Geographies of Exclusion: Rebuilding Collective Responsibility in a Fragmented School System

ABSTRACT The goal of equity in education in England is damaged by regional disparities in outcomes and a marked social gradient in school exclusion. The most vulnerable groups are disproportionately represented in in-year transfers. Drawing on 24 interviews with school leaders and education decision-makers in a socioeconomically deprived area, this study examined institutional strategies to promote inclusion by reducing pupil mobility in an area-based initiative. The analysis highlights the interaction of administrative, professional and market logics, and the significance of the “middle tier” in mediating inter-local tensions. Further research is needed on “hidden” pupil moves and diverse forms of within-school segregation-reintegration.


Introduction
In many countries school reform has sought to achieve a better balance between centralization and decentralization (OECD, 2015).Devolving greater responsibility for operational and governance processes to schools/communities is associated with more efficient and responsive use of resources, increased innovation, and adaptability of services to local needs (Au & Ferrare, 2015;Cheema & Rondinelli, 2007;Moe, 2019).However, more decentralization within national school systems has important implications for the policy goal of inclusive education (European Agency for Special Needs and Inclusive Education, 2016;Robertson & Dale, 2013).An erosion of coordination capacity, coupled with increased inter-local competition, may compound socio-economic and regional disparities in educational opportunities and outcomes (Bush, 2016;Crawford, Maxwell, Coldron, & Simkins, 2022;Greany & Higham, 2018;Lubienski, 2014;OECD, 2018).This article considers the unintended consequences of the promotion of self-managing schools on collective responsibility for high needs pupils in a disadvantaged area.We do this using one indicator of inclusion, namely the extent of inyear pupil mobility between neighboring education providers as result of disciplinary exclusion.Focused on a deprived northern town in England this paper explores the rise in school exclusion, pupil moves that fall short of exclusion, and the drivers and barriers to increased inclusiveness within the local school system.
Place-based oversight of practices of exclusion rely on access to accurate data.There are limitations to official data on the range and frequency of forms of school exclusion and subsequent in-year movement of pupils between education providers (Hutchinson & Crenna-Jennings, 2019;McCluskey, Cole, Daniels, Thompson, & Tawell, 2019).It follows that critical examination of patterns of exclusion demands careful attention to the multifarious ways in which pupil mobility between education settings is defined.The practice of exclusion from school is commonly understood as, "a disciplinary measure imposed in reaction to students' misbehavior (e.g., violations of school policies or laws) by a responsible authority" (Valdebenito, Eisner, Farrington, Ttofo, & Sutherland, 2019, p. 254).Pupil moves as a disciplinary sanction take a variety of forms of which the most widely recognized are formally recorded fixed-term (temporary) exclusions/suspensions (when pupils are withdrawn from school for a number of hours or days), and permanent exclusion/expulsion (when pupils will not return to the excluding school).However, pupil moves in England also include locally negotiated managed moves i.e., a "fresh start" transfer to another school (without a formal record of exclusion), continuing education through Alternative Provision (AP) i.e. placements outside of mainstream schools (which may include full or part-time placements in Pupil Referral Units (PRUs), further education colleges, independent schools and other provision such as home tuition services and voluntary or private sector providers), and transitions to Elective Home Education (EHE) (i.e., home schooling) (IntegratEd, 2020).
Pupil moves that fall short of a removal from school, such as in-school segregation/ seclusion, are less visible in mandatory records.Official exclusion statistics do not include unofficial exclusions where children are removed from school registers while continuing with their education through out-of-school independent and unregistered AP placements (Gill, Quilter-Pinner, & Swift, 2017;Partridge, Landreth Strong, Lobley, & Mason, 2020), or where parents feel they have little choice but to withdraw their child for EHE (Children's Commissioner, 2019) or "consent" to a managed move to avoid the damaging consequences of expulsion, or prosecution for school nonattendance.Similarly, pupils retained in school can be excluded from mainstream classes repeatedly, for extended periods of time, by being assigned to "alternative" or "therapeutic curricula" (Power & Taylor, 2020, p. 874).Despite its prevalence, there is little research on "managed moves" between schools (Gazeley, Marrable, Brown, & Boddy, 2015;Messeter & Soni, 2018), or the operation of "respite" via temporary alternative provision (Jalali & Morgan, 2018;Malcolm, 2018), or the impact of in-school "referrals" ranging from seclusion to schoolbased behavioral interventions (Stanforth & Rose, 2020;Valdebenito et al., 2019).
Obtaining a better understanding of factors influencing decisions to exclude, and the frequency, form, and duration of diverse forms of exclusion is important because the personal and social costs are high.Formal exclusion, persistent absence, and school changes are strongly associated with poorer educational and social outcomes (Partridge et al., 2020).Young people excluded from school have high incidences of mental health issues, (Ford et al., 2018;Tejerina-Arreal et al., 2020), involvement with law enforcement (Arnez & Rachel Condry, 2021), and fewer employment prospects (Gill et al., 2017).Moreover, the most vulnerable children are most likely to experience exclusion from school.Children from low-income families, looked after children (with experience of local authority residential care), minoritized ethnic groups, children receiving support for special educational needs and disabilities (SEND), especially those with Social Emotional and Mental Health (SEMH), needs are overrepresented in exclusion data (Thompson, Tawell, & Daniels, 2021;Timpson, 2019).In England, children from low-income families are four times as likely as other children to be permanently excluded and over three times as likely to receive one or more fixed-term exclusion (Fair Education Alliance, 2018).
The pronounced social patterning within exclusion data stands in stark contrast to an espoused commitment to inclusive education.For Slee (2018) inclusive education "seeks to identify and dismantle barriers to education for all children so that they have access to, are present and participate in and achieve optimal academic and social outcomes from school" (p.2).In the UK and elsewhere, awareness of the cumulative impact of multiple disadvantage has directed policy makers' attention to the importance of place in redressing regional imbalances in educational outcomes (Chapman, Drever, McBride, Orr, & Weakley, 2019;Duveneck, Grund, de Haan, & Wahler, 2021;House of Commons, 2021;Kerr, Dyson, & Raffo, 2014;Szelei & Alves, 2018).While evidence on the magnitude of neighborhood effects is mixed, it is generally agreed that living in a poor neighborhood for an extended period intensifies educational disadvantage (Nieuwenhuis, Kleinepier, & van Ham, 2021;Türk & Östh, 2019).As a consequence, education-oriented area-based initiatives have targeted additional resource to enhance professional collaboration across institutional and sectoral boundaries.
The locational challenges of socio-economic deprivation and a marginalized pupil body influence the work of school leaders (Ovenden-Hope & Passey, 2019).Winter, Hanley, Bragg, Burrell, and Lupton (2020) maintain that "schools serving areas of disadvantage and poverty do things differently in practical terms as compared to those serving more affluent areas" (p.392).Leaders exercise agency in prioritizing social and emotional wellbeing, and direct attention, energy and resource to a range of support services including breakfast clubs, food banks, uniform swaps, equipment/kit loans, and emergency loans to families/carers.Transient populations with high rates of in-year admissions require continual assessment of the needs of newly registered pupils, with attendant monitoring and responsive action.School leadership teams devote greater time to safeguarding and multi-agency liaison.Winter et al. (2020) describe how leaders engage in everyday "quiet (small p) politics" and "quiet activism" to sustain education and pastoral services in times of fiscal austerity (p.394).Leaders engaged in such tactics draw upon humanistic principles that attend to children's wellbeing, combined with an ecological understanding of the contexts in which children live (Hanley, Winter, & Burrell, 2020).Leading schools facing challenging circumstances requires critical reflexivity, moral literacy and ethical practice (Lowery, 2020).For example, recent work on relational justice (Laing, Mazzoli Smith, & Liz Todd, 2018) and curricular justice (Mills, Riddle, McGregor, & Howell, 2021) draw attention to affective dimensions of marginalization and the stigmatizing effects of exclusion from mainstream norms.
However, the prevailing policy context of test-based accountability, a more "rigorous knowledgerich, academic curriculum" (DfE, 2016, p. 24), and standardized assessments for the purposes of competitive comparison creates "perverse incentives" toward exclusion (Thompson et al., 2021, p. 34).A highly politicized imperative to "close the attainment gap" in social mobility "cold spots" paradoxically undermines inclusive approaches to behavior management.Pupils deemed troublesome may be subject to "off-rolling"/school push-out via punitive zero tolerance policies that do not consider individual needs (Done & Andrews, 2020).In England, the school inspectorate define off-rolling as, The practice of removing a pupil from the school roll without a formal, permanent exclusion or by encouraging a parent to remove their child from the school roll, when the removal is primarily in the interests of the school rather than in the best interests of the pupil (Ofsted, 2019, n. pag.).
Exclusion levels peak as the national assessment period approaches (Partridge et al., 2020).Geographical proximity of schools may permit greater churn where other providers are within acceptable traveling time.Funding constraints alongside escalating demand for additional support creates difficult choices.A decision to formally exclude will transfer the costs of pupil support from the excluding school to the Local Authority.Between 2014-15 and 2020-21 despite increasing cost pressures (associated with rising staff costs) average per-pupil funding to schools barely increased (0.4%) (Davies, 2021, p. 6).Moreover, from 2017 there has been a relative re-distribution of funding from the most deprived schools to the least deprived schools (op cit., p.10).In England, pupils with SEND account for 45% of permanent exclusions and 43% of fixed term exclusions (DfE, 2020).The requirement for self-managing Academies to report directly to the Department of Education (rather than their Local Authority) reduces available mechanisms to assure place-based accountability.

Research Aim and Questions
This study explored how school leaders in an area of acute deprivation cooperated to reduce school exclusion rates through an area-based initiative, the Continuum of Provision (COP) project, 2019-2020, 2020-2021.In the context of an increasingly fragmented school system, this is important because it casts light on the potential for divergent interpretations of inclusion among key stakeholdersindividual headteachers, Multi-Academy Trust (MAT) leaders and Local Authority officers -with practical implications for how inclusive practice is measured, supported and sustained in the longerterm.The analysis goes beyond topline findings expressed in terms of pupil mobility between education settings, to explore in greater detail how educational experiences were reconfigured for high needs pupils with previous records of exclusion or deemed "at risk" of exclusion.It offers a nuanced examination of how in-year school transfer and reintegration processes were mediated by strategic actors within one locality and during a period when continuity of education was further challenged by pandemic-induced intermittent school closures.It examines how competing motivations were reconciled including professional, institutional, regional and sectoral concerns.The analysis explores how leaders generated bespoke responses to locally assessed needs and how far these reflect a commitment to public leadership for the common good.
The purpose of this paper is not to reinforce deficit or salvation narratives around failing schools, troubled/ troublesome children and problem communities but to explore the complex interaction of policies for schools that purport to address and yet may compound educational disadvantage.In this article we purposively report instances of affirmative action as well as identification of constraints.As Thomson (2002) has observed, schools are context-generative as well as context-derived.We follow Passy and Ovenden-Hope (2020) in directing attention to, "school leaders' enactments of social justice, carried out as they negotiate a path through their own moral imperative to improve young people's lives with government demands and their school's own particular situation" (p.223).Using the case of a place-based approach to reduce school exclusion in an area of socio-economic disadvantage, we examine the different ways in which school leaders in individual schools are positioned, and position themselves, in relation to the espoused goals of a time-limited project.We consider the interaction of headteacher agency, school context and national policies for schools.We explore how school leaders established an uneasy local alliance amid a steady state of "coopetition" (Muijs & Rumyantseva, 2014) while subject to close external scrutiny.
In interrogating the rationale, design and enactment of the Continuum of Provision project, the following questions are addressed: (1) How is inclusion-exclusion understood by school leaders?
(2) How have school leaders approached the development of more inclusive practice?(3) In what ways is collective responsibility constrained or enabled?

Context
This research was conducted in a coastal town in the north of England which has a predominantly white working-class population of approximately 140,000.Once a popular seaside resort, the town is an area of acute socio-economic deprivation with eight of the ten most deprived neighborhoods in England.Compared to local authorities with similar characteristics, the area has a higher proportion of workless households, households experiencing in-work poverty, and the worst life expectancy in the UK (Chief Medical Officer's Annual Report, 2021).Referrals to Children's Social Services per annum exceed 1,000 per 10,000 (compared with 544.5 per 100,000 for England).There are high levels of population inflow and outflow due to low paid seasonal and intermittent employment generated by the tourist economy and the availability of cheap housing stock in poor condition (ONS National Statistics, 2020).Around 14% of the total school roll per annum moves in and out of schools outside the usual times of joining and leaving.Over 30% of children live in income-deprived families.In 2021, the percentage of secondary pupils eligible for free school meals (eFSM) (i.e., pupils whose parents/ carers receive income-related benefits) was around 40%, compared to 18% for England.
In 2019, the town's secondary schools Progress 8 scores were among the lowest for all English local authorities.Progress 8 is a value-added measure of the progress a child makes from the end of primary school (Key Stage 2) to the end of secondary school (Key Stage 4).Since 2016, under 45% of sixteenyear-olds in the town achieved a grade C or above/now grade 4 (of 9) in the General Certificate of Secondary Education (GCSE) examination in English Language.In 2019, at the end of Key Stage 4 (age 16) a disadvantaged pupil resident in the town was on average 26-months behind their peers in overall attainment for mathematics and English.This is consistent with growing evidence that disadvantaged pupils in coastal areas in England fare less well than disadvantaged pupils in other geographical areas (Harrison, Patel, Francis, & Hicklin, 2019;Ovenden-Hope & Passy, 2015, 2019) inviting critical scrutiny from the schools' inspectorate (Colman, 2021).
The town is served by eight academy high schools.The schools are in close proximity, under five miles apart, with pupil rolls between 670 and 1200.All the schools moved from local authority control between 2012 and 2016 as a result of the national policy of academization i.e., the promotion of a statemaintained yet independent schools' sector (Courtney, 2015;West & Wolfe, 2019).Sustained low performance in some schools resulted in mergers from a stand-alone academy to a Multi-Academy Trust (MAT), or re-brokering to an alternative MAT initiated by the Regional Schools Commissioner (RSC).In England RSCs support schools to address underperformance and are accountable to the Secretary of State for Education.Such moves often involve renaming a school, instigating turnaround headship, and sometimes relocating and/or partially or completely rebuilding the school estate (Simon, James, & Simon, 2021).In 2021, the town was served by three converter academies (schools that convert to academy status based on good performance), four sponsor-led academies (underperforming schools forced to become academies and run by sponsors), and a new all-through free school that opened in 2018.At the time of the research, this small group of schools belonged to five different Multi-Academy Trusts (MATs) that varied in size from just three schools to 28 schools across north-west England.Two MATs operated solely in the town.Between 2014 and 2019, six of the eight schools (or their predecessor schools) were judged by the school inspectorate, Ofsted, to be less than Good (i.e., were deemed Inadequate or Requires Improvement).The poor reputation of several schools led to declining school rolls.Parents able to exercise choice can apply for admission to publicly maintained schools in neighboring boroughs or fee-paying independent schools.
In addition to high rates of population in-and out-flow, there was a precedent of high rates of pupil mobility ("churn" or turnover) between academies within the town over an extended period.From 2014 to 2019, compared with primary schools (pupils aged 4-11 years), secondary schools in the town had persistently high rates of permanent and repeated fixed-term exclusion (i.e., where a pupil is excluded from a school for a set-period of time, between one day and 45 days per school year) and absenteeism (particularly persistent absence i.e., the percentage of pupils missing 10% or more of the mornings or afternoons they could attend).For example, persistent absence in 2018/19 ranged from 15% to 40% across the schools.A significant number of pupils disengaged early in their secondary school careers.Between 2014 and 2019, the town's high schools recorded twice the rate of permanent exclusions compared to neighboring authorities, and four times the national rate of permanent exclusions in England.Admissions to the local authority's Pupil Referral Unit (alternative provision for pupils who are not able attend mainstream school) rose year-on-year from 2015, peaking in May 2019 at over 280 pupils.During the academic year 2018/19, 45 pupils were permanently excluded, and 73 pupils transferred to Elective Home Education (EHE, home schooling).The local authority retains control over the allocation of school places outside normal admission times using an In-Year Fair Access (IYFA) protocol.The IYFA Protocol aims to secure access to education quickly for unplaced and vulnerable children, and those who are having difficulty in securing a school place inyear.In the 2018/19 academic year there were 54 re-integrations via the IYFA Panel.The proportion of pupils registered for Alternative Provision (AP) outside mainstream schools was higher than comparator local authorities.
The high levels of permanent exclusion, noted above, placed considerable pressure on the Local Authority and its High Needs Block (HNB) budget.High needs (or "top up") funding leverages additional per-pupil support for pupils with identified additional support needs and alternative provision for pupils who are excluded from mainstream schools.In 2019 the Chief Executive Officers (CEOs) of the five Trusts operating in the town committed to working collegiately with each other, the local authority and the Department for Education to reduce pupil mobility and initiated the Continuum of Provision (COP) project.The COP project was intended to (a) prepare the ground to devolve arrangements for Alternative Provision (AP) to the MATs (i.e., further devolution of funding from central government directly to schools); (b) inject capacity into the school system to develop inclusive practices that reduce exclusion; and (c) support the Local Authority in improving processes for In-Year Fair Access (IYFA) (i.e., supporting the re-integration of excluded pupils back into mainstream schools).From the outset, the COP project was framed by competing logics, balancing institutional autonomy and community benefit.The project sought to combine high levels of MAT autonomy and moves toward higher levels of devolved school-level funding for high needs, alongside a commitment to equity-oriented area-based inclusion and more efficient management of pupil movement.
The COP initiative spanned the academic years 2019/20 and 2020/21 at a cost of one million pounds per annum.The project was jointly funded by the Department for Education Opportunity Area Programme and the local authority (LA).The Opportunity Areas (OAs) programme was introduced in October 2016 to raise education standards and broaden the horizons of young people in areas struggling with social mobility (DfE, 2017).A Service Level Agreement (SLA) was put in place to confirm match funding arrangements.Under the SLA, if an academy were to exceed one permanent exclusion in a term, three in total for the academic year (aggregated at a cluster level for Trusts with multiple academies in the town), they would be held responsible for commissioning a place in alternative provision for the additional permanently excluded pupils.Commissioning costs of £10,000 per pupil (USD 12,300) would be charged by the Local Authority to incentivize schools to retain rather than to exclude pupils.In agreeing to participate, each school used logic models to produce a locally authored plan of bespoke interventions that were designed to promote inclusion, reduce school exclusion, and encourage successful re-integration from Alternative Provision.

Methods
The study employed a two-stage sequential design to examine outcomes and processes (Bryman, 2006;Teddlie & Tashakkori, 2006).An initial quantitative review served to enhance its interpretive intent (Creswell & Plano Clark, 2018).A new dataset was constructed to establish patterns of pupil movement, which then informed the qualitative strand of the enquiry.During the initial scoping phase, researchers liaised with the Local Authority Business Intelligence Team to access and review schoollevel administrative data: number of permanent and fixed-term exclusions, moves to elective home education, number admitted/retained via IYFA (2019-2020).De-identified pupil mobility data for all eight schools was shared securely by the Council (via a data-sharing agreement in compliance with 2018 UK General Data Protection Regulation, GDPR).In addition, the researchers compiled a secure corpus of documents generated during the COP project including school behavior review reports, revised behavior policies, project logic models and monitoring reports.
The second stage involved semi-structured interviews with school and MAT leaders, and council officers engaged directly in managing the COP project from 2019 to 2021.Preliminary analysis of pupil movement data and a review of documentary data informed the interview topic guide.Interviews were conducted between May and July 2021 with twenty-four key informants in the following roles: Multi-Academy Trust (MAT) Chief Executive Officers (5), Headteachers/Executive headteacher (8), Deputy headteachers (5), school inclusion/behavior leads (2), and Council officers (2) and Pupil Referral Unit staff (2).Secondary sources (i.e., project documentation) were used as evidence-based prompts to focus reflection.All the interviews were conducted remotely (seven via telephone and 17 via video call) and were typically sixty-minutes duration.
The research was conducted in compliance with the BERA Ethical Guidelines for Educational Research (BERA, 2018) and approved by the university institutional ethics review committee.Given the time of the research (during a pandemic) and the politically sensitive research context (a designated Opportunity Area), participants' voluntary informed consent was actively sought and revisited regularly.Researchers were aware that headteachers in deprived Northern contexts in England were facing pressure to "close the gap" before the pandemic (Jopling & Harness, 2021;Thompson et al., 2021) and approached potential participants with a duty of care.The limits of anonymity (Dawson, 2014;Tilley & Woodthorpe, 2011) and the risk of deductive disclosure (Tolich, 2004) (or re-identification) were explicitly addressed within the consent process pre-, during and postinterview.This was important given the small participant pool and the public profile of the initiative.To protect participant confidentiality in reporting, direct quotations are identified by a numerical code (in parentheses), and contextual identifiers are removed from descriptions.Comparative school-level data is not included to reduce the risk of identification.Interim and final reports were shared with participants as verbal and written briefings as a member check and to elicit community feedback.
All interviews were audio recorded and transcribed in full for analysis using NVivo, qualitative data analysis software.For the purposes of this paper, the transcript dataset was analyzed through the lens of affirmative action, leader agency and ethical public leadership.This is consistent with a commitment to asset-based (Thomson & Hall, 2017) rather deficit views of "problem" communities, a theorization of agency as achieved (Priestley, Edwards, Priestley, & Miller, 2012), and an acknowledgment that scope for individual and collective agency is influenced by the contingencies of the environment in which it occurs (Priestley, Biesta, & Robinson, 2015).Following an overview of outcomes in reducing pupil movement, the results are reported in three related sections that address the respective RQs: (1) How is inclusion-exclusion understood by school leaders?(2) How have school leaders approached the development of more inclusive practice?( 3) In what ways is collective responsibility constrained or enabled?First, an account is offered of how inclusion is conceptualized (as alternative within-school provision or additional whole-school provision), giving weight to actions in the two schools that showed most improvement in reducing permanent exclusions.Second, consideration is given to how different conceptualizations of inclusion connect with different forms of accountability, specifically the metrics of success that are deemed most appropriate in holding schools to account for progress.The third section considers the implications of how inclusion is defined and measured for the longer-term goal of sustaining collective responsibility for high needs pupils across the town.The themes were identified using both emergent patterns (inductive) and a priori codes (deductive) drawn from the literature on school-to-school cooperation in the context of devolved governance.

Pupil Movement
Rates of permanent exclusion and referrals to the Local Authority Pupil Referral Unit fell in seven of the eight schools between 2016 and 2020, and particularly sharply from the launch of the COP project.The permanent exclusion rate for the town's high schools exceeded 0.85 in 2018/19, falling to 0.20 in 2019/20 (i.e., the number of permanent exclusions as a proportion of the overall school population in an academic year).Rates of fixed-term exclusion (not limited via the COP Service Level Agreement) and number of days lost through fixed-term exclusion fell in just two of the eight schools.Rates for elective home education (EHE) plateaued at the start of the COP project and escalated during the pandemic (from March 2020), in line with the national trajectory.Analysis of movement by pupil characteristics shows a strong and persistent association between exclusion and eligibility for Free School Meals, Special Educational Needs and Disabilities (SEND), and deprivation.A very high number of excluded pupils have recognized additional support needs.Between 2016/17 and 2019/ 20, over 80% of excluded pupils were designated as SEND and 50% were eligible for free school meals.Initial referrals to secondary academies during 2019/20 through the IYFA process were fairly evenly distributed but a high withdrawal rate (over 58%) undermined confidence in equity in admissions.
There are significant blind spots in the official data, notably around managed moves between schools, and on-roll moves to (and from) in-school Alternative Provision (AP).Due to variability in recording it is not possible to determine with accuracy how many managed moves and moves to onroll AP occurred.The validity of pupil movement data assumes consistent use of attendance, absence, and transfer codes (e.g., formal managed moves were also recorded as "school transfer," "back to mainstream" or "other") and this was affected by changes to the way information was recorded over time and between data entry operators.In comparison with off-roll moves, there is scant data on young people supported through in-school AP or dual registration (supported by home and subsidiary school) to enable analysis by pupil characteristic (e.g., by gender, ethnicity, education, and welfare needs).Data from school census records does not capture the non-linear journeys of high needs pupils.Census data is limited where pupil movement rates are high and characterized by multiple short-term moves.The limitations of the data in terms of quality and coverage gave greater weight to the qualitative inquiry reported below in relation to each research question.

RQ1 (Re-)Conceptualizing Inclusion
The following section outlines expressions of leader agency supported through the COP project that were directed at organizational change.Schools varied in the extent to which they operationalized and embedded four key strategies aimed at reducing barriers to inclusion: better use of diagnostic assessment tools; the development of graduated pathways to personalize support; the provision of teacher professional learning for inclusion; and a revised curriculum offer to promote engagement.

Diagnostic Assessment to Address Unmet Need
Before 2019, school-level systems for behavior management and special educational needs were separate.For example, it was not uncommon for the Special Educational Needs and Disabilities Coordinator (SENDCo) to be excluded from deliberation on the strategic direction of behavior support.From 2019, formal COP project monitoring processes (termly meetings) encouraged schools to make better use of learning analytics, to examine behavior referrals for recurring incidents and screen pupils for learning, social and emotional needs.All the schools engaged in the process of data review, but responses varied in terms of the depth of critical scrutiny and how best subsequently to address identified need.Seven schools showed a clear commitment to regular assessment of pupils who were beginning to show disengagement from learning, as well as pupils with registered SEND needs and Pupil Premium pupils (i.e., additional funding to improve the attainment of children from low-income families).Two schools embedded diagnostic assessment activity within a larger cyclical framework or spiral of inquiry (Kaser & Halbert, 2017).In these schools, persistent high levels of challenging behavior were reconceptualized as not residing with individual pupils but as a school-level issue of unmet need.
We needed to home in on unmet need.A significant number of referred pupils showed signs of social, emotional, and mental health needs that we didn't know about previously, because it was entangled with misbehaviour [. ..]We don't want a solely punitive model.The next step is to create an inclusion Center that provides a transformative positive experience for those struggling to self-regulate in school.(10) In the first two years, we were firefighting behaviour.We were very sanction-driven, which worked to some extent in reducing lower-level behaviour but wasn't reducing fixed-term and permanent exclusions.The new mantra of inclusivity is not reducing the number of challenging students we have, but how we manage them and how we help them to progress despite their challenges.( 22)

Graduated Pathways for a Personalized Response
Each school developed graduated responses to address the needs of pupils identified by teachers as "at risk" of permanent exclusion.These ranged from continued referrals to external alternative provision provided off-site for an agreed period, through to time-limited interventions managed within recently established on-site inclusion centers, and blended part-time/flexi-programmes.The range of provision offered in bespoke school "inclusion centers" extended from a separate "school within a school," through to staggered pathways to partial or full re-integration to mainstream classes.For example, one school developed four pathways to support a return to learning for pupils with identified learning, social and emotional needs: (1) Pupils access external alternative provision.
(2) Pupils are taught within the in-school center for all aspects of their curriculum (academic and enrichment).
(3) Pupils access some mainstream lessons, as well as completing sessions (academic and enrichment) within the center.(4) Pupils are fully integrated into mainstream lessons but are withdrawn to undertake specific targeted intervention as part of their identified plan.
Another school offered three options: (1) Bridge -a six-week intervention via in-school Alternative Provision for pupils designated at risk of permanent exclusion.
(2) Nurture -a small group intervention where pupils have an individualized timetable and curriculum, most full-time.
(3) Hybrid -a blended timetable of mainstream classes and bespoke provision.

Professional Learning for Inclusion
The effectiveness of in-school centers was deemed dependent on staffing by professionals with appropriate training, expertise and positive attitudes toward inclusion.Bespoke in-house provision in two schools was supported by access to mental health workers, a specialist counselor, speech and language therapist and educational psychologist, as well as the school SEND team.Space was re-designated for a sensory room and a therapy room, in addition to small group teaching facilities.Educators assigned to these centers needed to demonstrate high-quality pastoral and teaching skills to support engagement.Resources were committed to maintaining a stable team of Higher-Level Teaching Assistants (HLTAs) and key workers with specialist skills to work in school-based AP (6).Care was taken to select and then support staff to work with high needs pupils (7,22).Teachers new to the AP received additional inclusive education training before working with pupils in that setting.In promoting positive outcomes through internal AP, interviewees noted the importance of "trust between pupils and staff" (2), "a mindset shift in terms of delivery" (10).
What we don't want is to put a teacher down there who is going into battles.We need it to be a positive experience for everybody (7).
In more expansive approaches to inclusion, school leaders also stressed the importance of developing capacity across the wider school staff to support pupils with diverse needs.As one interviewee tersely observed, inclusion is, "not about developing a room for a child to sit in a punitive environment and staff it with £40,000" (11).Conceptualizing behavior in terms of unmet need refocused leader efforts beyond individualized student support for problem behaviors to enhancing school capacity to support inclusion.Priorities for professional learning included the development of trauma-informed practice, awareness of Adverse Childhood Experiences (ACEs), restorative practices, and confident use of a range of de-escalation strategies.Transitioning to a relational behavior management policy required targeted staff development and changes to schools' behavior policy.For example, the "on call" policy (where members of the Senior Leadership Team (SLT) are called to extract pupils from lessons for temporary periods of isolation) without attendant teacher development or pupil support was increasingly viewed as perpetuating a cycle of challenging behavior.
All you're doing is abdicating responsibility to manage the needs in that classroom from the teacher by SLT coming down the corridor to take a child out.That child is struggling to develop a relationship with that member of staff so they're kicking off, and you're reinforcing that and undermining the system.You've got to take time to upskill staff, not take the problem away.( 11)

Revised Curriculum
Beyond efforts to re-culture and foster relational trust (between leaders and teachers, and between teachers and pupils, families and carers), the emphasis on inclusion encouraged four schools to review their curriculum offer to "fit the curriculum to the needs of the child, not the child to the curriculum" (4).While continuing to provide the statutory broad and balanced curriculum, schools exercised local discretion in meeting the perceived needs of individuals and groups of pupils considered most at risk of disengagement and exclusion.A CEO noted, "this temptation to chase a narrow academic agenda is a risky one if schools genuinely want to engage all learners" (18).Two schools removed a requirement for all pupils to choose an English Baccalaureate (EBacc) subject to provide "more appropriate" pathways for pupils disaffected with standard curricula (7, 4).The EBacc is a set of subjects at GCSE: English language and literature, mathematics, the sciences, geography or history, and a language.
We are not chasing Progress 8 scores.This is about engaging children who need something different, who could potentially be on the cusp of a permanent exclusion.( 8) Schools expanding curriculum choices re-developed on-site facilities to accommodate horticulture, hospitality, hair and beauty, sports coaching, and ASDAN (Award Scheme Development and Accreditation Network) programmes, (which aim to develop skills for learning, work and life, with practical and accessible pedagogy).Student support, nurture and vocational provision were purposely co-located in three schools alongside senior leadership team offices on the ground floor (1,5,8) because "We wanted people to see that these students belong in the School" (1).The physical location of student support was cited as signaling the value attached to inclusion and parity of esteem between flexible and academic pathways.For some schools the COP project was part of a wider commitment to foster a sense of belonging and community, a social covenant to "keep our sheep in our fold" (15).

RQ2 Metrics and Marginality
Success criteria for the COP project were stipulated by the Department for Education Opportunity Area team and the Local Authority as a condition of funding.Pupil mobility metrics were the primary measure i.e., reductions in the number of permanent and fixed-term exclusions, moves to other schools and moves to Elective Home Education.At school-level some participants were keen to assert a more sophisticated understanding arguing that the drive for inclusion was inevitably multi-faceted, and required motivation beyond cost reduction.

Multifaceted Nature of Inclusion
The social complexity of working with high level needs in an area of acute deprivation might include a wider range of indicators such as the number of children on education, health and care plans (EHCPs), number of children supported through early help assessments, number of designated "children in need," and the number of cases escalated through the children's social care system.A child in need is defined under the Children Act 1989 as a child who is unlikely to reach or maintain a satisfactory level of health or development, or their health or development will be significantly impaired without the provision of children's social care services.In addition, school leaders acknowledged the potential for topline exclusion rates to obscure and misrepresent the level of disengagement among pupils.
There are so many different ways to skin a cat, other ways to change the dynamic of your school without necessarily permanently excluding a child.We could look at the number of part-time timetables, at alternative use of register codes, at the flow of children out of the school in other ways, at the numbers of pupils with below 60% attendance, or numbers of in-year admissions that go to appeal.( 16) Interviewees noted that self-evaluation might usefully consider not just the extent of pupil mobility, but also which practices are associated with successful reintegration in specific cases.For example, one headteacher noted that topline project data did not afford, "some acknowledgement, or way of factoring in schools taking on more challenging students from elsewhere and what they do with those" ( 22).The production of pupil case studies was valued as yielding important qualitative insights but was regarded as too demanding in terms of teacher time (18,24).Consequently, the project yielded limited teacher-generated information on which components of targeted interventions produced positive outcomes in particular settings.

Valuing Inclusion
There was a perception in some quarters that the COP project was primarily oriented to cost reduction for the Local Authority, and that meeting high needs would require significant levels of sustained additional support.Progress toward the espoused goals of the COP project were reportedly undermined by funding cuts in other areas.There was a perception among school leaders that finite funds were shuffled between multiple streams rather than increased overall in a coherent and targeted strategy.Planning was impaired by insecurity regarding the sustainability of funds and contested accountability measures.
There's a disconnect between professionals' understanding of inclusion, meaning not practising exclusion and other people's perception of inclusion as working with family support programmes, the EHCPs in the system, a diverse curriculum offer.Sometimes people come at things from entirely different directions.( 16) The metrics they're measuring are different.So, they're measuring from reducing high needs spend, the number of children who are in the Pupil Referral Unit, funding different places in different schools.The project's been talking about the need for schools to get support for Education, Health and Care Plans then just when we can submit a lot of applications, the funding is reduced.That's a hard pill to swallow.( 7)

RQ3 from Inter-Local Competition to Collective Responsibility
The COP project was initiated to re-build collective responsibility.Leaders' engagement with the project was inevitably influenced by a history of inter-local competition in a crowded and turbulent market for secondary school places, set against a prevailing discourse of high stakes accountability.Fostering a culture of inclusion required school leaders to look beyond short-term institutional advantage.The following section reports the context of institutional competition, the impact of areawide data sharing, and the prospects for sustainable change beyond the project lifespan.

Fragmentation and Competition
In the period preceding the COP project a degree of "ill-feeling" (16), "animosity" (1) and "tension" (6) had developed between schools in relation to in-year admissions.Schools with lower enrollments (below their anticipated Published Admissions Number, PAN) felt under pressure to accept pupils permanently excluded from other schools in the town.One headteacher reflected, "these children were someone else's problem, just cast aside" (1).Another commented, "people are very protective of their own academies and Trusts" (6).Town level deliberation was impeded by regressive inter-local competition and a lack of collective responsibility, which led to the IYFA process "breaking down and being taken over the Local Authority" (22).Meetings between secondary headteachers before 2019 were described as "very limited and sporadic at best, almost non-existent" (5).Intermittent or nonattendance at headteacher meetings was described as "taking the ball home and not playing, just not valuing the process" (7).

Transparency and Openness
Greater transparency was cited an important catalyst of change.Official records of pupil movement between schools were shared to cast light on the extent to which each school was a sender (excluding pupils) or receiver (accepting pupils excluded from elsewhere, supporting re-integration).Data sharing was described by one CEO as "the bow on the icebreaker" (24).Another suggested, "You can't hide away.If somebody is not in the mix around their contribution to the project, everybody can see it."(16).
However, moving from a history of institutional guardedness was an evident challenge.The extent to which leaders experienced discomfort in the new context of openness was influenced by their professional capital (signaled in published school performance ratings), the strength of community relations and their positional power e.g., the relative size and economic standing of their Multi-Academy Trust.Initial engagement by veteran senior leaders was defensive around both the scale of challenge and the looked-for pace of change.High rates of fixed-term and permanent exclusion before 2019 were justified in terms of efforts to improve behavior in response to poor inspection outcomes and the threat of intervention action, including re-brokering.Incoming headteachers and CEOs noted a need to first ensure the learning environment was "under control and safe" (10).

Sustainability
While pupil movement between schools fell during the lifespan of the project, interviewees expressed concern about sustainability.Three participants suggested that pupil moves may rise again on conclusion of the COP project (1,7,15).Commenting on improvement in the in-year Fair Access process one headteacher suggested, "it is only working better because we've been told we've got to take children" (7).A minority of headteachers (and no CEOs) suggested that the project was being pursued for instrumental gain, rather than an enduring commitment to inclusion.
People are putting strategies in place because there's money and there's a limit on three permanent exclusions a year.Once the Continuum of Provision project disappears, and the money starts significantly reducing, I can see a situation where some schools will say, we'll do four or five exclusions, because I can afford to pay the price.( 7) There are some schools and Trusts that will try and keep to the three because that's the right thing to do.There are others who will think there's no speed camera, and that opens the door again to potential inequalities in the system.( 15) Over-subscribed schools and larger Trusts with greater reserves were better placed to set longerterm priorities.Most schools had limited capacity to sustain the high staff to pupil ratio in recently established in-school inclusion centers.Some headteachers, not all, anticipated an increase in support needs, particularly around mental health issues, following the pandemic (4, 7, 8).
Frustration was expressed at delays in accessing funding for additional support.The process undertaken to obtain resources for secondary pupils was deemed too late (i.e., would benefit from early identification in primary school), too lengthy and too costly in terms of pupil progress and school resources (2,22).
The length of the process to get an EHCP can be 20 weeks.Well, that's half a school year.If you've got a student coming in presenting extreme behaviours from day one, that's a long road to go down.(22) In a competitive admissions market, headteachers of schools below their Published Admission Number (PAN) continued to voice concern about in-year transfers that might require substantial levels of additional support.The resource attached to in-year admissions was regarded as insufficient to support high needs.Reservations were expressed about moves falling short of exclusion that suggested "gaming" the system (i.e., manipulation of official records of exclusion).
IYFA provides an extra nominal £5k.Often these are pupils who have been permanently excluded from another school.It feels like a lot of repair work.That restorative work will happen in our inclusion center.If you're getting a lot of pupils coming in through that route, that can overload the system.We need to think about that risk.(10) A student who has lots of issues in one school, potentially getting close to permanent exclusion could suddenly put a transfer request in.We have to pick up some students with quite serious issues that are challenging and labor intensive.( 22 The school is not excluding anyone, but we keep getting children from that school who are presenting with very significant behavioral issues from day one.( 7)

Limitations
Before considering the key learning and implications of this study, it is important to acknowledge some limitations regarding the data and approach.As all eight secondary academies in the town participated in the programme there is no comparison or control group within the local authority.Baseline data for pupil mobility in 2018 were compared with annual rates before and after programme implementation.Exclusion rates were compared with a comparator Local Authority selected for size, demographic and socio-economic profile and a neighboring Authority with cross-border pupil mobility.While pupil exclusion rates fell in 2019/20, caution should be taken when comparing figures across years.Solo causal attribution is not possible given the social complexity of the problem, multiple OA projects operating in the same space, and possible alternative explanations which include recurring school closures and high levels of pupil absence (self-isolation) due to Covid-19 from March 2020.

Discussion and Conclusion
This research advances knowledge of school exclusion in a number of ways.It offers a critical investigation of area-level pupil moves in the context of devolved school management, specifically academization.Its highlights significant limitations of official records of in-year pupil moves and offers qualitative insights into the practices beneath topline data.While between-school segregation as an outcome of school choice has been a focus of much research, far less is known about the in-year allocation of places, or different types of within-school or flexi-school arrangements.Such practices are not choice-based in the accepted sense of parental choice and disproportionately affect the most vulnerable groups.
Future research needs to focus not just on formal exclusion from school, but on the rapidly developing models of within-school segregation-reintegration. Research-based evaluations of initiatives to reduce school exclusion need to address blind spots in official data arising from ill-defined inyear transfer categories and inconsistent reporting.It might then be possible to: (i) test anecdotal evidence of practices including the under-recording of the marginalization of female pupils through "soft" forms of exclusion; (ii) the dynamics beneath the post-pandemic rise in elective home education; (iii) the complexity of individual learner journeys; and (iv) young peoples' experience of withinand between-school transitions.
The study findings are consistent with earlier research that indicates that the creation of a quasimarket for school places has the potential to exacerbate existing inequities in high poverty, high needs settings (Burgess, Greaves, & Vignoles, 2020).While feted as promoting new freedoms for teachers and school leaders (DfE 2010), academization in England proceeded within a context of stringent centrally orchestrated accountability.High stakes accountability systems increase the attraction of controlling retributive approaches to behavior (e.g., zero tolerance, warm-strict protocols) that ratchet up sanctions culminating in removal for noncompliance (Barber & Sackville-Ford, 2021).In the period preceding the project, pressured school leaders in the case study town had elected to manage disruptive behavior by outsourcing provision for "troublesome" pupils off-site and off-register.The Continuum of Provision project was needed to tackle off-rolling and "gaming" within the in-year admissions process, and to re-build collective responsibility.A tendency to resort to disciplinary exclusion was enabled by "disintermediation" or the fragmentation of the local landscape of schooling in England with an attendant loss of place-based accountability (Courtney & McGinity, 2020;Greany, 2022).The COP project temporarily re-inserted a stronger intermediary role for the middle tier Local Authority, scaffolded by additional resource secured through designated Opportunity Area status.
Social mobility approaches advanced by the UK Government (such as the Opportunity Area programme and before that Education Action Zones and Educational Priority Areas) typically start from an assumption that inequality can be made equitable accordant with individual merit.Such logic distracts attention from persistent economic inequalities and regional disparities in educational opportunities (Thompson & Ivinson, 2020).From this perspective the goals of the OA are, at best, over ambitious.At worst, area-based initiatives can be viewed as "placebo" interventions (Giannakaki, McMillan, & Karamichas, 2018, p. 193) that reinforce social reproduction and competitive individualism.The prevailing logic underpinning intervention through the COP project was individualistic and corrective.Pupils are rewarded with a graduated return to mainstream by demonstrating compliant behavior and attitudes or are retained within a holding unit for further personalized ameliorative support.Short-term psycho-pedagogical support acts as a sticking plaster rather than serious remedy for multi-faceted complex needs that extend beyond school boundaries.As Pomeroy (2020) notes, "by mobilizing a discourse of 'failing schools' and providing seemingly straightforward interventions for reducing educational inequality, the UK government and Department for Education keep the focus squarely on school performance rather than poverty reduction" (p.341).
When self-managing schools in close proximity are placed in competition with one another school leaders face with difficult choices, often within ambitious timescales for turnaround.Critical scrutiny of localized in-year admissions brings the micro-political decision-making of headteachers and MAT leaders to the fore.Downey and Condron (2016) argue that schools are "refractors" of inequality and "decisions made by school leaders will influence the extent to which their schools will be 'neutral (no change to inequality), exacerbatory (makes inequality worse), or compensatory (reduces inequality)" (p.211).The headteachers in this study responded in ways that may be characterized as both welfarist (oriented toward inclusion) and "corporatized" (Courtney, 2017(Courtney, , p. 1054) (oriented to school and/or Multi-Academy Trust advantage).Schools' strategies to become more inclusive of pupils with challenging behavior ranged from the restrictive (individualizing) to the expansive (contextualizing).While all the schools made use of data from learning analytics, behavior referrals and diagnostic assessment, strategies to address need varied in terms of institutional preparedness to review systems, structures and policies around behavior and schools' capacity for change.
How schools engaged with the project was influenced by the history of relationships between schools and with the Local Authority, the prevailing context of accountability and the culture of individual schools.While formal exclusion rates fell sharply, the longevity of these gains is likely to reflect the extent to which leaders radically reconceptualized the problem of behavior and initiated a longer-term process of reconfiguring provision to promote inclusion at whole-school level.Analysis of school outcomes (in terms of available mobility data) in some settings encouraged a focus on school processes and investment in teacher development.Retributive approaches were supplemented by guidance approaches (e.g., restorative practice, relational approaches, pupil voice) drawing on humanistic principles to rebuild relational trust.Empathetic approaches sought to establish pedagogies of care (Riddle, Howell, Mcgregor, & Mills, 2021).In other settings, responses were more superficial, piecemeal, and suggested strategic participation or "lip service" to school-to-school cooperation.
While schools adopting a more expansive approach connected pupil engagement with curriculum design, equity-oriented curriculum making needs to challenge rather than reproduce hierarchies between academic and vocational knowledge.In examining moves toward a socially just curriculum, Mills et al. (2021) insist, "questions have to be asked regarding who gets what type of curricula, how categories of people are constructed in those curricula, and what input do marginalised groups (including students) have into what is enacted in the classroom" (p.6).High quality, high equity curricular provide rich and diverse opportunities for all.Access to less challenging curricula might potentially compound the marginalization of disadvantaged groups (Pendergast, Allen, McGregor, & Ronksley-Pavia, 2018).Equity-focused policy and research would benefit from widening the concept of school exclusion.Adopting Malcolm's (2018, p. 77) definition of exclusion as encompassing all forms of "education outside the mainstream norms" affords consideration of practices within mainstream schools that restrict access to experiences and outcomes.This would address the criticism leveled by Huilla (2020) that much research on disadvantaged schools is framed by regimes of accountability and consequently reinforces test-based accountability practices.
The challenge facing school leaders should not be underestimated.Equity-focused leaders must reconcile competing and seemingly incompatible demands (Dulude & Milley, 2021;Wang, 2018;Yurkofsky, 2022).The headteachers leading MAT schools in this study were caught between the devil and the deep blue sea.At a local level they were committed to swiftly reducing unacceptably high rates of school exclusion, at regional level they were contending with public disparagement of underperformance regarding attainment measures (Clifton, Round, & Raikes, 2016), at national level they remained subject to close external scrutiny and possible sanctions including re-brokering.Central policy and legal frameworks influence how the school system is configured locally and creates incentives for action (Crawford et al., 2022).This study of high needs support in one locality illustrates the negative consequences of a fragmented school landscape on the coordination of resource intensive services.The COP project was a centrally managed, locally delivered initiative to rebuild place-based collaboration -or "coopetition" (Muijs & Rumyantseva, 2014) -among eight autonomous neighboring schools in a one small Northern town.The threats to sustainability noted by the school leaders suggest the persistence of competitive values in a school-led system where the costs of inclusion exceed the penalties for exclusion.In a competitive schools market compounded by contracting public resource there are few incentives to gain a reputation for taking or retaining pupils with additional support needs (National Audit Office, 2019).Local demands for area-based inclusion were counterbalanced by scrutiny of institutional performance from external agencies (Regional Schools Commissioner, school inspectorate Ofsted, Department for Education).Competitive comparison is a disincentive to collaboration that drives inter-local guardedness rather than the openness associated with area-level change.In the complex and politicized terrain of school improvement there remain significant drivers of tactical engagement.Within ambitious reform projects, as Sam (2021) notes, "Effective leadership and ethical leadership are not necessarily the same" (p.304).
Overall, this study of school exclusion and high needs support in one locality exposes the unintended consequences of dismantling local systems of accountability in education.The study highlights the dangers of "disintermediation" (Greany, 2022;Lubieski, 2014) for community outcomes, equity and collective capacity.The putative development of a schools-led collaborative system requires co-production of aims and processes, and deep and sustained engagement over time.To be judged effective in the COP project leaders needed to demonstrate compliance with objectified measures under the auspices of the OA programme and its local Steering Board.At the design stage, individual school leaders, teachers, parents/ carers and support services were not involved in deliberation on the success criteria.This is in contrast to Hargreaves (2016) model of collective autonomy characterized by "constant communication and circulation of ideas in a coherent system where there is collective responsibility to achieve a common vision of student learning, development and success" (p.131).Similarly, Gunter and Courtney (2021) argue that educational leadership that is educative is "communal and shared " (p.194).They go on to assert, "The forms of leadership required to exclude children because they do not fit the school brand, to close down a school because of how the market is working, or to academize are not educational leadership that is actually educative " (p.196).This study demonstrates how leadership committed to inclusion is vulnerable to co-option and the consequent adoption of piecemeal strategies to wicked problems.This is frequently evident in the hopeful application of imported pedagogical strategies to remediate entrenched inequalities (Allen, Evans, & White, 2021;Hayes et al., 2017).High needs support for the most vulnerable young people in publicly funded schools demands more than a "pick and mix" approach (Hopkins & Craig, 2017).Finely grained portraits of the micro-politics at play in local school systems have value in appraising the outcomes of fast reform and the equity costs of disintermediation.Rebuilding mutual responsibility in a context of structural autonomy requires a sustained focus on equity in admissions (especially in-year transfers) and equity in outcomes for marginalized young people.Within the complex social ecology of local school networks, space for maneuver remains.School leaders who made most progress in reducing pupil moves promoted whole-school approaches to inclusion and demonstrated public leadership in resisting institutional incentives toward exclusion.

Disclosure Statement
No potential conflict of interest was reported by the author(s).