Promoting Equity in Market-Driven Education Systems: Lessons from England

: There is a global trend towards the use of market-driven approaches as a strategy for educational reform. However, this is creating new barriers to the promotion of equity in some countries. Focusing on England as an extreme example of this approach, this paper points to some possibilities for addressing this concern. It reports ﬁndings from a series of studies in high poverty contexts in England. These studies have typically involved local educational practitioners and university researchers working together in ways designed to support equitable developments. Lessons from these experiences are identiﬁed for market-driven systems internationally. They suggest that to create more equitable arrangements, schools need to work together, and with other organizations, both within and beyond their local areas. They also point to the value of surfacing and using the rich experiential and contextualized knowledge held by practitioners to inform these collaborative developments. Acting on these lessons would mark a signiﬁcant shift for systems whose current emphasis is on schools working competitively and in isolation, often to the detriment of disadvantaged children and young people.


Introduction
Equity in education is high on the policy agenda of most countries. This is not least because of the impact of the United Nations Sustainable Development Goal 4. This key international policy document aims to 'ensure inclusive and equitable quality education and promote lifelong learning opportunities for all' [1] (p. 6). However, there is considerable evidence that this is not being achieved for many children internationally. For example, UNESCO's Institute for Statistics [2] estimates that, globally, some 258 million children are not in school and more than 617 million do not achieve minimum proficiency levels in reading and mathematics. Moreover, the most economically disadvantaged children and young people are systematically overrepresented in these groups. This situation is inequitable, with the distribution of poor educational outcomes appearing socially unjust [3].
This problem is most acute in low-and middle-income countries. However, there are also entrenched patterns of inequity across higher-income countries, where poverty has long been the strongest predictor of poor educational outcomes [4]. England-one of the four constituent nations of the UK and the focus of this paper-is typical in this respect. Figures for the 2019-2020 academic year suggest that 31 per cent of school-aged children in England were living in poverty [5]. As a cohort, those eligible for free school meals-an indicator of low family income -were the equivalent of 18.1 months behind in their learning than their more affluent peers at age 16 [6]. England is also typical of many countries in having strong spatial concentrations of poverty and poor educational outcomes. These typically co-occur in places with weak physical, economic, and service infrastructures [7]. Furthermore, studies have highlighted the roles that regional economics and underinvestment, and national health, housing and welfare policies, have played in helping to create and sustain poor educational outcomes in particular places [8,9].
A recent report from the European Commission [10] reflected on the prevalence of similar patterns of educational inequity to those in England across its 27 member states and in other OECD countries. It reached the somewhat damning verdict that 'education is failing to reduce inequalities linked to socio-economic status, despite the fact that the highest performing national education systems are those that put a premium on equity' [10] (para 2.2). It is, therefore, imperative to find ways to promote greater equity across education systems internationally.

The Impact of Market Forces
In recent decades, a common feature of many the 'highest performing national education systems' has been the use of market forces in an effort to drive equitable change. These countries have set out with the dual aims of raising standards overall, while also narrowing gaps between the most and least advantaged learners. In doing so, they have assumed that market forces can provide schools with the freedoms and incentives needed to ensure that all learners succeed academically, regardless of their personal backgrounds and where they live. The introduction of academies in England, charter schools in the US, free schools in Sweden and voucher systems in Chile, are all part of this international direction of travel.
In practice, Salokangas and Ainscow [11] argue that this has widely resulted in a combination of measures to: • Increase school autonomy. This assumes that schools will be better able to meet the needs of their own particular pupils if more decisions about how schools operate are taken locally by the people involved in the schools, rather than by district administrators.

•
Involve new 'actors' in the management, administration and governance of schools. These 'actors' often include organizations and individuals not traditionally involved in the management of publicly funded education, such as private businesses and philanthropic and third sector organizations. They are seen as being able to improve schools' performances by bringing fresh ideas and more efficient ways of working into public education systems. • Introduce (or reinforce) heavily regulated quality assurance systems. Typically, this involves holding schools publicly to account on attainment-focused outcome measures. • Promote parental choice. Often this involves establishing mechanisms to allow parents some choice about the schools they wish their children to attend. It is generally assumed that they will choose the highest performing schools locally available.
The potential merits of this global trend towards marketization remain a matter of considerable debate and there have been varied views as to its impacts. Reviewing the available international evidence, Muijs and Rumyantseva [12] concluded that competition between schools 'leads to modest improvements in pupil attainment, but also to modest increases in socioeconomic and ethnic segregation' (p. 3). This indicates that while standards may have risen overall, the most disadvantaged appear to have benefitted least, and may even have been further disadvantaged. More recent evidence has added weight to these concerns [6,11,13]. This has led Fullan [14] to claim that market-driven systems have become 'catastrophically ineffective' (p. 4) in addressing inequities and that nothing short of a 'radical paradigm shift' is needed to tackle these. In what follows, we consider the kinds of shifts that might be achievable.

Setting the Agenda
The international trends we have identified point to one of the most pressing educational challenges of our time. This can be expressed (or appreciated) as a tension as follows.
On the one hand, there is good reason to believe that education systems working on market principles are failing to reduce inequities and will continue to do so. Yet on the other hand, the political and economic capital invested in market-driven arrangements is now so great that there is likely to be little political will to fundamentally disrupt or dismantle them. Even in Nordic countries that have historically operated on more collectivist and universalist principles, and enjoyed high attainments and relatively narrow gaps, market principles are becoming increasingly dominant. This is despite the resulting growth in inequalities [15,16].
This leads to the question at the heart of this paper: 'how can greater equity be promoted within education systems based on market forces?' Here we set out to address this question from an applied, empirical stance. That is to say, we are concerned with the possibility of change in practice, and in particular, in high-poverty contexts in high-income nations. It also means that, in contrast to Fullan's call for a radical paradigm shift, we adopt a more pragmatic position. We suggest that whatever is to be done, it will almost certainly have to work within the existing constraints of "normal politics".
In these respects, we have long argued that some of the most valuable knowledge available about how to move market-driven education systems in more equitable directions is likely to come from within these systems. More specifically, it is likely to come from those working directly in and around schools serving high-poverty populations in high-poverty places [17,18]. We also know from experience that while these local professionals can be expert in finding opportunities to do things differently, their efforts are rarely visible beyond their immediate local contexts. This means that there is often little opportunity to learn from them. Research therefore has a vital role to play in identifying and drawing out internationally relevant lessons from places where local system arrangements are being remodeled gradually, from the bottom-up, to promote greater equity.
With a focus on England as an extreme case of the application of market principles, in what follows we explore the implications of this thinking. We begin by outlining some of the main challenges to equity in the current English context. We then present some locally emerging responses to these, as identified through our own research. We argue that these suggest ways of working that run counter to market-driven concerns and have the potential to promote greater equity. We then draw out the lessons from these developments for market-driven education systems internationally.

The English Context
For the last 30 years, England has vigorously pursued a market-driven approach to educational policies [11,19]. Reforms have included the introduction of a national curriculum, coupled with central-government mechanisms to hold schools publicly to account. These moves have enabled the creation of public league tables based on schools' performances in national-level standardized attainment tests. New forms of centrally managed school inspection have also been introduced, with punitive measures applied to those schools inspectors deem 'underperforming'. These range from changes in school leadership, to school closure in extreme cases.
All of this has resulted in the increased scrutiny of schools. It has also served to (re)inforce a narrow understanding of what counts as valuable educational outcomes. This has led to a government preoccupation with identifying tightly defined, replicable interventions, that 'work' to improve attainment outcomes rapidly and at scale [20].
England has also promoted school autonomy as core to its reforms. It has done so perhaps more intensively than in any other country, primarily through its national academies program. Academies are state-funded schools that remain subject to regular inspection. However, they sit outside local government control-in England, through Local Authorities (LAs). Instead, academies operate as self-governing, not-for-profit charitable trusts. They also have some freedoms from the national curriculum and national agreements on teachers' pay and conditions [11]. As originally conceived, academies were designed to replace inner-city secondary schools (for ages [11][12][13][14][15][16] judged as underperforming. In these situations, they were intended to use their additional freedoms to bring about significant improvements. Since 2010, however, this focus has shifted to encouraging all schools to become academies. Over the last decade, the pace of academization has been fierce, with Ladd and Fiske [21] dubbing England as the 'wild west' of academy reforms. While there were just 272 academies in England in the 2010-2011 academic year, by 2020-2021 this had increased to 9444, with academies accounting for 78 per cent of all secondary schools and 37 per cent of primary schools (for ages 5-11) [22]. The current government target is for all schools to become academies by 2030 [23].
As academy numbers have grown, government has also encouraged the creation of multi-academy trusts (MATs). These are groups of academies with shared governance arrangements. Schools with poor performance on national measures have been pressured to join MATs. It has been assumed that sharing governance arrangements with higher performing schools will support their improvement. MAT-level performance league tables are now also published by central government.
Emerging MAT arrangements have proved highly diverse. A wide range of 'new actors' have become involved in managing schools who have differing motivations, expertise and understandings of education [24]. For instance, while some MATs are locality-based, others operate nationally. Some are school-led, while others, led by church diocese, are exclusively for faith schools. Others are led by organizations with strong social responsibility remits, but which may have little experience of managing schools. These range, for example, from universities to registered social landlords who provide low-cost housing. Meanwhile, some MATs are led by non-profit organizations created specifically to form MATs with the capacity to manage large numbers of geographically dispersed schools. As all of this illustrates, the English education system has become increasingly diverse and fragmented.
In what follows, we demonstrate how this has had wide-ranging implications for achieving greater equity.

The Changing Role of Local Authorities
As academy numbers have grown, the role of local government in educational arrangements has shrunk. Historically, the English school system has relied on LAs to manage all the schools in local administrative areas. They have managed school admissions, supported school improvement, and provided specialist services. Some LA services were supplied directly to schools-from educational psychology and behavior support, to administrative and financial functions. Others were in the form of wider 'children's services', a term commonly used to refer to social care and welfare supports for vulnerable children and families.
However imperfect these arrangements may have been, they positioned LAs to address wide-ranging inequities, experienced by individual children, families and schools, and at an area-level. By contrast, LAs' positions are now much less clear. They are just one of a number of organizations managing the schools in an administrative area, and have no direct influence over academies or MATs. In addition, over the last decade, national government austerity measures have led to public sector funding cuts. These have severely depleted LAs' abilities to provide specialist services. For example, the Association of Directors of Children's Services [25] (pp. 9-10) recently reported that: Most public services were faced with growing levels of need in the community and year-on-year funding reductions. In 2019, the Local Government Association estimated that children's social care was facing a £3.1 billion funding gap by 2024/25.
Together, these shifts mean that LAs typically have very limited capacity or opportunity to respond to growing fragmentation and levels of need.

Impacts on Educational Inequities
There is much evidence to suggest that the changes outlined above have had perverse effects for England's most economically disadvantaged children and the schools they attend. These are seen particularly in relation to attainment outcomes, access to support services, and school admissions.
Over the last five years, attainment gaps between the most and least advantaged have begun to grow noticeably at both primary and secondary level, despite having been relatively stable prior to this [6]. Spatial concentrations of inequity have also deepened.
For example, the Social Market Foundation found that 'comparing the performance of 11-year olds born in 2000 with those born in 1970 reveals that the geographic area a child comes from has become a more powerful predictive factor for those born in 2000 compared to 1970 [26] (p. 5). Quantitative evidence has also shown a deepening of existing inequities arising from the COVID-19 pandemic [27]. Qualitatively, there have been increasing reports of schools serving high-poverty contexts struggling to provide both basic welfare and more specialist health and welfare services for children and families [28]. Often they are attempting to fill the gaps left by depleted LA services, but without the funding, capacity or expertise needed to do so. By contrast, some school leaders have responded to growing challenges by becoming expert at 'protecting' their own schools. Often this has been by intentionally recruiting more advantaged student cohorts to their schools. An independent commission into the implications of academy expansion [29] reported that some academies were using their freedoms over admissions effectively to exclude children from poorer backgrounds. For example, it found some schools intentionally made the language on application forms inaccessible to some parents.
Another concern has been the emergence of 'off-rolling', defined as 'when a child is removed from the school roll for the school's benefit, rather than in the child's best interests' [30] (no page number). A recent study [31] reported that while behavior was widely cited as the reason for removing a child from school, 'most teachers feel confident in agreeing that off-rolling is done to fix statistics for the benefit of the school' (p. 8) in high stakes attainment tests. This study also found that disadvantaged students were perceived to be disproportionately affected by off-rolling. Furthermore, they were then at risk of becoming trapped outside the school system because LAs were often unaware they were not in school and had limited influence to arrange new school places in any case.
Poorly performing schools within MATs have been affected by similar practices. For example, there have been instances of schools serving intensely disadvantaged populations that have been pressured to become academies and join a MAT. They have then subsequently been removed from their MAT for continuing poor performance. Mansell [32] has termed these 'orphan' schools-cut off from their LAs and unable to find other MATs to join. Reflecting on such practices, a recent government report stated: 'The Department for Education . . . has not set out a clear plan for how . . . it will ensure that trusts are being set up to best support pupils' and local need' [33] (no page number). It also notes reports of dubious financial practices in some MATs.
This picture of the current English context is far from comprehensive. Nonetheless, it serves to show that the English case is an extreme example of a growing international direction of travel: a competitive and fragmented school system, operating in an unequal society, in which public services are being reduced and inequities are increasing. This also makes England an important case for the purposes of this paper. If, as we will demonstrate, there are still spaces within these current arrangements where local practitioners can begin to remodel some aspects of the system on more equitable lines, it seems highly likely that similar, and perhaps greater, opportunities will exist in other market-driven systems internationally.
With this in mind, we turn now to outline the nature of our engagement with emerging local developments in England. We then present four examples, each of which demonstrates a different source of leadership for equitable reforms.

Methods
The examples we report are drawn from a 15-year research program on equity within the English education system. This has been underpinned by a series of evidence-based working processes, designed to support equitable developments in the field. Fully explicated elsewhere [18,[34][35][36][37], in summary, these assume that efforts to promote greater equity-and therefore the research that supports them-must:

•
Seek to surface and challenge taken-for-granted assumptions about how education systems should operate; • Focus specifically on the poorest learners, places, and schools that serve them; • Value different forms of knowledge and expertise held by multiple, and often unheard, professional and community stakeholders; • Intentionally move this knowledge and expertise around within and across systems to enhance learning about possibilities for more equitable practices; • Support forms of collaborative action that seek to use such knowledge and expertise to stimulate more equitable practices; • Build the capacity of local stakeholders to develop and sustain collaborative action.
Often, we have been commissioned by emerging initiatives to support them in realizing these principles. In doing so, we have typically employed a design-based implementation (DBIR) research approach [38,39]. For our purposes, this approach has four important characteristics aligned with the principles above. First, it assumes that the context where an intervention operates-in our research, high poverty places in England-is integral to its design. Second, it treats interventions as dynamic, assuming that they will learn and develop over time through taking action. Third, it involves intervention leaders and researchers working in partnership and bringing their collective expertise and critical scrutiny to bear on developments. Fourth, these practitioner-researcher partnerships can support an iterative process of moving back and forth between context-specific and more general concerns. This can both strengthen local interventions by drawing on wider knowledge, while also supporting the development of transferable theories from local practices.
Operationally, we have employed a wide range of qualitative and quantitative methods as suited to individual project requirements. We have also worked with samples of diverse stakeholder groups, sometimes training them as co-researchers to lead research activities locally. Our engagement with local developments has also tended to be long-term. Often there has been an initial period of quite intense research activity over a six-month to twoyear period. This has then been sustained through ongoing 'light touch' engagements over a number of years. This prolonged engagement has enabled us to identify emerging bottom-up reform efforts that, while they inevitably remain works in progress, also show some evidence of sustainability and gains in equity over time.

Emerging Equitable Practices in England
In what follows, we present four examples of efforts to develop more equitable educational arrangements, emerging in high-poverty contexts in England. The first three examples have been selected to illustrate that, in our experience, leadership for such developments tends to come from one of three sources: (i) schools working in more-or-less formal local alliances; (ii) LAs seeking to redefine their relationships with schools; and (iii) third-sector organizations lying outside the education system, but that are already working to address wider determinants of poor educational outcomes, for instance, in health or housing. While these examples have built, to some extent, on pre-existing relationships, the fourth has been selected as indicative of the role that universities, or other similarly placed research organizations, can play to catalyze developments in contexts without favorable foundations.

Schools Working in More-or-Less Formal Local Alliances
There is increasing evidence that collaboration between schools has enormous potential for fostering system-wide improvement, and particularly so in high poverty contexts [40]. This can be about transferring knowledge and expertise between schools, or by generating shared context-specific knowledge on which schools can act collectively.
We have worked in a variety of places where groups of schools within an areawhether defined administratively, or by the way people live locally-have developed such collaborative strategies to support their collective improvement. In some instances, these have focused most strongly on providing comprehensive pastoral and welfare supports. In others, as in the following example, they have maintained a tighter focus on improving attainment.

Example 1: An Area-Based Partnership Developed by Schools
This example involves a local area-based partnership of 15 schools in the same LA: 12 primary, two secondary and one special, including a diverse mix of faith schools, academies and LA-maintained schools. Stimulated and strengthened at various times by government initiatives targeting high-poverty places, the partnership has been maintained by participating schools in various forms for over 20 years. Throughout, it has maintained a focus on school improvement.
Its latest iteration is in response to the LA reducing its role as a direct provider of school improvement services. The partnership decided to step into this role for its schools, though other schools can also buy into its offer. The 15 school principals meet monthly to make collective decisions about the partnership's offer, their aims being to: • Secure the highest standards of teaching, learning and achievement for all learners in the area; • Improve opportunities for all learners and ensure the effective use of resources; • Provide evaluation and challenge based on trust and reciprocity; • Contribute to the professional development of all staff; • Disseminate good practice across the partnership of schools.
To realize these aims, the partnership has developed a complex and continually evolving model for school improvement. For example, it has worked with a leading national literacy charity and local library services to co-create a reading quality mark for its schools to follow which emphasizes reading for enjoyment. It has also created 'tripods' of primary schools, grouping these together by geographical proximity, with representatives from the partnership's two secondary schools also invited to join. Each tripod works on a particular school improvement issue and then feeds back to the whole partnership. This arrangement has enhanced the partnership's knowledge and capacity to learn. The schools all make an annual financial contribution to running the partnership. This covers the cost of employing a full-time partnership coordinator who manages its activities on a day-to-day basis.
This example demonstrates that even when the schools serving a high poverty context have different affiliations and are concentrated on improving attainments, local collaborations can still powerfully contribute to more equitable arrangements. These can include the pooling of resources for mutual benefit. The small-scale geographical nature of the partnership also appears fundamental to this. The partner schools all have a shared interest in the distinctive challenges and opportunities of the area they serve. This simply cannot exist in the same way in their wider affiliations. Shared leadership has also been an important feature in developing a sense of common purpose.
It is important too, that the schools involved have not had to compromise their aims to work within existing inflexible structures. Rather, they have developed funding and operational arrangements that support their co-determined aims. Although these are largely informal and there is no requirement to comply, the schools have clearly found value in these as evidenced through their ongoing voluntary adherence.

LAs Redefining Their Relationships with Schools
While LAs have often largely stepped back from managing educational arrangements, they nonetheless remain distinctively placed within the English education system. However emasculated, they still operate at a broader scale than local school collaborations, so have the potential to engage with a wider range of schools and services. Furthermore, unlike MATs, some of which have large numbers of schools in different parts of the country, they remain geographically anchored, with some understanding of local places within their administrative boundaries.
LAs also still provide public services, however reduced, and not just to children and schools, but also in relation to housing, the environment, transport, leisure, business, and local democracy, for example. This means that they are still one of very few organizations with the potential to engage with what Anyon [41] (p. 5) calls the 'educational policy panoply'-i.e., the range of policies and services distal to, but nonetheless influential in, shaping and responding to educational inequities.
The following example illustrates how some LAs are seeking to capitalize on their distinctive position. They are using the breakdown of their historic roles as an opportunity to redefine their relationships with schools and remodel their services accordingly.

Example 2. Redefining LA-School Relationships
This example focuses on a relatively small LA, with approximately 150 primary schools and 50 secondary schools. The local area has seen a sustained widening of attainment gaps in recent years, with the number of students meeting national expectations being about 25% higher in the LA's most advantaged neighborhoods compared to its most disadvantaged.
Schools serving the LA's poorer communities have also faced pressure to become academies. While this seems to have worked well for some, others have been left feeling isolated and ill supported. In addition, funding cuts have considerably reduced the LA's capacity to provide specialist support services. When schools have bought in external services, these have sometimes been of poor quality.
In response, the LA has begun to redefine its role, taking on three main functions. First, it has positioned itself as a 'match-maker' between schools and the MATs they could potentially join. It is actively seeking out MATs whose operating principles are well aligned with commitments to: (i) supporting schools to work together on an area-basis; and (ii) maintaining supportive relationships with the LA. It sees this as important to ensure that schools remain connected to wider service and policy developments in their administrative areas.
Second, the LA has grouped all schools (academies and LA-maintained) into what it terms 'area-based hubs'. These are intended to serve as a platform for schools, services and wider stakeholders to share local intelligence, and develop collective strategies and actions in response. Practically, the hubs are led by local steering groups, including staff from schools, services, community organizations, and local residents. To facilitate this, the LA has funded the appointment of a coordinator for each hub. It has also appointed a senior member of the LA's Children's Services department to each steering group. This supports the hubs' access to wider services and resources, whether held by the LA or other organizations. As hubs have started to work on local priorities, the LA has also established a program of cross-hub meetings to enable shared learning.
Third, the LA has begun to develop plans to reconfigure the services it offers to schools to ensure they reflect actual needs within the local hub-areas and are accountable to those they are designed to benefit. To this end, it has piloted the introduction of communitybased accountability processes alongside statutory reporting requirements. The LA plans to operate these services through some form of specially created not-for-profit company, which hubs collectively, and schools individually, can buy into. The LA anticipates that, through such arrangements, it can: (i) increase its current capacity to support schools and families; (ii) quality assure its services, so schools are not vulnerable to investing in poor external support; and (iii) enable swift access to services. This example suggests that LAs have potentially important new roles to play within the English education system. In particular, it indicates that they can create new partnership arrangements for their own work with schools-both across the whole LA-area and local neighborhoods within this. This suggests that LAs retain at least some potential to build capacity within local education systems to support school improvement and meet children's and families' wider needs effectively. Moreover, they may be able to act in various ways to counter growing fragmentation, whether by catalyzing local collaborative arrangements, or by brokering positive school/MAT/LA relationships.

Third-Sector Organizations Outside the Education System
The previous two examples have explored how, on different scales, schools and LAs can work to counter the fragmentation of the school system caused by market-driven policies. However, two challenges remain. The first is how to re-connect schools to learners' neighborhood communities where the link between these is broken. The second is how to make visible and then better support children within these communities who have become 'hidden' to the system-whether through off-rolling or in other ways.
We have found many examples that suggest that third sector organizations can have particularly important roles to play in addressing these issues. These may include non-profit organizations with expertise in providing particular kinds of services. For instance, they might provide specialist support for speech, language and communication development, or in working with specific groups, such as young people at risk of gang membership.
The following example is illustrative of this approach. It focuses on a housing association's role in supporting school-readiness and early school success for children who might otherwise be overlooked within the system.

Example 3: Improving Outcomes on a Social Housing Estate
This housing association provides accommodation for low-income families at social rents through housing stock that it owns on a small estate within a city. This is similar to a housing project in the USA.
In recent years, the inner-city borough in which the estate lies has been gentrified. While the borough has benefitted from the city's economic growth, the estate has been largely isolated from this. As a result, it remains characterized by poor outcomes across multiple domains including health, crime, economic activity, and education. Its residents also often lack the necessary skills and income to access new opportunities and facilities in the borough. Around 50 per cent are first generation migrant families who can easily find themselves socially isolated. Typically, they also have limited knowledge of how to access services.
One of the housing association's central goals is to achieve long-term reductions in child and family poverty on the estate. It has therefore invested significant time and resources in establishing partnerships with a range of LA and voluntary and community services. Together they work on increasing: (i) access to employment; (ii) the help available to parents through informal support networks and training; and (iii) the school-readiness of the estate's kindergarten-aged children (ages 3-5).
However, the housing association has faced particular challenges in trying to connect with education providers to support this third aim. While there is a kindergarten on the estate, there are no schools. The area school system is also extremely competitive and fragmented, with schools tending to focus intensively on their internal improvement at the expense of partnership activities. Children also typically travel around the city to access schools, which often have a geographically disparate intake.
The housing association's own research has revealed that although around 600 children live on the estate, because they attend such a wide range of schools, they are only ever a small proportion of a school's total student population. This means they can easily become 'hidden' within these. It has also found that schools often do not recognize that these children may face barriers to learning relating to their residential context, or appreciate that their parents might have little knowledge of the school system and its expectations.
In response, the housing association has set out to create formal mechanisms for linking schools to the range of support services working on the estate. As a first step, it has identified a local primary school that has about 20 children from the estate in its annual intake. It has then used the shared goal of promoting school readiness as the foundation for a three-way partnership, underpinned by a Memorandum of Understanding, between itself, the school, and the estate's kindergarten. Together they have developed a 'Ready for School' (RfS) project, with the housing association successfully applying for funding for a dedicated teacher to work on this.
This project provides support for children on the estate who secure a place at the school for the six months before they are enrolled, through to the end of their first year. Its aims are to: (i) support successful transition to, and learning and well-being in, school; (ii) help parents to better support their children's learning; (iii) reduce wider barriers to family well-being and ensure families can access and take up services; and (iv) strengthen school-home-neighborhood links.
To realize these aims, the RfS teacher works with families to prepare their children for school, linking them to the estate's kindergarten and other learning activities. In addition, one of the housing association's parent advisors works alongside the RfS teacher to link parents to the housing association's wider services as required. These range from support with budgeting and accessing employment, to informal networking. Once a child starts school, the RfS teacher works closely with the class teachers to help the child settle in school, support learning, and provide an active link between school and home.
This example suggests that leadership for equitable educational developments need not always come from within an education system. As this example shows, external organizations operating in high poverty contexts can have the potential to catalyze new approaches and enrich the system. They can bring additional expertise, knowledge and resources.
Our experiences suggest that only particular kinds of external organizations may be able to play these roles. In this example, staff working for the housing association have significant understanding of the experiences of children and families living on its estate. This points, again, to the need for those leading equitable developments in high-poverty places to be closely engaged with their local contexts.
Based on this knowledge, the housing association chose to make itself accountable for improving school readiness and supporting transition to school. It has committed its own resources and partnership arrangements to this, despite there being no requirement for it to do so. It is worth noting that this is a very different scenario to the entry into the English education system of what may be seen as 'rogue players', who run MATs with dubious financial practices.

Building the Foundations for Local Collaboration
The three examples we have reported so far all build on pre-existing foundations, whether informal school networks, LAs' previous relationships with schools, or a housing association's close relationships with tenants and wider services. However, there will be places where such foundations are weak and need to be developed.
At times, we have taken on the challenge of bringing disparate and competing organizations together to stimulate developments to enhance equity. Often this has involved us initiating DBIR projects where we support multiple schools within an area-individually, in pairs and trios, and as area networks-to interrogate their own internal practices, explore how these might be made more equitable, and share their learning and expertise [34,35]. Sometimes this has been at the invitation of LAs, and they too have engaged in these processes. One such example is outlined below.

Example 4: A Research-Based Teaching School Alliance
This example involves a group of around forty primary schools in one LA working in partnership with a team of researchers from the University of Manchester. The schools followed a model of collaborative action research to explore new ways of supporting disadvantaged learners (see [34] for full details). This was designed to draw together teachers' professional knowledge with the researchers' wider research knowledge.
In practice, each school determined its own research focus, for instance, identifying issues of concern around school attendance, or the integration of newly arrived students. To help them explore these issues in depth, the schools were encouraged to use an 'ecology of equity' developed by the university team, to help guide their thinking [34]. This focuses attention on:

•
Within-school factors, arising from school and teacher practices. These include: teaching strategies, how teaching is organized and students' engagement in learning; the ways in which the school responds to student diversity; and the relationships the school builds with families and local communities.
• Between-school factors, arising from the characteristics of the local school system. These include: the ways in which schools compete or collaborate; the processes through which students with similar backgrounds are concentrated in different schools; and the distribution of educational opportunities across schools.

•
Beyond-school factors. This far-reaching arena includes: the demographics, economics, cultures, assets and histories of the areas served by schools; and the wider policy contexts in which the schools operate.
Teachers and university researchers worked together to collect and share evidence about the school's practices and from the wider research-evidence base. This allowed the schools to develop a richer understanding of what was happening to learners in their particular contexts to lead to inequitable outcomes. It also helped them to think critically with university researchers about how they might need to respond. In addition, the project was structured to support participating schools to share their learning and practice, and act as critical friends to one another.
As the project evolved, the schools increasingly took ownership of and led their own research activities. They also orchestrated opportunities to visit, work with, and learn from one another. These collaborative dimensions were sustained long beyond the university's involvement.
This example suggests that even in contexts where schools do not have strong preexisting relationships, universities-or other similarly placed research organizations-can help to develop strong and potentially sustainable networks of research-engaged schools. Notably, the research process employed in this example also requires schools to act against their competitive instincts. It encourages them to move away from a dominant view of teaching as a 'personal and private activity' and to adopt the 'more risky view that teaching activity can be continuously improved if it is made public' [42] (p. 13). In this instance, this has been through schools sharing with others locally, in a safe and reciprocal learning process.
This example also shows research as a process that can work for and with schools. It both values teacher knowledge and has the potential to create powerful interruptions in schools' existing practices. These can draw attention to new possibilities, rather than leading to summative judgements about what is or is not 'working' [35].

Drawing Out the Lessons
Reflecting on these four examples it is important to acknowledge that each remains fragile and imperfect, with its own limitations, difficulties and risks. It would, moreover, be extremely naïve to underestimate the political complexities entailed. Much still depends on maintaining goodwill, sufficient stability and resourcing, and favorable inspection outcomes, among many other factors.
Despite these potentially unfavorable circumstances, the examples nonetheless indicate that professionals working in and around schools can influence local educational arrangements to make them more equitable. Together they suggest that:

•
Area-based school-to-school collaboration can better support improvement for all; • This requires an emphasis on collective action, not just individual accountability; • LAs, rather than being seen as a 'dead hand' that slows system improvement, still have vital strategic, coordinating and quality assurance roles to play; • In particular, LAs can help to counter some of the vagaries of opening service provision and school management to the market; • Leadership for equitable developments need not always come from those directly engaged in schooling. Some third sector organizations may be well-placed to do so and less constrained by education system arrangements; • Equitable reform efforts have to recognize the value of local professionals' knowledge of working in particular high-poverty contexts, and of university research that can help to surface this and enable its use.
Our examples also appear to speak to Valli, Stefanski and Jacobson's [43] argument that school leaders working to enhance equity 'must learn to navigate 'enduring' dilemmas, not simply solve problems' (p. 34). These dilemmas include the need to open school boundaries to parents, the local community and other agencies, while also 'protect[ing] the core . . . of teaching and learning against unnecessary and unproductive intrusions' (p. 34). Leaders' skills in managing such tensions and building consensus and reciprocity around a shared sense of educational purpose therefore appear crucial. The more extensive developments become, and the more widely leadership is shared, the more these skills will be tested.
Keeping this analysis in mind, we turn now to consider the transferable principles for promoting greater equity within market-driven education systems, that can be generated from our empirical examples.

Transferable Principles
Fundamentally, we conclude that market-driven education systems must do more to recognize the importance of local contexts in shaping system arrangements. Whilst the introduction of market forces may have been widely seen as a means to break the link between education, disadvantage and place by allowing parents to choose schools outside their local areas, and in England, by enabling schools to join MATs with others from outside their locales, the link remains and cannot be ignored. Many children in high poverty contexts will still attend schools in those contexts. Whatever their wider affiliations, schools serving those contexts will still need to address barriers to learning arising from children's home and community contexts. Our third example suggests that even when children travel to schools outside their local neighborhoods, the challenges they face relating to these contexts remain, even if schools do not recognize them.
All of this strongly suggests that equitable reforms must actively engage, both conceptually and pragmatically, with the challenges and opportunities presented by high poverty contexts. Conceptually, an active engagement with places as they are 'lived' appears vital. Whatever tangible or measured characteristics high-poverty contexts share, they are also to some extent unique. They are continually shaped and reshaped by the ways in which their geographical, physical and service infrastructures, histories, cultures and demographics, and the norms, networks, and behaviors of the people who live there, interact and change over time [44,45]. In addition, pragmatically, however austere the times, high-poverty contexts still hold resources that can valuably support children's education. For example, these may be public or third sector services operating locally, or other kinds of tangible and intangible assets [37].
Drawing these points together, Smith [46] has argued that while some inequalities can only be addressed at the national government level, there are others that can only be understood and addressed at a local level, by local actors with a deep knowledge of the places where they work, and who can access local resources. Our examples point to the power of education and related professionals acting on their local knowledge and harnessing local resources to develop coordinated and strategic actions. In doing so, they also diverge markedly from the approach currently being pursued by the English government as part of its new 'levelling up' strategy [7]. This does more to treat high-poverty areas as 'containers' into which interventions that 'work' can be dropped, rather than as dynamic entities that present both challenges to, and opportunities for, equitable reforms.

Further Implications
We have established the basic premise that close engagement with high-poverty contexts, understood as dynamic entities, is essential to equitable reform efforts. If this is accepted, it points to further finer-grained principles needed to guide equitable reforms.
Here we outline four principles that appear of particular importance.
First, our findings suggests that schools that have become disconnected from their local contexts need to find ways to reconnect, whether by (re)engaging with other local schools, LAs (or their equivalents in other contexts), and other local service providers. For schools that draw their student cohorts from across diverse geographical areas, connecting with organizations active in the neighborhoods where their most disadvantaged students live, appears particularly important. We are not suggesting that this reconnection should be at the expense of schools' wider affiliations and the benefits they can gain from these. However, we are arguing that schools' wider affiliations cannot substitute for the value of local collaborative activity.
Second, our experiences suggest that local collaborative activities require schools and local government (through LAs or their equivalents) to exercise some collective responsibility for all the children in an area, not just those attending particular schools. Notably, in our examples, the arrangements made to support this are not hard bureaucratic structures with associated punitive accountability regimes, but are 'softer' and more flexible in nature. Although, on the one hand, this makes them inherently fragile, on the other hand, they appear to be better able to support ways of working that can develop trust, reciprocity and collective responsibility between partners [47].
Third, as we have indicated, local collaborative activities do not always need to be initiated and led from within the education system, whether by schools or local governmentthough some may wish to lead. External organizations with deep knowledge of highpoverty contexts and particular expertise in working within these, might equally take a leading role. This may be especially so with regard to addressing inequities that could otherwise be overlooked within current educational arrangements. Schools and LAs (or their equivalents) must, however, be active in helping to develop and sustain partnerships between and beyond schools, where partners multiply the impacts of each other's efforts.
Fourth, all of our examples suggest that to develop meaningful partnerships, schools and their partner organizations must be willing to work beyond their traditional institutional remits. They indicate the importance of investing in staff who can work to span the boundaries between schools, and between schools and local government services, housing providers, and universities. These 'boundary spanners' [48] (p. 532) appear essential to establishing new shared spaces in which partnerships can work, that lie beyond restrictive organizational boundaries. Integral to this, they need to be able to develop strong relational skills [49] and exercise these consistently over time. This is important because partnership working is as much a social as technical exercise. Partners have to be able to build trust, deepening this through their ongoing actions. As previous research has shown, partnerships typically fail when they are short-term and contrivedi.e., when partners are primarily focused on how they can use others' resources to help meet their own performance targets, rather than on a shared commitment to achieving equitable change [18].
To help to realize these four principles in practice, national policy makers will need to recognize, value and 'unlock' the potential power of local collaborative activity. Key to this, they will need to create and protect spaces for local policy development and to trust local government, schools, and third sector organizations to act within these to pursue greater equity, rather than seeking to micro-manage the details of local strategies and actions. They will also need to encourage greater boundary crossing and joining-up in national and local government policy, so that connections are also made at these levels between education, housing, health, and so on. This is necessary to reflect challenges and opportunities experienced locally, and to help to create the conditions in which these can be addressed more effectively [9].
This, in turn, will require governments to develop accountability arrangements that can accommodate local aims and more complex collective ways of working. While market systems demand a focus on single institutions held individually to account, to move in more equitable directions, centralized accountability systems will have to develop a greater tolerance for what Honig [49] (p. 531) terms 'means-end ambiguity'-i.e., a lack of clarity about individual contributions to particular outcomes and shared goals. She suggests that one way in which this might be addressed is for accountability processes to focus as much, if not more, on partners' equitable intents and evidence of these being operationalized, as on outcomes. This could also help to support 'soft' collaborative arrangements, rather than cutting across them.
To be clear, we are not suggesting that centralized accountability should be any less important in pursing more equitable arrangements-and the development of MATs in England serves as an indication as to why this is required. However, we are arguing that accountability arrangements must be appropriate for the ways of working required to promote greater equity, and the risks inherent in these.
Following from this, we wish to propose a final overarching principle. This is the need for intentions and actions to promote greater equity to be aligned at all levels within education systems-from the very local, to the national. Without this, equitable developments of the kinds we have detailed here will inevitably be exceptional rather than commonplace. They will be confined to those spaces carved out by local practitioners who are willing to take risks as they seek to square-the-circle of trying to achieve academic excellence alongside greater equity.
This, in turn, points to the need for a commonly articulated sense of ethical leadership for more equitable system arrangements, underpinned by the principles identified above. This is important because while there is a substantial literature on ethical leadership, this is widely understood in terms of 'normatively appropriate conduct' [50] (p. 120). However, what is normatively appropriate is contested, and especially likely to be so in market-driven systems where demands for greater equity and excellence can be contradictory in practice. Indeed, we suggest there may be a stronger public sense of what unethical leadership practices look like, whether manifest through off-rolling or cheating scandals on high stakes standardized tests [51], than of practices that can move market-driven systems in more equitable directions. Our empirical examples, and the principles we have drawn from these, offer a first step in addressing this situation. If these were widely articulated and acted on, emerging developments could powerfully reinforce one another, and begin to establish distinctive normative expectations for future equitable reform efforts.
Our work also suggests an important role for design-based research approaches in supporting this. We have briefly outlined how the distinctive principles underpinning our own DBIR studies have enabled us to become more-or-less active partners in local systems change efforts. They also indicate how policy makers and research funders may need to change their thinking to support equitable developments more widely. For example, they may need to: (i) fund longer-term light touch research engagement, not just initial activity; (ii) value processes as much as outcomes; and (iii) be less prescriptive about the outcomes, timescales and target groups on which they expect researchers to focus. Funding partnerships between universities and school districts or other collaborative arrangements, in addition to more tightly specified, fixed-term and experimental projects, appears a valuable way forward.

Conclusions
At the outset of this paper, we posed the question 'how can greater equity be promoted within education systems based on market forces?' The examples we have reported here may be fragile, localized, and inevitably limited in what they can achieve. Nonetheless, they point to how more equitable arrangements could be developed and the conditions needed to support this. They suggest that equitable systems must be connected systems, and that schools need to work together, and with other organizations, both within and beyond their local areas. They also point to the vital role of local collaborative activities in promoting greater equity, and the value of knowledge held by practitioners working in high poverty contexts in informing these developments. They show, too, that pursuing arrangements that are more equitable does not always mean schools individually doing more. However, it does imply partnerships beyond the school, where partners multiply the impacts of each other's efforts.
There are significant implications here for national policy makers. To support greater equity, they need to develop 'intelligent' policy that can learn from and nurture innovative developments. They also have to determine how they can foster the conditions needed to encourage the interpretation of policy on the ground in ways that promote greater equity, rather than being manipulated for institutional gain. To inform this, it will be important to commission research that can identify, support and learn deeply from emerging developments, without being unduly restricted by narrow concerns about 'what works'. Indeed, if Fullan's call for a paradigm shift is to begin to be realized, it will most likely be through finding, supporting and sharing learning from a growing number of 'anomalous developments' [52], enacting the kinds of principles we have proposed here.