or passionate knitters, large parts of the characters of the citizens they meet will probably also not be evident in interactions with police.

The coordinator’s above description of the patrolling police’s autonomous working routines and excision of estimates fit very well with Holmberg’s observations of the same arena. The interview fragment especially relates to offenders known by the police, hence belonging to the customers’ category. Several part-takings in police-offender scenes will likely have pre-enhanced the characteristics relevant to this particular setting of both parties in the eyes of the other. Goffman distinguishes between communication in or out of character. In order to avoid inappropriate VOM pre-screening, it is relevant to examine and possibly challenge and develop police officers’ ability to imagine offenders communicating out of what they perceive as the character.

Various research has indicated that compared to patrol policing, local policing can be a way of widening the character scope for both police and those in the limelight of police attention, whether clientele or customers.¹⁷ In their 2003 study of police and youth interaction in a somewhat troubled Danish residential area, Ansel-Henry and Jespersen demonstrate how the relations between youth and local police are significantly more nuanced than the often very tense relations between youth and police patrols, especially canine patrols.

The more nuanced relations between local police and youth contribute to the de-escalation of conflict situations on the part of both police and youth, whereas without the presence of local police, conflicts between youth and patrolling police tend to escalate from both sides (Ansel-Henry and Jespersen, 2003). Similarly, Petterson’s study (2012) on Stockholm local police interactions with youth describes how conversations between local police and youth including joking and sharing of personal stories might lead to positive interrelations. Drawing upon these relations, in situations of escalating conflict the local police can choose to give young people more space and deescalate the conflict or to mark their authority in a more traditional police manner, for instance by arresting the young person.

Put in Goffmanian roleplay terms, local policing provides the option for the participants in detached scenes to evolve into recurring characters in a full movie or maybe even a multi season series in the eyes of each other. The insight into the many facets of the characters can grow accordingly. Think of the initial villains, Jamie Lannister of Game of Thrones-type characters, whom we occasionally over time can grow to sympathise with or even cheer for. Not due to their harsh, dehumanising actions, but due to the scenes where they get to show their humanity and be heroes in a part of their own story. In this way, police officers with local policing tasks might arguably be better equipped to avoid exercising inappropriate pre-screening than those of their colleagues primarily engaged in patrolling activities – at least when it comes to VOM parties they know.

Coordinator screening

As described, according to the legislative framework of the Danish VOM programme the screening that takes place at VOM coordinator level is supposed to provide a ‘fine-meshed’ estimation of whether a case is suitable for VOM. This estimate is generally based on case files, telephone conversations with parties, and possibly background checks.

Generally, the VOM coordinators in my data set fell into one of two categories when it came to screening criteria. Around half of the coordinators did not vocalise any specific screening criteria for parties. This group generally tended to only point to other reasons than their own screening (for instance a decision by police lawyers) as factors, which could lead to a potential VOM case not making it to an actual mediation. I experienced the approach of this informant group as being similar to that of the informants I interviewed and observed in the Norwegian Mediation Service. I will get back to this point later in the article.

The other half of the interviewed VOM coordinators also primarily pointed to external reasons for a potential case not resulting in a mediation. However, this group did further mention criteria, which could lead to their dismissing a case themselves with the purpose of ensuring a safe and sound meeting environment. As an example, one coordinator explains how background checks are a part of the normal coordinator routine and further uses the access to background information as an argument for the police affiliation to the VOM programme:

VOMC: We can look it up in the systems. We usually do a background check on the parties we contact. […] The police can find this [information]. Other institutions can’t do that. Then they would need access to the police reports.

Two recurrent criteria mentioned by this group of informants were mental illness and insincere motives among offenders. A repeated reason for being aware of these criteria was to avoid disappointing or perhaps even re-traumatising victims. This is in accordance with the aforementioned statement of the recommendation for the screening process to – among other – eliminate offenders not found psychologically fit to participate in VOM. In relation to offender insincerity, several coordinators mentioned stalking cases as possible examples, and one coordinator highlighted, how the Danish Stalking Centre advises against mediation in stalking cases.

I had access to internal VOM case files from three police districts. For two of the districts (A+B) the files also included VOM cases, which had not resulted in a mediation. Most of these were not due to dismissal by the coordinator, but for instance due to one or both parties declining to participate. Yet, as the following table demonstrates, coordinator dismissal due to unfit offenders had happened in both districts, and the dismissals tended to fall into the two aforementioned categories: mental illness or offender insincerity. I will get back to the last district (C) later in this section.

In District A, I reviewed case files from a period of 12 months. In this period, 192 cases were registered by the coordinator. Of these 192 cases, 41 cases resulted in VOM and 151 cases did not result in VOM. Out of the 151 non-VOM cases, 7 cases were dismissed by the VOM coordinator due to unfit offender. Hereof 4 were due to offenders’ mental illness and 3 due to offender insincerity.

In District B, I reviewed case files from a period of 21 months. In this period, 519 cases were registered by the coordinator. Of these 519 cases, 117 cases resulted in VOM and 402 cases did not result in VOM. Out of the 402 non-VOM cases, 8 cases were dismissed by the VOM coordinator due to unfit offender. Hereof 5 were due to offenders’ mental illness, 1 due to offender insincerity, and 2 for unspecified reasons.

The review of case files in District A and B illustrate several differences between them as well as exemplify the diversities at district level in the Danish VOM practice at large.¹⁸ There is a large difference in the number of registered cases in the two districts, even when considering that the files from District A cover 1 year and those from District B 1.6 years. Yet almost the same proportion of registered cases result in VOM (District A 21%, District B 23%). In total numbers, the coordinator has dismissed almost the same number of cases due to unfit offender. However, the proportion of cases dismissed by the coordinator due to unfit offender differs considerably with 3.6% of cases dismissed for this reason in District A and 1.5 % in District B. However, even if the relative share of cases dismissed due to unfit offender differs in the two districts included in this review, the numbers do illustrate that dismissal of cases on this background does occur in both districts.

In line with this, in the following interview fragment the national secretary of the VOM programme talks about possible criteria excluding from mediation:

NSVOM: If it is someone who is mentally ill, we tend to say no, we cannot do this. Or if the offender’s motive for participating is somewhat questionable. If the first question is ‘will I get off with a lesser sentence?’ it is debatable whether it is reasonable to put an offender in the same room as a victim again. Because we do not want to re-traumatise the victim and say ‘oh, then you got “smacked” by the offender in VOM too’. So, this is what we look at.

The following interview fragments with three different coordinators are along the same lines. The first two focus on mental illness:

VOMC: This is why we have discussed how it makes a lot of sense to have some police officers, who can screen the parties. Regarding the safety at the meetings and such. There can be different things at stake, especially regarding – how do you put it – mental illness and such. It is simply reassuring to have some sort of feeling about which sorts of characters we are putting together for a meeting. Then we can work with matching the expectations regarding the parties’ intentions for agreeing to the meeting, so we don’t risk for one party to become more offended by meeting the other. This is why it makes a lot of sense to place the VOM programme in the police. We have a fairly good insight into the different cases, and we know when it can be really relevant and very useful for the parties to meet.

And:

Researcher (R): And those in the ‘no’-pile – are there any explanations as to why a mediation did not happen?

VOMC: Yes, we usually write that.

R: So, it could say something like ‘the offender said no’, or?

VOMC: Or that the offender is schizophrenic. We are not allowed to have mediations with mentally ill persons.

Mental illness and especially ex-mental patients are recurrent examples in Goffman’s Stigma. Goffman uses ex-mental patients as go-to examples of how information controlled by others can lead to an otherwise invisible stigma – like that of former mental illness – becoming visible and an access-limiting factor in the lives of the affected. In the case of the Danish VOM programme, it is the police who become controllers of this information. They have access to personal records potentially including information about former mental illness and they can decide whether the presence of such information disqualifies for VOM or not. In other words, in the era of an otherwise increasingly recovery ¹⁹ oriented mental health sector, police officers who have no professional prerequisites for estimating the current relevance of a former mental diagnosis potentially get to decide whether to evoke the Stigma or not. This – I believe – can be viewed as an anachronistic and problematic state of affairs. Yet, this is in line with the recommendation behind the legislative framework of the VOM programme.

The third interview fragment focuses on offender sincerity:

R: Do you often discard potential parties after a background check?

VOMC: Yes, it happens occasionally. […] It could be something like – let’s say the offender wants to be in a mediation. He wants to apologise. But you know that his motive for doing this is perhaps not that he really wants to apologise. He just wants [the justice system] to treat him less harshly. So if his intentions for wanting to meet the offended party are not all sincere, you need to be careful to make sure that the offended party is aware that this might be why he has agreed to meet. That maybe he is not really sorry.

R: How can you tell this in a background check?

VOMC: […] Well I would say, you cannot tell from a background check how this person would act and what his intensions are for entering this meeting. […] But it can give you an idea of what he has done before and how he has behaved in the past – when the police have had a hold of him at least.

I believe this interview fragment highlights the screening dilemma of the Danish VOM programme: casting for the role as an appropriate and sincere offender based on the offender’s previously recorded interactions with the police. Many offenders will likely not have the same attitude – or what Goffman (1959) calls facework – in situations where they meet the police, as in situations where they meet offended parties in VOM. In this way, the police might end up taking misguided considerations to the victims in the sense that they want to spare the victim of the interaction with what they see as an inappropriate offender, who perhaps would not come across as ‘inappropriate’ at all in a mediation setting.

So, in this regard too, the double role of the police can be viewed as problematic. As mentioned in the introduction, the Danish legislators have decided that VOM cannot substitute for a trial. Hence, in the case of criminal offences – regardless of whether the offender decides to participate in VOM or not – he or she still has to perform in the legal system. This can lead to the offender and the police being positioned in what Goffman (1959) calls discrepant roles. Both offenders and police have different roles to play in relation to a trial than in relation to a mediation. These different stages for performance can call for different strategic secrets as well as different facework that does not mix well. With the police acting as a go-between in Danish VOM there is a risk that offender face work, which might be fully appropriate in relation to different levels of the trial roleplay could be interpreted as a disqualifying trait of insincerity in relation to VOM participation.

In this way, an offender’s ability to apply confidence-inspiring facework in his or her interactions with the police and justice system can be decisive in future VOM castings. Notably, castings which neither the potentially recording police officer nor the offender are necessarily even aware exist.

However, a coordinator explains how neither an offender’s mental illness nor a notion that the offender might have ‘insincere’ motives for participating in VOM necessarily means that the case will be rejected:

VOMC: Then we will have a long talk with them. These initial conversations allow for us to find out what their intentions are. We try to work with matching the expectations of the parties and see ‘why do you want to meet? What do you hope to gain from it? And we have held mediations where the one party is perhaps a little mentally ill. Then we have just prepared the other party that ‘you should know, this meeting could become a little – this and this – because he has a mental illness, which makes it difficult for him to… etc.’ […] And if they say that they want to meet anyway we will take the circumstances into consideration. You just need to set the scene, so people know what they are entering into. And so, they do not risk sitting in the meeting and becoming more offended, if the offender for instance is cold and cynical and does not care about them. Then it would be putting them through something, which is not very pleasant. Unless they want it themselves and still want the opportunity to say some things to him. But, by matching the expectations we can have some really good meetings anyway. So, this is why in order to match the expectations we have these thorough conversations to find out what kind of characters they are.

Thus, according to this coordinator with the right preparation the parties might still benefit from a mediation, even if the offender did not necessarily pass the initial casting with flying colours.

My review of case files in District C provides examples of this. In this district, I only reviewed files from cases where a mediation had taken place. While I cannot tell how many cases might have been rejected due to unfit offenders in this district, out of 177 completed mediations, 4 were stalking-related and 1 had an offender described as mentally ill.

The options of the discarded

So, what about those parties who due to conscious or unconscious screening processes are not offered the option of participating in VOM? The National VOM secretary acknowledges the significant organisational challenges of the Danish VOM programme in relation to ensuring that all potential parties are actually informed about the option of VOM. Yet, the secretary also claims that the public awareness of the VOM programme is increasing to the extent that interested parties would ask for VOM themselves, even if they had not been directly informed about VOM by the police. Or – put in Goffmanian terms – had not received the casting call.

NSVOM: […] and then we were so lucky as to have Danmarks Radio [Danish Broadcasting Cooperation²⁰] make three programmes. […] Documentary. Really good. Very decent and nice, right? And station 2 has also been at it a few times with flashing police lights and all that. That was more… well okay. So, we have done a lot to make the public aware. We have created a separate homepage. It is not hidden under the main police webpage. […] Our coordinators in the districts they put themselves out there, telling about the VOM programme at length at any place – fairs and such. […] We are part of Folkemødet [The People’s Political Festival²¹] in Bornholm with a stall telling about VOM. And by the way, it is my experience from last time we were there that people knew what [the VOM programme] was. They had heard about it. They had seen those programmes. And they knew someone who perhaps had been in contact with it. […] So, I would say that we have managed to make this offer [the VOM programme] known to the public. And that helps me sleep at night. Because I can say that if the citizen wants to be in a VOM meeting this citizen will know that he or she merely has to contact the police and ask if it is possible.

In the following interview fragment, the national secretary similarly reverses the responsibility regarding potential VOM parties, which might – most likely unknowingly – initially have been screened unsuited for VOM:

NSVOM: But it is not just the policeman who decides [whether a mediation should be held]. Because it is voluntary for the parties. […] And as a party you can come to the police and say ‘I have had this case. Can I be part of a mediation?’ And then the police should check if the other party wants a mediation. And if both parties want it a mediation will take place.

I believe this to be a problematic stance regarding a service offered by an institution, which is to adhere to the key standards for public administration, predictability and uniformity. Yet rather than emerging from an ideological position, I also believe this stance to be born from the position of powerlessness, in which the Danish VOM programme is left regarding legislative, organisational and resource-oriented challenges.

In the following section, I focus on the screening practices of the Norwegian Mediation Service as these differ noticeably from those of the Danish VOM programme.

Screening in the Norwegian Mediation Service

According to my findings, the overall approach of the Norwegian Mediation Service seems to be significantly more incorporating than that of the Danish VOM programme. Not only does the Norwegian Service complete around 12 times as many mediations per year, it also does not have a victim-oriented point of departure. Rather, the focus – also legislatively – is on the needs of the affected parties alike. This includes penal cases, where an information video on the service webpage²² focuses on the needs of the parties rather than those of the victim – as opposed to the formerly mentioned internal promotion video for the Danish VOM programme.

In an interview, the head of the Oslo Mediation Service noted that the term ‘mediation’, which could possibly invoke notions of the quest for some kind of common ground, might create inappropriate connotations in the case of some offences. Hence, the head stated, the current trend in the service is to talk about meetings rather than mediations. This is also visible in another information video about the Service entitled ‘Nothing to mediate over but perhaps something to talk about’.^{23 24} The topic of this video is a rape case. It has been internationally debated whether sexual assault/rape cases are suited for mediation. This is also reflected in the practice of the Danish VOM programme, where – according to the national secretary – mediation generally only happens in such cases if one of the parties requests it themselves. In this way, the existence of a Norwegian Mediation Service promotion video focusing on rape in itself underpins the relatively wider portfolio of this service’s practice.

In Norway too, the Mediation Service has the option of dismissing a case if one or more parties are considered unfit for mediation. In contrast to the Danish programme, the Norwegian Mediation Service generally does not have access to Stigma-related information on potential parties. A case cannot be dismissed without involving the department head. The Oslo department head lists two possible reasons for such dismissal: heavy substance abuse-related and/or mental illness-related concerns. These reasons for dismissal are like those mentioned in the recommendation behind the Danish VOM programme. Yet, the practice of the Norwegian Service does not include a record-based screening of the parties so a dismissal would be based only on the contact between the Mediation Service and the party in question. In the three years the current department head has held the position, this person has only dismissed one case on this basis. In this way, judging from my data, in the Norwegian Mediation Service it is predominantly the ideals of impartial mediation which are applied for the screening of cases.

There are no national records of the total number of coordinator-dismissed cases in the Danish VOM programme. Yet, my data indicate a much higher dismissal rate here than in Norway: Even with approximately 12 times fewer cases, various reasons for dismissal are mentioned with several accompanying examples by the Danish management and several coordinators. Furthermore, the aforementioned analysed records for potential VOM cases from two police districts include a total of 15 cases dismissed by coordinators. As demonstrated, some of these reasons play well into the rationale of police typologisation, indicating that sometimes the police gaze triumphs over the ideals of impartial mediation in the screening processes for the Danish VOM programme. Yet notably these dismissals are also in line with the Danish VOM legislation.

Furthermore, in Norway I had the possibility to observe the screening process for the two restorative youth sanctions. These sanctions are based in the Norwegian Mediation Service and include for the young offender to participate in mediation.²⁵ In this way, I was able to observe discussions among professionals on the individual suitability of a restorative sanction in different cases.²⁶ The screenings for the restorative youth sanctions happen at Coordination Group Meetings. While the Mediation Service does not have access to personal information on the young offenders, in the Coordination Group Meetings other professions can provide such personal information when co-screening for these sanctions.

The professionals present at these meetings are: Mediation Service Youth Coordinator, Police, Prosecution Service, Children’s Services, and Probation Service. Other relevant parties such as Health Service and others can attend depending on local organisation (Konfliktrådet, 2019). Based on their different knowledge of the situation of the young offenders, it is for the Coordination Groups to decide whether these sanctions are an appropriate option in each case.

The Coordination Group meetings provided very interesting perspectives to my research, as the formal screening process in the Danish VOM programme is not very observable, since it is a process that largely happens in the head of the individual coordinator.

Notably, at the meetings I observed none of the professionals essentially went against offering a young offender the possibility of a restorative sanction, even if they had maybe vocalised their doubts that the young person would be able to complete the sanction and/or desist in the future. Yet, in the six Coordination Group Meetings I observed, there was a clear tendency of the youth coordinators being the main source of vocalising ideals of impartiality and of preserving a resource based focus, among other by focusing on the deed rather than the doer. At the other end of the spectrum, in these meetings, the police professionals were the most frequent to problematize the young offender, his/her background, and the outlook of the young person desisting from future offending. At some meetings, the professionals from social services tended take the same position as the police in this regard.