The shared principalship: invitation at the top

Abstract

A school principal’s workload is recognised as being heavy, with an imbalance between demands and resources. This paper contributes to the development of collective leadership. The principalship constellations of six schools in Sweden were studied with the aim of strengthening the current knowledge about structures and experiences of shared principalship. The empirical basis is qualitative data from interviews with principals and vice-principals. The analytical focus was on how the sharing structures were organised and how the shared principalship was experienced. The results point to a considerable variation in the organisational structures of shared principalship. Despite the type of model, form and constellation, the principals and vice-principals voiced a striking sense of relief in not feeling alone in their duties, as problems and troubles became manageable. An intensified interaction level in the principalship constellation created opportunities to develop competence. Theoretically, this study broadens the invited leadership concept to include horizontal invitations across unit boundaries between principals in different units within the same school. The knowledge contribution of this study is useful in discussing the legal possibilities for shared principalship, which may be especially relevant in times when the Swedish school system is being criticised for not delivering good student outcomes.

Introduction

The concept of leadership in the research literature is increasingly being described as collective (Bolden, 2011; Denis, Langley, & Sergi, 2012; Gronn, 2015; Harris, 2013; Ulhøi & Müller, 2014). Shared leadership between managers is part of this trend (Denis et al., 2012) – even in school settings (Eckman & Kelber, 2009; Gronn, 2008). The idea of a ‘super-principal’ – an individual carrying the full weight of responsibility of operating and developing a school – belongs to a traditional organisational model (Eckman, 2006). A school principal’s workload is recognised in the literature as being heavy, with an imbalance between demands and resources (Berntson, Wallin, & Härenstam, 2012). The sharing of a principal’s position by two people is described as a way of decreasing the often overwhelming burden (e.g. Court, 2003b). The isolation that can be felt by school principals (Kelchtermans, Piot, & Ballet, 2011), is contrasted with the sense of reassurance and collective responsibility that is felt in shared leadership: ‘coprincipals value not being alone at the top’ (Eckman, 2007, p. 26). Shared principalship is also suggested to increase the practice of democratic principles in schools (Wilhelmson & Döös, 2016b).

The case study featured in this paper originates in the recent changes to the Education Act that lessen the legal opportunities for shared principalship in schools in Sweden (Örnberg, 2016). These changes stand in contrast to findings of the benefits of shared leadership between managers (including school principals) identified in previous research (e.g. Döös, 2015; Rosengren & Bondas, 2010; Wilhelmson, 2006). However, legal openings do exist for shared leadership, and we consider it important to understand more about how principals who share principalship accomplish and experience this. Furthermore, there are knowledge gaps when it comes to identifying structural forms and constellations of shared leadership between managers and, in addition, further knowledge is needed about sharing with clearly divided responsibilities and work tasks. In this paper, we investigate structures and experiences of shared principalship in local schools. We look upon schools as organisations and study the way they operate from action theoretical and experiential learning perspectives, as these give an understanding of how working forms function and develop.

Below, we give a brief introduction to the Swedish school system and recent legal changes governing its principal assignment. We then explain the paper’s theoretical point of departure concerning shared leadership as well as the previous research on shared principalship. Thereafter, we present the aim and research questions, followed by two result sections and the discussion.

A brief introduction to the Swedish school system

Schools in Sweden are publically financed but run both by municipal and private education organisers.¹ To a large extent, the same rules apply to both of them. For all children 6–16 years of age, there are 10 years of compulsory schooling and, after that, three to four years of upper secondary schooling that most children attend, either as preparation for university or vocational training. Swedish schools’ budgets depend on the number of students choosing and attending the school, as each student brings with them a set amount of funding. See Blossing and Söderström (2014) for a description of the Swedish school system.

The current The Education Act (2010:800) came into force in July 2011. It details the responsibilities of school principals, introduces the concept of the ‘school unit’ and stipulates that each school unit shall have only one principal. These changes to the law effectively closed the legal opportunities for joint principalship² in Sweden (Örnberg, 2016). However, the legal prohibition against joint principalship was a side effect rather than being intentional. The law was primarily motivated by the state’s wish to lessen the influence of the local authority, increase the degree of national governance and therefore hand specific responsibilities to principals (Madestam, 2016). Also, the law was proceeded by many years of investigatory work lead by governments with different political views. In the legislative process, it was taken for granted that appointing one person as principal would guarantee clarity over where ultimate responsibility lay (Örnberg, 2016).

The Education Act establishes that each principal makes decisions about his/her unit’s inner organisation, and is responsible for allocating resources within the unit according to the children’s and students’ abilities and needs (2 Chapt. 10 §). By international comparison (Blossing, 2013; Pont, Nusche, & Moorman, 2008), principals in Sweden have a vast amount of autonomy (e.g. concerning budget, employment of teachers and salary setting). Also, Sweden is an example where ‘to lead and not to administer the work in schools’ is an explicit leadership requirement’ (Pont et al., 2008, p. 77). Apart from the requirement of having only one principal per school unit, the state gives autonomy to organisers and principals over how to organise their schools, which results in a diversity of organisational models and leadership forms. For example, large schools can be organised into more than one school unit (usually 2–4) and thus have one principal per unit. While the work of principals is highly regulated in the Education Act, the function of vice-principals is not mentioned. Yet, vice-principals are common in schools in Sweden. When used in this paper, the term vice-principal refers to the title held by a principal’s close colleague/colleagues who are not formally principals but who nevertheless take responsibility as informal equals.

Shared leadership – theoretical point of departure

Collective leadership is an umbrella term for shared responsibility in an organisation and is here thought of as consisting of two distinct but connected subsets: distributed leadership with a spreading of responsibility and power to those not in management positions (e.g. Jones, 2014; Spillane, 2005), and shared leadership between managers (e.g. Döös, 2015), with the latter being the focus of this paper. Leadership within such a managerial team can be said to occur when interactions create direction, co-orientation and space for action (cf. Crevani, Lindgren, & Packendorff, 2010).

In the specific case of shared leadership between managers, a range of different terms are used (Alvarez, Svejenova, & Vives, 2007; Gronn, 1999; Heenan & Bennis, 1999; Järvinen, Ansio, & Houni, 2015; Rosengren & Bondas, 2010). The phenomenon belongs to Denis et al.’s (2012) second stream of research on leadership in the plural: ‘pooling leadership capacities at the top to direct others’ (p. 213). Typical claims of this stream are that ‘pooled leadership bridges expertise and provides legitimacy’, is sustained by ‘role specialization, differentiation, complementarity, and mutual trust’ (p. 216), but is also understood as potentially fragile because of internal competition among co-leaders.

In this paper, we use shared principalship following Döös (2015) conceptualisation of three identified forms of shared leadership between managers within an organisational unit: joint leadership, invited leadership and functionally shared leadership. This conceptualisation is built on an interest in work-based experiential learning processes in organisations (e.g. Döös, Johansson, & Wilhelmson, 2015; Ellström, 2001; Kolb, 1984) and focuses on how tasks and responsibilities are shared between the managers involved. Joint leadership (Wilhelmson, 2006) is understood as complete cooperation, where formal hierarchic equality is in place and work tasks are merged. Invited leadership points to the sub-form of shared leadership, where there is one official manager and a lower ranking partner’s share in decision-making takes place ‘with the permission of the first’ (de Voogt & Hommes, 2007, p. 2) – a vertical invitation thus takes place similar to Heenan and Bennis (1999) co-leadership. Functionally shared leadership is where hierarchical equality exists but managers have separate professional areas and daily tasks (Järvinen et al., 2015; Wilhelmson, Döös, Backström, Bellaagh, & Hanson, 2006). The term functional refers to the principle of structuring the formal organisation around employees’ specific skills (Bratton, 2010).

There are basically two organisational models of sharing between managers: either having one unit with staff, activity and a budget in common (Organisational Model I), or each manager having his/her own unit and cooperating across unit boundaries within a larger organisational context (Organisational Model II) (Döös, 2015; Wilhelmson et al., 2006). The three forms of shared leadership presented above concern shared leadership according to Organisational Model I. Besides form and model, the concept constellation is used in order to fully capture the structures of shared leadership between managers. Following Hodgson, Levinson, and Zaleznik (1965), the term ‘leadership constellation’ refers to the collective leadership group. We use the concept to point to the variation in the number of sharers and their formal hierarchical relations.

Successfully sharing managers draw attention to three qualities that together form the bedrock of sharing: trust, a lack of pretention and common values (Döös, 2015). Common values constitute a foundation for the building of trust, and concern two equally important aspects: the goal and vision for the activity, and how to lead and treat human beings.

In sum, shared leadership between managers is the common taking of responsibility for the tasks that comprise the manager’s assignment, such as administration and management, leadership towards goals and in organising working conditions for others. We understand the sharing of leadership as (a) taking responsibility together for the organisation or organisational part as a whole, and (b) as the mutual and shifting influence between those managers who lead together. In schools, the manager is known as the principal, which is why this paper also uses the term shared principalship.

Previous studies of shared principalship

Shared principalship greatly resembles co-principalship – which is the term used more commonly in the school leadership literature (e.g. Court, 2003b; Eckman, 2007) – and we use their term when examining this stream of research. The co-principalship research field has grown considerably since West (1978) identified the demands and pressures faced by principals and tried to address what he saw as an excessive burden by dividing the task into two. West presented detailed job descriptions for the two separate roles (principal for instruction and principal for administration) with separate responsibilities, and also a list of shared duties where both principals needed to collaborate, such as staff meetings, personnel evaluations and long-term planning. More recently, Flessa (2014) describes a similar way of sharing between a principal and a pedagogical head. Aside from that, the division of responsibilities and work tasks as suggested by West (1978) is not common in the co-principal literature, which tends to focus on merged responsibilities and work tasks.

Different aspects of co-principalship have been empirically investigated over the last three decades, mainly focusing on a variety of potential benefits of co-principalship, such as sustainable working conditions for the principals themselves (Eckman & Kelber, 2010; MacBeath, 2006), advantages for principal recruitment (Brooking, Collins, Court, & O’Neill, 2003; Eckman, 2007), the practice of democracy (Thomson & Blackmore, 2006; Wilhelmson & Döös, 2016b), description of work models (Bunnell, 2008; Gronn & Hamilton, 2004) and work processes (Paynter, 2003). This collaborative form of leadership in schools is praised by researchers, who report findings of successful schools and content principals (Court, 2002, 2007; Eckman & Kelber, 2009; Gronn & Hamilton, 2004; Grubb & Flessa, 2006; Paynter, 2003). A few studies, however, document failure and despair; Eckman (2006) points to difficulties in sharing power and positions that concern ‘problems in communicating, defining responsibilities, developing trust, presenting a unified front, and being ‘played against each other’ by parents, teachers and community members’ (p. 102). Court (2003a, 2003b, 2004b) gives an example of how a co-principal partnership fails, mainly because of difficulties in cooperating within the co-principal partnership, but also due to confronting resistance from the authorities.

The main argument for co-principalship is the demanding work task, which is often impossible for one principal to manage alone (Brooking et al., 2003). The task of principalship is too burdensome and stressful for a single person according to Eckman (2006, 2007), who points to co-principalship as an alternative with the potential to facilitate a high level of job satisfaction and diminish stress. Instead of ‘heroic leadership-individual accountability’, Brooking et al. (2003, p. 154) argue that ‘the emerging alternative paradigm of shared leadership-collective responsibility […] where two or more people share, on a full-time or part-time basis, the position of principal’ should be considered.

The qualities of co-principalship that seem to be crucial are excellent communication skills and an ability to put the ego aside (Eckman, 2006, 2007). A flat management structure built on open and honest communication, the sharing of information, and effective decision-making, improves the conditions for dealing with dilemmas and difficulties (Court, 2002, 2004b). Court (2004a) points to the importance of this kind of open and ongoing communication for the development of the shared meanings and understandings that serve as a foundation for successful collaboration in schools.

Finally, there is little research that explicitly approaches shared principalship by identifying the importance of vice-principals. When Lee, Kwan, and Walker (2009) summarise previous studies about vice-principals, they describe ‘a misalignment between the time spent on various duties and the perceived importance of these roles to school success’. (p. 190). However, Cranston, Tromans, and Reugebrink (2004), who previously claimed that vice-principals were the ‘forgotten leaders’ (p. 225), found in an empirical study that it was significant for their level of satisfaction ‘how well the notion of team among school administration team members (e.g. principal, deputy principals) was developed’ (p. 233). Also, Abrahamsen (2017) examines a redesigned vice-principal role in Norway with expanded responsibility for performing instructional leadership, which also affected the relationship to the principal.

Aim and research questions

The aim of this paper is to strengthen the current knowledge about the structures and experiences of shared principalship – a specific type of shared leadership between managers. Three research questions are posed:

(1)	How can the structures (models, forms and constellations) of shared principalship be described?
(2)	How is the shared principalship collaboration experienced by principals and vice-principals?
(3)	How can structures and experiences of sharing principalship be understood?

Method

The leadership constellations of six schools were studied, using qualitative methods. The choice of schools, as well as the data collection and analysis methods used, are presented below.

Choice of schools

The choice of schools required that the school principal shared leadership, and that we were informed about this by a principal or a vice-principal at the school, or by local school organiser’s administration. This can be assumed to mean a positive selection, in that the informant saw a value in their sharing leadership. A risk of a negative selection also existed, in that principals working particularly closely across unit boundaries would not make themselves known for fear of being recognised and stopped from continuing their collaboration by the organiser or the Swedish Schools Inspectorate (Wilhelmson & Döös, 2016a). We strived for variation in schools, according to whether they consisted of one or several school units and whether they were publicly or privately run. Both compulsory and upper secondary schools were chosen and schools with differences in applicants per place and students’ socio-economic background.

Data collection and analysis

The data were collected between 2014 and 2016, in four compulsory schools³ and two upper secondary schools where the principals said they shared their principalship with one or more others. The data consist of recorded semi-structured interviews with nine principals, one head of unit and eight vice-principals. They had worked as school leaders for between 3 and 21 years; for some, it was their first principal assignment.

There is a total of 21 interviews. Six group interviews with the school’s leadership constellation, and 15 individual interviews were done. Interview guides were used and focused on how the school and the sharing was organised and specifically done in terms of work-task division within the leadership constellations, and also their experiences of this. Paralleling Argyris (1990) words, we interviewed about the work-in-use, not the espoused work. The interviews were transcribed.

The analytical focus was on how the sharing structures were organised and how the principals and vice-principals experienced their shared principalship. The analysis was a continuous process that took place in the research team both during data collection and afterwards. In the final stage, the interview material was analysed by the first and second author according to a developed thematic structure. When analysing the experiences, we kept in mind the occupational categories and organisational models. This provided a possibility for identifying relevant aspects in order to add to and strengthen current knowledge about shared principalship. Attention was paid to both similarities and variations in the material.

This study does not, according to the Swedish Ethical Review Act (2003:460), handle sensitive personal information, so approval according to Swedish law is not required. Actions to secure ethics have, however, been taken. Participating organisations and individuals are kept anonymous. The presentation of the results is based on all interviewed people and all are cited, except one who was not interviewed individually. Each quote is followed with a letter indicating whether it was a principal (P), vice-principal (VP) or head of unit (HU) who spoke. It was common for interviewees to use the word ‘principal’, which also included the vice-principals. We have not changed their wording in this respect as it reflects the informality that existed. They also called each other by their first names, which has been replaced in the quotes with an X (and a Y if a second person is mentioned) followed by a square bracket indicating whether it refers to a principal, vice-principal or head of unit.

Variations in structure and division principles of shared principalship

In this first result section, the variation in the structures of shared leadership in the six schools (A–F) is displayed. Three concepts – introduced in the theory section – are used to describe the shared principalship structures of the schools: organisational model, form and constellation of shared principalship (see Table 1). All nine principals in the studied schools had chosen to collaborate closely with one or more others in sharing their principal function, thus forming the sharing constellation in each school’s leadership. The invited leadership form (see theory section) is here extended to include sharing across unit boundaries and not only within one unit. Therefore, the concept vertically invited is used within a unit to illustrate shared principalship between principal and vice-principal (in Organisational Model I) and the concept horizontally invited is used to illustrate shared principalship between principals in different school units (Organisational Model II).

Table 1. The structures (i.e. the models, forms and constellations) of sharing principalship in the six schools of the study.

Organisational model	Organisational Model I: Sharing within one school unit				Organisational Model II: Sharing across school unit boundaries
School	A	B	C	D	E	F
	Upper secondary	Compulsory	Compulsory	Compulsory	Upper secondary	Compulsory
Form of sharing	Vertically invited	Vertically invited	Vertically invited	Functionally shared and vertically invited	Horizontally invited	Horizontally invited
Constellation of shared principalship

Note:

Table symbols: Principal: Vice-principal: Head of unit:

The principals of four schools (A–D) shared within one school unit, and thus worked according to Organisational Model I. In two schools (E–F), the principals shared across school unit boundaries in working according to Organisational Model II. The main form of sharing within Model I was in three schools (A–C) with vertically invited leadership, and in one (D) it was functionally shared. The invitation in Model I was made by the principal to one or more vice-principals. Also, in the functionally shared leadership of school D, there was a vertical invitation from the principal to the vice-principal and, in addition, a mutual horizontal invitation between the principal and the head of unit who were equals in terms of organisational hierarchy. School D was an example of where the school administration tried a new organisational solution to solve problems concerning school results and the principals’ workload and health. A new managerial position, head of unit, had been introduced and was accountable for the municipal tasks, whereas the tasks regulated by the Education Act remained with the principal. The solution aimed to lift the burden of administrative and budget issues from the principal and so free time for the principal to fulfil the Education Act’s requirement of pedagogic leadership. The term head of unit was chosen by the local responsible authority to circumvent the legal hindrance of having more than one person appointed as principal.⁴ In addition to the ongoing sharing, school C was an example of a lost joint principalship due to the restrictions imposed by the Education Act. Originally, this school had two principals who invited development leaders to share their principalship. The development leaders were later appointed as vice-principals. Because of the Education Act, the principals were forced to end their joint principalship and one of them remained as the single principal.

As can be seen from Table 1, the shared principal constellations comprise duos (B, E), trios (A, D, F) and a quintet (C) – some of which take place with equal hierarchical lines of division (D–F) and some with vertical formal hierarchical lines of division (A–D).

So, how is the inner organisation of these schools structured? The division at hand in the six schools varies markedly; the freedom in how to organise schools is apparent and the division has been organised according to different basic principles, both at the local school level and at the organiser level. Firstly, two overarching types of division existed that were made at the organiser level: the division of a school into different organisational units (E, F) and the hierarchically equal function of head of unit (D). These lines of division have a formal sharpness that is not present when a principal decides how to structure a school unit’s inner organisation. That said, we found three main bases for the division of the inner organisation in terms of the shared principalship of the six schools. They were either mainly organised:

(1)	by levels, school years or programmes (C, E, F),
(2)	by functions of school leaders (D), or
(3)	by interests and competence of principal and vice-principals (A, B).

Organising a school along the lines of levels, school years or programmes (1) is traditional and common. In the three examples, this meant that the principals and vice-principals to a large extent performed the same type of work tasks. The division by function (2) of school leaders was a kind of organisational invention on the part of municipal organiser, i.e. the local school authority. Work tasks were divided. The principal was ultimately responsible in accordance with the Education Act, while the head of unit was responsible to the local school authority. The division by interest and competence of principal and vice-principal (3) was organised according to the individual preference of each principal and her/his vice-principal(s). Within a school unit, this, for example, meant that one vice-principal could take care of student health issues, while another took responsibility for timetabling. In sum, all six shared principalships were clearly divided in terms of both work tasks and responsibilities, and always with the formal correctness of one principal for each school unit who was ultimately accountable for all decisions made.

Experiences of the shared principalship

This second result section details how the principals and vice-principals, identified as the school’s principalship constellation, experienced their collaboration. It describes what they did and how they thought about it. The overall pattern is that they highly appreciated their sharing as it made an otherwise burdensome assignment manageable. Principals and vice-principals both emphasised that they valued not feeling alone and having sounding boards. There are some specific aspects of this generally positive experience that either strengthen or add to previous knowledge. Six subsections will detail such experiences of the constellations: (1) Informal interaction coordinates the divided, which deals with the coordination of what was in the first result section described as divided. (2) Collaboration against a demanding background details some of the reasons that motivate the sharing of responsibility and influence. (3) The easiness of a shared physical space concerns a physical condition that facilitates sharing. (4) Forming a solid platform continues the line of prerequisites but with a relational view. (5) To be different is useful highlights how the sharing principals and vice-principals thought about their differences in competence and personal characteristics. (6) Getting important things done sets out a number of advantages connected with their sharing of principalship.

Informal interaction coordinates the divided

As set out earlier, there are organisational divisions between the people in the principalship constellations. The potentially negative consequences of these divisions were counteracted through coordination by interaction and involvement among the sharers. The coordinating access to close colleagues was highly appreciated by the sharers. It included daily informal interaction, scheduled work meetings and helping each other in specific matters.

Continuous daily contact was vital in creating the coordination and involvement of each other to handle the stream of issues that needed to be taken care of. This involvement was largely informal and occurred naturally during a workday. It was mostly about bandying ideas around, and discussing problems, solutions, and decisions. They would think and talk a lot together, but when acting they were mainly apart. They regarded the continuous flow of informal communication and checking as tight and well functioning, and it occurred regardless of organisational model or shared leadership form. Involvement in each other’s tasks was explained as necessary if they were to lead successfully together.

Then it happens quite naturally, because it’s not something we put on our schedules. We meet several times a day and discuss all the things that need to be done. How do we best present this or that, in a way that is clear to staff. We have that kind of a discussion ahead of all activities. (VP)

There were several concrete examples of how involvement occurred, such as the salary-setting process where the principalship constellation prepared and made decisions together, and in between they acted apart in holding salary meetings with their own subordinates individually. In that way, they acted individually but it was grounded in a common plan.

We work together on staff performance reviews and setting salaries. X [vice-principal] has performance and salary review discussions with those who report to him/her, I have discussions with those who report to me, and Y [principal] has discussions with those who report to him/her. This adds up to a collective picture. The three of us get a shared picture, and on that basis we can set the salaries of all the staff of the school. (VP)

Also, the functional division of duties and work tasks between principal and a head of unit was perceived as well functioning. It was essential for them to think together to produce a shared picture of where to take the school. An important task for the head of unit was to create structures and conditions that would facilitate the principal’s legal assignment, especially concerning pedagogic leadership. In launching this new leadership solution, it became vital to hone the division of responsibilities, even in the middle of the day-to-day goings-on. For example, when confronted by new problems where they had to decide who would own the problem:

It’s essential to work closely as a team. There was one occasion when we needed to report an incident to the police, and the question was, ‘Who should do that?’ We discussed it, and concluded that in this case, which involved a student hitting another student … it should be the principal. But if on the other hand, it had been a case of damage inflicted on the school, I would be the one to step in and report it. (HU)

Some schools had regular leadership meetings in addition to the daily interaction. These meetings were described as vital forums, where a large number of issues are dealt with. This was where problems were identified and understood, tasks coordinated and responsibility allocated. Keeping the school together and leading in a common direction required striving towards joint understanding.

There is a huge range of different questions, big and small. There is so much we need to talk about. We process a lot. Sometimes we let a week pass and we all reflect and then we see things a bit differently … we are the ones who need to steer this ship together. We need to talk until we reach a common position, we need to be in agreement, we need to be able to respond in a fairly like-minded way. It wouldn’t work if we didn’t have these leadership team meetings. (P)

Within the principalship constellations, they helped each other in several concrete and practical ways – even between school units. For example, budget costs were moved between the units, and one principal explained IT and budgetary matters to another principal. The support could also concern difficult staff conversations, managing when a principal colleague was ill by, for example, responding to calls from parents, or giving each other staff resources in a crisis situation.

And you recruited a staff member that was supposed to work for you, and then I had a problem because of a late notification of sick leave, and you just handed the new staff member over to me in that huge crisis. I’m madly grateful for things like that. (P)

An important aspect of the principalship collaboration was to make the collaboration visible to the staff, else there could be staff unrest. It was therefore important for each principalship constellation to send signals to the staff that the principal constellation was cooperating, and that they were on the same level of authority. It was also important for the sharing constellation to be physically seen together.

The collaboration between X [principal] and me and Y [vice-principal] sends an important signal to others, because I’ve heard that they see us as a happy team that work well together, that check in with one another, and this creates a feeling of security. (HU)

The unit-divided schools had a formal structure that kept the units apart. Therefore, it was particularly important that they as principals acted in the same way to prevent discord between units. That the unity of one school was largely dependent on the principals’ informal collaboration was described as a weakness in the long-run if, for example, one of them was to leave and be replaced by a new person. When a school is not only divided into units but is also located in a difficult socio-economic area with many consequent problems for the principals to handle, then the time for coordination becomes even more important, but also more difficult to find.

We ought to have a sacred time set aside for reflection. But the time that we should have had has been taken up by so many staff issues and student issues this autumn that we … and if we couldn’t meet face to face, we called each other instead. (P)

Collaboration against a demanding background

There emerges from the data an essential reason behind the sharing of leadership that points to the heavy demands associated with the Swedish principal assignment: the collaboration of shared principalship makes the impossible possible. The sharing was portrayed as a lightening of the burden from duties that were wide-ranging and demanding. A sharp difference was experienced between the potential loneliness lurking around the corner and the appreciated collaboration.

A recurring pattern in these schools was that the principals and vice-principals strongly, and in a number of ways, expressed that the existence of close colleagues brought about joy and security. This was communicated as general expressions that both reflected the importance of leading as a team and the easiness it provided. Job satisfaction, a healthy work situation and not being alone seemed to be the main driving forces for collaboration as sharing principals. Some examples of this recurring pattern follow below, reflecting two vice-principals’ and a principal’s sense of relief:

A problem is never MY problem, it’s OUR problem. That’s somehow the big thing. And when you know it’s our problem, then the feeling that it’s a problem isn’t so overwhelming from the start, it’s more like here are a few issues we need to address. (VP)

I never feel alone. I think that’s an important part of how you see your workload. (VP)

It’s all those millions of tasks you don’t need to keep track of yourself. (P)

The collaboration offered strength when the task was tough and all involved principals and vice-principals found that sharing relieved workload and feelings of isolation. Yet, the sharing solution was not always enough and the assignment could sometimes still be too tough to manage, even putting the principals’ own health at risk. In one school, a vice-principal described the overwhelming workload and a resulting pressure on the chest. In addition, the vice-principal also worried about the principal’s health, as it was burdensome for the principal to be ultimately responsible for so many issues. These more problematic sides were primarily associated with schools located in areas with low socio-economic status, which influenced both the school’s economic resources (as there was more unplanned influx and outflow of students) and the social climate as there was a large proportion of immigrant families with a traumatic background. There could thus be serious worries about the risk of teachers leaving a burdensome work situation and about student unrest in and around the school.

This autumn has been pretty rough. We’ve had a lot of tough student issues where I had to step in a lot, I mean not just managing the students but also frustrated teachers who need support, and then you find yourself completely exposed in the middle of all this and you have to be this strong, secure person. (P)

Obviously I have problems sleeping and wake up at night and so on. I don’t really know how to handle it, and I feel like I can’t really do anything about it either. (P)

The traditional principal assignment was described as sprawling, in that it included too many and diverse tasks, and one principal emphasised that it was necessary to be allowed to concentrate, which is something that definitely occurred more as a result of their new functionally shared leadership. Schools demand a stable principalship for the sake of the students. Therefore, dependence on one single person was described as dangerous. Some expressed serious concern about the task of principals in the Swedish school system, and that they should be better supported by society so normal working hours could be kept and focus could be maintained on teaching quality and school development, instead of overloading one person and thereby risking the stability of the school.

Schools are important organisations that have to be very stable because we’re dealing with so many small children. We can’t have things fall apart because of one person, whether it’s one person’s lack of performance or one person disappearing. (P)

Furthermore, one principal compared their current situation with how it used to be when they worked in a joint principalship. The principal described that it was difficult to have lost the closest colleague because of the new regulation. The colleague’s chair stood empty during the leadership team meetings for a long time, and the principal remained for a long time in an old tiny room instead of moving into the big one where the colleague used to sit. The loneliness felt by this former jointly sharing principal was described as being related to the loss of the colleague, or rather to the knowledge about what could exist.

I think it’s primarily because I’ve been a sharing principal in the past. I know how it used to be. I could not wish for a better leadership team than the one I’ve got, but … I can never duck nowadays, the way I could then, because then I could say ‘X [principal], maybe you could handle this issue?’. But the way it is now, if someone here comes and needs my help, or I sense that there’s a problem somewhere … then I can’t duck. I’ve known what it was like to be two principals, and that’s why there’s a difference. (P)

The easiness of a shared physical space

A recurring theme identified in our interviews concerns the positive aspects of being located in a shared physical space. These physical workspace solutions, however, were results of practicalities and not of intended organisation or design. With one exception, the sharing constellations either had their offices adjacent to each other with an inner door between, were sat in the same room, or were very close and with easy access. This easy access to one another was highlighted in the interviews and enabled solutions to be found to many small matters during the working day. Their spatial arrangements facilitated being in joint control. Smoothness and security thus existed despite the otherwise demanding principal task.

Our rooms are next to each other and usually the door is open, so we use the board in X’s [principal] room so we don’t miss anything. It feels like we rarely miss things now. (P)

One constellation trio shared the same room into which they had squeezed their three desks, a tiny round table for leadership team meetings, and a projector in the ceiling to project documents being used during meetings. This solution signalled both cooperation and easy communication.

All three of us sit in the same room. This brings us very close physically. I sit between the two of them. This makes us into a team that can bounce most questions off one another in a very quick and simple way, which is a big advantage. (VP)

In another school, the principal and vice-principal had their offices close to each other and could be seen by students and staff passing in and out of the school or visiting the cafeteria; they were seen sitting and talking in the principal’s office. The vice-principal was confident that staff knew about their tight collaboration, and referred to a staff survey where this was made known.

The whole staff can see our teamwork. They see us as one entity and they see our teamwork. It’s visible. And the things we want to be visible are indeed visible, which is great. The whole staff have made that known. (VP)

In a school that was divided into school units, the principals chose to have their offices co-located. The reason for this was to counteract the organisational division by school levels, but it was at the expense of not sitting close to their own staff and students, since the school consisted of several buildings.

There’s always a bunch of minor things, so this means we can talk to each other about those directly and then there is less us and them [between the school levels]. If the ‘us and them’ talk were to catch on, I think there would be deep divisions. (P)

In one school, the leadership team discussed office-sharing as something that could relieve pressure on the principal. They compared their current situation to a prior period in which they had joint principalship at the school and when the joint principals used to sit in adjacent rooms. Two vice-principals, who at the time of the interview shared a room, found their room-sharing valuable and suggested that the principal would be relieved if the principal were to share a room with one of the vice-principals.

I think the physical aspect is actually quite important. You used to sit in separate rooms with an open door in between. You were able to close the door on the difficult world outside, but you could talk to each other. (VP)

By the same token we share a room, we don’t sit and talk to each other all day, but we do meet and sometimes we look up and talk. This daily flow. I don’t think we should under-estimate it. … If you [turns to the principal] had one of us sitting here, maybe it could partly fill that role, even if formally there wouldn’t be two principals. (VP)

Forming a solid platform

The bedrock of successful sharing, identified in previous research, was also expressed in these leadership constellations and, taken together, it tells us a lot about the importance of good relationships; a lack of pretention, trust in one another and common values were described as essential qualities for collaboration to take place, and as a reason why it worked. They were not described as separate qualities, but rather that the qualities interacted and manifested in several ways and areas.

Lack of pretention means to share all information and not to compete against each other. Vice-principals thus do not aspire to the title or task of the principal, and the principals do not hold on to tasks and power. Thus, vice-principals stated that they were not interested in taking command and to be the figurehead; instead, they were content with a formally subordinate position. Lack of pretention was present in the everyday work, but could also be reflected in how the school was divided.

In terms of pride, it doesn’t matter who’s taken the decision, the important thing is that we’ve taken a decision that is the best possible one for the school. (P)

We’re very unassuming towards each other and we can reveal ourselves to each other, and X can bring her side of the equation and that gives me strength. (VP)

Trust is a quality that is developed with time and creates, for example, a feeling of security that important tasks are being performed well. It originates from competence as well as from liking each other as people. Trust comes out of knowing that the other is competent, that he or she knows the school system, and understands the other’s way of reasoning – all this makes it possible for a principal to let go of work tasks without being afraid of how these will be performed.

Now it’s me who has the formal responsibility for student health issues and X [principal] has entrusted me with the task of making it very independent. If there is a verdict against us over something we didn’t handle correctly, it’ll be X who’s standing there. This requires a lot of trust between us. (VP)

Trust is needed, and may at times also be put to the test when there is a difference in views. In one example, conflicting views that potentially could threaten the bedrock could be dealt with because of the trust that emanated from working together for a long time.

Common values were described in the interviews as the ground for mutual trust and confidence in each other. This was an overall pattern that showed in recruitment, in formal as well as informal collaborations, and which permeated the everyday work and the decisions made.

It’s based on equal values and similar ways of dealing with staff. (P)

A concretising example of this was when a vice-principal and principal both emphasised their common values and outlook on people, and described how these permeate all their conversations and decisions. Their common values brought about trust and an accompanying security.

We have very similar values, a shared view of humanity, which we actually bring to all decisions, all the discussions we have. (VP)

We have the same basic values in how we see our students and what we think is important to them, how we see our staff group, what we want and how we respond to staff in various situations, when there are problems, when there is joy, when there is sadness or whatever it may be. And this has resulted in us having a great deal of trust in one another, and when one of us is absent, the other can easily answer questions from the staff group regardless of what it may be about, because we are quite confident about how the other would have answered. (P)

To be different is useful

In addition to the essential bedrock qualities – lack of pretention, trust and shared values – it was common for the sharing principals to characterise themselves as being different. To have different qualities within the leadership, constellation was found to be enriching by the sharers. The differences were mainly of two types: personal characteristics and competence areas. They described being different and complementing each other as being important for successful collaboration. They also thought that access to each other’s capabilities and knowledge made them better as individuals and more stable as a constellation, and that broader competence was generally of benefit to the organisation.

I’m quite a go-getter and want things fast. X [vice-principal] is reflective and analytical and a highly skilled mathematician, who can see patterns in the schedule and things that I’m seriously bad at, and I happily hand this over to X. With Y [vice-principal], there are reasons why Y specifically works on student health issues, Y is very people-focused that way, the soft values are truly Y’s big strength, it’s one of my strengths too, but Y can work better on it over time. I wouldn’t enjoy digging into it like that, I’d rather have more overarching responsibility. (P)

Differences in personal characteristics were described by the sharers as positive; in one interview, the sharers described themselves as one being proactive and another reflective, one structured and the other visionary, one liking to be at the centre and the other wanting to act behind the scenes.

That’s also our strength of course, that we complement each other very well. X [principal] sort of needs a bit of peace and quiet. So maybe I can provide that […] And vice versa. (P)

X [principal] appreciates that I am quite direct, I’m extremely proactive while X is a bit… but like ‘No, now we have to think it over’. And through that, I can go for it, but at the same time I have to remember to be in listening mode as well. (P)

The principals and vice-principals also learned from each other and developed new competencies. In some cases, they said that a learning process had taken place when differences were a bit difficult to handle and that they had learned how to make use of their different strengths. Below follow two descriptions of the differences and of the process that was necessary in order to establish what became a truly appreciated collaboration.

The three of us are very different, and it hasn’t always been straightforward to work together and build a team, we had to work at it […] it has evolved quite naturally over time. It started with our respective areas of expertise. X [principal] who is very good at organising and finance. Y [principal] who has an innate leadership that’s a bit stronger, a natural leadership that I don’t have. And I have very deep expertise about schools and I’m a wiz at organising things and thinking things through and making sure. So there’s a bit of us using each other […] So somehow I think it was from the starting point of our respective strengths that we found a common basis. (P)

We give and take because we’re good at different things, I think that’s a prerequisite for everything to work well, being different from one other, and that we don’t dwell on factors that may be annoying, but rather on being different and that this is a good thing. (P)

It was, however, mentioned that the differences between the people in the leadership constellation that can be fruitful are not always seen as such by others. According to the principals, some teachers, for example, felt that such differences may actually intensify the division between a school’s units.

Getting important things done

The principals and vice-principals experienced a number of advantages connected with their sharing of principalship. In addition to the relief felt for themselves, and the advantages described above, they regarded their collaboration as beneficial for the whole school in that they got important things done together. When working as sharing principals, they found more room for achieving quality, and it became easier to correctly prioritise work, to be proactive and to solve crises.

In five of the six schools, they described experience of handling serious troubles and found their shared principalship to be a unifying force. One example was when principals, vice-principals and special teachers laid aside their regular duties for a couple of weeks. With daily coordination, they were present around the school, tackling conflict with troublemaking students, calling parents, being present in the dining hall, and so on. They called themselves the Task Force. The initiative created peace and prevented teachers and students from leaving the school.

We dropped everything else we were doing and just stayed out there in those levels. We were in on the teaching, we were in on the breaks, we held a lot of talks with the parents. I spent lots of time in the dining hall, just stayed there, welcomed all the students when they came and like that, yes, we rebuilt a calm school environment. And we managed to turn things around. (P)

The principal and vice-principals in two different schools explained how they constantly supervised what was going on in the school, and how they were prepared to act on any problem that may arise.

The big dividend when we work together is still ... when we can feel our way with our tentacles out together, and as soon as we start to feel that something is going on, some sort of tension, we can focus on that at once and tacticise for ‘How should we address this?’, and then we address it at once, so it somehow rarely gets traction. (VP)

We’re remarkably successful at capturing what’s going on and having a plan and an idea about what we want to do about something before it turns into a conflict or a problem in the house. (P) If a colleague isn’t doing well or we need to do something before something happens, I can usually spot that. (VP)

The united leadership constellation created a sense of stability and security for the staff, and decisions were considered and strong. Also, they described how their leadership created clarity and confidence.

A lot of people mention that we’re very clear in what we say, and if we say we’re going to do something we do it. Several people have brought this up in evaluations, and directly with us as well, so there’s a very strong feeling of security in the staff group today. (VP) They felt that they could turn to either one of us, that we were well coordinated […] this is something that has evolved very of much of itself, organically you might say. (P)

Sharing principals may use their collective power and responsibility to act strategically towards the teachers, to achieve the goal of the school. To be outwardly united gives a lot of strength to the decisions made by the leadership constellation.

We may know that the organisation [i.e. the organiser] has decided to go in a particular direction, but how should we work to implement that here in this school to ensure it is received positively and how do we roll it out? Or should we do it in some other way to make people feel ‘we were part of the process’? Are there a few key people who we actually feel that on these issues they would be better if they got into these issues and then they got to present it somehow. So we can work on strategies like that, and I think we’re relatively successful at it. (VP)

Finally, the sharers felt that the new functional sharing in one school brought about higher quality and made it possible for the principals to work considerably more with pedagogic leadership issues, which was in line with the intention with the model. Of considerable importance for the principal, in this example, was to be allowed to concentrate. The collaboration meant it was no longer necessary to jump in and out of highly diverse tasks and responsibilities, that the focus could be kept, and that someone else could hold the economic perspective. In turn, the head of unit emphasised the benefit of having the authority to make final decisions without having to wait for, or disturb, the principal.

I can take a decision that ‘Now we’re going to do this.’ I don’t need to wait for approval from X [principal]. … you need to have a mandate to put your foot down, even if we think differently. (HU)

The upside is that you manage to achieve something of high quality and that the size of the task is reasonable. That’s absolutely the advantage. (P)

Now we have a pedagogical mission that we really can get our teeth into. [The teachers] see me around much more. That’s something they noticed at once. (VP)

Discussion

Shared leadership between managers demonstrates an interactive way of organising leadership within a managerial position. In this study, the principalship constellations of six schools in Sweden were studied with the aim of strengthening the current knowledge about structures and experiences of shared principalship – a specific type of shared leadership between managers.

The results point to a considerable variation in structures (i.e. in organisational models, forms of sharing and constellations of shared principalship), with accompanying differences in work task division. The sharing constellations varied in numbers between two and six sharing principals and vice-principals. Overall, and despite the model, form and constellation of sharing, the principals and vice-principals gave voice to a striking sense of relief in not feeling alone in their duties; their problems and troubles became manageable, their ideas and decisions better developed. The significance of having close access to each other was stressed, as was the importance of the bedrock qualities. They appreciated the personal and competence differences within the constellations, and experienced more room for achieving quality, being proactive and solving crises. Yet, the sharing was not always enough to bring about a healthy work situation. So, how then can the structures and experiences of shared principalship be understood in relation to previous literature? Below, the diversity in structures and work task divisions of shared principalships are discussed, as are the overall positive findings concerning sharing principalship of those taking part.

A structural diversity and accompanying work task divisions

While plurality in leadership is often associated with conceptions of democracy, empowerment and participation among members of a team who mutually lead each other, this plurality is, in some situations, ‘clearly more structurally embedded’ in the category of pooled leadership in order to lead others (Denis et al., 2012, p. 231). To Denis et al.’s category, this study contributes an understanding of the complexity in its structural diversity. The shared principalship structures in this study were either characterised by formal equality or formal subordination. Cases of formal equality existed either as boundary crossing or within an organisational unit. When crossing organisational boundaries, we deal with a mutual invitation between equals in the organisation hierarchy, conceptualised as horizontally invited leadership. Formal equality also existed at the top of one and the same organisational unit – functionally divided leadership, and in addition, below the principal level as joint leadership between vice-principals. The cases of formal subordination, by definition, occurred within an organisational unit – vertically invited leadership. See Figure 1, which attempts to clarify the conceptualisation of shared leadership forms.

Figure 1. Conceptualisation of four shared leadership forms in relation to formal position hierarchy and organisational boundaries.

In terms of work tasks and responsibilities, and despite the diversity in how the schools and their leadership constellations were organised, the principals in the study highlighted many advantages for themselves and their work situation in shared leadership. Here, the The Education Act (2010:800) defined, through the requirement that one school unit can have only one principal, organisational conditions that meant restrictions in how work tasks and responsibilities could be distributed inside the leadership constellations. In vertically invited leadership within one school unit, the principal and the vice-principal(s) could utilise their individual competence and interests to divide work tasks and responsibilities between them, and at the same time have the overall leadership of the school as an important shared task. And a full merging of tasks and responsibilities could take place between vice-principals in a joint vice-principalship. Instead, in the equality of functionally shared leadership within one and the same school unit, a clear division of tasks and responsibilities was the required arrangement. In horizontally invited leadership between school units, two or more principals could stabilise and support each other to a certain limit, as all principal duties had to be carried out by each individual. Together, they could to some extent strive to bring the units together to develop the school as a whole and to make teachers cooperate. In both horizontally and vertically invited leadership, the formal requirement of division was to some extent revoked. In so doing, the principals and vice-principals have walked a tight rope to balance the practical and legal considerations.

Distributed leadership is in several countries described as a model for improved system performance and in Scandinavia also associated with principles of democratic education (Harris, 2013). However, the relation between shared leadership between principals and the wider distribution of leadership and democratic practice is seldom focused. Yet, the qualities in sharing principals’ interactive ways of working serve as a model of democratic practice with the potential for wider distribution. Being two (or more) at the top also means extended possibilities to take part in and influence ongoing everyday work interactions in the school. This is of importance to the quality of the school’s competence networks (Wilhelmson & Döös, 2016b). In accordance with Spillane (2005) and Brooking et al. (2003), this study points to the possibility of reconsidering the ‘heroic leadership-individual accountability’ (p. 154) along the lines of an alternative paradigm where two or more people share the principal position. The Education Act is built upon the notion that overall responsibility is an individual undertaking, but, in the schools studied here, responsibility is definitely a collective undertaking.

Voicing the strengths of shared principalship

This study supports the findings of previous co-principal research that recounts principals’ strongly positive experiences of sharing their leadership – usually in a pair constellation, thus being in line with Court (2003b), Eckman and Kelber (Eckman, 2007; Eckman & Kelber, 2010), Gronn and Hamilton (2004), and MacBeath (2006) and others. Also, the everyday informal work processes (Bunnell, 2008, 2015; Paynter, 2003) are highly recognisable from earlier studies, as are the qualities needed to build successful managerial sharing (e.g. Döös, 2015; Eckman, 2006, 2007).

Furthermore, the relief felt by principals and vice-principals in this study points to shared principalship as a partial solution to the overly taxing and unhealthy work situation of school principals as found in previous research (e.g. Berntson et al., 2012). The results of this study make it possible to identify shared leadership as an important health and well-being resource. According to the Job Demands–Resources model (Demerouti & Bakker, 2011), in work environment research, resources are thought to buffer or balance stressing demands at work (Wallin, Pousette, & Dellve, 2013). It is evident that the sharing of principalship functioned as a stress-reducing resource, making the assignment possible to manage. However, the results also indicate that numbers may be a problem. When a constellation consists of many vice-principals and only one principal, the workload in one school was excessive for the principal. Gronn (2008) states that numbers count, and so it seems in a negative sense in this study. A study of multi-shared leadership in intensive healthcare also indicated that nurse managers who shared leadership between four and five people found it taxing to handle the constellation’s internal relationships (Döös, Vinell, & von Knorring, acc.). In the specific case in this study, it is not possible to know with certainty if the principal’s partial loneliness and tough workload was related to numbers. Other contributing factors were the socio-economic conditions in the area where the school was located, but also the fact that this principal was able to compare the situation with a previous situation of formal joint leadership before the current Education Act took effect.

Leadership within the constellations can be understood as occurring when interactions create what Crevani et al. (2010) term as direction, co-orientation and space for action. There is an ongoing and valuable discussion between the sharing principals and vice-principals. In this interaction, they influence each other reciprocally, adjust their understanding and form ideas about how to influence the different kinds of school activity. This view of the importance of managers’ daily work is also highlighted in the leadership research (e.g. Alvesson & Sveningsson, 2003; Holmberg & Tyrstrup, 2010; Staunaes, Juelskjaer, & Knudsen, 2009).

Why, then, is the overall picture painted in this study so positive? Most likely because the voices heard in this study come from leaders who talk about themselves and their own ways of organising leadership. They work in this way as they, as sharing constellations, appreciate it and find it relieving and generally beneficial. It remains to study how people outside the leadership constellation (e.g. teachers or teacher team leaders) experience shared principalships. Also, a study of other organisations could have revealed more criticism of the idea of shared leadership. Cases that report problematic sharing are another suggestion for future research, as they would help us to grasp whether there are specific circumstances that facilitate or make sharing of leadership difficult. This study contributes knowledge about how more divided types of shared principalship function and succeed, while previous studies have shown the potential of joint principalship (Wilhelmson, 2006; Wilhelmson & Döös, 2016a, 2016b).

Conclusion

This paper gives an understanding of the potential of collective leadership, which is an essential counterweight in the midst of a climate that believes in heroic singularity as a prerequisite for clarity in schools (Madestam, 2016; Örnberg, 2016). The study contributes to the development of the understanding of collective leadership, which emphasises the importance of a solid relational and value-based platform, the division and coordination of work tasks, the advantage of a shared physical space to reduce workload and counter isolation among principals and vice-principals. In addition, the concept of invited leadership has been somewhat adjusted and now encompasses invitation both within one organisational unit (vertical) and across organisational unit boundaries (horizontal). Vertical invitation occurs between an appointed manager and a formally subordinate person within the unit. Horizontal invitation takes place as a mutual invitation process between two or more appointed managers at the same level, thus the invitation mostly crosses the organisational boundary. The knowledge of the different forms of shared leadership is useful in order to discuss the legal possibilities for shared principalships in schools. With a general view of leadership as a social process that has a cultural context (Hoy & Miskel, 2008), a potential topic for future research would be to study the surrounding conditions for shared principalship in terms of both hindrances and support. It could also be of value for the development of knowledge about what type of management organisation to support and develop in schools – knowledge that is essential for politicians at all levels, as well as school authorities and organisers, in making decisions concerning the governance of schools in the future. This way of organising leadership functions may be especially relevant in times when the Swedish school system is being criticised for not delivering good student outcomes.

Notes on contributors

Marianne Döös is a professor in Education and a researcher within Organisational Pedagogics at Stockholm University since 2008. Her research deals with the processes of experiential learning in contemporary work settings, on individual, collective and organisational levels. Topical issues concern interaction as carrier of competence in relations, shared and joint leadership, conditions for competence, organisational learning and change.

Lena Wilhelmson is an associate professor in Education and a researcher within Organisational Pedagogics at Stockholm University since 2008. Studies have been conducted concerning: dialogue and learning, learning processes in organisational renewal work, shared and joint leadership, enabling leadership, co-principalship and learning-oriented leadership.

Jenny Madestam, PhD in Political Science at Södertörn University. Her research deals with leader ideals leader recruitment processes in Swedish parties. Current studies concern collectivistic forms of leadership between principals in schools and among leaders in Swedish parties as well as social media and political leadership.

Åsa Örnberg, PhD in Public Law at Stockholm University. Her research deals with municipal services and activities. A current study concerns the legal prerequisites for shared principalship in compulsory and secondary schools in Sweden.

Funding

This work was supported by Forskningsrådet om Hälsa, Arbetsliv och Välfärd [grant number 2013-0774].

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. We use ‘organiser’ for the Swedish juridical term huvudman. Such an organiser – whether municipal or private – is ultimately responsible for running the school according to the directions in the Education Act, curricula and other regulations. This responsibility involves allocating resources, organising activities according to local conditions, and following up, evaluating and developing the operation so that the national goals and quality demands are fulfilled.

2. A form of sharing where formal hierarchic equality is in place and work tasks are merged (see theory section).

3. Three schools for students aged 6–16, one for students aged 7–10.

4. The Swedish term is enhetschef and we have chosen to use a direct translation. The head of unit function greatly resembles West’s (1978) term ‘principal of administration’, but with respect to current legislation, cannot be called principal.