Vulnerable children’s right to education, school exclusion, and pandemic law-making

ABSTRACT This article draws on the impact of the ongoing pandemic to highlight the failure of the English legal regime to adequately protect children’s right to education, particularly equal access to education by especially vulnerable children. Ifirst outline key domestic and international legislative provisions positioned as securing children’s and parents’ rights in this context. Prior to the pandemic, there was growing recognition of the current regime's failings regarding illegal exclusions from school, children missing from education, and the lack of inclusive education for children with special educational needs and disabilities (‘SEND’). The protection of children’s rights relied on the benevolent exercise of discretion and key decision-makers not exploiting limited oversight and scrutiny in order to meet results-driven accountability measures. Second, I critically analyse pandemic law-making and regulation, particularly in relation to the exclusion process, the legal duty to provide education in an online environment, the law on Education, Health, and Care Plans (‘EHCPs’), and the de-registration and fines for non-attendance. Third, I argue that the educational impact of the pandemic highlights the need for law reform, rather than merely revisions to statutory guidance and focus on best practice. Such reform may also trigger improvement via the ‘reflexive regulation’ of the education system.


Introduction
Prior to the first lockdown in March 2020, when challenged on the most recent exclusions statistics and the problems concerning the link between crime and school exclusions, Nick Gibb, MP, Minister for School Standards, rejected criticism that 'something was wrong with the system' given that school exclusions had increased by 70% since 2012, 1 replying that the permanent exclusion rate of 0.1% was 'extremely low, and [.] actually lower than it was in 2006-07ʹ 2 and that absence, not exclusion from school was 'the key factor' in relation to the connection to knife crime. 3 When further criticised on unexplained exits from school and asked by Mike Kane, MP, to commit to implementing all recommendations in the Timpson Review, the Minister simply noted that the Government took the issues raised in the review seriously and was continuing to work with Ofsted to tackle 'off-rolling'. 'Off-rolling' is the process by which schools instigate the removal of, and then remove children from their admission register when it is primarily in the school's 'best interests' to do so rather than the child's. The exchange on 'unexplained exits' that undermine the headline figures upon which the Government relies, highlights the extent to which the Government avoids looking beyond the overall rate and confronting the reality of the significant inconsistencies in exclusion rates between local authorities and between schools. Nick Gibb, MP, concluded as follows: We know that we have to give headteachers the tools to ensure that we have safe, calm environments in our schools. No headteacher excludes without giving the matter very careful consideration, with permanent exclusions used only as a very last resort. What is key is that exclusion from school must not mean exclusion from education, so timely access to high-quality alternative provision plays a critical role in improving excluded children's outcomes. Our objective is to improve the quality and capacity of alternative provision. 4 This statement embodies the Government's preference for protecting school autonomy regarding student discipline and behaviour and the assumption that exclusion of itself is not a harm to be avoided. For some children, exclusion may be beneficial, whether because of the move to highquality alternative provision or otherwise. But there is no data to support the Minister's assumption that there is no need to focus on exclusion itself. There is a critical need to understand exclusion within its broader education and social policy context, which does not mean seeing exclusion as a gateway to alternative provision but as a site of the expression of that broader context. This further explains the importance of an intra-UK comparison of the contextualisation of school exclusion, although my principal focus in this article remains on the English position.
In what follows, I draw on the impact of the ongoing pandemic to highlight the failure of the English legal regime to adequately protect children's right to education, particularly equal access to education by especially vulnerable children. First, I outline the key legislative provisions, which are positioned as securing children's and parents' rights in this context. Prior to the pandemic, there was a growing recognition of the failings of the current regime in terms of illegal exclusions from school and the lack of inclusive education for children with special educational needs and disabilities. The protection of children's rights relied on the benevolent exercise of discretion and key decisionmakers not exploiting limited oversight and scrutiny in order to meet results-driven accountability measures. Second, I critically analyse pandemic law-making and regulation, particularly in relation to the exclusion process, the legal duty to provide education in an online environment, the law on Education, Health, and Care Plans (EHCPs), and the de-registration and fines for nonattendance. Third, I argue that the educational impact of the pandemic highlights the need for law reform, rather than merely revisions to statutory guidance and focus on best practice. Such reform may also trigger improvement via the 'reflexive regulation' 5 of the educational system.

The legal framework
In this section, I outline the current legal regime governing children's exclusion from school and highlight its key pre-pandemic deficiencies. The legal framework is sparse and discretionary. The governing statutory provision on exclusion is section 51A(10) of the Education Act 2002 in respect of England (section 52(10) adopting identical language in respect of Wales). It provides that those in charge of schools or alternative providers 'may exclude a pupil . . . for a fixed period or permanently' and defines 'exclude' as meaning 'exclude on disciplinary grounds'. 6 This legislative provision is supplemented by Regulations, but it is the statutory guidance that is relied on by key decisionmakers, particularly schools.
Such reliance on statutory guidance is technically to assist understanding of the legal framework but, to the extent that the guidance provides detail that cannot be assumed to be implicit within the language of the legislation, the guidance itself effectively comprises the legal framework. I suggest that evidence to justify such a conclusion lies in the 2015 reform to the statutory guidance that contemplated exclusion for other than 'disciplinary' reasons, namely 'where a pupil's behaviour mean[t] allowing [them] to remain in school would be detrimental to the education or welfare of the pupil or others in the school', 7 and was withdrawn only after legal challenge was threatened. 8 The current, 2017 statutory guidance sets out two conditions to specify when a head teacher should, as a matter of last resort, 9 permanently exclude a pupil. The decision 'should only be taken', firstly, 'in response to a serious breach or persistent breaches of the school's behaviour policy' and, secondly, 'where allowing the pupil to remain in school would seriously harm the education or welfare of the pupil or others in the school'. 10 Moreover, exclusion is unjustifiable if it discriminates against the child within the terms of the Equality Act 2010.
This exclusion framework applies within a broader context that seeks to give effect to the child's right to education and specific rights that protect children with disabilities. The right to education is incorporated into domestic law through Article 2 of Protocol 1 of the European Convention on Human Rights ('A2P1ʹ), which provides as follows: No person shall be denied the right to education. In the exercise of any functions which it assumes in relation to education and to teaching, the State shall respect the right of parents to ensure such education and teaching in conformity with their religious and philosophical convictions.
To justify omitting examination of head teachers' powers to exclude from his review of exclusion, Edward Timpson reasoned: It is the right of every head teacher to enable their staff to teach in a calm and safe school, just as it is the right of every child to benefit from a high-quality education that supports them to fulfil their potential. 11 It is clear, however, that this cannot be a reference to any A2P1 right to education. In practice, A2P1 provides only weak protection, as evidenced by three leading House of Lords/Supreme Court decisions on exclusion, namely Ali v Headteachers and Governors of Lord Grey School, 12 R (JR17) (Northern Ireland), 13 and A v Essex County Council. 14 In Ali, a boy (13-14 years old at the time) was subject to successive fixed term exclusions from school over the course of three months that were agreed to be unlawful as a matter of domestic law because the procedures set out in the governing statute and regulations were not followed. When the maximum period for fixed term exclusion was reached, the school did not exclude the child permanently but continued to wait for the outcome of criminal proceedings. In the House of Lords, it was agreed that A2P1 did not guarantee compliance with domestic law, such that exclusion that is unlawful under domestic law, may not -and was held not to in Ali -infringe a child's A2P1 right. 15 When the boy was subsequently offered a place at a pupil referral unit ('PRU'), which his parents did not accept, that was held by a majority not to amount to a denial of the right to education since that meant that 'the necessary minimum of education was available'. 16 In JR17, the Supreme Court was concerned with a case in which a Year 12 boy had been subject to a series of fixed term exclusions ('suspensions' in Northern Ireland) whilst the school investigated a complaint about him made by a girl pupil. As in Ali, the fixed term exclusions failed to comply with the requirements of the governing legal framework. The boy complained that the exclusions amounted to a denial of the A2P1 right to education, and the Court unanimously concluded that they did not, a conclusion that was upheld by the ECtHR. 17 It was not material that the standard or quality of the education provided whilst excluded may have been low; the pupil was offered access to those materials/arrangements that were offered to all pupils subject to fixedterm exclusions. 18 In A v Essex, a boy (12-13 years old at the time; 21 years old at the time of the Supreme Court hearing) had been out of school for 18 months. His behaviour was extremely challenging; he had special educational needs, was severely autistic and had a severe communication disorder, suffered from epilepsy, and had various other conditions. He had been attending a school for children with severe learning difficulties. The school asked his parents to remove him because it was concerned that his behaviour posed a risk to himself and others. The boy was out of education for 18 months because of both the lack of a suitable place at a residential school and the local authority's inadequate resources to undertake the requisite medical testing. The boy claimed that this infringed his right to education as guaranteed by A2P1.
The lower courts held that the boy's claim should be struck out because it had no realistic prospect of success. The boy made two arguments, namely that (i) A2P1 imposed a minimum, positive obligation on local authorities to provide a child (including him) with an effective education, taking into account any special needs, and regardless of any resource implications; and (ii), in the alternative, A2P1 entitled him to be provided with such resources as were available during the 18 months, even though they were inadequate to meet his needs. The Supreme Court dismissed the boy's appeal. A majority of three dismissed argument (i). A different majority of three held that argument (ii) might be able to be established on the facts, yet the case did not so proceed because a majority of four held that there was no basis for extending the one-year time limit applicable to the boy's ECHR claim. In the majority on this point, Lord Clarke reasons that the A2P1 right to education requires 'effective access to such educational facilities as the State provides for such pupils', such that a child 'was only denied effective access if he was deprived of the very essence of the right'. 19 This highlights the pragmatic, fact-specific examination of the minimal requirements for educational provision sufficient to meet a child's right to education. 20 In this way, the rights-based approach of A2P1 in practice at least currently requires no more for children than the duties imposed on local authorities under domestic law, particularly those set out in section 14 of the Education Act 1996, which has been interpreted as requiring local authorities to reasonably seek to 21 ensure there are 'sufficient' schools in their area to give children the 'opportunity of appropriate education', and in section 19(1) of the same Act, which requires local authorities to make alternative provision available.
Drawing on more recent A2P1 case law centred on discrimination under Article 14 of the European Convention on Human Rights in relation to the financial conditions for access to higher education, Harris suggests that ' . . . one nevertheless detects signs of an increasing judicial willingness to raise the bar for the state' 22 in terms of what is required as a minimum under A2P1. I would remain more cautious, however, particularly in relation to exclusion. Firstly, it is notable that the existence of other international expressions of children's right to education, such as Articles 28 and 29 of the United Nations' Convention on the Rights of the Child, 23 have not been relied upon to raise the threshold of minimum education to be provided to children so as not to unjustifiably violate their right to education. Secondly, the right to education, pre-pandemic, has not had any measurable impact on reducing the range of ways in which children 'drop out' from school and become disengaged from the education system beyond legal fixed-term or permanent exclusion, which includes illegal exclusion (that does not follow the legal processes, as evidenced in Ali and JR17), together with . . . being "internally excluded" from mainstream regular classes while remaining in the school; undergoing a "managed move" to another school or "alternative provision"; otherwise "missing education" by not being on a school roll and not being educated other than at school, including not receiving full-time provision when able to do so 24 ; and being in full-time education but disengaged from the education offered regardless of the precise cause. 25 26 Thirdly, and relatedly, the lived experience of the legal framework remains unchanged, as highlighted by Oxfordshire Safeguarding Children Board's Child Safeguarding Practice Review into how Jacob was exploited, leading to his death at 16 years of age. During almost two years prior to his death in April 2019, Jacob was not on any admissions roll and did not attend any education provider, with professional practice being 'impacted by ineffective systems which resulted in much too slow a response to Jacob's need for an education provision'. 27 Four educational settings were asked to take Jacob on roll but refused mainly because of his perceived behaviours and risks to other students. No referral was made to the Education and Skills Funding Agency so that an Academy admission direction could be made, and no progress was made by the local authority's In-Year Fair Access Panel and Children Missing Education teams; the oversight provided by the local authority's Virtual School and Independent Reviewing Officer Service resulted in no significant improvements. Children's Social Care raised no formal disputes. The review summarised the impact as follows: The on-going lack of daily education for Jacob will have played a significant role in not identifying levels of risk and leaving him vulnerable to further episodes or incidents of harm. The possible impact on Jacob is one where he may have felt disheartened as he repeatedly asked for an education provision and was not provided with the opportunity for this as he should have been. Not being on roll or in any education provision left Jacob highly vulnerable to coercion and being criminally exploited as he had nothing to fill the majority of his day or week with and left him with much time on his hands in the community. Not being in a learning environment denied Jacob the many daily experiences, and possibly a relationship with a teacher, which would have given him opportunities and hopes for a different day to day life and for a better future. There would have been a need for creative thinking and some flexibility in supporting Jacob in an education setting but he also wanted this opportunity, often asking to be in a school setting, talking of missing his education and sharing his dreams of what he wanted to become later in life. 28 There is no reference to A2P1 in the OSCB review. The only discussion of the child's rights centres on his voice and his wish to return home from care to his mother. As a child missing education, Jacob was clearly failed by the pre-pandemic legal regime and practice, including beliefs held about the value of making a referral to the ESFA and the accountability/implementation gap between local authorities and Academies. The success of the discretionary legal framework relies on the benevolent exercise of discretion; key decision-makers are incentivised to be intransigent within a discretionary regime premised on cooperation 29 through Academies working within In-Year Fair Access Panels and with local authorities. Jacob's effective exclusion from education is a product of the current legal framework, and there is no sense in which available rights arguments have been posited as demanding a better outcome.

Pandemic regulation -and law-making and school exclusions
Pre-pandemic, recent government reviews had shown significant concerns over both exclusion from school and the treatment of children and young people with disabilities and special educational needs within the education system more generally. The Timpson Review identified too much inconsistency in exclusion in practice and noted there '[was] more we can do to ensure that every exclusion is lawful, reasonable, and fair; and that permanent exclusion is always a last resort, used only where nothing else will do'. 30 Summarising its analysis of the current educational experience of children and young people with disabilities and special educational needs, the House of Commons' Education Committee reasoned that 'the [Children and Families Act 2014] reforms have resulted in confusion and at times unlawful practice, bureaucratic nightmares, buck-passing and a lack of accountability, strained resources and adversarial experiences, and ultimately dashed the hopes of many'. 31 Moreover, there was already significant awareness of the 'concerningly disparate' effects of exclusion that disproportionately impact on children with disabilities and SEN. 32 The pandemic impacted children's education within this broader context.
Temporary changes were introduced to the process on exclusion from school, as well as to attendance, the nature of the legal duty to provide education in an online environment, and the law on Education, Health, and Care Plans (EHCPs). The Government had previously announced that 'clearer, more consistent' guidance on exclusions and behaviour would be published by summer 2020, 33 but this has been halted by the pandemic.

The exclusion process
Changes to the exclusion process centred on practicalities such as permitting governing boards and independent review panels to meet via videoconferencing or telephone conferencing, and the applicable timescales for such meetings. 34 The Department for Education ('DfE') has also informally changed the language it employs, reverting to the term 'suspension' for fixed-term exclusion and 'expulsion' for permanent exclusion. This change has not (yet) been implemented in the statutory guidance or legislation, but in all DfE website references to exclusion and has been justified as 'preventing confusion and conflation between the two terms'. 35 Whilst well-intended, such change in the language is unhelpful. First, there is no suggestion that the legislative language will change, which both supports the conclusion that the statutory guidance and other non-legal guidance is unjustifiably treated as if it were legislation and invites greater confusion of the relationship between the legislation and broader policy context. Second, it is unclear why it is considered important to signal a distinction between fixed term and permanent exclusion given that the two are necessarily related: both are governed by the same legal 'disciplinary' restriction and any excess to 45 days of exclusion in a school year is lawful only if formalised via permanent exclusion.
Importantly, no changes were made to the 'disciplinary' ground for exclusion. Neither the legislation nor the statutory guidance require any more precise basis for the exclusion beyond that it is on 'disciplinary' grounds. The annual School Census does require the provision of a single 'reason for exclusion' and aggregate school responses are published. Twelve 'reasons' were in effect until the end of July 2020; the most frequently named 'reason' was 'persistent disruptive behaviour', which, in recent years, has been cited as the 'reason' for more than one-third of permanent exclusions. 36 As of August 2020, five additional 'reasons' were added whilst 'other' was removed; the five 'reasons' added were as follows: first, 'use or threat of use of an offensive weapon or prohibited item'; second, 'abuse against sexual orientation and gender identity'; third, 'abuse relating to disability'; fourth, 'inappropriate use of social media or online technology'; and, fifth, 'wilful and repeated transgression of protective measures in place to protect public health'. These changes are not framed as temporary.
Little critical attention has been paid to the changes to the 'reasons' for exclusion. Whilst the giving of 'reasons' is not part of the legislation or required by the statutory guidance, I suggest that the role played by 'reasons' within the School Census creates a site of soft power that influences schools' and head teachers' approaches to exclusion and their understanding of when exclusion is justified. As a consequence, I here highlight a number of aspects of these changes.
The final new 'reason' relates to the Covid-19 pandemic and the need for the fourth new 'reason' has understandably been enhanced by the increase of home learning. It remains to be seen whether schools adopt a consistent, proportionate approach to pupils' breaching of pandemic-related protective measures. It might be anticipated that its application will raise similar concerns to 'persistent disruptive behaviour', the flexibility with which it can be interpreted underpinning its role as the most commonly cited 'reason' since schools are able to interpret it to suit their own 'best interests'. Would coughing in the direction of a fellow pupil more than once be sufficient? What of choosing to walk the wrong way through part of the school's one-way system in order to take a shortcut?
The removal of 'other' as a category of reason suggests that the DfE is of the view that it is not justifiable to exclude a child from school for any reason except those on the list, an approach that conflicts with the only constraint on exclusion being that it must be on 'disciplinary grounds'.
The other three new 'reasons' are not Covid-related. The first new reason, centred on an offensive weapon, plainly makes sense in light of increased awareness of and responsiveness to the issues surrounding knife crime and the importance of accurate data, 37 which would not be possible were relevant exclusions coded as 'other'. The discrimination-related 'reasons' build on the pre-existing 'reason' centred on 'racist abuse'.
Whilst there is a larger body of understanding upon which head teachers should be able to determine whether a child abused another on the basis of their disability, it is arguably more contentious how one identifies 'abuse against . . . gender identity' given the ongoing debate over the relationship between transgender rights, gender critical feminism, and freedom of expression. That the correct balance to be struck is not yet settled is evident from recent and ongoing court challenges, such as Harry Miller's actions against the Humberside Police and the College of Policing, 38 and Oxfordshire County Council withdrawal of its Trans Inclusion Toolkit, which stated that transgender children should be allowed to use the school toilets, changing rooms, and dorms of the gender with which they identified. 39 The 'reason' centred on gender identity abuse implies that a head teacher would have been justified to permanently exclude the 13-year-old girl who challenged Oxfordshire CC's Trans Inclusion Toolkit, yet the council's subsequent withdrawal of the toolkit indicates that any child excluded for such 'reason' may well be able to successfully judicially review their exclusion. That such exclusion would be unlawful highlights the difficulty in policing the line between freedom of expression and a 'disciplinary' offence, and that the new 'reasons' risk suggesting that exclusion is justified for lawful behaviour protected by the Article 10 ECHR right to freedom of expression. Were this change introduced outside of the pandemic context, it may well have received greater critical attention.

The legal duty to provide education in an online environment
Changes not directly focused on exclusion may also have a significant impact on exclusion, particularly if one understands exclusion to include the multiplicity of ways in which a child can be in fulltime education but disengaged from such education. The pandemic has expanded involuntary disengagement brought about through circumstances outside of the child's or the family's control such as the digital divide and the need for vulnerable learners to stay home due to their or other family members' health needs with no adequate in-home teaching provision. The necessities of the national health response to the pandemic emphasised facets of exclusion that had been previously neglected, particularly by governments and policy-makers.
During the early parts of the pandemic, the digital divide meant that just under half of secondary school pupils and over 58% of primary school pupils had no access to online education. 40 Estimates vary, but it is clear that a very large number (estimated at some 1.78 million pupils 41 in the early stages and at some three-quarters of a million 42 by Autumn 2020) missed school work due to lack of either an appropriate device or an internet connection. The pandemic has made more explicit the connection between disadvantage and exclusion.
In September 2020, the Government brought in a new legal duty under the Coronavirus Act 2020 to provide online education to pupils at home on the same terms as if they were in the classroom. 43 The extent to which that improves the education provided to children is an open question. Where schools simply lack the resources to provide sufficient laptops and so forth to enable children to adequately engage with online learning, the legal requirement to do so does not cause them to exist. Making more specific the duties under which schools operate does not make it more likely that schools will be able to fulfil those duties. Moreover, the pragmatic, resource-bound A2P1 jurisprudence discussed above, implies that a rights-based perspective would not make it more likely that children in deprived circumstances would receive more online lessons, formal learning, or better technology at home.
It is an open question whether an argument framed in terms of Articles 28 and 29 of the United Nations' Convention on the Rights of the Child ('CRC') might be more successful. Article 28 addresses access to education and Article 29 the goals of education, yet Article 4 recognises that limited resources may justifiably limit the implementation of economic, social, and cultural rights set out in the CRC. The Office of the Children's Commissioner for England relies on Articles 28 and 29, together with other CRC rights, to argue that children have 'fundamental rights . . . to be in school, and to receive the wider benefits this provides'. 44 The CRC is of persuasive value only in English courts. Further, it is unclear whether a Court would rely on the CRC to suggest a different outcome to that indicated by A2P1.
I suggest a clear possibility for doing so lies in the Article 28(1) reference to achieving to the right to education 'on the basis of equal opportunity'. Whereas A2P1 has been interpreted as requiring no more than, as set out in section 14 of the Education Act 1996, that local authorities reasonably seek to ensure there are 'sufficient' schools in their area to give children the 'opportunity of appropriate education', it may be arguable that equal opportunity requires more than a local comparison of opportunity. This would enable argument that the relative denial of access to online, formal education, whether as a result of insufficient teaching staff to provide online lessons or inadequate technology for children to use at home, violates children's Article 28 right. It is not a foregone conclusion, however, that a court would both interpret Article 28 more expansively than A2P1 and rely on Article 28 to develop A2P1 in the context of the pandemic.

The law on education, health, and care plans
Schedule 17.5(6) of the Coronavirus Act 2020 introduced temporary changes to the obligations imposed in terms of implementation of the terms of EHCPs, namely: first, that local authorities and health commissioning bodies must only use their 'reasonable endeavours' to discharge any duty to secure special educational and/or health care provision in accordance with a child's EHCP in order for the duty under section 42 of the Children and Families Act 2014 to be treated as discharged; and, second, that public bodies need only comply as 'soon as reasonably practicable' with any timescales set out in regulations in relation to the implementation of EHCPs if the reason for delay relates to the pandemic. 45 It is not surprising that there was some recognition of the practical limits of securing education in the pandemic. Article 23 of the CRC recognises 'available resources' limits on the disabled child's right to effective access to education. The pragmatic 'reasonable endeavours' limit, however, is problematic for its uncertainty in practice. It also goes further in relaxing the absolute duty owed by the relevant public bodies than was necessary. It has never been sufficient for the local authority to argue that the overall objective of the EHCP is being achieved such that it is not obliged to provide the specific support set out in the plan. In R(N) vNorth Tyneside Council, 46 Sedley LJ reasoned There is no best endeavours defence in the legislation. If the situation changes there is machinery for revising the statement, but while it stands it is the duty of the LEA to implement it. In a margin of intractable cases there may be reasons why a court would not make a mandatory order, or more probably would briefly defer or qualify its operation. 47 'Reasonable endeavours' plainly demands less than 'best endeavours' and it is not clear that there is justification for this additional concession other than to reduce the likelihood that parents challenge the lack of provision in the courts given the uncertain standard to which local authorities and health commissioning bodies are being held. Whilst the relaxation was temporary, it arguably serves as a key indicator of the extent to which the government prioritises the interests and rights of vulnerable children when more restrictive approaches were available. This attitude may be said to underpin the seemingly-intractable over-representation of children with special educational needs amongst those excluded from education. 48 It is common knowledge that children with EHCPs are amongst the most vulnerable and most likely to be excluded from school. In 2018-19, the permanent exclusion rate for children with SEN with an EHCP was 0.15 and 0.32 for children with SEN without an EHCP, compared to 0.06 for children without SEN. 49 In such a context, it is particularly disappointing that, even if some relaxation were considered necessary, a 'best endeavours' approach was not preferred.

Attendance
As of June 2020, three months into the pandemic, only about 18% of children with an education, health and care plan or a social worker attended an education setting. 50 Whilst this relates to attending school, it is clear that for many vulnerable pupils, especially those with EHCPs, attendance is critical to accessing and not being effectively excluded from education.
The government issued a number of notices 51 under Schedule 17 of the Coronavirus Act 2020, which disapplied section 444(1) and 444(1A) of the Education Act 1996, such that parents did not commit an offence if their children did not attend regularly at school. The period from September to December 2020 was not included within any issued notices such that parents committed an offence and were liable to penalty notices and prosecution in respect of unauthorised absences under s444A of the 1996 Act. Prior to the first notice coming into operation on 1 April 2020, the government 'set out an expectation' to local authorities not to penalise parents whose children were eligible to attend school. 52 In answer to a petition, the DfE asserted 53 that it deemed that absence from school related to the coronavirus pandemic would prevent a school from being able to lawfully delete a child's name from the school admission register under Regulation 8 of the Education (Pupil Registration) (England) Regulations 2006. Presumably pandemic-related absences not directly attributable to sickness are intended to qualify as 'any unavoidable cause' within the terms of Regulation 8(1)(h).
Local approaches to the implementation of s444A of the Education Act 1996, which permits authorised officers of the local authority, head teachers, and the police to issue penalty notices to parents in respect of their children's unauthorised absence from school, have emphasised the possibility of new forms of exclusion. In a November 2020 debate in the House of Commons, Wes Streeting, MP highlighted that one of his constituents, a mother of a terminally ill three-year-old was forced to deregister her older daughter from school to avoid being charged weekly non-attendance fines. 54 Rather than being used to support and give effect to the daughter's right to education, the fine served as means to exclude her from accessing that education. This is a legal problem. Wes Streeting called for the Government to be clearer in its guidance about the approach to be taken if there are vulnerable family members at home such that a child returning to school might present a risk. But such additional guidance should not be necessary. As noted above, a child may not be removed from a school register unless the school lacks 'reasonable grounds' to believe the pupil is unable to attend because of sickness or 'any unavoidable cause' and both the local authority and the school have failed, after reasonable enquiry, to ascertain the pupil's whereabouts.
In the particular case of Wes Streeting's constituent, the school clearly knew that the pupil was at home and why. It was arguably a wilful misinterpretation of the 2006 Regulations and unlawful 'off-rolling' to require the mother to withdraw her child from school in such circumstances. The fact that it was not highlighted in the parliamentary debate as unlawful is indicative of the lack of understanding of the legal framework. Further, I suggest this example exemplifies the tensions in the relationship between the legislation, regulations, and statutory and non-statutory guidance: as schools are accustomed to relying on guidance as if it were legislation, insufficient attention is paid to the wording of the actual statute or regulation in question. This is deeply problematic given that non-statutory guidance and policy, such as the School Census, discussed above, is changed without the same oversight and scrutiny as applies in the context of legislative and regulatory revisions and reforms. The response to the constituent mother's older daughter's circumstances should have been for her school to support her being on-roll and educated at home, with the school fulfiling its duty to provide remote education under the Coronavirus Act 2020, as discussed above.

The case for law reform
In the conclusion to its 2019 review of the implementation of the Children and Families Act 2014 as it relates to children with special educational needs and disabilities, the House of Commons' Education Committee recommended that . . . when the Government makes changes to address these challenges, it should avoid the temptation to address the problems within the system by weakening or watering down duties or making fundamental changes to the law. 55 Whilst there may be good reason to reach such a conclusion regarding the treatment of SEND outside of the exclusion context, I suggest that it is clear that children's right to education will not be sufficiently protected in relation to exclusion without law reform. Responding to the 2018-19 data on permanent exclusion from school, Anne Longfield, former Children's Commissioner for England, commented: The majority of schools do excellent work supporting children and just 10 per cent of schools have been responsible for nearly 90% of exclusions. Exclusions should always be a last resort and the Government must give schools more support to keep children in school wherever possible, as well as doing more to identify the small number of schools who use exclusions to manipulate exam league tables. 56 I argue that Longfield's suggested improvements are unable to provide the necessary solution to unjustified permanent exclusion for the reason that underpins the statistics she cites. The broader discretionary legal regime within which children are excluded from school led to 10% of schools in 2018-19 being responsible for nearly 90% of exclusions. Even if we identify the schools that are free riders within the discretionary exclusion framework, how can such actions be sufficiently disincentivised without law reform?
The Timpson Review notes four fundamental drivers for current practice: the differences in school leadership teams; the insufficient support for schools to manage disruptive behaviour; the lack of any incentives to be more inclusive at potentially increased expense and negative performance; and the lack of safeguards to protect against schools not following the formal exclusion process. 57 The pandemic has enhanced the scope for these drivers to lead to negative outcomes for children via formal and other forms of exclusion from school. The additional pandemic-related 'reasons' for exclusion risk inconsistently and unjustifiably increasing permanent exclusion rates since there is not yet any detailed guidance on matters such as what qualifies as 'inappropriate' use of social media and online technology or which 'transgression[s]' of pandemic-related protective measures qualify and the extent of repetition that qualifies as justifying exclusion.
One obvious response would be to update the statutory guidance, as is currently in progress and was proposed in the Timpson Review, 58 albeit not in relation to this particular issue. This might be said to also represent the straightforward solution for the difficulty that arose in relation to attendance and the application of the 2006 Regulations to circumstances where the pupil herself was not ill. By contrast, I suggest that the better solution is in fact a legal one. I contend that the discretion built into the 'disciplinary grounds' definition of exclusion in the Education Act 2002, discussed above, should be supplemented by additional legislative detail or new regulations, rather than merely revisions to statutory and non-statutory guidance and focus on best practice. This would provide greater clarity to key decision-makers and prevent undue weight from being placed on guidance that could unjustifiably depart from statutory requirements, such as is risked in relation to the new gender identity 'reason' for exclusion. Further, a legal response would also ensure sufficient oversight over the reform process. The importance of this oversight is demonstrated by the changes introduced during the pandemic without such scrutiny, including the new 'reasons' for exclusion, which may in their uncertainty risk unlawful, unjustifiable exclusions being wrongly treated as lawful. Would these reforms have been introduced if they were considered part of the legislation rather than simply a matter of policy?
As the apparent uncertainty over how to approach the issue of non-attendance for pandemicrelated reasons other than illness demonstrates, school decision-makers rely on guidance to the exclusion of the authoritative legislation and regulations. The education system can be understood as governed by 'reflexive regulation' 59 whereby, over time, the 'self-referential capacity' of the system can serve to underpin internal conversations and collaborative learning through best practices. 60 Operating within a self-regulating, autonomous system may be justifiable where best practice progressively winnows the range of potential outcomes for children such that consistent, child-centred decisions result in exclusion and inclusion where it is in the child's 'best interests'. Reflexive regulation posits that the cause of regulatory failure in the past is attributed to a failure to appreciate the limited role that law can play in bringing about change directly in other social sub-systems because of the limited openness of these other sub-systems to external normative interventions. 61 There is a further, particular difficulty in the operation of the education system. Decision-makers internal to the education system see themselves as implementing and giving effect to external legal obligations when they are in fact operating internally, drawing on self-referential understandings of how to balance the competing rights and interests of children at risk of exclusion and other parties.
Policy on the implementation of legislation and regulations is treated as if it comprises the entirety of those external legal sources, at least when it comes to application. This creates a particular difficulty for reflexive regulation. Law's role is to stimulate self-regulation and act as a catalyst for new approaches, yet how can it 'provoke . . . [this] reconfiguration of self-regulation' 62 if decisionmakers within the education system fail to have sufficient regard to the external law? I suggest that the misunderstanding within the education system of the nature of statutory and non-statutory guidance has arisen because the governing legislation both contains little detail and has not been reformed when critical changes have been introduced to its intended application in practice. Recent experience demonstrates that, whilst reform to policy or even statutory guidance can affect change within the closed system, it cannot stimulate self-regulation and act as a catalyst for new approaches in the same way as external legal reform. The need to focus on law reform is evidenced by the lack of action and absence of legal consequences for the LEA and schools refusing admission in Jacob's case, discussed above. As part of its recommendations, the OSCB's review . . . asks the Department of Education to provide statute and guidance to local areas and their communities on how to manage the Governance arrangements with academy run schools and local education departments who currently cannot be mandated to accept children on roll. 63 Legal reform to the admission direction process, including tight time limits for action by the Education Skills and Funding Agency, is capable of achieving much more for children excluded from the education process and subject to the vagaries of free-riding academy schools via In-Year Fair Access Panels than can be achieved by change to guidance alone. Where there are critical legal rights to be accommodated, the shift in approach may be best triggered by law reform that demands co-option into the education system's internal language and processes. This may enable greater integration of the demands of A2P1, as well as Articles 28 and 29 of the CRC. With the rights recognised, the self-referential system can then develop and cement a child-centred approach to those rights.

Conclusion
The pandemic has demonstrated the importance of adopting a broad understanding of exclusion from school. In a pandemic-driven, home-schooling context, reference to the formal exclusion process captures only a very partial view of the circumstances in which children are excluded from education. Being unable to access education through lack of online technology or lack of adequate support for one's special educational needs results in effective exclusion. With the pandemic bringing together social exclusion and exclusion from school, reform must focus on more than the formal exclusion process.
Yet, pandemic law-making also reveals the need not to overlook the detail of the formal process since it provides key sites of soft power for shaping and influencing head teachers' and schools' understanding of when exclusion is justified. This is especially evident in the potentially contentious changes to the 'reasons' for exclusion to be provided in the School Census. Further, pandemic lawmaking outside of the exclusion context highlights the extent to which exclusion as an outcome should not be considered in isolation from the regulation of access to educational resources, the duties that relate to implementing EHCPs, and attendance at school.
The ease with which the Government relaxed absolute duties owed to children with EHCPs beyond what was made strictly necessary by the pandemic context demonstrates the need to increase attention paid to children's rights, both in strengthening the formal legal implementation of A2P1 and CRC rights (including Articles 23, 28, and 29) and in integrating a truly child-centred approach to children's rights into the internal logic and language of the self-regulating education system. Whereas the development of best practices might be the appropriate response to difficulties encountered elsewhere in the education system, law reform may be the most effective means of stimulating the necessary shift in attitude and approach to exclusion where the over-representation of vulnerable children, such as those with special educational needs, remains seemingly intractable.
The rapidity with which legislative and regulatory changes were introduced to both relax and extend duties owed to children is a reminder of the legislature's capacity to respond as necessary to emergent concerns. Unjustifiable outcomes such as parents feeling forced to de-register their children for non-attendance because of misunderstanding and misapplication of the governing legal framework are alleged to be avoidable with improved guidance; I have suggested that improved drafting of both primary and secondary legislative provisions makes such avoidance of unjustifiable outcomes more likely. We risk failing to learn the lessons of the entrenched unjustifiable data on exclusion from school if we continue to rely exclusively on revisions to statutory and nonstatutory guidance to restrain free-riding head teachers' and schools' unjustifiable exercises of discretion at the expense of vulnerable children's right to education. The pandemic has further exposed that failings in the legal framework on exclusion may need not only policy but also technical legal solutions.