More fragmented, and yet more networked: Analysing the responses of two Local Authorities in England to the Coalition’s ‘self-improving school-led system’ reforms

This paper explores school reform in England under the Conservative-led Coalition government, elected in 2010, through a focus on the changing roles and status of Local Authorities (LAs). The Coalition’s stated aim was the development of a ‘self-improving, school-led’ system in which LAs should become ‘champions for children’. The paper draws on two locality case studies and a set of future scenarios and policy narratives to analyse the ways in which LAs and school leaders are responding to reform. The paper concludes that the Coalition focused its attention on structural reform, but that this placed an additional onus on leadership agency within local school systems to shape contextually appropriate solutions. The schools in the two areas studied appear to have become more fragmented and yet – paradoxically – more networked; however, they are not yet working in the ‘deep partnerships’ envisaged by Hargreaves (2010). This has meant that the LAs have needed to sustain their traditional roles (for example, in providing challenge and support to schools), whilst simultaneously evolving new ways of working (for example, providing ‘bridging social capital’). These roles may sometimes be in tension, but are driven by different factors: LA-level accountability in the case of challenge and support, and reduced funding in the case of ‘bridging social capital’. This suggests that the Coalition’s conflicting policy narratives were in tension and that the notion of LAs as ‘champions for children’ requires review.


Background
This paper explores the implications of the structural reforms in English education introduced by the Conservative-led Coalition Government elected in 2010. The election of a Conservativemajority government in May 2015 means that this broad policy direction can be expected to continue. This paper draws on two local area reviews that the author led or participated in during 2013-14 to analyse the ways in which Local Authorities (LAs) and school leaders are responding to policy reform. The findings from the reviews are analysed using two theoretical constructs: the first is a set of scenarios developed by the author in consultation with a group of school and LA leaders in 2012; the second is a set of four policy narratives identified by the author through discourse analysis of Coalition speeches and publications.
The two reviews drawn on for this paper were commissioned by the respective LAs -the London Borough of Brent (Gilbert et al., 2014), and Coventry in the West Midlands (Greany and Allan, 2014). In Brent the evaluation team gathered evidence on the LA's current approach to school improvement and support through several means: a call for evidence; analysis of LA and national reports and data; an independent review of school performance data; discussions with a range of stakeholders, including the trade unions and the Leader and Lead Member for Education in the Council; and visits and interviews with 20 schools. The schools were selected to represent both key 'system leaders' (such as Teaching Schools, National Leaders of Education (NLEs), and academy sponsors -see below for definitions) and a more representative spread of schools with different performance profiles from primary, secondary, and special phases. In each school the headteacher and, in most cases, other senior staff and governors were interviewed using a semi-structured interview protocol. In Coventry the review included analysis of publicly available and LA documentary evidence and performance data, interviews with LA leaders (the CEO, Lead Member for Education, and Shadow Lead Member for Education), interviews with external stakeholders (Ofsted and the National College for Teaching and Leadership), an online survey of all Coventry headteachers (24 responses received), and interviews with 28 school leaders (headteachers and school governors). The schools were selected to include both key 'system leaders' (e.g. the headteachers leading Coventry's 12 school networks, Teaching Schools, and NLEs) and a more representative spread of schools with different performance profiles that were drawn from five school networks that were studied in more depth.
The selection of the two LAs is a result of the author being commissioned, via a competitive tender in Coventry and an invitation from the review chair in Brent. The two reviews thus provide illustrative case studies (Thomas, 2011) that can provide 'naturalistic generalizability' (Stake, 2000). The evidence from the literature reviewed below and from separate, independently funded research across a larger sample of LAs that is currently being undertaken by the author indicates that similar developments are in train nationwide. The literature referenced in this paper was identified by searching Google Scholar and the IOE library ejournal resource using the keywords of the present article.
As can be seen from Table 1, Brent and Coventry have some broad similarities, in terms of their size, urban context, and above-average levels of deprivation. The majority of secondary schools in both LAs had adopted academy status by April 2014. Both LAs were behind the national picture in terms of primary academy conversion rates: for example, by November 2014 36 LAs nationally had 20 per cent or more of their primary schools working as academies (Bolton, 2014). Both LAs ranked highly for the proportion of schools that hold formal system leadership designations (i.e. as a Teaching School or headed by a National or Local Leader of Education), although this is true for many urban areas. In terms of pupil outcomes, Brent performs above the national average while Coventry performs below it. Additional features of the two LAs and their schools are described in Box 1.

Brent and the London Challenge
Brent is an outer London borough formed in 1965. It has been controlled by Labour for around half its history. It is the most densely populated of the outer London boroughs and continues to increase in density -its population grew by 18 per cent between 2001 and 2011. The ethnic make-up of the population changed over the same period, with a 9 per cent reduction in the White ethnic group and a corresponding increase in the Black and Asian Minority Ethnic (BAME) population from 55 per cent to 64 per cent. Brent's pupil population continues to increase: for example, the number of primary age children grew by 13.7 per cent between January 2010 and October 2013.
Brent's higher-than-average performance in terms of school and pupil attainment reflects the wider picture across London, which saw dramatic improvement from being the lowest to the highest performing English region during the 2000s. Debates continue around the causes of London's improvement (Baars et al., 2014;Greaves et al., 2014), with the role of primary school performance and immigration both highlighted, although most commentators agree that Labour's London Challenge programme played a significant role (Ofsted, 2010;Rudd et al., 2011;Hutchings et al., 2012).
While the Challenge was a pan-London programme, the role of the individual Local Authorities responsible for the city's 32 boroughs remained important (Woods et al., 2013). The London Challenge finished before the Coalition was elected in 2010, yet London's schools have largely continued their upward trajectory since then. Meanwhile, Brent's improvement has slowed, taking it from being one of London's highest performing boroughs in 2009 to a middle performer by 2014.

Coventry
Coventry is a city in the West Midlands with a strong industrial tradition as a centre of the British car industry. It suffered heavy bombing during the Second World War and then embraced an ambitious town planning and rebuilding programme (Kynaston, 2008). Coventry LA needed to find savings of over £60m in total between 2011-12 and 2015-16. The LA was judged 'Inadequate' by Ofsted in March 2014 for the quality of its social care services for children.
In terms of educational performance, Ofsted's 2011-12 annual report included a league table of Local Authorities, ranked by the proportion of schools that were 'Good' or 'Outstanding': Coventry came bottom overall, with 42 per cent. The picture differed between phases: 78 per cent of secondary schools in Coventry were rated 'Good' or 'Outstanding', but just 41 per cent of primary schools were. Concerted action by the LA and schools since then has improved this picture, but the LA still performs below its equivalent peers and national rates.
The Coalition's reform programme: The 'self-improving school-led system' A number of researchers on school system reform refer to 'tri-level reform' (Fullan, 2010;Levin, 2012) as a common approach. Tri-level reformers attempt to focus and align activity and resources between central government, district (i.e. LA) and school levels, around a shared set of goals with a significant focus on capacity building. By definition, the 'tri-level' model assumes a strong role for what is variously called the 'district level', 'middle tier' or 'mediating layer', which is largely synonymous with local government in most school systems. Mourshed et al. (2010) found that every international system they studied had some type of intermediate layer between central government and schools, playing three important roles: As the school systems we studied have progressed on their improvement journey, they seem to have increasingly come to rely on a 'mediating layer' that acts between the centre and the schools. This mediating layer sustains improvement by providing three things of importance to the system: targeted hands-on support to schools, a buffer between the school and the centre, and a channel to share and integrate improvements across schools. (Mourshed et al., 2010: 22) The Coalition government in England resisted both tri-level reform as a model and a strong mediating layer as a design feature. Instead, the Coalition's broad approach was to maximize school autonomy while raising the accountability bar for schools, increasing diversity and choice for parents and reducing the role of central and local government where possible. This reform programme has been radical and widespread, affecting almost every aspect of school life (see Lupton and Thomson, 2015). It has come to be known as 'the school-improving, school-led system' (see Greany, 2014Greany, , 2015aGreany, , 2015bGreany and Scott, 2014;and Greany and Brown, 2015 for more detailed analyses of this policy). The academies programme has been a key element of this reform, with particular implications for LAs, since academies are companies and charities that are funded directly by central government and are outside LA control. By December 2014 there were 4,344 academies open; these included over half of all secondary schools in England (HoC Education Select Committee, 2015), although around four in five schools were still maintained by their LA. Successful schools were encouraged to convert voluntarily to academy status, while schools judged to be failing by Ofsted (which inspects schools, LAs, and wider children's services) were forced to become 'sponsored academies', meaning that they were removed from LA control and run as part of Multi-Academy Trusts (MATs, or academy chains). Free Schools are new academies that can be proposed and developed by parent groups and other providers who want to challenge existing local provision. There were just over 150 free schools open in 2014.
A further innovation has been the expansion of system leadership and school-to-school support, whereby high-performing leaders support under-performing schools and promote wider system improvement (Higham et al., 2009;Hill andMatthews, 2008, 2010;Ofsted, 2010). For example, the Coalition doubled the number of NLEs to 1,000 and introduced 500 Teaching Schools. NLEs are headteachers who are designated to work with their school leadership team to provide support to schools that are struggling. Teaching Schools are designated by the government to co-ordinate initial and continuing professional development, school-to-school support, and research and development across an alliance of partner schools (Matthews and Berwick, 2013). By the end of the Coalition's time in office it could be argued that school-toschool support was the primary mechanism for school improvement in England (Sandals and Bryant, 2014;Earley and Higham, 2012;Education Select Committee, 2013).
Few studies are assessing the ways in which attitudes and practices are evolving on the ground as a result of Coalition policies. Earley and Higham's 2012 research suggested that headteachers fall into one of four categories: confident, cautious, concerned, or constrained. 2 Research with 'wellpositioned' headteachers (i.e. those from schools that are Ofsted-rated 'Good' or 'Outstanding') suggests they see the world as increasingly hierarchical, indicating the risk of a two-tier system in which the weak get weaker and the strong stronger (Coldron et al., 2014).

The changing role of Local Authorities in a self-improving school-led system
When universal primary and secondary education was developed after the Second World War, Local Education Authorities were given the prime responsibility for shaping and overseeing provision in their area -a 'national system, locally administered'. The power and authority of LAs was then steadily eroded from the 1970s onwards as school autonomy increased, with Grant Maintained and then academy schools accelerating this shift towards central funding and oversight from the 1990s onwards (Volansky, 2003;Newsam, 2014). By 2009 school leaders in England were ranked among the most autonomous in the world in terms of their decisionmaking powers (OECD, 2011). Nevertheless, under the 1997-2010 New Labour Governments, LAs largely retained -and in many cases extended -their powers and responsibilities in relation to schools, for example through the National Literacy and Numeracy Strategies. The effectiveness of LAs in fulfilling these roles varied, which the Labour Government sought to address by outsourcing some or all underperforming LA functions to private companies and charities.
The Coalition's 2010 white paper envisaged LAs playing: a critical new role -as strengthened champions of choice, securing a wide range of education options for parents and families, ensuring there are sufficient high-quality school places, coordinating fair admissions, promoting social justice by supporting vulnerable children and challenging schools which fail to improve. (DfE, 2010: 65).
The onus was on LAs themselves to shape this new role -for example by offering school improvement as a traded service and/or beyond their immediate boundaries -with the Coalition promising to free them from unnecessary bureaucracy to enable this entrepreneurial diversity to emerge. In practice, LA spending on education reduced by around 18 per cent between 2009 and 2013, largely as a result of schools becoming academies and a sharp reduction in ring-fenced grants administered by LAs (Hastings et al., 2013). As a result, most LAs have had to reduce their staffing and capacity, while the focus of most Coalition policy activity and effort has been on expanding the number and performance of academies, and on designing a national infrastructure to fund and oversee so many academies directly from Westminster (Greany and Scott, 2014). A number of commentators have explored questions such as how local democratic accountability might best be secured and how support and challenge for all schools might be orchestrated in a fully academized system. Arguments range from increasing accountability through the market (O'Shaugnessy, 2012), to a re-conceptualized role for LAs (Gilbert et al., 2013), to the need for a new middle tier to manage school provision, challenge, and support (Hill, 2012;Blunkett, 2014). By 2014 the Coalition had accepted the need for increased regional capacity to oversee academies, and appointed eight Regional Schools Commissioners (RSCs), supported and advised by elected Head Teacher Boards. The RSCs report to the Secretary of State and are charged with monitoring the performance of academies and free schools in their area and intervening where standards drop. Thus the mediating layer in England today is made up of multiple players. Ball describes this move away from oversight of all schools by democratically elected LAs and towards a more complex, heterarchical model of governance as characteristic of a 'reluctant state' that is deliberately reducing its role in publicly funded schooling (Ball, 2012). Interestingly, when it comes to impact, the most recent analysis of MAT and LA performance (DfE, 2015;Cook, 2015) shows an overlapping range of performance for both models.
The focus on building an academized system has meant that LAs have largely been subject to policy neglect by the Coalition (see Rogers, n.d., for a summary of key policy developments in relation to LA oversight of schools). For example, LAs have retained nearly two hundred statutory duties 3 in relation to education, including a requirement to appoint a Director of Children's Services (despite the fact that few can now afford or justify such a role). A subset of LAs appeared after 2010 to take an initial view that they should encourage all schools to become academies as quickly as possible, thereby freeing up LA resources and minimizing risk since the schools would be accountable to the Secretary of State rather than the LA. Othersparticularly Labour-controlled LAs in the north -sought to prevent any schools converting to academy status and to retain their existing school improvement teams as far as possible. A third group -such as Wigan (Aston et al., 2013) -reduced their core teams and asked all schools to join a school-led network with a designated lead school responsible for school improvement.
By 2012 Ofsted appears to have become concerned that the lack of focus on LAs coupled with their reducing capacity was leading to a loss of momentum in terms of school improvement. In response Ofsted began inspecting LA school improvement services from 2013, publishing damning critiques and requiring detailed action plans where performance was found to be weak. 4 In practical terms the task for LA leaders responsible for oversight of schools since 2010 can therefore be seen as threefold: • to continue to fulfil their statutory obligations with regard to maintained schools in the context of sharply reduced funding; • to continue to fulfil their wider statutory obligations relating to education, for example to ensure a sufficient supply of school places; and • to shape a meaningful new role as a 'champion for children' as more schools become academies.
As yet, there has been relatively little research to understand the ways in which LAs are responding to these challenges. Ainscow (2015) provides an insider's perspective on the often painful and variable process of change for 10 LAs involved in the Greater Manchester Challenge, a Labour initiative that continued into the early years of the Coalition and that informed the focus on 'school-led' reform. Aston et al. (2013) drew on case studies of innovative LAs and other local arrangements to identify some of the ways in which local school systems were beginning to work. Simkins (2015) analysed the responses of three contrasting LAs and their schools with a focus on the types of school partnerships emerging in each locality. Sandals and Bryant (2014) tracked progress across ten LA areas over one year, with a focus on three areas of practice that were identified from an earlier phase of DfE-sponsored action research (Parish et al., 2012). Whereas the 2012 report signalled concerns about a lack of capacity for self-improvement and a lack of buy-in from autonomous schools to areas where collective action and decision making is required, the 2014 report painted a more positive picture. It found that confidence and commitment to partnership working were growing amongst school leaders and cited many specific examples of how this was developing in different contexts. Confidence in the capacity of local systems to meet the needs of vulnerable children was less strong. Sandals and Bryant also hinted that some LAs were struggling to engage schools in a collaborative dialogue through its characterization of them as either 'timely adapters' (i.e. strong LAs that have worked with schools to shape more school-led approaches); 'slow movers' (lower-performing and less credible LAs that have not managed to respond to the changing environment); or 'sudden reactors' (LAs that have withdrawn from provision with little attention to capacity-building in schools).
Hatcher (2014) offers a critical perspective on the examples presented by Aston et al. (2013) and also draws on his own research into the development of a headteacher-led partnership in Birmingham. He identifies the trend towards developing new partnership arrangements that attempt to secure collaborative engagement and strategic oversight, but argues that these are 'closed managerialist networks' that exclude or minimize the legitimate democratic role of LAs as well as other stakeholders such as parents. He views them as focused on delivering narrow, government-prescribed school improvement priorities rather than a more developmental or critical agenda.

Analysis and discussion
The approaches to school improvement and reform adopted by Brent and Coventry are briefly outlined in Box 2. At the simplest level, Brent's approach could be described as 'pared back monitoring and intervention', while Coventry's was 'intensive monitoring, brokerage of schoolto-school support, and capacity building through school networks'. Both LAs continued to offer some services to schools, generally on a traded 'buy-back' basis (meaning the schools could choose whether or not to purchase the services). Both LAs had mixed -often poor -relationships with their academies, though the Lead Member in Coventry was continuing to provide active challenge to its academies. Meanwhile, in both areas and with varying degrees of LA support, a number of schools had begun to shape a more 'school-led' approach to improvement. The Brent Schools Partnership (BSP) was initiated by system leader schools, although it had received some funding from the LA, and its membership did not include all LA schools. Coventry's 12 school partnerships had been initiated and funded by the LA, so did nominally include all schools. The city's Teaching School Alliances had developed separately, but had recently agreed to align themselves with the networks.

Brent's approach to school improvement and development
Headteachers and other stakeholders in Brent told the Commission that the Council's approach towards many issues was not strategic. The Commission concluded that 'education has not been a sufficiently high priority for the Council for many years' and identified a number of areas, such as pupil place planning, where the quality of education was at risk as a result. Many senior positions in the education department had been undertaken by interim staff for an extended period. Relations with academies in the borough were often poor. Although the LA's standards of performance remained high by national standards, it was slipping compared to other London boroughs and the totals masked significant issues, such as the fact that one in five secondary schools in the borough were ranked 'Inadequate' by Ofsted in 2014 -the highest rate in London.
The LA had recently moved to a new model that focused attention on monitoring and supporting lower-performing maintained schools, drawing on a mixture of LA staff and brokered school-to-school support. Ofsted 'Good' and 'Outstanding' schools in Brent were now largely left alone, subject only to a light-touch annual monitoring visit. Many higherperforming schools in Brent were actively looking beyond the LA for partnerships and support from regional and national organizations. Meanwhile, a number of Brent schools had come together to form the Brent Schools Partnership (BSP), which saw its aim as 'not to replace the services provided by the LA, but to work together to ensure that all services provided to our schools are of high quality and have impact'. The Commission concluded that 'Head teachers want the BSP to work but they are not yet seeing enough of the kind of support that they want.'

Coventry's approach to school improvement and development
Ofsted undertook an area inspection of Coventry LA's schools in early 2013, which found that 'there is still some way to go in establishing a widely understood and methodically delivered strategy for improvement'. Coventry's primary schools have improved rapidly since then, with 74 per cent judged 'Good' or 'Outstanding' by summer 2014, although the city's 2014 KS2 national test results remained below national averages and significantly below the highest-performing similar authorities. Secondary schools in the city declined in Ofsted terms and in terms of overall attainment between 2012 and 2014, after several years of steady improvement, leaving them below national averages and in the bottom quintile of authorities nationally.
The Lead Councillor for Education started monitoring the progress of all schools and students in the city following the Ofsted review, with a focus on the 30 schools most at risk. The LA now commissions experts to help schools that are struggling, wherever possible from other schools within the city. Views expressed by interviewees included that the LA was good in the past but had 'lost its way' a few years back, that it has been very inwardly focused as it has grappled with cuts and personnel changes, that it has sometimes provided insufficient leadership and challenge, and that it has been overly focused on primaries. More positively, most school leaders felt that the LA had improved in the past 18 months, and that the new model of differentiated challenge and support for schools was credible.
The LA had started funding school networks before 2013, as a way to build capacity as LA resources decreased. The evaluation found that these networks were now developing well, although some -generally those that had been established for longer -were stronger than others. The review concluded that 'at present, it appears that the challenge role is too firmly driven by the LA and the support role is not sufficiently led by schools'.

A framework for analysis: Four scenarios
One framework for assessing the Brent and Coventry cases is a set of scenarios that I developed with a group of primary school and LA leaders through a series of four workshops in late 2012. The framework for the scenarios is shown in Figure 1 and the four scenarios are sketched out in Annex A. The framework includes two variables: competition versus collaboration between schools, and whether the LA remains a significant force or disintegrates in the face of academization.
Chermack (2005) offers a theoretical framework for scenario planning, wherein one of the roles of scenarios is to enable learning by helping leaders to 'reperceive the organization and its environment' (Chermack, 2005: 62). He outlines a series of testable hypotheses that can be used to assess the impact of scenario planning on organizational performance. Our intention in developing the LA scenarios was certainly to enable leaders to 'reperceive' their relationships and the potential implications of different courses of action at a time of rapid change. However, rather than seeing the scenarios as a means of improving performance, our aim was more limited: we wanted to map out potential outcomes if different courses of action were taken, and we wanted to emphasize that no outcome was given -i.e. that leaders could choose to act in different ways to secure different desired outcomes.
Analysing the Brent and Coventry examples against the four scenarios, it is clear that there is no neat fit. This is not surprising; the scenarios were written to caricature extreme possible outcomes, for example by positing that all schools might collaborate all the time, when the reality is that schools operate on a spectrum between competition and collaboration. As a result, while both Brent and Coventry can be seen to have aspects of scenario A (LA remains; schools compete), because both have had to move away from universal services towards a traded 'buy-back' offer (Glover et al., 2014), neither LA could be said to fit only into this scenario. Nevertheless, it is possible to see features of Brent or Coventry more clearly in some scenarios than others. For example, Coventry is an authority that has been highly active since the 2013 Ofsted review and has sought to proactively build school networks, placing it most obviously in scenario B (LA remains; schools collaborate). Brent is an authority that, at least prior to the Commission report, has lacked a clear vision and role beyond a focus on the lowest-performing schools. Many of its secondary schools are fiercely competitive with each other and dismissive of the LA, actively looking beyond LA boundaries for inspiration and support. This seems to put Brent most obviously in scenario C (LA disappears; schools compete), although the development of the BSP is more characteristic of scenario D (LA disappears; schools collaborate), while the Commission's recommendations appear to be an attempt to move the LA into scenario B.
What seems most interesting in reviewing the scenarios is that, back in 2012, the school and LA leaders were hypothesizing the disappearance of LAs altogether. At that time, the expansion of academies and the policy neglect of LAs made this seem a real possibility. For example, the RSA was undertaking a review at that time to understand what a fully academized system might look like (Gilbert et al., 2013). Yet, three years later, the two LAs studied appear to be sustaining their roles (for example, by continuing to monitor and intervene in underperforming schools), whilst also beginning to shape new ways of working (for example, providing 'bridging social capital' by funding and supporting school networks). A second framework for analysing the two case studies is a set of four parallel narratives in Coalition policy that I identified through a discourse analysis of policy documents and speeches (Greany, 2014): • World-class/no excuses: includes policies such as a more demanding National Curriculum and assessment framework, and rising floor targets for schools. Schools that do not achieve these higher benchmarks are taken over by academy sponsors. • Freedom to teach: includes increased autonomy for academies and new powers for all schools, for example on performance-related pay and classroom behaviour, as well as reduced bureaucracy. • Market-based approach: includes increased parent choice and competition for schools through Free Schools, as well as a more equitable and national funding formula. • System leadership approach: includes mechanisms for school-to-school support, such as Teaching Schools and NLEs, as well as school-led solutions such as Headteacher Boards and a Royal College of Teaching.
I have argued elsewhere that these competing narratives create tensions and contradictions for school leaders and that they accentuate local hierarchies of schools (Greany, 2014(Greany, , 2015aColdron et al., 2014), but here I want to assess their impact on LAs.

Three themes from the analysis
Three themes emerge from comparing the scenarios and narratives with the real-world examples of Brent and Coventry, which I explore below.

The LAs have sustained but reshaped their monitoring and intervention roles, driven primarily by accountability pressures
Comparing the scenarios with the real-world LA examples reveals that the scenarios underplay the continuing importance of monitoring and intervention for maintained schools by LAs, and to a lesser extent academies. This is the classic LA school improvement role, as defined over several decades in reams of legislation and guidance, but in the Coalition's rush for a fully academized system after 2010 this role was at risk of neglect or disintegration in many LAs. The four narratives arguably define the possible drivers for this LA behaviour: • accountability through Ofsted reviews -world class/no excuses • school empowerment -freedom to teach • reduced LA core budgets -market-based approach • the need to draw on the most credible and expert leaders -system leadership.
The pressure from school empowerment (i.e. that schools might choose to leave the LA and become academies if services were not good enough) does not appear to have been a primary driver of the reshaped offer in either LA: if it were, then Brent would surely be doing more to ensure Ofsted 'Good' and 'Outstanding' schools were receiving a more comprehensive service, since these are the schools that could convert to academy status most easily. The reduced resources available to LAs certainly appear to have driven change: both LAs have moved from a universal offer to focus their resources on schools causing concern, but this appears to be a case of 'cutting the cloth' rather than responding to market pressures. Both LAs had increased the use of serving/recently retired headteachers to provide monitoring support and both were commissioning school-to-school support interventions instead of LA advisors, suggesting that the system leadership narrative has had an impact. However, it was Ofsted's decision to focus on the LA school improvement role from 2013 onwards that appears to have had the greatest impact on LA decisions and action in this area: for example, the Chief Executive of Coventry described the 2013 Ofsted review as a 'burning platform' for change.

The LAs have sought to increase leadership agency, in particular by providing 'bridging social capital', which strengthens school networks
Both LAs have sought to enhance 'bridging social capital' (Mulford, in Ainscow, 2015) between schools, most obviously by establishing and funding the school networks in Coventry and through support for the BSP. This strengthening of relational capital between schools appears to be seen by the LAs as a necessary foundation for 'school-led improvement', thereby enabling the LA to 'get off the pitch' (Coventry LA CEO). The theory of action underpinning such a shift was not fully developed in either LA, although the fact that they both commissioned external reviews might indicate a recognition of the need for a more strategic approach. What seems clear is that the combination of contradictory national policy narratives coupled with limited capacity-building for school leaders at LA level meant that schools were responding in widely differing ways to the change agenda. This response was also influenced by each school's existing position within the local hierarchy, which depended in part on its Ofsted profile, since this determined whether or not it could become an academy or Teaching School for example. These leadership responses appeared to fall on three broad spectra: reactive (fatalistic) versus proactive; individual versus collective; and focused on shared values versus focused on compliance with external requirements. All these responses were apparent in both Brent and Coventry, from the system leaders working to shape the BSP and the Coventry networks (proactive, collective, shared values), to the headteacher who was fearful of peer review, sceptical that the networks would make a difference for his school, and focused on his next Ofsted inspection (reactive, individual, compliance-focused).
The LAs appeared to play a role in influencing the ways in which this school leadership agency was enacted, with a subtle balance to be struck between shaping a collective approach without dominating and thereby preventing the emergence of school-led solutions. For example, there was a consistent commitment to sustaining 'the Coventry family of schools' from both school and LA leaders there, and the LA had been proactive in funding the school networks. This suggests a commitment to the 'system leadership' policy narrative, yet the LA was also supporting the development of new free schools, albeit in a strategic way (for example, working with the secondary schools to develop a shared proposal for a new special school). This indicates that the LA was supportive of the other policy narratives, but was working to mediate their potential negative impacts (since a new Free School could potentially increase competition between schools). The possible downside of Coventry's proactive approach, in particular to the monitoring and intervention of schools, may have been that schools did not feel a great need to develop their own solutions: for example, there was no equivalent to the BSP in Coventry. By contrast, Brent had been far less proactive in shaping and implementing a shared vision among its schools and had adopted a reactive approach to new Free Schools. This may explain why the borough felt more fragmented, with many strong schools looking elsewhere for inspiration and collaboration, while others were hoping the BSP could provide a more collective solution.
Thus it can be seen that leadership agency is as important for the LAs as for their schools: agency here denotes the extent to which the LA has the capacity and credibility to shape collective solutions to shared challenges. This LA leadership appeared to range along the same dimensions as for schools -reactive/proactive, individual/collective, and values-based/compliance-focused. What was notable, however, was that LA leadership agency was critically influenced by whether or not the LA was seen as competent and credible: that is, who was involved and how they worked were as important as what they did. For example, both LAs had reduced their internal adviser teams in favour of buying in serving heads or HMI (Her Majesty's Inspectors -i.e. Ofsted) to undertake monitoring visits, which was seen as more credible and therefore welcomed by the schools. Yet Brent in particular was an LA that had struggled to recruit and retain high-quality staff, with a succession of interim appointments in key roles. This appeared to be less to do with its ability to pay competitive salaries so much as a sense that LAs were not seen as exciting, career-enhancing places to work at a time when policy attention was focused on academies. These staffing and leadership challenges arguably affected the LA's ability to shape and enact a collectively held vision for education across the borough, particularly in the face of a number of high-performing and highly autonomous schools.
The interplay between the four competing national policy narratives, the mediating layer, local context, and leadership agency at school level is shown diagrammatically in Figure 2. Critically, LAs appear to play a role in mediating the competing narratives with and on behalf of schools.

Deep partnerships, peer review, and the self-improving system
The third area illuminated by the scenarios, the policy narratives, and the case studies is school partnerships and support. Both LAs were brokering arrangements for strong schools to support struggling schools rather than drawing on in-house LA expertise where possible. What seems notable, though, is that neither LA had been able to fully step back in the ways envisaged in scenario D (LA disappears/schools collaborate). Their schools simply did not feel ready or able to undertake the LAs' monitoring and intervention roles.
This raises a question around whether these roles could ever really move to schools in a 'self-improving system'. David Hargreaves (2010Hargreaves ( , 2011Hargreaves ( , 2012aHargreaves ( , 2012b argued that schools must operate in 'deep partnerships' for a self-improving system to be successful. However, Hargreaves acknowledged that building deep partnerships is difficult. In particular, he saw robust 'evaluation and challenge' between schools as particularly difficult to develop: put simply, it is hard for a school leader or teacher from one school to tell their counterpart in another school that their work is not up to scratch. But Hargreaves argued that such peer challenge is possible if sufficient social capital (characterized by mutual trust and reciprocity) and collective moral purpose have developed between the partners.
Certainly, some of the schools in both LAs were participating in some form of peer review. At one end of the spectrum this was primarily about learning from practice in other schools (for example through Learning Walks focused on a particular aspect of practice), but at the other it was more akin to peer inspection and rigorous feedback on strengths and weaknesses. In general, though, schools in both LAs saw such work as about enhancing mutual learning and continuous improvement, not as a replacement for external scrutiny and challenge from the LA or Ofsted. In fact, they saw peer review as more difficult with schools close by in their own LA, where the competitive pressures were greater, than with schools further afield. Equally, many schools were nervous about even this level of peer review, either because they felt it would break the fragile trust between them or because they wanted to protect their trade secrets in a competitive environment. Perhaps as a result, most heads still saw themselves as fundamentally responsible for their own school and their own destiny. Coventry's networks and the BSP might help in some respects here, but they were not yet the 'deep partnerships' that Hargreaves envisaged that could replace the traditional intervention roles of the LA.

Conclusion
This article has explored the role of LAs in the Coalition's 'self-improving school system', drawing on illustrative case studies from two LAs, a set of future scenarios exploring how LAs and schools might respond in the face of academization, and a set of four policy narratives extrapolated from Coalition documents and speeches. In undertaking this analysis, the article has focused on the practicalities of the LA role, rather than debating the wider issue of democratic legitimacy in the oversight of schools.
Three key findings emerge: • LAs have been under four pressures to change that align with the four policy narratives: a pressure to reduce budgets, a school empowerment pressure, a credibility pressure, and an accountability pressure from Ofsted. The accountability pressure appears to have been the most influential in changing practice. • The LAs have sought to increase leadership agency, in particular by providing 'bridging social capital', which strengthens school networks. While this has helped strengthen some school networks, their capacity and reach are limited and both localities were also becoming more fragmented. Progress appears to have been dependent on credibility and relationships between LA staff and schools, rather than statutory positions or roles. This need for credibility places a premium on the quality of leadership in LAs and schools, but, partly as a result of policy neglect, LAs have struggled to recruit and retain high-calibre staff, thus hampering their leadership capacity. • Schools in these LAs are not yet operating in the 'deep partnerships' envisaged by Hargreaves (2010), with peer review (which Hargreaves terms 'evaluation and challenge') proving the most difficult aspect to achieve. The accountability pressure on LAs to fulfil their statutory obligations on school improvement may be hampering the development of robust peer review arrangements between schools. This raises questions about whether and how a fully 'self-improving system' can emerge or whether a new, more coherent but strategic mediating layer will be required under the new government.
The analysis suggests that the scenarios and narratives resonate with current practice on the ground, although the scenarios underplay the continuing importance of school oversight and intervention by LAs. The scenarios also appear to underplay the importance of leadership agency, both for school and LA leaders. It is clear that the new Conservative Government elected in 2015 will want to increase the number of academies and thereby reduce the role of LAs further. Since the two studies outlined in this article were completed, a number of LAs across the country, including Birmingham, have moved further towards commissioning school partnerships to take on the lead role in monitoring and intervening in schools. Others, such as Lincolnshire, are requiring all their maintained schools to engage in peer reviews, which might help build the 'deep partnerships' envisaged by Hargreaves. It is unclear how these emerging roles will sit alongside the Regional Schools Commissioners created by the DfE to oversee academies. Some argue that the answer is to move all schools into a MAT, with the trust board responsible for monitoring and intervention, but this would require a massive expansion in the number and capacity of such trusts. Thus, although the picture continues to evolve, it seems that LAs remain a significant force within England's complex mediating layer, not least because they shape the ways in which local leadership agency is framed. Whether LAs remain or disappear as a force within local school systems remains crucially important and worthy of more than mere policy neglect over the coming five years.

Notes
Annex A: Four potential future scenarios for primary schools and LAs in a 'self-improving system' Schools are encouraged to work together in families and clusters, taking responsibility for their collective improvement, with the LA as broker. Many partnerships thrive, but others collapse or simply fall into neglect.
In a world without local authorities, schools become increasingly competitive, whether they have joined chains or multi-academy trusts, or remained independent. Local DfE offices sort out the problems of market failure.
School partnerships become the building blocks of a self-improving system, with headteachers as architects. In the absence of LAs, schools are directly accountable to the DfE.

Responses to change
• Core school budgets may have been protected, but most services that were formerly free must now be bought in • Many schools, dioceses, and private providers enter the market. Most LAs shrink their core teams to a minimum and focus on running viable traded services. Most work hard to maintain the link with primary schools, and some compete for business beyond their borders • In many areas, competition has the effect of improving the range, quality, and value for money of services to schools. Schools become more accomplished at solving problems in house, and knowing when and how to buy in external support • Where the new arrangements work less well, single schools are left with minimal support, chains and federations become insular, and LAs lose their capacity and expertise to intervene with failing schools • The DfE responds by requiring LAs to earn a 'licence to operate' and making sure that chains are more inclusive and less protective of their brand and intellectual property rights.
• Most LAs turn their traded services into fully independent businesses. They focus their remaining capacity on brokering effective school partnerships. These partnerships are given responsibility for collective improvement, often based on the teaching school/NLE model • Where existing school partnerships are strong and there is a culture of peer challenge and support, this works well. In these areas the LA has fostered schoolto-school support for some time and there is a cadre of system leaders • In other areas the LA does not have the credibility, expertise, or capacity to broker and quality-assure partnerships. Schools object to being put into forced partnerships, but lack the leadership to make it work themselves. With no support or challenge the partnerships are weak, but nonetheless time-consuming • Concerned that schools will not be held rigorously to account, the DfE introduces new governance legislation, including paid chairs of governors.
• Once LAs have disappeared, schools find themselves facing some difficult decisions in a competitive market • The great majority of church schools join diocesan chains. Maintained schools have the option of converting on their own, becoming part of a chain, or joining a different kind of grouping • Localism, diversity, and competition are the hallmarks of the new system. Many parents welcome the choice that comes with this, as schools compete for custom • In some areas, however, the big chains, often based on large secondary schools, start to dominate. This prompts the emergence of innovative primary chains and groups committed to maintaining the culture and ethos of member schools • The residual powers of the LA transfer to regional commissioners appointed by the DfE to deal with market (i.e. school) failure and pupil exclusion. Over time, this role is expanded to cover aspects of pupil place planning and admissions.
• Following the introduction of GP-led Clinical Commissioning Groups (CCGs) in the NHS, the DfE legislates to remove the statutory powers of the LA and requires all schools to join formal school commissioning partnerships • As LAs disappear, heads and governing bodies take a lead in making the new arrangements work, despite the additional work that this requires • The best partnerships design in quality assurance through peer challenge, often using teaching school alliances as the basis for a new, more formal approach • The less successful partnerships become overwhelmed by the bureaucracy associated with many former LA functions, and some heads go it alone on the grounds that the partnerships they had joined are now too 'comfortable' • Facing criticism that these alliances are just groups of providers getting together and that they lack rigour, the DfE legislates for all partnerships to appoint 'an independent partner'.