British Educational Research Journal

Introduction

The recruitment and retention of teachers has become a perennial policy concern within the UK. While the ‘crisis’ in teacher recruitment has not been as pronounced in Wales as in other UK countries (Davies et al., 2016), there has been increasing concern in relation to the recruitment and retention of Welsh head teachers (HTs) (BBC, 2016; Flint, 2016). Rather than ‘misrecognising’ Welsh policy (Power, 2016) by treating it as a replication of issues in other areas of the UK, this article contextualises this recruitment shortfall within the altered professional role and identity of Welsh HTs. The study shows how these changes in HTs’ professional practice are framed by the wider social, cultural and, particularly for this study, political context of Wales. The article illustrates how these debates are more intense within a small education system undergoing rapid educational reform, where the previously elite professional status of HTs is being increasingly challenged.

The changing professional role and identity of the head teacher

The restructuring of professional work has been both critiqued and embraced: some identify this with a general ‘deprofessionalisation’ (Ball, 2003), while others argue that this is merely a rearticulation, often categorised as ‘reprofessionalisation’ (Noordegraaf, 2007). The causes and consequences of such restructuring have been outlined by Freidson (2001), who traces the paradigmatic shift in how professional work is conceptualised as the ‘logics’ of managerialism and commercialism ‘encroach’ upon the professional work organisation. Freidson categorises such logics as ‘organisational professionalism’, which manifest themselves in: the proliferation of bureaucracy; attempts at standardisation; moves towards hierarchical structuring within professional work organisations; and techno‐rationalist understandings of knowledge, associated with technicist approaches to practice.

From a deprofessionalisation perspective, the results have been the ‘proletarianisation’ of professional work under ‘neo‐liberal governance’ (Collet‐Sabe, 2017). This governance is applied through regimes of accountability and ‘performativity’ (Ball, 2003), which are antithetical to the discretion, autonomy and collegial practices within horizontally structured organisations which form the basis of classic or, as Whitty (2008) suggests, idealised accounts of occupational professionality. A key element within such accountability regimes is governance through data (Ozga, 2009) or, as Ball (2015, p. 299) argues, the ‘tyranny’ of numbers. It is claimed that such governance works not only at state level (Neave, 1998), but also through international educational assessment characterised as the ‘hegemony of PISA’ (Collet‐Sabe, 2017, p. 143).

Parallel to this intensification of accountability have been decentralisation reforms which have provided HTs with more autonomy and responsibility. Within this paradoxical or ‘oxymoronic’ (Wilkins, 2014, 2018) position, HTs are seen as both ‘determinant agents’ of success/failure and implementers of policies which ‘dis‐empower and mistrust them’ (Collet‐Sabe, 2017, p. 148). The expansion of responsibility has resulted in the locus of headship expanding to include budgeting, human resources, management of staff and the school's physical spaces, moving the HT role beyond its ‘core’ function of leading teaching and curriculum development (Brundrett & Rhodes, 2011). A consequence of these changes to HTs’ professional practice has been the intensification of their perception of risk and heightened individual reflexivity (Beck, 1992). This risk perception has been accentuated by contemporary hyper‐anxiety in relation to children's lives and safety (Connolly & Haughton, 2017).

The de/reprofessionalisation binary has been superseded by research which has considered the interplay between competing understandings of professional practice categorised as occupational and organisational professionalism. Moving away from a deprofessionalisation perspective, Evetts (2012) traces how organisations draw on these competing understandings, while Noordegraaf (2011, 2014) describes roles where organisational and bureaucratic logics are incorporated into professional practice—categorised as ‘hybrid’ or ‘organised professionalism’. Such an approach emphasises the agency of professionals to endorse, resist or subvert these changes (see Biesta et al., 2015; Connolly et al., 2018 or, in relation to HTs as ‘agentic dissidents’, Courtney, 2017). Priestley et al.'s (2015) reflection that education systems differ in their fostering or, conversely, limiting of teacher agency is particularly apposite in a UK policy context where, following devolution, approaches to teacher professionalism have diverged (Davies et al., 2016). This stress on agency has been emphasised by academics, who see it as an essential component within wider school reform (Hargreaves & Fullan, 2012), or as forming the basis of alternative forms of professional capital (Fullan et al., 2015).

The Welsh context

Since 1999 and the devolution settlement in the UK, responsibility for education policy in Wales has resided with the Welsh Government (WG). The post‐devolution education system has differentiated itself from other parts of the UK—especially its closest neighbour, England (Rees, 2012). These policy divergences reflect political differences between the two countries where the ruling Welsh Labour Party has eschewed the rhetoric of choice and market‐driven reform (Power, 2016). The relative uniformity within the Welsh system is a reflection of the WG's commitment to a comprehensive system (Power, 2016), which has resulted in less polarised or, from a critical perspective, ‘stretched’ educational outcomes (OECD, 2014).

This Welsh ‘alternative’ extended to the relationship between government and professional work generally. According to Power (2016, p. 286), the WG's approach to professional work rejected the ‘challenge’ rhetoric of England, espousing what she suggests is a discourse of ‘trust’ and ‘collaboration’. However, this uncritical acceptance of the Welsh approach has given way to a countervailing narrative that the Welsh education system was ‘failing’ (Dixon, 2016; Evans, 2016). This failure narrative draws on a number of sources, most especially PISA rankings, where Wales has consistently performed less well than other UK countries. The post‐2009 PISA crisis narrative resulted in the reformist Education Minister, Leighton Andrews, adopting a suite of educational reforms aimed at systemic improvement of the Welsh system (Andrews, 2011a,b). While the WG continued to invoke Power's (2016) ‘discourse of trust’, these reforms leveraged accountability and challenge in the Welsh system, which, as Andrews claimed, needed a ‘wake‐up call’ (Andrews, 2014, p. 108).

The alarm for Andrews’ ‘somnolent’ system was provided by a series of reforms set out in a 20‐point plan for improvement (Andrews, 2011a), which (re)introduced national tests for numeracy and reading from Year 2 through Key Stages 2 and 3; baseline assessments within the Foundation Phase; and instituted a form of school grading—banding—for secondary schools. When introducing the banding system, Andrews—using what Dixon (2016, p. 42) refers to as ‘linguistic contortions’—claimed that it was not about ‘crude league tables’ and naming and shaming schools but, instead, concerned with identifying schools in need of additional resources and support. Following widespread criticism of banding, a reworked ‘traffic light’ system of categorisation was extended to Welsh primary schools in 2015; however, the discourse within government literature continued to counter the rhetoric of accountability, instead suggesting that schools were categorised according to the ‘level of support’ they required.

To facilitate these school‐level reforms, the WG initiated a series of structural changes to the Welsh education system. Fundamental to this was the establishment of a School Standards and Delivery Unit within the WG, and the reconfiguration of the delivery of school improvement services from 22 local authorities (LAs) to 4 regional consortia.¹ According to Andrews, these accountability mechanisms were essential intermediary bodies in a low‐capacity system moving from ‘fair to good’ (Andrews, 2011a).

However, there has been an increasing recognition that while accountability and challenge have been introduced into the Welsh system, the support needed to meet these increased challenges has been missing (Hill, 2013; OECD, 2014, 2017). This has been compounded by meso‐level bodies employing ‘parallel efforts and strategies’ (Hill, 2013; OECD, 2014): the central tension was between the newly formed consortia, which feed indirectly into the WG in relation to standards and performance and the LAs, whose role and remit have become less clearly defined and understood. While the WG has provided more clarity in relation to responsibilities (Welsh Government, 2014), the meso‐level educational landscape continues to evolve.

School leadership and recruitment in Wales

Within Welsh school improvement initiatives there has been increasing emphasis on leadership as the essential and necessary catalyst for change. This has resulted in a revision of the Leadership Standards in Wales in both 2011 and 2016. Following the Hill Report (Hill, 2013), the National Leadership Development Board (NLDB) was formed and a revised model for the National Professional Qualification in Headship (NPQH) initiated. The ongoing development of leadership capacity or capital has been emphasised in both OECD reports (OECD, 2014, 2017), which highlighted issues around succession planning, professional development opportunities for senior and middle leaders, and capacity within the Welsh system for leadership. Consequently, in 2016, the cabinet secretary Kirsty Williams announced the development of a National Academy of Educational Leadership, an ‘arm's length’ body with an initial focus on headship issues, with the intention of expanding its remit to address support for all leaders within the education system.

Many of these recent reforms of headship were, and continue to be, precipitated by the perception that Wales was experiencing a ‘crisis’ in headship recruitment. The ‘crisis’ narrative emerged from the high levels of HT vacancies observed across Wales, but was especially acute in certain areas and contexts (BBC, 2016; Flint, 2016). The precise measures which might indicate a recruitment ‘crisis’, or deficit, have been the subject of debate (MacBeath et al., 2009, p. 407). Nevertheless, key indicators typically used to quantify the vitality or otherwise of HT recruitment include the number of unfilled vacancies, applications received per vacancy ratios and re‐advertisement rates (MacBeath et al., 2009).

An analysis of such indicators does suggest some cause for concern in Wales, if not the sort of acute ‘crisis’ that characterises lively public debate. A report published by NAHT Cymru in 2016 presented data relating to HT recruitment in Wales relating to the period from 1 September 2014 (based on Freedom of Information requests to all 22 Welsh local authorities, 18 of which provided data). This report examined each of the key indicators of recruitment vitality outlined above, and drew attention to comparatively high rates of HT vacancies (as a proportion of all HT posts in each authority) in three LAs: Monmouthshire (42%), Cardiff (35%) and Carmarthenshire (30%) (NAHT Cymru, 2016, p. 8). The re‐advertisement of posts, and the repetition of the recruitment cycle due to a failure to recruit or attract applicants, is a common feature of HT recruitment in Wales, and all but five LAs, for whom data was available, reported that they were forced to repeat recruitment processes due to a failure to recruit HTs in the initial phase (NAHT Cymru, 2016, p. 10).

The number of applicants per post in Welsh local authorities would also appear to be a cause for legitimate concern, with all but six local authorities, for whom data was available, noting that the majority of HT recruitment processes they had initiated over this period had failed to attract more than five applicants (NAHT Cymru, 2016, p. 10). Across Wales, the last five years have witnessed a significant fall in the average number of applications per post for HT or deputy positions (without a teaching commitment): in 2012, a Wales‐wide average of 29.5 applications were received per post, compared with 9.6 in 2015 (Education Workforce Council, 2017a, p. 17). In the secondary and middle phases specifically, an average of just 5.6 applications per post were received for headship or deputy vacancies (with no teaching commitment) in 2016, compared with an average of 18.5 applications in 2014 (Education Workforce Council, 2017b, p. 14).

Methods

A qualitative methodology was employed for this research study, with data drawn from (n=30) semi‐structured interviews with HTs and deputy HTs throughout Wales (identified in the data as P‐HT for primary heads and S‐HT for secondary heads, and similarly with deputy heads as P‐DHT and S‐DHT). Two sampling methods were used: the first level of sampling was based on a wider study involving a sample of schools throughout the country (n=12), participating in a multi‐cohort longitudinal research project which has been surveying pupils at 12 schools across Wales since 2012. HT interviews were conducted at all the case study schools, following the last round of school data collection in 2017. These case study schools serve a range of different communities, including urban/rural and advantaged/disadvantaged areas. They include Welsh‐medium, English‐medium and bilingual schools, and the size of the pupil population varied considerably. This range of schools enabled us to consider the particular issues facing head teachers in different settings, and the ways in which the roles of head teachers might vary within these institutions. Eleven of the participants were head teachers, and one was a deputy head. The interview schedule was semi‐structured: it asked participants about their views on the challenges facing schools and on recent Welsh education policies. This provided context for us to understand their responses to the following questions about their perspectives on the role of the head in Wales, and the possible barriers to the recruitment of head teachers in Wales.

The second level of sampling was purposive and designed to address a specific issue which had arisen from the preparatory review of contextual data, namely the higher rates of HT vacancies in some local authorities (Teddlie & Yu, 2007, p. 77). Based on the NAHT Cymru data (NAHT Cymru, 2016, p. 8), schools were sampled from within those LAs with the highest percentage of HT vacancies: Monmouthshire (42%), Cardiff (35%) and Carmarthenshire (30%). Within these three authorities, (n=10) HTs and (n=8) deputy HTs in nine primary and two secondary schools. Within the second phase of sampling we attempted to create a representative sample by school phase, language of instruction and socio‐demographics. Of the total number of schools (n=23), five were Welsh‐medium and the remainder were English‐medium. For the interviews we used a semi‐structured interview schedule which began with questions in relation to respondents’ professional role and identity before we asked them to consider broader questions in relation to political changes and reforms. All interviews were conducted between January and July 2017. We recognise that there was a potential for sample bias, as some of the more experienced head teachers within phase two of our sample may have been particularly critical of recent system reforms. As a qualitative study which drew on 30 head and deputy head teacher interviews within a system which has over 1,600 schools, we make no claims in relation to the wider demographic generalisability of phase two of the sample (Table 1).

All interviews were recorded, with the permission of participants. They were then transcribed and, in the case of the Welsh‐medium transcripts, also translated. Data was analysed using a graduated set of coding protocols, based on the typology outlined by Miles et al. (2014), moving from initial ‘close‐to‐data’ descriptive codes through to inferential codes, identifying broader themes. The research team met at key junctures during the coding process to maintain shared and co‐constructed understandings of these themes, and to agree the directions of the analysis as it progressed from specific units to overarching themes, and finally to broader theory building and abstraction. Within this final coding phase, we tried to eliminate bias from our analysis by closely interrogating both the data and our own positionality (Richards, 2005): for example, one of the team was a former acting head teacher and deputy head teacher so we wanted to ensure that our interpretation was not coloured by a romantic or idealised (Whitty, 2008) view of headship. Through this process we identified seven key themes and organised our data into seven documents under each overall theme. These were:

1. Accountability
2. Changing attitudes towards professionals
3. Changing nature of accountability and performance
4. Headship role
5. LA/consortium support
6. Leadership development
7. Mentoring
8. Risk

We then scanned the data under each heading to ensure that they reflected the key themes and discussed which were the main messages emerging under each one, enabling us to begin planning our article's analytical structure. While the data extracts we have chosen are particularly insightful or engaging, each was chosen to illustrate a theme that emerged from our coding and reflected the views of the majority of respondents within the sample (Richards, 2005).

Findings

The data illustrate that the professional work and identity of HTs has undergone profound change in Wales in the past decade: the system has been, as the former Education Minister Leighton Andrews predicted, ‘woken up’. A key element of this change—and one of the objectives of the restructuring of Welsh education—was the leveraging of layers of external accountability into the system:

the job has become more – the accountability side has got larger, looms larger. I think it's always been there but it's certainly talked about more it's about that standards agenda. (P‐HT)

the level of scrutiny is … is just unbelievable, isn't it? (S‐HT)

We are very, very driven now by system and process and the accountability that we quite rightly have … They are very demanding of paperwork and of targets and of data and one of the things that's changed over the years is data and the amount of data and the accountability to that data. (P‐HT)

These extracts illustrate that, from a HT's perspective, the changes in their professional role have been the result of external accountability mechanisms and the data generated within these. From a deprofessionalisation perspective, this ‘governance’ (Ozga, 2009) would be evidence of the ‘colonisation’ of HTs’ professional work by managerialist logics. However, the participants were clear that they were not resistant to the concept of accountability in itself, though this was not exercised through external scrutiny but through collegial approaches within an occupational approach to professional practice:

I would say I've always felt a sense of accountability for the standards in the school. I don't think that's changed. I think the way that it's been measured has perhaps changed … I think the way that the process in which schools are supported and challenged has changed over that time, but there's always been an accountability. (P‐HT)

The myriad of overlapping external accountability mechanisms were seen to be inconsistent between various agencies, and the HTs claimed that the data generated was often used to judge/sanction rather than support them. In addition, HTs argued that the accountability structures were volatile, resulting in them having to adapt continually to new structures which, in turn, made planning particularly challenging. The sense of frustration was apparent in the use of a sporting metaphor to compare new accountability structures to a game, and the metrics used to judge success as ‘moving goalposts’:

… there is very high stakes accountability which the goal posts keep changing a lot, don't they, you know? (S‐DHT)

So, sometimes I feel that with categorisation the goal posts aren't so much being moved, I think they're on castors and they're being rolled about … around so they miss the ball all the time. (P‐HT)

There was also scepticism about what was being measured, with the suggestion that ‘not everything that can be counted counts and not everything that counts can be counted’ (S‐HT), so that accountability frameworks were governed by what could easily be assessed, or what was meaningful only in the wider meso‐level or political contexts—echoing Strathern's (1997, p. 308) claim that ‘when a measure becomes a target it ceases to be a good measure’. In HTs’ views, this was detrimental to other contextual aspects which are more difficult to evaluate, such as children's ‘aspirations and self‐belief’ (S‐HT).

From these HTs’ perspectives, the ephemerality and volatility of these accountability mechanisms were compounded by the overlapping layers of meso‐level scrutiny: these include Estyn,² the LAs and the regional consortium. Despite attempts by the WG to clarify the roles of each body, it remains unclear to our participants where each body's function and responsibility rests, in particular the relationship between the LA and the consortium:

You're almost seen to be having to sort of battle against the region [consortium] and the local authority and then, you know, the political party in power in your local authority as well … and you have to decide who to listen to sometimes. Because they do sometimes give you conflicting opinion. (S‐HT)

We've had the Consortia, which is another tier if you like of leadership bureaucracy so we get the Welsh Government, we get the Consortia, and we get the Local Authority … it's taking you away from the … our core purpose. (P‐HT)

There was also a lack of clarity about where the support and sanction were located within the various accountability structures, with many heads suggesting that, initially, meso‐level structural changes introduced too much sanction into the system. This was particularly true in relation to the consortia, whose remit is both to ‘challenge’ and ‘support’; however, the HTs believed that the support element within the remit was not fully developed and that challenge was often seen as a sanctioning ‘stick’:

Constant challenge … constant, constant challenge. (P‐DHT)

I don't know how the Consortium can do challenge and support at the same time, because you're talking about the same people … (P‐DHT)

HTs emphasised how they were inundated with short‐term and fast‐paced initiatives introduced to meet changing regional and national policies, with seemingly insufficient consideration given to how these would fit into the local context and interact with existing initiatives. The outcome of this was a limited time for embedding good practice, evaluation and reflection:

I think that there's absolutely no opportunity for schools to build on the good practice that they've got. I think that they are so inundated with national and local initiatives, that some of the good practices that they have got then go on the back burner. (P‐HT)

While these accounts do suggest that these external accountability mechanisms were ‘disciplinary tools’ (Ball, 2015), participants did not represent themselves as passively accepting changes which conflicted with their core beliefs and values. Some of the more experienced HTs insisted on exercising agency in resisting initiatives (Courtney's, 2017 ‘agentic dissonance’), based on their embedded and local contextual knowledge of what was appropriate for their local community, school, pupils and staff:

We've got about 10 initiatives going on [but] I know we can say no to things and I think that's crucial as well in the role because I think they can throw everything at you and you need to pick and choose what's best for your school because fundamentally as a head, you know what's best for your school and your children. (P‐DHT)

As well as accountability at a meso‐level there has been increased public accountability driven through the publication of ranking under the various articulations of the school categorisation system. While the revised model does attempt to take a highly contextualised approach, many HTs within schools—including those ranked high (green)—criticised it for being a blunt measurement that lacked contextual subtlety or nuance:

because of the accountability factor, schools … are judged harshly through the categorisation system, which again is I think absolutely a negative, it's having a really negative impact on schools, and recruiting school leaders into certain schools in Wales. (P‐HT)

… if you are in a green school you think, I might be better off just sitting here waiting until potentially the head may retire in four years’ time … the stakes are quite high and it tends to be very personal, they tend to be more conservative really about what they apply for. (P‐HT)

This participant's claim that choice is ‘personal’ reflects the heightened perception of risk and hyper‐reflexivity (Beck, 1992) experienced by HTs. This fed into both existing and aspiring HTs’ strategising in relation to applying for posts, as they were unwilling to expose themselves to the risks and scrutiny of leading a ‘red’ school; conversely, the metric of continuous improvement embedded within the categorisation system intensified the risks and precarious nature of being held accountable for maintaining standards and ranking in a ‘green’ school. All participants agreed that public categorisation also led to a general undermining of the professional standing of teaching and a resultant loss of esteem for the profession:

the public categorisation has been disastrous for those schools which are amber and red. I [in a green school] don't get parents questioning my decisions or my policies or strategic plans. If you are in a red school or an amber school it's almost like a public statement and so as far as parents and other people they always question absolutely everything. (P‐HT)

Within this account the HT rejects the claim that Welsh school categorisation's primary aim was to provide support for struggling schools and aligns it more with the ‘naming and shaming’ which its initiator disavowed (Andrews, 2011a). This intensification of risk resulting from external and public accountability has militated against deputy heads seeking promotion:

people don't want to be part of the blame culture and to be sued – people don't want the hassle and the level of personal responsibility … (P‐HT)

It is not just risk in relation to academic results but, also, the intensification of risk generated by child protection and safeguarding protocols which rest with the HT. This hyper‐sensitivity to risk is accentuated by a perception of the highly litigious society in general and parents specifically. This can be limiting for professionals, where they feel vulnerable within a culture of blame:

We have senior staff in schools put off by the accountability regime – the scrutiny yearly in terms of data but also the safeguarding of a child is becoming more of an issue – it's difficult to sleep at night with these situations. (S‐HT)

Deputy Heads … they're shielded from some of the huge responsibilities, certainly around safeguarding, you know, it's only one little … one incident and that's enough to get somebody on the front page of the Echo … (P‐HT)

The reluctance of deputies to move into leadership roles has been outlined previously (MacBeath et al., 2009), but as the headship role changes and the risks associated with headship intensify, this has become more pronounced. The reconfiguration of the HT role has resulted in a separation between the traditional locus of responsibility for deputy and head:

many years ago the job was around teaching and learning. The job now is a Business Manager … many deputies they fall flat on their face … And they say: ‘Well, hang on a minute now, I'm … I'm all about teaching and learning’. (P‐HT)

This participant illustrates a perceived disconnect between the professional practices and identities associated with headship and deputy headship. The latter is characterised by a personal commitment to, and excellence in, teaching and learning in practice, while headship is increasingly aligned to managerialist and bureaucratic logics: it has moved towards ‘organised professionalism’ (Noordegraaf, 2011, 2014).

This reframing of headship has not only resulted in a misalignment between the role of deputy and headship, but has also resulted in a disjuncture between many of the head's core professional values—rooted in an occupational account of professional practice—and the managerialist role they inhabit as managers of organisations, aligned with an organisational account of professional work. One of the consequences of this altered professional work was the movement away from working with people (adults and children) towards following systems and processes and working with and performing to data:

We're about young people, developing them and I've got a passion for developing staff and ensuring that staff perform to the best possible potential that they can … That's the difference, I think that there's a lack of emphasis now on people, people skills. And that's created by the over‐emphasis on system and process. (P‐HT)

This move towards an ‘organised professionalism’, governed through process and data generation, appears to be fundamentally changing the nature of the job and, importantly for this study, the type of person who wishes to move into headship:

… looking at colleagues coming in now, they're far more business managers and they manage a business, they're like business directors. And of course, I've had to adopt along those lines … adapt along those lines myself. But they are a very different breed coming through. (P‐HT)

The perspectives outlined above pose a question about the traditional manner in which succession planning in schools is organised, often through the appointment of an excellent classroom teacher or year leader to the role of deputy headship (if being a deputy is seen as preparation for headship). Within an organisational account of headship, the skills and aptitudes needed are more of a generic managerialist nature, rather than those of leading teaching and learning. In the following extract this HT claims that their former training as a corporate manager had a closer relationship to their HT role than HT training:

I can absolutely be confident that I draw more of my experience … for some of the tasks that I'm asked to do now than I ever did on any training that I had as a Head Teacher. You know, things around the budget, things around policy and procedure. (P‐HT)

This position was supported by heads who outlined how the role had mutated into that of organisational leadership and management:

They want strategic. Business manager. (P‐HT)

You need to be business minded. (P‐HT)

They want strong managers – who micro‐manage everyone – but we haven't got the tools to do that – we can't hire and fire. (P‐HT)

This move away from the locus of teaching and learning is not only due to increased managerialism, but to the multifarious responsibilities within the HT role. These multiple roles have moved some of the heads into a category of work that is not consistent with their own perception of what the core purpose of headship is, or ought to be:

We have to be social workers, bouncers, health and safety managers. (P‐HT).

I still think there is a huge piece of work to be done about freeing up Head Teachers … for example, what I've done over the last week … I have, seen a man about a broken heater upstairs because it keeps sparking and the plug's broken. I've had a conversation with a painter and decorator. I have, seen somebody this morning because there's a tear in the gym mat and we're worried about health and safety. Not that the health and safety is not important, but actually I don't think the Head Teacher should be being bogged down with things like that. (P‐HT)

These wider societal issues that result in the heightened perception of risk are compounded by structural issues where the support mechanisms which may have been formally in place within LAs are less accessible or available. As a consequence, HTs are responsible for a range of child‐ and site‐centred issues which are often beyond their area of expertise. To mitigate some of these vulnerabilities, many of the participants suggested that an experienced, external mentor would be able to provide pastoral support and a ‘safe space’ outside the formal educational structures and accountability mechanisms:

I definitely think there should be a formal system of mentoring which is a special relationship – Head Teachers would then have a confidential relationship with an experienced Head Teacher who could then be signposting them. (P‐HT)

In a system with multiple accountability structures, the data emphasised the need for mentors to be impartial and non‐judgemental, providing ‘safe spaces’ where HTs could seek both practical support and pastoral advice, without bringing additional scrutiny upon themselves:

You want someone you can ask stupid question to and not be judged. It's daunting. (P‐DHT)

I think the National College, if it takes off, is really important and I think that anybody who is aspiring to headship in Wales should at least have access to various modules through the National College of School Leadership. (P‐HT)

Some of the HTs and deputies identified the introduction of the National College for Educational Leadership as a welcome step, hopeful that it will bring a more coherent approach to leadership support across Wales.

Discussion

The data represents HTs’ accounts of the changing nature of their role in Wales. Of course, this could be interpreted from a ‘professionalisation’ perspective, which seeks to excavate the ideological components within claims to professionalism which can serve to protect that groups’ occupational interests (Ozga & Lawn, 1981). From this perspective, these HTs’ accounts could be dismissed as a rearguard action from a privileged professional group to protect their occupational boundaries. This approach could be applied to the Welsh context in particular, where there was formerly a system of small LAs within which local HTs had considerable power and influence. However, while claims to professionalism certainly involve social contestation and struggles often motivated by self‐interest, the reduction of professional practice to only ideological struggle driven by self‐interest results in a relativism which disallows a consideration of the value of the professional work of HTs (Evetts, 2012).

Such value is often realised through professionals using their discretion and exercising their agency. Biesta et al. (2015) and Sachs (2016) argue that educational systems can either foster or stifle teacher agency. The data from this study suggests that the more dirigiste system introduced post‐2011 within Wales reconfigured HTs’ role and identity and, as a result, their sense of agency. While this may not have been an intentional outcome of the reforms, it was a predictable consequence of increased centralised control and accountability. Although the system adopted in Wales falls short of the highly prescriptive educational policy decried by Sahlberg (2011, p. 99), it did introduce top‐down approaches and accountability mechanisms (Harris et al., 2018, p. 2).

These increased accountability structures have reframed the professional role and identity of Welsh HTs, moving headship towards being an ‘organised professional’ role (Noordegraaf, 2011). While this was endorsed by some HTs, most found this anathema to what they considered the core role of headship, which they articulated within an occupational account of professional work. HTs in Wales are not the first to experience the recalibration of the professional work of headship as a consequence of reforms which intensified accountability and scrutiny. However, there are a range of contextual factors which make the experience of Welsh HTs unique: the speed of change; the proliferation and disconnected nature of reforms; the overlapping accountability structures; and the reduction of support structures for HTs at a local level. While these factors may exist in other educational contexts, they have been exacerbated within what is a relatively small system.

One of the consequences of these reforms was that HTs had an intensified and personalised sense of responsibility and risk (Beck, 1992; Connolly & Haughton, 2017). This risk emanated from a hyper‐sensitivity to issues around protection and litigation but, more acutely, levels of public accountability. All the participants rejected the disingenuous argument (Dixon, 2016) that the school categorisation system adopted in Wales was aimed at channelling support and identified a number of negative consequences of this level of public scrutiny. Of particular relevance when considering a recruitment ‘crisis’ is how this intensifies the risk of becoming a head, not only of an underperforming, but also a highly performing school.

Categorisation was one of a raft of reforms precipitated by the disparagement of the Welsh education system as a result of a universal PISA ‘fetishisation’ (Harris & Jones, 2017). These reforms were both ambitious and relentless, with the OECD itself identifying ‘reform fatigue’ within Wales (OECD, 2017, p. 44). The accountability mechanisms embedded within these reforms, and the data generation which accompanied them, intensified vertical work structures within schools and, increasingly, isolated HTs at the apex. Within their newly configured hybrid role, HTs’ work was increasingly driven by the need to generate, analyse and account for data. However, when accountability mechanisms and data generation become the sine qua non of the work of headship, many of the talents of these heads are lost (Hoyle & Wallace, 2005). While some HTs endorsed this hybrid identity, research suggests that where leadership has its greatest impact is in schools where leaders have direct and sustained involvement in what most of these heads considered to be their ‘core’ function—professional development and curriculum planning (see, for example, Bush & Glover, 2014).

If headship continues to move towards organisational professionalism, with HTs being de facto business managers at the head of a hierarchical organisation, many of the practices for induction into headship become anachronistic: for example, classroom‐based and child‐focused skills and dispositions, which are often used to recruit deputy heads, would be redundant. The data also suggests that the reconfiguration of the HT role as a hybrid organised professional (Noordegraaf, 2011) has resulted in a disconnect between many HTs’ core occupational values and, in some cases, moral purpose. Where educational leadership is reduced to educational ‘management’, there is a danger that moral purpose can be sacrificed for technical efficiency (Bush & Glover, 2014).

To foster agency and develop professional capital (Hargreaves & Fullan, 2012), the promotion of mentoring opportunities for HTs is imperative. Lejonberj et al. (2015) highlight the need for mentor development programmes that challenge mentors’ existing belief systems. In high‐stakes educational contexts, the professional learning needs of external mentors may benefit from being built around principles outlined by Daly and Milton (2017), which include harnessing diversity, working with an inclusive design, problematising notions of consistency, exploring complexity and developing an inquiry orientation, in order to support HT mentors in meeting agentive goals with their mentees.

Biesta et al. (2015) correctly argue that successful schools and systems promote the agency of teachers. However, the data within this study suggests that, from a Welsh HT's perspective, the post‐2011 reforms were delivered to, rather than with, educational leaders, thus reducing their sense of agency. This was compounded by the perception that these reforms disregarded culture and context, both at school and, as Harris and Jones (2017) argue, at system and country level. Classic accounts (Rhodes & Brundrett, 2009) and most recent research (Myers & Darwin, 2017) have illustrated how effective school leaders are sensitive and responsive to the cultural and wider contextual elements which might influence any intervention. Recognition of culture and context relies on promoting the agency of HTs and creating stability within the system to allow HTs to make decisions in the best interests of their schools. Such changes would foster not only agency, but also collaboration within the Welsh system: what Hargreaves and Fullan (2012, p. 9) describe as a move from power over to power with fellow professionals.

Conclusion

This article has considered the changing professional role and identity of HTs in Wales. It illustrated how political changes which introduced increased accountability into the Welsh education system recast this role as ‘organised professionalism’. While some HTs and deputies adapted and endorsed these changes, many argued that these conflicted with their core values and understanding of headship. In addition, these changes intensified these HTs’ perceptions of personal risk. While the study was specifically about Wales, the findings have resonance with reforms to the head teacher role in other devolved areas of the UK: for example, Scotland has had longstanding issues in relation to head teacher recruitment and retention (Christie et al., 2016) and is experiencing similar political changes at a meso‐level with the introduction of Regional Improvement Collaborations as well as changes to the professional training and development of head teachers within the Scottish College for Educational Leadership.

One of the consequences of these changes has been difficulty in recruiting and retaining HTs. This can be explained in a number of ways. Firstly, there are higher attrition rates where the professional role and identity of HTs no longer align with existing heads’ professional values, and where there are increased accountability pressures. These pressures have been intensified by the volatility and instability of accountability mechanisms and the lack of clarity in relation to the remit of various agencies. As well as attrition rates of existing heads, there has been insufficient succession planning within the system to allow for deputies to move into headship. Indeed, the move to organised professionalism has resulted in an increased bifurcation between the traditional role of the deputy and the reconfigured role of the HT. The data suggested that the intensification of risk associated with headship, and the disconnect between the two roles, militated against deputies moving into headship. To mitigate some of these problems, the WG needs to introduce policies which, as Biesta et al. (2015) argue, foster teacher and head teacher agency and, by doing so, reinforce their professional capital. One way of developing HT professional capital is the promotion of mentoring for aspiring but, in particular, new heads—mentors need to be external and provide a ‘safe space’ which is beyond formal accountability frameworks to help contextualise and problematise issues and nurture an agentic disposition within HTs.

The WG should also explore ways to reduce HTs’ administrative burden. One way would be to consider a reduction in accountability mechanisms or a reconsideration of the consequences of public accountability structures. While such a retrenchment is unlikely, they should certainly ensure that all meso‐level agencies cohere and that requisite support is provided when metrics are not realised. There is evidence that the WG has recognised that the bureaucratic and administrative burdens of headship not only act as a disincentive to recruitment, but militate against effective school performance: as a consequence, they have committed to recruiting business managers and administrators to ease this administrative element (OECD, 2017; Williams, 2017). While this is a welcome move, they should also consider introducing external contextually sensitive support for the diverse responsibilities now within HTs’ locus of responsibility.

The WG has initiated reforms which may ameliorate HT recruitment issues, such as the establishment of the National Academy for Educational Leadership (NAEL). Whilst the introduction of the NAEL is welcome, this should focus on the promotion of agentic heads and not technicist policy implementers: as Harris et al. (2018, p. 1) assert, ‘all too often teachers are the passive recipients of mandated reforms’. Embedded within the academy must be mechanisms for leadership development at all levels of the profession, in order to develop future leadership capacity. To foster agency and develop professional capital (Hargreaves & Fullan, 2012), the promotion of mentoring opportunities for HTs is imperative (Barber et al., 2010). These could take the form of a national mentoring scheme for aspiring HTs with professionally trained mentors (Daly & Milton, 2017), who can support the transition into new headship roles. To realise this, the ‘arm's length principle’ must be realised to allow for the development of head teachers’ professional capital and not merely the narrower conceptions of capabilities, capacities or skills found in other models of professional training (Hargreaves & Fullan, 2012).

At a system level, the WG should resist ‘obsessing’ (Harris & Jones, 2017) about international comparisons and adopting knee‐jerk reactions and policy borrowings which may not fit the country's context. The data from this study supports the OECD's (2017) conclusion that the WG should resist implementing any new reforms and both resource and engage the teaching profession in attempts to embed the reforms that are already in place. There is evidence that there has been a retrenchment from the more prescriptive elements of Welsh educational policy governance, and that Wales is moving towards a third phase of policy‐making (Egan, 2017). Within this ‘phase’, the WG should attempt to foster the agency and develop the professional capital of its existing and potential heads and, by doing so, re‐orientate the role of headship towards an occupational understanding of professional practice.

Acknowledgements

Thanks to our colleagues for their advice and feedback on previous versions of this article. We would also like to thank the BERJ reviewers for their insightful advice. This article is partly based on research supported by the Wales Institute of Social & Economic Research, Data & Methods (WISERD), a collaborative venture between the universities of Aberystwyth, Bangor, Cardiff, South Wales and Swansea. This research was undertaken through WISERD Education and was funded by HEFCW (Higher Education Funding Council for Wales). Daniel Evans and Constantino Dumangane Jr. assisted in the collection of data.