Managing development plans in the Norwegian healthcare system

Background: As part of a national plan to govern professional and organizational development in Norwegian specialist healthcare, the country’s hospital clinics are tasked with making development plans. Using the development plan as case, the paper analyses how managers navigate and legitimize the development plan process among central actors, and deals with the tension between governmental control and organizational autonomy in such strategy work. Method: This study applies a qualitative research design, triangulating different kinds of data. The study was performed in two steps: a) an analysis of various guidelines for development plans and how they are translated through the different levels of Norwegian specialist healthcare, and b) a single-case study of the process of developing a development plan at the clinical level. Findings: Findings shows that the development plan was shaped through a multilevel translation process consisting of different contending rationalities. At the clinical level, the management had diculties legitimizing the process. Underlying tension between top-down and bottom-up steering challenged their involvement and made it dicult to manage the contingency of decisions. Conclusions: The ndings are relevant to public sector managers working on strategy documents, as well as to policymakers identifying challenges that might hinder fulllment of political intentions. This paper uses a case from Norway; however, the ndings are of general interest. It contributes to the academic discussion on how to take account both the health authorities perspective and the organizational perspective in order to understand tensions between control and autonomy, applied to a functionally differentiated health care system.

on clinic level shall make the foundation for regional development plans. Joint planning is highlighted as important for improving integrated care [5], and is from this perspective connected to the intentions in the Norwegian coordination reform. On the organizational level, the development plan also includes a "mission statement", stating the organization's purpose and identi es the scope of its operations in service and market terms [6]. Public management scholars have argued that de ning a strong mission is a useful step toward improving the performance of a public organization [7].
Development plans are interesting in light of current public sector reform changes, because they challenge the tension between control and autonomy. On one side, they are used by governments to control organizational and professional development. On the other side, the development plan represents an organization's autonomy; its own de ned purpose and the plan is supposed to be developed within the organization. This paper argue that one should take account both health authorities-and an organizational perspective to understand the tension between control and autonomy. Using the development plan as case, the aim is to show how managers navigating and legitimizing the process among central actors, and deals with the tension between control and autonomy in such strategy work.
The overall aim is divided into two research questions; How is work with development plans organized through the Norwegian health care system, and how does this effect the work on a clinic level?
How do local managers navigate and legitimize the development plan process in a functionally differentiated health care system?
In order to analyses the complexity the health care system entails, the paper combines [8] systems theory with public sector management theories on management strategies. A system theory inspired perspective on the functional differentiation of the healthcare system is used to analyze the structural couplings between stakeholders with different rationales and the tension this might bring in the plan work [9]. Management theories, speci cally those regarding inclusive management, are analytical tools that can be used to analyze the way in which a manager facilitates participation and inclusion when working with an organization's development plan.
A functionally differentiated healthcare system Functional differentiation is central in [10] system theory. Functional differentiation argues that modern society is divided into several functionally specialized and autonomous subsystems, all managing speci c functions for society as a whole. In this paper, we argue that functional differentiation is the primary form of differentiation in the healthcare system. Describing the healthcare service as functionally differentiated implies that it is understood as one system comprising several autonomous and selfreferencing subsystems, each maintaining its own function for the healthcare service as a whole.
"Subsystems" refers to the various professions, organizational units, and administrative levels in the overall healthcare system. Such perspective has a stronger focus on the subsystems autonomy then formal and hierarchic organizational structure in the health care system [9].
The cornerstone of [8] systems theory is its distinction between a system and an environment.
Furthermore, Luhmann describes a system as being operatively closed, implying that any operation is always the result of conditions of possibility determined within the system itself. Any action or decision taken by the system is based on the system's own logic and understanding. This implies that at all development plans are created inside the various subsystems. In its most basic form, development plan can be understood as a way of operationalizing an organization's function; de ning it's purpose, addressing why the organization exists, and what it wants to achieve [11]. Hence, using [8] concept, the organization can use the development plan to de ne "it self" in relation to its environment.
Functional differentiation also points to the structural coupling between the healthcare service and the various functional systems in society. A functional system is an abstract communication system that an organization attaches itself to by making decisions. Each functional system refers to its own logic, rationality, and communicative structures [10]. Economic, political, and health systems are examples of functional systems that employ different criteria for observing the world. Table 1 shows the various codes, mediums, programs, and functions that organizations can activate through functional systems. Source: Adapted from: [12] The basic premise of [10] work is that the differentiation process of modern society entails the crystallization of organizations that are attached primarily to one functional system. Political parties and public administration would communicate through the political code while banks and businesses would communicate through the economic code. Many modern organizations do not attach themselves to one main functional system, but to multiple functional systems. This is especially evident in public sector organizations. This "polyphony" means that organizations of the same type could, in principle, attach themselves to different functional systems with crucial effects on their communicative structures. The functional system to which an organization chooses to attach to when making a decision will have consequences for how it communicates and for how the organization fundamentally evolves [13]. In the present case, it would make a difference whether healthcare organizations attach themselves to the health, economic, or political system when communicating their development plans. Different functional systems cannot understand one another's rationalities and evaluation criteria, and this indicates a possible central tension when constructing a development plan in a functionally differentiated healthcare system. In the process of making a development plan a healthcare organization must address the needs of stakeholders representing different functional systems. This is central in light of systems theory because an organizational system creates itself by forming an internal structure that mirrors its environment [14]. An organization's self-description thus greatly depends on how it constructs its environment. By constructing an image of its surroundings, the organization concomitantly constructs an image of itself. In the present case, a healthcare organization is not only one system within one environment but often operates within several systems/environmental constructions. One must accordingly observe how such organizations communicate multiple system/environment demarcations in their development plan work.

Mangers role in coordinating development plan processes
This study especially observe how managers handle the tension between communicate multiple system/environment demarcations and the tension between control and autonomy on an organizational level, using development plan work as an case. In systems theory, organizations are seen as social systems that can stabilize forms of action and behavior by deciding about stronger or weaker conditions for practices and procedures [14]. In other words, organizational systems operate through decisions and decision communication.
A key theme in [14] organization theory concerns how organizations manage the contingency of decisions. The contingency of decisions points to the paradox of decisions and the fact that a decision is neither necessary nor certain, but could always have been made differently. This contingency makes connectivity less likely because it calls into question the notion of connecting to a decision that could inherently have been made differently [15]. Connectivity is essential to decisions because it is only the connection to further decisions that can turn a decision into a real decision. A decision to which no further decisions are connected turns out to not be a decision, just noise.
[15] point out that a main strategy for how organization manage the contingency of decisions is displacement. For example, an organization can "deparadoxify" its decisions by interpreting them as necessary responses to external stakeholders, displacing the contingency to the environment. Another way is to displace the contingency to decision-makers, such as the managers of the organization.
To observe how organization handle the paradox of decisions this paper uses the concept of inclusive managers and brokers. Inclusive managers brings together participants from different practices in collaborative settings. Such a collaborative setting could be important in the process of creating a development plan since the statement should be anchored in the organization's employees to be legitimate and successful [16].
A manager can assume the role of promoting, as well as inhibiting, inclusion. An inclusive manager tries to design inclusive processes and create a community of participation in which people can share information and perspectives and work together. However, it is not enough simply to bring people together; the people must also be willing to listen and be engaged in ways that advance the collaborative process, in which inclusive managers play a key role. Inclusive managers identify various relevant areas and know the problems faced in each one. They encourage people to see different perspectives in discussions or meetings, fostering an atmosphere in which problem-solving occurs. However, these meetings do not necessarily enable people to feel connected or to trust each other, so joint activities in this sense can be either constructive or destructive. Managers must try to create joint activities that provide a shared experience and transcend boundaries between participants [17].
[18] demonstrated that inclusive processes do not have to be driven by managers and emphasized the role of "brokers." The work of brokers is similar to that of inclusive managers, as they make coordination possible by opening up new possibilities for learning and exchange. Brokers help other actors transfer, translate, or transform the meanings encountered during joint activities [19]. A development plan in a functionally differentiated healthcare system must address the needs of different external and internal actors. Translating and transforming meanings and knowledge between these actors is essential in the process of making and legitimizing the plan.
A broker translates knowledge created in one group into the language of another so that the new group can integrate it into its cognitive portfolio. To do this, brokers must be able to manage the relationships between individuals as well as act as translators. The broker's role necessitates a delicate balancing act.
To be effective, brokers must have authority in all of the groups to which they belong. They must be able to evaluate the knowledge produced by the different groups and to earn the trust and respect of the various parties involved. Over time, the broker's activities may lead to the development of a repertoire of shared resources, such as the rules and procedures used by the group [18].
The theoretical perspective show that development plan work in a functionally differentiated healthcare system is complex and consist of tensions between different logics. The concepts inclusive management and brokers are used when analyzing how organizations handle this complexity, and how mangers navigate and legitimize decisions on the development plan within the organization.

Methods
The research design is structured as a two-step process. First, a larger content analysis of central documents (see Fig. 1) related to development plan work in the context of Norwegian specialist healthcare was conducted. In this part of the study, the national government's intentions for the development plans were analyzed, as was the way in which these intentions were translated through various organizations and government levels. The second step was to perform an in-depth analysis of development plan work at the clinical level. The selected case is that of a clinic dealing with mental health issues and drug addiction. This clinic was "under construction," meaning that it had recently been reorganized with a new structure, as two clinics (a mental health clinic and a drug addiction clinic) had been merged into one. This organization is segregated into ve units and 27 sections with approximately professional tension between the geographically separate organizational units. This new clinic was, therefore, seeking to de ne its culture, work processes, visions, etc. The case can therefore be seen as a «extreme case» [20, p. 23]. An extreme case for this study is relevant because such case has the possibility to construct a development plan based on a "blanc sheet". A more established organization might have used a copy-paste" strategy, reproducing older strategy documents/plans.
At the clinical level, group observations, individual semi-structured interviews, and public documents were used as data sources. Observations were made at three strategic management meetings in which the development plan was discussed. Group observation made it possible to capture ongoing discussions within the development plan work, especially problems with the work and content negotiations, as well as the managers' various roles and in uences in the process. The meetings were held 2-3 months apart and represent different parts of plan process.
Ten key people involved in the work process were selected for in-depth individual interviews: the clinic's adviser (who was given internal responsibility for development plan work), four unit managers, and ve section managers. Interviews with the unit and section managers were conducted during the process of developing the plan. An interview with the clinic's adviser was conducted after the evaluation of the work process. The interviews were conducted in Norwegian and were audio-recorded and transcribed. The project was ethically evaluated and approved by the Norwegian Centre for Research Data.
This research employs a qualitative research design, which combines different data gathering methods (interviews, document research, and observation). Qualitative data triangulation has several purposes. The combined data sources contribute to a "thick" and complex description of the studied case or phenomenon. Data triangulation is also an important method of ensuring validity [21]. In our research, interviews were used to further explore discussions that were observed or to investigate how statements in public documents are operationalized by different actors. [22 p. 96] discuss the challenges of combining qualitative approaches and distinguish between "mixed qualitative methods" and "synthesizing methodologies." Our approach uses a mixed qualitative method, which is a single qualitative study operating within a singular methodology but using more than one method of data collection. This means that data from different methods are analyzed through the same analytic framework, and are thus epistemologically congruent. This theoretical framework guided the analysis of the empirical data; for example, the emphases of the different codes and system boundaries were identi ed, as were actions taken by the managers and presumptive brokers to achieve an inclusive process at the clinical level.

Results
Multilevel translation: a local focus, but a national work process As part of implementing the National Health and Hospital Plan, all health authorities have had to create their own development plans. This section examines how government intentions are translated and operationalized through various government guidelines within Norwegian specialist healthcare. [8] theoretical concepts of connectivity and contingency are used to analyze decision communication regarding development plans in what we call a multilevel translation process. Figure 1 presents the guidelines used and illustrates how they are translated vertically in the Norwegian specialist healthcare system. In this process, different health authorities try to ensure their decisions' connectivity and limit the contingencies by implementing the various guidelines. The overall national goal is to facilitate good future-oriented patient treatment, as well as the rational prioritization of resources at every level of Norwegian specialist healthcare (see no. 1, Fig. 1). Guidelines (see no. 2, Fig. 1) on how to manage development plans were sent by the Ministry of Health and Care Services to the ve regional health authorities. According to the Ministry, the national guidelines for development plans should help navigating the practical work. The political goal of the guidelines is to achieve collectively binding decisions through the development plan formulated by the various authorities, clinics, and hospitals. The guidelines present a recommended thematic structure for the development plans, whose main components are to be historical background, current situation, contextual change, desired situation, and strategic choices. The guidelines also emphasize transparency and stakeholder involvement. All documentation connected to the development plan process are therefore published, and at all levels there should be broad involvement of users, patient organizations, professionals, unions, municipalities, and private actors. The emphasis on involvement and transparency can be understood as a way of legitimizing the political system's function of achieving collective binding decisions regarding healthcare strategy.
The regional health authorities then translated the national guidelines according to their mandate (no. 3, Fig. 1), serving as a premise for the work on local development plans. In this case, the regional health authorities stated that local development plans must follow national and regional instructions issued through various parliamentary reports and reforms. This includes guidance from the National Health and Hospital Plan [3] and the Coordination Reform [23]. The regional instructions concentrate on how local health services can be adapted to improve regional economic conditions, e ciency, capacity, and competence. They also emphasize how to improve internal coordination and collaboration with other health regions. Through the regional mandate (no. 3, Fig. 1), the health authorities try to ensure connectivity to both the National Health and Hospital Plan and previous reforms. At the same time, translation is done through a shift in focus from a national and political context to a regional and economic context. The goal of the development plan is no longer merely to create binding decisions regarding strategy but to create binding decisions regarding strategy based on regional economic conditions. In light of systems theory, the regional mandates entail incorporating the economic code into development plan work.
Based on their mandates (see no. 3, Fig. 1), the local health authorities created their own guidelines (see no. 4, Fig. 1), which ensure connectivity by stating that the local development plans must follow the thematic structure recommended in the national guidelines (see no. 2, Fig. 1). The local guidelines (see no. 4, Fig. 1) also state that local development plans must include three types of perspectives: local health service, coordination, and clinic perspectives. The local health service perspective should take account the various clinic perspectives, and the coordination perspective is to be developed by a working group consisting of representatives from various clinics and municipalities. This differentiation into perspectives allows space for contingency because there is no longer any integration between decisions regarding the different perspectives. This means that the clinic perspectives do not have to connect themselves to those of the local health authorities or the coordination function.
At the last level, the various clinics, who had responsibility for the clinic perspective, makes their own notes (no. 5, Fig. 1) on how the work process should be structured in their own organizations. These notes state that the plan should be based on previous strategy work, and on regional and national guidelines. Neither the National Health and Hospital Plan nor the Coordination Reform is mentioned in the note (no. 5, Fig. 1).
This section has analyzed the development plan from a top -down perspective. It shows that intentions and rationalities from the national government level are translated through the government system. The translation process has strong connectivity, enhancing the top-down control element. The process also contains room for contingency in the form of local interpretations, highlighting the autonomy of the different subsystems. After nalization at the clinical level, the development plans are sent to the upper organizational level; these plans are then used as a foundation for the local and regional development plans (see nos. 7-9, Fig. 1). This process is not covered in this article.
A development plan in a clinic under construction The following section presents the work with the development plan at a clinic for mental health and drug addiction. This organizational level was selected for a in depth study since it holds most tension between control and autonomy. The section is divided into three parts: 1) presentation of the development plan and of the clinic's reactions to the requirement of making a plan, and it's interpretations of the process; 2) how the clinic organized the plan work and 3) mangers managing reactions on the nalization of the plan.
Introducing the work with a development plan at clinic level The rst time the clinic manager presented the idea of making a development plan was at a leaders' meeting in June 2017 at which various managers and union representatives were present. The clinic manager began by informing the participants that the clinic needed to address strategy and development.
He then presented his view of the development plan, which he called "a political requirement," the aim of which was to obtain an overview of how the various health organizations organize their work at the national, regional, and local levels. The clinic manager noted the three perspectives from the regional mandate (see no. 3, Fig. 1) and highlighted the thematic structure in the national guidelines (see no. 2, Fig. 1). He speci cally highlighted the clinic and the local health authority's economic conditions and proclaimed that "a key question when working on the development plan is how we can get as much health for the patient given the economic constraints that we work under. There have to be strategic prioritizations" (Observation 1).
The clinic manager's presentation illustrates how the various guidelines have ensured connectivity. The main translations at the different levels of the health authorities, shown in the previous section, were all covered in the presentation. By calling the development plan a political requirement, the clinic manager activated the political system, placing the contingency of further decisions within the context of a concrete decision premise, namely, the political guidelines. The clinic manager's emphasis on economic conditions also implies that the economic code should be activated when working on the development plan.
During and after the presentation, there were many reactions from both the union representatives and various managers. There were negative reactions related to the political establishment of guidelines through a top-down process lacking professional involvement. For example, one section manager asked why politicians, instead of their own healthcare professionals, were allowed to formulate guidelines for the clinic's development plan. Another section manager argued that "decisions on the future direction of the clinic can't be made only at the top of the system, as the professionals working with the patients must be involved in the process." These reactions were connected to the strong use of the political code when presenting the development plan. The way the clinic managers set the contingency of decisions within the context of the politically de ned guidelines also resembled a top-down process. By emphasizing the role of healthcare professionals within the organization, the reactions were intended to place decisions regarding the plan inside the organization itself, in this way activating and reinforcing the functional system of health in the ongoing process.
Based on observations of the presentation of the development plan (Observation 1), there was clearly a gap between the organization's top management (e.g., clinic and unit managers) on one hand, and the section managers and union representatives on the other hand. Top management was more comfortable with political rhetoric and working to achieve political and economic goals set at the national and regional levels. Using [8] concepts attaching their communication to the political and economical function system. Section managers and union representatives on the other hand was more link to functional system of health. Discussion in the meeting addressed the fact that the organization had not yet decided how to manage the contingency of decisions connected to the development plan, or chosen the functional system through which the decisions should be made.

Making our plan
This section presents the "hands-on work" of the development plan, constructing the content of the plan and attempting to engage healthcare professionals. In this section, systems theory will be complemented with the theoretical concepts of "broker" and inclusive management to analyze the managers' role in facilitating participation and involvement when formulating a development plan.
Four weeks after the development plan work was rst presented, a new meeting was held, led by the clinic manager and a clinical adviser, who were responsible for writing and coordinating the clinic's development plan. The clinical adviser started the meeting by stating that it was important that the development plan was based on a mutual understanding of how it can be used in the clinic. The work process must be driven from within the clinic, and the local context must be in focus. He also emphasized that the work must end up with something more than merely a political document (Observation 2). Here, the clinical adviser was alluding to inclusive management [17] and attempting to translate the development work from a political requirement to an internal collaborative practice. To reach this goal, he had to translate and coordinate the political and economic aspects of the various guidelines into a language with which the various organizational actors could relate. In the interviews, performed after the meeting, the clinical adviser followed up on his role and re ected on the process of making the development plan: For me it was very important that our plan re ect our organization and the professional work done throughout our organization. I was concerned that this was not happening. Several times I tried to make the point that it had to be our process. Our process of actually guring out what our patients' needs would be in the future, and how we should meet them. I felt in the meeting, and several times afterwards, that this point wasn't getting through. (Clinical adviser) At the meeting, the clinic manager stressed the importance of having a development plan by pointing out that the clinic lacked a clear purpose and direction and often had to make reactive and rash decisions instead of being in front of situations as they arose. The clinic manager argued that the development plan could be a tool for turning around this dynamic. In addition, the clinic manager wanted the development plan work to be a collaborative practice and emphasized the importance of involving the whole organization. He stated that it was a personal choice to become involved, but that the managers should try to facilitate discussion about where the focus should be regarding local challenges and areas of improvement. He also pointed out that the management group (clinic manager and unit managers) could not make a good product by themselves (Observation 2).
Both the clinical adviser and the clinic manager emphasized the importance of translating the development plan work from the political to the local context. They also pointed out that all managers were responsible for encouraging involvement in and enthusiasm for the plan process throughout the organization. This meant that all managers (unit and section managers) in the organization were encouraged to assume the role of "brokers" [19]. This role was also highlighted through the method by which the clinic organized the work process of making the development plan. To involve the whole organization, it was decided that the work process should follow the "line principle." The goal was to use the existing organization and areanas within the organization to develop the development plan. This meant that all unit managers were to use their hierarchical lines of authority to produce a description of their current situation, possible challenges, and future goals, and to ensure broad involvement. In this work process, the unit managers had to involve their section managers, who would then involve their healthcare professionals through various staff and union meetings.

Top-down or bottom-up?
Despite the management's goal of involving the healthcare professionals in constructing the development plan, this did not quite work out. The line principle did not have a positive impact on involvement, and the managers did not succeed in their job as brokers. Based on the interviews and observations, it seems that the line principle enhanced the opinion that the plan work was a top-down process. In the interviews, the section managers responsible for involving the healthcare professionals said that it was di cult to get any feedback and enthusiasm from them: There has been little talk of the mission plan in our section. The goal of the work process was that it should be bottom-up. Yes, it was, and there has been a lot about that goal and I think many are tired of the whole thing. Especially those at the bottom. I found it hard to get any feedback or engagement.
(Section manager 2) For the development plan, everybody was supposed to be involved, and our clinic has over 1000 employees … There are guidelines from the Ministry of Health and Care Services about what we are to deliver, and from Central Health Norway through the mandates. This quickly met with skepticism and indifference. For the employees and many of the managers, the work feels like a duty, and then people lose their commitment and enthusiasm. This has been top-down, not bottom-up at all. We are invited in, but too late and everything is already been de ned. (Section manager 3) The ambivalence evident at the meeting was due to the fact that the development plan was part of the larger political health policy project while being an essential part of the clinic's strategic development. In response to criticism that too much planning and too many processes were happening at the same time, the clinic manager argued that a new process must not displace old processes. There is a political demand that the clinic must deliver on, and it should be integrated into the clinic's strategy work. The clinic manager noted that the work can also be used to get a full picture of the main challenges and collective goals to address (Observation 2).
Here the challenge of translating the political requirements into local strategy work is obvious. The clinic manager attempted to stress the importance of the development plan, but he also called it a political demand that they might as well try to use positively. The clinical adviser also noted the contradiction between a top-down demand and a bottom-up strategy process, feeling that this had made it di cult to foster involvement and commitment from the professionals: Another element of this is that the plan is part of a bigger political order. Everybody knows that our plan will be almost invisible in the local and regional plans. Maybe we will be able to nd traces of it, but its essence will disappear. This makes it more di cult to get commitment from the professionals. But it doesn't change the fact that we need this kind of plan for ourselves. So for me, the process in the clinic of working on such a development plan may be more important than the document that we will send to the local health authorities. (Clinical adviser) Achieving a shared mission in a functionally differentiated organization The last phase was to nalize the development plan. As in previous phases, to obtain broad input and legitimize the content, the nalization of the development plan was discussed at a managers' meeting. At this meeting, all the managers and union members representing the various professions were present. The objective was to present and discuss the feedback on the development plan work process, and to discuss which areas should be prioritized in the nal version of the statement.
The meeting started with a presentation by the clinic manager. He stated that the development plan had been introduced as part of a governing process for professional and organizational development in clinics all over the country. This meant that decisions regarding professional and organizational change must be re ected in the development plan before they could proceed. The development plan should, therefore, be the basis for future decisions. This illustrates how the governing bodies and healthcare authorities want to use development plans to ensure connectivity in decisions concerning organizational and professional development. For the clinic manager, it was therefore important that certain key areas in the organization be prioritized. The clinic manager argued that the work that had been done so far did not constitute a good basis for assessing what professional and organizational changes were needed for the future. He stated that to make this assessment, the professionals must be more involved in the ongoing development work (Clinic Manager 2, Observation 3).
One section manager questioned the clinic manager's thoughts on involvement when it came to decisions on prioritizing: The national government has given us an order. We cannot prioritize 19 areas. It is a managerial responsibility to decide what is to be prioritized. We cannot have a democratic process on this ... There are so many different motives and wishes in the clinic, that it must be up to the leaders to decide what our focus should be. At the same time, there are quite clear [political] guidelines about what we really should prioritize. (Section manager 2) This discussion was concerned with how the clinic should manage the decision contingencies connected to the development plan. The section manager made the point that without displacing the paradox or handling the contingency [15], there will be no decisions: the various subsystems are not motivated to understand one another and are therefore incapable of reaching an agreement on what should be prioritized in the organization. In the interviews after the meeting, the section managers criticized the work process, calling it "skin-deep democracy": I think the process is being contaminated by "skin-deep democracy," when it should have been [a matter for] good strategic management. The belief in democracy and involvement is "in the time," but not all administrative decisions should be made bottom-up. Administrative changes and organizational goals should be decided at the top. It could be unpleasant, but you cannot make good strategic decisions if you expect everybody to be involved. (Section manager 2) The clinic manager responded: I have to follow the assignment given to me. If I don't, I have to nd another job, but it is also important to get the process right. I can't and won't decide everything alone. That's why I think it's important to involve the whole clinic. When the time comes, everyone should have had the opportunity to get involved, and I will make a decision. (Clinic manager 2) At the meeting, the clinic manager opened up a discussion of how the organization should follow up on the development work. He asked whether there was any real desire to get involved, and what was needed to get the healthcare professionals involved. These questions started a discussion of how the clinic should proceed to ensure the engagement of the healthcare professionals in the future process (Observation 3).
Several union representatives expressed their views on how to involve the healthcare professionals. One union representative stated that they found it di cult to get involved in the process because they could not relate to the general matter of clinic strategy. Another union representative added that there was no shortage of commitment from the professionals in regards to working with patients, but that it could be di cult to get them involved in general organizational matters (Observation 3). This feedback shows that the managers did not succeed in their brokering role [18]. In the previous meeting, it had been stated that it was the managers' responsibility to translate the "general" aspects addressed by the various guidelines into more concrete elements concerning the professionals' local context and practices.
A unit manager then pointed out that the clinic has many areas in which to facilitate broad and open processes and asked how these could be better used to achieve broader involvement in nalizing the development plan. Another union representative agreed, believing that it would be easier to involve the healthcare professionals in strategy and development questions if they had areas in which they could discuss them, rather than simply being told to get involved in something by the management (see Observation 3). It was also pointed out that the healthcare professionals cannot be seen as a homogeneous group. For example, one section manager stated that it was important to involve the professionals, but did not believe that meetings between the various specialists would lead to any unity or mutual understanding because there was too much professional disagreement in the organization. One unit manager responded by stating that it would be unfortunate to gather the different specialist groups separately, as this would only reinforce the differences between them (see Observation 3).
The point being discussed here concerns the functional differentiation in the clinic and how this challenge the goal of having a "mutual" mission and development plan. A mission statement should de ne an organization's unique and enduring purpose [11]. The problem for the clinic was that the various subsystems represented by the organization's sections and professional disciplines operated according to different purposes and understandings based on their functions in the clinic. This theoretical point was exempli ed in the interviews when two of the unit managers re ected on the heterogeneous group of professionals working in the clinic: To achieve good collaboration in the clinic, we have to break down the professional boundaries between drug addiction, psychiatry, and rehabilitation and the geographical boundaries between north and south.
(Unit manager) The problem is that our focus is on ourselves and not on the clinic as a whole. Everyone looks at the clinic based on their own sections and unions. The goal must be to achieve a shared understanding... Or that we should at least relate to the clinic as a whole. The goal must be to bring about a common culture with the patient in the center. (Unit manager 2) Both quotes show that the mental health and drug addiction functions consist of different subsystems, both organizationally and professionally. These different subsystems operate according to different understandings and cultures, making it di cult to achieve uniform understanding and consensus when it comes to describing the organization's challenges and long-term goals. Elements that are essential for developing a strategic plan [24].

Conclusion
This paper have focus on development plans as part of reform work in the health care sector.
Development plans are interesting in light of current public sector reform changes, because they challenge the tension between governmental control and organizational autonomy. This paper argue that one have to take account both the health authorities perspective and the organizational perspective to understand the tensions and complexity regarding strategic planning in the public sector. Using the development plan as case, the aim has been to show how managers navigating and legitimizing the process among central actors and deals with the tension between control and autonomy in such strategy work.
The results show that a clinic's development plan work is part of a broader context and that the intentions of the national government in uence the work at the clinical level. Involvement and transparency were highlighted in the national guidelines, and these values permeated the different health authority levels. Hence, the values and ideas identi ed at the highest level of government and have been addressed through the multilevel chain.
The paper identi ed examples of inclusive management, particularly at the clinic level [17]. Clinic managers and the clinical adviser emphasized on involving the whole organization in the meetings connected to the development plan formulation. However, despite this, the managers did not succeed in involving the healthcare professionals in the development plan process.
The study identi es two reasons for di culty with inclusion; a) challenges in managing the contingency of decisions and b) the tension between top-down and bottom-up steering. Communicating a development plan work both inside and outside the organization was challenging for the managers.
Various stakeholders, such as national, regional, and local health authorities, as well as patients and healthcare professionals, activated different functional systems when seeking to understand the development plan, meaning that there was no shared understanding of what a development plan is or should be. In line with [25], there was a gap in how the organization's top management perceived the development plan, and how the section managers and union representatives did. Top management was more comfortable working to meet political and economic goals, while these goals were too abstract and general for the professionals. Despite this gap, the managers assumed the role of brokers and tried to provide a shared experience and transcend the system boundaries between the participants so that the development plan statement process could bring about shared meaning.
The problem was that the managers were not clear on what the decision premises for the development plan should be. In the plan process, several strategies have been identi ed, by which the organization tried to manage the contingency of decisions. The managers tried to displace this contingency to the political system by calling the development plan a political order, and they displaced this contingency to the economic system by referring to the economic frames of the local health authorities. The emphasis on involving the healthcare professionals meant that this contingency was also situated inside the organization and was therefore not managed. As a result, some union representatives and section managers also tried to displace the contingency to the clinic management by urging them to clarify the decision premises for work on the development plan. When the decision premises are not managed, it is di cult to foster involvement, as it is not clear to the actors on what basis they should get involved.
Another reason why involvement was di cult to achieve was the tension between control and autonomy.
The process of making a development plan is part of a bigger political health reform project involving all levels of specialist healthcare. Because the goal of involvement was set by the health authorities, it felt more like an obligation than the result of willing participation in a collaborative setting. The clinic's use of the "line principle" in involving the professionals as well as the various prede ned guidelines enhanced the sense that the development plan work was a top-down process. By taking a multi-level translation approach in this paper, and including an analysis of the higher level of governments work, the ndings identi ed elements of control that are translated down to the clinic level, con rming the control dimension in the policy documents as well. The case studied here was that of an organization under construction.
This could imply that the organization lacked a repertoire of shared resources such as rules and procedures that are important in realizing collaborative processes 18). At the same time, the process of making a development plan could have been an important step for the organization in developing shared resources. As the clinical adviser noted, the process they underwent was more important than the development plan itself. This can indicate that control mechanism can facilitate con ictual collaboration [26]. This might be important in developing shared resources between the different subsystems in a functionally differentiated health care system.
The ndings from this paper comes from a case from Norway, however many countries undergo reform changes in the public sector, which are to be implemented on an organizational level. Making a development plan is often one aspect in structuring organizations that are undergoing changes. The conclusions presented in this paper are of general interest and can be used in discussions with public sector managers working on strategy documents as well as policymakers when identifying challenges that might hinder the implementation of political intentions.

Abbreviations
Not applicable

Declarations
Ethics approval and consent to participate The researchers have gotten written consent from the participant that their a willing to be a part of the interview study.The interview process has been approved by the Norwegian Data Protection Services (NSD). They did not give a speci c motivation why the oral consent was approved. However the procedure is made in accordance to national ethical standards regarding qualitative data methods, approved by the Norwegian authority NSD. The observation data was orally approved on the meetings and a written consent was given from the organisations top management.
Data availability: The datasets used and analyzed during the current study is available from the corresponding author on reasonable request.

Competing interests
No competing interest