Shared Principalship: The Perspective of Close Subordinate Colleagues

ABSTRACT

This article presents a study of five schools with shared principalship. It contributes knowledge about how shared principalship is experienced by people who work closely (“close subordinates”) with the shared principalship constellations. Data consists of 20 semi-structured interviews. Close subordinates describe that the shared principalship meant a reasonable workload for their principals and welcomed their level of accessibility. Confidence in the leader collaboration was heightened if close subordinates were able to witness that collaboration in action with their own eyes. The findings are discussed in relation to the conditions introduced by current school legislation in Sweden.

A school principal’s workload is recognized in the literature as being a heavy one, 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 professional isolation that can be felt by school principals (Kelchtermans, Piot, & Ballet, 2011) contrasts 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). The idea of a “super principal”—an individual carrying the full weight of responsibility for running and developing a school—belongs to a traditional organizational model (Eckman, 2006). 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 study featured in this article originates in the recent changes to the Education Act that reduce the legal options for shared principalship in schools in Sweden (Örnberg, 2016). These policy changes stand in contrast to findings about the benefits of leadership shared between managers (including school principals) that previous research has identified (e.g., Döös, 2015; Rosengren & Bondas, 2010; Wilhelmson, 2006). Legal openings do still exist for some forms of shared principalship, making it important to understand not only how the sharing principals themselves experience these ways of working (Döös, Wilhelmson, Madestam, & Örnberg, 2017), but also how it is experienced by people working closely with them. Such experiences are essential when determining whether or not to advocate shared principalship as organizational solutions in schools. In addition, there is generally speaking a knowledge gap in the research literature when it comes to staff views on leadership shared between managers. The aim is to contribute knowledge about how shared principalship is experienced by people who work closely with the shared-principalship constellations and to discuss this knowledge in relation to recent policy changes. In this article, we refer to the study subjects as “close subordinate colleagues,” mostly abbreviated to “close subordinates.”

This article represents a study of five schools with shared principalship, and the findings are discussed in relation to the conditions introduced by current school legislation. While this article focuses on the perspective of close subordinates, the experiences of the sharing principals themselves are reported elsewhere (Döös et al., 2017; Wilhelmson & Döös, 2017). In sum, the sharing principals point to a considerable variation in organizational models, forms, and constellations, with accompanying differences in work-task division. Overall, and irrespective of the type of sharing, the principals and vice principals gave voice to a striking sense of relief at not feeling alone in their duties; their problems and troubles became manageable, their ideas and decisions better developed. There was an emphasis on the significance of having easy access to each other and on the importance of the bedrock qualities (mutual trust, lack of pretension, and values held in common). They appreciated the personal differences and differences in expertise within the leadership constellations, and felt they had more space to achieve quality, be proactive and solve crises.

We give below a brief introduction to the Swedish school system and to recent legal changes governing the role of the principal, followed by an account of previous research. We then set out this article’s theoretical point of departure in relation to shared leadership. This is followed by a presentation of methods, the findings, and a discussion.

A brief introduction to the Swedish school system and recent changes in legislation

Schools in Sweden are publicly financed but run both by municipal and private education organizers.¹ They are generally subject to the same rules. For all children between six and 16 years of age, there are 10 years of compulsory schooling. This is followed by three to four years of upper-secondary schooling that most young people take part in, either as preparation for university or for vocational training. Swedish school budgets are dependent on the number of students choosing and attending the school, as each student brings with them a set amount of funding.²

The current Education Act (2010, p. 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. The title principal is mandatory and reserved for this person. These changes to the law effectively closed off the legal options for joint principalship³ in Sweden (Örnberg, 2016). However, the legal prohibition of joint principalship was a side effect of the legislation rather than intentional. The law was primarily motivated by the state’s wish to lessen the influence of the local authority, increase the extent of national governance, and assign specific responsibilities directly to one principal within each school unit (Madestam, in press). During the legislative process, it was taken for granted that appointing one person as principal would guarantee clarity over where ultimate responsibility lay (Örnberg, 2016). A school unit may also have vice principals, but those are not mentioned in the Education Act.

Under the terms of the Education Act, the principal makes decisions about the internal organization of the school unit, and is responsible for allocating resources within the unit in accordance with the abilities and needs of the children and students (chap. 2, sec. 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). Apart from the requirement that there should be only one principal per school unit, the state gives organizers and principals autonomy over the organization of their schools, resulting in a diversity of organizational models and forms of leadership. For example, large schools can be organized into more than one school unit and thus have one principal per unit.⁴

Previous research

Shared principalship greatly resembles co-principalship, the term used more commonly in the school-leadership literature (e.g., Court, 2003b; Eckman, 2007). Different aspects of co-principalship have been empirically investigated over the course of the past 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 recruitment of principals (Brooking et al., 2003; Eckman, 2007), school development (Møller et al., 2007), the practice of democracy (Thomson & Blackmore, 2006; Wilhelmson & Döös, 2016b), challenging the ideals of heroic leadership and new public management (Court, 2004a), description of work models (Bunnell, 2008; Gronn & Hamilton, 2004) and work processes (Paynter, 2003), and bridging cultural differences (Bunnell, 2008, 2015). This collaborative form of leadership in schools is praised by researchers, who report findings of successful schools and contented 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 relating to “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 can fail, mainly because of difficulties in cooperating within the co-principal partnership, but also due to having to deal with resistance from the authorities.

Empirically based literature concerning staff perspectives on shared leadership in management constellations is scarce.⁵ The few studies identified come from health care (Choi, Holmberg, Löwstedt, & Brommels, 2012; Rosengren, Bondas, Nordholm, & Nordström, 2010), schools (Glenny, Lewis, & White, 1996; Gronn & Hamilton, 2004; Grubb & Flessa, 2006; Paynter, 2003), various municipal activities (Wilhelmson, Döös, Backström, Bellaagh, & Hanson, 2006), and theater and art (Järvinen, Ansio, & Houni, 2015; Reid & Karambayya, 2009). The studies report both positive and negative staff experiences of leadership shared between managers.

Where staff had positive experiences, it seemed to be a result of the sharing managers experiencing less stress and a reduced workload. They had time to handle tasks more effectively, so that staff felt supported, fairly treated, and empowered. The sharing managers are felt by staff to be accessible and complementary to each other, leading to positive qualities such as flexibility and creativity (Grubb & Flessa, 2006; Paynter, 2003; Rosengren et al., 2010). The sharing between managers also promoted sharing ways of working amongst employees, encouraging participation, teamwork, and models for shared decision-making (Glenny et al., 1996). Seeing two people interact as peers and equals was seen as beneficial by staff and students (Grubb & Flessa, 2006), and feeling involved and valued made staff more willing to cooperate in a health-care merger (Choi et al., 2012). Boundaries in organizations were bridged, such as between the artistic and managerial functions (Järvinen et al., 2015) or between administrative and professional functions (Choi et al., 2012). One study (Wilhelmson et al., 2006) relates the experiences of staff to the managers’ own experiences of sharing and finds that the experiences largely match. Interviewees say that their managers cooperate well and are more accessible than solitary managers. The staff members interviewed also say that their managers differ from one another, and this is largely seen as beneficial as the managers are felt to complement each other well.

Negative experiences of shared leadership among staff might be due to the fact that, although the leadership constellation of the leaders may function well, it might be impenetrable to the staff, “creating a sense of otherness” (Järvinen et al., 2015, p. 24). Also, employees sometimes have to ask or inform both leaders to be sure that both know what is happening (ibid.). For the most part, negative experiences are due to the sharing managers not getting along (Grubb & Flessa, 2006; Paynter, 2003; Reid & Karambayya, 2009). In such cases, the sharing may be perceived as unproductive in that it does not result in the managers having a common voice, or, if it becomes apparent to staff that there is disagreement, it may be seen as energy-absorbing, time-consuming, and confusing; this may make teachers keen to return to a more hierarchic structure (Paynter, 2003). In one case, destructive conflict in a leadership duo´s relationship was found to have spread into the organization, so that members of the organization became dispirited and less optimistic. Dual leadership thus lost effectiveness as a motivational force for members of the organization (Reid & Karambayya, 2009). Wilhelmson et al. (2006) found that staff report disadvantages that are not recorded by the sharing managers themselves, such as things falling between two stools or managers forgetting to provide an answer they have previously promised. Other concerns raised are a lack of clarity in communication and drawn-out decision-making processes.

Shared leadership—theoretical point of departure

Collective leadership is an umbrella term for shared responsibility in an organization and is here thought of as consisting of two distinct but connected subsets: distributed leadership with responsibility and power being spread to those not in management positions (e.g., Jones, 2014), and shared leadership between managers (e.g., Döös, 2015), with the latter being relevant to this article. The number of sharers varies, and the two-person instance is regarded as a “special case of shared leadership” (Pearce & Conger, 2003, p. 8), as power and responsibility are located and shared between two managers instead of only one. Thus, the phenomenon focused on in this study belongs to Denis, Langley, and Sergi’s (2012) second stream of research on plural leadership: “pooling leadership capacities at the top to direct others” (p. 213). Denis et al. include studies of co-leaders, co-managers, and co-principals in this stream.

In this article, we use the term shared principalship, following Döös’ (2015) conceptualization of three identified forms of leadership shared between managers within an organizational unit: joint leadership, invited leadership, and functionally shared leadership. This conceptualization builds on an interest in work-based experiential learning processes in organizations (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 there is formal, hierarchic equality and work tasks are merged. Vertically invited leadership points to a sub-form of shared leadership, where there is one official manager and a lower-ranking partner shares in decision making “with the permission of the first” (De Voogt & Hommes, 2007, p. 2). This vertical invitation operates in a similar way to Heenan and Bennis’s (1999) co-leadership. Within a school unit, an example of this form is when a principal shares leadership with one or two formally subordinate vice principals. Functionally shared leadership is where hierarchical equality exists but managers have different and separate professional areas and daily tasks; this form of sharing has been reported in studies of leadership pairs, and terms such as dual leadership have been used (Järvinen et al., 2015). Within a school unit, an example of this form is when a principal shares leadership with an administrative director at the same hierarchical level. When the sharing crosses organizational boundaries, it is conceptualized as horizontally invited leadership (Döös et al., 2017). In the school context, an example of this is when a school consists of two different school units and the two principals collaborate in a shared leadership.

There are basically two organizational models of sharing between managers: in the first, there is one unit with staff, activity, and a budget in common (organizational model A), and in the other, each manager has their own unit and cooperates across unit boundaries within a wider organizational context (organizational model B) (Döös, 2015; Wilhelmson et al., 2006). The first three forms of shared leadership outlined above relate to shared leadership according to organizational model A, whereas the last form is related to organizational model B.

Managers who share successfully highlight three qualities that together form the bedrock of sharing: mutual trust, a lack of pretension, and values held in common. It is common for managers who share leadership to underline the benefits of their differences as individuals and stress that extensive communication is essential. Cracks in the bedrock are one of the two main reasons why shared leadership fails (Döös, Wilhelmson, & Backström, 2013). The second main reason is that obstacles are somehow put in the way by the environment (ibid.).

Method

The study was designed as a qualitative exploratory study and data was collected between 2014 and 2016 in three compulsory schools and two upper-secondary schools where the principals shared their principalship. Data consists of a combination of interviews, micro-reflections, and meeting observations. The study is part of a wider research project on shared leadership. We outline below the background to the choice of schools, describe data collection and analysis, and also provide a short introduction to the five schools.

Choice of schools

Our requirement was for schools where the principal was sharing leadership, and the relevant information came from a principal or vice principal at the school or from the administration department of the local school organizer. This can be seen as constituting a positive selection of shared principalships, in that the principals in the selected schools saw a value in their sharing of leadership. There was also a risk of negative selection, in that principals working particularly closely and effectively across unit boundaries would not make themselves known for fear of being recognized and prevented from continuing their collaboration by the organizer or the Swedish Schools Inspectorate (Wilhelmson & Döös, 2016a). We aimed to achieve a variety in the schools selected, on the basis of whether they consisted of one or several school units and whether they were publicly or privately run. The study included both compulsory schools and upper-secondary schools and also schools which differed in terms of the number of applicants per place and the socioeconomic background of the students.

Data collection and analysis

The data used in this article consists of 20 recorded, semi-structured interviews (approx. 60 minutes each). Semi-structured interviews allow for a combination of structure and opportunities for open, follow-up questions (Kvale & Brinkman, 2008). The interviews were carried out either by the first, second, or third author. Interview guides were used and focused on the interviewees’ experiences of the shared principalship of their school. The interviews were transcribed. The interviewees worked closely with the shared principalship constellation. They were direct reports who were members of the school’s management team but not part of the principalship constellation. Most of them were teacher team leaders (18) and two worked as vice principals with specific tasks outside the sharing constellation. Of those close subordinates, 15 were women and five were men. The teacher team leaders had mainly teaching roles with a reduction in teaching hours ranging from zero to 20%.⁶

The focus of the analysis was on how the work of shared principalship constellations was experienced by the interviewed close subordinates. The analysis was a continuous sense-making process taking place within the research team both during and after data collection. As a final stage, the interview material was analyzed by the first and second author. This identified the themes displayed in the findings section and collated instances where the themes were expressed. Attention was given to both similarities and variations in the material. We were mindful during analysis of whether the interviewee worked in a school of organizational model A or B. This is shown in the findings section in that each quotation is followed by a letter indicating whether the cited interviewee worked in a school of model A or B. The quotations are chosen in order to substantiate the themes presented and to illustrate whether the aspect is at hand in both organizational models or not. In the findings, the quotations from both models are displayed together when similar. This way of presenting is chosen to facilitate the understanding of similarities and differences between model A and B schools. All interviewees are cited (1–3 times).

This study does not handle sensitive personal information under the terms of the Swedish Ethical Review Act (2003, p. 460), and so approval is not required under Swedish law. However, action has been taken to ensure ethical conduct. Participating organizations and individuals remain anonymous. Personal names and “he/she” have been replaced by X, Y, or Z, and quotations used in the findings have been slightly edited. It was common for interviewees in some model A schools to use the word “principal” when also referring to the vice-principal(s), and we have not changed their wording in this respect.

Introduction to the schools studied

All five schools involved in the study (see Figure 1) have at least one principal, a management team, teacher teams, and teacher team leaders. Beyond that, they differ from each other a great deal. Three schools consist of a single unit (organizational model A) that is led by a principal and one or two vice principals; thus, school and school unit are one and the same. Two schools consist of two or three units (organizational model B), with each unit managed by one principal. Schools that consist of several units have some joint functions and some joint staff. Each principal has worked with their sharing colleagues in the leadership constellation to determine the extent of their shared principalship and the way it will work. Of the five schools, three are medium-sized (400–650 pupils) and two are large (900–1,000 pupils). It is the two larger schools that are divided into units.

Figure 1. The principalship constellation and organizational model of sharing in the five schools in the study.

In schools of model A, the constellation consists of either two or three people—either one principal with one or two subordinate vice principals or one principal with an administrative director on equal footing and a subordinate vice principal. In schools of model B, the constellation comprises two or three principals on equal footing.

Findings

The following is an account of how the study subjects experienced their school’s shared principalship. The experiences include both similarities and differences, and we have chosen to start with the theme where the views are most similar and finish with the one where they most differ. The overall pattern discernible is that variations in how the experiences are valued are largely related to whether the school consists of one (model A) or more (model B) school units. This is also pointed to throughout the presentation below. The title of each section is a quotation from the interviews that encapsulates the content of that section. The fifth and final section gives an account of two distinctive patterns, which are reflected in the two quotations of the title.

I. “Being a sole school principal is too big a task”

A principal’s working day consists of a large volume of completely disparate and sometimes conflicting issues to manage, some short term, some long term. For the schools in the study, this might involve cleaning contracts, pupil or staff matters, new educational findings, refurbishment, complaints about the school food, fire prevention, disputes with guardians, budget issues, contact with the media, school organization, dealing with superiors, statistics, documentation, or visits from the Swedish Schools Inspectorate. One view consistently expressed by the close subordinates is that shared principalship at their school means the principals have a more reasonable workload and greater security. For those close subordinates, the idea of having a single principal at the school is seen as problematic. They observe, for example, that a principal who has been at risk of burnout has recovered due to being able to share the principalship, and that the shared leadership arrangement also enables principals to change up or down a gear in terms of their workload depending on the situation in school and in their own private lives. These kinds of advantages for the school leaders were voiced from both model A and model B schools:

They give each other security. (B)

I think that the support they get from each other is almost essential to enable them to cope with the job. (B)

Our principal is happier, no longer having to do everything, a little bit of everything, but more able to focus on specific bits. (A)

I think they take it in turns a bit … it’s about making the school work, and whatever is happening just now [in their private lives]. (A)

Thus, close subordinates in both models observe how this form of working reduces the workload of their managers and provides stability. What is being described is both the everyday security for the principals in having one or two responsible colleagues by their side and the situation where one of the principalship constellation steps up and carries the main burden when needed but then is given space to recover—they take it in turns.

II. “There’s always someone you can go to”

Close subordinates stress the importance of leaders being accessible, and therefore highlight the value of having more than one person to go to. Despite the division into school units or the fact that people have different areas of responsibility in management, it is thus possible to get answers to questions quickly while the negative consequences of one principal being temporarily absent are diminished. The principals are felt to be up-to-date and able to cover for each other. Amongst the observations made are that principals have an open door, and that it is easy to put questions in passing and to get help when needed. It is also pointed out that shared principalship is good because there is better contact with the principal. As illustrated in the quotations, this accessibility is appreciated in both model A and model B schools:

If I’m in a fix, I can go to any one of them and we deal with it as best we can; if one isn’t there, I can go to another, they’ve got their finger on the pulse. (A)

We can approach any of the principals if we want to at any time. [While X is] on sick leave, there are still the others. It’s actually really good for staff at the school. (B)

You would never be able to have that level of contact with your principal if they had 800 people to keep track of. (B)

They aren’t as stressed, they have time to talk properly. It’s extremely important for a principal to have time for their employees. (A)

Thus, within a school unit, you can choose to approach the principal or vice principal, or the principal or administrative director, depending on what the issue is. You get to learn the routes to use to take an issue forward; you can go to “the great strategist” (A) or the “sounding board” (A). In the schools in the study that are divided into units, the normal procedure is for staff to turn to their own principal. However, it is regarded as an advantage that there are several people to approach if the unit’s own principal is absent. Sometimes contacts with a different principal are based on previous working relationships. It is also clear that approaching a principal other than your own is not always a straightforward step to take:

I have done it a couple of times, but have felt afterwards that I have perhaps just created a bit of a burden. (B)

III. “I could see them in the mornings sitting and talking to each other”

The sharing constellations were located in a shared physical space. They either had adjacent offices with a connecting door, were in the same room, or were very close and with easy access to each other. One aspect that became apparent in the interviews is the importance of concrete evidence of the principals’ collaboration. It matters, in both model A and model B schools, that close subordinates can see the leadership constellation together with their own eyes and also see that they communicate extensively with each other:

So they sit there, all three of them, going through the issues that they as principals have to deal with … They do work really, really well together. (A)

I can see that…. They are very in tune with each other, constantly updating each other…. they do talk a great deal. (A)

I think they bounce ideas off each other to make things work more smoothly. (A)

And it’s often a visual thing. When you’re down there, you can see that the door between them is open and that they check things off with each other. You see that they … they sit next door to each other, after all. (B)

In some cases, the close subordinates’ ability to see the leadership constellation collaborating is facilitated by the design of the premises. For example, some schools have glass walls, and in some cases their rooms are located so that they are clearly visible to staff. One constellation trio shared the same room into which they had squeezed their three desks, a tiny round table for meetings, and a projector in the ceiling to project documents being used during meetings. This solution signaled both cooperation and easy communication:

They sat really close to each other, in one room, with their tables opposite each other, so that alone says a lot about the way they worked. (A)

The other aspect of visibility mentioned was that the leadership constellation also appeared together to jointly lead meetings where, for example, both close subordinates and other staff participate. This was mainly found in model A schools:

Both of them are always at these [workplace meetings]. They are there together and they don’t leave any gaps, so that each of them covers the other’s areas of responsibilities; it’s a bit like a set of cogs working. (A)

Conversely, the absence of visible cooperation may mean that some people do not believe it is happening, or do not see the outcomes of that cooperation. For example, some close subordinates would like a more fixed structure so that they know when the principalship constellation is meeting. The principals should work with each other more obviously, meet according to a schedule, and disseminate information on what has been discussed. This was only voiced in model B schools:

As it is now, when there is no sense of “us,” rumors spread really quickly—“I heard this”—and it might not be true, but it’s already too late because the rumor has spread. So, staff get worried. I think they [the principals] could have created a clearer method of collaboration and got that out to us … letting us know what they talk about. That’s where I think something’s lacking. (B)

However, the mere existence of regular management team meetings, which in schools divided into units (model B) means one meets across unit boundaries, does not imply that these meetings have a unifying function for close subordinates. For example, the role of the educational management team was described as difficult to understand, even for those involved. In some cases, the principals’ workload leads them to use the shared leadership to take action separately, so as to create space for all the tasks that have to be completed. Sometimes they are able to check things off with each other in advance of a meeting where only one is going to be present and sometimes not:

Sometimes only X or Y is at a particular meeting. And then I know that the person leading the meeting will have checked things off with the other. But sometimes they haven’t spoken to each other, and have to go back to the other one to do that; but they tell us that, they’re not secretive about it. (B)

There was a time when they would forget to talk to each other about certain things. They weren’t prepared for the sort of educational management team we had…. But I think they talk an awful lot to each other, you can tell that they are in agreement. (B)

IV. “They’re really different from each other”

The leaders in the same school are often described as being very different, both as individuals and in terms of their experience and skills. One overall pattern is that the differences between the individuals in the principalship constellation are described in positive terms, and as being of benefit to the school and staff, where there is close and visible collaboration within a school unit (model A). On the other hand, similar differences can become a problem when they are felt to be further apart and are working in schools with several units (model B).

In cases of close and clearly visible collaboration, differences seem to have the potential to become a common strength. School leaders in a unit become known in the school for their close collaboration and are sometimes described as being very different but complementing each other very well. Differences of this sort may have advantages when one principal’s visionary ideas are boiled down by the other into concrete action:

They are very different, but they complement each other incredibly well. X is structured and analytical, and thinks before [X] makes a decision. Y is visionary and incredibly inspirational … really good at enthusing people. (A)

The individuals in a shared principalship might be referred to as a mathematician who likes to find solutions to timetabling problems, a person-focused individual who likes to find solutions to pupil health issues, and a leader who likes to have an overview and develop their staff. Or, from another point of view, as one person who wants to avoid figure work, one who does not have an in-depth educational background, and one who does not want to deal with all the stress alone:

X is good at personnel issues, even when they’re awkward. Perhaps a bit more receptive than Y. But Y is good at things that X is not good at, such as legislation, orderliness, clear strategies, and looking at figures. X isn’t a mathematician. Z is brilliant, must be the best person in Sweden on pupil health. And Y wouldn’t be able to do that side of things…. It’s a completely united form of leadership…. They speak through one channel. (A)

Some respondents refer to the advantage of having both sexes represented in the leadership constellation. Close subordinates at the schools consisting of one unit describe how these differences have positive consequences, in that the leaders develop better together than they do individually.

The other type of pattern relates mainly to the two schools divided into units (model B); the differences are seen as disadvantages, particularly if the principals are thought to collaborate too little within the constellation. When principals in schools divided into units are described as different types of people and as doing things in different ways, one example given is of principals providing different information about the same things to different parts of the school. The differences may be described as differences in leadership styles or as differences in personalities. The principals are described as having some fundamental differences in terms of, for example, preference for firm or loose school structures, how to calculate hours, etc.:

In fact, they pull in different directions. (B)

They lead in different ways; I think it’s extremely regrettable. I have indicated very clearly that I would like them to represent one school. (B)

X is more philosophical, a bit more afraid of conflict—that’s how it seems to me—not as quick to act … with Y things are a bit too hasty, a bit too impulsive. (B)

One of them emails every bit of information, all the time. One is clear and direct and gets things done quickly—this is how it’s going to be … and one who is a bit … perhaps not so structured and not so clear about what’s going to happen…. There has been some frustration, a lot of frustration, because they are so different. (B)

While not an issue in everyday situations, such differences can be particularly problematic when they coincide with other boundaries in the school, such as when programs and school-year divisions coincide with the division into school units. The differences are reinforced by the fact that the nature of the work in the various units is different.

V. “Everyone knows how things work” versus “actually it’s troublesome”

So, do close subordinates feel that the principalship constellation takes joint responsibility for the leadership of the school, and how useful is shared principalship felt to be for the school’s activity and the school as a whole? As has started to become apparent above, in some respects there is a dividing line between schools that consist of one unit (model A) and those that are divided into several units (model B); a dividing line that is partly about clarity versus a lack of clarity, about boundaries that hinder, and about differences in options for the division of labor within the principalship constellation.

For close subordinates, shared principalship seems useful for the school’s activity and the school as a whole when school leaders’ voices are united, when they work together closely and visibly, while also having clearly defined areas of responsibility. There is clarity, and people know the rules. One example noted is that better decisions are made. From the point of view of the close subordinates, it seems beneficial for the school’s activity that the principals become skilled together by exploiting each other’s strengths, thereby preventing a single person’s inadequacies from being detrimental to the work of the school. Four voices from model A schools:

I think many decisions are much, much better. (A)

They have pretty much the same ideas, the same thoughts about where they want the school to go; they have meetings and they talk to each other. (A)

They keep track of everything to do with the whole school, and then they have their own areas of expertise, but they share this information with each other…. It feels like it works really well. (A)

I like the division between the principal and the administrative director. It means X does what X is really good at, and Y can do what Y is really good at, because Y is really good at the educational and leadership sides and such like. (A)

In the schools that are divided into units (model B), there are more doubts and more questioning of the value of shared principalship to the work of the school. This may relate to instances in which the principals have collaborated closely in preparing for, say, a staff meeting or a project, but later still act separately with respect to the specific implementation in their own school unit. This makes their collaboration difficult for others to perceive. Close subordinates make it clear that they would like a more unified school and that dividing the school into units of lower, middle, and upper school, or by different types of program, is problematic. They miss the sense of connection with the other parts of the school, which can cause problems in, say, the canteen (cafeteria) if younger pupils become envious of liberties taken by upper-school pupils. The interviewees would like to see more coordination between the school units to help make the day-to-day running of the school smoother. The division into units also seems to be a hindrance for functions common to the whole school, such as the Pupil Health Service. Three voices from model B schools:

Things are bound to be a little different, because there are different programs and different circumstances. (B)

There’s no longer a feeling of “us” in the school; we want to know more about what’s happening in the other parts of the school. (B)

In our heads we still think we’re one school, but we don’t function as one school any more. The only thing we try to keep together is the Pupil Health Service, and that’s not easy either because we have different managers. (B)

At the same time, the number of pupils at a school has increased so that the school is too big and can’t be held together, even though it is felt that the principals try to uphold the idea that it should be, and be perceived to be, one school, particularly for pupils and parents. Sometimes the close subordinates say that the principals ought to prioritize differently. Differences between principals that are perceived as problematic may also be about wanting more classroom visits by one principal, while another principal at the same school is described as already doing visits:

X very rarely comes along to listen to lessons, and I don’t care if it’s because X hasn’t got time … obviously you can make time if you want to. (B)

Y is very engaged and involved … has been out observing our teaching several times … really has been out a lot and talked to a lot of people. (B)

In schools divided into units (model B), there are examples of close subordinates who highlight operational shortcomings connected to the shaping of the leadership function. For example, one thing that some interviewees felt was lacking was the work of agreeing on a common thread that specifies what pupils should have learned in different subjects in their earlier school years. One may be aware that the principals are trying to treat the school as one whole, but can see that it is nonetheless being pulled apart.

Pupils should learn certain fundamentals, which should be in place before they continue on to the middle and upper stages of compulsory schooling. Certain things need to be dealt with, and things like assessments should be completed. There should be certain things that they must know, so that someone doesn’t get to the age of 14 before we discover that they can’t read or do maths. (B)

They are focused on the school as a whole, or they try to focus that way. [And yet] there is a lot of focus on individual units…. The principals pull in different directions. Obviously, they have their own pet projects. (B)

Discussion

In this article, we have cast light on the phenomenon of shared principalship from the point of view of people who work closely with shared leadership constellations in five schools, in the circumstances arising out of the changes introduced by the 2010 Education Act. With reference to the pressurized work situation, principals are acknowledged to have, with a substantial workload, feelings of isolation and numerous responsibilities and tasks (e.g., SOU, 2015, p. 22); the principals themselves felt that the favorable consequences of the option to share principalship were striking (Döös et al., 2017). Close subordinates in this study agree on this point. They describe the value of the principals having a more reasonable workload as a result of not having sole responsibility for the school. Having more than one leader in a school meant that there was a welcome level of accessibility. Confidence in the collaboration between the leaders was heightened if close subordinates were able to witness that collaboration in action with their own eyes. Internal differences between the different individuals in the principalship constellation were described in both negative and positive terms, depending on whether or not the school was divided into several units. In the eyes of the close subordinates, the borders created in schools divided into units seem to contribute to a lack of clarity, both with respect to the shared principalship and the work of the school.

With respect to the few earlier studies of staff experiences of shared leadership in management constellations that we found, one aspect that we recognize is the value placed on the managers’ accessibility (Grubb & Flessa, 2006; Paynter, 2003; Rosengren et al., 2010). The conclusions previously outlined about boundaries being bridged through the managers’ shared leadership (Choi et al., 2012; Järvinen et al., 2015) related to a division between different types of tasks and not, as in the present study, to a division between units. Also, previous studies have identified lack of clarity as a problem—although within one organization (Järvinen et al., 2015). In terms of perceived difficulties and negative experiences, the issue mostly raised in previous studies was that the managers did not get along with each other (Grubb & Flessa, 2006; Paynter, 2003; Reid & Karambayya, 2009), which was not raised at all in the present study.

In the following, and based on the perspective that the experiences of close subordinates have helped to provide, we discuss three interrelated aspects that have been affected by the policy changes introduced by the current Education Act: the new concept of school units, the issue of shared leadership and clarity, and the form of shared principalship.

The concept of school unit

The changes introduced by the 2010 Education Act (2010, p. 800) indeed seem to have complicated matters when it comes to school cohesion and how the principal function may be organized. While intended to bring about clarity (Madestam, in press; Örnberg, 2016), as can be seen above the law has had the opposite effect in its introduction of the concept of school units. The resulting division into units introduces organizational boundaries within an individual school, which creates a lack of clarity and causes staff to question the extent to which their school forms a cohesive whole; an artificial, but very real, inner boundary leads to frustration.

In schools consisting of one unit (model A), close subordinates view a principal’s choice to share his/her principalship with one or two other people as both feasible and beneficial. Differences between the individuals in the leadership constellation are seen as positive and a strength, and the collaboration of the shared principalship is visible; it demonstrates solidarity between the members of the constellation and generates a feeling of security for their close subordinates. In schools that were divided into several units after the introduction of the Education Act (model B), the organizational model has itself instead contributed to what the close subordinates describe as difficulties in achieving a sense of “us” and a unified school. The internal differences between the principals are here described as disruptive and problematic. The close collaboration that the principals themselves say exists (Döös et al., 2017) is not visible and is viewed as insufficient; the differences between the principals are perceived as disruptive. The school is pulled apart because of the divided organizational structure.

In the study, the division into units is associated with size, measured in terms of the number of pupils and thus also the number of teachers; it is the larger schools that are divided into several units. The problems with shared principalship that the close subordinates perceive in the divided schools might thus be thought to be connected to the size of the school, a point that, as can be seen above, some of them in fact draw attention to. One common pattern is that the difficulties arise in the interspaces or gaps created by the boundaries arising from the division into units. Indeed, Tyrstrup (2007), with his concept of “organisational inbetweens,” highlights the negative consequences of boundaries between the parts of an organization for expertise and collaboration.

Shared principalship—visibility and clarity

As has been demonstrated previously, the legislators assumed that appointing one principal per school unit would lead to clarity around responsibility (Örnberg, 2016). The idea that there might be advantages in shared principalship, or even in joint principalship, did not occur to the legislators (ibid.). However, our study has shown that having two or three sharing leaders in one school seems to have a stabilizing effect for close subordinates and to create a sense of security—assuming the sharing is visible to them. It is important for close subordinates to be able to experience and have confidence in the unity and collaboration of the sharing principals, which is aided by their being able to see and experience the principals’ collaboration with their own eyes. The perceptions of close subordinates of the consequences of differences between the shared principalships for them and the school’s work thus seem to depend on whether or not the collaboration between the principals is visible and whether or not the school is divided into units. The organizational conditions have an impact on the principals’ ability to use their differences in ways that are seen by their close subordinates as constructive for the work of the school.

A relevant question to ask, therefore, is under what circumstances shared leadership works well or not (Döös et al., 2013; O’Toole, Galbraith, & Lawler, 2002). This study has shown how clarity for close subordinates can be achieved in model A schools, both through vertical, invited leadership (Döös, 2015) and functionally shared leadership (ibid.), which here points to leadership forms with a high degree of task division. In these cases, close subordinates note that collaboration between the leaders can clearly be seen, something that is easier if their rooms are clearly visible, or even have glass walls, making it possible to see with your own eyes that conversations are taking place. Thus, the negative consequences of being drawn apart by the division of tasks are diminished when responsibility for the school as a whole is taken jointly, and when the principals’ competence-bearing relations (Döös, 2007) are visible, not only to themselves but also to their staff.

On the other hand, difficulties arise in cases where the visibility of the collaboration is disrupted by a division into school units. This means close subordinates experience a lack of clarity with respect to the school as a whole, although perhaps to a greater extent than they have cause to, since they do not see the collaboration between the principals; it exists, but is behind the scenes. That principals do not sufficiently co-act in front of subordinates has various reasons, as, for example, lack of time and energy or a lack of awareness on the part of the principals that their staff need to experience collaborative leadership with their own eyes and ears.

New future policy for shared principalship

It can thus be stated that shared principalship is perceived as beneficial both by the school leaders themselves and by their close subordinates in schools that are not divided into units. However, this is not the case in the schools in the study that are divided into units. This raises the more general issue of the form of shared principalship. The benefits of a principal within a school unit choosing to share their principalship with one or two other people are apparent to close subordinates too, both when principalship is in a vertical, invited form (Döös, 2015) and functionally shared (ibid.). These positive experiences from model A schools in this study suggests that future research should look into the question of a similar model in larger schools, i.e., some form of shared principalship but without division into school units. In view of the difficulties identified relating to horizontal invited leadership across an organization’s boundaries, there is an urgent need for broader studies of schools divided into units before an opinion can be expressed on how common the problems are. However, one complicating factor to consider is whether the number of participants in the constellation would increase too much in number. Studies tentatively suggest that the leadership constellation should not be greater than three, as this makes it more difficult to maintain close communication and the bedrock qualities (Döös et al., 2017; Döös, Vinell & Knorring, 2017). Other aspects to study that may affect the conditions for shared principalship are the type of school organizer and the differences between the socioeconomic nature of schools’ catchment areas.

In addition to vertically invited leadership, we can envisage two future options for principals to share leadership. One of them follows the rules of the current Education Act. The principal has by his/her side a person whose job title is not principal, who both has and takes management responsibility for duties not governed by the Education Act—the logistics of running an organization. In such a functionally shared leadership, the principal has and takes responsibility for the educational work and an administrative director for duties not governed by the Education Act or curricula. To be able to fully relieve the principal of logistics responsibility, the administrative director must have equal status to the principal in the organizational hierarchy. In a larger school, there may also be a need for one or more vice principals who either support the principal or administrative director or both of them. The other option for shared principalship is to change the wording in the Education Act so that it becomes possible for two or more principals to have and take on joint responsibility for a school, perhaps with vice principals for support. That would mean that schools would not need to be divided into units. The main principles of both these options are shown in Figure 2.

Figure 2. Two examples of how the leadership functions in a large school could be organized: a principal and an administrative director, or two principals, on equal footing share responsibility for the school as a whole. They are supported by one or more vice principals who take secondary responsibility for aspects of the school’s work.

We would argue that these equal forms of shared principalship have the potential to bring unity and clarity back to larger schools that are currently divided (model B schools). We suggest that the joint as well as the functionally divided principalship would make it possible to manage a larger model A school in tandem. The rationale behind the new Education Act (Madestam, 2016) seems based on a combination of ignorance—at times rooted in a conservative view of leadership, the desire of a newly appointed education minister to demonstrate purpose, and a concern for the legal rights of pupils. The latter concern is highly legitimate. However, it ought to be possible to resolve this issue by stating in the Education Act that each pupil must have a designated principal. In our opinion, however, this should not be brought about by accidentally prohibiting a whole working method that previous research has shown to have great potential for both clarity and democratic working methods and sustainable working conditions for managers (Wilhelmson & Döös, 2016a, 2016b).

Conclusion and a look ahead

We believe that it is important to understand that there is no single answer to how a shared principalship is perceived by close subordinates. The answer depends on how the principalship operates and on the organizational circumstances. Neither is there a single answer to the question of what a suitable organizational structure looks like. The latter question can be further addressed in future studies that shed light on the organizational circumstances in which close subordinates, and the managers themselves, perceive shared leadership as beneficial. In view of the acknowledged difficulty of the working conditions principals are faced with, we consider it a matter of concern that the changes introduced as part of the current Education Act have limited the scope for shared leadership for this particular group of managers. The problems that have been highlighted concerning the principals’ working conditions (SOU, 2015, p. 22), together with research-based knowledge about shared leadership, point rather to an increase in the scope for this organization of leadership positions. Dividing into school units leads to problems; an artificial boundary threatens to disassemble the big picture. These are problems that schools consisting of a single school unit have not had to deal with. The clarity that the Education Act hoped to achieve led in practice to a lack of clarity when the boundary between school units has to be managed in order for schools to be perceived as complete entities.

Acknowledgments

We would like to thank the schools who took part in the study.

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article.

Funding

The authors thank FORTE (Forskningsrådet för Hälsa, Arbetsliv och Välfärd), who provided funding for our work.

Additional information

Funding

Forskningsrådet om Hälsa, Arbetsliv och Välfärd

Notes

1. We use the term organizer for the Swedish juridical term huvudman. The organizer—whether municipal or private—is ultimately responsible for running the school according to the prescriptions of the Education Act, curricula, and other regulations. This responsibility includes allocating resources, organizing activities in line with local conditions, and monitoring, evaluating, and developing the work of the school in order to fulfil national goals and quality requirements.

2. See Blossing and Söderström (2014) for a description of the Swedish school system.

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

4. This is a solution that is used to enable there to be more than one principal in the same school. However, there is as yet no accurate official data on how many schools divided into units there are.

5. In order to find research in which staff views were expressed, we searched methods descriptions and empirical descriptions. Staff views mainly featured as a complement to the views of the managers themselves.

6. In general, a teacher team leader leads team meetings, takes part in the school’s management team, and is a link between the teacher team and the school’s leadership function.