Professional Values Challenged by Case Management—Theorizing Practice in Child Protection with Reflexive Practitioners

In this article, we theorize and reflect based on former research into professional practice and discretion as well as use some results from working together with practitioners in child protection services to explore the phenomenon of non-performing. Regulation lies at the heart of the contemporary child protection discourse. On the one hand we have seen a trend towards systematization of assessment content and procedures, on the other hand it is assumed that rational management approaches can secure consistency of performance. Social workers may be weary of the constraints all this imposes, but seem generally content to comply. Our reasoning was that social workers in child protection should be helped to get to grips with modifications to practice so that multi-challenged families could be accorded priority. These changes would include a reframing of assessment to take account of family needs as well as the needs of children. Follow-up would also require much more attention. Additionally, the choice of help provided for children and families would have to come into better focus, despite the limitations often experienced in practice. The question we asked was whether these types of reframing could be fostered within local child welfare units. We conducted a field trial in which child protection units were encouraged to reframe their practices, with the support of an expert group. The idea was to enhance and enable innovation through the combination of a more thorough dialogue with the families involved, as well as critical reflection based on available knowledge related to the identified challenges. We do a critical discussion of the work and the results from this in order to enhance knowledge on innovation in child protection.


Introduction
While carrying out research as well as examining researching practices and the world of social work for more than three decades, the authors have experienced the influx of neoliberal reasoning, especially involving an increased focus on metrics and performance in welfare services. This development has adversely affected social work service development in many ways, does not explain why social work often disapproved of by the clients who rely on it. Very often, we have seen qualified social workers who have clear ideas of the needs of clients and value the cooperation and responsibility that lies at the heart of social work as being relevant to people's lives and wellbeing, but in a reflexive mode must acknowledge their failure in practicing their professional values. In this article, we theorize and reflect based on former research into professional practice and discretion. We also use some results from working together with practitioners to explore the phenomenon of nonperforming.
Regulation lies at the heart of the contemporary child protection discourse. On the one hand we have seen a trend toward systematization of assessment content and procedures, on the other hand it is assumed that rational management approaches can secure consistency of performance. Social workers may be weary of the constraints all this imposes, but seem generally content to comply (Fauske et al. 2009;Marthinsen and Lichtwarck 2013;Clifford et al. 2015). Research in the New Child Welfare (NCW) found that efforts to develop a knowledge-based practice in Norwegian child protection with considerable attention paid to management have not resulted in a service with clear or even defensible priorities. The most marginalized and deprived children and families were those who received the least help designed to improve the care of their children. Parents were critical of the fragmented nature of the help they were provided with, and what they perceived to be inadequate follow-up and assistance provided.
On this basis we proposed an innovation project, which was accepted and financed by the Norwegian Research Council. During the years 2013 through 2016, we carried out the project "Innovative Approaches to Work with Neglected Families: Targeted intervention and support for high-risk families in Child Welfare". This project was part of the Program for Practice-based RandD for Health and Welfare Services (PraksisVEL).
Our reasoning was that social workers in child protection should be helped to get to grips with modifications to practice so that multi-challenged families could be accorded priority. These changes would include a reframing of assessments to take account of family needs as well as the needs of children. Follow-up would also require much more attention. Also, the choice of help provided for children and families would have to come into better focus, despite the limitations often experienced in practice. The question we asked was whether these kinds of reframing could be fostered within local child welfare units. The offices and their managers included in the trial took the normative position of acknowledging that work could and should be improved in order to deliver sustainable services to multi-challenged families and their children. Innovative actions would be supported from the outset. We conducted a field trial in which child protection units were encouraged to reframe their practices, with the support of a group of experts in child protection assessment and treatment that should act as a discussion partner in order to enhance critical reflection. The group was not to act as supervisors but should discuss challenges on a general level-not work with the families and children as such. The idea was to enhance and enable innovation through a combination of a more thorough dialogue with the families involved, as well as critical reflection based on available knowledge related to the identified challenges. Innovation here would mean developing actions to broaden the scope of investments in families with multiple challenges, and following up and evaluating how these innovations could be adapted to use when facing similar challenges with other families in need.
Another line of reasoning was that social workers in child protection by virtue of their training are unfamiliar with the mindset and skills that can support low-level innovation of the kind envisaged here. The university set up a trial master-level course designed to develop innovation skills. This was offered to social workers drawn from the municipalities participating in the field trial. These offices had taken part in the research program "the new child protection" where the lack of investment in multi-challenged families were documented from researching a representative part of their clients (>700).
New families with the same characteristics as the challenged families were identified to be followed up in the coming two years, and those in charge in four offices were offered the master course. The course proved successful and led to the initiation of a European project "Learning to Innovate with Families" (LIFE) funded by the Erasmus + program.
Innovation in practical settings, however, proved to be very difficult. The main reasons for this were not hard to detect. In the four municipalities that took part in the trial, major reorganizations of the child protection service began early in the trial period.
These changes were relatively drastic; one municipality dispensed with a considerable portion of its middle management to cut expenditure, another had a breakdown due to infringements of regulations and embarked on a lengthy reorganization and rehabilitation process under new leadership. Two municipalities had difficulty in meeting deadlines for assessments after referral. One of the municipalities abandoned a purchaser-provider model for organizing its service, while another adopted the same model. There were other signs of retrenchment due to cost pressures, with child welfare units removing vulnerable families from their caseloads. All the child protection units that took part in the trial experienced severe pressure and had great difficulty in complying with the trial design, which required that 12-20 families with multiple and complex problems be selected for a reframing of assessment and follow-up.
The field trial and the master´s degree course did however provide much interesting material that enabled us to trace some of the phenomena which played out in the organizations that could explain why families were not followed up with and the lack of innovation. Our article is based on our experience of organizational conditions that may preclude or hinder innovation in child protection and some examples of how innovation is viewed by the social workers themselves. We explore dilemmas that arise in giving priority to clients' needs in settings where managerial demands prevail and regulate social work. To support this discussion, we draw from the two-weeks-homebased exam papers and two days of structured discussions reflecting with the social workers/students upon their first year of experiences trying to innovate when working with challenged families.
We interviewed some of the chosen families for these services; these interviews are not presented here, but they do confirm stories or descriptions given by some of the social workers. We do not discuss other ways of working with families, such as projects based on emancipatory, feminist or critical social work. This research is to some extent focusing on how disciplinary traits in social work settings is acted out, without regarding or working with resistance.

Research and Theory Background
In child protection, discourses of management and of professional concern have been divergent for some time. Policy makers have increasingly favored approaches that are research-based and there is a pressure for innovation and accountability in practice (Parton et al. 1997;Gray et al. 2009).
Research on child protection practices based on information from clients as well as social workers has shown increasing conflict around assessing needs, responding to them and problems of handling caseloads under new public management regimes. Much social work literature points to how new public management and the neoliberal ethos challenges social work values and client focused practices (Lorenz 2005;Ferguson 2007;Garrett 2009;Featherstone et al. 2016). Social workers report a lack of time and resources while clients often express frustration about the lack of adequate response, fragmented services and lack of respect for their expressed needs. Juhasz and Skivenes (2016, p. 1) have studied care order preparation in several countries-including Norway-and they find that "roughly two out of three workers say they would experience obstacles, and the main obstacle by far is related to time and/or large caseloads. Lack of organizational structures or poor management is the second major obstacle, followed by collaborative problems with external partners and challenges related to providing evidence. Only a few workers mention individual factors".
Government papers as well as reviews and inspection reports point to high workforce turnover, lack of competence and the need for reorganization and stricter management. The discourse on child protection practice and its expansion reveals disagreement about possible solutions. These solutions range from demands for better management and stricter regulation to specific competence-building strategies and yet further toward enhanced reflexivity and more expertise in coping respectfully with uncertainty. Enhancing or promoting innovative practices has over the years surfaced as one way of coping with the increasing complexity.
If one particular discourse becomes hegemonic in a field, there are likely to be reasons for this. The dilemmas that arise from the coexistence of, and possible conflict between different value hierarchies will have to be negotiated (Garrett 2009;Rugowski 2010). Decisions have to be justified as social workers get on with their everyday tasks. Concern about accountability, threatening to overbalance into a moralizing culture of blame, permeates public service organizations nowadays. Parton et al.'s (1997) breakthrough contribution Child protection, Risk and the Moral Order was an early empirically based contribution questioning the development of a child protection that embodied social and cultural biases. A Bourdeuasian analysis of child protection in Norway (Marthinsen 2003) using the notion of accumulation of symbolic burdens among clients deployed a wholly different approach but drew a similar picture. Munro's report on child protection in the UK averred that instead of "doing things right" (i.e., following procedures) the system needed to be focused on doing the right thing (i.e., checking whether children and young people are being helped) (Department of Education 2011).
However, developments in Norway since 1990, in the child protection field, cannot be seen simply as a process in which tension between managers and professionals has escalated. The service can be seen as an example of a mixed knowledge regime (Sørhaug 2004). When knowledge is the organization's capital, knowledge, power and authority are also bound together, and as Sørhaug points out, different knowledge domains (or regimes) are involved. These include the organizational hierarchy (line management) the collegium (its interaction based on continual discussion) and the network in which personal alliances, reciprocity, antagonisms and trust are articulated. These knowledge regimes infiltrate one another. An analysis based on dichotomization of management and professional interests may be blind to important coalitions of interest that bring management and social workers together at particular junctures or even over longer periods.
Some earlier theorizing and research about this problem area will be explored here, as well as social workers' need to cope with questions of accountability and discretion. We must emphasize that the shortcomings of child protection services are interesting and relevant in terms of theory and formulation of research questions and approaches. But they are scarcely abstract. Our own research based on client's narratives concerning experiences with child protection services revealed frequencies of service-family contact that were very low, fragmented service provision from a range of services involved, high levels of turnover among caseworkers, a lack of continuity on the part of responsible social workers, frequent cancelling of appointments and promises not being kept (Clifford et al. 2015).
Our approach has applied complexity theory and critical social theory based on Bourdieu's 'toolbox'. We use the concept of bifurcation point (explained below) to illustrate how managerial values may challenge social work priorities regarding families in need. We found that this helped us to get a better grasp of the practices involved and could help the participating social workers to reflect upon why they were not able to give priority to the chosen multi-challenged families involved and thus why little innovation evolved.
Child protection and welfare services have been expanding in Norway as in many western countries, with an increasing emphasis on child protection as opposed to more general child welfare work (Lonne et al. 2008;Gilbert et al. 2011;Parton 2014;Christiansen and Kojan 2016). In Norway, the growth in volume of service as measured by numbers of children referred and children and families provided with services, has been considerably greater than the growth in numbers of staff employed in the services Clifford et al. 2015). As a result, there are more tasks for each worker than there were before. As services expanded (from 12,000 children in 1982 to more than 50,000 receiving services in 2010) workers and management found it difficult to maintain an effective overview of which children and families became clients and why did they come into contact with the service. Statistics have been vague and not able to count for the increase enabling the services to develop good strategies to cope (Marthinsen and Lichtwarck 2013). The services have become dependent upon databases to access information about their work and their users. The tools adopted have focused on client flow related to certain tasks from intake to output (controlling time limits for reports, assessments and actions), rather than the content of the operations carried out during working hours. Research carried out after the introduction of the new Children Act in 1991, which set up a system of time limits and deadlines for assessments, showed an emerging pattern in which the office became the prime site for social work, the child care workers made home visits and visited other contacts much less frequently (Clifford et al. 1996). The research revealed a lack of focus on some of the more time-consuming activities involved in social work, especially those tasks related to follow-up and practical work with families and children. The monitored tasks were mistaken as being the most timeconsuming ones, while the routine follow ups that were not monitored were the real time consumer (Clifford et al. 1996). The data recorded in computerized monitoring systems as they have evolved since the mid 1990's mainly does not concern direct client-related work, but rather task completions related to deadlines, legal grounds for intervention and reasons for intake and allocation of services. Databases regulate much of the working day and draw attention to the measured activities. When set goals or time limits are not fulfilled, these are often flagged and reported to management as flaws that have to be corrected, since much of this is later reported to the inspectorate and may lead to fines and criticism. Social work with families, follow-ups and responses to developing needs are seldom measured. Social workers have reported deficits of working time available for core client-related activities. Gautun (2010) found that 87% of municipal child protection workers claimed that there were serious bureaucratic constraints on their working practices. The same research indicated that four out of ten complained of a lack of necessary resources to practice early intervention, 34% said there was no opportunity to improve their own competence and more than half reported resorting to non-optimal handling of cases within the previous year.
Tension between managerial and professional interests has been a long-standing issue in social work, certainly predating the advent of New Public Management. Howe (1991) turned the awareness on the ongoing shift of power from the practitioner to the manager. Evans and Harris (2004) discussed the impact of managerialism and how it influenced and possibly reduced social workers' discretion. Norwegian social work has very similar development to that of the Anglophone world. Christiansen (1977) in a seminal article on "administration or treatment" showed how encroaching bureaucracy and older practices of management dating back to the days of the poor legal administration still haunted social services that were being increasingly professionalized with the employment of trained social workers. These professionalizing changings included working with people to enhance personal development, to facilitate social change and to engage in community work, but ended up with the administration of applications for social security benefits. She outlined the conflicts between professionals and the bureaucracy by pointing to the values of social work; promoting active involvement by the client, their right to influence decisions and the need for trust in relationships: This in contrast to the conditional management of economic support that often disempowered clients.
Billingsley's early work on the challenges facing most social workers operating within bureaucratic organizations is also of interest in this connection (Billingsley 1964). He discussed how social workers approach their work in terms of role orientation or attitudes towards the job. He also uses the notion of orientation patterns. This may be compared to Bourdieu's (1999) later concept of habitus within social fields that allows for value and identity distinctions to have different impacts within hierarchies of distinct groups. Within bureaucracies, professional workers seem to become "socialized according to a set of values that differ from those inherent in their professional calling" (Billingsley 1964). Social workers operate within "subsystems of society made up by social units with distinctive boundaries and patterns of interaction that are oriented toward rules" (op.cit 401). For social workers Billingsley used the notion of subsystems: the profession, the agency, the clients and the community. All these imply different obligations and make demands on their role orientation. The different systems may refer to or enhance different values, which we here have chosen to categorize as distinctions and attractors. Billingsley's work concluded that the clients and the community had less influence than the profession and agency on social workers' orientation at work. This state of affairs was quite general in spite of the social workers' intellectual and emotional commitment to meeting the needs of their clients' (ibid). He also found that social workers were more oriented to carrying out agency policies and procedures than toward carrying out their professional commitments, when these were in conflict. Billingsley developed Riesman's (1949) idea of role conceptions in his analysis, and created four orientation categories that seemed to be in operation. The first category, termed 'professionals' included those who tended to give primary allegiance to professional standards and who regarded their job as a setting to do their professional practice, and feeling free to leave or change jobs if their values and professional aims were challenged. The second category, 'bureaucrats' were primarily oriented towards carrying out agency policies and procedures. This group was more loyal towards their agency. These two are the roles dichotomized by Christiansen (1977) although her article is not based on research but is rather part of a debate on social work and its challenges in the evolving welfare state. Billingsley's third category is the 'conformists' with a relatively high commitment to conform to the expectations of both agency policies and professional standards. He compares this group to Riesman's 'service bureaucrats'. The last group is the 'innovators' who show low commitment to both agency policies and professional standards, and have a willingness to challenge or violate agency as well as professional standards in order to meet the needs of clients. This group is also more responsive to the community. Although this research dates back to the 1960s, it identifies some important patterns that are probably still in play in organizations today. As we came across this after our project was finished, we were not able to reproduce Billingsley's work, but it has certainly made us aware of these tensions appearing in our empirical data.

The Question of Discretion
The Oxford Handbook of Public Accountability (Bovens et al. 2014) operates with several prerequisites for accountability when viewed as a social as well as political mechanism; accountability is fundamental in influencing the extent and character of discretion. In an accountgiving discussion, the actor must inform about the conduct performed, its tasks, outcomes and procedures. Explanations and justifications become important-especially in case of failure. Secondly, it must be possible to question the actor about the legitimacy and the adequacy of the explanation given. Accountability relates to answerability, that is, specific questions and the feasibility of providing an answer to them. Lastly, we must be able to pass judgement on the conduct of the actor to facilitate reward as well as criticism (Bovens et al. 2014).
Evans (2016) has an interesting analysis of the influence of new management structures and on the continuing relevance of Lipsky's street level bureaucrat theory. Evans´s point is that new public management has given senior management greater influence in the sense of governing the room for discretion among those who work with clients (street level bureaucrats). Their working conditions are increasingly restrained by resource shortages and policy confusion as well as having to live with a political concern with the avoidance of blame and culpability (Evans 2016, p. 603). Evans shows that two strategies may be in use in order to manage risk. One is proceduralist; guidelines and detailed specification of procedures are used to achieve control. This is reminiscent of Munro's (Department of Education 2011) dictum about doing things right, but not caring about the right things. The other strategy Evans points to is 'service abandonment' which in child protection means that one closes a case and cannot be blamed, or one refers it to someone else. This may be done by using tools to assess needs and allocate services, and interpreting and applying them in ways that reduce demand for example through 'nudging' in certain directions by politicians and management (Thaler and Sunstein 2008). Høybye-Mortensen's (2015) study of decision-making tools in social work seems to indicate that in spite of the increasing use of such tools, discretion seems to still rest on the workers ability to maneuver inside or outside the services available, and that the real challenge is when clients present needs which are not included in the available 'service package', for instance to cope with loneliness or have more frequent contact with their family than time allows. Lorenz (2005( , quoted in Spolander et al. 2016 suggests that "social workers often withdraw to privatisation and therapeutic approaches or accept dictates of New Public Management without opposition". This may explain why social workers do not usually show open opposition towards management or policies that may inhibit efforts to improve the lives of clients or oppressed or marginalized groups. In the light of the older research referred to earlier about social workers' mode of operation or orientation patterns, Lorenz's expectations directed at social workers as political actors in everyday life may of course be unrealistic. Professional social work has to relate to the need for legitimacy of public or private services and has to comply with rules that regulate discretion and accountability. Its role conceptions have to evolve within this context. Molander (2016) discusses the anatomy of discretion in relation to the welfare state. He refers to Rawls (1993), who provides a list of six points describing non-eliminable hazards that influence both theoretical and practical reasoning involved in coping with the burdens of judgement (Molander 2016, p. 39 However, there is mostly a lack of clear and uncontroversial criteria for such rankings. To explain how people may cope with this Molander discusses heuristics, which is a form of reasoning that may remind us of populism (Müller 2016). These simple procedures for answering difficult questions allow us to replace complexity with simple questions and answers about what is, and why and how. Compared to more thorough reflexive reasoning in which we take precautions based on insecurity and lack of knowledge, heuristics may allow us to make (even if maybe false) conclusions about futures, for example. This fits well with the kind of discretion necessary in child and family work with families with children at risk. Heuristics may be a type of warrant or inference rules that bridge the gap between data and conclusions (Molander 2016, p. 41). For instance, this way of reasoning may explain why a social worker chose to cancel an appointment with a mother for follow up work with the family and give priority to bureaucratic work on the computer to meet deadlines for reporting to avoid sanction from superiors and inspectors. Usually nothing critical happens, but it might be that something really bad could have been avoided had she/he kept that appointment. Then she/he would have to be accountable for the exercise of discretion and the manager would have to account for this incident. This is what would be called a bifurcation point where different attractors/distinctions (symbolic capital) would be at stake, as demonstrated in the analysis we provide below. Clifford et al. (2015) see child protection as an organization on alert that always has to be prepared to intervene in its everyday practice to avoid and prevent public and political criticism directed at presumed mismanagement. This results in a strong, even exaggerated focus on assessment, categorization and goal attainment, quality assurance systems and fast track thinking regarding the seriousness/gravity of maltreatment or abuse, all conducted at the same time in an at times bewildering complex of case trajectories. We have seen this in these services elsewhere, not least in the UK over the same period of time (Garrett 2009, Department of Education 2011). As such, this notion of organizations on alert seems to correspond to Evans´s point about service abandonment.

Theorizing Practice in Everyday Life
Complexity theory operates with a nonlinear logic. It allows us to observe how certain decisions transpire based on judgement in everyday work. When do choices turn up and what choice is made in certain situations? A situation may escalate to what can be characterized as chaos. We experience dissipative structures and may reach a bifurcation point where several options may apply, or else someone will have to innovate. The choices made may be regarded as more or less attractive depending on the future they may contribute to within the organization and in relation to service for families (Stevens and Hassett 2007;Stevens and Cox 2008). The main questions involving making these choices are: Can we manage the situation, can we afford these changes, what kind of knowledge may support our decision-are there time and resources available, etc.?
We have chosen to combine complexity theory with Bourdieu's (1999) concepts of habitus, social fields and symbolic power or capital. The dissipative structures operate within social fields with kinds of habitus that may not be stable over time. What are at stake are attractors or distinctions that may represent different kinds of symbolic capital to agents within the social fields in operation. If management promotes change and innovative work, it also has to support and enhance the allocation of resources such as time, competence and funding to these activities. If not, then management is not in effect promoting the distinctions or the symbolic capital at stake that are about meeting needs and coping with fears about the future among those involved. At the bifurcation point, the dissipative structures (habitus with a certain hierarchy of distinctions defended) lead to a choice among attractors that will lead in certain directions-confirming the value positioning of the agents involved.
How is a situation perceived, what is the importance of certain actions available and what kind of futures are at stake? A constructed example based on non-linear logic could be: There is a telephone call from a mother about her concern for her child due to her new fiancé's interest in her teenage daughter. This call may result in an agreement to see whether her new boyfriend is on a register or has been convicted of any previous maltreatment or assaults. The check may result in two outcomes, either he has been convicted, or not. If the answer were "no", then the social worker might choose to leave the case for now, and give priority to other cases. A risk is of course that sexually abusing this girl might be his first offence of this kind. The criminal record versus no criminal record may be regarded as a bifurcation point, which imposes a certain action where risk has been operationalized to be equivalent to what is recorded. The social worker has chosen to believe that this man does have good intentions, wants to be a decent stepfather and so does not move on towards suspecting him of ill intention. The mother is thus regarded as too concerned and no action taken. The good/bad thinking is an attractor that weights the decision in certain directions. Of course, this may also be regarded as a risk assessment, but the decision is not based on linear cause-effect thinking. It is more about understanding processes that are in train in which prediction is not based on elaborate thinking-what is described here might well be thought of as a routine procedure.

Materials and Methods
Starting in 2007, we developed a research program aiming at expanding our knowledge about clients in Norwegian and later Swedish Child Protection services (Fauske et al. 2009;Marthinsen and Lichtwarck 2013;Clifford et al. 2015). The initial survey of a representative selection of children (N = 715) and families was followed by a longitudinal study based on 96 of these families. These families were selected as multi-challenged households based on interviews assessing the children and parents scores on certain psychosocial stressors. We used the concept of families with complex needs based on high scores on at least three of these variables reflecting different life areas. The conclusion was that the needs of these families were poorly met by services. We developed a research project based on this knowledge where four of the services that took part in earlier research were offered two kinds of support in order to facilitate innovation working with this group. One was a team of experts who could be used as discussion partners in working with these families, though not as supervisors. The services were supposed to develop their own actions in cooperation with the clients. The other resource was a master level year course focusing on innovation in child protection working with families with complex needs. The innovation team was offered only to Norwegian participants from four service units, but the master-course (20ect) included social workers from eight service units in Sweden. The Swedish social workers came from services who had been involved in a replication of the Norwegian survey and a comparative follow up (Davidson and Bredmar 2012). In each office, the system of selection developed earlier was applied to identify new families with complex needs. Each service unit was asked to appoint participants to the innovation course based on the premise that these participants should be in charge of the selected families, and were given the task of innovating in dialogue with the families and the innovation-team. The 16 students (1 male) attending the course were social workers, except for one occupational therapist. All of these students had more than five years of experience in child protection work, with ages between 30 and 50 years old. The course included theories and experiences of innovation, the setting of knowledge production, some relevant theory of science and the possible impact of neoliberal policies and management. Much of the teaching was focused on recent knowledge developments in working with challenged families and their children and the team of experts did some of this teaching.
Specific theorizing of the conflicting or competing values of management versus social work using research like Billingsley's were not presented during the course. This was immediate and evident in their everyday experiences and discussed as challenges in practice. Critical reflection was the mode for teaching, much like in the UK's Making Research Count projects and the Critical reflective practice forum (personal visit by authors 2020). The learning was shared among the researchers and students, more than being socialized into a certain way of thinking. The teaching during the course was very much focused on active student learning through presentations in groups and individually based on material chosen by students themselves, as well as active discussions in class and during meetings in setting at work and in gatherings and meetings related to the research project in practice.
The idea of using complexity theory came among the researchers during the course, and the referred article by Stevens and Cox (2008) was enclosed in the home exam as an analytic tool. Part of the home exam was to discuss what may support and what may restrict innovation in social work practices.
At the end of the course, two days included discussions with four of the researchers/teachers concerning the possible explanations and reasons for challenges met during the innovation project. This setting was similar to a focus group session with researchers taking notes during discussion. The number of participants attending classes varied due to turnover but 14 completed the examination. The exams used in this article was an essay written during a full week, based on applying theories and research presented during the course to their own experiences and ideas about how to enhance innovation, as well as what may restrict it. All students gave consent to the use of the examination material for research purposes.
The main project these social workers were a part of was a trial, and not an experiment or a variant of action research, since there was no feedback other than the communication facilitated during the master course. Our design was more user-focused, and included an initial interview with the families selected that contained a life history interview and survey data on living conditions, as well as health and psychosocial functioning and burdens for parents and children. These families were supposed to be interviewed again after two years, but due to the fact that many never received any follow up or cases were closed, we only managed to interview a couple for the second time. There were little traces of social work to look for, and we also lost contact with many of the participants. The motive for this article is to discuss some of the phenomena playing itself out that may explain why innovation in the sense intended did not succeed. We do not intend to present to what degree innovation practices have evolved in such a way that clients experience change but concentrates upon managerial issues evolving in relation to the realization of the project. Whereas the master-level course provided learning opportunities that were highly valued by the participants, our efforts to promote innovation opportunities in child protection units were much less successful. Access to the innovation team had little impact upon fieldwork and focus interviews with the innovation team confirmed this-their expertise might as well be used to comment on other challenges rather than the involved families. The discussion that follows is mainly based upon the discussions from the two days reflecting with the course participants and their exams where they elaborate upon their experiences with the project to enable learning. We do not think the setting of an exam is greatly influenced by the researchers/teachers based on looking at the experiences. The influence may be the tools they were able to incorporate in their thinking and the way experiences and context are theorized and discussed.
We did not regard these students' work as being representative of social workers in child protection in general, but we did see them as experienced (not experts in Dreyfus' terms) in situations where research and knowledge in general could be applied in a test to develop and implement innovative practices. They were strategic informants chosen for their particular experience, situational perception and professional backgrounds. Their experiences can provide a description of what happens in these organizations when change is at stake or made possible. There may be biases in their accounts and perceptions, but these are unlikely to have been systematic or be artefacts.
Rather than general knowledge, working with these social workers allowed for sharing of typical experiences in practice, especially with the intention to innovate and use and produce knowledge in their own organizations. The reflexive analysis and theorizing enabled by this cooperation provided an opportunity for developing a more thorough understanding of the possible relationship between social work and management.

Practicing Discretion
The examples cited below were drawn from examination papers and researchers written notes from the discussions focusing on recording the arguments presented, the challenges met and how they were interpreted and eventually shared among participants. The notes and the exam scripts were analyzed using concepts central to complexity theory, by identifying attractors in play as well as analyzing reasons for action. The arguments participants presented and the attractors were also analyzed in the light of professional versus managerial values. Junctures when discretion was exercised were identified as bifurcation points-they had to choose between at least two options, where one would regard the follow up of families, while the other is a managerial task. This is primarily what we mean by a 'dilemma' between managerial and professional interests.

Examples of Attractors-Distinctions at Play in Our Empirical Data
The list below is a selection of some of the arguments interpreted as representing attractors or symbolic capital. We choose not to comment on every one of these and rather present the data as examples of the points we want to make here. The different attractors may be discussed as values within certain social fields, but we want to present them as options rather than preferred directions of good practices. In the real world, the decision to move in a certain direction has to be seen and understood in relation to the whole context of the decision. While reading the extracts, one should bear in mind the six points referred to above as non-eliminable hazards. Students emphasized issues such as:


Time to do follow up with a few cases versus handling a large case load and box ticking  Looking at parenting skills rather than poverty  Risk assessment versus engaging with families expressing fear and concern of undesirable futures  Use of manuals (Kvello in Norway and BBiC in Sweden) versus listening to parents and children's own accounts of their needs  Enhancing user involvement and 'letting go' of certain routines and established ways of moving forward-relinquishing managerial control  Focusing on learning and evaluation rather than experiencing lack of success as a failure  Focus on cost of provisions/investments rather than families and children's needs  Evidence based practices hampered by costly learning and skills improving programs with certification  Reduction of compensating provisions (risk reduction and proactive actions), which suffer because priority is given to short term change investments  Retreat to a belief in resilience rather than taking responsibility for risk reduction  Class based, culturalized sets of thinking that involve treating the precariat in different ways to middle class families.
Höybye-Mortensen's (2015) point about moving outside the available service package may also be relevant here, as well as in the examples of bifurcation points provided below (This analysis does not take into account any actions or opinions of families, since this may be discussed elsewhere. The interviews with some of the families involved show some of the same as previous research-that many social workers often cancel appointments, there is a high turnover of workers, it is hard to build trustful relationships and not much new is tried-things may start or be discussed, but not followed by new case workers.).

Examples of Bifurcation Points
The following extracts are presented not as interpretations but as direct quotations from the texts analyzed. One should have in mind that there are reflections made after the social workers looked back and discussed what may have influenced them at different bifurcation points to the extent that the service discarded the family rather than choosing to follow up with them more closely. This included for families where no harm or injury threatened the safeguarding of the children. What the service did not manage was supporting needs and develop new resources (symbolic capital) to increase the life chances and access to symbolic capital for children and adults. The headings are our interpretation of what is at stake.

Strict Routines and Guidelines Versus Individually Crafted Social Work
At what point during the assessment do we start to consider a family as a complex problem as opposed to an easy one-a problem that is suitable for standard procedures and actions? A family included in the project may have complex needs, but these needs may not lead to any action, since there is no immediate threat to the child or the situation. Identifying a problem that has a known solution may simplify work with a child in a family with complex needs, without regarding the whole situation in all its complexity. But complexity does not necessarily result in complications.

Management Trumping Professional Judgement and Other Constraints
Based on my own experience, I wonder if we ever will have access to resources that may help the most complex families. Even if they were within reach, we are hampered by individual factors, framing by the organization and management as well as the culture at work. Resources seem to be regarded as the usually accessible routine solutions rather than as solutions that could be deployed to support innovative practices.
When you face resistance, regard it as being energy in play rather than undesirable and as a personal failure.
When a social worker sees matters in this way, it may lead them towards innovation in terms of assessment as well as in addressing questions as to what help a child and family may need. This may affect other decisions and their attitude to discretion in general.
Do we need to move some of the social work out of the bureaucratic child protection service to other parts of the organization?
This question arose partly on the basis of experience from a Swedish project in which a team was established to work only with families with complex needs, a unit to a considerable extent being allowed to dispense with the strictures enforced in the rest of the service. The reorganization here is becoming a bifurcation point.


Do we need to acknowledge the lack of knowledge about children at risk, and families with complex needs and start using scientific work in a more elaborate sense-to research our practices?
 If love and good intentions are virtues in social work practice, how is it possible to not give priority to children and families in need-and how do we organize the work?  Do we need to look at more interprofessional and systems interplay-and how might this improve our ability to provide appropriate responses?
These last points comply in a certain degree with the challenges presented in Rawls' list on noneliminable hazards that influence reasoning. A reductionism related to ideas of practical reason is also found in Bourdieu (1999).
A case presented by a student concerned a mother who confesses to a serious long-lasting drug problem. Her children are not removed, but her network and family are activated as supporting resources while she is undergoing treatment. This work goes on for months without any set goals besides working with the recognized problem-an approach that later is regarded as a decisive resource to stabilize the situation. The mother is also diagnosed with a neuropsychiatric problem; the family will be in need of assistance as long as the children are involved and probably even longer. This is not efficient problem solving, it involves allocating resources to a chronic situation-but placing the children in an adoptive situation would have been even more expensive, so it is defended as economically sustainable, and might leave the mother and the child happier for a while? Who knows if the child someday might sue the municipality for not removing him or her to provide better life chances in a more stable family?
The kind of work carried out in services seems to be of great interest for many, and several of the social workers talked about turnover as a result of personal interest in what kind of professional work they preferred. Those who moved into a 'task force' for families with complex needs valued their work as interesting and challenging in positive terms, while regretting the time they earlier had to spend on management of heavy caseloads. In complexity theory terms, a more professional focus on a few families as an attractor and the establishment of a new team in the service making this possible, became a bifurcation point where personal preferences influenced their decisions. In social theory, this is also a preference for a certain symbolic capital where values and professional interest can be identified as distinctions defining a social field. Those who worked in this 'family task force' team made up their own social field where there were integrated professional activities and processes of change rather than set goals that they worked to attain. User involvement and recognition were also central values in this field-values that may be part of the rhetoric in the system as such but could in this situation be realized with closer and more continuous involvement on the social workers' part.
An interesting comment made was that in order to apply new knowledge, take part in research and give recognition to others' knowledge, the service or organization has to be open to innovative practices and experiments. This corresponds to Eräsaari's (2003) idea of open rather than closed expertise as Karvinen Niinikoski has discussed: The position of scientists and professionals as experts and knowledge creators is re-constructed in relation to the expertise contained in the personal experience of practitioners and citizens, the users and providers of human services in our case. The new ideas of expertise and knowledge emphasize new kinds of negotiations, co-operational and networking environments in the processes of learning and of knowledge production. Expertise and knowledge are understood as more open and dialogical and even conditional for negotiation. They are seen to be context-dependent (Nowotny 2000), the context being an important source for generation and validation of knowledge. Additionally, the knowledge of street level professionals and the lay-expertise of service-users are seen as necessary parts in the dialogue. There is a need for new kinds of mechanisms of innovative knowledge production, forums for dialogue in promoting knowledge creation and an epistemic pluralism (e.g., Nonaka et al. 2000;Nowotny et al. 2004). All this sounds familiar from the angle of social work and could be called reflexive expertise (Karvinen-Niinikoski 2005, p. 264).

Discussion
According to Molander (2016), discretion involves possessing negative liberty that provides the actor with an area for choice and action consisting of those options that are neither forbidden nor prescribed (Molander 2016, p. 9). The involved 'risk assessments' may include both a possible client orientation as well as an agency orientation with different attractors involved. Since the managerial distinctions related to the symbolic capital identified by the checklists and monitored deadlines and actions so easily may backfire if not complied with, the non-feedback system from doing good social work with clients is left in a black box only partly known to families and children and the social worker involved. Based on the discussions in sessions with our students as well as the exam papers analyzed, the point of departure is an acknowledgement of the failure to meet needs for families and in general the time needed to do long term follow up of many kinds of problems that users may present. This is not in line with ideas of professional standards or best practices. These social workers looked for ways of coping and ways of overcoming the resistance they faced in wanting to promote change. The exam papers as well as the discussions had a strong focus on available time as a major deficiency in the services. Lack of knowledge does not seem to be regarded as a major obstacle, since it may be met by reviewing accessible knowledge and by supervision as well as practice research. Enhancing user involvement may also increase the ability to meet needs in better and more efficient ways. The social workers also looked at ways of reducing bureaucracy for social workers by using assistants in recording and reporting. There also seems to be some disagreement about why everybody within an organization should necessarily have to take part in innovations-why is it not sufficient to engage only those involved and management. It is argued that this may reduce tensions and resistance, since many do not want to rearrange their priorities (distinctions and different attractors).
It has to be pointed out that the social workers in our study are not representative of the child protection field in general, as they have seen a good deal of research involvement in their local practices over many years. On the other hand, the knowledge developed among staff members, and especially those attending the master course, may have enhanced their ability to identify and describe situations and settings. Another problem may be the fact that we have not tested where they have their loyalty according to Billingsley's concepts. This might have indicated problems and challenges that we do not have any knowledge about. Our data does not allow for any categorization of role conceptions such as professionals, bureaucrats, conformists and innovators, which may have been useful. Some workers seemed to fit into some category that may seem relevant or belonged to two, but we did not use Billingsley´s categorization.
Supplementing complexity theory with culture sociology (in particular Bourdieu's symbolic capital concepts) one can regard management's focus on goal attainment, certain ideas of efficiency related to quality and work within limited timeframes that result in desired change as distinctions counting as symbolic capital within the social field of management and in the interface with politics. Social workers who find themselves at these bifurcation points seem to be exposed to some kind of leverage where professional values related to individual needs that families have presented have to be put on hold or ignored in order to avoid criticism that may ensue from breaching efficiency standards. This seems to be one of the factors influencing the fact that families with complex needs are not given priority.
Innovative practices could in time lead to better social work, but taking risks with improvisation to devise new ways of being in dialogue with families may also lead to unwanted or unforeseen negative results, such as increased risk of domestic violence, child abuse or maltreatment. This would invite even more criticism and perhaps sanction directed at the service and the social workers involved. At such bifurcation points, it seems the system prefers safety rather than innovation.
Social workers describe what we have dubbed the 'vicious circle': they have a heavy workload, they end up falling behind in their work and risk getting behind, leading to criticism from the inspectorate. They see their workplaces as understaffed, leading to sick leave being taken and burnout. Research may offer some possible strategies to avoid this. One is to develop more precision and efficiency in getting on with cases and obtaining more effective outcomes. A second strategy is more elaborate and involves better prevention and early intervention. The problem with the latter is the extra resources needed in order to do good social work and also have a health promotion strategy, work inclusion, etc. This may require changes to society that might be thought unlikely under prevailing neo-liberal conditions (Garrett 2009;Featherstone et al. 2016).
Our experience is that one may move towards innovation and change during a project, but the system reverts to 'normal' or the status quo after some time due to a lack of integration of changes in the total infrastructure of the organization. Change thus becomes dependent on knowledge management, action learning as well as devoted and competent workers and the available resources to continue at a higher level of intervention. Discretion may also imply actual disobedience when work and practices run against a professional interpretation of the client's rights and legitimate demands in respect of services, but this requires the will to oppose management or institutional policies (Kadish andKadish 2010, in Evans 2016). Turnover, organizational change and continuous increasing pressure due to unrelenting reporting of new cases seem to force systems back to a restart. Processes of change may thus have to be regarded as the modus operandi, and in order to secure some impact from applied social and other research, the system will have to include research as an integral part of the organization-a complex practice with an interplay between child protection work, innovative research, evaluation and a management that has the resources to support both long term development and everyday priorities. A relevant question is how likely it may be that a focus on knowledge development and professional value based needs-orientations can be maintained in a situation where management seems to be ever more constrained by the downward pressure imposed by policy: the desired policy results being all the time operationalized in detail via computer systems that do not reflect the way that needs will have to be met in the 'real' world encountered in social work with children and families.

Author Contributions:
The text is produced in cooperation between all authors based on research by the same, E.M. rewriting the final text after supplements from all authors. All authors have read and agreed to the published version of the manuscript.

Funding:
The research is primarily funded by the Norwegian University of Science and Technology and Inland Norway University of Applied Sciences, in addition to a three year grant from the Norwegian Research Board (NFR).

Conflicts of Interest:
The authors declare no conflict of interest.