Organizational Professionalism: Social Workers Negotiating Tools of NPM

This article examines how social workers and managers perceive meaningful work and expertise in six care and treatment facilities in Denmark. Based on 29 interviews with social workers (n=22) and managers (n=7), the article shows how New Public Management-inspired tools such as scoring schemas align with social work values such as “client-centeredness” and working with the individual welfare recipient face-to-face. The article finds that fitting social work into organizational schemas changes the work practices of social workers and also the way members of this profession define meaningful work and expertise. In addition, the article also finds that scoring schemas cause conflicts among social workers regarding the character of expertise when values of social work (to meet a welfare recipient’s need) must be aligned with NPM-inspired values of organizations (to meet managers’ demand for documentation).

Since the late 1980s, European welfare states have undergone substantial changes concerning their objectives, areas of intervention and instruments of use (Bonoli & Natalie, 2012;Clarke, 2004). Public organizations are no longer primarily characterised by the classic traits of bureaucracy, for instance, routine-based office work and standardised administrative procedures (Sturdy, Wright, & Wylie, 2016). Public organizations have incorporated new systems of management, such as performance management, manual-based practices, New Public Management (NPM), scoring schemas, and other tools. These tools aim to provide high quality and effective work. A so-called managerial discourse has entered social work organizations (Banks, 2013;Clarke, 1996;Farrell & Morris, 2003;McDonald, Postle, & Dawson, 2008;Rogowski, 2010;Roysum, 2013;Shanks, Lundström, & Wiklund, 2015) as well as nursing practices (Carvalho, 2014;Debesay, Harsløf, Rechel, & Vike, 2014;Newman & Lawler, 2009) and doctors' work (Benish, 2014). Within the field of social work, research shows how a managerial discourse has suppressed classic social work values such as placing the welfare recipient at the centre of the work (Clarke, 1996;Ferguson, 2009) as well as focusing on the needs of the client (Rogowski, 2010).
The current study adds to this strand of research by investigating how social workers in their everyday work life manage organizational scoring schemas in care and treatment facilities housing vulnerable clients. The research question of the article is following: how do social workers perceive meaningful work and expertise in care and treatment facilities and how does this assessment relate to the increased influence of organizational scoring schemas and tables?

Negotiating meaningful work and expertise in social work organizations
Some scholars argue that modern organizations are "Kafkaesque" because they operate on a logic that integrates very different rationales (Hodson, Martin, Lopez, & Roscigno, 2012). This unusual description pinpoints the fact that professionals are simultaneously held accountable according to values stemming from both their disciplinary background and according to bureaucratic and NPM-tools such as scoring schemas. According to Hodson and colleagues (2012), the co-existence of these different values may result in "chaos" in modern organizations (see also Mik-Meyer, 2017). For instance, is social workers' expertise to be measured against their ability to meet the demands for documentation set by their managers or should their expertise be measured against their ability to actually help to solve welfare recipients troubles (Rogowski, 2011)?
Whether current organizations are Kafkaesque and chaotic or not, they provide a paradoxical reality for the professionals working in them (Clarke, Newman, Smith, Vidler, & Westmarland, 2007;Evetts, 2009b;Ferguson, 2009;Mik-Meyer, 2017). Thus, vulnerable clients complex troubles fit badly with the current scoring schemas and guidelines of social work organizations (Gubrium & Holstein, 2001;Gubrium & Järvinen, 2014;Loseke, 2001;Mik-Meyer & Villadsen, 2014).The apparent context-independence of these new tools challenge the professional values of what it takes to conduct meaningful work and what it means to be professional (Kallinikos, 2004). Even though scoring schemas, guidelines and other organizational tools "appear to hold out the promise of greater reliability by establishing systematic, objective and uniform practices," these tools also suppress the "real need of users" and the importance of working individually with each welfare recipient and his or her troubles (Rogowski, 2010, p. 159, my emphasis).
When social workers translate their work into scoring schemas, they meet documentation demands of managers, and when they engage in pedagogical relations with welfare recipients, they accommodate classic values of social work. The tension of these dual goals result in new forms of expertise and-consequently-the social work professions' perception of what it takes to conduct meaningful work. Eyal's (2013, p. 864) definition of expertise as "networks that link objects, actors, techniques, devices, and institutional and spatial arrangements," locates the arrangements that are in place for staff to complete an assignment as well as the effects that the particular techniques and tools have on the work. Following this definition, expertise is the things organizational staff "do" and is not (only) related to the particular profession's disciplinary background (who they "are") (Carr, 2010).
Sometimes organizational tools hinder professions in conducting what they consider to be meaningful work. This dilemma is one of the reasons why scholars today engage in lengthy debates about what expertise looks like in modern organizations (Aldridge & Evetts, 2003;Baines, Charlesworth, Cunningham, & Dassinger, 2012;Broadbent, Dietrich, & Roberts, 1997;Evetts, 2009a;Freidson, 2004Freidson, , 2014Liljegren, 2012;Noordegraaf, 2007;Saks, 2012). This current scholarly discussion of expertise emphasises that professions do not occur in a pure form in any organizational context, which is defined in the notions of organizational professionalism (Evetts, 2009a), pragmatic professionalism (Liljegren, 2012), and NPM-professionalism (Evetts, 2009a). These concepts show that professionals in their daily work combine the managerial tools of the organization they work in with the disciplinary knowledge of their profession (Noordegraaf, 2015). Evetts (2006Evetts ( , 2009aEvetts ( , 2009bEvetts ( , 2011 suggests the concept of organizational professionalism to pinpoint the important role of organizational factors-broadly speak-ing-in work, whereas occupational professionalism is a concept that directs attention to how professional values control the work. However, as Evetts explicitly states, both types of professionalism are ideal types. Everyday organizational work thus involves pragmatic professionalism (Liljegren, 2012, p. 309) in which staff combine the values and tools of their work organization with those of their occupational backgrounds.
Finally, the concept of NPM-professionalism (Evetts, 2009a) suggests that a managerial discourse has a strong effect on work conducted by professionals in modern organizations today and that the definition of expertise is therefore closely related to the values of NPM. The scholars that make this argument state that NPM values such as scoring schemas, performance measures, standardisation tools and so forth are very dominant in today's welfare organizations (Clarke, 2004(Clarke, , 2005Clarke et al., 2007;Denhardt & Denhardt, 2000Ferguson, 2009;Fountain, 2001;Jos & Tompkins, 2009;Kuhlmann, Allsop, & Saks, 2009;McCafferty, 2010;Rogowski, 2010). This branch of research also claims that NPM values create many dilemmas for professionals. For instance, social workers who give aid to homeless persons often work with highly ambiguous goals (Smith-Carrier & Lawlor, 2017;Stonehouse, Threlkeld, & Farmer, 2015) as they must sort out the complex troubles of the homeless individuals' situation as well as negotiate with them what a better life entails for them (Dwyer, Bowpitt, Sundin, & Weinstein, 2015). This relational work must hereafter be translated and documented into scoring schemas (Farrell & Morris, 2003;McDonald et al., 2008;Shanks et al., 2015). Hence, social workers cannot rely (only) on their professional skills deriving from their discipline when dealing with the imprecise nature of homeless persons' troubles; they are also dependent on organizational scoring schemas and guidelines as well as observations that can be "quantified" (McDonald et al., 2008. It is this tension in the work of the social work profession that is in focus in the upcoming analysis.

The study
This article's analysis is part of a larger project that examines vulnerable welfare recipients' situation in different welfare organizations in Denmark (Mik-Meyer, 2018). The larger project concerned how welfare recipients perceived their situation and how a broad range of professions and their managers described and tried to resolve welfare recipients' troubles in different organizational environments. In order to examine welfare recipient and staff perceptions of social troubles, I conducted interviews with 58 staff members, 21 welfare recipients, and eight managers. Most of the participants worked (staff) or were aided (welfare recipients) at care and treatment facilities such as homeless shelters, substance abuse clinics and social psychiatric clinics. The interviewed persons at these facilities included 22 social workers, eight nurses, six social and health assistants, seven managers who were mostly trained in the academia, and four doctors and police officers. After my research assistants had coded these 47 interviews, I discovered an issue that has been given little attention in empirical studies, namely, how the social worker profession and their managers working with vulnerable welfare recipients negotiate expertise and how their assessments relate to organizational tools such as scoring schemas and guidelines.

Data, procedure, and analysis
The current article's analysis is based on 29 interviews with social workers and managers working at six Danish care and treatment facilities. To engage participants in the study, I approached the managers who then provided contact with the social workers. All participants took part anonymously, which means that interviewees' names are fictional. The interviews were steered by an interview guide that included wide-ranging open-ended questions, and the focus of the interview was on how the interviewees perceived the scoring schemas and documentation work of the organization. For instance, among my questions were themes such as the organization of the work (e.g., How are the tasks organized?); documentation practices (e.g., How do you fit your observations in to the scoring schemas of the organization?), perception of professionalism (How are your disciplinary training and/or organizational factors affecting your work?), norms and ethics in the job (What norms and ethics do you consider important for your profession, the organization and/or the manager?), and the needs of the clients (What do you prioritise in your relation to welfare recipients?) After all interviews had been transcribed verbatim, I re-read each interview in its full length to obtain an in-depth understanding of the material at large, for instance, the retold story about what to report in the organizational scoring schemas and which observations were considered "trivial." Hereafter all interviews were coded using the NVIVO software programme. The initial coding process included a broad range of issues, which were followed by a focused coding process (Charmaz, 2006) that resulted in categories such as "documentation practices," "computer work," "social workers' opinions of meaningful work/expertise" and similar. These categories included interviewees' stories about how they managed their reporting, their opinions about their assignments, and their own and their colleagues' expertise. I then compiled a list of the main topics in each category, for instance, the choice of using third person abbreviation of "signed" (detachment) in the individual reports. These topics were then highlighted in the text, and this highlighting was the direct point of departure for the detailed-often verbatim-analysis of the specific interview sequence (which always included the interviewer's questions and responses).
In disseminating my analysis, I have included a table with "proof quotes" (Pratt, 2008), that is, the many short, single quotes that are the analysis's outset and justify it. The proof quotes make it clear that the analysis is not only based on the opinions of a few social workers (see Table 1). However, the analysis presented in the paper is predominantly based on "power quotes" (Pratt, 2008), that is, longer quotes and dialogue sequences that illustrate the point of the analysis in a striking way.
The structure of the analysis is threefold. First, I present a brief introductory analysis of how the interviewees explain what they consider to be the core aspects of social work when conducted with welfare recipients with complex troubles. Second, I conduct a more detailed analysis of how social workers from two participating homeless shelters work with the scoring schemas and guidelines of their organizations. Third, I look at seven reports on the attitudes and behaviour of homeless persons living at a shelter (each report is written by 510 staff members) to show in detail the challenges social workers face when they must document the troubles and situations of homeless people in accordance with organizational scoring schemas.

Relational work and scoring schemas
Using scoring schemas and other NPM-inspired organizational tools in public organizations is spreading and has steadily won recognition in social work that is directed towards vulnerable groups (e.g., Ferguson, 2009;Rogowski, 2010). Written reports of staff have long been the norm for the clinical treatment of welfare recipients who abuse drugs and in psychiatric diagnostic work. However, it is a newer phenomenon for social workers to use scoring schemas and other organizational guidelines in their work with this group of troubled persons. According to two managers at a homeless shelter, their shelters' scoring schemas result from an increased requirement for evidence-based practice in social work-including work in homeless shelters. This belief is supported by multiple resources that are used to develop and test these tools (and their content). For instance, the participating social workers working at the organizations for homeless, drug addicts or persons with mental health troubles explain that their main job is to engage in pedagogical rewarding relationships with vulnerable clients/patients in order to help this group of individuals solve quite complex troubles. However, they also explain that they spend a significant part of their jobs recording observations in the scoring schemas of their organizations. Even according to a middle manager, the amount of reports at her organization has "grown too much" and is "useless." The idea that "if it's not written, then it did not happen" [as our manager declared] is very problematic. This solution generates too many entries where staff try to prove what they do at work. In addition, we cannot use these reports at all. … We do not need documentation of "Oh great, you played a board game and had a cosy evening." In this version, descriptions of games and cosiness should not be in the client reports of the organization. Her opinion identifies a core tension for the participating social workers-they work within a managerial discourse that requires documentation ("If it's not written, then it did not happen"), but are at the same time expected to engage in a pedagogical relationship with the welfare recipients which may involve playing board games, and so forth. However, the scoring schemas of the organizations leave no room to record observations that originate from their "relational work" with welfare recipients, as social workers termed this pedagogical work. As Table 1 shows, relational work is a central characteristic of social work in the care and treatment facilities.
Relational work occurs in the daily morning gatherings of the organizations in which the social workers advise the welfare recipients to behave; in ball games and during other physical activities; and is the take-off point for computer activities where social workers collaborate with the welfare recipient on how to check e-mails, find letters from the municipality, and so on. "It is alpha and omega" as a social worker at a clinic for substance abusers explains. To him "relational work is the foundation for helping welfare recipients to develop." Relational work is also central to social worker Susan, who explains that she uses herself "a lot in order to become part of these peoples' lives." However, the breadth of the relational work also makes many social workers conclude that their work is diffuse, time-consuming and difficult to document. It is not always "that easy," as Alma says, it can be "a big task" and an "uphill" battle, as her social worker colleagues Sally and Laura explain. Staff's problem arises where not all welfare recipients are ready to enter this kind of pedagogical and psychological relation to the staff. However, the problems mentioned by staff also concern the fact that this kind of work is difficult to document in the scoring schemas of the organizations in which they work.

Alma 40s
"All that pain and mischief, they arrive with, has happened in a relationship, and this is why I think they can only be cured by working on the relation." "I have been struggling a lot with the reports because it is very important to have eye contact with the resident." "I think documentation is important." "If no one reads it, then there is no continuity."

Caitlin 40s
"The most important thing is to establish a relationship based on trust." "Documentation can be annoying … if you are in the middle of something with a resident." "You cannot remember everything, so it gives you the opportunity to go in and read."

Sally 30s
"You can never have enough of it [relational work.]" "Sometimes I think it is hard to find time for writing reports." "It gives a picture of what has happened [with the resident]."

Pam 40s
"It is good to be engaged in an activity with other people about something social." "No [documentation has not caused any problems]." "I rate on their stability, how often do they show up, how well do they perform, and such things." "If someone is feeling bad, it is important to give the person a high rating, because then you can pull it out and get an overall picture of the person. So it makes good sense."

Peter 20s
"By establishing a good relationship, it gets easier to work towards a range of ends." "You miss the time on the floor to do relational work because you have to document everything." "It is an exceptionally good work instrument. And a very good way to get an insight of what you are really dealing with. … But I think that health professionals are more used to use it from their training."

Tony 50s
"To gather around some kind of activity and have a relation.

Ben 40s
"Focusing on the relationship and why it went wrong." "The more we write, the further away we get from the person." "I think it is a big dilemma." "I think it is useful in order to pass information on to my colleagues. … But in the everyday work, I would like it to be an easier system."

Susan 40s
"To help them become part of society again. That it is okay to be a little different. That there is also room for [them]." "We use a lot of time on documentation. Sometimes more time than on contact with the person. It is not always connected." "It is okay. It is fine."

Margaret 40s
"Overall it is to connect these people to a range of public agencies that can help them out of the social status they are in."

Theresa 40s
"To focus on the resident and to promote development." "It takes a lot of time. It certainly does." "It is good and bad." "It enables you to get a view over the development since the last meeting. It is more concrete, whereas earlier it was broader."

Robert 40s
"Our most precious task is to meet them where they are. … It is the relation that is the most important. The presence is crucial in order to establish cooperation." --

Paul 30s
"We seek to help them by establishing a good relationship with them." --

Samantha 50s
"To me, it is really, really important that there is trust. That they feel safe around me; and to have time together. To have the time if someone comes up and needs to talk." "23 months ago 80 % of my work was administrative." "But it has changed, and I would say some of [the documentation], I have asked myself, 'What use is this for?'" "I would rather have conversations with residents. But it is not a big problem for me to sit behind a computer."

Vicky 20s
"We look at abuse, health … and finances. Sometimes it is about maintenance of basic skills, other times of course about personal development." "Yes, it is. I think documentation is difficult. I think the hardest part is to do it collectively as an organization, and use it in the same way. That is the hardest." "I think the idea about it is really, really good because you can extract ratings on something very specific. For example, abuse or contact."

Luke 30s
"To support the most basic everyday needs. Definitely food and medicine." "I am in my 30s; I am basically born with a computer, so I do not see any problems with it. It seems logical to me." "[It provides] control and structure of what is going in every individual's life."

Vanessa 50s
"We want to give our residents a sense of life quality and to a better life on their premises. What is good for them is probably not good for me, but when they are satisfied, we have to accept it." "Every time we are told to spend more time on documentation, it is taken from time with the resident whom we still do not treat properly. It is so frustrating." "It is important to journalize, and it is important to produce statistics, but let me 'have' fewer residents in my pool if I have to do both instead of making me run around beheaded."

Janice 40s
"Housing first and recovery. And also I think it is important to keep up their hope and faith that it can be better, that we can help them." --

Documentation practices in two homeless shelters
At two of the participating shelters, staff report observations of welfare recipients in both text and numbers. Staff "rate" the welfare recipients' need for support in different situations from 0 to 4. This numerical assessment can be printed as graphs that show progress according to several pre-defined parameters, such as "empowering activities" and "mental well-being" as it says in the guideline. Most staff find writing reports a time-consuming task. In the following sequence, social worker Samantha elaborates about the issue of time: Every time you jot down an observation, you rate it. In addition, then you can get such fine graphs, so you can determine if the progress [of the welfare recipient's situation] goes back or forth. … Therefore, I can see the intention in it, but Jesus Christ! [Both laughing]. Well, it takes much time and requires many resources too. We have been understaffed for a long period. People are tearing their hair out and having breakdowns due to stress.
Although Samantha provides a graphic account of the flipside of documenting their work, social workers' criticisms are generally also accompanied by positive assessments of writing reports as found in other research (e.g., Banks, 2013;Farrell & Morris, 2003). The participating social workers emphasise that the scoring schemas create continuity and coherence in their daily work. Many argue that reporting supports a consciousness of their methods, "you can extract ratings on something very specific" and clarify where to "put some effort," as Vicky, a social worker, explains. A colleague agrees and states that reporting "focuses" the work by giving staff "goals and a way to structure their work." However, the question is how the demand for increased reporting by the social work profession align with this profession's focus on relational work with the welfare recipient. The following dialogue provides an elaborate and illustrative description of how a social worker (Peter) perceives ratings and the use of this instrument (no words have been left out in the dialogue).
Interviewer: Do you think that ratings are useful in the work, from your perspective as a social worker? Peter: Yes, very much I think.
Interviewer: Yes. Ratings agree with your discipline?
Peter: Yes eh or [pause]. I'm having a hard time answering that question because I think maybe it… Therefore, I can easily see that ratings are very, very useful. Therefore, it's an extraordinarily good work tool. In addition, a very good way to get an insight of what you are really dealing with. Do you have to address someone who is well functioning? Where are the needs of support? In that way, it's a very, very useful tool. However, I think that nurses are more used to using it from their training. I mean, they address facts, and this is what you are aiming at. Where… In comparison, you are much more… I think in the study programme of social workers there is not really any… There is not really… Well, of course, there is some theory and some knowledge and… However, it is very much like, if you are good at arguing, then whatever you are saying is almost as true as what I am saying. Sometimes it is like, to put it in a very radical way… then I think that sometimes… then I sort of think that the study programme of social workers is not so fact-oriented. I mean not like that of nurses. Therefore, I believe in that sense that it is a good instrument for us social workers… That something more… Somehow something more… Of course, you can structure your work and set some goals for what you want to achieve. However, the very concreteness of scoring schemas allows you to say, "Well, that is the way it is, because that is what you can see." You cannot say, "Well, maybe because…" or "That is be-cause…" Therefore, in that sense, it's a very, very good tool. In addition, it's much more constructive to do it that way.
Rating and scoring in schemas provide something you can "see," as Peter emphasises, and this clarity is "really, really good." His long monologue shows that he is very appreciative of the organizational demand of documenting the work because scoring your work in schemas means that his profession's conclusions also will hold weight in the organization. However, his hesitant and constant self-interruption leads to another parallel analysis that points to another aspect of social workers' documentation. It may be the subjectivity of his professions' "argumentation" that makes it difficult for this profession to fill out the scoring schemas. While he is not proud of the fact that there is no universal way of deciding what is correct and false in social work, it is at the same time this exact focus on relational work and the context dependency of social work that defines meaningful work and expertise among social workers (see Table 1). It requires hesitation, pauses, and several attempts to draw a conclusion that his discipline is not "fact-oriented" ("then I think…," "then I sort of think…," "that some-thing…," "somehow something…").
Health professions, on the other hand, report their observations in short and precise language, as he explains. Their expertise is in controlling medicine and measuring symptoms in pulse and blood pressure, and so forth. In comparison, the expertise of social workers involves reflections on how individuals behave in situations, for example, when playing a board game. Social workers' jobs include collaborating with the welfare recipient to push for changes in his or her life. Peter's social worker colleague Tony calls these things "to feel and sense." Unlike Peter, Tony is not enthusiastic about the numeric rating and organizational scoring schemas in his work. He rolls his eyes when the interviewer talks about the categories that are the basis for their "rating." Interviewer: What about these categories? You rate from 0-4, right? Now you're rolling your eyes, why is that? Tony: It's hard to do and then why do we do it? Rating has to do with evaluating welfare recipients' behaviour against different criteria. Then you can do a graph that displays, for example, an increase in alcohol consumption followed by escalating levels of problems with social contact. You can register correspondence between the one curve that goes upwards and another curve that goes upwards. Then you can look at them. However, in my opinion, there are way too many possible ways to report and too little consensus on what we should measure, why and when. Interviewer: So, you can actually measure everything? Tony: Yes.
Tony's body and spoken language show doubt in how rating can contribute to clarifying the social problems of the homeless individuals at his shelter. He explains in addition that ratings contribute to extensive paperwork and reports. He problematizes the absence of a consensus of what, why, and when to measure and thereby identifies social workers' chief problem-in principle, all welfare recipients' behaviour can be reported in the schemas of the shelter he works in. Thus, his frustration concerns the complexity and ambiguity of social problems that poorly align with the organizational scoring schemas. Whether positive (Peter) or negative (Tony), when evaluating the effect of scoring schemas in social work in homeless shelters, the interviews with all participating social workers show that most often the individual interview person is both positive and negative, when reflecting on NPM-inspired work tools such as scoring schemas and other documentary practices. This ambivalence of what to think about these newer tools shows that expertise is no longer only-or first and foremost-evaluated according to values embedded in the disciplines of professions. Expertise and perceptions of meaningful work are also evaluated according to values embedded in the managerial discourse present at their work organizations. Thus, current NPM-tools changes the work and result in new perceptions of what social work entails and how to define expertise among this group of professionals. This is an example of what Evetts (2009a) call organizational professionalism or NPM-professionalism.

Writing reports in a homeless shelter
Social workers' measure the need for aid on topics that are hard to define such as "empowering activities," "employment activities," "mental well-being," and so forth, whereas nurses assess easy-to-define topics such as "medicine intake." For instance, a social worker notes about a welfare recipient, "Michael has asked for help to make sense of letters from the homeless unit" and rates her assistance to "3" under the topic "skills of perception," because there was a "significant need for social assistance." In other words, the expertise here has to do with measuring a quite complex area (e.g., skills of perception) in a relatively trivial way. Another distinct aspect of reports from social workers is the key role that he/she plays in the reports and its relation to social workers desired a focus on relational work with the welfare recipients. A third characteristic of the reports is social workers use of third person abbreviation of "signed" ("SN") in the report to hide the subjectivity of the many reported observations. Thus, social workers reports are stories of relationships that include both staff and welfare recipients as the main characters even though the staff's identity is often hidden under SN. The following extract of a report from a social worker (SN), illustrates this.
Ibrahim looks SN up in the kitchen and is obviously sad. Ibrahim says he needs to talk. SN agrees with Ibrahim to meet in the rainbow office. Shortly thereafter, Ibrahim arrives at the office, telling SN that in addition to mental health problems, he is also physically ill.… Ibrahim says he hates himself because he was once a nice guy with work, and so on. SN tells Ibrahim that even though he has mental challenges, he should not hate himself. SN comforts and tells Ibrahim that there is no reason to hate himself because he is still a nice guy. SN tells Ibrahim that he must keep fighting to get better and even though SN has a job now, then SN can also suddenly get sick and get sad, but that one must not give up.… Be-cause Ibrahim has not eaten for several days, SN explains the importance of correct diet for well-being.
This sequence from a report highlights how relational work looks like in practice. Note that this report includes descriptions of how the social worker thinks Ibrahim's situation can develop, and the social workers own role in this identity work. This is a story of Ibrahim as well as a story of the social worker's role in helping Ibrahim get better. When the social worker describes himself in the third person (SN) as a substitute for using the first person ("I" and "me"), he makes a strong case for detachment. It is not him as a named person who interprets Ibrahim's situation; it is any person ("SN"). The report of this client is accompanied by the score "2" in the category "psychological well-being" and is in harmony with this scoring schema's demands for numbers. By attributing the text to the score "2" an alternative way of interpreting the report is made. The added number shifts the focus away from the nuances and ambiguities of Ibrahim's situation (and social work) and underscores the detachment of the social worker's observations. His repeated use of "SN" is characteristic in the reports of the social workers and shows the basic dilemma of documenting relational work. Thus, the social worker's role is pivotal on the record-as social work requires-but as a person whose identity is hidden. When reviewing all of the reports of the 58 participating staff members, the profession of social workers uses SN much more frequently in their reports compared to other staff members with different training. This indicates that social workers attempt to conform to the neutrality of organizational guidelines. However, it is doubtful if the social worker's use of SN instead of "I" or "me" solves this basic problem of the lack of "fact orientation" in social work, as Peter was stating earlier. By using SN no less than nine times in the short (condensed) report, the reader is left with a very strong impression that the reported observation is a report in which both parties played an essential role.
The social workers who explained in interviews that they wished for a stronger fact-orientation may thereby demonstrate that they do not personally identify with the key characteristic of their profession (to work upon relationships). However, this might not be the only interpretation-or the most likely. Their wish for more facts may also indicate that the lack of factual orientations in social work causes the social work profession to be ranked lowest in the hierarchy of professions. Because staff's documentation of welfare recipients' situations is conducted by tools that favour objectivity and detachment from the writer, social workers' expertise may be threatened more than their colleagues with other training such as, for instance, nurses. This finding suggests that scoring schemas and other NPM-inspired tools have effects on the way that expertise is defined in this type of organization. The current strong culture of documenting the work in scoring schemas as demonstrated in the two participating shelters may, therefore, cause new battles of professions (Abbott, 1988) around whether and how professions are successful in integrating NPM-inspired tools in what they consider to be their core competencies.

Concluding discussion
This present study has shown how social workers negotiate what counts as expertise in care and treatment facilities today. The analysis here shows that the expertise of social workers concerns their success in jotting down their assessments of clients in organizational scoring schemas. Thus, expertise has to do with integrating and quantifying social workers' relational work with welfare recipients in schemas and reports. In line with McDonald and colleagues (2008), this study also found that expertise concerns a numerical assessment of client behaviour. Hence, assignments that social workers cannot "quantify" are automatically defined as less important and depict a low level of expertise.
The article's analysis also reveals how work in care and treatment facilities contains two opposing ideals. On the one hand, perceptions of social workers' expertise have to do with their success in building a personal relation with welfare recipients (relational work) such as playing board games, soccer, and so forth. On the other hand, however, another important ideal for the social work profession is present. Social workers have to meet the targets of their organizations and managers that stem from NPM and other management's tools focus on documentation of work in scoring schemas. These two ideals create tension for the social work profession, as discussed in the article, because they conflict. It is difficult for social workers to fit assessments of their relational work with welfare recipients into the scoring schemas of organizations.
The organizational scoring schemas present social workers with other problems as well. For instance, when they are successful in recording the complexity of the welfare recipients' situation in the schemas, then the subjectivity of their assessments become less clear because scoring schemas hide or blur that social work by nature is subjective and based upon a personal relationship between a social worker and a welfare recipient. The apparent neutrality of the schemas and reports (Rogowski, 2010) hide and even downgrade this central characteristic of social work mentioned by all participating social workers. This apparent neutrality of scoring schemas may also challenge the power of welfare recipients and their ability to make their potential disagreement with the assessments of them applicable (because the subjectivity of social work is now hidden).
In addition, the categories of the scoring schemas influence what staff can jot down, and thus which activities are meaningful for social workers to suggest to welfare recipients. Hence, the chosen activities of social workers may not reflect their assessment of clients' needs but instead, reflect which activities can be fitted into the categories of the scoring schemas provided by the institutions. In sum, social workers critique of the scoring schemas often had to do with what social workers (still) think is their core competency, namely to build a good relation to welfare recipients. For instance, in their view, nurses can easily translate their professionalism into the sections of an organizational scoring schema, while social workers' professionalism (relational work) is difficult to fit into the small sections of the schemas. Thus, the analysis suggests that that the current focus on NPM-inspired tools in social work may even affect the internal hierarchy of professionals.
Nevertheless, the analysis also revealed that social workers welcomed the organizational scoring schemas. These positive stories show that many social workers thought that the lack of factual knowledge in social work made it difficult to have a distinct voice in their organizations. Documenting the work made social workers' job visible for managers (who may not work side-by-side with the staff) and colleagues from other professions when discussing work and reports in the regular staff meetings.
To conclude, this article shows how current welfare organizations balance different rationales and dilemmas leading to new perceptions of meaningful work and expertise (Mik-Meyer, 2017). Social workers disciplinary training stresses relational work with welfare recipients, but at the same time, they are held "accountable" according to organizational systems of documentation (Ferguson, 2009;Hupe & Hill, 2007). The notions of organizational professionalism (Evetts, 2009a), pragmatic professionalism (Liljegren, 2012), and NPM-professionalism (Evetts, 2009a) all emphasise that professionalism and expertise today is indeed an organizational phenomenon. The current increased use of scoring schemas in care and treatment facilities strongly influence the way social workers in a more general sense organize their work. NPM-inspired tools thus seem to have a profound influence on how expertise looks like today. Ironically, the expertise of social workers may perhaps primarily be about being successful in translating their relational work with the welfare recipient into organizational scoring schemas and only secondarily about meeting welfare recipients' needs.