London Borough of Croydon v SK-A [2022] UKUT 106 (AAC)

IN THE UPPER TRIBUNAL
ADMINISTRATIVE APPEALS CHAMBER
Between
London Borough of CroydonAppellant
On appeal from the First-tier Tribunal (HESC Chamber)
- v –
Between
S. K-A.RespondentBefore: Upper Tribunal Judge WikeleyHearing Hearing date: 6 April 2022Decision date: 20 April 2022Appellant: Ms Amelia Walker of Counsel, instructed by the London Borough of CroydonRespondent: Ms Alice de Coverley of Counsel, instructed by the Coram Children’s Legal CentreDECISIONDate 2022-04-20

The legal issue raised by this appeal to the Upper Tribunal

[1]This appeal by the London Borough of Croydon (‘the local authority’ or ‘the LA’) against the decision of the First-tier Tribunal (‘the Tribunal’) concerns the proper interpretation and application of section 9 of the Education Act 1996. This section (as amended) provides as follows: Pupils to be educated in accordance with parents’ wishes 9. In exercising or performing all their respective powers and duties under the Education Acts, the Secretary of State and local authorities shall have regard to the general principle that pupils are to be educated in accordance with the wishes of their parents, so far as that is compatible with the provision of efficient instruction and training and the avoidance of unreasonable public expenditure.[2]It is not in dispute that section 9 also applies to the Tribunal insofar as it stands in the shoes of a local authority in deciding matters on appeal. It is also common ground that section 9 does not impose a duty to act in accordance with the parental preference, but rather to have regard to it.

The background to this case

[3]This case concerns a young boy who I shall simply call J, in order to protect his privacy and anonymity, as well as that of his family. At the time of the Tribunal hearing, J was aged 13. He has a diagnosis of severe bilateral spastic quadriplegia form of Cerebral Palsy. J is non-verbal, doubly incontinent, non-ambulant and a full-time wheelchair user, who requires hoisting for all transfers during the day. He has serious respiratory problems and profound and multiple learning difficulties. He is, inevitably, reliant on adults for all his personal care and mobility needs.[4]That summary is doubtless wholly inadequate in conveying the challenges faced both by J and by his mother on his behalf. I apologise in advance for the inordinate length of this decision, which may seem somewhat detached from the real life issues they face, but the appeal raises a wider point of legal principle that may affect other children with special educational needs.

The proceedings before the First-tier Tribunal

[5]In February 2020 the local authority decided to make and maintain an Education Health and Care (‘EHC’) Plan for J. By the time the matter came before the Tribunal, the local authority had accepted that J needed a 52-week per year residential placement. It proposed (in Section I of the EHC Plan) a placement for J at a school I will simply call ‘the Manor’, an independent special school. J’s mother felt that he needed to be at a different residential school (‘the Trust’), a non-maintained special school. In fact as it happened J had already been placed at the Trust for post-operative recuperation, but had then remained there in the light of the national lockdown caused by the Covid-19 pandemic.[6]J’s mother accordingly appealed to the Tribunal under section 51 of the Children and Families Act 2014, against the contents of the EHC Plan made by the local authority for her son. At the Tribunal hearing the live issues in dispute revolved around Section B (special educational needs), Section F (special educational provision) and, of course, Section I (educational placement). The Tribunal framed the issues raised by Section I as follows (the text has been suitably anonymised but the emphasis in paragraph 17 is in the original):16. … The LA accepts that the Trust is a suitable placement, but does not consider that J requires a waking day curriculum, and regards the Trust therefore as over-provision. Given the cost differential between the placements too, the LA says that placement at the Trust would not be compatible with the efficient use of its resources under s.39(4) Children and Families Act 2014. J’s mother says even if both schools are suitable, we should place J at the Trust notwithstanding the cost differential on the basis that it isn’t incompatible with the efficient use of resources, given the additional benefits the Trust could offer.17. As all children at the Manor receive a waking day curriculum in any event, in practice J will receive this provision on a 52-week residential placement, whatever we decide (though we must decide whether this is something that he reasonably requires as part of provision for Section F). The placement issues therefore are whether both schools are suitable for J, and if so whether it is incompatible with the efficient use of resources for him to remain at the Trust. We must also, if needs be, consider the application of s.9 Education Act 1996, which would require us to consider the wider costs to the public purse.[7]The Tribunal issued a careful and detailed decision on the appeal by J’s mother, running to a total of 38 pages and 152 paragraphs. It dealt with the issue of educational placement at paragraphs 97-129. In paragraphs 97 and 98 the Tribunal, having summarised the local authority’s objections, named the Trust (the preference of J’s mother) for the purposes of Section I. Paragraphs 99-104 and 105-109 of the decision reviewed what the Trust and the Manor schools respectively could each offer J. In its analysis at paragraphs 110-115, the Tribunal found that both placements were suitable in terms of meeting J’s special educational needs and provision. The annual costs differential was found to be £61,003, the Trust costing £268,393 p.a. and the Manor £207,390 (paragraphs 116-117). As such, the Tribunal found that placement at the Trust was an inefficient use of resources for the purposes of section 39(4) of the Children and Families Act 2014. Accordingly, it also concluded that the statutory presumption in favour of naming the Trust was displaced (paragraph 118).[8]The Tribunal then turned to consider the application of section 9 of the Education Act 1996. The Tribunal correctly recognised that “this principle is something to which we only need to have regard and it does not offer any form of presumption in favour of the parent’s position” (paragraph 119). However, the Tribunal went on to adopt the submission advanced on behalf of J’s mother that “we are required in having regard to the principle to take account of a wider range of public expenditure in considering the question of what is unreasonable, and to consider public expenditure more holistically” (referring to PD & AD v Stockton-on-Tees BC (SEN) [2019] UKUT 57 (AAC)). The Tribunal added that it “must also take into account a wider range of potential benefits to J, in terms of educational, social care and healthcare benefits” (citing K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) at paragraph 29).[9]The Tribunal went on to dismiss the argument put by J’s mother that there were several additional items of expenditure involved in moving J to the Manor, which would have had the effect of reducing the differential in costs between the two placements for the purposes of applying section 9 (paragraphs 121-125). However, the Tribunal then concluded as follows, again with the text appropriately anonymised (PEG means percutaneous endoscopic gastrostomy and NG means naso-gastric):126. It is the wider benefits to J from placement at the Trust, in particular healthcare benefits, that we consider to be substantial. Based on the evidence of the Trust’s head of therapy and J’s mother, which we had no reason to doubt, we find that J is likely to pull his naso-gastric tube out on a regular, at least weekly, basis. The Trust’s head of therapy indicated that this had happened 3 times in the week leading up to the hearing. This required it to be re-inserted by someone with a medical qualification. Although we accept that nursing support is available for J for a large part of the day at the Manor, the care is not 24-hour. The evidence before us was that J tended to pull the tube out at night, presumably in his sleep. There is some likelihood therefore that he will do this at night if not every week, then at least regularly, when nursing care is not available to him immediately at the Manor. As the Manor’s head teacher accepted, this would necessitate a trip to a local hospital in an ambulance.127. The LA downplayed the likelihood and significance of this in the hearing, and in their submissions. They noted that J might soon have a PEG tube rather than an NG tube. But this cannot be assumed, and he has not yet been assessed for this as J’s mother confirmed. We need to proceed on the basis of the child before us. It is also important to appreciate the significance of the removal of this tube. As J is at severe risk of aspiration, he cannot be given thin fluids like water orally. His NG tube therefore meets his hydration needs, as well as providing him with a range of medication, including any painkillers and medication presently taken to control gastroesophageal reflux (supplementary bundle page 79). He remains at risk both of aspiration, and pneumonia. Although the Manor’s head teacher is right in one sense when she said that the removal of such a tube was not an emergency situation in the same way that respiratory problems might be, he also could not, in our judgement, be left overnight and into the morning for this tube to be replaced if it was removed many hours before it could be replaced on site. He would lack medication and hydration. The risks to his health from this are significant. Should he need to be taken to hospital overnight, this is likely to be a distressing and uncomfortable experience for him. Evidence from J’s mother, and from the independent advocate, was that J does not travel well. We consider on balance that this is likely to occur regularly.128. We also take into account the social care benefits to J of the closer proximity of the Trust to his mother in south London than the Manor [situated north of London]. Although J’s mother would be able to stay at both schools overnight for free, she confirmed that driving to the Manor was much less of an option that she would therefore be more reliant on public transport for what was a longer, and more involved journey. J’s mother estimated that what was presently a 45-minute journey to the Trust would take 2-2.5 hours to the Manor. She said that it would also likely affect the ability of other members of the immediate family to visit, once they were allowed to do so following Covid-19 restrictions being lifted. We do not over-estimate this benefit – the Manor is still reasonably accessible from South London, and J’s mother can stay over. But we consider that there is some weight to be given to this factor and it is in J’s interest to see and have contact with his mother as much as possible.129. As we note above, in our view the health benefits of remaining in a residential placement at the Trust are considerable – in addition to the onsite 24-hour nursing team, the Trust has a significant additional medical team (bundle, pages 271-272) including consultants and registrars, to which the Trust’s head of therapy also referred; and the healthcare risks to him of moving to the Manor are also significant. We have to decide whether the additional expenditure involved in placing J at the Trust is compatible with avoiding unreasonable public expenditure, and whether, all things considered, having regard to the general principle, in this case we should decide to place J there. We have decided on balance that the healthcare benefits to J remaining at the Trust substantially outweigh the cost differential of the placement. The risk to J at the Manor is regular, and will cause him discomfort and inconvenience on a regular basis. Even 1 trip to a hospital every couple of months late at night would, in our view, be unacceptable for any child. Accordingly, we do not consider that the additional expenditure is an unreasonable public cost. On balance we also consider that taking all of the various factors into account, we should place J at the Trust. We order accordingly, and name the Trust in section I of his EHC Plan.[10]The Tribunal’s decision was therefore to allow the appeal. Its Order provided that the EHC Plan should be amended in various respects, including that the Trust should be named as the school placement in Section I.

The local authority’s application for permission to appeal

[11]The local authority applied to the Tribunal for permission to appeal to the Upper Tribunal under three broad headings, being(a) (A) errors of law (including error of law in application of section 9 of the Education Act 1996);(b) (B) inadequate and/or inconsistent reasoning; and(c) (C) an unreasonable decision.[12]On 2 July 2021 Deputy Chamber President Judge Tudur, in a comprehensive and fully reasoned ruling, refused permission to appeal on behalf of the Tribunal.[13]On 26 July 2021 the local authority applied direct to the Upper Tribunal for permission to appeal.[14]On 3 August 2021 Upper Tribunal (UT) Judge Hemingway granted permission to appeal, but confined exclusively to the section 9 point: What I shall call ground A1 is, essentially, a contention that the First-tier Tribunal (F-tT) erred through taking an impermissibly wide approach as to the range of factors it was able to consider when conducting the balancing exercise called for when applying section 9 of the Education Act 1996. The F-tT relied, at least in part, for its view upon what had been said in K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC). But what was said or what the F-tT thought had been said, in that case, arguably sits unhappily with what was subsequently stated to be the position in KE v Lancashire County Council(SEN) [2017] UKUT 468 (AAC). The ground, as set out in the passage from paragraph 12-19 of the above document [i.e. the application for permission to appeal], is arguable and I give permission with respect to it.[15]UT Judge Hemingway refused permission to appeal on the various other grounds, essentially on the basis that they were an attempt by the local authority to re-argue findings of fact properly made by the Tribunal.[16]The local authority subsequently applied for one of the remaining grounds of appeal to be reconsidered at an oral permission hearing. This ground (‘ground C’) was that the Tribunal had reached an erroneous decision on the weight to be attached to the parental preference against a costs differential of £61,003, and so had reached a decision which no tribunal reasonably directing itself could have reached. Following a hearing on 9 February 2022, and in a detailed ruling dated 21 February 2022, UT Judge Hemingway refused the local authority permission to appeal on ground C. This left the local authority’s appeal limited to the section 9 ground to go forward to the full appeal.

The Upper Tribunal oral hearing of the appeal

[17]Both parties have made comprehensive and helpful written submissions in accordance with UT Judge Hemingway’s case management directions. I held an oral hearing of the local authority’s appeal on 6 April 2022. This was a conventional face-to-face hearing so far as the representatives were concerned but was held in a hybrid format so J’s mother (the Respondent to the local authority’s appeal) could observe the proceedings by a video-link. The local authority was represented by Ms Amelia Walker of Counsel and J’s mother by Ms Lucy de Coverley of Counsel, both of whom appeared before the Tribunal below. I am grateful to them both for their helpful oral and written submissions.

An outline of the parties’ respective arguments

[18]Ms Walker’s primary submission on behalf of the local authority was that the Tribunal had erred in law in its application of section 9 of the Education Act 1996 in deciding that the healthcare and social benefits to J could outweigh the significant public expenditure differential between the two placements. In considering the advantages that are relevant to section 9 it was contended that the Tribunal should have considered only the educational advantages to J and not any wider advantages. Ms Walker argued this was evident from the wording of section 9 itself, which was framed in terms of parental wishes for their children’s education, and not any wider wishes around meeting their social or healthcare needs. This interpretation, she submitted, was supported by the decisions of the Court of Appeal in Oxfordshire County Council v GB [2001] EWCA Civ 1358 and W v Leeds City Council and SEND Tribunal [2005] EWCA Civ 988; [2005] ELR 617. The case law on what constitutes “unreasonable public expenditure” in the context of section 9 – e.g. Haining v Warrington Borough Council [2014] EWCA Civ 398; [2014] PTSR 811 – did not displace the proposition that the First-tier Tribunal’s function was to consider educational advantages set against public expenditure. Ms Walker argued the Tribunal’s reliance in the present appeal on the Upper Tribunal’s decision in K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC); [2011] ELR 165 was misplaced as HH Judge Pearl’s approach in that case had subsequently been disapproved by UT Judge SM Lane, especially in KE v Lancashire County Council(SEN) [2017] UKUT 468 (AAC); [2018] ELR 196.[19]Ms de Coverley, for J’s mother, submitted that section 9 required tribunals to take a holistic approach to determining placement for the purposes of Section I. In deciding whether a placement is incompatible with the duty to avoid unreasonable public expenditure, a tribunal had to take into account a diverse range of factors as part of a balancing exercise. The weight to be attributed to those factors was a matter for the specialist First-tier Tribunal. Neither statute nor case law confined that consideration to the educational advantages of any given proposed placement. This Tribunal’s approach was consistent with the Court of Appeal’s decision in Haining v Warrington Borough Council [2014] EWCA Civ 398; [2014] PTSR 811 and the Upper Tribunal’s decision in K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC); [2011] ELR 165. The Tribunal had identified additional advantages for J in a placement at the Trust in terms of his health, safety and wellbeing – see paragraphs 126-129 and paragraph 9 above – that were sufficiently beneficial such that the extra public expenditure involved was not found to be incompatible with the duty to avoid unreasonable public expenditure.

A review of the case law

[20]I was treated by both counsel to a veritable Cook’s tour of the case law on the proper interpretation of section 9 of the Education Act 1996. Naturally enough each advocate placed greater weight on certain of the authorities and within each authority counsel drew attention to different passages in support of their respective submissions. Broadly speaking, I adopt the approach to the case law urged by Ms de Coverley, and for the following reasons.[21]The proper construction of section 9 has been the subject of the case law of both the courts and now (at least since the implementation of the Tribunals, Courts and Enforcement Act 2007) the Upper Tribunal. So far as the case law is concerned, it seems to me the most convenient jumping off point is the decision of Mr Andrew Nicol QC, sitting as a Deputy Judge of the High Court (as he then was) in O v London Borough of Lewisham and SEND Tribunal [2007] EWHC 2130 (Admin), not least as it is arguably the most comprehensive consideration of section 9 in the case law of the courts. The question there was whether O, a boy with severe and complex difficulties, should go to a maintained day special secondary school (as the council contended) or a maintained residential special school (as O’s mother argued). The latter placement would have involved an extra cost of some £20,000 a year. If the cost of respite care provided by social services was deducted, on the basis it would no longer be needed, the differential shrunk to only about £3,500. However, the tribunal in that case had decided it could not deduct the cost of social care in considering “the efficient use of resources” for the purposes of paragraph 3 of Schedule 27 to the Education Act 1996 (now section 39 of the Children and Families Act 2014). On appeal, Mr David Wolfe, counsel for O’s mother, argued that the tribunal should have considered the effect of section 9 of the Education Act 1996. Allowing the mother’s appeal, the High Court held that “the term ‘public expenditure’ in section 9 is not confined to the expenditure of the Local Education Authority” (at paragraph 41). In a key passage, the Deputy Judge ruled as follows: 34 … The legislative changes that have already been made reflect the good sense of looking at the totality of the child's position under the alternatives being canvassed. Mr Wolfe referred to the decision of the Court of Appeal in R v Leeds City Council and Special Educational Needs and Disability Tribunal [2005] EWCA Civ 988 where Wall LJ said:
'[50] ….Because of his condition, C is manifestly a child with multiple needs who poses enormous challenges for those who have to attempt to care for him and provide him with education. Such a child's educational needs simply cannot be viewed in isolation; nor can his section 17 [a reference to c.17 of the Children Act 1989] needs; nor, for that matter, can his need for services provided by the Health Authority and CAMHS. A holistic approach is necessary, and with inter-agency co-operation, essential, particularly since two of the bodies with statutory responsibilities for (the LEA and SSD) are part of the same local authority. [51] At the same time, of course, the Tribunal is a creature of statute, and its powers are limited to the areas of responsibility given to it by the Education Act 1996 and the consequential regulations. Judge LJ has set out the relevant provisions in paragraphs 27-31 of his judgment and I will not repeat them. In a case, such as the present, the Tribunal, in my judgment, had to tread a delicate line between properly informing itself of the 'full picture' relating to C, and limiting its decision to a careful assessment of C's special educational needs within that full picture….' 35. In his judgment at para [43] Judge LJ recorded his full agreement with what Wall LJ said about the 'imperative, that so far as possible within the relative statutory frameworks, a holistic approach should be adopted by the various bodies with different responsibilities for C.'
Thomas LJ agreed with both judgments. 36. This guidance that LEAs and the Tribunal should inform themselves of the 'full picture' and adopt a 'holistic approach' accords with what I regard as the natural meaning of the term 'public expenditure' in s.9, namely that it is concerned with the impact of a parent's choice on the public purse generally and not exclusively with the cost to the local education authority. Of course, as the Court of Appeal warned in the Leeds City Council case, that approach must be constrained by the statutory framework within which LEAs and the Tribunal operate, but I accept Mr Wolfe's argument that for the Tribunal (or the LEA for that matter) to take account of savings to the local authority's social services budget does not require the Tribunal or the LEA to go beyond that legislative framework.[22]I turn now to the Upper Tribunal’s decision in K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC); [2011] ELR 165. Rather as in O v London Borough of Lewisham and SEND Tribunal [2007] EWHC 2130 (Admin), the dispute was over whether the child in question should attend a maintained special day school (as the council argued) or a residential school providing a waking day curriculum (the parents’ preference). The tribunal dismissed the parents’ appeal, who appealed further, in part on the basis that the tribunal had failed to consider the wider benefits that would attach to the parents’ choice of school. In that context, counsel for the local authority in that case specifically submitted that the O v London Borough of Lewisham case had been wrongly decided (see paragraph 18). HH Judge Pearl expressly rejected that submission in the following terms: 20. … Andrew Nicol QC, having looked with considerable care at all of the authorities, said that the LEAs and the tribunal “should inform themselves of the ‘full picture’ and adopt a ‘holistic approach’ …”. He considered that this was the natural meaning of the term “public expenditure” in section 9, namely it is concerned with the impact of a parent’s choice on the public purse generally and not exclusively with the cost to the local education authority. Andrew Nicol QC accepted the argument advanced before him by Mr Wolfe in that case that for the tribunal (or the LA) to take account of savings to the LA’s social services budget does not require the tribunal or the LA to go beyond the legislative framework. Having referred to the practical difficulties which might follow if the term “public expenditure” was given the meaning for which Mr Wolfe contended, the judge went on to specifically say as follows: “40. In my judgment these are not factors which alter the outcome of the exercise of statutory interpretation. Although the Tribunal’s focus is a child’s educational needs, his or her non-educational requirements are not concepts which are alien to it. The Tribunal may have to grapple with these for the purpose of deciding what should be included in Parts III and IV of the statement. And, as the Court of Appeal said in the Leeds City Council [[2005] EWCA Civ 988] case, as far as possible and consistently with the legislative framework it is desirable that a Tribunal inform itself of the ‘full picture’ and adopt a ‘holistic approach’. As for possible change in circumstances, this is inherent in the system already. Mr Wolfe gave the example of transport expenses. In deciding whether the parents’ choice would involve an inefficient use of resources, the Tribunal must take account of only the marginal cost of providing transport to the alternative schools – see, for instance, Oxfordshire County Council v GB [[2001] EWCA Civ 1358] at para [18]. The marginal cost would be nil if the LEA had to provide transport for other children in any case and if the car, bus or taxi could take an additional child without further expense. However, the position would be liable to change if those other children no longer needed the transportation. The possibility that the calculus of expenditure and savings may change is not a good reason for the Tribunal [not] doing the best it can to assess the position at the date of the hearing. 41. Accordingly, I respectfully differ from the conclusion which Sir Richard Tucker reached in the Somerset case [S v Somerset County Council [2002] EWHC 1808 (Admin)]. In my judgment, the term ‘public expenditure’ in s.9 is not confined to the expenditure of the Local Education Authority. Mr Wolfe referred me to the decision of the Court of Appeal on 12th February 2003 in Helen S v Somerset Council [2003] EWCA Civ 195. I observed that this was a refusal of permission to appeal and, as such, was not something which should ordinarily be cited. I had not appreciated at the time that it was a refusal of permission to appeal from Sir Richard Tucker’s decision in the same case. As such, it is right to note that Sedley LJ considered that it remained an ‘open question’ as to whether the term ‘public expenditure’ in s.9 was confined to the education budget of the LEA in question. Permission to appeal was refused because on the facts of that case, the Court of Appeal considered that, even if the Tribunal had taken into account the social services’ costs savings involved by choosing the parent’s preferred school, the decision would inexorably have been the same given the Tribunal’s views about the relative educational advantage of the two competing candidate schools.” 21. I have arrived at a similar view to the one arrived at by Andrew Nicol QC, and for the same reasons as those expressed by Andrew Nicol QC I respectfully differ from the narrower conclusion reached by Sir Richard Tucker in Somerset and the similar conclusion reached by Dyson J in C v SEN Tribunal (1997) ELR 390.[23]HH Judge Pearl went on to “agree absolutely” with what was said by Mr Andrew Nicol QC about W v Leeds City Council and SEND Tribunal [2005] EWCA Civ 988, namely that local authorities and tribunals should inform themselves as to the ‘full picture’ and adopt a ‘holistic approach’ to the term ‘public expenditure’ in section 9 (see paragraph 24). HH Judge Pearl then concluded as follows: 28. Of the many recent cases in this area, in addition to O, the other decision that seems to me to provide the most valuable assistance is that of Stadlen J in Hampshire County Council v R and Special Educational Needs and Disability Tribunal [2009] EWHC 626 (Admin); [2009] ELR 371. It is not necessary for the purposes of this case to set out the facts of that case in any detail. Suffice it to say that Stadlen J was concerned, as here, with the meaning of section 9. He said: “[35]… In my judgment the policy behind section 9 is that there are limiting factors on an unbridled regard being paid to the general principle that pupils are to be educated in accordance with the wishes of their parents, and those limiting factors are inserted by Parliament by reference to the knock-on effect that giving effect to those wishes might have generally. [36] It may be that having regard to the parents' parental wishes would be incompatible with the provision of efficient instruction and training to the parents' own child or children. It may be that it would be incompatible with the provision of efficient instruction and training for children with whom the parents' children would be educated if the wishes were to be given effect to. It may be that it would be incompatible with the provision of efficient instruction and training to other children who would not be educated together with the pupils of the parents in question, but who would be adversely affected, because the efficiency of their instruction and training would be incompatible with the parental wishes, in some indirect way, by reason of the arrangements that would need to be made to give effect to those wishes.” 29. In other words, one looks at the total picture. Applying this general principle, it is my view that a LA (and the tribunal on appeal) when conducting the balancing exercise, are obliged to take account of wider social and health benefits when deciding whether additional public expenditure is unreasonable. I do not consider that my approach is different from that as set out by the Upper Tribunal in Hampshire County Council v JP [2009] UKUT 239 which decided that the decision of the tribunal in that case was flawed in that it gave inadequate reasons for its conclusion that the child in that case required residential education. That case was not concerned with, and did not consider, section 9.[24]As regards this last passage, it will be recalled that the Tribunal in the present case specifically cited paragraph 29 of HH Judge Pearl’s decision in K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) in support of its approach. The proposition in paragraph 29 that “a LA (and the tribunal on appeal) when conducting the balancing exercise, are obliged to take account of wider social and health benefits when deciding whether additional public expenditure is unreasonable” self-evidently provides that support.[25]The next case to consider is Haining v Warrington BC [2014] EWCA Civ 398 (on appeal from the decision of UT Judge Williams in HS/1444/2013, [2013] UKUT 391 (AAC)). The council argued that the child should go to a maintained day special school while the parents wanted him to go to an independent residential special school. The council accepted that if its proposal was accepted it would also provide residential respite care. The Master of the Rolls helpfully identified the points raised by the appeal in paragraphs 2 and 7 of his judgment: 2. The question that lies at the heart of this appeal is how the words "public expenditure" should be interpreted. In relation to local authorities, do they mean expenditure incurred by local authorities in discharging their functions under the Education Acts as defined in section 573 of the 1996 Act ("education functions") (the narrow meaning); or do they mean expenditure incurred by any public authority as a result of the discharge by the local authority of the education functions (the wider meaning)? … 7. The issue that arises on this appeal is whether, in comparing the cost of placements at the two schools, Warrington (and on appeal the FTT and the UT) should have left out of account respite care and other costs that were to be met from public expenditure, and limited the comparison to the costs that were to be met from its education budget.[26]The Master of the Rolls reviewed the existing authorities at paragraphs 11-21 of his judgment, and in doing so expressly approved of, and adopted, the reasoning in O v London Borough of Lewishamand SEND Tribunal [2007] EWHC 2130 (Admin) (see paragraph 19). Decisions of the Upper Tribunal which had held that public expenditure in the context of section 9 included expenditure by public bodies other than the LEA were also noted (at paragraph 21) and by implication approved – including K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC). UT Judge Williams’s narrower reading of section 9 was held to involve an error of law, because “the correct meaning of the words ‘public expenditure’ in section 9 is expenditure incurred by a public body, as opposed to ‘private expenditure’ (ie expenditure incurred by a private body)” (paragraph 27). The Master of the Rolls (with whom Pitchford and Rafferty LLJ agreed) concluded as follows on the impact of section 9 in placement decisions: 40. … In this respect, I agree with the analysis of Stadlen J in Hampshire County Council v R and SENDIST [2009] EWHC 626 (Admin), [2009] ELR 371 at paras 59 and 60. In other words, the authority must ask itself the question whether naming the school preferred by the parent would involve incurring unreasonable public expenditure generally. In many cases, the only relevant public expenditure will be expenditure incurred by the local authority discharging its education functions. In such a case, it is difficult to see how the result of the section 324(4)/section 9 exercise can properly differ from the result of the para 3(3) exercise. But in a more complicated case involving, for example, the costs of respite care, the answer may be different. 41. This is a more complicated case, although all the relevant public expenditure will be incurred by the same local authority, namely Warrington. It is more complicated because substantial respite care fees are involved. For the reasons I have given, Judge Williams erred in holding that, for the purposes of section 9, the FTT was entitled to leave out of account "the respite and other costs that were met from public expenditure but were not met from the education budget of the Council" (para 80). If those costs were not left out of account, then the cost to the public purse of placing B at WHS (£61,238) was lower than the cost of placing him at GHS (£33,448). In these circumstances, it is impossible to say that, if Judge Williams had directed himself correctly, he would have reached the same conclusion as he did. In my view, this matter must be remitted to the FTT for reconsideration in the light of this judgment.[27]That consideration then takes us to the two decisions by UT Judge SM Lane upon which Ms Walker placed emphasis, with particular emphasis being reserved for the former decision: KE v Lancashire County Council(SEN) [2017] UKUT 468 (AAC); [2018] ELR 196 and PD & AD v Stockton-on-Tees BC (SEN) [2019] UKUT 57 (AAC); [2019] ELR 401.[28]In KE v Lancashire County Council(SEN) [2017] UKUT 468 (AAC) the council had proposed a maintained special school (cost £31,610 p.a.) as suitable for the appellant’s son while the parent argued for an independent special school (cost £102,572 p.a.). The differential was accordingly just under £71,000 a year. The tribunal agreed with the council that the parent’s choice of school need not be named because the cost of the boy’s attendance there was incompatible with the efficient use of resources for the purposes of section 39 of the Children and Families Act (CFA) 2014.However, as UT Judge Lane identified (at paragraph 9): The problem with the F-tT’s decision is that, although the F-tT dealt with section 38 [sic] of the CFA 2014, it did not direct its mind to section 9 of the Education Act 1996 (EA 1996), at all. This was a necessary step for the F-tT ... Its failure to do so was an error or law, but … I decline to set the decision aside. This is because no tribunal considering the issue properly could have come to any other decision.[29]UT Judge Lane confirmed that section 9 “has a role to play in determining the resolution of a conflict between a parent’s choice of a school for their child and an alternative school or institution proposed by the LA in Section I of an EHC Plan. So even though section 39(4) of the CFA 2014 appears to furnish a self-contained solution to that conflict, section 9 nevertheless requires the decision maker to have a further look at the dispute from the different viewpoint of whether public expenditure would be unreasonable” (paragraph 13).[30]She continued as follows (footnote omitted), before citing relevant passages from the Court of Appeal judgment in W v Leeds City Council and Special Educational Needs and Disability Tribunal [2005] EWCA Civ 988: 15. The test in section 9 is whether the parental choice would represent ‘unreasonable public expenditure’. This requires the tribunal to consider the impact of the parent’s choice of school on the public purse generally, and not just on the particular LA which has responsibility for the pupil. When weighing up the respective costs of the competing schools/ institution proposed by the parties, it is necessary to take a ‘holistic’ view of the particular pupil to get a full picture of his needs (O v London Borough of Lewisham [2007] EWHC 2130 [34] – [36] Deputy HC Judge Andrew Nichol QC. The tribunal is, however, ‘constrained by the statutory framework within which LAs and tribunals operate’: O v London Borough of Lewisham [2007] EWHC 2130 [34] – [36]; W v Leeds City Council and Special Educational Needs and Disability Tribunal [2005] EWCA Civ 988.”[31]UT Judge Lane further acknowledged that “the approach in O v London Borough of Lewisham and cases following it is binding following adoption in Haining v Warrington Borough Council [2014] EWCA Civ 398 [19][27]ff” (at paragraph 16). However, in a passage on which Ms Walker relied, UT Judge Lane made the following further observations:17. It is important to bear in mind, however, the context in which the case law has been developing in testing the elasticity of the holistic approach. The Leeds case, for example, did not involve section 9 at all. It mainly concerned the question of whether the special educational needs identified in Part 2 of the pupil’s Statement of Special Educational Needs were properly supported by provision made for those needs in Part 3 of the Statement. The specific issue that required attention was whether the pupil needed a waking day curriculum. The tribunal rejected this need. It supported that view by having regard to the provision the social services department would provide for the pupil under its duties under the Children Act 1989. The ‘holistic’ approach that the Court of Appeal approved in Leeds was the tribunal’s use of information from social services to help determine where educational and non-educational provision should end in that case.18. In both O v Lewisham and Haining the issue was whether the identifiable cost of respite care that the local authority would otherwise have had to spend if the pupil attended their preferred school could be set off against the cost of the parent’s preferred school in order to reduce the difference between the two. These were straightforward quantifiable calculations. In CM v London Borough of Bexley [2011] UKUT 215 (AAC), the question related to the impact on the public purse where there was an arrangement between two local authorities regarding attendance by a pupil from one at a school in the other area. This was, again, a straightforward quantification.19. The above cases can be contrasted with K v London Borough of Hillingdon (SEN) [2011] UKUT 71 [29]. The parents wished the child to attend a residential independent school providing a waking day curriculum whereas the LA considered a maintained special day school to be suitable. The question before Upper Tribunal Judge Pearl was whether the F-tT erred by taking the position that ‘social and health provision, however compelling they may be’ must be excluded when considering the educational advantages of a placement. The answer to that question must have been yes, insofar as the public purse had to be considered more generally than just the LA’s education budget. Had the costs saved by the LA and health authority been set off against the cost of the parent’s preferred school, the cost of the latter might not have been unreasonable.20. If that is all Judge Pearl meant to say, it is plainly correct. But Judge Pearl referred to the LA and tribunal (on appeal) taking ‘account of wider social and health benefits’ in conducting the balancing exercise regarding unreasonable public expenditure under section 9 [29], and a ‘broader calculus’ [33]. He confirmed his view that this broader calculus was required by reference to a Statement in the case given by the chief executive of IPSEA. This is an organisation offering free legal advice on special educational needs, including special educational needs litigation, to parents [32]. Judge Pearl does not explain the reason for admitting this Statement. The chief executive deposed: ‘For all children with a statement it is our experience that any attempt to silo their educational needs from their social care needs or medical needs often prove impossible. For example, to attempt to isolate when learning is an educational need and when learning is a social need is a false exercise…It is therefore almost impossible for this group of children [with special educational needs] to separate, predict and assess the benefits that arise directly only in relation to formal educational needs as opposed to care needs. In many cases it is a false exercise to attempt to do so as, like with younger ordinarily developing children, they need to learn continually whilst awake. What is different however is that in order to make progress this has to happen in a more planned, structured and formalised way. To consider the wider benefits of a particular school placement is therefore essential when considering the special educational needs of a child.’21. This evidence is problematic if only because it is not clear how these views fit into the statutory framework within which LAs and tribunals operate. It appears to be aimed at changing the whole perspective from which a tribunal is to consider a placement from its statutory basis of suitability to meet a pupil’s special educational needs (subject to unreasonable expenditure), to one which takes the need for waking day curriculum as a starting point. There is nothing to suggest that the balancing exercise in section 9 is involved. I doubt whether Judge Pearl would have intended to change the nature of the exercise under section 9 by the side wind of this evidence. If he did, I would be unable to agree with him.[32]The last case that merits consideration is PD & AD v Stockton-on-Tees BC (SEN) [2019] UKUT 57 (AAC); [2019] ELR 401, another decision by UT Judge Lane. This was also a case in which the First-tier Tribunal had considered section 39(4) of the Children and Families Act 2014 but had failed to consider the balancing exercise required by section 9 of the Education Act 1996. The relevant passage in UT Judge Lane’s decision reads as follows:47. The long and the short of this is that there is a two-stage process for deciding whether parental choice is to prevail. If a Tribunal comes to the conclusion that an exception in 39(4) so that the parental choice need not be named in the EHC plan, it must nevertheless go on to consider whether the result is the same under the test in section 9.48. It has long been decided in both the courts and Upper Tribunal that the tests under the old Schedule 27 paragraph 3(3) (which has now morphed into section 39(4)) and section 9 are different. It is not necessary to extend the length of this decision by a detailed exposition of the well-known case law since the facts of this appeal admit of only one answer. It is enough to say that section 39(4) requires the LA or the Tribunal standing in its shoes to look at the resources of the LA itself whereas section 9 requires the Tribunal to take a ‘holistic’ look and take a wide view of public expenditure rather than just the resources of the LA.

The Upper Tribunal’s analysis

[33]The local authority’s case is that the Tribunal erred in law in its application of section 9 of the Education Act 1996. In taking into account the advantages relevant to the section 9 balancing exercise, Ms Walker submits that the Tribunal should have confined itself to a consideration of the educational advantages of each potential placement rather than any wider advantages (e.g. by way of health care or social benefits). Ms de Coverley resists so narrow a reading of section 9.[34]I do not accept Ms Walker’s analysis for the following reasons.[35]First, Ms Walker’s opening gambit is to rely on the wording of section 9 itself, which she characterises as being framed around parental wishes for the education of their children, rather than any wider wishes around meeting their healthcare or social needs. It is certainly true that section 9 requires decision-makers to “have regard to the general principle that pupils are to be educated in accordance with the wishes of their parents, so far as that is compatible with the provision of efficient instruction and training and the avoidance of unreasonable public expenditure” (emphasis added). However, Ms Walker’s reading places too great an emphasis on the verb “educated” and insufficient weight on the following phrase “in accordance with the wishes of their parents”. As Ms de Coverley correctly noted, in deciding their preferences parents routinely consider wider factors than purely educational issues (e.g. a school’s geographical location and whether a sibling is at the same school). Nor is the local authority’s obligation to consider factors in relation to the exercise of its statutory functions under section 19 of the Children and Families Act 2014 confined to educational issues.[36]Second, Ms Walker points for support to the statements of Sedley LJ in Oxfordshire County Council v GB [2001] EWCA Civ 1358, and in particular the passage at paragraph 16 of that judgment: 16. In cases like the present, the parental preference for an independent school over an available state school, while perfectly reasonable, may have difficult cost implications for the LEA. In that event it is for the LEA, or on appeal the SENT, to decide whether those cost implications make the expenditure on the independent school unreasonable. This means striking a balance between(a) the educational advantages of the placement preferred by the parents and(b) the extra cost of it to the LEA as against what it will cost the LEA to place the child in the maintained school. In cases where the state system simply cannot provide for the child's needs, there will be no choice: the LEA must pay the cost. In cases where the choice is between two independent schools, it is accepted on all hands that the second criterion is simply the respective annual fees, whatever the comparative capital costs or other sources of income of the two establishments: for example, the one with lower fees may have private or charitable funding, but this will have no bearing on the quantum of public expenditure involved in a placement there…[37]Understandably, Ms Walker highlights Sedley LJ’s reference to “striking a balance between(a) the educational advantages of the placement preferred by the parents and(b) the extra cost of it to the LEA as against what it will cost the LEA to place the child in the maintained school” (emphasis added). Ms Walker likewise draws attention to Sedley LJ’s reference to the need for “a consideration of the burden which the respective placements will throw on the annual education budget when matched against their educational advantages and drawbacks for the child in question” (at paragraph 18, emphasis also added). However, the Court of Appeal’s reference to “educational advantages” (which, of course, is not the statutory terminology) must be read in context. Oxfordshire County Council v GB [2001] EWCA Civ 1358 was essentially a case about the accountancy method by which the cost of a placement at a LA school was properly calculated. The nature of the advantages of one placement over the other was not really a live issue as both schools were appropriate to the child’s needs. Accordingly, the case did not involve consideration of the nature of those factors which were relevant to determining whether the costs differential was itself unreasonable. It follows that the Court of Appeal’s decision should not be read as strictly confining the consideration of one placement’s advantages over another by reference solely to the educational benefits involved. Furthermore, as Deputy Chamber President Judge Tudur noted, when refusing permission to appeal on behalf of the Tribunal below, the overall legislative landscape is now very different (e.g. as a result of the Children Act 2004 and Children and Families Act 2014)[38]Third, Ms Walker submits that Haining v Warrington BC [2014] EWCA Civ 398 merely confirms that the Tribunal is entitled to take into account wider public expenditure (beyond that of the LA) in deciding what constitutes “unreasonable public expenditure” for the purposes of section 9. In particular, she contends that authority does not displace the proposition that the Tribunal’s role is to consider educational advantages set against public expenditure; nor does it establish a principle that the Tribunal is entitled to consider anything other than educational advantages in the section 9 balancing exercise. It is true that the precise question raised by the present appeal was not directly in issue in Haining v Warrington BC [2014] EWCA Civ 398. However, Ms Walker’s reading of the decision is too narrow. The Court of Appeal expressly confirmed the broader ‘holistic’ approach to the construction of section 9 as advanced by the Deputy High Court Judge in O v London Borough of Lewisham and SEND Tribunal [2007] EWHC 2130 (Admin). In addition, K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) was cited by the Court of Appeal with implicit approval (paragraph 21), albeit on an allied point.[39]Fourth, and last but by no means least, Ms Walker submits that, insofar as there is any conflict between the authorities, I should follow Judge Lane’s decision in KE v Lancashire County Council(SEN) [2017] UKUT 468 (AAC) in preference to HH Judge Pearl’s decision in K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC). There is, plainly, a degree of common ground between the two cases. Both authorities accept that the section 9 balancing exercise requires the Tribunal to adopt a holistic approach and to take a wide view of “public expenditure” and not one confined to the local authority’s resources, let alone just its educational budget (see the Hillingdon case at paragraphs 20 & 24, the Lancashire case at paragraphs 15-19 and the Stockton-on-Tees case at paragraph 48). It is also noteworthy that in the Lancashire case UT Judge Lane characterised the issue in the Hillingdon case as being “whether the F-tT erred by taking the position that ‘social and health provision, however compelling they may be’ must be excluded when considering the educational advantages of a placement” (at paragraph 19). She continued “The answer to that question must have been yes, insofar as the public purse had to be considered more generally than just the LA’s education budget”. To that extent the Lancashire case is not wholly supportive of Ms Walker’s position. Indeed, UT Judge Lane specifically found that “it is not necessary to come to a final conclusion on how broad the calculus under section 9 is in deciding this case” (paragraph 22).[40]However, there is also equally plainly a fault line between the two authorities to the extent that, as UT Judge Hemingway observed, they sit somewhat uneasily together. K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) holds that in conducting the balancing exercise decision-makers “are obliged to take account of wider social and health benefits when deciding whether additional public expenditure is unreasonable” (paragraph 29) and to consider “the broader calculus” (paragraph 33). The decision in KE v Lancashire CC (SEN) [2017] UKUT 468 (AAC) case questions and indeed throws doubt on this approach, at least to some extent (at paragraphs 20 and 21).[41]There are several reasons why I consider it appropriate to follow K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) over KE v Lancashire CC (SEN) [2017] UKUT 468 (AAC) case to the extent that they diverge. In no particular order, those reasons are as follows.[42]First, K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) is consistent with the strong emphasis in both O v London Borough of Lewisham and SEND Tribunal [2007] EWHC 2130 (Admin)and the Court of Appeal’s decision in Haining v Warrington BC [2014] EWCA Civ 398 on the need for a holistic approach to section 9 (albeit those authorities related to a separate matter of construction of section 9).[43]Second, UT Judge Lane’s observations in KE vLancashire CC (SEN) [2017] UKUT 468 (AAC) case at paragraphs 20 and 21, to the extent that they question the approach of K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) at paragraph 29, are strictly obiter (not necessary for the Upper Tribunal’s decision). UT Judge Lane had already held that the First-tier Tribunal in that case had erred by not having had regard to section 9 at all. Moreover, on the particular facts of that case, it was “inevitable that any reasonable tribunal properly directing itself to the law and facts” would have found the costs differential to be unreasonable (paragraph 25).[44]Third, UT Judge Lane brought into question the approach of K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) at both paragraphs 29 (the reference to the “wider social and health benefits”) and 33 (the reference to the “broader calculus”). However, the Judge’s reasoning for doing so is focussed exclusively on what she describes as the problematic admission in the Hillingdon LBC (SEN) case of evidence from the IPSEA chief executive. But HH Judge Pearl had reached his conclusion in paragraph 29 before there was any consideration in his decision of the IPSEA evidence. Instead, HH Judge Pearl had framed his finding in paragraph 29 against the backdrop of the developing jurisprudence, and in particular O v London Borough of Lewisham and Hampshire County Council v R and SENDIST [2009] EWHC 629 (Admin). UT Judge Lane had discussed the former case but not the latter so did not fully address HH Judge Pearl’s reasoning. I also do not accept Ms Walker’s submission that HH Judge Pearl had made an unjustifiable “leap” from his discussion of the previous case law to his conclusion at paragraph 29. Rather, his approach was both a natural development of and consistent with the holistic approach to section 9 more generally.[45]There is a further consideration on which I do not rely in reaching my decision but which provides ex post facto (after the event) support for the view I have already arrived at.

A further consideration: the status of Upper Tribunal reported decisions

[46]I have already given my reasons for following the decision of HH Judge Pearl in K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) in preference (insofar as there is any conflict) to the decision of UT Judge Lane in KE v Lancashire CC (SEN) [2017] UKUT 468 (AAC). There is, however, a further consideration that has come to light which reinforces this decision. I did not raise the matter with counsel at the oral hearing of the appeal as it only came to light later when I was in the process of writing up the decision which I had already reached for the reasons set out above.[47]The further consideration is that K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) is a reported decision of the Administrative Appeals Chamber of the Upper Tribunal (UTAAC) whereas KE v Lancashire CC (SEN) [2017] UKUT 468 (AAC) is not, strictly speaking, a reported decision. The joint agreed authorities bundle, helpfully provided by the local authority for the oral hearing, listed all the decisions by reference to their NCNs (and conveniently listed in date order). However, it was only when I visited the UTAAC decisions website in order to copy and paste a passage from the decision that I realised that the Hillingdon LBC (SEN) case is reported in the Administrative Appeal Chamber Reports (AACR) at [2011] AACR 31 whereas KE v Lancashire CC (SEN), although published on the UTAAC decisions website and so available on Bailii,is not reported in that series. A ‘reported decision’ is a term of art in this Chamber’s jurisprudence. A decision is only reported (in the AACR) if the decision and its reasoning commands the broad assent of the majority of the Chamber’s judges. Other decisions, which are accorded an NCN and available on the UTAAC decisions website but not included in the AACR, are perhaps best regarded as ‘published’ rather than as ‘reported’.[48]This difference involves an important distinction in the doctrine of precedent operated in this Chamber. The point is well illustrated by the decision of UT Judge Wright in ET v Secretary of State for Work and Pensions [SSWP] (UC) [2021] UKUT 41 (AAC). In deciding that appeal (which was about a claim to a social security benefit) UT Judge Wright was faced with a conflict between two previous Upper Tribunal decisions (the point at issue in those decisions is not material for present purposes). The first case was a decision by UT Deputy Judge White, reported in the AACR, namely NS v SSWP (ESA) [2014] UKUT 115 (AAC); [2014] AACR 33. This was therefore a reported decision. The decision in conflict, which was not so reported, was MW v SSWP [2015] UKUT 665 (AAC) (coincidentally a decision by UT Judge Lane). In resolving the conflict, UT Judge Wright held that MW v SSWP was wrongly decided for two reasons. The second reason was that the decision in MW v SSWP was inconsistent with prior Court of Appeal authority in Charlton v SSWP [2009] EWCA Civ 42 on the point of social security law in question. But the other reason was as follows: 25. The first reason is that NS is a reported decision of the Upper Tribunal. The rules of precedent of this chamber of the Upper Tribunal provide that the starting point should be that I follow the decision in NS: see R(I)12/75 at paragraphs [17] and [21]-[22] and Bury MBC v CD (HB) [2011] UKUT 43 (AAC) at paragraph [7]. Put another way, and adopting the language of Upper Tribunal Judge Mesher in paragraph [7] of DM v LB Lewisham and SSWP (HB) [2013] UKUT 026 (AAC) (not doubted or overturned on this point on the subsequent appeal to the Court of Appeal in Mahmoudi v London Borough of Lewisham and the Secretary of State for Work and Pensions [2014] EWCA Civ 284; [2014] AACR 14):
“….[NS]…must have been regarded as rightly decided by at least a majority of the [UTAAC judges] at the time. For that reason and because of the desirability of certainty about the legal position, under the authority of decision R(I) 12/75 an individual judge of the Upper Tribunal should not depart from the legal principles for which [NS] stands unless satisfied that to do so would perpetuate error.”
For the reasons given below I do not consider that following the part of NS that endorsed IJ would be to perpetuate any error, even though this part of the decision in NS did not grapple with Charlton. 26. Further, there is no later decision that expressly declines to follow NS on the basis it was wrongly decided on the “IJ” point. The main, if not sole, contrary decision is MW, which was decided on 1 December 2015 and so after NS had been decided and reported. As set out above, the reasoning in MW is similar to (though more emphatic than) the arguments the Secretary of State initially made against this appeal being allowed on the second ground of appeal, but the Secretary of State has now abandoned that as a line of argument. MW was decided at the end of 2015 and thus well after NS had been decided and reported, but MW makes no reference to NS and so does not say why it did not follow a reported decision of this Chamber. I note too that other subsequent decisions have been decided on a basis more consistent with the IJ/NS line of authority than with MW: see, for example, the ‘by reason of’ analysis in paragraph [10] of JT v SSWP (ESA) [2018] UKUT 124 (AAC).[49]I simply observe that UT Judge Wright’s decision in ET v SSWP (UC) [2021] UKUT 41 (AAC) has itself been selected for reporting as [2021] AACR 6.[50]Accordingly, the reported status of K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) means at the very least that it carries more weight than KE v Lancashire CC (SEN) [2017] UKUT 468 (AAC). However, as explained above, I had independently arrived at the conclusion that insofar as the decisions conflict I should in any event follow the Hillingdon LBC (SEN) decision.

Drawing the threads together

[51]I conclude that the Tribunal did not err in law in taking into account, when considering the section 9 balancing exercise in relation to the prospective placements for J, the “wider social and health benefits when deciding whether additional public expenditure is unreasonable”, as HH Judge Pearl put it in K v Hillingdon LBC (SEN) [2011] UKUT 71 (AAC) at paragraph 29. I have reached this conclusion on the basis of the proper construction of section 9 of the Education Act 1996 in the light of the case law. Both Ms Walker and Ms Coverley also advanced a number of practical and policy arguments to support their respective positions. However, these considerations need not be explored, and so make an already over-long decision even longer, given my conclusion on the interpretation of section 9 in the context of the case law. I simply make two further observations in this regard.[52]The first observation is that there is an obvious conceptual symmetry in considering any wider health care and social advantages attaching to the parental preference under section 9 while at the same time having regard to public expenditure in a broad sense (so including e.g. health care and social care costs and not just educational expenditure by a council). The whole point of section 9 is that it involves a balancing exercise that accordingly requires a cost/benefit analysis of the relevant factors. A balancing exercise by definition necessarily involves weighing up the advantages that accrue on one side of the equation against the public expenditure involved on the other side. If that is to be meaningful, the same potential constellation of factors has to be considered on both sides of the equation. In sum, to use a rather strained metaphor, Ms Walker’s approach would require one to weigh educational advantages (apples) against educational, healthcare and social costs (apples, pears and oranges). I consider Ms de Coverley’s analysis to be more consistent with the holistic approach required by the case law – so the educational, healthcare and social advantages (apples, pears and oranges) are likewise weighed against the educational, healthcare and social costs (apples, pears and oranges) on the other side of the scales.[53]The second observation is that the weight to be attached to the respective benefits and costs under section 9 is axiomatically a matter for the expert evaluation and judgment of the specialist First-tier Tribunal. Furthermore, the outcome of this process of assessment is in effect advisory rather than determinative. As noted at the outset, section 9 does not impose a duty to act in accordance with the parental preference, but rather to have regard to it.[54]In the present case, in a factual context in which both placements would cost the local authority in excess of £200,000 a year, and the differential between them was in the order of £70,000, it might be expected that quite exceptional circumstances would be needed to justify the significant extra public expenditure involved in a placement at the Trust. However, the Tribunal’s approach discloses no error of law. It correctly directed itself as to the limited and non-binding nature of the principle of parental preference. It made factual findings about J’s tendency frequently to pull out his naso-gastric tube at night. Against that background, it made clear findings about the type and quality of health care available on site 24/7 at the Trust when compared with the need, when this occurred out of hours at the Manor, for an uncomfortable and distressing emergency journey by ambulance to a local hospital (see paragraphs 126 and 127). These were quintessentially issues of fact for the specialist Tribunal to take a view on. I do not accept Ms Walker’s submission that the Tribunal focussed on risk rather than health care advantages and so blurred the line between the two. That is, in essence, a challenge to the Tribunal’s fact-finding, but the local authority failed to secure permission to appeal on those grounds. The Tribunal also properly had regard to the social advantages associated with placement at the Trust, while recognising these considerations were by no means as compelling as the health care advantages (paragraph 128). In its concluding passage the Tribunal both correctly framed the issue it had to determine and reached a decision on placement that was sustainable in the light of its carefully evidenced and detailed factual findings (at paragraph 129): We have to decide whether the additional expenditure involved in placing J at the Trust is compatible with avoiding unreasonable public expenditure, and whether, all things considered, having regard to the general principle, in this case we should decide to place J there. We have decided on balance that the healthcare benefits to J remaining at the Trust substantially outweigh the cost differential of the placement. The risk to J at the Manor is regular, and will cause him discomfort and inconvenience on a regular basis. Even 1 trip to a hospital every couple of months late at night would, in our view, be unacceptable for any child. Accordingly, we do not consider that the additional expenditure is an unreasonable public cost. On balance we also consider that taking all of the various factors into account, we should place J at the Trust. We order accordingly, and name the Trust in section I of his EHC Plan.[55]Thus, in the exceptional circumstances of this case, and having carried out the correct weighing exercise under section 9, I am satisfied that the Tribunal had identified what UT Judge Lane described as “sufficient ballast to justify the extra cost” (PD and AD v Stockton-on-Tees BC (SEN) [2019] UKUT 57 (AAC).

Conclusion

[56]I therefore conclude that the decision of the First-tier Tribunal does not involve any material error of law and its decision stands. The local authority’s appeal is accordingly dismissed. Nicholas Wikeley 57. Judge of the Upper Tribunal Authorised for issue on 20 April 2022