Professional responses to exogenous change: the social work profession and the jurisdictional domain opened up by the Norwegian welfare-to-work reform

ABSTRACT This paper examines the social work profession’s responses to the organizational reform of the Norwegian Labour and Welfare Administration, which opened a new jurisdictional domain of employment services to groups marginalized in the labour market. This domain was not colonized by any established professions, and social workers were the only occupation in the reformed organization with the character and identity of a profession. With existing research as the main source, the paper shows that despite having influenced the reform policy, during reform implementation, agents of the profession refrained from expanding the jurisdiction of social workers and instead protected and maintained its established jurisdiction based on municipal social services. Potential explanations behind the response are discussed: the pressure from the organizational context, combined with the profession’s ingrained distrust of ‘the system’; the ‘dirtiness’ of welfare-to-work tasks, around which some scepticism existed; and the profession’s strong value base and lack of capacity to articulate its knowledge base and, subsequently, theorize its knowledge and skills as resources that were key to meeting the reform’s welfare-to-work goals. All three explanations have some explanatory power but must still be considered tentative and in need of more research.


Introduction
Whether the implementation of welfare-to-work constitutes professional work or deprofessionalized administrative rule-following is contested (Nothdurfter and Olesen 2017;van Berkel and van der Aa 2012).Also contested is the role of the social work profession in implementing welfare-to-work policies where support is often combined with conditionality and sanctions towards disadvantaged unemployed people considered too far from the labour market (Caswell and Larsen 2017;Hasenfeld 1999;Kjørstad 2005).However, it is also proposed that social workers are particularly fit for managing activation work because social work is developed out of the dilemma of being linked to policy but also striving for autonomy to meet the needs of the target groups (Nothdurfter 2016).
In Norway, social workers have traditionally been the dominant occupation in social services to clients that now are encompassed by welfare-to-work policies.In Norway, like the other Scandinavian countries, welfare-to-work policies have primarily had an enabling profile and CONTACT Tone Alm Andreassen toaa@oslomet.noorientation towards enhancing the human capital of marginalized citizens.In 2005, the organizational reform establishing the Labour and Welfare Administration (NAV) introduced a new organizational context for social work.The reform aimed to provide integrated services to groups at the margins of the labour market, increase employment rates, and reduce benefit recipience.
The reform proposal was closely connected to social work ideologies of improving the life situation and employability of vulnerable, jobless clients by providing coordinated, personalized services to those with complex problems (Erlien 2017).The profession's association had succeeded in influencing the reform design to also include vulnerable citizens not ready for the labour market (Messel 2013).
The reform opened a potential new jurisdictional domain of employment assistance not only to social assistance recipients but also the much larger group of citizens entitled to social security allowances.In the reformed organization, social workers were the only occupation with the character and identity of a profession -that is, a specialized knowledge base from higher education, connection to a specific jurisdiction (social services), discretionary power, and a code of ethics.
Based on the literature on professions, one possible response of the social work profession to the reform would be a professional project to expand its jurisdiction (Abbott 1988) and define this new practice area by the use of their expertise and knowledge base (Suddaby and Viale 2011).In this case, the profession could have done so without fearing interprofessional conflict or counteraction from other professions.Another possible response would have been to protect and maintain established jurisdictions when confronted with bureaucratic organization and management placing pressure on the professionals' position and autonomy (Evetts 2003(Evetts , 2011;;Freidson 2001).
The possibility for diverging responses is the point of departure for our exploration of the following question: How did the social work profession respond in the face of the exogenous change introduced by the reform?Furthermore, we discuss the more difficult matter of potential reasons behind the profession's response.We base our exploration and discussion primarily on scientific literature on the social work profession in the context of this major organizational reform.We focus on the period of reform creation and implementation because this is the central period where a profession can position itself as decisive for the success of the reform.While we cannot offer definitive conclusions, we believe our exploration contributes to improving the understanding of professions' responses to exogenous change in general and of the social work profession in particular.

Perspectives on professions and exogenous change
A key theme in the study of professions is how they pursue professional projects and seek to expand control over jurisdictions of work, often in interprofessional disputes with adjacent professions (Abbott 1988).Professional projects take place in workplaces as well as in the political and legal systems, where claims for jurisdiction are most often made by the professionals' association (Abbott 1988).A key characteristic of professions is their capacity for 'theorising', meaning 'the development and specification of abstract categories, and the elaboration of chains of cause and effect' (Greenwood, Suddaby, and Hinings 2002, 60), which can later diffuse and result in the development of new practices across organizations.Accordingly, professional claims for jurisdiction are based on the possession of a body of abstract knowledge.
Exogenous changes, such as public sector reforms, may affect established jurisdictional boundaries, open new spaces for professional projects, or force professions to respond to disturbances in their jurisdictional control.
The literature highlights the potential for an expanding response, in which professions can be powerful change agents, operating as 'lords of the dance' (Scott 2008), given their capacity to influence organizations.They can do so by articulating and justifying the need and directions for change, creating or opening up new spaces for their expertise, and populating existing social spaces with new actors (Greenwood, Suddaby, and Hinings 2002;Muzio, Brock, and Suddaby 2013;Suddaby and Viale 2011).Such influence may be found prominently among high-status professions, such as medical doctors, but may also often be found among lower-status professions and organizationally based professions, such as museum curators (DiMaggio (1991), occupational health and safety managers (Daudigeos (2013), personnel staff (Dobbin 2009), and nurse practitioners (Reay, Golden-Biddle, and Germann 2006).Efforts to expand may nonetheless be hindered by interprofessional competition, status, and power hierarchies.For example, when faced with the expansion of new nursing and medical roles, medical doctors responded by delegating routine tasks and co-opting other powerful professional groups (Currie et al. 2012).Similarly, a professional project of paramedics was curbed as they became caught between management dynamics and a dominant elite of medical professionals (McCann et al. 2013).
Whether a profession will seek to expand its jurisdiction might be affected by the tasks being included through expansion, specifically whether they are considered respectable and prestigegiving.The concept of 'dirty work' (Hughes 1984) refers to elements of an occupation's bundle of tasks that are physically, socially, or morally dirty (Ashforth and Kreiner 1999;Hughes 1984).Dirty tasks are disgusting (such as the filth and dirt social workers encounter on home visits (Muzicant and Peled 2018)), degrading (for example, contact with stigmatized clients, as social workers have), or activities the workers are morally ashamed of (for instance, being intrusive or confrontational).Hughes argued that all occupations have tasks that they perceive as dirty work, which, as a stage in their upward occupational mobility, they may seek to drop to other occupations.Against that background, since studies have shown that a policy's value affects whether it is willingly implemented by professionals (Tummers 2011;Tummers, Bekkers, and Steijn 2009), one might expect that a profession will avoid expanding its jurisdiction if expansion involves 'dirty tasks'.
At the other end of the spectrum of responses lies a protective response, seeing exogenous change as a threat.This may be triggered, for instance, if bureaucratic organizations and management place pressures on professionals' position and autonomy (Evetts 2003(Evetts , 2011;;Freidson 2001).New Public Management reforms are often seen as impulses behind such pressure, as well as criticism of professionalization as a self-serving project that advances the interests of professional workers, instead of clients and society at large (Noordegraaf and Steijn 2014).From this perspective, professions can be victims of demands for regulation and increased bureaucracy, transparency, and accountability (Muzio and Kirkpatrick 2011;Noordegraaf 2011), or they can hold the power to resist organizational or managerial requirements (Ackroyd, Kirkpatrick, and Walker 2007;Hendrikx and van Gestel 2017).To do so, professionals may apply defensive or conservative strategies, seeking to protect their jurisdictions and privileges and maintain stability in the system.Classic professions such as medicine have discretely deployed various strategies to maintain their control (Evetts 2002;Numerato, Salvatore, and Fattore 2012), although members of high-status professions, such as family doctors (Reay et al. 2017), can also alter their role identity despite initially being reluctant to change.Furthermore, contrary to the assumption that management values and priorities colonize professional work, professional values have been found to remain robust in the wake of management reforms in health and social care (Ackroyd, Kirkpatrick, and Walker 2007;Evans 2010).
Since status differences exist between groups of professionals, powerful groups may be in a better position to withstand pressure.In particular, professions with strong occupational boundaries, underscored by a strong technical knowledge base, are seen to be well-positioned to challenge managerial control, while professions with weak occupational boundaries and limited resources in a scientific base, such as social workers, are seen to have fewer opportunities to resist organizational or managerial reforms (Healy 2009).Accordingly, social workers are depicted by many as subjected to pressures from managerialist reforms (Caswell and Larsen 2017;Evans and Harris 2004;Harlow et al. 2013;Healy 2009;Healy and Meagher 2004).
In sum, when responding to the welfare-to-work reform, one could expect the social work profession to either grasp the opportunity to expand its jurisdiction or alternatively seek to protect and maintain its established jurisdiction.

Analytical approach
This paper is a result of a long-standing research interest in and a larger project on the topic of professionalization of welfare-to-work services aiming to synthesize the vast literature produced about the organizational reform and the role of the social work professionalism in the Norwegian Labour and Welfare Service (Breit et al. 2018(Breit et al. , 2022)).Following this investigation, we became interested in the response of the social work profession to the new domain opened by the reform.The apparent lack of efforts to grasp the opportunity to expand was a 'mystery' (Alvesson and Kärreman 2007) to us and became the research puzzle we wanted to explore.
We pursued the research puzzle by tracing the agents of the profession.In line with Freidson (1985), we distinguished between the profession as its formal association negotiating the profession's interests; its elite members who establish, advance, and communicate the body of knowledge and skills claimed by the profession (university teachers and researchers); and the individual rank and file practitioners (professionals within the frontline organizations).In Norway, where associations of professionals, in general, do not possess authority over matters concerning their professions, the responsibility for supervision and professional development belongs to national authorities (direktorat, helsetilsynet) and their regional branches (statsforvalteren, previously fylkesmannen), which, therefore, may also act as agents of the profession.
Our approach was inspired by ideas of an interpretive synthesis since we aimed to build a comprehensive understanding of a complex phenomenon from a large and diverse set of literature (Depraetere et al. 2021;Dixon-Woods et al. 2006).We wanted to explore the social work profession's responses to the exogenous change caused by the reform.
Our analysis is not a systematic literature review aimed at giving an overview of the literature about social workers in the context of the reform.Rather our aim was to search for published literature that could help us understand social workers' reform responses and pose assumptions about possible reasons behind those responses.
The literature was included purposefully based on its likely relevance to our research interest.With the 67 texts from the larger research project (Breit et al. 2022) and our own knowledge about the research context as a point of departure, we focused on those texts that particularly addressed social workers perspectives on, experiences with, and responses to the reform.Continuously throughout the exploration, we have added other literature that has come to our knowledge from discussions with research colleagues or from being referred to in the literature already explored.
Our interest was responses during the period of reform development and implementation when the new organizations and work forms were unsettled and the jurisdictional domain was still open, which might be considered to last some years after the opening of the last frontline organization in 2011.Therefore, we have not included research that describes recent changes in the profession's responses.Studies on social work with clients were also outside the scope of this analysis.
There are limitations with using scientific literature as the basis for our exploration, most importantly that we are left with the topics and findings covered by the literature.Much research was qualitative case studies and do not answer questions about the distribution or representativeness of their findings.Only a few studies directly addressed the question of why social workers responded as they did, which is why the possible explanations we discuss are only tentative.Furthermore, the frontline workers were subject to considerably more research than the association of social workers (FO), university teachers and researchers, and national authorities.To compensate this limitation, we have drawn on textbooks, pamphlets, consultative statements, and media reports, while also being aware that these sources cannot be given the same status as the scientific sources.An overview of the sources is provided in the online Appendix.
Despite these limitations there are unique benefits with using existing research as sources.They offer access to observations and findings covering a relatively long period, including historic analyses, The value of the scientific sources is also that they cover studies from researchers with different research questions and perspectives than our own, who we consider substitute observers of the process.Hence, in total they give a much richer picture than would have been the outcome of a single study performed by us.
We aimed at a lines-of-argument synthesis, building a general interpretation grounded in the findings of the separate publications and documents, thereby unifying disparate aspects from the literature in an explanatory way.Findings were extracted with an orientation to the material, seeking to place the literature within its context.
Our 'litmus test' of having captured the process sufficiently was if the literature allowed us to develop a reasonably consistent detailed story as a preliminary step for the discussion about possible explanations for the responses (Langley 1999).Seeking the best explanations we could offer for the responses meant looking for several possible explanations, combined with contrasting reasoning: why this explanation and not another?(Harley and Cornelissen 2022).The theoretical perspectives provided contrasting explanations and supported us to identify the most likely explanations: a) the pressure of the organizational context, b) the perceived character of the tasks involved (seen to be prestigious or 'dirty'), and c) the knowledge base and capacity of the profession to theorize its knowledge and skills in relation to the reform goals.We believe that the analytical approach allowed us to form reasonable explanations to inform understanding of professions' response to organizational reforms.Yet, the explanations are not to be regarded as conclusive as they are tentative and must be further refined in future analyses.

The organizational reform and the responses of the social work profession
The social work profession in Norway involves credentials from formal higher education, a code of ethics, a professional association, and a professional identity.However, the profession has not been guaranteed occupational closure in the form of state-sanctioned certification or authorization for specific jobs.Instead, its position was connected to the institutionalized infrastructure around social services (Andreassen 2015;Røysum 2017): the Social Service Act ascribing the legal responsibility for social services to the municipalities; an occupational dominance of social workers in municipal social services, with managers mainly recruited from the profession; and professional supervision and support from regional and national authorities, including a series of development projects.
The profession arose to a certain extent from the development of and expanding need for qualified workers to staff the social services, along with the development of a more academic social work education (Erlien 2016;Messel 2014).Following the 1964 Social Care Act, social service administrations were established in the municipalities, with financial support from the state conditioned on those being staffed with educated social workers (Messel 2014).
The welfare-to-work reform disrupted this institutionalized structure.From 2006 to 2011, new frontline organizations were established based on a mandatory partnership that integrated social services and social assistance (the responsibility of local governments) with employment services and social security allowances (the responsibility of the national authorities).
In this context, expanding the jurisdiction of social workers would mean expanding their client group to include not only social assistance recipients but also recipients of social security allowances.Such an expansion would have required social workers to assert that their knowledge and skills were useful to broader target groups and to theorize what employment services for marginalized clients needed to entail.
Initially, social workers saw the reform as an opening for and valuation of social work (Røysum 2009b).The association viewed it as necessary and overdue, and asserted (in its consultative statements to the reform proposal) that all clients must be guaranteed a social work educated case worker securing them coordinated assistance.Social workers believed that service integration would provide more coordinated services and offer access to activation programmes to social assistance clients, groups which the previous employment service considered too far from the labour market (Schafft & Spjelkavik, 2006).Social workers with development positions in the first implementation period voiced a belief that the reform provided space for social work (Glemmestad 2011;Halås 2010;Øyehaugen 2006;Stolanowski 2009), but also fear that the profession's jurisdictional base will be lost (Røysum 2009a).
During the implementation period, key agents of the profession, however, held on to the established jurisdiction and a perception of social work as a specialized municipal service targeted at social assistance recipients.
The national social service authorities, in 2006, concomitant with the reform's implementation, launched a research and development programme involving municipalities' social services, social service clients, and universities and university colleges providing Bachelor's degree in social work (Johannessen, Natland, and Støkken 2011).Its inattention to the reorganization and the wider target groups of the new frontline organizations led the programme to be built on the pre-reform institutionalized social work model, rather than to fit the reform context (Andreassen 2015).
Concerning university teachers, textbooks from the early reform period seemed implicitly based on a perception of 'the social service within NAV' (our emphasis) (Dalen, Ranger, and Rytter 2014), as if the former social service agency still existed although within a different organizational context (NAV), and as if the only positions within the reformed organization of relevance to social workers were those associated with the social service legislation.Furthermore, a commissioned report on the reformed organization' needs for professional knowledge and skills complained that the totality of the Labour and Welfare Administration was invisible in the social work curricula, which paid attention only to municipal social services (Rambøll 2010).According to the report, some within the administration reported an antipathy towards working in NAV from both social work students and teachers.
The association agreed with the report (in its consultative statement) that stereotypical images of the administration and the educations existed on both sides but emphasized differences between education programmes.Still, the association linked social work primarily to social problems and those who cannot find a place in the ordinary labour market although it emphasized that the holistic approach of social workers was relevant for recipients of social security allowances too.Not until 2019 did the association distribute a booklet explicitly stating that social workers could operate within both state and municipal areas of responsibility, in doing so indicating an expansion of social work into the employment service domain.
Frontline social workers expressed and maintained a collective 'we' by referring to the ethical principles of social work and to their worries about their social assistance clients (Røysum 2017).They apparently reserved social work to social assistance clients, leading to frustration among frontline managers (Andreassen 2011).For example, a frontline manager -herself an educated social worker -said: Working with people is the core of social work, but now that means working with social assistance recipients . . .Social workers believe that they have municipal tasks, which is to take care that social assistance recipients receive what they are entitled to.(quote from Andreassen 2011) Frontline social workers found that the professional basis for their work was under pressure.They had to work in more standardized ways and spend more time on bureaucratic documentation and 'paperwork', which was coming to form the basis of administrative case processing (Røysum 2010(Røysum , 2013;;N. Skjefstad 2013).Although survey data indicated that the management model allowed autonomy in choosing means and measures in various fields, and that collegial support played a significant role in providing professional standards for decision-making (Jessen 2015), disapproval and disbelief seemed widespread among social workers, who considered the organizational conditions incompatible with social work professionalism (Rambøll 2010;Røysum 2009b).Some left the organization; others tried to embrace their 'municipal' tasks and clients, emphasized that if municipal social workers spent too much time on 'state' tasks, social services would be made to suffer, and found support in a national audit in 2010 criticizing the municipalities' administration of social assistance (Andreassen 2011).
The frontline workers' response may be interpreted as an effort to protect social workers from being pressured to fall in line with bureaucratic accountability and resource constraints, as mandated by the state administration.Social workers were exposed to new forms of organizational control from a state administration characterized by hierarchical accountability, performance management, demands for equal treatment, digitized case processing, new work tools, and standards for encounters with clients (Christensen and Laegreid 2012;Fimreite and Laegreid 2009;Jantz, Christensen, and Laegreid 2015), and with no tradition of basing its work on professional discretion and autonomy (Berg, Heen, and Hovde 2003).
Furthermore, reorganization of the benefits administration in 2008, concomitant with the financial crisis, caused a considerable increase in the benefits caseload and case-processing duration (Andreassen, Legard, and Lie 2011).Attention, resources, and social workers were absorbed in caseprocessing of social insurance benefits, rather than the reform goal of integrated welfare-to-work services (Andreassen and Fossestøl 2011).Additionally, the central administration depicted the frontline organizations as single-purpose organizations with a standardized management model and task portfolio oriented towards employment, without services for vulnerable, less 'job-ready' clients with major social problems (K.Fossestøl et al. 2015).Apparently, clients and services that social workers considered essential had no place in the reformed frontline organizations.
The association and its members were worried that practicing social work professionalism was difficult in the new frontline organizations (Røysum 2013;N. S. Skjefstad, Kiik, and Sandoval 2018).Their complaints concerned a lack of time available for comprehensive follow-up of clients, increased standardization, and less professional autonomy.At stake was the holistic orientation of the social work profession, which was threatened by the logic of bureaucracy, as Caswell and Innjord (2011) argued (in works republished in an anthology by the association).
This image of frontline NAV organizations as incompatible with social work was also demonstrated by a shift introduced in 2014.In a report published by the association ('The positive NAV'), the secretary general described an 'educational journey' motivating the association to develop its professional and political standpoint and consider NAV an organization in which high-quality social work is possible (FO 2014).The association's efforts signal restorative identity work on behalf of their tainted members working in a discredited organization (Breit 2014).
To sum up, although the social work profession initially responded positively to the reform, during its implementation, key agents of the profession refrained from expanding its jurisdiction.When faced with the threat of reformed organizations, the profession instead protected its established jurisdiction.Seeking possible explanations for this reactive response, we turn to the literature of professions.

Possible reasons behind the profession's responses
The organizational context, at a time when the whole NAV organization was geared towards efficiency and control over case processing of social insurance benefits, undoubtedly affected the frontline professionals and made them concentrate on organizational conditions for social work.However, neither the association nor the university teachers were encroached in this context, which therefore cannot explain their lack of agency.Neither the association, which succeeded in influencing the reform policy, made any significant attempts to remind the politicians of the reform goal of attending to the labour market's marginalized citizens and promote social work knowledge and skills as important for reaching that goal.When seeking explanations for these agents' reluctancy to advance the profession's jurisdiction, what comes to the fore is ingrained perceptions of the organization as a context for good social work.
The final section of the abovementioned 2014 report produced by the association, titled 'Workplace NAV', tells a story of three social work students agreeing never to work in NAV.Nevertheless, they end up doing that, although without pride, as the story tells, because among social workers outside NAV, NAV is not portrayed as a workplace you want to brag about at a party, or to other social workers, for that matter, as working in NAV is defined as a betrayal against values such as 'empowerment', human rights, and 'the good relationship'.This image of NAV, against which the report protests, is depicted as an element of the educational journey towards the belief that good social work is possible within the organizational context of NAV.
Clearly, this story must not be read as an accurate description of an empirical reality, though it was not out of touch with that reality either.As Rambøll (2010) reported, such an image existed among members of the profession (see f.ex.N. S. Skjefstad, Kiik, and Sandoval 2018) and might represent a source of their reluctancy to expand its jurisdiction into the new domain.Furthermore, a dominant idea in the core literature of social work curricula was of organizational bureaucracy threatening to undermine important social work principles (Liodden 2021).Frontline social workers' identity of being the only profession in the organization acting in the clients' best interests implied that they felt as if they sometimes had to 'protect' their clients from the system, and possibly also from their new colleagues from the state administration (Røysum 2017).Such an identity aligns with the notion in curricula literature that social workers cannot assume the 'system' will be ethical, and, therefore, must prioritize loyalty to their clients above the demands of public bureaucracies (Liodden 2021).
Furthermore, the perceived dirty character of tasks involved in the reform may have affected the profession's response, as activation work by social workers is seen to de-professionalize social work and jeopardize social work traditions (see van Berkel 2017).It is argued that a conflict exists between social workers' traditional focus on clients' social needs and policies' concern instead with their employability, and that there is an incompatibility between the ethics of social work and the enforcing and sanctioning of activation policies (Hasenfeld 1999;Kjørstad 2005).The reform's welfare-to-work mission could, therefore, explain the professions' avoidance of the jurisdiction of activation work.
While arguing against demanding elements of activation (Messel 2013), the association approved the reform goals.Among the social work educated university teachers only 15% reported absolutely agreeing that except for those completely without work capability, having a job is better than receiving benefits, in contrast to 58% among frontline social workers (Terum 2014, 19).University teachers were also more sceptical of incentives, conditionality, and sanctions than frontline social workers (Terum 2014).Among frontline workers, the introduction of a designated welfare-to-work programme targeting long-term social assistance recipients (the Qualification Programme) stimulated an optimism.They saw the programme to offer opportunities for high-quality social work (Brodtkorb 2017;Gubrium, Harsløf, and Lødemel 2014;Røysum 2013).In 2011, social workers considered their practice activation-oriented to the same degree as other frontline workers (Terum and Jessen 2015).
Nonetheless, during implementation and beyond, textbooks were published about social work with activation tasks (Glemmestad and Kleppe 2019;Rytter 2008; N. S. Skjefstad and Marthinsen 2007).Rather than being unsupportive of the activation goal, these agents argued for an approach to activation that acknowledged some clients had more complex problems than their lack of employment (N.S. Skjefstad and Marthinsen 2007).A concern existed that too much of a focus on activation would leave behind more vulnerable clients (Røysum 2009b), such as 'burned-out' clients for whom there is little or no hope that they will be able to enter the workforce (Røysum 2013, 718).
Thus, concerning the reform's activation goals, there were diverging opinions among the profession's agents.Within the universities, scepticism was more widespread than at the frontline.Here, activation work might have been considered dirty work to be avoided or left to other occupations.
The knowledge base and capacity of the profession to theorize its knowledge and skills in relation to the reform goals may also explain why the profession refrained from expanding its jurisdiction.In social work research, it is argued that professionalization is an impossible strategy because the relational character of much social work makes it difficult to define a unique area of expertise shared across the profession (Healy and Meagher 2004).That is to say, the relational, practical, and moral character of social service work is incompatible with the technical and rational ideals of conventional models of professional expertise.
The ethical dimension seems to have been a key factor in frontline professionals' understanding, representing the defining characteristic separating social workers ('us') from other frontline workers ('them') in NAV (Røysum 2017).Social workers saw social work as 'good' and in the clients' 'best interests'.A grounding in values (wanting the best for others) was fundamental to their understanding of who they were as professionals (B.Fossestøl 2019).Professional ethics symbolized the exercise of their profession, understood as a value-based, practical-moral activity (Røysum 2017).Social workers believed they had distinct competence and ability to make ethical judgements but found the importance of these qualities in them was not recognized by co-workers and managers (Røysum 2014).
Social workers focused on how they performed their work, rather than what they did as social workers and had profound difficulties in articulating social work theoretically and spelling out the specific methods they used in practice (Røysum 2017).According to Røysum, social workers 'appeared to be uncertain about what social work is, especially in light of the theoretical knowledge they could potentially provide and integrate into the Nav system ' (2017, p. 150).
According to Fossestøl (2019), social workers under-communicated their professional knowledge and competence.They seldom referred to theories and methods developed with a scientific basis, and they de-emphasized the theoretical in favour of the principles and practical benefits of their profession.Due to social workers' inability to clarify their knowledge and competence, beyond general references to the core concepts, they appeared ambivalent, seeing themselves as professionals, yet being uncertain whether they were professionals (B.Fossestøl 2019).

Discussion and conclusion
The literature on the professions suggests that, when faced with exogenous change affecting their jurisdiction, professions, such as social workers, will respond by expanding their jurisdiction and positioning their knowledge and skills as essential expertise (Greenwood, Suddaby, and Hinings 2002;Muzio, Brock, and Suddaby 2013;Suddaby and Viale 2011) or by seeking to protect and maintain their established jurisdiction in response to changes perceived as threatening their professional values (Evetts 2003;Healy 2009).
In the implementation of the Norwegian welfare-to-work reform, the social work profession refrained from expanding its jurisdiction into the domain of employment service, and it also refrained from positioning itself as a profession with the expertise needed to meet the political goal of increased employment among groups receiving social security allowances.
Obviously, there are organizational barriers for the success of professional projects (McCann et al. 2013), in this case the scepticism towards social work educations (Rambøll 2010) and the absence of a tradition for employing professions in the central NAV administration (Berg, Heen, and Hovde 2003), as well as its knowledge hierarchy (Breit, Fossestøl, and Pedersen 2018).A professional project might not be effective.Yet, an intriguing matter is why the profession placed so little effort in expanding its jurisdiction.
We have explored three possible explanations inferred from the literature on professions.First, the explanation of pressure for hierarchical accountability and performance management within the reformed organizational context is primarily relevant for the responses of rank-and-file social workers on the frontline.However, organizational pressure can hardly explain why neither the profession's association nor the university teachers operating outside the organization took any significant action to expand the jurisdiction of social workers.
For the university teachers our exploration points to a more general perception of organizational bureaucracy as incompatible with good social work and as a 'system' clients have to be protected from (Liodden 2021).Rather than equipping social workers to constructively handle the dilemma of being linked to policy but also striving for autonomy to meet the needs of the target groups (Nothdurfter 2016), the social work educations seem to have hosted an image of NAV as a morally dirty organization in which it was shameful to work.The association's 'journey' report indicates that this image existed beyond the universities, although the association after a while tried to distance itself from it.
The second explanation revolves around the perceived dirty character of activation -i.e. the potentially intrusive or confrontational tasks associated with conditionality and sanctions required in the welfare-to-work reform.Such a perception was most dominant among the university teachers.The association was critical of conditionality and sanctions, but generally approved the reform goals of employment.There was scepticism among frontline social workers although it was neither strong nor widespread.While value incompatibility might lead to implementation resistance (Tummers 2011), here, the reform's welfare-to-work character does not seem decisive of the workers' reluctance to engage in its implementation.Social workers were concerned for the most vulnerable clients, for whom they believed that employment was not a realistic option.However, instead of seeing the work with stigmatized and disadvantaged clients as dirty work (Ashforth and Kreiner 1999;Muzicant and Peled 2018), they embraced this work as prestigious (in contrast to the dirty 'system').However, they did so without a concomitant conceptualizing of activation as dirty work.
The third explanation highlights the capacity of a profession to articulate its knowledge base and subsequently to theorize (in this case, what employment services for marginalized clients must entail) its knowledge and skills as key resources in the reformed organization.The ability to theorize professional work is an important prerequisite to legitimize professional skills and expertise in relation to new occupational domains (Greenwood, Suddaby, and Hinings 2002).The profession's reservation or reluctance to articulate the concepts and practices underpinning their existing work in more scientific terms (B.Fossestøl 2019; Røysum 2017), may have excluded the profession from extending and reframing those concepts and practices for the new occupational domain of employment services, and positioning itself as a dominant actor in the cognitive map of the field.
The social workers emphasized the profession's strong value base (B.Fossestøl 2019; Røysum 2017).However, a key feature of any profession is a code of ethics obliging professionals to act in the best interest of their clients, and even elite professions struggle with balancing science-based knowledge with care for the clients (Dunn and Jones 2010).It seems the specific challenge faced in the social work profession is that the professionals have failed to acknowledge and articulate social work's specific knowledge dimension (Fjeldheim, Levin, and Engebretsen 2015).This struggle with articulating and theorizing social work is arguably not a defining characteristic of the social work profession in Norway or NAV specifically, but seen in other contexts (Morriss 2017), and presumably a characteristic of the profession more broadly (Webb 2016).
Researchers and university teachers have a special responsibility for theorizing by virtue of being elite members who establish, advance, and communicate the body of knowledge and skills claimed by the profession (Freidson 1985).There were textbooks written in this period that embraced welfare-to-work tasks, but they did seemingly not address work beyond social services and were not made relevant for other occupations.However, as most frontline social workers probably were educated before the reform, their struggle with theorizing the practice and expertise of social work in the new organizational context assumingly refers to a more long-term and inherent characteristic of social work education.
Rather than a theoretical knowledge base, it was instead the institutionalized infrastructure that upheld the professional position and jurisdiction of social workers.Within this institutional infrastructure, the profession's relevance seemingly was obvious and not in need of theorizing.
On this basis, we arrive at two alternative conclusions.One conclusion is that when exogenous change disrupts an institutionalized infrastructure upholding a profession, and the professionals lack a sufficient knowledge base to theorize its importance for the emerging institutional domain, the profession may see no options other than to refrain from advancing a professional project and thereby expanding its jurisdiction.However, recent signs of shifts or evolvement of the profession's strategy over time -textbooks, pamphlets, the 'journey' reportunderscore that a profession's institutionalized perceptions is indeed resistant to change, especially within the universities.Thus, a different conclusion is that professions are more conservative or incapable to recognize opportunities for jurisdictional expansion than theories of professions suggest.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Appendix 1 Overview of sources
The table includes sources examined in the larger project and sources added through the exploration of the research questions on the social work profession's responses to the welfare-to-work reform.The table shows the character and content of each source.
The sources are accessible through university libraries, journal websites, the National Library, or the websites of national authorities and of the association of the social work professions (FO).
A discussion of incompatible models of governance in the implementation of the NAV reform at the street-level, based on the data sources described in Andreassen and Fossestøl (2011).Andreassen (2011) A discussion of efforts to enhance efficiency by organizing specialized units for case processing.The empirical material is key documents from the national administration and interviews from 2010 with all regional directors of NAV about the establishment process.Andreassen (2011) Forvaltningsenheter i Nav.
[Specialised units in Nav.Establishment process and consequences for the handling of tasks].Work Research Institute Report, 2011:5.
Three analyses of the reorganization of the NAV administration with centralized case processing units for national insurance benefits, based on interviews with all regional directors of the NAV administration and an electronic questionnaire to managers and employees at selected frontline offices and at the centralized units, and interviews with managers and employees of three case processing units.Andreassen, T. A. ( 2019 Examination of organizational identity work among members of the publicly criticized and discredited NAV organization.The empirical base is interviews conducted in 2011 with managers of regional NAV administrations (4), administrative personnel in the central administration (2), managers of local NAV offices (4), communications managers in the central administration (3) and communications, managers in the local administrations (3), and front-line personnel (5), recruited through a snowballing method, as well as internal archival data such as press releases, internal magazines, and transcribed speeches by the top management to members of the organization.Breit et al. (2018).From pure to hybrid professionalism in post-NPM activation reform: The institutional work of frontline managers.Journal of Professions and Organization, 5(1), 28-44.
An examination of the strategies of local managers to handle demands imposed by central reform efforts to transform professionalism (post-NPM).The method is an interviewbased case study, and data sources are 86 interviews with managers and frontline workers in seven Nav offices located in Norway's largest cities. Breit et al. (2018) 2018) Berg et al. (2003).Kvalitetsbyråkratiet -mellom autonomi og kontroll [The quality bureaucracy -between autonomy and control] Skriftserie, Arbeidsforskningsinstituttet.
Case studies of selected public service organizations, including the Public Employment Service (now merged into the NAV organization).Christensen, T., et al. (2007)."Reform of the employment and welfare administrations -the challenges of co-ordinating diverse public organizations".International Review of Administrative Sciences 73(3): 389-409.
Examining the decision-making process and the organizational thinking in the NAV reform.The empirical data are mainly a combination of personal interviews with central political and administrative actors, public documents and several theses and studies of the reform process and effects of other relevant reforms and reorganizations.Erlien (2017).Nav-reformen som politisk prosjekt: Bedre enn sitt rykte, men dømt til å mislykkes?[The NAV reform as a political project: Better than its reputation, but doomed to fail?] Nytt norsk tidsskrift, 34(4), 366-378 The article discusses general features of Nav and the background to the reform in 2006, shows how the establishment of the social sector in the 1960s, as a political reform, has clear similarities with the Nav reform, and how the social sector from the beginning struggled with problems that the Nav reform can have contributed to solve.Fimreite and Laegreid (2009).Reorganizing the welfare state administration.Public Management Review, 11(3), 281-297.
A discussion of the network-like organizational structure, based on interviews with forty-three central political, administrative actors and stakeholders at central government level, and extant empirical studies.Christensen, T. and P. Laegreid (2011).Complexity and hybrid public administration -theoretical and empirical challenges.Public Administration Review 11(4): 407-423.
A discussion of the increasing tendency towards hybrid administration, the attempts to balance NPM and post-NPM, and the major mechanisms producing hybridity in a modern reform, with NAV as the case.

Scientific publications and research reports about reform context and implementation trajectory
The publication's aims and methods Christensen and Laegreid (2012).Competing principles of agency organization -the reorganization of a reform.International Review of Administrative Sciences, 78(4), 579-596.
Analysis of how welfare administration changes in the implementation process through changing of organizational principles and design.Data sources are elite interviews in the relevant ministries and agencies, in two rounds: the process leading up to the reform (before 2005) and the reorganization (in 2008).Christensen, T., Fimreite, A. L., & Laegreid, P. (2014) A discussion of different interpretations of the Norwegian policy for labour market (re)integration and how to achieve quality in frontline vocational rehabilitation.Analysis of the problems with integrating three different knowledge systems in the former employment, social insurance and municipal social assistance services), and of a possible common basis for a new occupational role in the Nav offices.Data sources are a qualitative case studies and interviews in four Nav offices of different sizes and geographical locations, collected in two periods (2008-2009 and 2011).Helgøy, I., Kildal, N., & Nilssen, E. (2010) Terum, L. I., et al. (2017).Når vilkår og aktivitetskrav brytes.
Analysis of how frontline workers approach situations where clients do not behave according to the terms, and to what agree this behaviour is sanctioned.Draws on a survey of frontline workers in NAV, with nearly 1500 respondents, collected from 113 frontline offices in spring 2015.Terum (2014).NAV, arbeidslinja og sosialarbeiderutdanningene.Skriftserien.Høgskolen i Oslo og Akershus An investigation about what the teachers who teach and supervise at all university colleges and universities in Norway that offer education in child welfare pedagogy, social work, and social pedagogy, think about the educations and their relevance for NAV.The basis is a survey with a response rate of 67% of the population of 714 teachers.Terum and Jessen (2015).(Continued)

Fossestøl
et al. (2015).Managing institutional complexity in public sector reform: Hybridization in front-line service organizations.Public Administration, 93(2), 290-306 An analysis of how local Nav offices respond to institutional complexity between NPM and post-NPM demands in the reform.Based on a four-year research-based evaluation of the Nav reform (2007-2010) involving qualitative case studies of 10 Nav offices.Fossestøl, K., Breit, E., & Borg, E. (2014).NAV-reformen 2014: En oppfølgingsstudie av lokalkontorenes organisering etter innholdsreformen.[A follow-up study of the organization of the local Nav offices after the 'innholdsreformen'].Work Research Institute, Research report.A follow-up study of the initial evaluation of the street-level implementation of the Nav reform.Data sources, collected in 2014, are (a) survey to the same 57 Nav offices and (b) qualitative case studies in 10 (of the original 18) Nav offices.Fossestøl, K., et al. (2016).Hvorfor lykkes ikke NAV-kontorene med å jobbe mer arbeidsrettet?[Why do the Nav offices not succeed in working more labour market oriented?]Søkelys på arbeidslivet 33(01-02): 5-23 Research article based on the follow-up study (see Fossestøl et al., 2014) Hagelund, A. (2016).The activating profession: Coaching and coercing in the welfare services.International Journal of Public Sector Management 7(29): 725-739 Exploration of the consequences of more active and individualized welfare policies for conceptualizations of professional competence in the frontline services.Data sources are 25 interviews with frontline workers in two Nav offices, in the period 2012-2013.Helgøy, I., et al. (2013)."Utvikling av en arbeidsrettet spesialistrolle i Nav".[Development of a work-oriented specialist role at the Nav-offices] Tidsskrift for velferdsforskning 16(3): 141-156 Den tvetydige aktiveringen: en studie av veiledere ved lokale Nav-kontor [The ambiguous activation: A study of frontline workers in local Nav offices].Tidsskrift for velferdsforskning, 18(2): 96-109.An examination of how supervisors at local NAV offices balance the potentially contradictory objectives of, on the one hand, control and paternalistic pressure and, on the other, flexibility and user influence, and to what extent the balancing is influenced by the supervisors' educational/ professional background.The article is based on a nationwide survey of supervisors at 115 local NAV offices, undertaken in 2011.Terum, L. I., & Sadeghi, T. (2019).Medarbeidernes kompetanse ved NAV-kontorene.[Frontline workers in the Nav offices' competence] OsloMet Skriftserien.Describes changes and variation in the educational background among the frontline workers in Nav in the period 2011-2018.Draws on three surveys, in 2011, 2015 and 2018, in the same Nav-offices.
. "A knowledge hierarchy in labour and welfare services?Evidence-based and practice-based knowledge in frontline service innovation".International Social Security Review 71(4): 13-32 Research article based on the evaluation of the R&D project 'Practice and knowledge development in Nav offices'.Empirical material is appr.80 interviews with involved participants on different levels in Nav, and a survey to the involved participants (2015-2017).Breit, E., Fossestøl, K., & Pedersen, E. (2019).Kunnskapsbasert praksis innenfor en samstyringsmodell.En analyse av satsingen «Praksis-og kunnskapsutvikling i Nav-kontor».
Tidsskrift for velferdsforskning, 22(03), 184-197.Research article based on the evaluation of the R&D project 'Practice and knowledge development in Nav offices'.Same empirical material as Breit et al. (

publications and research reports about the social work profession in the reform context and about their responses to the reform The publication's aims and methods
Collection of the results of 13 task forces in different universities, proposing views on common professional competence for educations of social workers, child welfare workers, and social educators.Describes distinctive features and common features in terms of their purposes, which value-based and/or academic foundations the education has, which target group it has and the field of work it is aimed at, and finally which knowledge, skills and methods are particularly prominent.Hansen, H. C., & Natland, S. (2017).The working relationship between social worker and service user in an activation policy context.Nordic Social Work Research, 7(2), 101-114.Scientific publications and research reports about reform context and implementation trajectory The publication's aims and methods Røysum (2013).The reform of the welfare services in Norway: one office -one way of thinking?European Journal of Social Work 16(5): 708-723.Analysis of the professional role of social workers in the new Norwegian Nav system.Data material are (a) editorial content of the journal produced by the Norwegian Union of Social Educators and Social Workers (FO) 2005-2011, (b) a fieldwork study of a Nav-office, 2007 and 2008, and (c) observations at conferences on the reform and the role of social work, 2005-2011.Røysum (2014)Ubehaget ved å forenkle det komplekse [The discomfort of simplifying the complex] In A. Ohnstad, M. Rugkåsa, & S. Ylvisaker (Eds.),Ubehaget i sosialt arbeid (pp.141-159).Gyldendal Akademisk.A discussion of how social workers at the NAV offices encounter expectations that the clients' complex problems must be simplified.Based on the same empirical material as Røysum (2013) Røysum (2017) 'How'we do social work, not 'what'we do.Nordic Social Work Research, 7(2), 141-154.An exploration of how Norwegian social workers in NAV offices experience and explain social work as professional knowledge and theory, and examines examples from their experience working in new, multi-professional settings.Based on the same empirical material as Røysum (2013).Skjefstad (2013)."Er det rom for sosialt arbeid i Nav?" [Is there room for social work in Nav?] Fontene forskning 1(13): 76-88.Analyses whether there is room to work based on social work principles with users with complex and holistic service requirements.Data material are parts of a follow-up study in the period 2008-2010, involving participant observation in meetings concerning the Qualification Programme and group interviews with social workers.Skjefstad et al. (2018).The role of social workers under neoliberal ideology at the Norwegian Labour and Welfare Service (NAV).In Neoliberalism, Nordic welfare states and social work (pp.137-147).Routledge A discussion of the following questions: How has the role of social workers changed under the influence of neoliberal ideology during the past decades?What implications do these changes have for the practices of social work?Social work at the Norwegian Labour and Welfare Service (NAV) is used as an example.The empirical material consists of interviews with 15 social workers, conducted from March to April 2012.