Flexible Rigidity and Caring Distance: How Discretion Works in Compulsory Care

The aim of this article was to develop the understanding of discretion in compulsory care for youth and the influence of emotional aspects in developing a collective discretionary practice. The empirical material consists of group interviews with staff at these institutions. The theoretical concepts of emotional labour and emotional energy are used to understand discretion and face-to-face interaction in this specific setting. The results show how the staff navigate between rigidity and flexibility in order to uphold institutional rules, whilst also being able to meet youths’ different needs in specific situations in a personal, yet not private, manner. Professionalism is understood as providing a caring distance, in other words, a temporary emotional involvement with the youth. The staffs’ reasoning and understanding of their work show how to manage emotional labour in a setting where the youth can be seen as both dangerous and vulnerable. The emotional energy is primarily connected to the long-lasting relation amongst staff and not to the youth, and hence, it is of primary importance to form a shared collective discretionary practice.


Introduction
Discretion is a way to handle work in complex organisations where detailed regulations cannot cover the whole practice. Discretion has been said to be the main component of professionalism (Molander, 2016), and uncertainty has been said to provide discretion (Goodrick and Salancik, 1996). Due to the innate uncertainty in social work, discretion is both necessary and situational. This article concerns discretion amongst practitioners working at secure accommodations for youth (compulsory institutional care). It adds knowledge on emotional aspects to the understanding of collective discretion, but also on how emotions influence the work.
These institutions are what Goffman (1961) called 'total institutions' and the power dimensions are obvious. Care in secure accommodations for youth comes with rigorous control and monitoring of activities and provides physical restrictions and discipline as tools for what Hasenfeld (2010) calls a people-changing practice. In this setting, staff continuously struggle with how to relate to the incarcerated youth and in which cases to follow the rules and when to make exceptions. Staff are supposed to act in a friendly manner and to create a good climate at the institution, which is part of the 'clinical attitude', Albert Cohen's concept for the collective expectations on how to act (Hochschild, 2012). Trotter (2015) has pointed out that working with involuntary clients means also handling the sense of betrayal felt when clients lie and still managing to build trust in relation to the client, and thus, it is a more tense emotional setting. Cousins (2020) argues that the involuntary setting promotes hostility and frustration from the clients, which on different levels affect the workers' personal feelings and reactions and lead to a focus on conflict management in the practice.
Earlier research has shown that staff in institutional care of youth see their personality as the primary working tool and emphasise personal competence over theoretical knowledge (Kallenberg 2016). Hill (2005) discusses the two main professional dilemmas of balancing control (between being rigid and boundless) in order to maintain professional authority and of balancing contact (between being distanced and private) and to be able to create contact. Staff can accordingly apply rules and regulations in different ways when conflicts arise at the institution depending on how they navigate between these dilemmas and their professional ideals. Franzé n (2017) claims that 'care' must be understood as something that is constructed in the interaction between the teenagers and staff at the institution, and that the actual methods are of less importance. For instance can humour be used by staff in interaction with youth in order to confirm and negotiate hierarchies related to authority, generation and age (Gradin Franzé n and Aronsson, 2013). Still, the staff are often ascribed a groupthink similar to what is said for inmates (Crawley and Crawley, 2012).
Working in close contact in a locked institution is demanding because staff should show emotions but control personal feelings. Nevertheless, colleagues are very important because they can provide chances to share experiences and feelings, what could be called an emotional zone (Crawley, 2003). Prison staff have shown to have the ambition to create a 'pleasant atmosphere on the wing', whilst still claiming to expect compliance and acceptance of their authority (Liebling, 2000), whilst belonging to a supportive working group enhance the ambition to continue working in prisons (Nylander et al., 2008). One's colleagues in prisons help to neutralise emotions at work by expressions like 'don't take things personally' and 'don't get sucked in' and help distance themselves from the prisoners (Tracy, 2004). Liebling (2000) also showed that the prison staff saw peacekeeping as their primary role. Stabilisation is essential in a tense situation, and peacekeeping is a way of controlling the emotion-laden situations. In a similar way, practitioners in institutional care for youth see violence and threat as part of the professional role and stress the importance of not showing fear in front of the youth (Andersson and Ö verlien 2018). In threatening situations, colleagues' emotional reactions and practical responses are, however, important for actually feeling secure, and might differ between individual colleagues (Andersson, 2020). When the interaction with and belonging to colleagues are so important, what happens with discretion?

Aim
The aim of this article was to develop the understanding of collective discretion in compulsory care for youth and the influence of emotional aspects in developing a collective discretionary practice. The empirical material was analysed following two empirical questions: How do practitioners in special residential homes for youth describe and relate to the teenagers they are working with? and How do they reason about their professional role and work in everyday practice?
Based on the stories told by the practitioners, we analyse how staff reason about how discretion works in face-to-face interaction in a compulsory care setting and how the staff reason about the values in their work.
The idea of collective discretion relates to a kind of groupthink and colleagues as supporters who help to neutralise emotions is closely related to Randall Collins' (2004) way of understanding emotional energy. He argues that emotions are not only 'the dramatic ones: fear, terror, anger, embarrassment, joy and so forth' (p. 105), but also 'undramatic, they are long-lasting, underlying tones or moods that permeate social life' (p. 106). These long-lasting emotions are what Collins calls emotional energy, and they develop through human intersubjectivity. When groups interact continuously, they develop a mutual rhythm and the participants become caught up 'in the swing of things', as Collins puts it (p. 108). This is how we can understand the collective of staff and their mutual reasoning about their work at special residential homes for youth. In addition, also Hochschild's (2012) concept of emotional labour and aspects as emotion management helps us to understand the stories told about the work because they allow us to see the personal aspects of working in this setting.
Before going into details on the results, we will present the theoretical framework and the material we use, as well as the special residential homes where we have collected the data. With support from secondary sources, we describe the homes' mission; the youth placed there, etc. We also include quotes from staff to give a more vivid description. The results are thereafter presented in two sections, followed by discussions and reflections on the study.
Discretion in face-to-face work Lipsky (2010) related discretion to the space given by laws and regulations. From that, the understanding of discretion is often reduced to being regarded as 'an all-or-nothing good thing, as a phenomenon that either does or does not exist' (Evans and Harris, 2004). Already Dworkin (1963) made it clear that being eligible to choose between solutions does not mean that you are free to make any choice. Evans (2013) has shown that professionals' ideas about how to relate to organisational demands and discretion include pragmatic elements related to their situation and that discretion, therefore, has to be understood from specific situations. Hupe (2013) distinguishes between different disciplinary understandings of discretion and their characteristics (see also Evans and Hupe, 2020). Depending on the perspective, the role of the discretionary actor differs. Here, we take a starting point in a sociological perspective, where the practitioner acts according to occupational standards, and the study concerns how these standards are seen in the discretionary practice. Evidently, as it is a setting with youth in locked up institution, the power dimensions are explicit. Molander (2016) distinguishes between discretionary space and discretionary reasoning. The discretionary space is the area 'where we can choose between permitted alternatives of action on the basis of one's own judgement'. The discretionary reasoning is the epistemic dimension and is 'a kind of reasoning that results in judgement about what to do under circumstances of indeterminacy' (Molander, 2016). Thus, discretion on the one hand means that there is room given for different judgements, whilst on the other hand, there are different ways to reason about what judgements are possible in specific contexts and settings. In this study, we focus on discretionary reasoning in an occupational group in compulsory care of youth. Whilst most studies of discretion focus on formal decision making by professionals, our contribution is to focus on the micro level, where lower-level workers make decisions on everyday issues in face-to-face work, and in this case with involuntary clients. Moreover, we centre the study on collective discretion. Rutz and De Bont (2020) point at collective discretion as group decision making, where the involvement of the others strengthens the decision that is made and thereby the action that is taken. Collective discretion is in practice controlled through peer review and adjusted according to rules, etc., by the workers who collaborate in practice. Still, management and rule makers also control the collective discretionary space because there are organisational frames for the collaboration between workers. Nothdurfter and Hermans (2018) argue that studies often connect discretion and professionalism but remain rather vague about how professionalism interplays with discretion. Focussing on low-level workers makes it even more important to clarify how professionalism is understood because the concept of professionalism is close to the conceptualisation of professions, whose definition in turn has been much discussed. Workers in social work and social care have the task of upholding the organisation's policies in close contact with the people in care, regardless of their level of education or professional status in any other sense (Lipsky, 2010). Staff in secure accommodations are seldom professionals as they most often have lower level of education and they do not have a distinct jurisdiction; they are rather the organisations' actors. Still, they have a certain discretion on a micro level, where they manage situations in everyday practice, and they can claim to be professional when they are doing good work. Lipsky (2010) includes, for example, prison guards in his concept of street-level bureaucrats. He claims that the situations where street-level bureaucrats work are often too complicated to be reduced to programmatic formats, and they need to respond to the human dimensions of situations. Moreover, discretion could be seen as having an emotional function for lower-level workers, as Lipsky (2010) says, '[it] promotes self-regard and encourages clients to believe that workers hold the key to their well-being' (p. 15).

Methods and material
The material used in this article derived from a research project on the practice in special residential homes' from the perspective of staff, young people and the social services. The project has been approved by the Regional Ethical Board in Lund, Sweden (dnr. 2015-98). The material in total consists of field visits to the homes, interviews with staff, youth and social workers from the social services, and online surveys to staff in residential homes and in the social services. The project was finished and presented in an overarching report in Swedish (Ponnert et al., 2020). Thereafter, further analyses have been made on separate parts of the material.
For this study, a separate analysis was made of six group interviews with sixteen employees at two special residential homes for young people under compulsory care in Sweden. Two or three employees participated in each group. Three interviews were held at both homes in March and August 2016. Of the sixteen interviewees, nine were men and seven were women. One of the homes was for girls, and here the female staff dominated the groups (six women and three men) and the other home, for boys, had more male interviewees (six men and one woman). The interviewees had a wide variation in work experience within these homes as well as previously. Six of the interviewees had been working for more than ten years in residential homes for compulsory care for youth, and those with the shortest experience from these homes had been working 1.5 years. All but one had some education related to the work, mainly in social care, and in some cases through shorter academic studies, but none of the interviewees held a higher academic degree. A few had shorter educations in basic medical care, such as nursing assistants.
The group interviews were conducted during the staff's normal working hours, and thus, they had to be kept to a maximum of one hour. The participants had been suggested by their managers and accepted to participate. Information had been sent out beforehand, and before the interviews started information about the aim of study and its context was given both orally and on a paper that was handed out. All participants gave informed consent before the interview started.
The group interviews can be described as formal field interviews. Frey and Fontana (1991) argue that this kind of group interviews is the most relevant when you aim towards grasping how groups reason together in their practice. Formal field interviews should be conducted in the specific field of practice and use semi-structured interview guides (Frey and Fontana, 1991). Our guide was based on eleven themes, and eight of these were formulated as statements that the participants should reflect on, based on possible differences between working at different units at the institution (e.g. 'Staff at acute divisions and treatment units have equal influence over the care of the youth'). The statements concerned the youth, the way they worked and their mission, their relation to the teenagers, and their own influence. Finally, they were also asked if they would like anything to be different, either for themselves or for the teenagers. The interviews were transcribed verbatim. The analysis was made through several readings of the transcripts by both of us, in order to identify themes. The data were then coded and sorted thematically. The main themes concerned their professional role and discretion, which is why they are in focus in this article.
In the presentation, the quotes are connected to the six groups. Because we interviewed in groups with the ambition of grasping how the staff at the specific units reason, we have not separated the sayings from different persons. The groups are named A for the home for girls and B for the home for boys and numbered 1-3. This way, we also keep the interviewees anonymous. No information that can reveal any specific person is included in this article.
We sought support for the analysis on the one hand in Hochschild's (2012) work on emotional labour, and on the other hand in Collins' (2004) work on emotional energy. Hochschild's (2012) concept of 'emotional labour' covers the tension between the private sphere and performing the tasks required by the job. She states that 'it is the pinch between a real but disapproved feeling on the one hand and an idealized one, on the other, that enables us to become aware of emotional labor' (Hochschild, 2012, p. xi). Pointing at the emotional aspects of work adds an important understanding to the situations where work is performed. In their reasoning, the interviewees revealed the emotions embedded in their arguments. Collins' (2004) interaction ritual theory can help to understand how the workers cope. He describes the central mechanisms as a mutual focus of attention and a high degree of emotional entrainment. Collins' concept of 'emotional energy' is related to what Durkheim called 'moral sentiment' and includes feelings of what is right and wrong. Emotional energy exists in a continuum from confidence, enthusiasm and good self-feelings at the one end and depression, lack of initiative and negative self-feelings at the other end. Collins (2004, p. 108) compares this with the psychological term 'drive', with the difference that emotional energy has no direction.

The homes and the youth
The practice in focus here is 'special residential homes' for youth in Sweden where young persons are placed after a decision by the Social Welfare Board and the Administrative Court based on an investigation and suggestion by the Social Services. One of the key components in the decision is that the care needed is not possible to give on a voluntary basis, which is why the compulsory measure is used according to the Care of Young Persons Act. Placing a young person in a special residential home is 'a last resort' and is used when no alternative is seen as possible (Emerson, 1981).
There are twenty-three homes of this kind in Sweden, all run by The National Board of Institutional Care. All together these homes have room for 659 young persons, of which 90 per cent of the places are or can be locked, which means that those who live at these homes cannot leave them without permission (National Board of Institutional Care [Statens Institutionsstyrelse], 2019). Each year around 1,000 young persons are admitted to special residential homes, whereof two-thirds are male and one-third are female. At intake, 93 per cent of the admitted are put under compulsory care due to the risk of harming themselves through drug abuse, criminality or other socially destructive behaviours (3 § the Care of Young Persons Act). The young persons are mainly teenagers, almost half of them are sixteen to seveenteen years old, a fourth are fifteen years or younger and the remaining fourth are eighteen to twenty-one years old. On average, the stay at the home is around six months. After the stay, two-thirds continue on to care or treatment in some other form (National Board of Institutional Care [Statens Institutionsstyrelse], 2019). At the homes, the staff have the authority to lock the section, but also to use other compulsory measures, such as isolating the young persons in solitary confinement in certain situations or demanding urine samples for drug analysis. These formal measures are to be used in a setting where staff also is supposed to create a positive climate and good relations to the teenagers.
These homes are not prisons, still, the teenagers have most often been involved in criminal activity and they are incarcerated. Whilst a prison sentence in Sweden is always fixed in time, the Care of Young Persons Act only states that there has to be a new decision every sixth months and that the care cannot continue after the person turns twenty-one. However, compulsory care and treatment do not have to be carried out in these homes, and there are also other alternatives, which is why the time for the compulsory care and the time at the institution do not have to be corresponding. Research on this practice shows poor results. Several studies show that it is hard to form working alliances between staff and youth and that this kind of practice shows strategic alignment rather than the results that are formally mentioned (Hill 2005;Polvere, 2014;Enell, 2015). One barrier is that special residential homes provide an artificial environment with artificial relations over a limited time (Ahonen, 2012).
When staff at special residential homes in our study describe their workplace, they switch between talking about the home, the activities and the teenagers placed there. Descriptions of the teenagers tend to focus on their bad situation in society. Statements, such as that they 'come directly from everything in chaos and misery' (A3), '..have lived a life with drugs and turned day into night and night into day' (B1), or 'have been close to dying several times' (B1), are repeated, or as concluded in one group: 'Everyone comes here, those who have abused children, raped, there are those under prosecution for murder and everything, then there are drugs, then crime and everything, and disabilities. So we get it all' (B2). The compulsory and locked setting creates a specific environment and newly arrived teenagers have to adjust to the situation. A metaphor often used by staff is 'landing'. One story goes: It is tough for many of them, those who haven't been in placement before, you know, for them it is really tough. I have been talking with them here, some, for three, four hours. They have been laying on my lap crying, as they don't dare to go out and meet the others. And then I have to walk along with him, but it eases up after some hours (B2).
The story about landing is a story of the institutional setting, where the teenagers are to adjust to the governing rules, but it also includes their vulnerability. There are also stories about taking the teenagers 'in secluded care, until they are reachable' (B2), which also relates to sayings like: 'Where else should they be? Should we let them out among the others? No one wants them' (A1). Therefore, in their descriptions the teenagers are regarded as dangerous and without self-control as well as vulnerable and excluded. In different ways, both arguments justify a need for locked care and show how staff handle their emotions by taking the measures used to be inevitable.
The primary role of the home is to create a safe and calm environment, which is the basic precondition for everything else. When staff talk about this, they ascribe this as a need also for the teenagers: 'That is what they really want, peace and calm' (B2). The process for the teenager therefore starts with adjusting to the institution's structures. 'The daily structure is incredibly important for these youngsters, as they haven't had it before' (A3). In this structure, there are times set for meals and activities and there are rules to be learnt by newcomers. 'Later on, in the treatment section, there are programmes and [the teenagers] "are more responsive towards [our programmes] if they have first landed"' (A3).
When the teenagers leave the home, the staff will most often not get any information about what happens to this person. A new one is taken in and the system continues. No interpersonal relations are supposed to last longer than then and there, and staff also argue, 'You cannot manage to follow the process, you have enough work in dealing with what is happening in your section, so being part of the next step in addition, that is something you don't want' (B3). Through such statements, they also show that their emotional engagement with the youth is situational and is not a personal emotional engagement. The emotions are part of the work, just as the teenagers are. Still, the interactions with their colleagues continue over a longer time, why the bonds between colleagues in general are stronger than the bonds between staff and teenagers.

Findings
The main themes for this study, flexible rigidity and caring distance, developed during the analysis of the staff's reasoning about their role and Flexible Rigidity and Caring Distance Page 9 of 18 work. The concepts are contradictory because of the contradictory practice from which they derive. The concepts mirror the practice. We relate caring distance to being human at work and flexible rigidity to doing the actual work.

Being human through caring distance
In the groups, the theme of 'being personal, but not private' reoccurred. This was about how to relate to the teenagers, but was also about being 'yourself'. In these statements, the staff talk about relating in a friendly way to the teenagers, but also about keeping distance, to protect yourself as well as the teenager from forming too personal bounds. The fact that the relationship should not be too close manifests in statements such as 'We are just a milestone in their life. They will be moving on. You should never befriend them. Be kind, but firm, I always say' (A2). The distance approach includes not telling too much about yourself as well not asking the teenagers too much (unless you have a specific treatment assignment) to enhance integrity for teenagers in vulnerable situations, as one person put it: 'Of course we ask them how they feel and we can sit in the room and talk to them. But it is not like we start asking about her childhood' (A2).
Being able to connect and form relationships with the teenagers is one of the essential parts of being professional in this context. This is emotion management in practice, where acting is emotional, but the emotions should not be personal but part of the role. One of the interviewees said: 'Being private is a caveat; you should not step over that line. You should not reveal yourself in front of the teenagers and you should be professional, but still accommodating' (A3). The distance and formalities are the important characteristics of the professional role. So is the ability to be human, to make relationships in the specific situations and to be flexible, and one of the interviewees said, 'We are damn good at creating relationships' (B3).
It is not only the relationship with the teenagers that matters, being collegial and continuously communicating and collaborating with one's colleagues is just as important because it upholds the collective emotional mood. The risk of violence reoccurred in the stories told and tended to be a way of manifesting the importance of collegial trust. Here, the importance of belonging to a collective stands out. One groups discussed this as follows: it is a tough environment to be working in; very much. Because there is a risk here, there are violent situations in this, eh, you don't really have to be involved yourself, it could be between two teenagers; yes, yes, sure; but, you have to interrupt; that is how it is; so. . .; and we have this joke that often, often we know more about each other and how we react in different situations than those at home [know about us]. That is because you work so close together (B3).
In that way, being professional is about distancing from the human interpersonal interaction and focussing on the fact that you are a part of a system, even if you are a human in face-to-face interaction with teenagers. As one put it: 'I am a representative for a public authority, no matter what I do. . .Like when I answer the phone, you answer correctly, your section, and name, and so on. That is professionalism as I see it' (B1). Going to work means going into the role, which might also include what Hochschild (2012) refers to as surface acting, where you need to display certain emotions without actually feeling them. 'Of course, we are humans. Sometimes we do have bad days. But I hope the kids don't notice' (A2). Being human thus also includes a forgiving approach towards yourself and colleagues if you should fail to manage emotions in a professional way.
Doing a good work is discussed as the part of the flexible caring relationship that is supposed to develop with each teenager, and knowledge is built up on how to deal with youth who often find it 'very hard to trust adults' (B1). This is sometimes described as fruitful in challenging situations: We can have a hysterical boy who threatens to smash windows, but if you then have a relation, it might be enough if you just approach him, put your arm around him. That's enough, he calms down. And if you don't have that relation, well, then that chair goes through the window (B1).
The quote above shows how staff talk about using what Hochschild (2012) calls emotion management. The emotions expressed by the teenagers have a signalling function that the staff believe in, so that they manage through deep acting, showing kindness and closeness. A story about flexibility reappears in the staffs' reasoning, showing the use of the right kinds of emotions and managing to take control and achieve a calm situation. This appears to be an ideal in their work.
In the stories told about being professional, the staff show an image of their collective emotions by talking about the importance of being personal, but not private, and by showing feelings, but not connecting other than in a professional way. They shared their different views and actions in order to find collective ways of keeping calm and handling the situations that arise. This leads us to Collins' (2004) way of talking about interaction ritual chains, where groups form ritualistic ways of interacting. The group of staff strengthen their emotional energy by developing collective ways of acting, and by joking, but also knowing that you will get support from your colleagues when you need it.

Doing the work with flexible rigidity
Working in special residential homes for youth is described as having a watchful eye all the time, as well as relating in a positive way to the boys and girls living there. You have to be strict and uphold the rules, but you also have to be flexible and open for making exceptions. When the teenagers enter the home, 'It is all about being sure. Did she bring drugs? What has she taken? What has she been through? Does she need to see a doctor?' (A2). During the first days, the work could be related to very basic everyday things like initiating and monitoring eating, showering, tooth brushing, sleeping. As time goes on, more and more conversations can be held, and the staff talk about having motivational conversations.
Balancing flexibility and rigidity, and closeness and distance, is important for taking care of the situations that arise and for upholding some kind of stability in an environment where rapid changes occur. 'As when I went for lunch, everything was calm, and half an hour, 45 minutes later, something has happened, something small that someone has done or something. It changes very quickly' (A1).
A key issue is to maintain unity amongst the staff and to uphold the collective emotional labour and a collective mood in order to develop a solidarity that opens for flexibility and the use of the discretionary space in similar ways. The group of staff is the main source for charging emotional energy, and conformity amongst staff increases the level of anticipation and thereby stability. It is about being able to take on a role and function together, despite personal differences, to keep up the work and to create a good ambiance. This was discussed in one group: Professionalism, it is our differences, and that we work together, I would say. Maybe we do not do things exactly the same, but we are really good at getting it to work out there. Yeah, exactly, as professionals we are doing the same, even if we have different views. And we can discuss it, when they can't hear.
If you and I have something to discuss concerning the girls, it is professional not to talk about it in front of them. Well, then I also think that we have different responses towards the girls; you have yours, and I have mine (A1).
In this discussion, they say that they acknowledge their differences and that they have a 'back-stage' where they coordinate their actions or can charge the group with emotional energy. This is a part of the institutional emotion management where the preconditions for deep acting are formed. In a collective discretionary room, the focus shifts from compliance (connected to individual discretionary room) to outcome, and experiences in the team can be used to adjust the rules (Rutz and De Bont, 2020, pp. 289-290). Collins (2004) talks about this as a mutual focus of attention that produces situationally dominant emotional moods. The interviewees describe how 'sometimes, a decision can be made in the specific moment' (A2), and this implies that decisions might not always be so substantiated. Nevertheless, because it is important for the staff to stand up for each other and try to hold a straight line, 'You have to keep on top of that, and so does the rest of the staff' (A2), they said. These are the kinds of situations where the collegial exchange is important for developing and upholding the collective emotional mood, and thus, discretion also becomes a collective tool in line with the group's sense of what is OK and what is not. This is the basis for the feeling rules in the group (Hochschild, 2012). Shifting between rigidity and flexibility in relation to youth and different situations can also be understood as a way for staff to navigate within a collective discretionary room where freedom and control tend to merge and 'discussion on when to use the discretion and when to adjust rules are all in the hands of the team' (Rutz and De Bont, 2020, p. 289).

Discussion
The aim of this article was to develop the understanding of collective discretion in compulsory care for youth and the influence of emotional aspects in developing a collective discretionary practice. From the stories told by staff at special residential homes for youth in Sweden, we have shown that they elaborated on their collective emotional mood. They talked about using their flexibility to create a positive ambiance and thus defuse threatening situations and prevent hard feelings. This is how they reason about using discretion in their micro-decisions. Because they talked about themselves as a group of employees, they also claim the right to be human in the interaction-a claim that no one opposes because it is rather a given part of the role. Being human and interacting with the teenagers facilitates the day-to-day work, which is important for the institution's legitimacy.
Research does not show that this kind of care is effective, but still it is used because the least that could be done is to reduce the negative aspects of the incarceration. From the interviews, but also from other studies (Hill, 2005;Kallenberg 2016;Liebling, 2000), we know that staff in institutional care of youth see their personality as the primary working tool and that they emphasise their personal competence over their theoretical knowledge. When staff use their discretion in order to make the stay at the home more humane, the practice as such upholds its legitimacy. However, they emphasise the importance of not being too personal, or not being too human, you could say. Thus, they uphold a service work position where they use deep acting, and they really do think that they care for the teenagers. The care and the interest for the specific individual teenager is, however, situational. The moment one leaves and another comes, the humanistic approach gets another teenager to engage in. This way it is obvious that this human service work is more of a 'service work', where emotional labour is practiced, than 'human work' where the individual persons are acknowledged. This situational care approach is also a prerequisite for successfully engaging in emotional labour because too much engagement could result in burnout (Hochschild, 2012, p. 187).
The emotional aspects come from the human interaction. The teenagers are both dangerous and vulnerable, with the dangerousness as a frame, and the staff focus on the vulnerability and seek to reach the teenagers on a personal, emotional level without being personally involved themselves. Hochschild (2012) calls this 'display work', where emotions are shown whether they are felt or not. The collegial setting helps to create an environment where emotions are managed so that they refer to the work, not to the persons, and in the collegial exchange they develop their 'feeling rules' that develop into an institutionalised emotion management that forms 'customs designed to manage the human feeling that threatens order' (Hochschild, 2012, p. 50). This feeling could either be the anger and frustration expressed by a teenager or the overly warm emotions a member of staff might feel towards a specific teenager. The feeling rules rely on a caring distance where the rigidity is flexible. How does this collegial setting develop? Collins (2004) argues that the central mechanisms for emotional energy are a mutual focus of attention and a high degree of emotional entrainment.
We have added the influence of emotions to the understanding of collective discretionary reasoning. More research is needed on discretionary practice and emotional aspects in micro-decisions in practice, for instance, on what happens in a more open and voluntary-based setting or in work with other groups of clients. There are many questions to be answered before we can know how the emotional aspects in general influence discretionary practices, but we do know that there are reasons to include the emotional aspects in trying to understand why people act the way they do at work.

Methodological limitations
This study is based on group interviews with staff, which limits the understanding into what they are saying, not what they are doing. In the full project we had several other sources, worth mentioning is the interviews with youth and some observations at the homes (Ponnert et al., 2020). These have served as a framing for our analysis, but also a validation of the statements in the interviews. Nevertheless, a more explicit and thorough ethnographic study would allow a more detailed information and developed analysis. A study similar to Liebling's (2000) prison ethnography and with focus on staff use of collective discretion would further the knowledge.

Implications for policy and practice
Compulsory care sends contradictory messages. In compulsory settings, used as a last resort, care that presupposes building interpersonal relationships is ideally hard to implement, or even impossible, as the power dimensions are obvious. Still, this kind of institutions exists, and striving to do at least the best possible in each situation is valuable. Even if the young persons are there primarily for being under control, they should have as decent conditions as possible. Having staff preventing violence and following the set rules might give some sense of protection. Meeting people who are kind and respectful might affect the young persons' selfesteem.
We have pointed at the staff's engagement in managing each specific situation in ways that promote stability by being friendly and supportive. The collective discretionary practice based on emotional ground is essential for this kind of institutions; it makes the everyday life bearable for all parties. It does not solve the problems for the youth placed there, but it can at least make them cope. This might be 'best possible practice' in that kind of setting and ought to be acknowledged and developed as such. Standardised treatment programmes and named methods might be good complements in some cases. Nevertheless, the everyday life goes on and is essential. This kind of institutions has shown strong evidence of being harmful. As long as they are used, the work that is possible to perform should be acknowledged. Upholding a calm and protected situation is an important kind of harm reduction when young persons, children, are being locked up.

Conclusion
Although the discretionary space in institutional care is formally granted to the staff as a team, it cannot formally be decided beforehand on how to deal with all of the micro-decisions that might arise in individual meetings with individual youth. In institutional care, the composition of staff members in the section is constantly changing during the day as different workers change shifts. Although there are shared feeling rules amongst staff in how to understand and interact with youth, their own personalities can be seen and used as a part of the individual discretion in concrete interaction with youth without endangering the emotional energy and loyalty amongst the staff as a team. This results in a 'rigid flexibility approach', where rigidity is connected to organisational aspects and to talk about the importance of upholding equal structures and rules, whereas flexibility is needed in face-to-face interactions between individual workers and youth in specific situations. Providing a 'caring distance', where the emotional engagement with youth is perceived as real but situational, can be seen as part of coping with this professional dilemma.

Funding
This study was funded by The Swedish National Board of Institutional Care, Dnr 2.6.1-647-2014.