Separate counselling services in Norwegian upper secondary schools. A possibility for a collective holistic approach?

Guidance counsellors hold an important role in youth’s educational and vocational choices, well-being and future careers. Central Norwegian policy documents recommend dividing counselling practices in schools into two functions: (1) career guidance and (2) educational welfare guidance. Policy documents also promote a holistic view of the pupil and the value of safeguarding the pupils’ best interest. This study explores how counsellors interpret these potentially conflicting recommendations for counselling practice. Our findings indicate that counsellors act as street-level bureaucrats and interpret the recommendations in relation to professional values and local contexts. Our analysis reveals three perspectives on how separate forms of counselling can be practised. The first view suggests an integrated, holistic approach. The second suggests a collective holistic approach. The third view indicates a reductionistic approach to career guidance. Counsellors appear to base their views on occupational professionalism, with the ideal of guarding the pupils’ best interest at its centre. Hence, the different understandings of separate counselling services, a holistic approach and of what is in the pupils’ best interest appear significant for implementing the policy recommendations in schools.


Introduction
In Norwegian schools, counsellors hold an important role in youth's educational and vocational choices and well-being. Because the counsellor's interpretations of policy form an important basis for guidance practice, the current study explores how counsellors interpret two central policy guidelines. The first guideline is an organisational International Journal for Educational and Vocational Guidance (2022) 22:557-576 recommendation on how to organise guidance in school, and the second is the regulation for professional codes of conduct amongst guidance counsellors.
The counselling services in Norwegian schools have two defined functions: (1) educational welfare guidance and (2) career guidance. The primary task for career guidance is to provide pupils with information and guidance regarding educational and vocational choices, whilst the primary task of educational welfare guidance is to help pupils experience social and academic success (FKunnskapsdepartementet, 2009). Furthermore, counsellors are expected to focus on both the system for all counselling activity at the school and guidance of individual pupils (Mordal et al., 2015). Most counsellors in the Norwegian school system hold part-time positions as counsellors and part-time positions as teachers . Since early 2000, policy has promoted separate counselling services (NOU, 2016, p. 7, p. 154) as the recommended form (Buland & Havn, 2003;OECD, 2002;St.Meld. nr.16, 2006-2007. Separate counselling services imply splitting career guidance and educational welfare guidance into separate functions that are attended to by different counsellors (Buland & Havn, 2003). Separate counselling services were initially promoted as a solution for safeguarding career guidance because stakeholders were concerned that this task was being neglected because of the expanding needs in the area of educational welfare guidance, which include pupils mental and social challenges (Buland & Havn, 2003). The Norwegian Ministry of Education (2009, Sect. 22-1) has stated, 'In order to preserve the pupils' best interests, there should be a holistic view of the pupil and the educational welfare guidance and career guidance should be integrated'. However, there is no specification of how counselling services in schools should be organised to ensure a holistic guidance practice. Currently, the variety in organising is vast; some schools have assigned both functions to the same person (integrated services), whilst others have split the functions across several persons (separate counselling services).
A 'holistic philosophy of counselling emphasises helping "whole" persons' (Betz & Corning, 1993, p. 137) and seeing career and personal counselling as inextricably intertwined (Krumboltz, 1993, p. 143). Thus, we see a holistic or whole-person approach as a guidance practice where the counsellor recognises the intertwinement of personal challenges and career decisions, treating them as a continuum rather than dichotomous concerns (Krumboltz, 1993;Super, 1993;Zunker, 2012). The holistic perspective is in accordance with Bronfenbrenner's (1977) classical concept of community collaboration, hence promoting a collective-oriented holistic approach. Kettunen and Makela (2019) found that ethical practice amongst others was conceived as stemming from a collaborative orientation. According to Low et al. (2013), collaboration and cooperation are essential to ensure the successful implementation of counselling services. Thomsen (2017) provide interesting ideas on career guidance as a collective practice in communities. Thus, recommendations that promote separating the two main tasks of counselling may create a conflict with regulatory and professional emphasis to achieve a holistic approach towards counselling. To be explicit, regulations provide guidelines on how the counsellors and schools should conduct counselling, whilst policy recommendations to separate the services address how the services should be organised. Nonetheless, how the work is organised has implications for its practice [e.g., Irgens (2010)]. In the gap between 1 3 policy and practice, the decisions made by practitioners influence how policy is realised (Hooley & Rice, 2019).
According to Watts (2008), most guidance services are the object of and a potential instrument for public policy; he claims that research within the field of career guidance pays little attention to public policy and implementation. Furthermore, Haug et al. (2020b) and Hooley (2019) address the difficulties of policy implementation in the career guidance field in Norway and how policy implementations' depend on professionals' faith and goodwill. This dependence could also be a consequence of policy having interpretative flexibility (Pinch & Bijker, 1987). Interpretative flexibility is the notion of how something is interpreted into local knowledge and how it is interpreted differently by different actors. This concept is appropriate for guidance counselling because counsellors need to understand policy in relation to their own local context (Frostenson, 2015;Irgens, 2010). This is supported by the Expert committee on the role of teachers (2016, p. 208), which is explicit on how professional communities have an important mediating function in giving meaningful content to external demands and expectations in development and change.
Prior to year 2000 Nordic research in guidance has focused mainly on the sociological aspects of guidance, its societal links and possible impact (Plant et al., 2003, p. 116). As a developing field of research, Haug et al. (2020a) finds a great variance of thematics in the specific context of Nordic guidance counselling. These include reflections on the Nordic model and the notion of career, gender equality and career guidance and professional development of career guidance practitioners, only to mention a few. However, no contributions of the anthology discuss how counsellors' interpretations and understandings affect the implementation of policy recommendations. The aim of the current study is to fill this gap in the literature.
We seek to expand the knowledge about what guides counsellors' professional practices by exploring how counsellors interpret and understand conflicting regulations and policy recommendations. Our approach to exploring this is based on three complementary theoretical contributions. We combine profession theory and the concepts of occupational and organisational professionalism (Evetts, 2009), boundary work (Liljegren, 2012) and hybrid professionalism (Noordegraaf, 2015), here with Lipsky's (1980) theory of street-level bureaucracy and Pinch and Bijker (1987) interpretative flexibility. Lipsky's (1980) contribution is relevant for understanding counsellors' attempts at managing the difficulties and ambiguities in conflicting recommendations. Interpretative flexibility addresses how policy intentions and content can be interpretatively flexible (Pinch & Bijker, 1987) when counsellors implement them into local school practices (Irgens, 2010). Insight into counsellors' interpretations and professional judgements may give us knowledge about their perspectives and reasoning for-or not-implementing policy. Such knowledge can be important for future policy making, implementation and development in professional guidance practice. Thus, our research question is as follows: How do counsellors interpret and understand the recommendation of separating career guidance and educational welfare counselling in schools?
The theoretical approach of the current study can contribute to novel knowledge of how counsellors in Norwegian schools understand the contradictions in policy. Exploring this can also provide additional knowledge of counsellors' general professional rationales for their practices. This will be further elaborated upon after a short introduction to the research on the Norwegian context of school counselling.

School counselling in a Norwegian context
The core values of Norwegian school counselling are egalitarianism and contain a moral and political commitment to the welfare state (Bakke, 2020;Haug et al., 2020b;Hooley, 2020). New public management (NPM) has a major impact on the world of work and, consequently, on the way career guidance is conceived, practised (Hooley et al., 2018, pp. 5-8;Hughes et al., 2015;Sultana, 2018). Accordingly also the development of education for career guidance professionals (Andreassen et al., 2019). The Norwegian school system is affected by NPM rationales, with a goal-oriented focus on efficiency, quality and the cost-effective delivery of public services (Hooley et al., 2018;Nilsen, 2021). As a result, counsellors in Norway have reported insufficient resources and that extensive goal orientation and standardised numbers of hours for pupils in each subject are obstacles for guidance activities; they have also stated that their work is rated as second in line and a 'time thief' (Mordal et al., 2015, p. 112). Counsellors understand themselves primarily as the pupils' supporter (Mordal et al., 2015, p. 116). According to Irving (2017, p. 57), there is a degree of flexibility in counsellors' freedom to review, reframe and reposition career/education discourse when organising their work. Also, as argued by Thomsen (2014), flexible career practices that adapt to the needs of participants have the potential effect of greater social justice. Indeed, the room for local variations in forming and developing the counsellor's role is considerable Mordal et al., 2015). Further, it is important to notice that understandings of competence have specific meanings in particular contexts (Sultana, 2009). Yet the role itself appears to be complex and stressful Mordal et al., 2015). This points towards a knowledge gap concerning counsellors' perspectives of their own professional practice because earlier research contributions in this field appear to have focused mostly on the outcome of guidance practices (Haug et al., 2018).

Guidance counsellors as professional street-level bureaucrats engaged in interpretative flexibility
To illuminate the guidance counsellors' professionalism within an NPM rationale, Evetts' (2009) ideal forms of professionalism is useful. Evetts (2009) argues that there is a development of two different and contrasting forms of professionalism. Organisational professionalism is manifested in a discourse of control that involves increasingly standardised work procedures and practices. At the other end of the spectrum, occupational professionalism is oriented towards authority and autonomy and emphasises relationships; here, organisational professionalism is more dependent on structures (Evetts, 2009, p. 248). An important difference between these logics is that occupational professionalism builds authority on trust in professionals' education and ethics, whilst organisational professionalism authority is grounded 1 3 in regulations and control (Liljegren, 2012). According to Abbott (1988), this distinction helps when we make sense of professional craft, which is understood as using abstract knowledge to solve specific issues. Practitioners' mandates and tasks are regulated by public policy and local conditions and can be seen as collective actors who manoeuvre within their specific contexts (Abbott, 1988;Liljegren, 2012). This process can be understood in terms of boundary work (Liljegren, 2012), where professions create, maintain and break down boundaries to separate 'us' from 'them' and claim professional turf. Uncertainty of boundaries and a weakening of autonomous spaces appear to attack professional work and harm professional values (Hermstad et al., 2020;Noordegraaf, 2015). Hence, Noordegraaf (2015) suggests the emergence of hybrid professionalism, which arises when professional and managerial principles-here concerning coordination of work, the establishment of authority and the values at stake-come together. Hence, hybrid professionalism occurs when professional and managerial boundaries become blurred, when professional logics are combined and reconfigured and when professionalism and managerialism are combined (Noordegraaf, 2015). Lipsky's (1980Lipsky's ( , 2010 concept of street-level bureaucracy can be helpful for understanding how counsellors manage conflicting or ambiguous policy guidelines. Lipsky's contribution provides a framework for understanding professionals' attempts to cope with difficulties and ambiguities. Street-level bureaucrats are 'public servants who work with residents daily and who must constantly make discretionary assessments' (Lipsky, 1980, p. 3). We understand counsellors as streetlevel bureaucrats because their priorities in daily counselling practice put policy into action. Street-level bureaucrats are 'public servants who work with residents daily and who must constantly make discretionary assessments' (Lipsky, 1980, p. 3). In addition, 'they cannot do the job according to ideal conceptions of the practice because of the limitations of work structure' (Lipsky, 2010, p. xvii). Street-level bureaucrats' actions, appraisals and priorities determine what services recipients receive (Lipsky, 1980). Furthermore, they are mediators between two worlds that are out of tune: the state and society (Zacka, 2017, p. 24). There is a growing body of policy-focused studies; however, they rarely indicate opposition or resistance to policy aims as street-level practitioners understand them (Brodkin, 2012, p. 944). Tummers and Bekkers (2014, pp. 540-541) find that street-level bureaucrats' discretion influences their willingness to implement policy in two ways. First, discretion allows them to tailor their decisions and practices to the needs of their clients. Second, discretion has a positive effect on bureaucrats' perceptions of client meaningfulness, which strongly influences their willingness to implement policy. The street-level bureaucracy theory is relevant for analysing counsellors' views as they participate in discussions and decision-making about how to organise the counselling services at their schools. In this context, counsellors are actors who translate policy into action; they decide how and in which form pupils receive guidance.
Hence, from their unique professional point of view, counsellors interpret and adjust their local practices to policy and ethical guidelines. Interpretative flexibility (Pinch & Bijker, 1987) is a notion used to describe how artefacts are interpreted differently by different actors into local knowledge. By this, Pinch and Bijker (1987) mean 'not only that there is a flexibility in how people think of or interpret artefacts but also that there is flexibility in how artefacts are designed '(p. 40). They further hold that 'a problem is defined as such only when there is a social group for which it constitutes a "problem"' (Pinch & Bijker, 1987, p. 30). Buland (1996) shows how interpretative flexibility can be used to understand how policy is interpreted differently by different actors when applying the concept to the analysis of the implementation of new technology in Norwegian public sector. Applying the concept of interpretative flexibility in the current study provides access for exploring counsellors' responses to policy and offers insights into how policy is interpreted differently in this field. In the counsellor context, policy needs to be operationalised within the schools where the counsellors work [e.g., Frostenson (2015), Irgens (2010)]. Thus, when policy recommendations are not clearly formulated, they have interpretative flexibility. Counsellors practice within different local contexts and, therefore, might have different understandings of both policy content and intentions. This is supported by Frostenson's (2015) discussion on how the local organisational context impacts professional autonomy. Also, Rosvall (2019) emphasises the importance of local conditions for pedagogic practice. In extension, Irgens (2010, pp. 133-134) emphasises how 'a good school is not created by individual work alone' and that 'the teachers at collectively oriented schools are more positive about collaboration than their colleagues at individually oriented schools'. Thus, policy's interpretative flexibility might result in different practices and understandings.
Counsellors are a part of the Norwegian school system and engage with pupils on a daily basis as street-level bureaucrats (Lipsky, 1980); they convert policy into action within institutions governed by neoliberalistic goal orientation. In some respects, they are in limbo between their organisational professionalism and occupational professionalism (Evetts, 2009). They use their professional knowledge to analyse each pupil's needs and devote specific meaning to their observations through their professional craft (Abbott, 1988). Furthermore, counsellors are largely left to their own assessments to balance between different expectations. However, both separate counselling services and a holistic approach are interpretatively flexible (Pinch & Bijker, 1987) concepts that counsellors must interpret into their local context (Irgens, 2010).

Participants
Thirty-five guidance practitioners representing 18 upper secondary schools in Norway participated in the interviews. The participants practised guidance for pupils between the ages of 16 and 19. Table 1 provides an overview of the participants' professional responsibility.
The 35 participants were from schools with both integrated and separate counselling services. Sixteen participants practised integrated services, 11 practised only career guidance, and four practised only educational welfare counselling. In addition, there were three counselling service leaders and one participant from the follow-up service 1 (see Table 1). Their experience as counsellors varied from 1 to 15 years. The data originate from two research projects: an evaluation project of quality in counselling services in three counties  and a project on the counsellor's role in schools (Mordal et al., 2015). The interview guides were almost identical. This means that the interviews were part of a wider research context where other thematics were also researched. Although, the data were rich in information when it came to the counsellors' understanding and interpretation of separate counselling services, the two original projects' time limits did not provide the possibility to explore this theme in depth. Research assistants transcribed the interviews. Not having transcribed the material ourselves, made the coding and analysis process more extensive. The research was conducted in accordance with research ethical principles for gathering and saving data and was approved by the Norwegian Centre for Research Data. The transcribed interviews were anonymised and stored for further research, for which the interviewees gave their informed consent.

Focus group
The data of the current study were collected through focus groups. A focus group method was chosen because participants' discussions could produce data that would not have been possible otherwise (Morgan, 2019, p. 5) and because of its capacity for exploring the logic of the phenomenon (Demant, 2006), that is, counsellors' reflections and understanding of school guidance. Data were generated through the participants' discussions and were determined and guided by the researchers (Morgan, 2019). We conducted focus groups according to a funnel-shaped strategy (Morgan, 2019, pp. 68-69) on nine predefined topics. Because of the small size of three schools, three of the interviews had only two participants. These three interviews, with two participants each, differ from the majority interviews by being more of a guided conversation rather than a discussion (Morgan, 2019). We conducted 10 focus group interviews and two individual interviews.

Analyses
We have analysed the data following an abductive approach (Timmermans & Tavory, 2012, p. 180), which is understood as a systematic process of meaning-making aimed at theoretical generalisations that nurture theoretical pluralism (Tavory & Timmermans, 2014, p. 123). Analyses were conducted by looking at the interaction between data and theory, and different categories have emerged from this interaction. The interviews were coded with a thematic analysis approach (Morgan, 2019), here with the intent of interpreting the meaning and creating codes that captured the discussed themes. We applied codes manually using NVivo to organise the data and obtain an overview.
During the interviews, reflections on separate versus integrated counselling services kept emerging. This caught our attention and made us interested in exploring the counsellors' views on this policy recommendation. The interviewees provided rather strong opinions on the subject, which gave us the notion that this was an important point from which the counsellors defined their professional practice. Further, whilst making their arguments-for or against separate counselling services-they frequently started discussing what was in the pupils' best interest in general. Further, the discussions of the pupils' best interests brought in the concept of a holistic view. As the findings in the next sections will show, a holistic view can both be related to a holistic view of the pupil and the holistic view of the counselling practice, implying that the counsellors have different points of departure for their arguments for or against separate counselling services. We will elaborate on this in the following section, where we present the findings.

Results
When asking counsellors how they felt about separate counselling services, we noticed that the views on how to organise the two functions of counselling services in Norwegian schools and the counsellor's basic values were closely intertwined. Three different lines of argument dominated their responses: (1) integrated counselling, a holistic approach; (2) separate counselling, a collective holistic approach; and (3) separate counselling, a reductionist approach. In the following, we will describe these three views, as discussed in the focus groups.

Integrated counselling, a holistic approach
The first stance comprised counsellors who were negative when it came to separate counselling services. They were concerned about safeguarding holistic values in counselling and expressed scepticism towards the separation of the services, regardless of how it was organised. One counsellor in a focus group said, 'Educational welfare, career guidance and future planning are all closely connected'. The rest of the focus group acknowledged her viewpoint. This view resembles a holistic approach. She explained, 'It's like this: when a pupil comes to us, then he comes with an easy question and a harmless problem, but it turns out to have a more complex background'. The counsellor further argued that when counsellors' functions are integrated as one person's responsibility, it is easier for the counsellor to safeguard the whole pupil. One counsellor stated that 'an individual is a whole after all, and then, it is both the educational welfare and [career guidance] […] everything is connected, you know'. This indicates flexibility as a prerequisite for the services. It also appears to highlight the importance of addressing concerns about how separating services could affect the accessibility of both career guidance and social welfare guidance to pupils. The counsellors who supported this stance appeared to assume that separate counselling services would equal little collaboration between the functions. Furthermore, they seemed to believe that integrated services were the only way to safeguard the pupils' best interests and a holistic approach. One counsellor explained, 'We follow the pupils. The whole pupil. We think it is an unnatural choice to separate educational welfare and career guidance'. Their assumption of separate counselling as a limitation regarding seeing the whole pupil led to their professional consideration of separate counselling services as a poor way to organise the services. Thus, this stance appears to gather a holistic view of the pupil, a holistic approach to counselling services and safeguarding the pupils' best interests in general, as integrated, intertwined and interdependent parts of counselling as professional practice.

Separate counselling, a collective holistic approach
The second line of arguments was positive towards separate counselling with extended collaboration found in teams of counsellors. The different parts of the team would safeguard different functions, thus securing a holistic approach towards the pupils and counselling. In one focus group, the following statement about experience with separate counselling was presented and gained consensus: It works well. It is effective. Because we use each other, [...] as a 'happy street', right? Including all pupils' services, counsellors, practical pedagogical services, the follow-up services, the health nurse and so on. We all sit over here together, and then, we can use each other so much more.
[…] It is much more effective when we can use our competence on the pupil instead of having to sit with everything ourselves.
In her case, all the student services were colocated, which made it easier to collaborate and offer different expertise to the pupils. Collaboration was essential in the provision of counselling services to pupils. In another focus group, a counsellor argued this in the following way: 'I believe that we can work with the same person in different ways.
[…] And then we balance it'. It appeared to be important to have a balance between the functions, thus safeguarding all aspects of a pupil's life through teamwork. This stance views separate counselling as positive yet with a holistic approach to counselling through extensive cooperation. The counsellors clarified this view in a focus group: 'There can't be watertight bulkheads because, […] it must be a collaboration. The fact that both [functions] may well join in when we have summoned a meeting and then get the overall picture [is crucial]'. These counsellors were concerned about how to collaborate to ensure a holistic approach; they presupposed that separate counselling denoted collaboration between functions.
At another school, they explained how they work: 'We have our clear areas, but it is the sharing culture that is so important to be able to be together in holistic thinking'. Thus, these counsellors valued the need to see the whole pupil and were positive towards separate counselling. Yet this approach depended on collaboration and could lead to performing tasks not formally assigned to their function. In one focus group, a counsellor told us, 'Pupil issues are often a volunteer job. Everyone must contribute. Often, you may need to do a thing that may not be within your job description. That's because it's no doubt who's in focus: it is the pupil'. She valued separate services because 'with separate counselling services, you get the opportunity to specialise in your field'. According to this, separate services can contribute to counteracting resource scarcity and creating room for personal development through interpretative flexibility. This suggests that separate counselling services with extensive collaboration can safeguard a holistic approach to counselling and pupils.

Separate counselling, a reductionistic approach
The third stance suggests that a separation of the services could shield career guidance from the extensive need for educational welfare guidance. This was one argument for separate counselling services in NOU 2016:7. Limited resources appeared to cause frustration amongst the counsellors, and separate counselling services were suggested as a possible solution. One counsellor addressed this in a focus group: His statement indicates an overwhelming workload within educational welfare guidance, and he was supported in the group discussion. Some informants expressed scepticism towards a holistic approach. This group argued that a counsellor does not need to know everything about a person to provide guidance. One career counsellor stated the following during a focus group: It was good for them to come here just to talk about educational choice, not to talk about 'how do you feel inside now' […] that part was put away, so we only talked about their resources for further education.
The interviewee valued the opportunity to represent a place where the students could talk about their future in a positive setting without having to address other difficulties in their lives, which was understood as of little importance when it came to career guidance. She argued that a holistic approach should be voluntary and that a reductionistic approach is sometimes in the pupils' best interest. Further, this argumentation promoted separating counselling services on the grounds of a need for 1 3 professionalising the services. These counsellors were concerned about developing the skills and practices of counsellors. In a focus group, they argued, 'We talk of professionalisation, you know.
[…] If we are going to get better in our functions, then of course we must have a separation. We cannot be equally good in both areas'. The reasoning was that the separation makes room for focus and professional development. Furthermore, as one participant stated, 'You get more time for career guidance, you know… educational welfare guidance takes all the time and is very unpredictable'. This can also be connected to counsellors safeguarding their own work environments and mechanisms for handling extensive workloads.

Interpretative connections
Our findings indicate that negative or positive attitudes towards separate counselling can be explained by the combination of interpretations regarding the implications of separate counselling services for a holistic approach. Table 2 depicts the contextualised analysis of the holistic versus reductionistic approach to guidance in the context of separate or integrated services.
As seen in Table 2, in the first combination, the counsellors were negative towards separate counselling services and appeared to assume that this organisational form equals no collaboration between the various functions. This stance viewed integrated services as the only way to safeguard a holistic approach in the pupils' best interests. The second combination represents counsellors who were positive towards separate counselling but with extensive collaboration. They believed that several counsellors could collectively safeguard a holistic approach. The third combination was counsellors who promoted separate counselling services, arguing that a reductionistic approach with room for focusing solely on career guidance was in the best interest of pupils.

Discussion
In the following, we discuss our findings in the light of professionalism, interpretative flexibility and street-level bureaucracy. The analysis shows that a holistic approach (Betz & Corning, 1993;Krumboltz, 1993;Zunker, 2012) has several nuances to it, and we argue that we can see the outlines of two slightly diverse understandings because of the interpretative flexibility (Pinch & Bijker, 1987) of the concept. Thus, the implementation and understanding of holistic approaches in counselling appear to depend on the local context. The third understanding of best practices in counselling resembles a reductionist approach.

Holistic approach
The emphasis on a holistic view appeared in many of our interviews. However, the counsellors did not define this concept, other than stating that it is a central aspect of good professional practice. Yet the holistic approach appears to represent a holistic Counselling approaches Holistic approach It is in the pupil's best interest that one counsellor sees the whole of the pupil's life. Therefore, we need integrated counselling services Collective holistic approach It is in the pupil's best interest that an extended team sees the whole pupil. Therefore, we need separate counselling services with extensive collaboration

Reductionistic approach
It is in the pupil's best interest to have a free space where they can talk of future career choices disconnected from all the bad things in life and to ensure that the resources for career guidance are not absorbed by other tasks. Therefore, we need to have separate services with only a minimum of collaboration view of the pupil. This is similar to how the holistic approach is defined in the literature as career and personal counselling being inextricably intertwined (Krumboltz, 1993) and not easily separated on different persons. Thus, the holistic approach constitutes a pupil-centred holistic approach, where the counsellors aim to take all aspects of the pupils' lives into account during counselling. This implies that each counsellor's professional turf includes both educational welfare and career guidance.
The collective holistic approach also emphasises a holistic view, yet it seems to orient more towards a holistic view of counselling. A collective approach to counselling can safeguard both career and educational welfare through teamwork. This interpretation facilitates collective counselling services where several counsellors implement holistic counselling through extensive collaboration. The implication of this interpretation is that it allows counsellors to specialise in specific areas whilst still delivering holistic guidance to pupils. One might see it as a combination of a holistic approach (Krumboltz, 1993) and a Bronfenbrenner-inspired (1977) collective holistic approach. The counsellors acknowledged that guidance should take into account the whole pupil's life, and this is solved through collective collaborative efforts. Zunker (2012) uses the term holistic approach when discussing counselling, but as we understand his accounts of it, it is still a 'people-focused' approach. What our analyses reveal is a holistic approach to the whole of the counselling practised in school, not only the counselling of 'whole' pupils. This finding relates to what Low et al. (2013) describe in their context of counselling in schools in Malaysia as the importance of collaboration between school counsellors.
The last and reductionistic approach to career guidance seems to see potential personal and health issues amongst pupils as outside their responsibility. The argument here is that there is no need to know everything about a person to give good career guidance. This appears to be a reductionist approach to counselling and contrary to what Betz and Corning (1993), Krumboltz (1993) and Zunker (2012) define as a holistic approach integrating personal and career guidance. However, this might be more related to resources and NPM than actual differences in professional values. Viewing guidance counsellors as street-level bureaucrats (Lipsky, 2010) who are managing their work situation, as described in Mathiesen et al. (2014), can explain the reductionist approach. In circumstances with scarce resources, the counsellors seem to respond to the requirement of a holistic view by reducing or confining their definition of a holistic approach to career guidance alone. Thus, the reductionist approach may be a strategy for coping with scarce resources and high demands. Yet it might raise concerns about whether the pupils in schools with reductionist approaches toward counselling are receiving guidance in accordance with their legislative rights.

Separate counselling services
The three understandings of separate counselling services described above can be described as the results of interpretative flexibility (Pinch & Bijker, 1987). This flexibility is visible in how the counsellors, as street-level bureaucrats (Lipsky, 2010), reflected on the practical consequences of the separation. Separate organising of guidance services is interpreted as a threat to occupational professionalism amongst the counsellors who emphasised the value of a holistic approach to the pupil (Evetts, 2009;Liljegren, 2012). Separate services are viewed to be in conflict with professional codes of ethics and in the pupils' best interest (Liljegren, 2012). Thus, these counsellors defined separate counselling services as a problem because they are an organisational measure that can potentially work against their occupational professionalism. Yet their responses can also be viewed in terms of boundary work, here suggesting that 'pupil-oriented' counsellors are using arguments for a holistic approach to claim their professional turf (Liljegren, 2012). However, the scepticism towards separating the services can also be related to resistance to increased organisational professionalism (Evetts, 2009) and a perceived threat to the counsellors' occupational professionalism (Evetts, 2009), accordingly, their authority and autonomy in professional practice. Further, the recommendation to separate can be seen as a further development towards an individualistic orientation in schools (Irgens, 2010).
Further, the counsellors appeared to assume interpretative flexibility in their understanding of how separate counselling services are 'designed' (Pinch & Bijker, 1987). Our findings indicate two interpretations of the designs. One is separate counselling services with collaboration between educational welfare services and career guidance. This design can be related to occupational professionalism, which is founded on a collective holistic approach; this approach seems to be related to a collective professional competence and a common code of ethics that defines counsellors' professional turf and college in schools. The sense of 'us' in the 'collective mode' can contribute to safeguarding the pupil's best interests overall. Further, this mode could contribute to integrating professional (holistic approach) and managerial principles (separate counselling services) in what can resemble a form of hybrid professionalism (Noordegraaf, 2015). Such hybrid professionalism depends on collective-oriented schools (Irgens, 2010). Thus, the local conditions at each particular school will most likely have an impact on the counsellors' understanding, interpretations and practice (Liljegren, 2012). Hence, a collective-oriented school might contribute to a fruitful merge between policy measures with the aim of occupational professionalism and policy regarding organisational professionalism, creating a professional practice founded on a holistic view of counselling services as holistic practice.
Second, there is the separation designed with 'watertight bulkheads' and minimal contact between counsellors with different areas of responsibility. In terms of professional turfs and boundary work, we see the contours of a struggle taking place between career guidance and educational welfare. There appears to be a marked 'us' and 'them' between counsellors regarding these tasks rather than between the counselling services and the rest of the school. This seems to be conditioned by the organisational environment and the collaborative environment at the local school (Irgens, 2010). Zacka (2017, p. 241) points out that 'whilst public service agencies actively depend on the moral agency of street-level bureaucrats, they place these bureaucrats in working conditions that tend to undermine that very agency'. Zacka emphasises that we need to 'evolve conditions where it is possible for them to fulfil their role adequately' (p. 241).
Additionally, the resource situation in counselling services may affect the counsellors' interpretations. Their interpretations of and responses to policy recommendations may be coloured by scarce resources and the need to prioritise between the two equally important and demanding tasks of career counselling and educational welfare; this also comes into play when policy is to be implemented in the local school context.
The counsellors struggle with limited resources; they have a mean of 2.4 min per pupil per week in total at their disposal . Also, the narrative of educational welfare guidance with the ever-expanding needs related to pupils' mental and social challenges might cause conflicts between two equally important functions of counselling, which both have a part in a holistic approach to career guidance. This could eventually reduce the possibilities for synergy effects between the two functions of Norwegian school counselling, thus providing poorer services for the pupils because scarce resources force the street-level bureaucrats to prioritise their professional conduct.

The pupil's best interest
Although the outcome of the counsellors' interpretive processes are quite different, the ethical value of safeguarding the pupils' best interests appears as a common denominator. Hence, our analysis indicates that counsellors are dedicated to putting the pupils' needs at the core of their practices, despite differences in views on organising and scarce resources. Perhaps, this is what defines the core of occupational professionalism amongst Norwegian school counsellors: concern for the pupils' best interests. This would imply a common agreement of safeguarding the pupil's best interest as a moral compass for counsellors' professional conduct and in accordance with policy guidelines. Tummers and Bekkers (2014) find that bureaucrat's perceptions of client meaningfulness strongly influenced their willingness to implement policy. Our study shows how the opposite of what Tummers and Bekkers (2014) have found: when counsellors interpret the recommended policy as a measure that will not be meaningful to their clients, they are less willing to implement it. Those counsellors who did not see separate counselling services as an organisational measure that is in the pupils' best interest or did not see how it could ensure a holistic approach were less willing to implement it. By resisting policy recommendations, they were attempting to safeguard the pupils' rights. Hence, the implication for policy intervention in school counselling is an orientation towards local understandings, here seeking to implement policy in ways that are recognised at the local schools as in the pupils' best interests. This resembles what the expert committee on the teachers role (2016, p. 208) recommends.

Conclusion
In the current study, we have explored how counsellors interpret and understand the recommendations for separating counselling services in schools. Our findings reveal three different approaches to these recommendations. The approaches appear as different outcomes of counsellors in encounters with interpretative flexible policy recommendations. The counsellors utilise their positions as street-level bureaucrats in their processes of interpretation, responding to the recommendation based on their judgement of what is in their pupils' best interest.
The present study was conducted in Norwegian schools. Hence, the specific characteristics of the Nordic context for guidance counselling may contain some limitations regarding the transferability of our findings. Further, participation was voluntary, and data were collected at a single point in time. Separate counselling services were only one part of the themes discussed in the focus group interviews. Thus, the current study might not capture every critical aspect of counsellors' interpretations of separate counselling services.
Areas for further research are to further explore the concept of interpretative flexibility in relation to policy implementations and street-level bureaucracy; this could provide insights into how to ensure the effective implementation of policy recommendations. Another interesting topic for research is whether individual-or collective-oriented schools have significance for practice development and policy implementation. Implications for practice are that school leaders, counsellors and policy makers need to recognise the ambiguities between policy and ethical values in guidance counselling and consider how local characteristics such as collectiveness or individualisation in schools might affect interpretation and responses to new policy interventions. Finally, we suggest that the findings of the present study may be relevant for the current process of implementing a new quality framework (Haug et al., 2019) for career guidance in Norwegian schools. This is because a successful implementation process would need to consider the interpretative flexibility of policy, counsellors' professional and ethical values and the local school contexts for implementation, along with how these elements affect how politics become a practice in each school.