Understanding School Leadership and Management Development in England

The Era of Administration: 1944 until the mid-1980s

This era covers a period that stretches back to before the journal was founded. It was embodied in the administrative structure of English education which was created by the Education Act 1944. This period was characterized by what Clarke and Newman (1997) have termed a ‘bureau-professional regime’, through which the welfare state, including education, was structured through a combination of high degrees of professional autonomy enacted within a broad framework of administrative rules and procedures. The system was often described as ‘a national system locally administered’, and the key players were local education authorities (LEAs) and the schools themselves. The underpinning values of the bureau-professional model were derived from the assumptions and practices of the constituent professions of the human services and, consequently, administrators were socialized within these values, which typically were based in ideas of equity, care and social justice (Clarke and Newman, 1997; see also Gewirtz et al., 1995; Simkins, 2000).

Within this context, ‘educational administration’ was typically conceived to encompass those in positions of responsibility both in schools and LEAs. ⁴ However, as the period progressed there was increasing uncertainty about the nature of these roles and the preparation necessary for them. For example, Glatter (1979: 19) alluded to the low key role of the ‘humble administrator’ and drew attention to the distinction made by Andrews (1980) between the ‘generic’ and ‘specialist’ professional models of administration. The former involved administrators being ‘promoted from the pool of experienced teachers … [and] where training does exist it is not usually extensive and reliance is placed primarily on experience in the job’; the latter saw administration as a sub-profession growing out of the teaching profession with ‘appointments to senior positions [being] made from the pool of those who have successfully completed an extensive graduate programme in educational administration’ (Glatter, 1979: 22–23). Analysis of the dilemmas involved here was taken further in Hughes’ influential work which, in relation to headship, developed the idea of ‘professional-as-administrator’, a role that involved seeking accommodation between the complex and sometimes conflicting sub-roles of ‘leading professional’ and ‘chief executive’ (Hughes, 1985), a distinction that prefigured key aspects of the debates of later eras.

In this period, the English system of administrator development, insofar as it existed, reflected both the ‘generic model’ and the locally based system that it supported. The custodians were seen to be the LEAs and universities, although these two players approached the problem from very different positions. University provision was typically embedded in social science theory, ⁵ while LEA provision, in contrast, comprised short courses that were primarily based in practice. The result was a somewhat uneasy alliance between what were essentially two parallel forms of LMD provision, within a diverse, fragmented and relatively low key ‘patchwork quilt’ (Glatter, 1979: 27). Reporting on his 1980 survey of provision in England and Wales, Hughes (1982: 14) was very critical of this fragmentation of short course provision in particular, calling for ‘a bold and imaginative initiative … to stimulate, but not in any way compel or control, the growth of a country-wide network of training activities in school management’. Echoes of this view were to resonate more and more loudly in the ensuing decades.

The Era of Management: The Mid 1980s to 1997

By the mid-1970s the bureau-professional regime was increasingly being challenged by the suggestion that school standards were too variable or, indeed, inadequate, and that this was the result of insufficient pressures within the system for effective performance. This was reflected in the gradual introduction of management ideas into the discourse, but the defining legislative moment of this era was the passing of the Education Reform Act 1988, which fundamentally redistributed power in the English educational system. Powers over the school curriculum were drawn to the centre, while major powers over staffing, finance and other matters were devolved from LEAs to schools ( EMAL Special Edition, 2008). These changes were part of a wider movement towards a ‘managerialist’ reconstitution of the public sector more generally which has been discussed widely elsewhere (Pollitt, 1993; Simkins, 2000; Learmonth, 2005). The discourse surrounding these changes produced a form of resolution of the ambiguities surrounding the administrator role that was characteristic of the previous era. Administrators were reconstituted as ‘managers’ with responsibility for the performance of the devolved school system against criteria that were centrally determined in Westminster.

Three particular developments in relation to LMD can be identified during this period. First, the expectations placed on LMD became more explicit, with an increasing focus on the identification of standards and competences for managers and the development of technologies, such as assessment centres (Lyons et al., 1993) associated with these. General occupational standards, such as those developed through the Management Charter Initiative (MCI) were soon imported into education, with specific standards developed by School Management South (SMS) and later by the Teacher Training Agency (TTA). In a detailed review of these approaches, Earley (1993: 235) suggested that many potential benefits in relation to both personal and organization development had been identified – for example, in relation to accreditation, self-assessment, appraisal or institutional/team review – and stated ‘it could be argued that they are particularly valuable for both institutions and individuals as the move towards school-based management development takes place’.

Second, conceptions of what LMD comprised became both sharper and broader. For example, the National Development Centre (NDC), which was nationally funded to promote the ideas of management development to LEAs and schools, outlined a ‘model for managing the development and training process’ (Wallace and Hall, 1989: 166), emphasizing the importance of balancing individual, school and broader system needs, and encouraging

The model did not include any explicit reference to the values or philosophical principles that should underpin such activities, although it did lay the foundations for a wider concept of management development that was continuously reasserted over the ensuing periods (Bolam, 1997; Brundrett, 2001; Bush, 2009). More generally, however, there was a shift in the kinds of knowledge that underpinned the discourse of development, with the social sciences initially being complemented by and then increasingly replaced by management theory, often adapted from business contexts (Gunter and Thomson, 2010). ⁶

Third, during this period central government attempted for the first time to influence provision of LMD and produce a somewhat more co-ordinated approach. Initially central control remained weak. LEAs were given responsibility for determining how the central funds – which were provided not just for management and leadership development but for in-service education for teachers more generally – would be spent. This led increasingly to a market developing with a range of providers competing for government funding and a continuing focus on courses external to the school (Wallace and Hall, 1989). However, by 1989 Wallace and Hall noted in many LEAs:

Many of these needs now arose from the demands made on heads and senior school staff by the Education Reform Act. Direct central intervention also increased. Compulsory appraisal of headteachers was introduced in 1991; the TTA, which was set up in 1994, embraced the competencies approach by establishing a set of ‘tasks and abilities’ required of headteachers; a development programme for newly appointed heads – HEADLAMP – was introduced in 1995 (Gunraj and Rutherford, 1999; Blandford and Squire, 2000); and in 1997 the TTA introduced the National Professional Qualification for Headship (NPQH), the first of a collection of centrally designed and controlled programmes that were increasingly to dominate the leadership development landscape in England in the years to come.

By the mid-1990s, then, the reforms of the late 1980s were embedded. As Bolam (1997: 271) noted, headteachers and other senior staff:

The history of this period, therefore, is one in which central government is active in the policy area of LMD for the first time. However, its approach embodied an on-going tension between the desire to co-ordinate provision in response to the perceived needs of fundamental policy changes for schooling on the one hand and a belief in the effectiveness of markets in organizing provision on the other. The outcome initially was a classic case of ‘policy steering’: central funding symbolized the policy significance of management development and training, decentralization of provision reflected market principles, but the severe pressures placed on the system by legislation had the effect of forcing provision in the desired direction. Alongside these developments, however, governments were clearly considering a more interventionist stance and by the end of the period the first elements of that approach were in place.

The Era of Leadership: since 1997

As these developments were occurring the discourse was changing. ‘Management’ of devolved responsibilities was no longer seen to be enough. Despite the reforms of the late 1980s, political concerns about the performance of the school system had not gone away – if anything they had increased. The Labour government, which came to office in 1997, articulated a ‘modernization’ agenda, which built on the Conservatives’ framework of school autonomy and accountability, but added a further ingredient – ‘leadership’ – which was seen as providing a much stronger agentic thrust in instituting change and improving performance in the public sector in general and among schools in particular than the more conservative idea of ‘management’ (Newman, 2005; O’Reilly and Reed, 2010, 2011). This move was embodied in the establishment of the National College for School Leadership (NCSL) in 2000.

The NCSL soon developed a ‘leadership development framework’ comprising five stages of school leadership around which it built a comprehensive programme of courses, with at their core a suite of programmes linked to the framework, comprising Leading from the Middle (LftM), Leadership Pathways for senior school leaders, NPQH (taken over from the TTA) and the Leadership Programme for Serving Headteachers (LPSH).

Since its foundation, the NCSL has been the focus of increasing numbers of papers on LMD in England published in EMAL and elsewhere. A special edition of EMAL was devoted to the College in 2004, and since then there have been many other papers, especially evaluations of particular programmes offered by the College. A number of writers have noted ways in which the National College presented a break from most of what had gone before, citing the leadership development framework, the advocacy and support of networked learning communities, the extensive use of modern technologies, and the comprehensiveness of its programmes (Bush, 2004; Bolam, 2004; Riley and Mulford, 2007). In responding to the articles in the 2004 special issue, Geoff Southworth (2004: 345), Director of Research at the College, noted that, with an estimated potential clientele of 250,000 leaders and potential leaders, ‘programmes alone cannot meet or serve the needs of all these leaders: the scale of the task is too great’. In responding to this challenge, he (2004: 346, 349) emphasized a number of themes that have resonated in the College’s work over the years: the significance of distributed leadership and the importance of developing the capacity of schools to contribute as major players in meeting the scale of the leadership development challenge; the need to consider leadership at system as well as school level in order to address concerns about fragmentation in an increasingly diverse school system; and, ‘avoiding any one approach’, the ‘need to develop a far more discriminating and differentiated view of school leadership’ and, associated with this, leadership development activities ‘which are personalized to [leaders’] individual needs and circumstances’.

By 2010, therefore, the landscape of LMD for schools in England has been transformed. To quote Bolam (2004: 255, 260) again:

However, with hindsight it can be seen that the NCSL was the natural continuation of earlier developments, albeit representing a step change from previous arrangements. As government concerns about the performance of the school system grew, and the relatively ‘soft’ concept of administration was replaced first by ’management’ and then by ‘leadership’, a fragmented system of LMD operating independently of government support or interference had over a period less than 20 years been completely transformed. First, a market system had been introduced, part-funded by and then increasingly steered by the centre, initially through the NDC but then increasingly by the power of legislation to influence provision. Alongside this a number of more interventionist initiatives were experimented with, particularly the NPQH, which was phased in as a mandatory qualification for new headteachers from 2004. Finally, these isolated initiatives provided the basis for the establishment of the National College.

The second part of this article explores how writing about leadership development, especially in EMAL, has attempted to make sense of this experience.

The Functionalist Perspective

In Mabey and Finch-Lees’ (2008: 52) words:

Most of the literature which has been referred to so far embodies an essentially functionalist approach. It does this in three ways. First, it largely accepts that the purposes of LMD can be objectively determined and agreed – for example, through the specification of competences, qualities or skills that effective performance requires – and ‘best practices’ sought. This leads to an acceptance also that the evolution, described earlier, of the assumptions underpinning these purposes and practices from ‘administrative’ to ‘managerial’ to ‘leaderist’ ones represented desirable functional responses to changing society needs. Second, it sees the primary role of research and evaluation to explore the causal relationships between practices and outcomes in order that purposes can be better achieved. And third, it tends to marginalize potential debates about the values that underpin both these purposes and the practices that are designed to achieve them.

These assumptions largely underpin the movement, led by the NCSL, towards a significant increase in the evaluation of LMD programmes over the past 10 years in England. In 2004 Southworth asserted the NCSL’s commitment to external evaluation of its programmes, and since then a very large number of such programme evaluations have been undertaken. Unlike its research studies, the College does not normally publish programme evaluations. Nevertheless, a number of papers derived from such studies have appeared in the academic literature. These include studies of: the New Visions programme for early headship (Bush and Glover, 2005; Bush et al., 2006); NPQH (Brundrett, 2006; Rhodes et al., 2009; Simkins et al., 2009; Crawford and Earley, 2011; Cowie and Crawford, 2009); LPSH (Brundrett, 2006; Luckcock, 2007; Simkins et al., 2009); Fast Track to headship (Churches et al., 2009; Jones, 2010), Leadership Pathways (Simkins, 2009) and Leading from the Middle (Brundrett, 2006; Simkins, 2009; Simkins et al., 2009). These studies almost always report generally positive responses to the programmes. However, the range of data on which they are based varies: some are very small scale, often only reporting the responses of participants; others are multi-method, using some combination of questionnaires, interviews with a range of respondents (not just participants) and case studies of individuals or schools. Studies are rarely longitudinal, nor do they attempt to compare programmes (but see Brundrett, 2006, Simkins, 2009; Simkins et al., 2009). Bush et al. (2007: 74), in a review of 35 evaluation studies for the NCSL, found that these were ‘much stronger in describing what was valued by participants, rather than linking this finding to how and why they work well’.

Perhaps of most concern is that very little of the published work on evaluations is theorized. Occasionally articles include critiques of programmes, but these are often separate from, rather than grounded in, the evaluation data. The dominant theoretical framework used has involved functionalist ‘level’ models of evaluation (Coldwell and Simkins, 2011; Kottcamp, 2011). Drawing on the work of Kirkpatrick (1998), Guskey (2000) and, more recently, Leithwood and Levin (2005), the results of LMD programmes are conceived to occur at a number of levels of impact: on participant reactions, participant learning, leader behaviour and student outcomes. In relation to evaluations of NCSL programmes, Bush et al. (2006) refers to the use of this approach in their evaluation of the New Visions programme, but it is explicated most clearly in Simkins et al.’s (2009) comparative study of three programmes: LftM, NPQH and LPSH, which will be referred to below.

Studies based on functionalist assumptions can be defended from both theoretical and pragmatic positions. Questions of causation are clearly significant if we are to understand development processes better; but pragmatically, it can also be argued that LMD provision designed on functionalist lines needs to be evaluated in its own terms, not least because only such evaluations are likely to be taken seriously by governments. Nevertheless, the functionalist perspective can be challenged from a number of directions. There are significant problems in designing and implementing research or evaluation designs using ‘level’ approaches and these can lead to concerns about the robustness and usefulness of their conclusions (Earley and Evans, 2004; Bush, 2009). More fundamentally, it can be argued that the perspective ‘diverts attention from the important management development processes that include unanticipated outcomes and informal, more covert episodes’, while its over-preoccupation with performativity ignores major ethical and moral questions (Mabey and Finch-Lees, 2008: 74). These issues lead us to explore more constructivist approaches to the study of development programmes.

The Constructivist Perspective

The evolving discourse about appropriate LMD activities, which has been led by the NCSL, has itself challenged simplistic functionalist assumptions with an increasing emphasis on constructivist approaches to programme design. Thus an early emphasis on blended learning, typically combining face-to-face and online provision, sometimes supplemented by in-school activity, have evolved more recently into an increasing focus on ‘personalization’ of provision. This can take many forms. At the most basic level it can simply involve choice in the activities that participants may engage in, through, for example modularization of content or providing a range of optional development activities. More fundamentally, it can emphasize process approaches to development, such as mentoring, coaching, action learning, projects, shadowing, placement or other activities that enable the participant to engage directly, to varying degrees, in negotiating the kinds of practical activities they engage in or the support they receive. At the extreme it may involve claims that programmes are ‘co-constructed’ with participants and providers ‘creating knowledge together’, as for example with the NCSL New Visions programmes (Bush and Glover, 2005; Bush et al., 2006).

These examples imply two further ways in which approaches to LMD have evolved. First, there has been an increasing emphasis in the provision of direct personal support, especially through coaching and mentoring both within participants’ schools and by ‘expert’ leaders and managers such as serving or retired headteachers (Bush, 2009; MacBeath, 2011); and second, much greater use has been made of the school as both a location for LMD and as a significant organizational actor in the process (Simkins, 2009). The latter development emerged initially for reasons of principle: the idea that practice-based activity is an essential aspect of effective LMD. Recently, however, it has also been accepted that school-led work is essential if the scale of the development challenge identified earlier by Southworth (2004) is to be adequately addressed.

Many research studies, too, employ a constructivist perspective, drawing much of their data from interviews and case studies which seek to explore the perspectives of participants and other actors. This work has highlighted a number of themes. The first concerns the degree to which understandings of development activities are shared among actors. For example, some less experienced heads participating in LPSH seemed more instrumental in their expectations of the programme – in terms of enhancing their capacity to do their job – than some very experienced heads who wanted to take personal stock and consider where they wanted to go next. Similarly, on middle leader programmes, such as LftM and Leadership Pathways, participants varied in the degree to which they wished to engage in depth with the ideas they were presented with or do the minimum necessary to complete the programme (Stroud, 2006; Simkins, 2009; Simkins et al., 2009). Such variations in expectations challenge simple functionalist assumptions about pre-specified programme purposes.

However, when our gaze moves beyond the individual participants themselves, the potential for multiple interpretations of LMD activities increases. For example, in relation to NCSL blended programmes, the College, the school and the participant may each have different expectations (Simkins, 2009), while within schools coaches may be unclear whether they are expected to act as ‘support tutors’ to the programme, as mentors whose main role is to support their coachee in their development, or as school advocates whose primary purpose is to ensure a ‘match’ between the development of the programme participant and the school’s needs and priorities (Simkins et al., 2006). Similarly, where participants are expected to undertake in-school activities as part of their programme, the experience may be designed and reflected on very differently if it is seen as a ‘project’ whose primary focus is the needs of the school or a ‘developmental activity’ whose prime purpose is the participant’s leadership learning (Simkins, 2009).

Recasting LMD participants as subjects rather than objects of programmes and activities as the constructivist perspective does, suggests a second theme, namely the need to see development as a longitudinal process. Such a perspective can manifest itself in various ways: for example through concepts of socialization (Crow, 2007); of role transition (Browne-Ferringo, 2003; Simkins and Close, 2009; Crawford and Earley, 2011); or of career stages of leaders in general (Gronn, 1999: ch. 2) or of incumbents of specific roles such as headteachers (Stroud, 2006). Central to each of these manifestations is the issue of identity formation. For example, in their study following up participants on the NPQH and the Scottish Qualification for Headship (SQH), Cowie and Crawford (2009) suggest a number of variables that contribute to the construction of the identity of new headteachers: the role of reflection, the power of the language embodied in the programmes and in the standards on which both are based, and the development of networks of colleague heads. They (2009: 16) conclude:

Finally, and complementing multi-perspective and longitudinal takes on LMD there is a further theme. This decentres the individual learner and the specific course or programme. It considers LMD to be concerned as much with ‘leadership’ as with ‘leader’ development and sees the workplace, and interactions within it, as a key arena for this. Spillane et al. (2009) identify two issues here: the ways in which LMD can support distributed leadership, viewed both in terms of formal top-down distribution and less formal emergent patterns of teacher leadership, and the wide range of opportunities to learn, both formal and informal that occur for leaders in school contexts. There is a large and growing literature on work-based learning, but, although it is increasingly influencing work on teacher development (for example, Hodkinson and Hodkinson, 2005; Burn et al., 2010; Opfer and Pedder, 2011), its influence on the literature on school leadership development has been relatively limited. Turner (2006) has analysed the role of different types of informal learning in the development of secondary subject leaders; while Simkins (2009) has explored the ways in which the individual participant, the programme and the workplace interact in NCSL blended learning programmes, emphasizing the challenges presented by the variation and complexity of local school contexts. In other work on the Scottish Qualification for Headship, Reeves and her colleagues provide a rich analysis of the tensions that arise when attempts to enable teachers to reframe their thinking and develop new identities as leaders through similar programmes come up against both the social and political realities of the workplace, whether these arise from specific school contexts or from broader aspects such as the cultural differences between primary and secondary schools (Reeves et al., 2003; Reeves and Forde, 2004; Reeves et al., 2005).

The Critical Perspective

Not surprisingly, perhaps, much of the writing on LMD in England over the past 40 years has been located within the two consensual paradigms of functionalism and constructivism. For example, in Bolam (1997: 277) suggested that there was:

As suggested earlier, this position, perhaps with the words ‘management training’ replaced by ‘leadership development’, is still the consensus position in much of the literature (see, for example, Bush, 2009).

Nevertheless, there is a powerful critical discourse that challenges these perspectives. There was relatively little hint of this in the ‘era of administration’, despite sporadic, but declining, discussions about the significance of social science theory, and theory more generally, in informing LMD programmes (Hughes, 1982; Bolam, 1997). The critical perspective emerged from a wider policy debate, which has its origins in the ‘era of management’, concerning the purposes of education, the processes through which schools are organized, and the consequences of choices made in these areas for the distribution – and redistribution – of power in schools and beyond. An early hint of things to come was Holmes and Neilson’s (1989) indication of the dilemmas to which the changes embodied in the Education Reform Act 1988 gave rise. They (1989: 68) asked ‘Which client and whose agenda?’, arguing:

Over the next 10 years the debates about ‘managerialism’ grew, both in general terms and in relation to more specific issues such as the competency movement. In 1993, for example, Ouston (1993: 214) critiqued this movement, distinguishing between ‘fundamentalist’ approaches which focus on skills and performance and ‘liberal’ approaches that focus much more on personal qualities, concluding that a narrow competence-based approach does not take account of the fact that managers work in an integrated way, and underpinning values are crucial, as are aspects of individual psychology – ‘what has become known as “personal effectiveness”’.

It could be argued that such critiques were essentially constructivist in their underlying assumptions. The ‘era of leadership’, however, has seen such developments examined much more thoroughly from a critical perspective. For example, O’Reilly and Reed (2010, 2011) have coined the term ‘leaderism’ for the core legitimating narrative of recent public sector reform. They (2010: 962) argue that replacing the instrumental rationality of managerialism with value rationality of ‘leaderism’ addresses continuing concerns about the limited effectiveness of improvement strategies by ‘making the issue the establishment of a passion for a common goal between leaders and led’. In relation to education, Gronn (2003), for example, has developed the idea of ‘designer leadership’ to describe standards-based approaches to the selection of development of school leaders, while Gunter and Thomson (2009) have explored the metaphor of the ‘makeover’ to characterize the approaches to LMD that have been emerging. The underlying idea here is that policy discourses and associated technologies – such as the NCSL’s online presence (Thrupp, 2005) – serve ‘to constitute actors [in this case leaders] in particular ways, which preference particular forms of judgement and which are based on particular forms of power and knowledge’ (Newman, 2001: 30; see also Newman, 2005; Gunter and Forrester, 2008). This means, for example, that:

Detailed, grounded applications of these ideas to LMD in education are rare. An exception is the work of Luckcock (2007, 2008), ⁹ who deconstructed the assumptions underpinning the design of the LPSH, arguing (2008: 545) that the programme:

Writing from the critical perspective has been dominated by such post-structural approaches that emphasize the power of discourse and processes of identity construction. Nevertheless, there has been little evidence in the journals reviewed here of socially critical approaches to leadership development which address either general issues of social justice or more specific questions concerning, for example, gender and ethnicity. There is nothing to compare, for example, with some US studies that explore in a grounded way how such perspectives might be effectively embedded in leadership development activities, the kinds of alternative leader identities that this might imply and the barriers that exist to such developments (for example, Rusch, 2004; Brown, 2006; Young et al., 2006; Collay, 2010; Furman, 2012).

However, it is important not to lose sight of more concrete manifestations of power in relation to developments in the LMD field. Who controls LMD? – the government and its agencies? Local authorities? The universities? The profession? Or private organizations and individuals? And what are the levers and mechanisms through which this control is achieved? Broad trends here were identified earlier: the first two eras were associated with fragmentation in provision among a range of providers, but with LEAs dominating short course provision and universities award-bearing provision. This reflected the dispersal of power in the education system under the bureau-professional regime. When central government entered the arena through the provision of funding, it was initially to encourage a market in provision. However, this market was increasingly ‘steered’ as radical government policies, introduced during the ‘era of management’, created a climate where demand for provision was driven by pressures to respond to central initiatives and imperatives. Eventually, the central government increasingly reverted to more direct intervention through the creation of the TTA and then the NCSL, which is now the dominant actor in the field.

Concerns were expressed about NCSL dominance from an early stage (Brundrett, 2001; Bush, 2004; Crow, 2004). In his historical review in 2004 Bolam reflected on the temporary nature of previous developments at national level and wondered whether the political will to embed the College in the national scene would be maintained. Now (in 2011), with the College’s position reconfirmed by the current Conservative—Liberal Democrat government, its position seems secure – and, furthermore, securely located within the ambit of central government. This has, in fact, always been the case, as the annual ‘Remit Letter’ from the Secretary of State setting out his expectations for the year ahead has always made clear. However, it has been reinforced by the College’s transformation in April 2012 from an ‘arms length’ non-departmental public body to an Executive Agency of the Department for Education in order to be ‘directly accountable to ministers’ (NCSL, 2010).

Nevertheless, the issue of changing power relationships is not a straightforward one. It is undoubtedly the case, as already suggested, that the College cannot possibly meet the scale of need implied by the demands of ‘leaderism’ across the whole educational system. Indeed, it has stated its intention to modify the way it works, moving from direct commissioning to a greater emphasis on brokerage, facilitation, accreditation and quality assurance (NCSL, 2011). A less directive philosophy might also be indicated by the ministerial decision that holding the NPQH will no longer be a mandatory requirement for headship from 2012. However, it seems likely that the level of ‘steering’ that this will involve will continue to be high. Furthermore, its provision is dependent on an army of lieutenants comprising private companies, individual consultants, and schools, all of which play active parts in the delivery of its programmes. Wallace and his colleagues (Wallace, O’Reilly et al., 2011; Wallace, Tomlinson et al., 2011) have explored the degree to which the NCSL and other national public sector leadership development agencies act effectively as ‘conduits for reform’ through these mechanisms. Their conclusions are complex. On the one hand, evidence was found of the ability of intermediate actors – leadership developers, headteachers – to mediate the ‘message’ that NCSL programmes propounded. Conversely, it was clear that ‘their power to mediate … was nevertheless delimited by the imperative to demonstrate compliance under the relevant accountability regimes’ (Wallace, Tomlinson et al., 2011: 278). In other words, the power of national leadership development initiatives to ‘acculturate’ leaders into the discourse of reform lies less in the programmes themselves than in the wider policy discourse and the incentives embodied in the accountability structures with which they are associated.