England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM Chancery Division
Mr Nicholas Strauss QC

		Royal Courts of Justice Strand, London, WC2A 2LL
		09/11/2015

Lord Justice Underhill:

INTRODUCTION

1. The Respondent to this appeal, Mr Nicholas Rayner, was the defendant in an action brought against him by Mrs Kumari Murphy, who became his carer after he suffered a serious stroke. Part of her claim in the action was for a share in his flat and in his investments, title to which was vested in two off-shore companies which he owned and controlled and which were accordingly also defendants in the action. Mr Rayner counterclaimed for money and other valuables misappropriated by Mrs Murphy.
2. Mrs Murphy's legal representation was – subject to the hiatus referred to below – funded by the Legal Services Commission ("the Commission") under Part I of the Access to Justice Act 1999 ("the 1999 Act").
3. The claim and counterclaim were heard by Mr Jeremy Cousins QC, sitting as a Deputy High Court Judge, in the Chancery Division. The trial started on 19 July 2010, was adjourned at the end of the month and was resumed in September, with a further hearing in November. By a judgment handed down on 18 January 2011 Mrs Murphy's claim was dismissed and an account was ordered: following the taking of that account she was ordered to pay Mr Rayner just under £1m. Mr Rayner was awarded the costs of the proceedings on the indemnity basis. There is thought to be no realistic prospect of recovery of any substantial sums from Mrs Murphy.
4. There are two features of the legal aid system under the 1999 Act, and the predecessor legislation at least as far back as the Legal Aid Act 1964, which are central to this appeal. In summary:

(1) Cost protection. A funded party against whom an order for costs is made enjoys a (qualified) protection against having to pay those costs. I set out the relevant provisions at paras. 15-18 below.

(2) Liability of the funding authority. A non-funded party in whose favour an order for costs has been made against a funded party enjoys a (qualified) right to recover costs from the funding authority – that is, under the 1999 Act, from the Commission. At the time with which we are concerned the relevant provisions appeared in regulation 5 of the Community Legal Service (Cost Protection) Regulations 2000 ("the 2000 Regulations"), which were made under powers conferred by section 11 (4) (d) of the 1999 Act. I set out the provisions in full at para. 19 below, but the most crucial for the purpose of this appeal is regulation 5 (4), which defines the costs in respect of which recovery can be sought as "so much of [the non-funded party's] costs as is attributable to the part of the proceedings which are funded proceedings [my emphasis]".

These two features are closely related in as much as the justification for allowing a non-funded party to recover against the Commission is that the funded party is (normally) impecunious and it is only because of the Commission's involvement that the non-funded party has had to incur the costs of the funded litigation. But the extent to which they can be treated as precise corollaries of one another is one of the key issues before us.

5. The regime established by the 1999 Act was altered by the Legal Aid Sentencing and Punishment of Offenders Act 2012, and the 2000 Regulations have been replaced with effect from 1 April 2013 by the Civil Legal Aid (Costs) Regulations 2013 (though it appears that the replacement provisions are in substantially identical terms). At the same time the Commission was abolished (see section 38 of the 2012 Act), with responsibility for civil legal aid passing to the Lord Chancellor (who discharges his functions through the Legal Aid Agency). All the Commission's surviving liabilities have been transferred to the Lord Chancellor: see paragraph 7 of Schedule 4 to the 2012 Act.
6. Mr Rayner made a claim against the Commission under regulation 5 of the 2000 Regulations for the totality of the costs awarded against Mrs Murphy: the sum claimed is over £700,000. The Lord Chancellor accepts that Mr Rayner is in principle entitled to an order, but he does not accept liability for the full amount claimed. The issue arises from the fact that there was a period in the course of the proceedings during which Mrs Murphy was represented by solicitors who were not named in the funding certificate issued by the Commission. The facts can be sufficiently summarised as follows:

(1) Mrs Murphy initially instructed a firm called Ismail & Co., and a funding certificate naming them as the "current supplier" was issued prior to the commencement of proceedings.

(2) In May 2010 Ms Murphy decided that she wished to be represented by a different firm, Smithfield & Partners. Smithfield did not have a legal aid contract. They and Mrs Murphy evidently hoped nevertheless to be awarded an "exceptional case" contract, but she was not prepared to continue with Ismail & Co. while the appropriate application was made. On 18 May notice was served that Smithfield had been instructed under a conditional fee agreement, and on 19 May a formal notice of change of solicitors was filed. From that point Ismail & Co ceased to provide any services to Mrs Murphy. Smithfield proceeded under a CFA pending the determination of the application.

(3) The funding certificate in favour of Ismail & Co was not at any stage withdrawn. It was in due course amended to identify Smithfield as the supplier, but that was specified to be only with effect from 1 September 2010.

The period between 19 May and 31 August 2010 has been referred to before us as the "hiatus period". It will be noted that it covers the two months immediately before the trial and the first two weeks of the hearing, which were inevitably the period of most intensive pre-trial preparation and during which liability for brief fees would be incurred. It is estimated that about half of Mr Rayner's costs were incurred during that period.

7. On 18 July 2011 Master Haworth, sitting in the Senior Courts Costs Office, ruled that Mrs Murphy did not have cost protection during the hiatus period: I refer to this as "the 2011 decision". I need not set out the Master's reasoning in detail, but the essential point was that "during [the hiatus] period … [Mrs Murphy] did not have the benefit of funded proceedings" because she had ceased to instruct Ismail & Co, who were the only solicitors whose work the Commission had agreed to fund.
8. The Commission contended, and the Lord Chancellor now contends, that it follows from the 2011 decision that no order could properly be made against it under regulation 5 in respect of the costs incurred by Mr Rayner during the hiatus period: since the proceedings were not funded during that period, such costs were not "attributable to [a] part of the proceedings which [were] funded proceedings". Mr Rayner disputes this: his case is that the costs incurred during that period were "attributable to" the earlier stages in the proceedings, during which Mrs Murphy was funded, because they had to be incurred in response to the steps taken by her in that period. In a nutshell, the Lord Chancellor's case is that the relevant criterion is purely temporal – costs are attributable to the period of the proceedings when they were incurred – whereas Mr Rayner's case is that the criterion is one of causation – costs are attributable to the part of the proceedings which resulted in them being incurred.
9. That issue came before Master Haworth on 28 February 2013, in the context of an application by Mr Rayner for disclosure of papers held by the Commission about the work funded by it: this was said to be necessary in order to enable him to identify to what extent his own costs were "attributable" to the funded part of the proceedings. (There was also an issue about the extent to which costs claimed by him had in fact been incurred by the corporate defendants; but that is not before us.) The Master upheld the Commission's contention. His order records his finding that:

"[Mr Rayner] is not entitled to recover any costs incurred by him during [the hiatus period] when (as has already been determined [i.e. by the 2011 decision]) [Mrs Murphy] did not have 'costs protection'".

10. Mr Rayner appealed against that decision, with the permission of the Master. The appeal was heard by Mr Nicholas Strauss QC, sitting as a Deputy High Court Judge, on 10 October 2013. At the hearing Mr Rayner was represented by Mrs Teresa Rosen Peacocke and the Lord Chancellor by Mr John Gimlette; but when, following the hearing, the Judge asked for further written submissions on the part of the case with which we are concerned the submissions on behalf of the Lord Chancellor were drafted by Mr Guy Mansfield QC. The same counsel appeared before us.
11. Mr Strauss handed down a reserved judgment on 2 December 2013 ([2013] EWHC 3878 (Ch)), now reported at [2014] 1 WLR 677. He allowed Mr Rayner's appeal on the issue of principle, though he held that it remained necessary for him to show in respect of any costs incurred during the hiatus period that they "were caused by the [Commission's] funding of the earlier stages of the proceedings" and that that issue would require to be decided by the Costs Judge as part of his detailed assessment. He summarised his decision near the start of his judgment, saying at para. 9 (p. 680 H):

"I hold that Mr. Rayner's claim is not limited to the periods in which Mrs. Murphy was in receipt of funding. The effect of para. 5(4) is that Mr. Rayner is entitled to claim all costs, including those incurred in the period of the hiatus in funding, which are attributable to (in the sense of caused by) the part of the proceedings for which the claimant received funding."

Para. 2 of his order reads (so far as material):

"In relation to the … issue … whether, under regulation 5 (4) of the Community Legal Service (Costs Protection) Regulations 2000 (as amended), the costs 'attributable to the part of the proceedings which are funded proceedings' are limited to those costs which are incurred by the non-funded party during the period in which the funded party was in receipt of funded services:

(a) The appeal is allowed in that it is determined that the costs are not so limited.

… ."

(The remainder of para. 2 directs the Commission to make the disclosure sought by Mr Rayner.)

12. The Lord Chancellor appeals against the decision of Mr Strauss with the leave of Christopher Clarke LJ given on 24 April 2014.
13. In addition to that appeal there is, at least potentially, an application before us from Mrs Peacocke. I explain it more fully at paras. 57-63 below; but in bare outline it arises as follows. At a very late stage in the proceedings before the Judge Mr Rayner sought permission to appeal out of time against Master Haworth's 2011 decision. The decision on the primary appeal meant that the Judge did not have to reach a decision on that application, and he declined to do so. By an Amended Respondent's Notice Mr Rayner belatedly applied for permission to appeal against that refusal. That application came before Christopher Clarke LJ on the papers on 6 August 2014. He refused permission. He also directed that if the application were to be renewed it should be decided at the full hearing, with the appeal to follow if permission were granted. No formal renewal application has in fact been made, but Mrs Peacocke has made it clear in her skeleton argument that she wishes to pursue the application.

THE RELEVANT PROVISIONS

14. The relevant provisions are distributed between the 1999 Act and the 2000 Regulations in a manner which is rather clumsy and involves a good deal of cross-reference. (The predecessor provisions were rather more straightforwardly arranged, but there is – subject to one submission which I will need to consider in due course – no reason to suppose that it was intended that the 1999 Act should substantially alter the law in the respects with which we are concerned.) Although the primary issue before us concerns the liability of the Commission, it is convenient to start with the provisions about cost protection.

Cost Protection

15. Section 11 (1) of the 1999 Act reads as follows:

"Except in prescribed circumstances, costs ordered against an individual in relation to any proceedings or part of proceedings funded for him shall not exceed the amount (if any) which is a reasonable one for him to pay having regard to all the circumstances including—

(a) the financial resources of all the parties to the proceedings, and

(b) their conduct in connection with the dispute to which the proceedings relate;

and for this purpose proceedings, or a part of proceedings, are funded for an individual if services relating to the proceedings or part are funded for him by the Commission as part of the Community Legal Service."

At the risk of spelling out what is known to every civil practitioner, I should say a word about how that provision operates in practice. In the typical case the funded party is practically impecunious – otherwise they would not have received funding in the first place – and the Court does not normally regard it as reasonable to make any order for immediate payment of the non-funded party's costs. But there will be cases where their financial circumstances have improved in the meantime; and in such a case an order may be made. Provision is normally also made for the possibility of their financial circumstances changing in the future – what used to be called a "football pools order".

16. More detailed provision in relation to cost protection is made in regulation 3 of the 2000 Regulations. For present purposes I need only refer to paragraphs (3) and (4), which read as follows:

"(3) Subject to paragraph (4), cost protection shall apply only to costs incurred by the receiving party in relation to proceedings which, as regards the client, are funded proceedings, and:

(a) where work is done before the issue of a certificate, cost protection shall (subject to paragraphs (2) and (5)) apply only to costs incurred after the issue of the certificate;

(b) where funding is withdrawn by discharging the client's certificate, cost protection shall apply only to costs incurred before the date when funded services under the certificate ceased to be provided.

(4) Where funding is withdrawn by revoking the client's certificate, cost protection shall not apply, either in respect of work done before or after the revocation."

17. The phrase "funded proceedings" which appears in paragraph (3) is defined in regulation 2 (1) of the 2000 Regulations as "proceedings … in relation to which the client receives funded services or, as the case may be, that part of proceedings during which the client receives funded services". "Funded services" is defined as "services which are provided directly for a client and funded for that client by the Commission as part of the Community Legal Service under sections 4-11 of the [1999] Act".
18. The "certificate" referred to in both paragraphs is defined in regulation 2 (1) as "a certificate issued under the Funding Code certifying a decision to fund services for the client". The Funding Code is a code issued by the Commission under sections 8 and 9 of the 1999 Act: Part C of the Code Procedures requires the certificate to specify the essential terms of the funding, including the identity of the solicitor supplying the funded services.

Liability of the Commission

19. The liability of the Commission for the costs of a non-funded party is not dealt with in the 1999 Act but in regulation 5 of the 2000 Regulations, which reads (so far as material) as follows:

"Costs order against Commission

(1) The following paragraphs of this regulation apply where:

(a) funded services are provided to a client in relation to proceedings;

(b) those proceedings are finally decided in favour of a non-funded party; and

(c) cost protection applies.

(2) The court may, subject to the following paragraphs of this regulation, make an order for the payment by the Commission to the non-funded party of the whole or any part of the costs incurred by him in the proceedings (other than any costs that the client is required to pay under a section 11(1) costs order).

(3) An order under paragraph (2) may only be made if all the conditions set out in sub-paragraphs (a), (b), (c) and (d) are satisfied:

(a) a section 11(1) costs order is made against the client in the proceedings, and the amount (if any) which the client is required to pay under that costs order is less than the amount of the full costs;

(b) unless there is a good reason for the delay, the non-funded party makes a request under regulation 10(2) of the Community Legal Service (Costs) Regulations 2000 within three months of the making of the section 11(1) costs order;

(c) as regards costs incurred in a court of first instance, the proceedings were instituted by the client, the non-funded party is an individual, and the court is satisfied that the non-funded party will suffer financial hardship unless the order is made; and

(d) in any case, the court is satisfied that it is just and equitable in the circumstances that provision for the costs should be made out of public funds.

(4) Where the client receives funded services in connection with part only of the proceedings, the reference in paragraph (2) to the costs incurred by the non-funded party in the relevant proceedings shall be construed as a reference to so much of those costs as is attributable to the part of the proceedings which are funded proceedings.

(5)-(7) …"

20. A "section 11 (1) costs order", as referred to in paragraph (2), is defined in regulation 2 (1) as "a costs order against a client where cost protection applies". "Cost protection" is defined as "the limit on costs awarded against a client in section 11 (1) of the [1999] Act".

THE PARTIES' CONTENTIONS

21. Since the issue is one of pure law, it is most convenient to begin not with the judgments below but with the rival contentions of the parties. I take the Lord Chancellor's case first.

THE LORD CHANCELLOR'S CASE

22. The Lord Chancellor's core submission is, as I have said, that the test of whether the costs of a non-funded party are "attributable to the part of the proceedings which are funded proceedings" is whether they were incurred during the period that the proceedings were funded – and in this case it has already been held, by the 2011 decision, that the proceedings were not funded during the hiatus period. Mr Mansfield submitted that that was the natural reading of the words as a matter of language and having regard to the scheme of the Regulations; but he also relied on a number of authorities.
23. So far as the language and the scheme of the Regulations are concerned, Mr Mansfield's submissions can be summarised as follows:

(1) The key phrase was "attributable to the part of the proceedings which are funded proceedings" – not, N.B., "attributable to … the funded services". That takes us to the definition of "funded proceedings" in regulation 2 (1) – "that part of proceedings during which the client receives funded services". That is temporal language, and the meaning of "attributable" must accordingly likewise be temporal. That is clear if the words of the definition are brought into regulation 5 (4), so that it reads:

"The reference in paragraph (2) to the costs incurred by the non-funded party in the relevant proceedings shall be construed as a reference to so much of those costs as is attributable to [that part of proceedings during which the client receives funded services]."

It was this point that Master Haworth regarded as decisive. He said, at para. 27 of his judgment:

"To my mind it is simply necessary to add the definition of 'funded proceedings' at [sc. to] the end of regulation 5 (4) to conclude that … the only relevant period of time that the Defendant can recover any costs at all is when funded services are 'funded proceedings'. In other words costs cannot be recovered during the period when the Claimant … was not covered by public funding."

(2) Since the liability of the Commission for costs is intended to compensate the non-funded party for the fact that (normally) no order will be made against the funded party, the criterion for that liability should mirror the criterion for cost protection. The latter is purely temporal – that is, the funded party enjoys (qualified) protection against liability only for those costs of the non-funded party which were incurred during the period for which he is funded. (That is obvious as a matter of common sense but is confirmed by the terms of regulation 5 (3), which excludes liability for costs incurred "before" the issue of the certificate or "after" its revocation – an explicitly temporal test.) The criterion for the Commission's liability should thus also depend on the question whether the costs claimed were incurred during the funded period.

(3) Mr Mansfield drew attention to the parenthesis at the end of regulation 5 (2). This is concerned with the case – untypical but perfectly possible – where the Court has considered it reasonable to make an order for costs against a funded party. Its effect is that an order cannot be made against the Commission in respect of the costs so ordered, even though they were incurred during a period when proceedings were funded and cost protection applied. That showed that the draftsman was alert to avoid any possibility of dual liability in respect of the same costs. Yet that was precisely what Mr Rayner's case involved: if he was right, he could recover in respect of costs incurred in the hiatus period either against Mrs Murphy (because she did not enjoy cost protection) or against the Commission (because the costs were attributable to the funded period).

(4) The temporal criterion was straightforward to apply. All that is necessary is to identify the date on which the costs in question were incurred and see whether the proceedings were funded at that date. By contrast, the criterion advanced by Mr Rayner would involve the Court having to resolve what would often be difficult questions of causation and would lead to uncertainty. Master Haworth attached importance to this point, adopting at para. 22 of his judgment Mr Gimlette's submission that "there is a need for certainty in relation to these regulations". He continued:

"It is important to define when cost protection applies and when a successful defendant is able to recover costs against the Commission can know with some certainty what they can recover."

(Something has gone wrong with the wording there but the sense is clear.)

24. I turn to the authorities on which Mr Mansfield relied. He referred us to S v S [1978] 1 WLR 11; Turner v Plasplugs Ltd [1996] 2 All ER 939; Burridge v Stafford [2000] 1 WLR 927; and Mohammadi v Shellpoint Trustees Ltd [2009] EWHC 1098 (Ch). However, he accepted that Turner did not assist him as directly as Burridge, and that Mohammadi, which is a decision of Briggs J at first instance, did not in any material respect take matters further. I will accordingly confine myself to S v S and Burridge.

S v S

25. S v S arose out of an application made by a wife for the custody of her children. At the material time the statute governing the provision of legal aid was the Legal Aid Act 1974. Under the 1974 Act legal aid was granted by the Law Society, out of the legal aid fund. The provisions entitling a successful non-assisted party to recover costs from the fund were contained in sections 13 (1) and 14 (5) of the Act. Section 13 (1) read:

"Where a party receives legal aid in connection with any proceedings between him and a party not receiving legal aid (in this and section 14 below referred to as 'the unassisted party') and those proceedings are finally decided in favour of the unassisted party, the court by which the proceedings are so decided may, subject to the provisions of this section, make an order for the payment to the unassisted party out of the legal aid fund of the whole or any part of the costs incurred by him in those proceedings."

Section 14 (5) read:

"Where a party begins to receive legal aid in connection with any proceedings after those proceedings have been instituted, or ceases to receive legal aid before they are finally decided or otherwise receives legal aid in connection with part only of any proceedings, the reference in section 13 (1) above to the costs incurred by the unassisted party in those proceedings shall be construed as a reference to so much of those costs as is attributable to that part."

It will be noted that as a matter of drafting section 14 (5) covers three situations – where legal aid is received after the start of proceedings; where it is discontinued before the end of proceedings; or where it is "otherwise" received "in connection with part only of any proceedings". As regards the third situation, the drafting appears substantially identical to that of regulation 5 (4) of the 2000 Regulations. In fact, however, it seems to me that all three situations are simply variants of the same basic situation, namely where legal legal aid is received in connection with part only of proceedings – NB the word "otherwise" – and the operative provision of the section applies equally to each.

26. The husband was not legally aided, but the wife was granted legal aid on, literally, the eve of the hearing. The husband successfully resisted the application. It was not in dispute that he was entitled to an order under section 13 (1), but the issue was whether that order should cover counsel's brief fees and the costs of pre-trial preparation incurred prior to the grant of legal aid. Latey J held that they should. He believed that "attributable to" must have been intended to mean something more than "during" (see p. 16H). His essential reasoning appears at p. 17 E-F, as follows:

"In this case what is at stake is the work done in preparation for what was the part of the proceedings in connection with which the mother received legal aid, namely, the hearing. To what was that work 'attributable'? In the words of Chambers Twentieth Century Dictionary: To what was it ascribed, assigned, or considered as belonging to? To what was it inherent in or inseparable from? In the words of the Shorter Oxford English Dictionary: What was it owing to or produced by? Surely the hearing of this application. If it was attributable to anything else what was that something else? An academic exercise? A vacuum? Without that work the part of the proceedings in question could not have been properly conducted or presented (and, incidentally, children might have suffered). With it it was."

27. That decision was reversed by this Court (Stamp and Ormrod LJJ). The reasoning of Stamp LJ, giving the judgment of the Court, is at pp. 20-22. It is quite lengthy, but I think I need to set it out in full (I have added paragraph numbers for ease of reference):

"[10] … With all respect to the judge, he places upon the word 'attributable' in the context of sections 13 (1) and 14 (5) a construction which it will not tolerate. The word is one which takes its colour by reference to the context in which you find it.

[11] As we have already indicated, section 14 (5) is designed to limit the effect of section 13 (1). For subsection (5) to bite at all – and bite it must – the case must be one in which, in the terms of the last two words of the subsection, an amount of costs is attributable to that part. So one must look at the earlier part of the section to see what is connoted by the words 'that part'. Referring back to the introductory words of the subsection, one finds that the draftsman postulates three situations in which the subsection is to operate. He postulates a situation in which the assisted person begins to receive legal aid 'after the proceedings have been instituted', he postulates a situation in which the assisted person ceases to receive it 'before the proceedings are finally determined' and he postulates a situation in which the assisted person 'otherwise receives legal aid in connection with part only of any proceedings'. Accordingly the words 'that part' at the end of the subsection can mean only that part of the proceedings 'in connection with' which the assisted person receives legal aid. And so one must ask the questions, (1) in connection with what part of the proceedings did the assisted person receive legal aid and (2) are the costs of the unassisted person attributable to that part?

[12] It is common ground that a legal aid certificate cannot operate retrospectively and where, as here, the assisted person begins to receive legal aid after the proceedings have been instituted the 'part' of the proceedings 'in connection with' which he receives legal aid can be only that part of the proceedings which takes place subsequent to the time when he begins to receive it. He does not receive legal aid 'in connection with' that 'part' of the proceedings which are anterior to the time when the legal aid was received. So the inquiry in the instant case becomes an inquiry how much of the father's costs is attributable to the part of the proceedings subsequent to the time when the mother received legal aid. In any natural sense of the word 'attributable' the costs incurred by the unassisted party prior to the receipt of legal aid were attributable to the part of the proceedings anterior to the receipt of legal aid by the assisted person; and in our judgment in the context of a dichotomy between the periods before and after the receipt of legal aid they cannot also be 'attributable' to the part of the proceedings subsequent thereto.

[13] Accordingly if the matter rested there we would, as a matter of the construction of subsection (5), conclude that in the instant case (a) the part of the proceedings in connection with which the assisted person (i.e. the mother) received legal aid was the part thereof subsequent to 5 p.m. on December 2 and (b) that the costs of the unassisted father incurred prior thereto were not attributable to 'that part'.

[14] Nor in our judgment does that construction of the legislation produce an anomalous or unfair result or one which is in any way contrary to any conception of natural justice. On the contrary, in our judgment, it is logical, fair and sensible as long as one party is being 'maintained' (we use this word in its ancient sense) in the litigation out of a fund provided by Parliament; it is fair, right and just that the costs to which the other party, being successful in the proceedings, is put during the period of such maintenance should be met by the fund. There can however be no good reason for making the fund liable for the expenses to which a party was put before ever the fund came on the scene. Here the father would have incurred the costs if the mother had never become legally aided but had submitted to having her application dismissed with costs on the morning of December 3, 1975.

[15] Far from being a fair or sensible construction of section 14 (5), that for which the father contends would introduce an anomaly and injustice in a case where the legal aid certificate was withdrawn on the eve of the trial instead of issued at the time: for Mr Hollis [counsel for the father] was constrained to concede – and rightly – that on the father's construction the costs incurred by the unassisted party in preparation for the trial at a time when the state was maintaining the litigation on the other side would not be recoverable from the fund. If subsection (5) was open to more than one construction, that which produces such an anomaly is to be rejected in favour of one that does not.

[16] We would only add that if any of the costs of the father incurred prior to December 3 could be regarded as attributable to the part of the proceedings on and subsequent to that date, the taxing masters would, we think, find great difficulty in distinguishing the costs so attributable from those not so attributable. Unless the line be drawn at the point of time at which the assisted party 'begins' or 'ceases' to receive legal aid, a choice of some other point of time could hardly do otherwise than produce an arbitrary result.

[17] It is not necessary for the purposes of this judgment to form a view whether the third case postulated in subsection (5) by the words 'or otherwise receives legal aid in connection with part only of any proceedings' is intended to do more than cover the case where the assisted person receives legal aid intermittently during the proceedings."

28. Mr Mansfield acknowledged that the specific question in S v S was not the same as that in the present case, because it was concerned with costs incurred before the grant of legal aid rather than during a hiatus period. But he submitted that the Court's reasoning was nevertheless equally applicable in such a case. The question of principle which it addressed was what was meant by "attributable" in a provision which is the predecessor to regulation 5 (4) and in which the key words are identical. The Court explicitly rejected Latey J's broader interpretation (see para. [10]) in favour of one which depended on whether the costs had been incurred before the grant of legal aid: that is, it applied a temporal criterion (see paras. [12]-[13]). It went on to explain why such an approach was fair (paras. [14]-[15]) and made the same point as Mr Mansfield about the practical difficulties to which the alternative approach would give rise (para. [16]).

Burridge v Stafford

29. This case in fact covers two appeals – Burridge v Stafford and Khan v Ali – which were heard together because, as Lord Woolf MR put it at the start of his judgment (p. 929 C-D), they both raised the question of "when a person ceases to be a legally assisted party for the purposes of sections 17 and 18 of the Legal Aid Act 1988". He continued:

"The question is of some considerable significance because while a party is legally assisted, his liability for costs is limited under section 17 and under section 18 an unassisted party has certain rights to recover costs from the Legal Aid Board."

30. I should set out the relevant parts of those two sections of the 1988 Act. Section 17 (1) read:

"The liability of a legally assisted party under an order for costs made against him with respect to any proceedings shall not exceed the amount (if any) which is a reasonable one for him to pay having regard to all the circumstances, including the financial resources of all the parties and their conduct in connection with the dispute."

I need not set out the parts of section 18 conferring the power on the Court to require the Board to pay the unassisted party's costs. The only part relevant for our purposes is sub-section (6), which read (so far as material):

"… where a party begins to receive representation after the proceedings have been instituted, or ceases to receive representation before they are finally decided or otherwise receives representation in connection with part only of the proceedings, the reference in sub-section (2) above to the costs incurred by the unassisted party in the proceedings shall be construed as a reference to so much of those costs as is attributable to that part."

It will be seen that those provisions are in materially identical terms to those with which we are concerned.

31. The procedural histories in both cases are a little complicated, but for present purposes the way the question identified by Lord Woolf arose can be sufficiently summarised as follows:-

In Burridge the unsuccessful defendant had been granted legal aid to defend proceedings but her solicitors had come off the record, because she had rejected their advice that she should accept a settlement offer, and she had continued to defend the claim in person; her legal aid certificate was not, however, discharged. The issue was whether she was entitled to cost protection for the period after the solicitors ceased to act.

In Khan Mr Khan had been granted legal aid for an appeal. Before the hearing, however, his solicitors came off the record. He pursued the appeal in person and lost. Again, however, his legal aid certificate had not been discharged. The issue was whether the respondent to the appeal could recover from the Board in respect of costs incurred after the date on which his solicitors came off the record.

It will be seen that it is in fact Khan which is more directly analogous with the present case because it was concerned with the liability of the Board – though an important part of Mr Mansfield's case is that the Court regarded the same approach as necessary in both situations; but for convenience I will refer to both as Burridge.

32. At paras. 26-30 of his judgment (pp. 933-5) Lord Woolf, with whom Butler-Sloss and Robert Walker LJJ agreed, reviewed a number of possible points in time at which – to quote from the heading to this part of the judgment – "a party cease[s] to be an assisted party for the purpose of sections 17 and 18 of the Act". He rejected (following Turner v Plasplugs) the argument that he remained legally assisted until the legal aid certificate was discharged. He was reluctant to accept the Board's submission that the crucial date was when the solicitor named in the certificate ceased to act, partly because that would be something of which the other party was unaware. But he held that, on the facts of the two cases before the Court, the date could be no later than the date on which the previously assisted party started to act in person: "[f]rom that date onwards he loses the protection of section 17 and the board ceases to be liable for him under section 18" – see para. 30 (p. 935 B-D).
33. Mr Mansfield relied on Lord Woolf's statement, quoted at para. 29 above, that "while [my emphasis] a party is legally assisted … an unassisted party has certain rights to recover costs from the Legal Aid Board" and on the fact that the entire analysis in paras. 26-31 was conducted in temporal terms, as regards section 18 just as much as section 17.

MR RAYNER'S CASE

34. As already noted, Mrs Peacocke sought permission to challenge the 2011 decision – that is, Master Haworth's decision that Mrs Murphy was not entitled to cost protection in respect of costs incurred during the hiatus period. I consider that application below, but at this stage I will proceed on the basis that the Master's decision stands.
35. Mrs Peacocke's primary submission in support of a causal approach is that as a matter of ordinary English the word "attributable" connotes a relationship of causation. It is not apt to a relationship based simply on things happening during the same period. If the draftsman had meant what the Lord Chancellor contended it would have been very easy for him to say so. He could for example, have referred to "so much of those costs as were incurred during the part of the proceedings which are funded proceedings".
36. Mr Mansfield's submissions to the contrary depend on the illegitimate importation of temporal language used in other provisions for different purposes. Specifically, taking the points which I have enumerated at para. 23 above, her submissions were:

(1) The language of regulation 2 (1) was neutral as regards regulation 5 (4). No doubt it defined a part of "funded proceedings" in temporal terms; but that had no bearing on the separate question of which of the non-funded party's costs are "attributable to" that period.

(2) There was no need for a match between the criteria governing eligibility for cost protection and those governing the right of recovery against the Commission: they are different rights.

(3) The possibility of "dual liability" did not undermine her case. Both the power to make a cost protection order and the power to make an order against the Commission were, in substance, discretionary: as regards the former, cost protection will only extend as far as the Court considers reasonable having regard to all the circumstances (see section 11 (1)); and as regards the latter the Court will only make an order against the Commission where it is "just and equitable" to do so (see regulation 5 (3) (d)). Those powers would be exercised so as to ensure that that no overlap arose in practice.

(4) She acknowledged that Mr Mansfield's approach would be more straightforward to apply, but she did not accept that her own approach was impracticable. She submitted that, whatever may have been the position previously, the regime for the public funding of civil litigation introduced under the 1999 Act made it perfectly possible for a costs judge to identify steps taken by the non-funded party in response to the steps funded by the Agency. As she put it in her skeleton argument:

"Under the 1974 legislation funding for legal representation was a blanket provision, covering matters arising in the litigation during the currency of the legal aid certificate. Funding representation was not restricted to specified services, but simply to "representation". The same was essentially true under the Legal Aid Act 1988. A substantial change was introduced in the Access to Justice Act 1999, by separate regulations under which the [Lord Chancellor] funded specific legal services limited in amount and/or by reference to the act(s) authorised to be performed by those representing the funded party."

37. More generally, and importantly, Mrs Peacocke emphasised the unfairness to which the temporal approach could lead. There is an inevitable lapse in time between the taking of a procedural step by a funded claimant – most obviously, the issue of proceedings – and the steps which the non-funded defendant is obliged to take in response. If the Commission has enabled an unfounded claim to be brought, the justice of requiring it to meet the defendant's costs of responding to it is the same whether or not funding has ceased in the interval before those costs are incurred. The present case is not a bad example, but still starker instances can be posited. For example, where a funded claimant serves a claim form with elaborate particulars requiring an expensive and fully-pleaded defence but ceases to be funded before that work is done, it would be arbitrary and unfair for the defendant not to be able to recover his costs of drafting that defence from the Lord Chancellor – and the need for the defence is, on any natural use of words, "attributable to" the funded part of the proceedings. She submitted that the importance of ease of application relied on by the Lord Chancellor could not outweigh the interests of justice and fairness.
38. As regards S v S, Mrs Peacocke's primary submission was that it was, as indeed Mr Mansfield accepted, concerned with a different situation from the present case. At the time when the costs in question had been incurred the wife had not been legally assisted: what the husband was trying to do was to treat them as "attributable to" the subsequent part of the proceedings following the grant of legal aid. It was unsurprising that the Court had held that, as she put it, "causation does not operate backwards". The present case was the reverse. But she also submitted that the scheme of the 1999 legislation was materially different from previous versions of the legislation: see para. 36 (4) above.
39. As regards Burridge, Mrs Peacocke submitted that the observations of Lord Woolf relied on by Mr Mansfield were not binding. The case proceeded on the assumption that the criteria required both for cost protection and for the liability of the Commission were temporal – that is, were the costs incurred during a period when the plaintiff was legally assisted ? – and thus that the only issue was how that period was to be defined. There was no argument about the meaning of "attributable" in section 18. She submitted that in such a case the reasoning could not be regarded as authoritative. She referred to the decision of this Court (Schiemann and Buxton LJJ and Jacobs J) in R (Kadhim) v Brent London Borough Council Housing Benefit Review Board [2001] QB 955.^[1] The relevant part of the judgment of the Court (given by Buxton LJ) is sufficiently summarised in the headnote, as follows:

"… a court was not bound by a proposition of law which, although part of the ratio decidendi of an earlier decision, had been assumed to be correct by the earlier court and had not been the subject of argument before, or consideration by, that court; that that exception to or modification of the strict rule of precedent was to be applied only in the most obvious cases and limited with great care, and would not normally apply unless the point had not been expressly raised before the court ... ."

THE JUDGMENT OF MR STRAUSS

40. The Judge's principal reasoning on this issue appears at paras. 80-103 of the judgment (pp. 703-9). But I should refer first to his review of the authorities, which is at paras. 58-79 (pp. 690-703). At paras. 67-70 (pp. 693-5) he considers whether he is bound by Lord Woolf's observations in Burridge. He rejects a submission by Mrs Peacocke that they were distinguishable because they were concerned with the 1988 Act. He says, at para. 67 (pp. 693-4):

"What now appears in para. 5 (4) of [the 2000 Regulations] can be traced back to section 2 (5) of the Legal Aid Act 1964, which introduced protection for defendants to legally assisted claims, where there was severe financial hardship and it was just and equitable, via sections 14 (5) and 18 (6) respectively of the 1974 and 1988 Acts. The wording has been streamlined, but I can detect no change of substance."

However, at paras. 68-70 (pp. 694-6) he holds that Burridge "falls within the narrow class of cases in which a decision is not binding because it has been reached on a basis which the court assumed to be correct without any argument to that effect", citing Khadim. At paras. 76-79 he quotes extensively from the judgments of Latey J and this Court in S v S: he does not at that point enter on any analysis, but see para. 46 below.

41. Turning to the Judge's principal reasoning, his starting-point, after some preliminary observations about the principles of statutory construction, is to conclude that the phrase "attributable to" is, as a matter of language, capable of bearing either of the meanings contended for by the parties: para. 83 (p. 703 F-H). At para. 84 (p. 704 A-B) he describes as "telling" Mrs. Peacocke's point that it would have been easy for the draftsman to use language explicitly adopting a temporal criterion.
42. At para. 85 (p. 704 B-E) the Judge identifies the rationale behind regulation 5 (4) and its predecessors as being that "if the [Commission] funds proceedings by an impecunious person, from whom it is unlikely that the other party could recover costs, it is only fair that the other party should have a remedy against [it]", supporting that proposition by reference to the judgment of Lord Phillips MR in R (Gunn) v Secretary for State for the Home Department [2002] 1 WLR 1634 (see at para. 50)^[2]. At paras. 85-87 (pp. 704-5) he draws an analogy with the policy behind the jurisdiction to make orders against third party funders of litigation under section 51 of the Senior Courts Act 1981 and points out that in the case of such orders it is recognised that "a causal link is an essential or at least important ingredient". (It is convenient to note at this stage that I do not regard this as a pertinent point. It is hard to see how a court exercising the general discretion conferred by section 51 could hold a third party liable for a party's costs unless he had caused or contributed to those costs being incurred.^[3] But on both parties' approaches in the present case the Commission will, by its funding, have caused the non-funded party to incur the costs in question: the issue is whether that is only the criterion for liability.)^[4]
43. At paras. 88-90 (p. 705 A-F) the Judge rejects the Lord Chancellor's argument that the criteria for identifying the costs in respect of which an assisted party enjoys cost protection and those in respect of which the Commission is potentially liable must be the same. He describes the argument as "at first sight persuasive" and as being supported by Lord Woolf's observations in Burridge, but he had already held that those observations were not authoritative: see para. 40 above. He continues:

"… [I]f one considers what is likely to be the position in most cases, it does not make much practical sense despite its logical symmetry. In most cases, including the present one, where the funded party loses funding by changing solicitors, or decides to act in person, this will not be because his financial circumstances have dramatically improved, and (assuming the causal link is proved) the non-funded party will be no less financially affected in consequence of the earlier funding; the right to proceed against the other party in respect of the period for which there is no cost protection is likely to be of little or no value, as in the present case. Mr. Gimlette has referred to the exceptional case, giving the inevitable example of the litigant who, having lost the litigation, wins the lottery. But in such a case the non-funded party could not recover against the [Commission], because he would be able to recover from his adversary, who has no cost protection and, anyhow, is now solvent; therefore he could not satisfy the conditions in para. 3 of the regulations, which require him to show financial hardship (which there would no longer be), and that an order is just and equitable."

44. At paras. 90-95 (pp. 705-6) the Judge considers what he describes as "the question of uncertainty", which he analyses as covering two points –

"… first, the uncertainty which may be felt by the non-funded party at the time and, secondly, the uncertainty involved in deciding after the event what part of the costs are attributable to the funded part of the proceedings or, to put it another way, does Mrs. Peacocke's construction lead to unduly complex and difficult issues after the event, when costs fall to be assessed?"

I need not trouble with the first of those points, which was not relied on by Mr Mansfield. As to the second, the Judge says, at para. 94 (p. 706 B-G):

"As to the point made by Mr. Gimlette that "a task-for-task analysis would be extremely impractical and difficult to implement, and would introduce uncertainty", hence the preferable time-related test, offering "certainty and finality at a time when the parties are making important decisions about litigation", there are in my view a number of answers:-

(a) Any assessment of costs is likely to involve detailed and intricate issues; in the present case the costs judge said that a bill of this size would take "between one and two weeks" to assess; investigating the extent to which the non-funded party's costs were the result of the funding of the other party is unlikely to be beyond the scope of the expertise of the costs judge, or to impose on him an unduly burdensome additional issue.

(b) The undertaking of such a task appears to be what has to be done, at least in some cases, by a court considering a section 51 application, without the benefit of the expertise of a costs judge; if it is appropriate on a section 51 application, it is difficult to see why it should be inappropriate here.

(c) The exercise would not normally be a task-for-task analysis, although it might in some cases be so e.g. if the funding was limited to particular issues such as the contempt of court proceedings in Littaur [1986] 1 WLR 287. Usually, it would be more likely to be a question of how long the funding had a continuing effect.

(d) It is difficult to see what important decisions would depend on whether the other party is funded; the non-funded defendant cannot escape from the proceedings either way.

(e) Even on Mr. Gimlette's argument, there could be difficult assessments to make e.g. what parts of the costs of continuous preparation fall within or outside a particular period; and there might have to be more than one assessment.

(f) More generally, if there is a choice between fairness and the avoidance of awkward issues as a factor in deciding what the legislative intent was, a court should unhesitatingly opt for the former; the legislature would assume that it was the function of the court to decide difficult issues, and to do so in a way that achieved a fair result."

45. There follows a long passage (pp. 706-8) in which the Judge explains why he regards Mr Rayner's approach as producing a fairer result than the Lord Chancellor's. I think I should reproduce it in full:

"96. Given that the object of para. 5 of the [2000 Regulations], and of the provisions of earlier Legal Aid Acts, is to provide fair protection for the other parties to funded litigation, subject to satisfying the financial hardship and just and equitable criteria, it seems to me that the construction of para. 5(4) contended for by Mr.  Rayner makes far better sense than the [Commission's] construction. The facts of the present case illustrate this. The probability is that there will have been a causal link between the finance provided by the [Commission] from the drafting of the proceedings down to a few weeks before trial and the costs incurred by Mr.  Rayner in the period in question, notwithstanding the intervention of a temporary CFA. It seems unlikely (although it is not for me to decide this) that Mrs. Murphy would have got as far as she did without the [Commission's] assistance. The fact that, as from 19^th May 2010 Mrs. Murphy was not legally assisted because she decided to change her solicitors, was completely outside Mr.  Rayner's control, and (assuming the causal link) is unlikely to have affected the amount of the costs that he had to incur as a result of the [Commission's] prior funding.

97. Further, the [Commission]'s construction would apply in many other situations. Changes of solicitors are common, arising as they do from client dissatisfaction, individual solicitors moving firms and other reasons, and amendments to certificates to cover the new solicitors are not retrospective (see Funding Code Procedures C38). In some cases, the new solicitors will do nothing until the certificate is amended, yet the other party may be forced by existing court orders, or the imminence of trial, to incur costs. In other cases, the claimant may represent himself for a time, or even for the rest of the case, having launched and formulated the case with funded legal help. If the [Commission] is right, the non-funded party would be unable to recover any costs incurred in any of these intervals, short or long.

98. The automatic effect of the [Commission's] construction would to my mind often work unfairly in circumstances such as these, and would also be random and arbitrary, in that the amount of costs rendered irrecoverable by the non-funded party is made to depend upon the stage of the proceedings at which the funded party decides to change solicitors, or act in person, whether the new solicitors have a contract with the Commission, the efficiency with which the new solicitors and the Commission go about reinstating … and other matters which are unlikely to affect the fundamental points that the non-funded party's costs will have been caused by the Commission having funded the proceedings, and that this will often not have ceased to be true merely because the funding has ceased, often for a limited time. In all cases, the extent to which the non-funded party was affected would be determined by what happened to be the steps which he had to take in the non-funded period. The Commission's construction will often result in the non-funded party losing out for no good reason, even if not always to the extent that Mr.  Rayner would in this case.

99. By contrast, the construction contended for on behalf of Mr.  Rayner works sensibly, not only on facts such as those in this case, but also in other situations. Essentially, as in section 51 cases, where the party is funded for part of the proceedings, and the non-funded party has incurred costs outside that period, attributability depends on a causation test. In the present case, it may well be that the costs judge will conclude that Mrs. Murphy would not have reached trial without the [Commission's] funding over most of the period of the case. In another case, in which the proceedings are funded over only a short period, and the claimant then manages to proceed to trial, either acting in person or with some form of private legal assistance, a costs judge might well reach a different conclusion. In all para. 5(4) cases, it is reasonable and fair that the [Commission] should be liable for the cost it has caused the other party, but for no more. To revert briefly to the example given earlier, it would be verging on the absurd for the defendant to be unable to recover the costs of preparing the Defence merely because, in the period in which it fell to be served, the claimant happened to be in the process of changing his solicitors."

46. At paras. 100-102 (p. 708 A-H) the Judge considers how the issue might arise in other factual situations. In that context he considers S v S. He observes (para. 100) that

"… Mrs. Peacocke's construction works, and is entirely consistent with the decision in that case: the non-funded party could not possibly say that any costs incurred before funding was provided were "attributable to" the funded part of the proceedings in the sense contended for; clearly, there could be no causal link between the funding and costs previously incurred by the non-funding party … ."

But he acknowledges that it is less easy to reconcile with the reasoning adopted by this Court, including its rejection of the reasoning of Latey J. He rejects Mrs Peacocke's argument that S v S could be disregarded because of the differences between the regime introduced by the 1999 Act and that under the predecessor legislation (see paras. 36 (4) and 38 above), saying, at para. 101 (p. 708D):

"While doubtless the new regime is different in some ways, the essential point, that proceedings may be partly funded by reference to issues, or funded only for a particular period or periods, has not changed; as can be seen from the judgment in Turner, precisely the same considerations arose at that time. Nor, as I have said [the reference is evidently to the passage quoted at para. 40 above], is there any significant difference between the wording of the relevant provisions of the Legal Aid Acts and the wording of the present regulations."

It is convenient to say at this stage that I agree with all of that. But he holds that the ratio of S v S is limited to the particular context in which the issue arose, i.e. where the costs in issue were incurred before the grant of legal aid. While the actual language used by Stamp LJ goes further and appears to espouse a temporal criterion in every case – that is, "that the non-funded party cannot make any claim for costs incurred outside the period(s) of cost protection" – that was to be treated as no more than an "assumption", on a point which did not arise on the facts of the case. He refers back to his reasoning in relation to Burridge: para. 102 (p. 708H).

47. The Judge summarised his decision at para. 103 (pp. 708-9) as follows:

"… For all the reasons I have set out above I consider that Mrs. Peacocke's construction of the regulations is correct. It is at least as consistent with the language of the relevant provisions as the rival construction, and accords much better with their purpose. Therefore, Mr.  Rayner is entitled to recover such of his costs incurred in the period of the hiatus as were caused by the [Commission's] funding of the earlier stages of the proceedings, to be decided by the costs judge in part of his detailed assessment. I allow the appeal on this issue and direct that the issue as to whether there was a causal link between the funding of Mrs. Murphy's case up to 19^th May 2010 and all or any of Mr. Rayner's costs between then and 31^st August 2010 be referred to the costs judge. I also allow the appeal against the refusal of disclosure of the case plans and bills of costs, which are relevant to this issue."

DISCUSSION AND CONCLUSION

THE APPEAL

48. I agree with the Judge that the authorities relied on by Mr Mansfield are not binding on us, essentially for the reasons given by him (which adopt some but not all of the points made by Mrs Peacocke). In my view the ratio of S v S goes no further than that costs incurred by an unassisted party prior to the grant of legal aid to another party cannot be regarded as attributable to the part of the proceedings subsequent to the grant. But that is not inconsistent with applying a causal criterion. Causation is necessarily temporal to the extent that events cannot be caused by a state of affairs that has not yet occurred; but Mr Rayner's claim involves attributing his costs to a past state of affairs. As for Burridge, the argument relied on by Mr Rayner was simply not run, and the case falls squarely within the principle explained in Kadhim.
49. It is accordingly necessary to consider the parties' arguments apart from authority. There are powerful points to be made on both sides, and I have not found deciding between them straightforward. I agree with the Judge that the phrase "attributable to", taken in isolation, is capable of bearing either of the constructions contended for. I feel the force of Mrs Peacocke's argument that the Lord Chancellor should as a matter of straightforward fairness be liable for costs which Mr Rayner was required to incur as a result of the Commission's funding of Mrs Murphy's case. But I have, on balance, come to the conclusion that Mr Mansfield's case is to be preferred. My reasons are essentially threefold.
50. First, I think that the phrase "costs … attributable to the [funded] part of the proceedings" does more naturally mean costs incurred during that part. I agree that it does not necessarily follow from the fact that a "part" of the proceedings is defined in temporal terms that the criterion for attribution is also temporal (see para. 36 (1) above). But the phrase does still seem to me to point in that direction. I take the point that the draftsman would have made life easier for us if he had simply referred to "costs incurred during the funded part of the proceedings". But he could also very easily have said "costs incurred as a result of the funded services" or "as a result of steps funded by the Commission". The fact that he used ambiguous language is the problem; it does not help in reaching the solution.
51. Secondly, I find Mr Mansfield's points summarised at para. 23 (2) and (3) above compelling. In my view it is reasonably clear that the draftsman proceeded on the basis that the criteria governing the qualification of the funded party for cost protection and those governing the liability of the Commission for the costs of the unfunded party should match. That would be the natural starting-point, since the two rights are essentially corollaries of one another: since cost protection is clearly defined by reference to the period during which funding is in place (see regulation 5 (3)), it would follow that the Commission's liability was equally so defined. I accept that the draftsman might nevertheless have taken the more sophisticated view that the two criteria should differ, in order to reflect the potential mismatch which is at the heart of Mr Rayner's case between the period when the funded party is protected and the period when the non-funded party may incur costs. But that would necessarily raise the problem of the Commission being potentially liable for the same costs as the funded party. Mrs Peacocke's answer, as recorded at para. 36 (3) above, is that the Court would not in practice make orders which resulted in dual liability. That may be so, but the point is about the scheme of the Regulations. In my view the whole structure of regulation 5 – including but not limited to the words in paragraph (2) to which Mr Mansfield attaches importance – rests on the basis that cost protection is in place as regards the selfsame costs of the unfunded party for which the Commission is to be (potentially) liable. Having reached that point without reference to authority, I do find some support in the fact that this Court in Burridge clearly regarded the criteria governing cost protection and the potential liability of the Legal Aid Board as mirroring one another.
52. Thirdly, I am concerned about the nature of the exercise involved in Mrs Peacocke's approach to "attributability". I take by way of example a case where a claimant is funded under the 1999 Act to bring proceedings and serves a claim form; where funding is discontinued immediately thereafter; where he nevertheless pursues the claim to trial without representation, putting the defendant to enormous cost; and where he loses and an order for costs is made against him (though subject to cost protection for the initial phase). Suppose that the non-funded party claimed the entirety of his costs from the Commission up to the point of judgment on the basis that they were all attributable to, in the sense that they were the result of, the funding of the initial steps in the claim. There seem to me to be in principle two possible responses to such a claim:

(1) It might be argued that the costs would have been incurred anyway because the claimant would if necessary have initiated proceedings without the benefit of any funding – in other words that the "but for" test was not satisfied; and so Mrs Peacocke's causation criterion was not met. There may be cases where that would be reasonably easy to demonstrate. But in other cases it would involve a difficult and speculative enquiry, far removed from the normal scope of a costs assessment.

(2) Even if the exercise considered above were performed, there would certainly be some cases in which the "but for" test would be passed – that is, it would be clear that the claimant would not have been able to get the case off the ground without the benefit of the initial funding, and that in that sense all the defendant's costs were the result of the initial funding. On the face of Mrs Peacocke's test, all those costs would be recoverable from the Commission. However, it might then be argued that they should only be regarded as "attributable" to the initial funding if the connection were sufficiently close – in other words, that some kind of directness and/or remoteness test had to be applied. The paradigm case of a cost that would pass the test would no doubt be that of preparing and serving the pleaded defence: that was a necessary step in response to the service of proceedings. But how much further would it go ? What about disclosure ? What about the preparation of witness statements ? It might be said that anything done in response to a procedural development after the claimant had ceased to be funded – say in response to a timetable set by the court – was too indirect or remote a consequence of the initial funding to be attributable to it; but that is far from self-evidently correct and might be thought somewhat arbitrary.

53. I do not regard it as likely that Parliament^[5] intended that the Court – more specifically, costs judges – would have to become involved in either of those kinds of enquiry. An enquiry into whether the costs in question (assuming them to be otherwise recoverable) would have been incurred but for the initial funding might often involve a substantial factual investigation. And even if that was acceptable, the criteria for distinguishing between those "post-funding" costs that are attributable to the initial funding and those that are not seem to me unacceptably uncertain.
54. I have thus far, in the interests of simplicity, been addressing the rather basic case postulated at para. 52 above. But my concerns are reinforced by a consideration of the circumstances of the present case. I am bound to say that if I were to accept Mrs Peacocke's overall approach my initial reaction would be to regard the entirety of Mr Rayner's costs of the immediate run-up to the trial and its first three weeks as being the result of the funding which Mrs Murphy had received up to that point, and thus to treat them all as attributable to the funded part of the proceedings. But that was not how the Judge proceeded. He remitted the claim to the Master to decide "whether there was a causal link between the funding of Mrs Murphy's case up to [the beginning of the hiatus period] and all or any of Mr Rayner's costs [during that period]" and ordered disclosure for that purpose (see para. 103 of his judgment – para. 47 above); and there is no Respondent's Notice challenging that part of his order. It is not at all clear to me what kind of distinctions he was expecting the Master to draw. The nearest he comes to a criterion is at para. 94 (c) of his judgment (see para. 44 above), where he says that it would "be a question of how long the funding had a continuing effect": that formulation, with respect, seems to me to raise more questions than it answers. I am also not persuaded by the other points that he makes in that paragraph in answer to Mr Gimlette's submissions on this question. Specifically:

(1) At (a) he says that the costs assessment in a case of this kind is bound to be intricate and time-consuming in any event. But in my view the difficulty created by Mrs Peacocke's approach arises from the nature of the exercise rather than simply the time that it might take up.

(2) At (b) he says that the Court sometimes has to perform at least as difficult an exercise when making a third party costs order under section 51 of the 1981 Act. But I do not regard the situations as analogous. The Court has a wide discretion in the exercise of its jurisdiction under section 51 to award costs against third-party funders and is developing its own approaches to the exercise of that discretion: cf. Arkin v Borchard Lines Ltd [2005] EWCA Civ 655, [2005] 1 WLR 3055. I am aware of no case where it has set itself the kind of difficult exercise required by Mrs Peacocke's approach, which is the product of specific statutory provisions.

(3) At (c) he refers to the case where funding has been granted in respect of a particular issue. But as he himself points out, that is a special case, which does not help on the principal exercise.

(4) I find his observation at (d) rather cryptic. But the thrust of it appears to be that once the proceedings were on foot Mr Rayner was in practice obliged to incur the costs of defending it. I rather agree, but that makes it more difficult, rather than easier, to understand what exercise he was asking the Master to perform.

55. Again, although my concerns about the uncertainty which Mrs Peacocke's approach would produce are not derived from the authorities, they do chime with the observations of the Court at para. [16] of the judgment in S v S – see para. 27 above – albeit that it did not have precisely the present point in mind.
56. I acknowledge that in the present case the construction of regulation 5 (4) which I prefer produces a result where Mr Rayner is unable to recover costs which he had to incur as a result of the Commission's funding of Mrs Murphy in the earlier stages of her case. I regret that, and I have already said that I see the force of the fairness argument which appealed to the Judge. But statutes do sometimes adopt rough-and-ready criteria which from time to time fail to produce a wholly just result in a particular case. It is important to appreciate that the present case is unusual: it must be rare for funding to cease, at so critical a stage, in circumstances where – as I am prepared to assume was the case, though the point was not demonstrated to us – the non-funded party had no option but to continue to incur very substantial costs. Typically where public funding ceases part-way through a piece of litigation it will be because it has become clear that the merits are insufficient to justify its continuation; in such a case it will generally not be possible for the previously-funded party to continue to prosecute (or defend) the proceedings, as the case may be, and the non-funded party will not have to incur significant further costs and can look to the Commission for any costs already incurred. There may also be cases where the financial circumstances of a funded party improve, so that funding is no longer justified; in which case the non-funded party is in principle at no worse risk of not recovering his later-incurred costs than is, alas, inherent in the system generally. Against that background, there is nothing surprising in Parliament proceeding on the broad basis that the interests of a non-funded party are sufficiently protected by his having recourse to the Commission for his costs incurred during the period of funding; and that as regards costs incurred thereafter he has to face the same risk as other litigants. That does not make the untypical case any more palatable to the victim, but it does put the claimed arbitrary effect of the temporal approach into perspective.

THE AMENDED RESPONDENT'S NOTICE

57. As indicated above, Christopher Clarke LJ refused Mr Rayner permission to appeal against the Judge's refusal to entertain a challenge to Master Haworth's decision, in July 2011, that Mrs Murphy did not enjoy cost protection in the hiatus period, and directed that if the application was renewed it should be considered at the hearing. As also indicated, there was in fact no renewal request, but it has always been clear that Mrs Peacocke intended to take the point, and if that were the only problem with the application I would have been minded to overlook the failure to lodge a formal request.
58. Having heard submissions from both parties we refused the application in the course of the hearing. My reasons are as follows.
59. Mrs Peacocke accepted that formally any challenge to the 2011 decision should have been made by way of an appeal lodged within 28 days of the Master's decision, and that this was not done. But she submitted that the failure to lodge such an appeal was venial. Mr Rayner was at that time seeking to persuade the Commission retrospectively to revoke the grant of funding; and indeed in September 2011 it decided to do so – though it changed its mind in August 2012 after Mrs Murphy had been given permission to apply for judicial review of the revocation decision. So long as the revocation decision stood the 2011 decision was of no practical significance. And indeed in any event a decision depriving Mrs Murphy (partially) of cost protection was, in itself, a decision in Mr Rayner's favour. It was only when it became clear not only that Mrs Murphy was impecunious but also that the Commission was relying on the 2011 decision as an answer to his claim against it that it became in his interests to challenge it.
60. For myself, I see the force of all that so far as it goes; but the real problem is the way in which the point was raised thereafter. We were told by Mrs Peacocke that Mr Rayner's solicitors intimated in correspondence in February 2013 (presumably at the time of Master Haworth's second decision) that they wished to challenge the 2011 decision. However, no Appellant's Notice was issued at that stage. Nor, most importantly, was the point taken during the hearing before the Judge in October 2013: Mrs Peacocke told us that she raised it with Mr Gimlette but that he pointed out that there was no Appellant's Notice. An Appellant's Notice, including an application for an extension of time, was eventually served, but that was not until after the conclusion of the hearing and the Judge had reserved his decision. Mrs Peacocke lodged written submissions in support of the application for an extension of time, which were answered by Mr Gimlette, who objected that it was made far too late.
61. The Judge dealt with the application at paras. 105-7 of his judgment (pp. 709-710) as follows:

"105. … [I]f my decision on the second issue is correct, there is no need for this application to be pursued. However, since there may be an appeal, I should deal with it in case my decision is wrong.

106. Mr. Gimlette's position, not surprisingly, is that he strongly resists the application, both because he contends that the decision was correct and because the application is made so long out of time, and he has indicated that, if I were minded to allow the application, he would wish to be heard at a further hearing.

107. In these circumstances, I think it best to do no more than indicate a provisional view, as follows:

(a) Although at one time I took a different view, I consider that the costs judge reached the right view on the material before him. Although para. 3(3)(b) refers only to funding being withdrawn by discharging the client's certificate, the position is governed by the opening words in para. 3(3), which provide that costs protection applies only to costs incurred in relation to proceedings which are funded proceedings i.e. where part of the proceedings are funded, only to that part of the proceedings "during which the client received the funded services". This wording is intended to replicate the decision in Burridge and the relevant date is the date when the other party is notified of a change of solicitors. See also Mohammadi. Therefore, in the present case, on the basis of the evidence before the costs judge, 19^th May 2010 was the correct date. Anyhow, his decision was reached over 2 years ago. For both reasons, I would refuse permission to appeal.

(b)-(c) … "

62. I should give the background to what the Judge says at para. 107 (a). Mr Rayner's case depended essentially on the fact that Mrs Murphy's certificate had never been withdrawn – see para. 6 (3) above – reinforced by various things the Commission is said to have said about it remaining "live" throughout the hiatus period. The Judge's point is that that was inconsistent with the ratio of Burridge, which held that the criterion was not dependent on the currency of the certificate but was, rather, the point at which the funded solicitor ceased to act, or at least when that became known to the other party (see para. 32 above).
63. Mr Mansfield submitted that the Judge had not in truth made a decision on this point at all and that all that he had done was to decline to decide on an application which in the context of his primary decision did not need to be dealt with; accordingly there was nothing to appeal against. I do not accept that. It is plain that the Judge refused the application: see the closing words of para. 107 (a). I agree that there may to the purist be an ambiguity about whether his only reason for doing so was that the application was redundant or whether it included also the "provisional" view expressed in para. 107. But the point is of no importance. Having overturned the Judge's decision on the primary issue, we need to decide whether he should have allowed the appeal against the 2011 decision to proceed, and for that purpose we need to consider all the reasons which he gave for not doing so, however he labelled them.
64. At para. 107 (a) the Judge gave two reasons for his "provisional" conclusion – that the challenge to the 2011 decision was bad on the merits and in any event that it was brought too late. In my view he was entitled to treat the delay in itself as a sufficient reason. The sequence of events set out above speaks for itself. The challenge to the 2011 decision was a potentially decisive point, and if Mr Rayner wished to deploy it before the Judge it was incumbent on him to do so in proper form, with Appellant's Notice and skeleton argument, in advance of the hearing. But I am inclined also to agree with him that the 2011 decision was right for the reasons which he succinctly gives.

CONCLUSION

65. For those reasons I would allow the appeal and restore the decision of Master Haworth; and I would refuse permission to the Respondent to appeal against the Judge's refusal to entertain an appeal out of time against the 2011 decision.

Lady Justice Gloster:

66. I have had the advantage of reading the judgments of both Underhill LJ and McCombe LJ in draft. I agree with them that the appeal should be allowed and that the decision of Master Haworth should be restored. I would also refuse permission to the Respondent to appeal against the Judge's refusal to entertain an appeal out of time against the 2011 decision. In so far as there is a difference in approach between that of Underhill LJ and McCombe LJ, I prefer the analysis of Underhill LJ.

Lord Justice McCombe:

67. I am most grateful to Underhill LJ for his comprehensive recitation of the factual background to this case, the statutory provisions and the competing arguments on this appeal. I entirely agree with him that this appeal should be allowed and that it was right to refuse permission to the Respondent to appeal against the Judge's refusal to entertain an appeal out of time against the 2011 decision. However, my reasons for agreeing with the conclusion that the appeal should be allowed are, in part, rather different from those expressed by Underhill LJ.
68. It seems to me that the present case turns upon the construction of a statutory phrase, which, while not entirely happily drafted, has been considered more than once by the courts over the years in the context of a number of different incarnations of the statutory code governing legal aid and in slightly differing factual scenarios. In the provision with which we are concerned in this case the phrase is, "…so much of those costs as is attributable to the part of the proceedings which are funded proceedings": reg. 5(4) of the 2000 Regs. By reg. 2(1) "funded proceedings" are…

"…proceedings…in relation to which the client receives funded services or, as the case may be, that part of the proceedings during which the client receives funded services".

69. While I accept that we are clearly not bound by the decision in Burridge to adopt the construction of these provisions advanced by Mr Mansfield (because the point now in issue was assumed rather than decided in Burridge's case) I am not confident that the same can be said of S v S. In my judgment, the reasoning of Stamp LJ in that case is as applicable to our problem as it was to that before the court in S v S. As Baroness Hale of Richmond put it in R (Smith) Secretary of State for Defence [2011] 1 AC at [135],

"In the words of Sir Frederick Pollock, cited by Lord Denning in Close v Steel Co. of Wales [1962] AC 367 at 388-389, "Judicial authority belongs not to the exact words used in this or that judgment, not even to all the reasons given, but only to the principles accepted and applied as necessary grounds of the decision""

70. It seems to me that the reasoning of the court as to the meaning of the word "attributable" in this series of enactments applies as well to our facts as to those of S v S. To that extent, I consider that we are bound by that case to allow this appeal.
71. Even if I am wrong about that, it seems to me that the reasoning of Stamp LJ in S v S, as to the meaning of the word "attributable" in the context of this legislation (set out in my Lord's judgment at paragraph 27), should be followed here unless it can be shown that either the word needs to be interpreted in an obviously different sense with regard to the different facts of our case or because that reasoning is manifestly inapplicable in our case. I do not think that that is shown. I also find the argument advanced on behalf of Mr Rayner, to the effect that the relevant test is a causative one, difficult to square with the rejection by this court in S v S of the approach of Latey J in that case: see paragraphs 26 and 27 above.
72. As a matter of construction of the language of reg. 5(4), I find the meaning of the word "attributable" decided upon by Stamp LJ to be applicable mutatis mutandis to the facts of that case and to those of the present case. There are differences of fact but, in my judgment, construction of the word in question must be the same in each case.
73. I am also persuaded that the submissions advanced by Mr Mansfield, as recited in paragraph 23 of Underhill LJ's judgment, are correct. I agree with Underhill LJ in his endorsement of those submissions in paragraph 51 above for the reasons there set out.
74. It seems to me that the alternative arguments, attractively presented as they were by Mrs Peacocke, do not undermine the logic underlying the construction of the statutory language adopted by Stamp LJ. Further, Mrs Peacocke's arguments appear to me, with respect, not to be primarily directed to the construction of the statutory language, but rather to potential anomalies of adopting a strictly temporal test to the question of whether the relevant party's costs are attributable to funded proceedings. The temporal test has, however, long been understood to be the applicable one in this context and I am not persuaded that we should differ from it now on the basis of perceived anomaly in some circumstances.
75. So far as application for leave to appeal out of time against the 2011 decision of Master Haworth is concerned, I am quite satisfied that the application had to be refused. I agree with Underhill LJ that the Judge was entitled to refuse the application simply on the basis of the delay in bringing it. I am inclined also to think that the decision was correct in any event.

Note 1    We were told that Mrs Peacocke did not rely on Kadhim before the Judge but argued that Burridge could be distinguished because the legislative scheme had changed (cf. paras. 36 (4) and 38 above); it was the Judge who came up with this line of reasoning, without notice to the Commission. But that is water under the bridge.     [Back]

Note 2    The passage reads: "49. It … will normally be just and equitable that when a costs order is made against a party who has been supported by public funds, the costs covered by the order should, insofar as they cannot be recovered from the funded party, be defrayed out of public funds. 50. … Costs judges should proceed on the premise that it is just and equitable that the Commission should stand behind their "client", by definition under the Regulations the individual who receives funded services, unless they are aware of facts which render that result unjust or inequitable.” But the context of those observations was quite different from that of the present case.     [Back]

Note 3    Mrs Peacocke in her skeleton argument referred to the issue debated in the authorities as to whether there may be exceptions to the general requirement of a causative link – referring to Total Spares and Supplies Ltd v Antares Ltd [2006] EWHC 1537 (Ch) and Jackson v Thakrar [2007] EWHC 626 (QB); but we need not go there.     [Back]

Note 4    We were told that it was, again, the Judge rather than Mrs Peacocke who came up with the suggested analogy with the jurisdiction under section 51 (which he also relies on elsewhere – see para. 44 below); and Mr Mansfield complains that Mr Gimlette had no real opportunity to deal with it. But this too is water under the bridge.    [Back]

Note 5    Very strictly speaking, we are concerned in the present case with the intention not of Parliament but of the Lord Chancellor as the maker of the 2000 Regulations; but he was reproducing the substance of provisions previously contained in primary legislation, and it is more convenient to refer straightforwardly to “Parliament”.    [Back]