Social workers as relational actors: descriptions of care order preparations in Finland

ABSTRACT This article examines what social workers do when preparing the removal of a child into public care in statutory child welfare. The focus is on how social workers describe both their ‘doing’ and themselves in care order preparations. Care order preparations take place at the crossroads of the different needs and rights of children and their parents as well as those arising from professionalism, law and bureaucracy. They involve the use of professional and public power given to social workers. The thematic analysis is based on interviews with Finnish social workers (29) in which they describe a preparatory process. The analysis focuses on the descriptions of ‘doing’ and those of the ‘doer’. Talking, writing and coordinating form an essential part of social work in the descriptions of care order preparations. When describing their work, social workers speak about themselves as relational actors and very little as agents of the statutory, bureaucratic system. This reflects the non-adversial nature of the Finnish child welfare system with its emphasis on consensus as well as the current dominant discourse on what ‘good’ social work in child welfare is. Given the nature of care orders and the restrictions on family life likely to arise as a result of the preparatory process, other approaches to describe ‘doing’ care order preparations could be expected as well.


Introduction
The question whether social work is -or should be -influenced more by professional theories and values than legal and bureaucratic norms has for a long time surrounded child welfare with new topics such as standardization being occasionally brought into the debate (Payne 2000;Burns et al. 2018;Ponnert and Johansson 2018). These questions are constantly relevant for child welfare. Working in this area requires particular professional skills as social workers look after the needs, wishes and rights of children and families in vulnerable and often conflict-laden situations; at the same time, practitioners need to consider the legal and organizational norms of the statutory system. The role of legal and organizational norms becomes even stronger when the threshold for the removal of a child is met and the preparations for a care order are started (Egelund 1996).
Social workers' responsibilities in care order preparations are regulated in a variety of ways, with considerable differences across countries. In Finland -the context of this study -care order preparations are regulated by legislation which gives social workers a relatively broad area of discretion to carry out the preparations (Berrick et al. 2015;Pösö and Huhtanen 2017;Falconer and Shardlow 2018). In addition to this professional discretion, social workers in managerial positions may make the decision to take a child into care without any involvement of the court, in this respect differing from other Nordic countries and elsewhere (Burns, Pösö, and Skivenes 2017;Pösö and Huhtanen 2017;Hultman, Forkby, and Höjer 2018). Consequently, social work in Finnish child welfare includes both tasks based on the profession and its theories and those relying on public authority, described in more detail in the following section. Although previous research has highlighted the obstacles social workers experience in their practice, such as time pressure and heavy workloads, (e.g. Mänttäri-van der Kuip 2016; Skivenes and Tonheim 2016), very little is known about what they actually do when they prepare a care order.
In this article, we are interested in the topic -often taken for granted -of 'doing' care order preparations. In care order preparations we examine how social workers describe their 'doing' and themselves as 'doers'. Social workers' descriptions are seen as forms of story-telling taking place in an institutional context (Polletta et al. 2011). When studying social workers' descriptions, we broadly make use of the concept 'institutional selves' (Gubrium and Holstein 2001). Institutional selves, in this case those of social workers doing care order preparations, are self-constructions lived by, storied about and embedded in institutional settings (ibid.). The focus on doing and doers thus informs about social work at this particular juncture of professionalism and bureaucracy in a type of a child welfare system which, in international comparisons, is known to be oriented towards family services and children's rights (Gilbert, Parton, and Skivenes 2011) and to employ a professionalized decision-making model (Hultman, Forkby, and Höjer 2018).

Care order preparations as an arena for social workers
Social workers constitute the key professional group in child welfare when the need to take a child into care arises (Burns, Pösö, and Skivenes 2017). They assess the child's and family's situations and make decisions regarding the services they need. Social workers need to balance between working in partnership with the families on the one hand, and using public power over them on the other (Buckley 2003;Kettle 2018). The views and interests of the parents and children may differ from those of the social workers, and the views, interests and rights within the family may be in conflict. It is not uncommon to talk about 'involuntary' or 'hard to engage' clients in child welfare to highlight the differences in service users' interests (Trotter 2008;Thoburn 2010).
Social workers are guided to encounter the complexity of tasks, uncertainty and possible conflicts in different ways. Ponnert and Johansson (2018) recognize that juridification and standardization are two evident yet different approaches to regulating present social work practice. In their view, 'a major challenge for professionals in contemporary child protection practice is to be able to navigate between different legal dimensions, juridification (hard law/legal discourse) and standardization (soft law/organizational norms), without losing their professional characteristics and accountability' (Ponnert andJohansson 2018, 2033). Despite attempts to standardize practice, assessments and judgements about families' situations tend to differ -not only across but also within countries (Benbenishty et al. 2015). The individual reasoning of professionals is obviously not easily unified.
Attempts to unify practice are also inevitably challenged by children and parents. There is no fixed category of 'clients', either involuntary or voluntary (Hall et al. 2003;Trotter 2008). Children and parents have different interests and wishes, which may change over time. If care order preparations take a long time, their situations and needs for services might change. Social workers need to react to these changes. Consequently, at a certain point, social workers may work in partnership with a family and at another point, the professional expectations may be in conflict with the family's or children's wishes.

The context: care order preparations in Finland
The role of a social worker in Finnish child welfare is shaped by the term 'a social worker responsible for the child's matters'. This term was introduced in Finland first in a narrow meaning in 1990 and then broadened extensively by the present Child Welfare Act in 2007 (HE 252/2006vp). The aim is to guarantee that every child in child welfare has one social worker who takes the overall responsibility for her/his matters. Both legal and bureaucratic as well as psychosocial elements of child welfare are thus integrated into the profile of a social worker. The role is attached to a child and not to a family. A social worker needs to make child-specific assessments and decisions, carry out care order preparations and be involved in court processes if necessary. Here, the need to balance between legal, bureaucratic, professional and psychosocial roles, responsibilities and frames is particularly clear.
According to child welfare legislation, a social worker responsible for a child is obliged to start care order preparations if the threshold for a care order is met (Child Welfare Act 417 2007, 40 §). The care order preparations are carried out by two social workers, one of them being the child's 'own' social worker, who may utilize the support of a multidisciplinary team. The preparatory process involves meetings with children, parents, and other professionals. Social workers draw up a client plan for the child and the parents, and they assess the network of people close to the child in order to find out whether the child could be accommodated with them. Further, social workers invite expert opinions and make arrangements for matching the child with a substitute home. The preparatory process is not time-limited unless the process has been started by an emergency removal of the child. In that case, the preparations must be carried out within 60 days.
The care order preparations conclude in a formal administrative hearing in which the child -if aged 12 or over -, the custodians and other people close to the child are officially heard. The aim of the hearing is to present the family with all the relevant information that has been gathered and used in preparing the care order proposal and the proposed substitute care placement and to hear their views on the care order proposal (Pösö and Huhtanen 2017). If the parents (custodians) and the child agree to the proposed care order and placement, the decision to take the child into care will be made by a social worker in a managerial position within the municipality. If one or more of the parties object to any part of the proposal, the decision will be made by the administrative court. In both cases, the legal implications of the care order decision are the same. The care order decision restricts parental rights to a great extent. However, cooperation between parents, children and the authorities is required by the Child Welfare Act while the child is in care.
The majority of care orders -1317 out of 1778 -were based on consent in 2018 (Lastensuojelu 2019). This means that the vast majority of care orders are prepared and decided upon in the social work domain, with social workers taking care of the preparations and eventually making the removal decision. Equally, in objection-based care orders, social workers are in charge of the care order preparations but a juridical body makes the decision. In both types of care orders, the care order preparations carried out by social workers are imperative.

Data, methods and limitations
We studied care order preparations as told to us by social workers and focus on the descriptions of 'doing'. We treated these descriptions as the presentations (and construction and re-construction) of the institutional selves of social workers (Gubrium and Holstein 2001), that is to say that the descriptions of 'doing' inform us about the 'doer'. Self-construction may lead to multiple institutional selves, which are of particular interest in the context of Finnish child welfare where practitioners have a certain amount of freedom to shape their practice. The analysis is based on interviews with social workers responsible for children's matters. The interviews were carried out as part of a larger project on care order decision-making, including also other material, such as case files, recordings of hearings and stakeholder views. The design for the study was planned together with social workers in three large municipalities (all of them having a population of more than 65 000) with a readiness to cooperate in a research project. The co-planning, following from previous joint projects, ensured that data collection was feasible and relevant from the practice point of view. Social workers especially wished to include free-format stories of care order preparations in the design as that format was seen as giving them freedom to speak in their own terms. The study was ethically reviewed by Tampere University and the municipalities gave their permission for interviews and other types of data gathering.
The social workers were invited to tell the story of one care order preparation process from the start of the preparations to the administrative hearing. The invitation letters were given after the researchers had met each of the five teams personally to inform about the study and to discuss the practicalities. We provided invitation letters to the team members and asked anyone interested to contact us to arrange the time for the interview. We encouraged them to tell the story as they saw it from the point of view of their practice ('professional narratives') as asking someone to tell a story is a way to search for an understanding of the teller's identity constructions (Riessman 1993;Larsson and Sjöblom 2010). First we asked the social workers about when the care order preparations had started and when the administrative hearing took place, illustrating this by drawing a time line on a piece of paper, and then asking them to tell what happened between those two points. This part of the interview aimed to encourage free story-telling, with only a few questions and interruptions from us as interviewers. At the end of the interviews, we asked a set of questions related to the consent and objection relating to the care orders.
The social workers (29) took part in the interviews on a voluntary basis, picking the case based on their own choice. This resulted in 30 stories which were recorded and transcribed. Three social workers narrated two stories each and two cases were told jointly by two social workers as they had been equally involved in the cases. When asked how they chose which story to tell, many responded that they told the story of a case which was recent (fresh in their memory) or one that had been quite unique in their professional careers. The majority of the social workers, aged 28-57, all females except one, were very experienced in child welfare: 16 of them had worked in child welfare for more than 5 years and had been involved in more than 10 care order preparations each, while the 13 interviewees with less experience had been involved in several care order preparations as well. Fifteen of them worked in the largest municipality and 6 and 8, respectively, in the smaller ones. Of the care order preparations in the data, 21 resulted in consent-based care orders and nine in objection-based care orders. The care order preparations lasted for 2.2 months on average and a typical starting point for the preparations was an emergency placement. The families and children had already been known to the child protection agencies much before the preparatory process began due to the misuse of alcohol or drugs, mental health issues and children's norm-breaking behaviour. The average age of children in the stories was 11, ranging from newborns to seventeenyear-olds. 18 children were 12 years or older, meaning that they were included in the administrative hearing process as parties whose consent or objection would direct the decision-making process.
The stories were rich in descriptions of the families but focused less on describing the work done by the narrators, a finding in its own right. In the analysis, we followed the principles of systematic coding (Coffey and Atkinson 1996), and collected, first, all descriptions of 'doing' (any sentence or paragraph saying something about the social workers' doing) in a separate document, and categorized them into thematic codes. The analysis of the descriptions falls into three categories of doing: talking, writing and coordinating. As they included a relational element, in the second round we looked at what kind of relations were spoken about with respect to the categories of doing and how they influenced 'doing'. This analysis falls into two categories, which will be presented in the second part of the findings: working in partnership with the parents and children and working in troublesome contact. Our interpretations of the institutional selves of social workers as 'doers', elaborated in the discussion section, are based on these two analytic rounds.
Three researchers, the authors of this article, were involved in the interviews, with one of them conducting the majority of the interviews (20). The preliminary analysis was done by one researcher having experience of five interviews, double-checked by another researcher (main interviewer) and discussed together to find consensus. The cross-reading of the stories was important for the analysis as many descriptions are so mundane that they are easy to ignore. It helped to have three researchers to pay sensitive attention to those mundane descriptions and to ensure that the main categories covered the descriptions and that their names were accurate.
Regarding the limitations of this study, the interview data is in many aspects narrow. First, the social workers come from three municipalities (out of 310 municipalities). It is known that child welfare practices differ between municipalities, and local cultures matter a lot in Finland (Heino and Pursi 2018); this small-scale data are obviously not nationally representative. Neither is it organizationally representative as some social workers in the agencies did not take part in the study. The reasons for not attending may vary: the agencies were very busy at the time of the data gathering and the social workers had to decide whether to spend their time with us or the families; there were many social workers working on temporary contracts without much experience of care order preparations and some social workers most probably did not see any point in joining the study. One limitation of the study may also be that it might not translate well beyond the borders of the country as care order preparations vary across countries and different child welfare systems (Berrick et al. 2015). Secondly, the stories were chosen by the social workers themselves and they rely on their memories of the case. Finnish social workers have university degrees in social work and are, through their studies, familiar with the power of story-telling and the choices the narrator can make -which they might have practised themselves when interviewed. The stories were told to three different interviewers and the interactions with them may have shaped the story-telling, especially as the interviewers were familiar to some interviewees as their past teachers of social work. When telling about the work done, it is probable that professionally and socially acceptable stories were told as people in institutional contexts often tend to conform to the stories that are expected of them (Polletta et al. 2011). Contested or poor practice presentations were likely to be excluded or marked as extraordinary events. In other contexts, other kinds of stories would be told. However, in our view, the stories provide us with an opportunity to learn about socially and professionally acceptable care order preparations, which serves our research interest well.

Talking together
When social workers describe their work in care order preparations, they speak a lot about 'talking'. They use a variety of verbs to describe activities which consist of verbal interaction to exchange views, information and opinions. It is a mundane description of work (e.g. calling someone), often so mundane that it is mentioned as a self-evident episode: I called the mother about our team's view. (2) Talking takes many forms: negotiations, phone calls, chats in the corridors or in a car, talking in a team meeting or in a supervision setting, or talking while lying on the floor and drawing a picture with a child, for example. Talking takes place in formal or informal settings. What is typical of different forms of talking is that it is described as a relational task: talking takes place with colleagues, team leaders, parents, children or external experts. The aim of talking is to learn about other people's views and to share information. It is rare in this data that a social worker would approach talking as her/his individual activity (e.g. 'I said'); rather it is approached as a joint activity (e.g. 'we said that') or it is described in the passive as is common in the Finnish language (e.g. 'it was said that . . . ').
The following extract demonstrates a common description of talking during care order preparations. The latter extract demonstrates how 'we' becomes the actor of talking.
These issues have been talked about, also among the colleagues and so on. (16) We have been in touch and they have accepted support, and there has been good interaction and talks about the needs of the child with both parents. They saw the seriousness of the situation in the same way as we did. And as I said earlier, the placement was on the agenda all the time, so of course when the referrals began, an emergency placement was considered, whether it was necessary, but the situations often settled quite quickly. (21) 'They' in the latter extract refers to the parents. Talking is described as being an activity among adults. If there is a shortage of time -as there often is in these descriptions -talking with children and teenagers is seen as being more time-consuming. Other professionals are mentioned as the ones talking with children. If the child is already in an emergency placement, social workers may ask the residential staff, close to the child and his/her everyday life, to speak with the child in order to inform him/her about the process and to share the child's views with the social worker. This is to say that talking with a child does not necessarily mean direct contact with the child. The stories include some obstacles to talking, especially with aggressive and absent service-users and those who withdraw from contact otherwise. We will return to these obstacles later. Yet, it is precisely 'talking' that social workers say they mainly do in care order preparations.

Writing
Social workers speak about writing -documentation relating to the case -as an element of their work in care order preparations. Especially when describing writing, legal and bureaucratic norms regulating 'doing' are spoken about: time frames for preparing the documents and the different types of documents required. Writing is more personalized than talking in these descriptions as the social workers use the expression 'I did the recording' instead of saying 'we did the recording' or 'the recording was done'.
If we have come to the conclusion that we have to take the child into care, when we have done all we can, I can write the care order decision with good conscience in the sense that I have tried out everything or that we here have tried to do everything to help the family and now we have to do this. (8) I struggled with the horrid application; it was this kind of pile of material which I had to look through. (13) The writing in both extracts relates to the actual decision or application, which in the first extract is approached as a legitimated result of supportive work and in the second one, as a difficult task to do as the case involves many previous documents to be included in the document at hand. The former views writing in relation to the family. In a similar way, some social workers describe that they write summaries of the events leading to a care order for the families to read. In these cases, the processes have been so complex that a summary -which some social workers approach as the child's biography -is seen to help parents and children fully understand the process and come to terms with the need for the removal, suggesting that writing is done for psychosocial reasons as well. These summaries are not required by legislation as the other documents mentioned earlier are. In fact, the social workers speak about their awareness of the time limits and formalities related to writing the documents as well as the personal and relational meanings which the documents carry, not least to the children and parents, and how they include those meanings in their work.

Coordinating
As part of 'doing', the social workers describe how they coordinate different issues and parties in the preparation process. When speaking about coordinating, they mention the norms set by the legislation in order to include every relevant party and view in the process. The task of coordination is not as prominent in the descriptions of doing as talking and writing but it is nevertheless evident. Social workers coordinate the times for meetings between the families and practitioners and invite different expert opinions. Again, coordination is more personalized than talking: it is the 'I' who asks for the statements from schools and health care. Coordination is clearly a task that cannot be transferred to other people but is the responsibility of the social worker in charge of the child's matters.
In the descriptions, social workers do coordination, first, by contacting parents, children, grandparents, relatives, foster carers and other people in order to hear their views and to get them to discuss the child's case together. Some of these discussions aim to establish an understanding of whether any form of kinship placement is possible or how the child's future is best supported, while other discussions take place for reasons of organizing formal hearings or other meetings. Finding times and places suitable for everyone is not always easy as people live apart, have their own schedules to follow and, especially, because some people do not turn up to the meetings or do not respond to contact requests. One social worker describes a father who never attended any meetings to discuss his child's case despite the social worker having tried to provide him with opportunities to be involved.
Second, coordination involves the inclusion of different kinds of expert opinions about the removal and placement. As in-home services have been provided for some time before the start of care order preparations, a single case may include several experts and service-providers whose views on the case are relevant to the preparatory process. Social workers ask for their written statements as described in the extracts below.
Always when we even start to consider a care order, we ask for expert statements. In this case, I asked for a statement from his schools, youth psychiatric care, family work, and he was placed out-of-home at the time, so the placement's view was included as well. (28) The treatment at the child psychiatric clinic started only this spring so they did not have such a clear understanding of the case but I still got a good enough statement from the doctor in child psychiatry. (14) Unfortunately, the interviewee did not specify and the interviewer did not ask what made the statement 'good' in the latter extract. We may speculate that it provided clear and detailed information for the purpose of preparing the care order. The extract does, however, imply that when coordinating expert views in care order preparations, the quality of statements matters and social workers need to contact different experts in order to include enough relevant expertise.

Relational actors in care order preparations
We have seen above that the descriptions of doing social work in the context of care order preparations -talking, writing and coordinating -emphasize work in relation to different parties. Thereby social workers position themselves as relational actors when describing care order preparations. This is especially prominent when 'talking' is described. Relations with parents and children in particular are given an important role although relations to colleagues and other professionals are present as well. We will examine the relational element further and focus on how the descriptions of relations with children and parents are included in the descriptions of doing. There are two main types of relations: partnership and troublesome contact. They both result in two different types of working: working in partnership and working with troublesome contact. Consequently, social workers position themselves either as partners, on the one hand, or contact seekers and interpreters of missing or vague messages, on the other.

Working in partnership with parents and children
'Talking' requires partners to talk with. Social workers speak a lot about cooperation and partnership between social workers, children and parents: cooperation is essential in order to do 'talking' in care order preparations. One social worker highlights how in the case of good cooperation, we could sit down and look at our diaries and decide together when the care order decision would be made and when and how the child would move into her foster home'. (17) The description emphasizes cooperation between the social workers, the parents and the child even to the extent that they can jointly decide when the care order decision will be made and when and how the child will enter foster care. The importance of partnership was elaborated further by another social worker who stated that I have promised to the parents that this care order will be terminated when you can look after your son. And I can keep my promise. I am, however, concerned that their case will be forgotten if I hand this case over to another social worker. So, I still keep them among my cases. (11) In this case, the social worker gives a promise of the termination of the care order and emphasizes that she will keep this promise. What happens here is that she promises to use her status as a public authority. However, according to the Child Welfare Act, the assessment of the continuation of a placement should follow certain legal norms. The promise does not follow those norms, and, in its emphasis on the relation between the social worker and the parents, it also excludes other actors such as the court and other social workers.
The descriptions about working in partnership do not necessarily exclude stories in which the parties express their objection to the care order proposal. The social workers speak about parents and children who object to the care order proposal but who still relate to social workers in a cooperative manner and thus the social worker is like a 'partner'. An example of this kind of a situation in the interviews is that of a teenager who disagreed with the choice of a certain institution as his placement because it was located far away. The social worker knew about the disagreement when preparing the proposal as they had spoken about it but could not suggest any other placement due to the shortage of places available in residential care. As the boy objected to the placement, the care order was formally taken to the court to be decided and was 'objection-based'. The role of the social worker is, however, described as working in close cooperative contact with the boy and the parents.
Partnership and cooperation are also talked about in situations in which social workers encourage the parents or the child to express their objection. In one of the descriptions, this happened when the social worker encouraged the parents to express an objection as she was concerned about the parents' full understanding of the care order. The social worker argued that the court process would guarantee a fair decision-making process for the parents. In other cases, the social workers suggested that the parties should object to the proposal if they themselves were not fully convinced about the need for a care order. In these situations, social workers described that they together with the family, as partners, asked the court to 'give an outsider view' on the case.

Working with troublesome contact
Difficulties in contact are presented as affecting the preparations as social work during the care order preparations is described to rely heavily on 'talking' and talking without contact is problematic. The difficulties in contact are talked about in two main ways: parents or children withdraw from contact or they communicate in an aggressive manner. In both cases of troublesome contact, there are difficulties in discussing the care order preparations with the children and parents. The social workers describe how absent service users, commonly parents, do not answer their phone calls or they do not turn up at the meetings set for 'talking'. What is left for the social workers to do is to try to chase the parents and to try to find the right ways to contact them, as described below: They never turned up at the meetings about the emergency removal when the plan was to discuss the children's issue and support for them and what was expected from the parents. And I never contacted them by phone after the meetings. I always sent a text message by phone and sent a paper letter. I am not convinced that they open the letters. The text message was more reliable as they sometimes responded to the message. (25) If the nature of the contact is described as aggressive, it refers to aggressive (non-verbal) behaviour. As words are not used, the social worker is left to interpret and translate the behaviour into words. This is highlighted by the following extract in which the hole in the wall was interpreted as the father's expression of objection. The social workers describe also other types of troublesome contact where they need to interpret children's or parents' behaviour. One description is about a 16-year-old girl who withdrew from talking and slipped into silence and therefore the social worker had to interpret the girl's view based on vague messages. She also described the girl's tired mother wishing to have the girl taken into care to get help for her. The social worker's interpretation was that the girl gave her consent by remaining silent. Another description of contact with a teenager without words is the following: When talking with him, it is likely that he is listening to us but he doesn't make any eye contact for example, so he sits in his room with his back towards us and then he has the hood on his head and answers by shrugging his shoulders. Nevertheless, I think that he listened to what was talked about. (14) In this extract, the social worker describes her attempts to create contact for talking. Attempts towards talking is thus evident in all types of troublesome contact. Social workers indeed describe themselves as seeking contact and interpreting non-verbal messages.

Discussion
In the descriptions of care order preparations, social workers' doing is spoken about as talking, coordinating and writing in partnership or in troublesome contact with parents and children. Talking per se is an important element of care order preparations and it is done to get the 'business done' (Spencer 2001). However, as care order preparations form a part of a decision-making process in which children's and parents' rights as a family are at stake, one could assume other types of vocabulary of doing, in particular that of decision-making (e.g. analysing options, doing risk analysis, referring to legal or formal terms, inviting legal representatives), to dominate the descriptions of doing or at least to be included in one way or another. Such vocabulary is almost entirely absent. It is fair to say that social workers do not employ the vocabulary of juridification or standardization (e.g. Ponnert and Johansson 2018). It is mainly the descriptions of writing and documentation that serve as a reminder that the work is done in a statutory context which sets external norms for social workers to follow. Consequently, the descriptions of 'doing' the care order preparations demonstrate that social workers position themselves more as relational actors of social work than as agents of the statutory, bureaucratic system.
When talking about talking, social workers inevitably orientate towards parents and other adults (e.g. colleagues). This finding underscores the adult-centredness of doing care order preparations. Furthermore, the nature of the contact between social workers and service users becomes an important dividing factor as talking takes place best in relations which are described as partnership relations, while it escapes those preparation processes in which the children and/or parents withdraw from contact or in which they are involved in an aggressive manner. In the latter case, social workers are left to seek contact or to interpret the vague messages to get their work done. Obviously, the troublesome or missing contacts pose a challenge, not only for social workers and their selfconstruction as relational actors, but for the quality of care order preparations. Although 'hard to engage' service users pose challenges to practice in every child welfare system (Thoburn 2010), in Finland they may pose fundamental challenges to the legitimacy of the dual decision-making system and its care order preparations. If contact is not feasible and social workers are seeking contact or interpreting vague messages, other types of vocabulary of 'doing' -relating to the representation of rights for example -could be expected to emerge, but they were absent from this data.
There are several possible reasons as to why social workers as relational actors are so evident in our analysis. The stories of practice are not only individual but collective (Misztal 2003, 11-19). During the history of child welfare in Finland, the legal, administrative and professional case work approaches have existed side by side, yet with varying degrees of prominence (Satka 1995;Harrikari 2019). At the time of data collection, the national child welfare reform programme by the Ministry of Social Affairs and Health implemented relationship-based training for practitioners in child welfare. The programme is based on the understanding that the Finnish child welfare practice is guided too much by bureaucratic norms and therefore it is in need of reshaping. The programme also cherishes the view that if relationship-based work is given a stronger role in child welfare, it would improve the quality of practice and there would be more work done in partnership with the families (Petrelius and Eriksson 2018). This is to say that the present ethos of 'good social work' especially emphasizes relations and talking. This may have influenced the social workers' descriptions of care order preparations and their institutional selves to some extent as the implementation of that reform programme was active in the municipalities involved in the study and elsewhere.
Moreover, the self-constructions of social workers as relational actors echo the long-lasting formal nature of Finnish care order decision-making. The very existence of consent-based care orders frames care order preparations so that agreement with social workers matters as the care order decision can be made only in the social work domain outside the court system. This is quite different from adversarial child welfare decision-making, which has been highlighted in several Anglo-American studies (e.g. Burns, Pösö, and Skivenes 2017) and which is characterized by such terms as non-voluntary and mandated. These terms are in line with the vocabulary of difficult, negative, hostile, resistant and uncooperative service users and related reactions from social workers (Barter 2008). In our data, in contrast, consensus among different parties is given prominence, and the nature of a care order as a form of service is emphasized. Indeed, in this data, the social workers' institutional selves as relational actors rest on the assumption of consensus being the standard option for care order decisions.
However, partnership relations with parents are not uncontested, not to mention those with children (Calder 2008;Broadhurst 2010;Venables and Healey 2019;Hultman, Höjer, and Larsson 2019). Although the views of children and parents are not studied empirically in this paper, one may assume that for them the care order preparations can offer a platform of partnership, but also one of resignation, fight or escape (e.g. Grell, Ahmadi, and Blom 2017). Procedures to remove a child from the parents to public care are, after all, socially, individually and legally complex, not least because the children and parents do not have full autonomy to decide how to be included in the removal process. The very notion of informed consent to a removal and the dual decision-making of care orders may be flawed in Finnish child welfare (Enroos, Korpinen, and Pösö in preparation) as the option for an involuntary removal exists alongside the consent-based one. Consensus paradigms with their emphasis on voluntarism have been criticized elsewhere as the power to intervene is seen to be camouflaged by the emphasis on consensus (Leviner 2014;Lynch and Boddy 2017). In a similar vein, the Finnish social workers' descriptions of relational work in partnership with children and parents could be seen as a way to camouflage the use of public power and to present partnership as more genuine than it is (Pösö et al. 2018).
On the other hand, looking at the findings from a different point of view, consensus as a platform could provide a 'friendly', non-adversarial process for parents and children in situations in which mutual agreement about the need for a removal and the removal itself exists (e.g. Venables and Healey 2019). From this perspective, the social workers' emphasis on talking and relations could be seen as a manifestation of a non-adversarial child welfare system and social workers' selfconstructions as relational actors as relevant implications of that system.

Concluding remarks
Social workers responsible for children's matters in child welfare 'do' care order preparations by talking, writing and coordinating and construct themselves mainly as relational actors and emphasize the relational aspect of their work. This reflects the non-adversarial Finnish child welfare system and the broad discretional space given to social workers to prepare care orders. In addition, the findings materialize the present dominant discourse on what 'good' social work in child welfare is like, emphasizing relations and separating the elements of bureaucracy, law, juridification or standardization from the descriptions of care order preparations. This is not, however, to claim that social workers do not employ the legal or bureaucratic norms in their practice. Rather, it is to say that the relational view on care order preparations is given priority when speaking about doing social work.
More research would be needed to examine how this prioritization serves children's and parents' rights and whether it treats all parties equally. This study can only argue that the vocabulary of relations can downplay the legal and bureaucratic norms when social workers describe their work. Given the nature of care orders and the restrictions on family life likely to arise as a result of the preparatory process, it can also challenge us to rethink, to what extent the relational view of care order preparations and social workers' institutional selves is, on its own, the ideal one.

Disclosure statement
No potential conflict of interest was reported by the authors.