EH v Kent County Council

It is sensible to start with the decision in Oxfordshire. Giving the judgment of the Court, Sedley LJ identified the issue before the Court in paragraph 4 of his judgment.

“The problem posed by this appeal, ……… can be stated in a narrow and a broad form. Put narrowly, it is whether in making a comparison between two appropriate schools, one an independent specialist school, the other a mainstream LEA school with a specialist unit, the cost of the latter is to be taken as the global cost of LEA provision (either in total or for the school in question) divided by the relevant number of pupils, or simply the additional budgetary cost of placing the child there. In its broader form, it is whether the cost of placing a child in the state sector should be taken to be an individual fraction of the global cost of local state provision, or whether that provision is to be regarded as given and the relevant expenditure quantified as the additional amount which the placement will cost the LEA.”

The fees for the school preferred by the childs’ parents were £16,800 per annum. The Tribunal considered that the annual cost of placing the child at school L, the chosen by the Local Education Authority (“LEA”) would be as follows:

The Tribunal concluded that the annual cost differential of £2,651 was outweighed by the educational advantages of the school preferred by the parents and ordered the LEA to amend the statement to name that school. It was submitted on behalf of the LEA that the Tribunal had understated the extent of the cost differential because it had erred in including items c and d:

“11.

Ms Karen Steyn for the LEA submits that the SENT erred in law by including in the cost of placing M at school L two elements (items c and d above) which were going to be incurred by the LEA whether he was placed there or not. The teacher of the deaf was on the staff of the hearing-impaired unit and would be paid the same regardless of whether or not M joined the unit. The taxi was already being used to carry two children to the school, and to add M to its passengers would cost the LEA no more. These elements she contrasts with the two genuine on-costs, items a and b. A personal learning support assistant was going to be needed for M in order to help him to cope with the mainstream part of his schooling if he went to school L. And the AWPU, a capitation fee paid to a maintained school by the LEA for each pupil placed there, would likewise be incurred only if M went there. If she is right about this, as we think she is, her other complaint of inadequate reasons becomes otiose.”

Paragraph 12 of the judgment summarises the response given on behalf of the parents:

“12.

Mr John Friel, for M’s parents, points out that there is no prescribed formula for determining unreasonable public expenditure for the purposes of s 9. As the judge held, it is a matter for the expert judgment of the SENT. Thus, says Mr. Friel, SENTs can and do fractionalise the entire country or borough education budget to arrive at the cost of educating a child in the state system. This exercise may be based simply on the annual running costs; or it may equally legitimately include the amortised cost of school buildings and so forth. It is a matter for the SENT in each case.”

The Court did not accept Mr. Friel’s submission. In paragraph 15 of the judgment Sedley LJ said:

“It seems to us that Mr Friel’s argument cannot be right, at least in the open-ended form in which he advances it. It is of course true that unreasonable public expenditure is not a term of legal art. But neither is it, in its present context, a protean concept capable of producing opposite outcomes on the same facts and figures depending on the individual tribunal’s choice of accountancy method. In our judgment the chief object of the last part of s 9 is to prevent parental choice placing an undue or disproportionate burden on the education budget. When one considers that a single placement in the independent sector may well cost a ring-fenced education budget more than a teacher’s salary, one can readily see why.”

Having said in paragraph 16 that where the choice is between two maintained schools the “efficient use of resources” will “intelligibly include comparative on-costs, such as transport and personal support, but in most cases it is unlikely to be helped by apportioning the LEA’s accounts or balance sheet”, Sedley LJ continued in paragraphs 17-19:

“17.

If so, there is no intelligible reason why a comparison of public expenditure as between an appropriate independent school and an appropriate maintained school should be at large. Mr. Friel, indeed, defends the quantification of the cost of school MH, the independent school, as the bare annual fee – that is to say, the cost to the LEA’s annual budget of placing M there. In our judgment exactly the same is true of the cost of placing M in the hearing-impaired unit of school L: the question is what additional burden it will place on the LEA’s annual budget. That means, generally speaking, that the existing costs of providing school L and of staffing it and its hearing-impaired unit do not come into account.

18.

This is not to say that there may not be particular cases in which some other method of comparison needs to be used in order to meet s 9. But as a matter of purposive construction of the section, it seems to us that what Parliament has called for in the ordinary run of cases is 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. Costs which either the private provider or the LEA would be incurring with or without the proposed placement are accordingly not in general relevant. This being so, it is not necessary to say anything about the accountancy problems which would bedevil any endeavour to quantify the per capita cost of providing for a child’s education in the state sector.

19.

It follows, in our judgment, that Ms Steyn is right in her submission that, absent any factors justifying a special approach, the SENT was wrong to treat items c and d, totalling £9,500 a year, as necessarily part of the cost of placing M in school L. We do not know, however, because the SENT has not told us in its reasons, whether it accepts the LEA’s case that the teacher of the deaf in the hearing-impaired unit would be present and paid in full whether M was placed there or not. This is a finding which should have been made, since the point had been argued, whichever way the decision was going to go. The same is true of the taxi occupancy, although here there seems not to be a dispute between the parties.”

In Coventry, the annual fees at the school preferred by the parents of the child, Georgia, were £11,500. The Tribunal concluded that there would be no unreasonable expenditure because the cost of additional teaching support, which Georgia would require if she remained at Potters Green school, the school in which she had been placed by the LEA, would be of the order of £13,000. The LEA appealed on the ground that the Tribunal had erred in including the cost of additional teaching support at Potters Green school. There would be no additional cost to the LEA because, under the “delegated arrangements” made by the LEA, the Individual Schools Budget for Potters Green school included a sum of some £93,000, out of which the school was expected to fund provision for all children with special educational needs, save in exceptional circumstances which did not apply in Georgia’s case. Relying on Oxfordshire, the LEA submitted that, because of the delegation agreement, there would be no additional burden on the LEA’s budget.

Underhill J did not accept the LEA’s submissions. In paragraphs 12 and 13 of his judgment he said:

“12.

I reject that submission. In referring to ‘the LEA’s annual budget’, Sedley LJ plainly did not have in mind the distinction on which Mr. Auburn [who appeared for the LEA] relies between payment by the LEA directly from its own pocket and payment by the school under delegated arrangements. That distinction had not featured in the argument before the Court of Appeal. The precise route by which payment is made out of what is ultimately the LEA’s budget is plainly immaterial to the purpose underlying s 9. The term ‘the LEA’s annual budget’ was no more than a paraphrase of the reference to public expenditure in s 9: Sedley LJ’s point was that the section was concerned with actual additional money paid out, no matter by whom.

13.

That conclusion seems to me to be right as a matter of principle, but it is reinforced by the terms of s 49(5) of the 1998 Act. Section 49 is part of the group of sections under Chapter 4 of Part II which establish the machinery for the financing of maintained schools. Section 49(1) provides that every maintained school shall have a delegated budget. Subsection (5) reads as follows:

‘Any amount made available by a local education authority to the governing body of a maintained school (whether under section 50 or otherwise)-

(a)

shall remain the property of the authority until spent by the governing body or the head teacher; and

(b)

when spent by the governing body or the head teacher, shall be taken to be spent by them or him as the authority’s agent.’

Mr Wolfe [who appeared on behalf of Georgia] submits that those provisions show that notwithstanding the power given to the school to spend the money under the delegated arrangements, the expenditure remains ultimately that of the council. In my view that submission is well-founded.”

The LEA also challenged the Tribunal’s decision on the ground that its reasoning was inadequate because it had failed to carry out any analysis of whether, and if so to what extent, Potters Green School could have met Georgia’s needs without having to incur additional expenditure. Responding to this ground of appeal, Underhill J said in paragraphs 16 and 17:

“16.

Although the council’s second ground of appeal is somewhat opaquely expressed, the essential point is that the tribunal failed to carry out any analysis of whether, and if so to what extent, Potters Green could have met Georgia’s needs without having to incur additional expenditure. As I have already had occasion to mention in connection with ground 1, the Court of Appeal in the Oxfordshire case has held that the equation required by s 9 in a case of the present kind involves the setting of the fees of the independent school against the actual additional cash expenditure which the maintained school will have to incur if the child in question attends it. Mr Auburn referred to this as the ‘marginal cost’: that is a useful shorthand, although I am a little wary of using too glibly economists’ technical terms, the full implications of which lawyers may not always understand.

17.

Assessment of the second element – that is to say the marginal cost – will depend upon the individual details of the provision required and the staffing arrangements at the school in question. To a greater or lesser extent it may be possible to meet the child’s needs by the use of staff who are already employed and to whom no further payment will fall to be made by reason of their making the provision in question – typically, for example, where a child is part of a group under a single special teacher or helped by a single assistant. That point is well illustrated by the facts of the Oxfordshire case itself. The court there held that the tribunal had been wrong to take into account the attributed cost of 5 hours per week of specialist teaching by a teacher for the deaf, because the teacher in question ‘was on the staff of the hearing-impaired unit and would be paid the same regardless of whether or not [the child] joined the unit’, and of transporting the child to school, because ‘the taxi was already being used to carry two children to the school, and to add [the child] to its passengers would cost the LEA no more’ (see paras [11] and [19]). That was in contrast to two other items, being the so-called ‘AWPU’ costs (a capitation fee paid to a maintained school by the LEA for each pupil placed there), and the cost of an additional learning support assistant: both of these were ‘genuine on-costs’.”

The LEA’s appeal was dismissed.

The report of Slough does not say whether the Court was referred to Oxfordshire. Although it is not cited in the judgment of Sedley LJ (with whom Rimer LJ and Sir Paul Kennedy agreed), it is most unlikely that the Court did not have Oxfordshire well in mind. Oxfordshire was not merely binding Court of Appeal authority; it was the leading authority on the proper approach to “unreasonable public expenditure” under section 9 of the 1996 Act. Sedley LJ was responsible for the judgments in both appeals, and Mr. Friel was counsel for the respondent in both appeals. Mr. Wolfe told us that it was his understanding (from which Mr. Rawlings did not demur) that the Court in Slough had been referred to Oxfordshire. It is sensible to proceed on the basis of that understanding.

In Slough the principal issue between the LEA and the child’s parents was the cost of a placement at the school preferred by the parents, the PACE centre: was the annual cost £36,000 or £10,000? The Court upheld the Tribunal’s conclusion that the correct figure was £10,000. A secondary issue was the cost of a placement at Arbour Vale, the school for which the LEA was contending. The LEA contended that that cost was only £4,161, and not the figure of £32,490 which had been adopted (albeit for reasons which it did not explain) by the Tribunal.

In paragraphs 9 and 10 of his judgment Sedley LJ said:

“9.

If the true cost of a placement at the PACE Centre were £36,000, it would without doubt be necessary to remit the case for an intelligible finding as to the true cost of a placement at Arbour Vale, for this could well be critical. But if, as I have concluded for reasons to which I will come, the tribunal was entitled to take the relevant fees at the PACE Centre as £10,000 a year, it is only if the fees at Arbour Vale were indeed as low as the £4,161 which Mr. Hyams advances that the s 9 condition for overriding parental choice would be met.

10.

The figure of £4,161 was advanced before the tribunal by the local education authority as the only additional expenditure which would be needed for this particular child at Arbour Vale, a maintained school in which there was a vacant place. It represented (somewhat surprisingly) the cost of additional full-time 1:1 support. For the rest, it was said, the cost was part of the provision already made at Arbour Vale.”

Sedley LJ then referred to the Tribunal’s conclusion:

“However, in the particular circumstances of this case, regardless of the staffing arrangements at Arbour Vale School, it is clear that the real cost of therapies, the scheme of delegated funding and the additional supplements committed by the LEA to this particular child in reality bring the full cost of a placement at this school beyond the £10,000 now being charged by the PACE Centre.”

He continued in paragraphs 12 and 13:

“12.

This appears to me to be a factual finding, made by a specialist tribunal with knowledge of the field, that the apportioned costs of providing for this child in a maintained school, whatever their precise amount, would inevitably exceed the £10,000 for which the PACE Centre was prepared to accept her. In my judgment this is a perfectly tenable finding unless M. Hyams is correct in his contention that admission to a maintained school with space for the child is cost-free apart from any special requirements that the child brings with her.

13.

The contention is not in my judgment sustainable. Every element of a maintained school carries a cost in public funds. The recurrent exercise for tribunals is to calculate what it is, because it is ordinarily only with such a calculation that the protection of public money to which the condition in s 9 is directed becomes possible. If it were not so, a like-for-like comparison between public and private provision could never be made. But here, because of the unusual facts, it was legitimate for the tribunal to take a short cut and to find, as it did, that whatever the notional per capita cost of the maintained school was, it must exceed the £10,000 with which it fell it be compared.”

In the present case, the FtT faithfully followed the approach in Oxfordshire. It included the cost of the AWPU (item b in Oxfordshire). It did not accept (see its conclusions K and N) the Appellant’s submission that there would be further costs, such as the additional learning support assistant who would have been required in Oxfordshire (item a). Mr. Wolfe submitted, correctly, that the cost of the AWPU was not in issue in Oxfordshire. The Court was concerned only with items c and d: whether the teacher of the deaf would be available, and paid, in any event, and whether there would be additional taxi costs. However it is clear that the Court accepted that the AWPU was a “genuine on cost”: see para. 11 of the judgment. If the Court had felt any doubt as to whether the AWPU was, on the facts of that case, a fair reflection of the additional burden that would be placed on the LEA’s budget by placing an additional child at school L (to which there had to be added the cost of a personal learning support assistant who was going to be needed for that particular child), it would surely have said so.

Mr. Wolfe submitted that the Tribunal should have concluded that there would be additional costs in this case (akin to the cost of the personal learning support assistant in Oxfordshire) because it had been “agreed on all sides that GM needs support from a learning assistant for 20 hours per week”: see the FtT’s conclusion D. However, the evidence of the Director of Learning Support at school H was that this need for support did not mean that GM would have to be withdrawn from his classes for up to 20 hours per week. Withdrawal would be limited to three sessions per week each of 40 minutes, and on this basis the Director “was able to say that [school H] would be well able to meet [GM’s] needs…. The learning support assistants had considerable skill and experience in dealing with the provision of support to pupils …..like [GM] …” (para. 29 of the FtT’s decision).

The FtT’s conclusion K – that additional costs should not be added to the AWPU so as to produce, for the purposes of proper comparison, a significantly higher figure for the cost of placing GM at school H must be considered against the background of the Director’s evidence. I do not accept Mr. Wolfe’s submission that the FtT was simply concerned in its conclusions K, N and P with the expertise of the learning support assistants within school H. It was also concerned with whether there were sufficient learning support assistants to meet a need for support which required GM’s withdrawal from classes for only three 40 minute sessions per week.

Mr. Wolfe submitted that there was a more fundamental error in the FtT’s approach. By including the cost of the AWPU the FtT was focussing upon the Local Authority’s (LA’s) budgetary allocation for H school, and not upon the expenditure of H school itself. The LA’s budgetary allocation for school H was not “public expenditure” for the purposes of section 9, the school’s expenditure of the budget allocated to it by the LA was the relevant “public expenditure” for the purposes of section 9. He submitted that Coventry (see paras 12 and 13) and Slough (see para 13) were authority for the proposition that in every case where the amount of additional public expenditure for the purposes of section 9 was in issue, the FtT had to consider, not the LA’s budgetary arrangements, but the school’s accounts, in order to properly carry out what Sedley LJ described in Slough as “the recurrent exercise for tribunals” to calculate the “notional per capita cost” of the maintained school. Such a calculation would exclude fixed costs, such as buildings (see para. 15 of Oxfordshire), but would include all variable costs, determined by reference to the school’s accounts.

In support of this submission Mr. Wolfe took us in some detail to the provisions in Chapter IV of the 1998 Act which deal with the financing of schools by local authorities. I do not think it necessary to set out all of those provisions, because I do not accept the underlying premise: that the LA’s budgetary arrangements for a particular school are not “public expenditure”, and that there is no public expenditure until the school spends the budget that has been allocated to it by the LA. It seems to me that there can be “public expenditure” for the purposes of section 9 at either or both stages of the process: when the LA allocates public funds to the school, and when the school spends those public funds.

Neither Coventry nor Slough is authority for the broad proposition advanced by Mr. Wolfe: that the FtT must always focus on the school’s accounts to the exclusion of the LA’s budgetary arrangements. Mr. Wolfe’s submissions treat the dicta of Underhill J and Sedley LJ as though they were enactments of general application, rather than responses to the particular circumstances of those two cases. In Oxfordshire the Court had rejected the, extreme, submission that for the purpose of section 9 the cost of sending a child to a maintained school could be obtained by fractionalising the entire county or borough education budget. In both Coventry and Slough the Court rejected submissions at the opposite end of the spectrum: that because of the way in which the local authorities in those two cases had made their budgetary arrangements with the schools in question, placing an additional child in the schools was, for the LEAs, “cost free” so there was no additional public expenditure for the purpose of section 9, whatever the cost to the schools might be: see para. 11 of Coventry and para. 12 of Slough. It is not easy to ascertain the facts in Slough, but it appears that the Court understood the LEA’s case to be that the placement would be “cost free” apart from the cost of additional 1:1 support.

There is nothing in Sedley LJ’s judgment in Slough which suggests that he was intending to depart from his earlier judgment in Oxfordshire, much less that he considered that Oxfordshire was wrongly decided. His statement in paragraph 13 of Slough that “Every element of a maintained school carries a cost in public funds. The recurrent exercise for tribunals is to calculate what it is …” (emphasis added) must be considered in the context of the arguments which were being advanced by the parties. It had not been suggested on behalf of the parents (nor was it part of the Tribunal’s reasoning) that the cost of sending the child to Arbour Vale School should be assessed by fractionalising the entire costs of either the LEA’s budget for the school or the school’s expenditure. Such a submission would have been contrary to Oxfordshire.

Although the SENT and subsequently the SENDIST had regularly been required to calculate the cost to public funds of sending a child to a maintained school, the “recurrent exercise” it had carried out on those occasions had, since the decision in Oxfordshire been carried out in accordance with the judgment in that case: that it had to determine what additional burden would be placed on the LEA’s (now the LA’s) budget. In some cases, of which Coventry and Slough were examples, the reference in Oxfordshire to the LEA’s annual budget had led certain LEAs to distinguish between payment by the LEA, and payment by the school, and to argue that even though there would undoubtedly be a need for additional payment by the latter, since there would be no additional payment by the former there would be no additional public expenditure for the purpose of section 9. Understandably, Underhill J in Coventry and this Court in Slough held that the Tribunal was entitled to reject such a “cost free” approach as artificial, and as not being a fair reflection of the cost to the public purse of placing a child in a maintained school for the purpose of comparison with the cost of placing the child in the parents’ preferred alternative.

It is one thing to say that the Tribunals in those cases were entitled to reject an approach that was plainly artificial, it is quite another to say that in every case the only proper measure of additional public expenditure is the expenditure of the school, by reference to an analysis of its accounts. Such an approach would place an unwarranted burden on the school, which would have to produce and explain its accounts, the LA and the parents who would have to respond to them, and the FtT which would have to consider them. While it is true that the FtT does, on occasion, have to examine financial issues in some detail – will the cost of a taxi service be increased by an additional passenger? - its primary expertise is in the field of education, not accountancy, and any general requirement for there to be a detailed analysis of a school’s accounts would be disproportionate. Whether a child’s parents’ choice of school would be compatible with the avoidance of unreasonable public expenditure is a question of fact to be answered by the FtT in a common-sense way. It is for the FtT to decide what evidence it considers most helpful in resolving this issue in any particular case. In undertaking that task it is entitled to have regard to other information, such as a school’s accounts, if it is not satisfied that the figures based on the LA’s budgetary arrangements are a fair reflection of the cost to the public purse of educating the child at the school in question.

Although they will not necessarily provide the definitive answer, the LA’s budgetary arrangements for an individual school will usually be a sensible starting point. The arrangements are made in accordance with Chapter IV of the 1998 Act and the School Finance (England) Regulations 2008, are readily ascertainable, and, as in the present case, can be easily (or at least relatively easily) explained by those who will be representing the LA before the FtT. If the LA’s budgetary arrangements do make provision for the payment of an AWPU to the school there is no reason why the FtT should not accept that the AWPU, together with any additional costs specifically incurred in respect of the child in question, for example transport costs or the costs of therapy or learning support if an additional therapist or learning support assistant has to be employed by the school, or if an existing therapist or learning support assistant at the school has to be paid to work additional hours (if it is satisfied that there are such additional costs), are a fair reflection of the cost to the public purse of educating the child at that school.

It should only be in those cases where there is no AWPU payment by the LA, or where the FtT is satisfied that, for some cogent reason, the AWPU plus any additional costs (as above) do not fairly reflect the cost to the public purse of placing the child in a particular school, that the FtT would consider it necessary to adopt some other method of calculating the public expenditure involved in that placement for the purposes of section 9. In the present case there was no need for the FtT to do other than adopt the AWPU. Mr. Rawlings demonstrated that in its evidence to the FtT the LA had described its budgetary arrangements, and had said that in the formula it had adopted “most funding” was allocated by reference to the number and ages of pupils at the school (the AWPU). The AWPU accounted for over £5.4 million of H school’s budget of nearly £6.7 million, of which premises costs accounted for nearly £560,000. It is common ground that, however the per capita cost is calculated for the purposes of section 9, fixed costs, such as premises costs, must be excluded: see Oxfordshire. In a case where the AWPU represented such a substantial proportion of the variable costs element of the LA’s budget for H School there was no reason to suppose that it did not fairly represent the cost of GM’s placement there, the FtT having rejected the submission that there would be any additional costs.

For these reasons I do not accept that the UT erred in its approach to Oxfordshire and Slough. As a matter of first impression, there does appear to be an inconsistency between the two decisions, but on analysis the inconsistency is merely apparent, and not real. The two decisions are consistent in that they both demonstrate that the question whether placing a child at a particular school would cause “unreasonable public expenditure” should be approached by the FtT in a common sense manner. Fancy accountancy footwork which produces an unrealistic result – whether an excessive figure based on global costs including fixed costs, or a “cost free” placement – is unlikely to be persuasive before the FtT. The decision of the FtT in the present case, to adopt the AWPU, was not unlawful. I would therefore dismiss this appeal.