Education-line Home Page

‘Distributed leadership’ in schools: what English headteachers say about the ‘pull’ and ‘push’ factors

George K.T. Oduro
University of Cambridge, email:; website:

Paper presented at the British Educational Research Association Annual Conference, University of Manchester, 16–18 September 2004


This paper addresses three main questions about distributed leadership. Firstly, what meanings are attributed to the term distributed leadership within headteachers’ working vocabulary? Secondly, do headteachers conceptualise and practice distributed leadership in the same way as suggested by experts in the field? Thirdly, what issues do headteachers face in trying to ‘distribute’ leadership or create environments in which leadership is dispersed? Data are drawn from an on-going National College for School Leadership (NCSL) commissioned project researching the state of distributed leadership in selected schools in Essex and Suffolk. It is anticipated that issues emerging from these data will provoke further debate about the practice of distributed leadership in schools and also set the scene for further research. The data were gathered through semi-structured interviews and the shadowing of headteachers. The findings demonstrate that (i) Although headteachers have an idea of the concept of distributing leadership, the term itself does not form an integral part of their day-to-day working vocabulary (ii) The process by which leadership is distributed in schools may be understood in terms of (a) the initiative headteachers take to share leadership responsibilities with teachers, (b) the creation of an environment in which teachers feel free to own initiatives and assume leadership responsibilities (c) ways in which headteachers, teachers, and students/pupils relate to each other in order to promote a greater sharing of leadership. The paper further discusses some factors that promote distributed leadership such as trust, confidence, communication, risk-taking and financial capacity. It also discusses some factors such as insecurity, structure of schools, dishonesty and external interference, which, from the perspective of the headteachers, inhibit the practice of distributed leadership. It concludes with a discussion of the implications of these factors for the professional development of headteachers, and as a historical footnote reminds us that contrary to the claim of existing research-based literature (e.g. Gronn, 2002) that ‘the first known reference to distributed leadership was in the field of social psychology in the early 1950s’ (p.653), the origin of distributed leadership can be traced to 1250 B.C.


This article is based upon the result of a National College of School Leadership (NCSL) sponsored-research project carried out between September 2003 and May 2004. The research investigated the practical implications of distributed leadership in 11 schools in Essex, Suffolk and Hertfordshire. It involved 302 teachers and 11 headteachers from whom data were gathered through questionnaires, shadowing and interviews. My interest in this presentation lies in some of the issues that emerged from the shadowing and interviews.

The shadowing technique was used to explore how each of the headteachers spent his/her day’s school time focussing on actions and transactions. Shadowing, as a data collection technique, has been used in major international projects such as the ongoing UK-based Carpe Vitam (Leadership for Learning) research and in the University of Cambridge’s Students as Researchers project. It involves ‘a researcher following those they are shadowing for a day, or two days or perhaps even a week to build up information, insight and crucially a sense of understanding that particular case’ (Sutherland & Nishimura, 2003, p.33). The interviews were used to explore, among other things, how the headteachers’ saw leadership, and those they considered as leaders in their schools; the meaning that they attach to the notion of ‘distributed leadership; those they consider to be initiators of distributed leadership; and factors that promote or inhibit the practice of distributed leadership in schools.

I thank the director of the project, Professor John MacBeath and Ms Joanne Waterhouse, a member of the research team, for permitting me to use material from the study for this presentation.

The centrality of leadership in school improvement

Pressures from changing government policies and competitive market demands have, over the last two decades, made challenges facing schools complex. This complexity has led to the identification of leadership as an indispensable coping strategy. Increasingly, educational researchers as well as educational policy makers have recognized the crucial role leadership plays in school effectiveness and improvement. As English writers such as West & Jackson (2001) observe, ‘whatever else is disputed about this complex area of activity known as school improvement, the centrality of leadership in the achievement of school level change remains unequivocal’. Similarly, Australian writers such as Macneill, Cavanagh & Silcox (2003) argue that ‘the effectiveness of schools in educating students is highly dependent upon the nature of leadership within the individual school’ (p.14). Simply put, ‘outstanding leadership has invariably emerged as a key characteristic of outstanding school’ (Beare, Caldwell & Millikan’s (1989,p.99).

The search for the best leadership model

The centrality of leadership in the attainment of the school’s mission has led to an extensive search for the best leadership model. Traditionally, the notion of school leadership tends to be limited to single individuals occupying formal headship positions, implying ‘lone leadership’ (Southworth, 2002). The perfect leader in this context is a headteacher who demonstrates heroic features such as authority, courage, control, confidence, the capacity to ‘ size things up and make them right, promote allegiance and compliance’ (Johnson, 1997). In this model, emphasis is placed on formality and opportunities for exercising leadership limited to hierarchical and structural positions. A teacher’s leadership therefore becomes feasible only when such leadership is exercised in his/her capacity as a deputy head, subject head or other formal leadership position holder or when exercised as a delegated responsibility. Leadership responsibility ‘is delegated either through formal post holding or in a more ad hoc ways according to the judgement of the headteacher or senior leaders […] usually with an attendant implicit or explicit accountability’ (MacBeath, Oduro & Waterhouse, 2004).

The shift from heroic to post-heroic leadership models

In recent years, the individual-focused heroic approach to leadership has been challenged. More and more, researchers and educational policy makers agree that the school’s ability to cope with the numerous complex challenges it faces requires more than reliance on a single individual’s leadership. Lashway (2003) articulates this as follows, ‘the common ideal of a heroic leader is obsolete […] the task of transforming schools is too complex to expect one person to accomplish single handedly’ (p.1). This is also reflected in Badaracco’s (2001) criticism that the heroic leadership model,

‘Fails essentially because it idealizes people, places a handful of individuals at the top of a "moral leadership" pyramid, and ignores the fact that human beings are inherently flawed […] It considers the majority of people as impotent, lazy and self-interested at the bottom […] it eschews the struggles of leadership and suggests that leaders have to be superhuman and presents a monopoly experience that is primarily male and for the most part aligned with the military model’.

It has also been argued that the success of contemporary organisations depends on leaders who are ‘humble rather than heroic, emotionally rather than intellectually wise, possess more "soft" than "hard" skills, people rather than system-oriented, and willing to celebrate failure as well as success’ (MacBeath, 2003).

As an alternative to heroic leadership, a post-heroic model that places school leadership ‘not in the individual agency of one, but in the collaborative efforts of many’ (Johnson, p.2) has been advocated largely because,

‘The wave of changes resulting from structural, financial, curricular and technological reforms as well as a growing demand for accountability impact powerfully on the working lives of not only headteachers but teachers, students and all others who are directly or indirectly involved in the continuity and improvement of the school’ (MacBeath et al, 2004).

The post-heroic model emphasises human relations-oriented features such as teamwork, participation, empowerment, risk taking and little control over others. In this context, school leadership ‘does not command and control, but works together with others, constantly providing relevant information regarding plans and operations’ (Eicher, 2003). In grappling with the challenges facing the school the headteacher is expected to work ‘alongside others, modelling the very interaction they seek to encourage,’ remembering that although he/she occupies a formal leadership position ‘the power needed to change classroom practices is widely dispersed, residing not in central office but in the many private lesson plans and staff conference rooms of the schools’ (Johnson, p.1). Deep-rooted in this model is the recognition that school effectiveness ‘depends less on individual, heroic action and more on collaborative practices distributed throughout the organisation’ (Fletcher, 2002).

One dimension of the post-heroic leadership model, which has gained much credence among English researchers, writers and educational policy makers in recent years, is distributed leadership. It has become a major focus of the National College for School Leadership’s (NCSL) research projects. As Southworth (2002) suggests, ‘today there is much more talk about shared leadership, leadership teams and distributed leadership than ever before.’ Bolden (2004) elaborates as follows,

‘An increasing awareness of the importance of social relations in the leadership contract, […] and a realisation that no one individual is the ideal leader in all circumstances have given rise to a new school of leadership thought, referred to as ‘informal’, ‘emergent’, ‘dispersed’ or ‘distributed’ leadership’. (p.12).

Distributed leadership: its origin.

Distributed leadership is one of the most ancient leadership notions recommended for fulfilling organizational goals through people. This notion has long been reflected in adages associated with decision making in societies. Examples of these are the English adage ‘two heads are better than one’ and two sayings from one former colony of Britain – Ghana: ‘etsir kor nko egyina’ literally meaning ‘problem-solving through consultation is impossible with a single person’s wisdom’ and ‘Nunya adidoe, asi mesu nei o’ literally translated ‘knowledge is like the baobab tree, no one person can embrace it’. The ‘baobab tree’ metaphor drives home the fact that leadership wisdom, knowledge and skills needed for solving an organisation’s problems go beyond the capacity of a single individual. In the context of administration, the notion of distribution leadership could be traced far back to 1250 B.C. The biblical Jethro, Moses’ father-in-law, used the idea as an alternative to Moses’ leadership style of not sharing administrative workload with others. This is reflected in Jethro’s advice:

‘Thou shalt provide out of all the people able men, such as fear God, men of truth, hating covetousness; and place such over them to be rulers of thousands, and of hundreds, rulers of fifties and rulers of tens. And let them judge the people at all seasons: and it shall be that every great matter they shall bring unto thee but every small matter they shall judge so shall it be easier for thyself, and they shall bear the burden with thee’. (Exodus 18: 21 & 22).

Jethro’s model was based on the principle that ‘Great men should not only study to be useful themselves, but contrive to make others useful’ (, 2003). This principle, as MacBeath et al (2004) explain, ‘implies not only a delegation of authority but the creation of an environment in which people are able to grow into leadership.’ The principle however remained dormant over the centuries appearing ‘not to have been explicitly theorised until the latter half of the last century’ (ibid), when according to Gronn (2002 ), it became important in social psychology and organisational theory.

The idea became an issue in school leadership literature around the late 1990s (Gronn). In the USA, as an example, one strategy that characterised the move towards improving the standard of school leadership was to ensure that educational institutions have ‘leaders working effectively in "multiple leadership" or "distributed leadership" teams’ (Council of Chief State School Officers, 2000:5). In the United Kingdom, the concept had not been given much prominence until recently when the NCSL resurrected the discourse and set it as an essential principle in its school leadership development literature (MacBeath 2003, Bennett et al., 2003).

Distributed leadership: the problem of definition

Although NCSL and writers on school leadership as well as educational researchers have commonly endorsed distributed leadership as the backbone to school improvement, describing exactly how the term differs in meaning from related terms such as ‘distributive leadership’ ‘dispersed leadership’, ‘shared leadership’, ‘collaborative leadership’ and ‘democratic leadership’ appears confusing and problematic. While in some cases, these terms are used interchangeably with distributed leadership to mean the same thing, ‘other writers are at pains to make fine distinctions among this ‘alphabet soup’ of descriptors ‘ (MacBeath et al.).

Reporting on her small-scale study of Primary School Management Teams in South West England, as an example, Kelly (2002), conceptualises both ‘delegated’ and ‘distributed’ leadership in terms of transfer and division, while ‘shared leadership on the other hand suggests collaborative responsibility.’ The definitional problem is further evident when one considers MacBeath’s (2004:4) distinction between ‘distributed’ and ‘distributive’ leadership, alongside the perception of other writers. On the one hand, MacBeath suggests that distributed leadership views ‘leadership roles as something "in the gift of the headteacher", which he/she allocates magnanimously while holding on to power’. On the other hand, distributive leadership ‘implies holding, or taking initiative as a right rather than it being bestowed as a gift’ (p4). Yet, analysis of Elmore’s (2000) and the University of Chicago’s Centre for School Improvement’s (CSI) use of the term ‘distributive leadership’ suggests that ‘bestowing leadership as a gift’ is not exclusive to distributed leadership:

‘Distributive leadership takes place when people who have been appointed officially as leaders (headteachers) become committed to ‘building learning organizations and providing opportunities for all […] to give their gifts, to develop their skills and to have access to leadership that is not dependent on one’s "place" in the hierarchy or formal organizational chart’

Similarly, the CSI’s school development initiative seeks to support principals to establish distributive leadership ‘where professionals with specific expertise and responsibility collaborate to strengthen teaching and learning across classrooms’ (CSI, 2001). The idea of ‘specific expertise’, in this context, according to MacBeath et al. (2004), ‘denote people collaborating across specified organisational roles and leadership being given or assumed relative to knowledge, competency or predisposition’ (p.). An examination of the following definitions for ‘dispersed leadership’, ‘collaborative leadership’, ‘democratic leadership’ and ‘shared leadership’ throws more light on the definitional problem because all of them project an element of distribution.

Figure1: Terminologies related to distributed leadership





‘Dispersed’ appears to suggest leadership as an activity that can be located at different points within an organisation and pre-exists delegation which is a conscious choice in the exercise of power. The idea of dispersed leadership is captured by David Green’s term ‘leaderful community’ which involves a community ‘in which people believe they have a contribution to make, can exercise their initiative and can, when relevant to the task in hand, have followers’ (Green, 2002).

Operates on the basis of 'alliance' or 'partnering' or 'networking.’ Network learning communities, sponsored by NCSL are an expression of collaboration across the boundaries of individual institutions. Collaborative leadership may also apply to an 'inter-agency context', expressed in schools’ joint work with community agencies, parents, teacher groups, and other external stakeholders.

Leadership as ‘democratic’ is by definition antithetical to hierarchy and delegation. Elsbernd (n.d.) suggests four defining characteristics (i) a leader's interaction with, and encouragement of others to participate fully in all aspects of leadership tasks (ii) wide-spread sharing of information and power (iii) enhancing self-worth of others and (iv) energising others for tasks Democratic leadership can either take the form of consultative (where a leader makes a group decision after consulting members about their willingness) or participative decision-making (where a leader makes the decision in collaboration with the group members - often based on majority rule) (Vroom & Yetton, 1973).

Shared leadership is best understood when leadership is explored as a social process – something that arises out of social relationships not simply what leaders do ( Doyle & Smith, 2001). It does not dwell in an individual’s qualities or competencies but lies ‘between people, within groups, in collective action, which defies attempts to single out ‘a leader’ (MacBeath, 2003).

It is built around openness, trust, concern, respect and appreciation.

A common message that runs through these definitions is that leadership is not the monopoly of any one person, a message that is central to the notion of distributed leadership. In distributed leadership, as Gronn (2002:655) suggests, it is not only the headteacher’s leadership that counts but also the leadership roles performed by deputy heads, substantive teachers, support teachers, members of school councils, boards or governing bodies and students. Leadership is ‘dispersed rather than concentrated’ and does not necessarily give any particular individual or categories of persons the privilege of providing more leadership than others. Thus, as illustrated in Figure 1 below, the notion of distribution permeates all aspects of post-heroic leadership techniques.

Fig. 2: The centrality of distribution in post-heroic leadership terms.

In this light ‘distributed leadership’ cannot be said to be a new leadership technique but rather an intellectual label that seeks to re-enforce the fact that leadership needs to be a shared activity in schools. It should therefore be conceptualised, as Bennett et al, (2003) put it, ‘not simply as another technique or practice of leadership, but, just as importantly as a way of thinking about leadership’ in post-heroic terms rather than a heroic phenomenon. In this way, we can distinguish clearly between delegation – as a heroic phenomenon – in which distribution is initiated solely from the top (headteacher) and distributed leadership – as a post-heroic phenomenon – in which distribution does not solely depend on the headteacher’s initiative. As Bolden (2004) elaborates,

‘this approach argues a less formalised model of leadership (where leadership responsibility is dissociated from the organisational hierarchy) […] individuals at all levels in the organisation and in all roles (not simply those with an overt management dimension) can exert leadership influence over their colleagues and thus influence the overall direction of the organisation’ (p.12).

Viewed this way, Bennett et al’s distinguishing notions - ‘doing to’ and ‘doing with’ - as illustrated in the statement below will become more relevant to our understanding of distributed leadership.

‘Distributed leadership is not something "done" by an individual "to" others, or a set of individual actions through which people contribute to a group or organisation. […]. Distributed leadership is a group activity that works through and within relationships, rather than individual action. It emerges from a variety of sources depending on the issue and who has the relevant expertise or creativity’. (Bennett et al., 2003:3)

How do heads perceive and practice distributed leadership?

The headteachers in MacBeath et al’s study expressed a strong belief in the notion of distributed leadership. This is exemplified in remarks such as, ‘I think everyone in this school should have the opportunity to do so [exercise leadership) from the youngest child through out and not just a selected few’ (Secondary head) and ‘[…] I try to openly and honestly deal with problems in this school with the involvement of other people (secondary head).

Their understanding and practice of distributed leadership were characterised by both formal and informal processes: ‘sometimes the business stops with me but it can stop with someone else as well, ’(Junior head) and ‘[…] It comes from the school’s culture where people can assume leadership roles’ (Secondary head). One secondary headteacher articulated his commitment to practising distributed leadership by suggesting that a teacher’s working experience in his school was not a barrier to his/her participation in leadership:

‘Staff who have only been in the school for a short time could also be school leaders in that they show by their personality, by their vision, by their jobs, commitment, expectations and values that they have got the capacity to lead […] In a sense, anyone can be a leader […] it s a process that a lot of staff can demonstrate.’

Yet, the term itself did not appear to form an integral part of their working vocabulary. Apart from three headteachers, who used the term spontaneously attributing their sources to the NCSL, the remaining eight commonly used terms such as delegated leadership, shared leadership, and democratic leadership in their discourses. One headteacher explained why the term was not part of her leadership repertoire as follows:

‘Until this research project, I wouldn’t have given it (referring to the term –distributed leadership) any attention but I think that’s what we need in our schools. It’s distributed (used word after prompting) at every level and its not delegated leadership. Equally, there’ll have to be opportunities for anybody who has ideas that fit in with the purpose of where we’re going. We’ve got leaders at every level whether in subject areas, whether members of our teaching assistant teams or the pupils.’ (Headteacher Junior School)

Formal distribution

Formally, the process of distribution is initiated by the headteacher who identifies and delegates leadership responsibilities to individual teachers. Schools in England, as MacBeath et al explain, are structured in terms of designated leadership and managerial roles through which the headteacher delegates responsibility. Such delegations may be driven by a headteacher’s recognition that others have expertise that he/she does not have. Distributed in this way, there is an expectation of delivery and the headteacher’s role is to ‘support and provide’

‘If I give somebody responsibility, I expect them to get on with the job. Ours is a very low attaining school when based on SATS results. I’ve been encouraging subject co-ordinators to tell me what needs to be done. I don’t know what to do in English to raise standards. There’re some generic things I can do but in terms of how to teach English better, it’s the English specialist’s job so I distribute responsibility. If they tell me what they need then my job is to provide’. (Middle school headteacher)

Leadership in this context is ‘seen as giving a sense of ownership but at the same time is constrained within the remit and boundaries of the respective designated roles of staff members’ (MacBeath et al). One primary headteacher had this to say

‘Well I think it’s still important to have structure in leadership but distributed enough so that everybody feels that they’ve got ownership of something and that they feel empowered to be able to do something that’s their own. I keep coming back to subject leadership. I can’t talk about it in any other context really’.

(SenCo, Primary school)

The formality characterising the distribution process gradually leads unto a less formal or informal approach as headteachers develop trust in their teachers and become more confident in teachers’ leadership capabilities. As MacBeath et al (2004) put it ‘as headteachers become more comfortable with their own authority and feel more able to acknowledge the authority of others they are able to extend the compass of leadership and to ‘let go’ the more. This is evident in one headteacher’s comment:

‘I think initially from top-down through delegation and as it progresses it becomes both bottom-up and top-down. People who show willingness to take some levels of initiative from any direction are really encouraged. And I love to see it really happen and that’s when I become happy’.

Informal distribution

Words such as ‘instinctive’ ‘intuitive’ and ‘internalised’ characterised the headteachers’ description of the informal process. The process of distribution in this sense ‘shifts from what the head does to what others in the school do’ and leadership ‘taken rather than given’ or assumed rather than conferred’ (MacBeath et al). One headteacher articulated this as follows:

‘Here we don’t work to a formula […] I don’t work with that idea in mind. I do think that it is so instinctive and its internalised. It’s like conducting an orchestra. I don’t go around thinking I need to distribute this or that. I don’t do that. It happens instinctively because I trust the people I work with and have confidence in them […]’ (Secondary Headteacher).

This headteacher was not a lone voice. A primary school head that described himself as ‘an intuitive leader’ stressed that ‘A lot of people exert leadership in the school having confidence to do that not because someone has told them to do that […] I think how I operate here is a intuitive way. I want people to be involved’. He illustrated the notion of ‘intuitive leadership’ using a football team as a metaphor:

‘When the ball goes out of play the nearest player runs to retrieve the ball and get it back into play. Players typically decide taking a free kick or penalty on the pitch opportunistically. The flow is within an overall strategy but in the event intuitive and inter-dependent’.

The views of the headteachers shows how committed they are in implementing distributed leadership in their schools. They do not only understand distribution in terms of roles performed by those in formal leadership positions such as school leadership management team members, subject leaders or year group leaders. They also view leadership as a collective responsibility involving teachers and pupils/students who do not occupy any formal leadership position (s) and yet exercise leadership in different ways, at different places and at different times within the school. During our shadowing of the headteachers, as an example, we found in some schools students performing classroom leadership roles with confidence as teachers allowed them to lead their peers in some learning activities. In one secondary school, students of Key Stages 4 and 5 were actively involved in reviewing activities of the school’s information and technology (IT) department as part of its development planning. Students confidently expressed their views on matters related to how their teachers motivate them to learn, the amount of homework they do, the amount of feedback they get, discipline in the classroom and many other leadership related issues.

Furthermore, they see distribution not only as having the strength of preparing teachers and students for leadership but more importantly as a means of reducing the pressure of overwhelming workload on them. Once leadership is effectively dispersed, teachers are able to attend to the needs of pupils thereby reducing the frequency and the amount of time headteachers would have to spend with pupils:

‘My leadership style of granting departmental heads and teachers free hand to carry out shared responsibilities enables them to resolve most issues affecting students […]’ (Secondary Headteacher)

Headteachers’ workload, as revealed in our shadowing of their activities involved complex simultaneous tasks: receiving visitors, attending meetings, handling discipline matters, monitoring teaching and learning, taking care of cleanliness issues, managing paperwork and many other incidental activities. Fundamental to these tasks was unpredictable interactions with different people. As an example, one primary headteacher was seen chatting with one visitor (A). In the process, another visitor (B) walked towards where he was and tried to attract his attention. On seeing visitor B he interrupted his conversation with visitor A, attended briefly to visitor B and then resumed conversation with the first visitor. Figure 2 illustrates time spent by the headteachers on some major interactions during a school day:

Implementing distributed leadership: the ‘pull’ and ‘push’ factors.

Distributed leadership does not develop in a vacuum. Its development and sustenance in the school may be either promoted or inhibited by internal and external factors. Where the factors are favourable, they tend to make distributed leadership attractive pulling headteachers, teachers and pupils closer to its implementation. Where the factors are frustrating, it does not make distribution appealing to heads, teachers and pupils and thereby push them away from participating in leadership. The headteachers, though committed to the implementation of distributed leadership in their schools, identified some conditions that could either promote or inhibit its implementation (Figure 4 below sums up some major promoters and inhibitors).

‘Pull’ factors

Most frequently and commonly mentioned favourable condition for promoting distribution of leadership in schools is trust. There must be trust among teachers, between a teacher and a teacher, between teachers and headteacher, between pupils and teachers, between pupils and headteacher and among pupils. Some headteachers said they would not encourage teachers’ participation in leadership unless such teachers demonstrate their trustworthiness:

‘You’ve got to be clear about those you can trust to do a good job. If all of them, that’s great, but that’s not possible. Bring the positive ones up with you and tap their talents, talk to the negative ones if possible. If they don’t change, ignore them because they can divert your energy’. (Primary headteacher)

Trust was again seen as a significant tool for creating the congenial atmosphere needed to develop confidence in teachers and for promoting good working relations among staff.

‘Trust, confidence, a supportive atmosphere, and support for risk taking –a culture that says you can take a risk – you can go and do it. If it doesn’t work, we learn from it. I think there’s a range of cultural issues that support distributed leadership and create a climate; high levels of communication, willingness to change and to challenge; a climate that recognises and values everybody’s opinion’ (Secondary Headteacher

Willingness to share and pursue common goals was also found to be critical to the success of distributed leadership.

‘There must be common goals and objectives in the school and people must agree to move towards the same direction. People must agree on things on which there can be compromises and those which there can’t ‘(Primary headteacher).

Moving towards the same direction may be difficult unless there is a culture that promotes teachers’ mutual acceptance of one another’s capacity to lead, and ‘an environment of reciprocated trust’ (MacBeath et al).

‘Coherent staff: a staff that trusts one another. Others must accept the leadership capabilities of others. I’ve no problem asking a newly appointed staff to lead but their colleagues need to accept him/her’ ‘(Middle Headteacher).

An enabling atmosphere for risk taking was also paramount. ‘I feel there must be a safe environment where people are secured enough to venture, where they know they’ll be encouraged’ (An infant school headteacher). People take risks in an environment where mistakes are not seen as a mark of failure but as a learning opportunity. A teacher is therefore more likely to take up leadership risks with confidence when headteachers and teachers treat his/her mistakes in ‘a non-judgemental manner and within a supportive atmosphere […] where there is the knowledge that all individual contributions are valued’. One headteacher had this to say:

‘When people come out with new ideas, I ask them if they’re prepared to carry out the idea. […] I try to make people feel confident about what they can do because most people have the ability to lead. What they need is confidence’.

Making people confident requires good interpersonal relations because people gain confidence when they are made to feel confident: ‘distribution can be seen in terms of how we relate to one another […] it’s about our attitudes which are more important’ (Secondary head). Hargreaves (1975) draws attention to the influence of relationships in promoting classroom leadership: ‘the creation of the appropriate classroom atmosphere, namely one that is non-threatening and acceptant, springs from the kind of relationship teachers establish with pupils’ (p.170).

‘Push’ Factors

The converse of the foregoing pull factors inhibits the implementation of distributed leadership in schools (see figure 2 below). One headteacher, as an example, emphasised distrust as a reason for which he might resist a teacher’s participation in leadership: ‘When there’s disagreement between a teacher’s vision and the school’s vision. […] I don’t suppose to have leaders in school where their visions undermine the shared vision of the school (Middle headteacher), while one other attributed teachers’ apathetic attitude towards leadership responsibilities largely to insecurity:

‘If staff are given a role, they need to feel secure with that role. For example, the ICT specialist will block other members from sharing his secret garden of knowledge if that person feels unconfident’(Secondary head)

Pressure from workload was another frustrating factor identified by the headteachers. Some of them explained that the overwhelming nature of workload on teachers tend to have an adverse effect on their motives about shared leadership. ‘I think it’s a motive issue […] when there’s so much pressure on teachers in the school they’ll definitely avoid taking leadership responsibilities’ (Secondary Head)

Another push factor which all the headteachers found worrying was the hierarchical structures of the school system and its associated demand for accountability. As one headteacher argued,

‘The structure of schools militates against distributed leadership. In my view, they’re Victorian in processes and structure. Often schools don’t focus on learning; they focus on control with 30 kids in a class, the bell going every hour to direct subjects; […] the control structure of school activities does not help pupils to acquire the skills to succeed in a world that is flexible, adjustable, free thinking, high level of communicative skills […] you’re controlling them and that militates against distributed leadership’ (Secondary headteacher).

Also frustrating is staff attrition. Some headteachers said they found the frequency with which teachers leave their schools after they had been helped to develop leadership skills frustrating and discouraging. In this sense, they were not enthusiastic to continue creating conditions for teachers’ participation in leadership because of budgetary constraints.

‘But one of my biggest worries, and I don’t think it will ever go away, is the thought that if you give a particular specialism to any one individual, that the institution is weakened – not necessarily because of the way that individual is fulfilling that role but the consequences of that individual, for whatever reasons, not being there next year or the year after to do that’. (Headteacher, secondary school)

Fig. 3: ‘Pull’ and ‘Push’ factors that affect distributed leadership

Discussion: Implications for professional practice and policy

Headteachers acknowledge distributed leadership as a tool for promoting pupils’ learning and improving the performance of their schools. Yet, the aspect of distribution that requires a headteacher to relinquish his/her role at times as ‘ultimate decision maker and trusting others to make the right decision’ (MacBeath et al) remains problematic. It places headteachers in a dilemma as they struggle between fulfilling external expectations characterized by accountability and creating an environment that will not give them (heads) the privilege of providing more leadership than others. As one headteacher argued, ‘I try to motivate people to take decisions but in the end I’m the one who is accountable, the one whose neck is on the line as it were’. In coping with this dilemma, this headteacher adopts a ‘benevolent dictatorship’ leadership approach, which, he explains, involves caring and being sensitive to people’s leadership capabilities without totally relinquishing ones control. The policy implication here lays in the provision of professional development strategies that will equip headteachers with the requisite skills for balancing the principle of distributed leadership with formal expectations. One headteacher ‘s concern was:

‘There is however a dilemma. If you give somebody a role and responsibility […], when or how far do you step back and not intervene and let them get on with the job so that in the end, the head becomes so removed from the school because you’re not intervening?’ (Secondary Head)

Resolving this dilemma has policy implication for school accountability. While I do not refute the fact that schools depend on external support and must therefore be accountable to external bodies, I believe strongly that subjecting the school to extreme compliance to external mandate threatens successful distribution of leadership. Once a school’s position on league tables continues to determine its success and for that matter the effectiveness of its leadership, headteachers will be cautious of how far leadership should be distributed. They cannot avoid making teachers accountable through monitoring, scrutiny of data and performance management. This tends to make teachers apathetic towards participating in leadership.

Successful implementation of distribution will therefore mean a reduction of external pressure on the school. Schools should be given greater autonomy in the determination of where they want to be, how they want to get there and when they wish to get there. Politicians and their agencies, such as the Office for Standards in Education (Ofsted), should allow schools more freedom to ‘speak for themselves’ (MacBeath) by taking ‘more responsibility for their own accountability, a greater role in steering and shaping their own improvement’ (MacBeath, Oduro & Lightfoot, 2004) instead of requiring them to respond to extreme external accountability. Achieving this will require Ofsted, as an example, to demonstrate trust in the professional leadership capabilities of headteachers and teachers in meeting the needs of the school’s clients. Ofsted must intervene to provide professional development support but not to interfere in the school’s desired pace of development. Without this the implementation of distributed leadership in the manner that experts define it will remain a mere rhetoric.

‘The beauty of a bird’, as a Ghanaian proverb goes, ‘lies in its feathers’. In the same vein, availability of resources is a necessary precondition of distributed leadership. Its success will largely depend on the strength of a school’s budget. Headteachers may create the necessary atmosphere for the active involvement of teachers and pupils in leadership at various levels. But if the school lacks the requisite resources, both human and financial to support its initiatives towards distribution, implementation will be problematic. As one Junior school headteacher explained, she could be more encouraged to involve people in leadership if her school enjoys ‘financial stability because it means when resources are needed I can provide’ while a Middle school head commented on the need for appropriate human resources: ‘provided I can assemble a staff that is skilled and efficient and trustworthy, then I’ll expect them to get on and do their jobs and to do them better than I can do’.

There are also implications for professional practice. Headteachers need to internalise the principles underlying the distribution of leadership and be prepared to share their authority with teachers with little control. This means headteachers must demonstrate practically that they respect and value each teacher’s leadership potential, appreciate efforts of individuals to take risk and also see a teacher’s mistake as an opportunity for learning. ‘The embracing of failure’ and ‘the positive celebration of error’ do not only help the weak to develop positive attitude towards seeking avenues for improvement but also makes people more willing to take on leadership initiatives in school. Heads need to remember that ‘mistakes’ they say, ‘are the bridge between experience and wisdom’ (MacBeath, 2003,p.1). Willingness of headteachers ‘to stand back and listen and allow other people to develop their talents as well’ can be possible when such heads trust the people they work with and do not find their positions threatened.


In spite of the confusion surrounding the definition of distributed leadership and the problems associated with its implementation, I believe strongly that it holds the key to the school’s capacity for meeting its complex challenges. Schools should therefore be encouraged to explore ways of counteracting factors that inhibit the implementation of distribution. Mutual trust should be at the center of interactions between and among teachers, headteachers, pupils and all stakeholders of the school. School accountability should be viewed from a more developmental perspective with the school itself playing a central role. Unless both the school and Ofsted make conscious efforts toward finding a more human-focused approach to the implementation of distributed leadership in schools, the phenomenon will remain a mere intellectual exercise.


Badaracco, J. (2001) ‘Beyond Heroic Moral leadership’. At

Beare, H., Caldwell, B., and Millikan, R. (1992). Creating an Excellent School. London: Routledge.

Bennett, N., Wise, C., Woods, P. and Harvey, J.A. (2003). Distributed leadership. Summary of a literature review carried out for the National College for School Leadership. (2003) ‘The Concise Matthew Henry Commentary on the Bible. Jethro’s counsel to Moses. At

Bolden, R. (2004). What is Leadership? Leadership South West Research Report. At

Centre for School Improvement (2001,2003). School Development. The University of Chicago: CSI

Eicher, J (2004). ‘Post-heroic. Managing the virtual leadership organization’. Performance Improvement Global Network. A chartered Chapter of the International Society for Performance Improvement.

Elmore, R.F. (2000). Building a new structure for school leadership.

Washington, D.C: The Albert Shanker Institute Fletcher, J.K. (2002). ‘The greatly exaggerated demise of heroic leadership: gender, power and the myth of the female advantage’. Center for Gender in Organizations (CGO) Insights. Briefing notes No.3

Gronn, P. (2002). ‘Distributed Leadership’. In Leithwood, K. & Hallinger, P. (2002). Second International Handbook of Educational Leadership & Adminstration. Kluwer.

Hargreaves, D. (1975). Interpersonal Relations and Education. United Kingdom: Routledge & Kegan Paul.

Johnson, S.M. (1997). ‘The Misguided Search for heroic Leadership: Brief Article’. School Administrator, Feb. 1997. At

Kelly, Allison (2002). ‘Team talk: Sharing Leadership in Primary Schools’. Available from NCSL web site at

Lashway, L. (2003). ‘Distributed leadership’. In Clearing house on Educational Management (CEPM). College of Education, University of Oregon, Research Roundup 19, 4.

MacNeill, N., Cavanagh, R.F. & Silcox, S. (2003). ‘Pedagogic principal leadership’. In BELMAS (2003). Management in Education, Devon: The EducationPublishing Company Ltd.

MacBeath, J. (2003). ‘Leadership: Paradoxes of leadership in an age of accountability’. Paper delivered at the Sixth World Convention of the International Confederacy of Principals, Edinburgh, 15 July, 2003. At http:

MacBeath, J., Oduro, G.K.T., Waterhouse, J. (2004). ‘Distributed Leadership’. An Unpublished Research Report Submitted to the National College For School Leadership.

MacBeath, J., Oduro, G.K.T., & Lightfoot, S. ‘School Self-Evaluation’. An Unpublished Research Report Submitted to the National Union Of Teachers.

MacBeath, J. (1999). Schools must Speak for themselves: the case for school evaluation. London: Routledge.

Southworth, G. (2002). Evidence-based Management and Leadership in the Primary School. London, Falmer Press.

Sutherland, G. & Nishimura, M. (2003). ‘Tools for schools’. In MacBeath, J., Sutherland, G. & Nishimura, M. (2003). Self-Evaluation in the Classroom. London: Routledge

West, M. & Jackson, D. (2001). ‘Developing school leaders: A comparative study of leader preparation programmes’. At

Vroom, V.H. & Yetton, P.W. (1973). Leadership and decision-making. Pittsburgh: University of Pittsburgh Press

This document was added to the Education-line database on 29 September 2004