England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM CHESTER COUNTY COURT
His Honour Judge Edwards
District Judge Wallace
AND ON APPEAL FROM LIVERPOOL COUNTY COURT
His Honour Judge Marshall Evans QC

		Royal Courts of Justice Strand, London, WC2A 2LL Tuesday 17th July 2001

SUMMARY (This Summary forms no part of the judgment of the court)

These two appeals were concerned with the new arrangements for financing the cost of personal injuries litigation which came into effect last year. Legal aid is now no longer available for most litigation of this type. Instead, it is possible for a claimant to enter into a conditional fee agreement ("CFA") with his solicitor under which his solicitor will be entitled to no fees at all if the claim fails, but an agreed uplift on his fees (a "success fee") if the claim succeeds. "After the event" ("ATE") insurance is also now available, which covers the claimant's liability to pay the defendant's costs if his claim fails, and the claimant is now able to include the cost of the ATE premium in the costs he is entitled to recover from the defendant if his claim succeeds.

Ever since the changes were introduced there has been disagreement between claimants' solicitors and ATE insurers on the one hand and defendants' solicitors and accident liability insurers on the other hand on four main questions. The first of these questions relates to the time at which it is appropriate to enter into a CFA and take out an ATE policy. The second is the reasonableness of the success fees charged, particularly when a claim is quickly resolved without the need for court proceedings. The third is whether claimants are entitled to recover an ATE premium at all in these circumstances. The fourth relates to the reasonableness of the ATE premiums for which claimants are seeking reimbursement by defendants when their claims succeed. The Court of Appeal was told that there are now about 100,000 cases in which the parties, or the lower courts, are waiting for an authoritative decision by the Court of Appeal on these issues.

These two appeals were concerned with claims for personal injuries suffered in road traffic accidents which were settled quite quickly without any need to bring court proceedings. In one case a circuit judge had upheld a decision of a district judge allowing a success fee of 40%. In the other a circuit judge had reduced a claimed success fee of 30% to 20%. The Court of Appeal reduced the allowable success fee to 20% in the first case and dismissed the second appeal. A further issue in the first appeal related to the amount of the ATE premium. The Court of Appeal directed an inquiry by a costs judge on the reasonableness of premiums charged for this type of insurance, and will deliver a separate judgment on this question when it has received his report.

In the present judgment the Court of Appeal now gives the reasons for its decision, announced during the hearing last month, to the effect that ATE premiums are in principle recoverable as part of a claimant's costs even though his claim is quickly resolved without the need for proceedings (see paras 42-55). It also gives its reasons for holding that it is in principle permissible for a claimant to enter a CFA with a success fee and to take out ATE insurance when he first consults his solicitor, and before the solicitor writes a letter of claim and receives the prospective defendant's response (see paras 101-116).

On the reasonableness of success fees the Court of Appeal gives some general guidance in relation to modest and straightforward claims for compensation resulting from road traffic accidents. It suggests that where a CFA is agreed at the outset in such cases, 20% is the maximum uplift that can reasonably be agreed in such a case. It stresses that it will be desirable to review this conclusion once sufficient data becomes available to enable a fully informed assessment of the position to take place (see paras 101-105).

The court also considers that it is open to a solicitor and his client to agree a two-stage success fee at the outset of proceedings. By way of example, an uplift might be agreed at 100%, subject to a reduction to a maximum of 5% should the claim settle before the end of the period fixed by a pre-action protocol. Such an uplift would more nearly reflect the risks of the individual case. The court suggests that once the necessary data becomes available, consideration will need to be given to the question whether the requirement to act reasonably mandates the agreement of a two-stage success fee in a case where a CFA with a success fee is agreed at the outset (see paras 106-115).

INDEX

Part 1	Background	1
Part 2	Developments between 1988 and 2000 in outline	8
Part 3	The early years of ATE insurance	14
Part 4	The legislative framework	24
Part 5	The three main issues	41
Part 6	The jurisdiction issue	42
Part 7	The CFAs in these two cases	57
Part 8	The development of success fees in CFAs	60
Part 9	ATE insurance	65
Part 10	The prematurity issue	80
Part 11	The reasonableness issue	101
Part 12	The Callery appeal: (i) The facts	118
Part 13	The Callery appeal: (ii) The judgment appealed	125
Part 14	The Callery appeal: (iii) Our conclusions	131
Part 15	The Russell appeal: (i) The facts	134
Part 16	The Russell appeal: (ii) The decision of the Judge
	and our conclusion	138

LORD WOOLF C.J.: This is the judgment of the court to which all the members of the Court have contributed.

1. Background

1. The judgment relates to two appeals which are confined to issues of costs. The Claimant in one appeal is Stephen Callery and we will refer to the appeal in his case as the "Callery appeal". The Claimant in the other appeal is Gregory Charles Russell and we will refer to the appeal in his case as the "Russell appeal". Both appeals are by the Defendants and arise out of modest claims for personal injuries due to minor traffic accidents. The amounts of costs in issue are small. However, the appeals are of very considerable importance to those members of the public who are involved in civil disputes, the legal profession and the insurance industry.
2. The appeals deal with two new powers of the courts. The first is the power to make an award of costs against a party to legal proceedings ("the paying party") in favour of a party ("the receiving party") which includes a sum paid by the receiving party for an insurance premium obtained to cover costs which the receiving party would have been liable to pay had he lost. This type of insurance is called after-the-event insurance ("ATE"). The second is the power of the court to make a costs order against the paying party which includes the amount of a success fee payable by the receiving party under a conditional fee agreement ("CFA").
3. ATE is to be distinguished from "before-the-event" ("BTE") legal expenses insurance (often provided as part of a broad range of indemnity e.g. for house owners) which is taken out prior to the event which gives rise to a possible claim.
4. Because of the importance of the issues, we agreed to hear representations from a number of bodies who have a direct interest in the matters we consider in this judgment. We received written and oral submissions from the Law Society, the Association of Personal Injury Lawyers ("APIL"), the Association of British Insurers ("ABI"), the Forum of Insurance Lawyers ("FOIL") and a Group (the "ATE Grouping") which consists of a large number of firms, organisations and insurers engaged in ATE. We received written representations only from the Motor Accident Solicitors Society ("MASS"). We were grateful for the assistance of these bodies. We also had the great advantage of having the Chief Costs Judge as our assessor. His advice was extremely helpful. Following the hearing, we received further written representations from the parties, which we have taken into account insofar as it was appropriate to do so.
5. Until Parliament intervened by legislation, it was always considered to be contrary to public policy, and therefore unlawful, in this jurisdiction for the financial reward which a lawyer received for his services in connection with litigation to vary depending upon the outcome of the litigation. The historic position and its development is lucidly described by Schiemann LJ in his judgment in Awwad v Geraghty & Co [2000] 3 WLR 1041. Schiemann LJ points out (at p.1044H) that there are three varieties of fee which a lawyer and his client may agree should be paid if, but only if, the claim succeeds:

i) a share of the recovery that the client makes in the litigation.

ii) the normal fee together with an uplift.

iii) the normal fee without any uplift.

6. The first type of fee is known as "a contingency fee". While it is common in the USA it is not lawful in this country and we are not concerned with it. It is now lawful to agree that the second and third types of fee shall be payable conditional upon success, subject to compliance with the requirements of the relevant statutory instrument. Each type can, where such an agreement is made, be described as "a conditional fee". Under modern CPR nomenclature an agreement for the second type of fee (but not the third) may also be described as "a funding arrangement" (see CPR 43.2(1)(k)(i)) . These appeals are concerned with this type of CFA.
7. The introduction of the legislation which made conditional fees lawful was motivated primarily by two problems in relation to the provision of legal aid for civil litigation. The first was that progressively fewer members of the public were eligible for legal aid to bring civil proceedings. It was thought that the introduction of CFAs would have the effect of enabling those who could not afford to bring proceedings without the benefit of legal aid to do so. The second problem was that the cost of providing legal aid was growing year on year. Accordingly the Government decided to reduce the areas of litigation which were funded by legal aid. It was considered that this would not reduce access to justice since those affected could bring proceedings using CFAs. The reason for reducing the areas of litigation eligible for legal aid was not, it was said, to reduce expenditure overall but rather to use the funds saved thereby to meet the need for publicly funded legal services to be provided in a different manner.

2. Developments between 1988 and 2000 in outline

8. It was in 1988 that the Report of the Review Body on Civil Justice (Cm 394, paras 384-389) opened up the desirability of re-examining the prohibition on what it described as "contingency fees and other forms of incentive scheme". In 1990 Parliament legislated for the first time to permit CFAs in certain narrowly prescribed circumstances. Section 58 of the Courts and Legal Services Act 1990 ("the 1990 Act") provided the enabling machinery. The Lord Chancellor's new enabling powers were exercised for the first time with effect from 5th July 1995 when the Conditional Fee Agreements Regulations 1995 and the Conditional Fee Agreements Order 1995 both came into effect. The former prescribed the form which a CFA had to take if it was to be legally enforceable, whilst the latter limited its availability to the six types of proceedings mentioned in Article 2(1) of the Order. These included, by Article 2(1)(a), personal injury proceedings. Article 3 of the Order prescribed the maximum permitted percentage by which fees might be increased (in the event of success) to be 100%. Schedule 17 of the 1990 Act added a new subsection 15(4A) to the Legal Aid Act 1988 to the effect that a person might not be refused representation under that Act for the purposes of any proceedings on the ground that it would be more appropriate for him to enter into a CFA.
9. In 1998, a new Conditional Fee Agreements Order revoked the 1995 Order. CFAs were now to be permissible in all proceedings (other than those specified in section 58(10) of the 1990 Act (see also section 58(1)(a)), including those concluded without the commencement of court proceedings. Article 4 of the new order retained 100% as the maximum permitted percentage increase.
10. The Access to Justice Act 1999 ("the 1999 Act") introduced major changes to the funding of civil litigation. Part I of the Act created a new Legal Services Commission (in place of the Legal Aid Board) with power to determine which types of litigation should qualify for public funding. (The history of the many relevant changes introduced by the Legal Aid Board before the 1999 Act came into effect on 1st April 2000 was described by the Divisional Court in R v Legal Aid Board ex parte Duncan and Mackintosh (unreported, COT 16th February 2000)). For present purposes, it is only necessary to record that from 1st April 2000 onwards, what used to be described as legal aid was no longer to be available for personal injury cases (although clinical negligence cases were excluded from this blanket ban).
11. The need to put alternative funding arrangements in place was recognised by Parliament in Part II of the 1999 Act, entitled "Other Funding of Legal Services". In this appeal, we are concerned only with the first five sections contained in this part (sections 27-31); the other three sections relate to legal aid in Scotland.
12. Sections 27 and 28 of the 1999 Act are concerned with the circumstances in which CFAs, and what are described as "litigation funding agreements", are to be enforceable. On this appeal we are not concerned directly with the latter, for which section 28 (which introduces a new section 58B into the 1990 Act) makes provision. So far as CFAs are concerned, the technique selected by Parliament was to substitute two new sections, 58 and 58A, in place of section 58 of the 1990 Act as originally enacted. We will refer to the detailed provisions of these sections later in this judgment.
13. The next two sections, along with section 31, are headed "costs" and deal with different matters. Section 29 is headed "Recovery of insurance premiums by way of costs" (for its text, see para 44 below). Section 30 is headed "Recovery where body undertakes to meet costs liabilities". Their evident purpose is to enlarge the scope of the items of costs which a successful party to proceedings (to use a neutral word) may recover from the paying party.

3. The early years of ATE insurance

14. In order to understand the purpose of section 29, it is necessary to consider its historical development. We have been greatly assisted in this regard by the submissions we have received not only from the parties and the Law Society, but also from the lawyers' groups and insurance interests that made written and oral representations to us.
15. The introduction of CFAs in 1995 still left a litigant at risk of having to pay the other side's costs. The Law Society therefore developed the ATE policy, with the help of insurance brokers, as a new form of insurance cover. Since about that time there have also been forms of ATE insurance which provide cover against other risks, but we are not concerned with such cover, whatever form it takes, in this judgment.
16. Written evidence was given by Mr Christopher Ward, the managing director of the specialist legal expenses underwriting agency which has run the Law Society's Accident Line Protect conditional fee insurance scheme since its inception in 1995. During the last six years, this agency has issued over 85,000 ATE policies. The insurance was always issued in conjunction with personal injury claims conducted under a CFA.
17. This agency is not itself the risk carrier in insurance terms. It manages the insurance on behalf of the underwriter who carries the risk. For the first five years, this was an American company. The Law Society told us that the ATE premiums for this type of cover were originally very modest, because the success rate of personal injury litigation was thought to be so high. Adverse claims experience, however, drove the premium up sharply, and in 2000 the original underwriters withdrew from the market after suffering major losses.
18. We have been shown the different stages of the public consultation process which took place between March 1998 and February 2000. This process started with the Government professing its keenness to encourage the wider use of legal expenses insurance. Premiums for BTE cover were now being paid by over 17 million people at a trivial annual cost to themselves, and the Government said it wished to encourage a varied market for insurance products which would enable people to go to law if the need arose.
19. None of the respondents to its first consultation paper appears to have suggested that an ATE premium should be recoverable from the other side in the event of success, but the Government eventually decided to introduce this provision by primary legislation. The debates in each House of Parliament in the 1999 Access to Justice Bill on this proposal, which we have seen, do not assist us in the task we have before us.
20. In September 1999 the Government initiated a further round of public consultation to address some of the firmer details of its proposals. We have read its response, published in February 2000 (and in particular paras 85 and 76(bis) of that response) which shows that it was willing to monitor the development of the insurance market and the level of premiums for legal expenses insurance quite closely. In this response it evinced for the first time its approval of a practice whereby claimants entered into an ATE insurance policy at an early stage, before embarking on the procedures required by a pre-action protocol.
21. On 3rd July 2000, the Civil Procedure (Amendment No 3) Rules 2000 came into force, supplemented by a new Practice Direction on Costs. We will discuss the salient features of these two documents later in this judgment. It is sufficient for present purposes to note that the additional rules numbered 44.3A and 44.3B made new provision for costs orders relating to funding arrangements and the limits on recovery under funding arrangements.
22. A novel procedure was introduced by a new rule 44.12A (for its text, see para 30 below). It provided for the initiation and scope of "costs-only proceedings" in those cases in which the parties had made a written agreement settling all the issues in dispute between them, including the incidence of liability for costs, without the need to initiate court proceedings, and when all that remained in issue was the amount of those costs. The relevant Practice Direction made it clear that on an assessment of costs between the parties, the court had power to reduce any ATE premium it considered unreasonable. It also identified some of the factors it might take into account when determining what was reasonable in this context (CPR PD44 para 11.10: see para 33 below).
23. This then, was the regulatory and legislative backcloth to the revival of the ATE market following the enactment of the 1999 Act.

4. The legislative framework

24. We will now turn to consider in detail the new framework of primary and secondary legislation, buttressed by practice directions and protocols, which marked the advent last year of the new arrangements whose effect we have to consider on these appeals.
25. The starting point is the effect of the changes the 1999 Act made to the conditions which must be fulfilled before entering into a CFA. The material changes are contained in:

(i) Section 27 of the 1999 Act, which inserted the substitute section 58 (referred to in (ii) and (iii) ) and a new section 58A into the 1990 Act.

(ii) Section 58(1), which requires a CFA to satisfy all the conditions of that section for it to be enforceable.

(iii) Section 58(3) and (4), which set out the conditions applicable to every CFA with a success fee. These conditions require, among other things, that such a CFA must be in writing, must comply with any requirement prescribed by the Lord Chancellor and must specify the success fee.

(iv) Section 58A(6), which states that:

"A costs order made in any proceedings may, subject in the case of court proceedings to the rules of court, include provision requiring the payment of any fees payable under a conditional fee agreement which provides for a success fee."

(v) Section 58A(7), which includes an express power for the rules of court to make provision for the assessment of any costs, including fees payable under a CFA.

26. Next came the Conditional Fee Agreements Regulations 2000. They provide considerable protection for a litigant entering into a CFA, by prescribing additional requirements with which CFAs must comply and stating the consequences if they do not do so. Regulations 3 and 4 provide:

"3(1) A conditional fee agreement which provides for a success fee-

(a) must briefly specify the reasons for setting the percentage increase at the level stated in the agreement, and

(b) must specify how much of the percentage increase, if any, relates to the cost to the legal representative of the postponement of the payment of his fees and expenses.

(2) If the agreement relates to court proceedings, it must provide that where the percentage increase becomes payable as a result of those proceedings, then—

(a) if—

(i) any fees subject to the increase are assessed, and

(ii) the legal representative or the client is required by the court to disclose to the court or any other person the reasons for setting the percentage increase at the level stated in the agreement,

he may do so,

(b) if—

(i) any such fees are assessed, and

(ii) any amount in respect of the percentage increase is disallowed on the assessment on the ground that the level at which the increase was set was unreasonable in view of facts which were or should have been known to the legal representative at the time it was set,

that amount ceases to be payable under the agreement, unless the court is satisfied that it should continue to be so payable, and

(c) if—

(i) sub-paragraph (b) does not apply, and

(ii) the legal representative agrees with any person liable as a result of the proceedings to pay fees subject to the percentage increase that a lower amount than the amount payable in accordance with the conditional fee agreement is to be paid instead,

the amount payable under the conditional fee agreement in respect of those fees shall be reduced accordingly, unless the court is satisfied that the full amount should continue to be payable under it.

(3) In this regulation "percentage increase" means the percentage by which the amount of the fees which would be payable if the agreement were not a conditional fee agreement is to be increased under the agreement.

4(1) Before a conditional fee agreement is made the legal representative must—

(a) inform the client about the following matters, and

(b) if the client requires any further explanation, advice or other information about any of those matters, provide such further explanation, advice or other information about them as the client may reasonably require.

(2) Those matters are—

(a) the circumstances in which the client may be liable to pay the costs of the legal representative in accordance with the agreement,

(b) the circumstances in which the client may seek assessment of the fees and expenses of the legal representative and the procedure for doing so,

(c) whether the legal representative considers that the client's risk of incurring liability for costs in respect of the proceedings to which agreement relates is insured against under an existing contract of insurance,

(d) whether other methods of financing those costs are available, and, if so, how they apply to the client and the proceedings in question,

(e) whether the legal representative considers that any particular method or methods of financing any or all of those costs is appropriate and, if he considers that a contract of insurance is appropriate or recommends a particular such contract

(i) his reasons for doing so, and

(ii) whether he has an interest in doing so.

(3) Before a conditional fee agreement is made the legal representative must explain its effect to the client."

27. The CPR contain specific provisions as to success fees and premiums payable for ATE insurance. Rule 43.2 has the following relevant definitions which apply to Parts 44 – 48, "unless the context otherwise requires":

"(a) 'Costs' include any additional liability incurred under a funding arrangement…

(k) 'funding arrangement' means an arrangement where a person has -

(i) entered into a conditional fee agreement which provides for a success fee within the meaning of section 58(2) of the 1990 Act;

(ii) taken out a policy to which section 29 of the 1999 Act (recovery of insurance premiums by way of costs) applies; or …

(l) 'percentage increase' means the percentage by which the amount of a legal representative's fee can be increased in accordance with a conditional fee agreement which provides for a success fee;

(m) 'insurance premium' means a sum of money paid or payable for insurance against the risk of incurring a cost liability and the proceedings, taken out after the event that is the subject matter of the claim;

(o) 'additional liability' means the percentage increase, the insurance premium, or …, as the case may be."

28. Rule 44.3A contains provisions as to costs orders relating to funding arrangements. It requires the court not to assess any additional liability until the conclusion of the proceedings, or part of the proceedings, to which the funding arrangement relates, and gives the court power to make the appropriate order summarily or on a detailed assessment.
29. Rule 44.3B places limits on what can be recovered under funding arrangements. Rule 44.3B(1) provides that a party may not recover the following as an additional liability (although this does not apply on an assessment of a solicitor's bill to his client):

"(a) any proportion of the percentage increase relating to the cost to the legal representative of the postponement of the payment of his fees and expenses;

(d) any percentage increase where a party has failed to comply with –

(i) a requirement in the costs practice direction; or

(ii) a court order,

to disclose in any assessment proceedings the reasons for setting the percentage increase at the level stated in the conditional fee agreement."

30. Rule 44.12A introduces a new concept which we will call "costs-only proceedings". It is in these terms:

"(1) This rule sets out a procedure which may be followed where-

(a) the parties to a dispute have reached an agreement on all issues (including which party is to pay the costs) which is made or confirmed in writing; but

(b) they have failed to agree the amount of those costs; and

(c) no proceedings have been started.

(2) Either party to the agreement may start proceedings under this rule by issuing a claim form in accordance with Part 8.

(3) The claim form must contain or be accompanied by the agreement or confirmation.

(4) In proceedings to which this rule applies the court-

(a) may

(i) make an order for costs; or

(ii) dismiss the claim; and

(b) must dismiss the claim if it is opposed."

31. Rule 44.15 requires information as to funding arrangements to be provided. In particular, Rule 44.15(1) provides that a party seeking to recover an additional liability must provide to the court and to the other parties information about the funding arrangement, as required by a rule, practice direction or court order. If the information changes, then a notice of change has to be filed and served on the other parties.
32. There is a Practice Direction relating to funding arrangements. Section 9 of the Direction supplements Rule 44.3A. It states categorically that:

" 9.1 Under an order for payment of 'costs' the costs payable will include an additional liability incurred under a funding arrangement."

33. Other relevant provisions of the Practice Direction are in these terms:

"11.4 Where a party has entered into a funding arrangement the costs claim may, subject to rule 44.3B include an additional liability.

11.5 In deciding whether the costs claimed are reasonable and (on a standard basis assessment) proportionate, the court will consider the amount of any additional liability separately from the base costs.

11.7 Subject to paragraph 17.8(2), when the court is considering the factors to be taken into account in assessing an additional liability, it will have regard to the facts and circumstances as they reasonably appeared to the solicitor or counsel when the funding arrangement was entered into and at the time of any variation of the arrangement.

11.8(1) In deciding whether a percentage increase is reasonable relevant factors to be taken into account may include –

(a) the risk that the circumstances in which the costs, fees or expenses would be payable might or might not occur;

(b) the legal representatives liability for any disbursements,

(c) what other methods of financing costs were available to the receiving party.

(2) The court has power, when considering whether a percentage increase is reasonable, to allow different percentages for different items of costs or for differing periods during which the costs were incurred.

11.9 A percentage increase will not be reduced simply on the grounds that, when added to base costs which are reasonable and (where relevant) proportionate, the total appears disproportionate.

11.10 In deciding whether the cost of insurance cover is reasonable, relevant factors to be taken into account include:

(1) where the insurance cover is not purchased in support of a conditional fee agreement with a success fee, how its cost compares with the likely costs of funding the case with a conditional fee agreement with a success fee and supporting insurance cover;

(2) the level and extent of the cover provided;

(3) the availability of any pre-existing insurance cover;

(4) whether any part of the premium would be rebated in the event of early settlement;

(5) the amount of commission payable to the receiving party or his legal representative or other agents."

34. Subject to these provisions, the approach of the courts in relation to funding arrangements should be the same as on any other assessment of costs of proceedings.
35. We should finally refer to the relevant protocols and the conditions which can appear in CFAs. The parties are required to comply with the objectives and terms of relevant protocols. Whether or not they have done so will be relevant when the court makes costs orders. This is made clear by the Protocols Practice Direction which sets out (at para 1.4) the objective of pre-action protocols as being

a) to encourage the exchange of early and full information about the prospective legal claim;

b) to enable parties to avoid litigation by agreeing a settlement of the claim before the commencement of proceedings; and

c) to support efficient management of proceedings where litigation cannot be avoided.

36. Paragraph 2.1 of the Practice Direction states that the compliance and non-compliance with an applicable protocol should be taken into account by the court when making orders for costs.
37. Paragraph 2.2 of the Practice Direction states that the court will expect the parties to have complied in substance with the terms of an approved protocol.
38. Paragraph 4 of the Practice Direction provides that where there is no approved protocol, the parties are expected to act reasonably by exchanging information and documents and generally trying to avoid proceedings.
39. The protocol which relates to personal injury claims ("Pre-action Protocol for Personal Injury Claims") applies to all claims which include a claim for personal injuries. It is primarily designed for those road traffic, tripping and slipping and accident at work cases allocated to the fast track which could include an element of personal injury with a value of less than £15,000. The protocol does, however, say that the approach advocated by the protocol is equally appropriate to some higher value claims.
40. The protocol is designed to enable the parties to settle a dispute without litigation or, failing settlement, to be in a position where they can dispose efficiently of any litigation which becomes necessary. The protocol gives a timetable for the exchange of information within 3 months. It is anticipated that as part of, or on completion of, the protocol process the parties will, if necessary, make formal offers to settle in order to benefit from the protection as to costs which such offers provide.

5. The three main issues

41. We have identified three main issues on these appeals:

a) Whether an ATE premium can be recovered in costs only proceedings ("the jurisdiction issue": see paras 42-55 below);

b) The stage of a dispute at which it is appropriate to enter into (a) a CFA and (b) an ATE policy ("the prematurity issue": see paras 80-100 below);

c) The reasonableness of the Claimants' (a) percentage uplift and (b) ATE premium ("the reasonableness issue": see paras 101-116 below).

6. The jurisdiction issue

42. This issue (which only affects the Callery appeal) differs from the other issues as it involves a pure point of law. Due to the uncertainty created by this point, we announced our decision on this issue during the course of the hearing. This part of our judgment sets out our reasons for that decision.
43. The Callery claim was settled on terms contained in a letter dated 7 August 2000 from Callery's solicitor. The terms were that the appellant would pay £1,500 by way of damages, together with the respondent's "reasonable costs and disbursements". This left outstanding the issue of what was payable by way of "reasonable costs and disbursements".
44. On 12th September 2000 the Respondent commenced costs-only proceedings pursuant to CPR 44.12A. No challenge is made by the Appellant to the Respondent's entitlement to make a claim for a costs order under that rule. It is submitted, however, that a claimant who avails himself of this procedure has no right to include the premium paid for ATE insurance in the costs claimed. This argument is based on the terms of section 29 of the Access to Justice Act 1999 headed "Recovery of insurance premiums by way of costs". The section states:

"Where in any proceedings a costs order is made in favour of any party who has taken out an insurance policy against the risk of incurring a liability in those proceedings, the costs payable to him may, subject in the case of court proceedings to rules of court, include costs in respect of the premium of the policy."

45. The Appellant's argument, as advanced by Mr Birts QC, can be summarised as follows:

(i) Prior to section 29 of the Access to Justice Act 1999 there was no right to recover by way of costs a premium paid for ATE insurance.

(ii) It follows that unless a Claimant can bring himself within the wording of section 29, he cannot recover the premium.

(iii) Section 29 only permits recovery of an ATE insurance premium where:

(a) there are or have been proceedings in which a costs order has been made in the Claimant's favour and

(b) those proceedings are proceedings in which the Claimant has taken out an insurance policy against the risk of incurring liability for costs in those proceedings.

(iv) The only proceedings that the Claimant has commenced are the "costs only proceedings".

(v) The Claimant has not taken out an insurance policy against the risk of incurring liability in those (i.e. the costs-only) proceedings.

(vi) It follows that the costs order made by the Court in the costs-only proceedings cannot include the ATE insurance premium.

46. Mr Birts accepted that where proceedings for substantive relief have been commenced, a costs order in those proceedings can include a premium in respect of ATE insurance against the risk of incurring liability in those proceedings, albeit that such insurance was taken out before such proceedings were commenced. He submitted, however, that the only way in which the ATE insurance premium can be recovered is by first commencing proceedings claiming substantive relief and then claiming costs in relation to those proceedings.
47. This submission is unattractive on its face for, if it is correct, those who have taken out ATE insurance will be disinclined to settle their claims before substantive proceedings have been commenced. Mr Birts submitted, however, that the wording of section 29 reflected a determination by Parliament to approach new funding arrangements cautiously and incrementally, and to discourage claimants from taking out insurance before it was plain that substantive proceedings would have to be commenced and that insurance cover was therefore really necessary.
48. Mr Birts was not able to support this submission by reference to any Parliamentary material admissible as an aid to interpretation under the principle in Pepper v Hart [1993] AC 593.
49. Mr Nice QC advanced an argument on behalf of the Respondent which required the introduction of a number of additional words into section 29 in order to provide clarity to a clause which he submitted was unsatisfactorily drafted. He submitted that section 29 should be read as follows:

"Where in [respect of] any proceedings, [whether commenced or contemplated,] a costs order is made in favour of any party who has taken out an insurance policy against the risk of incurring a liability in [connection with] those proceedings, [whether commenced or contemplated], the costs payable to him may, subject in the case of court proceedings to rules of court, include costs in respect of the premium of the policy."

50. This suggested interpretation involves implying words into section 29 in order to alter its natural meaning. More significantly, in our view, it requires a degree of clairvoyance on the part of the draftsman. The section, as interpreted by Mr Nice, contemplates a costs order being made in respect of 'proceedings' which were only contemplated. But when the 1999 Act was introduced, there was no procedure under which such an order could be made. A costs order could only be made in the action in which substantive relief was claimed. It was to meet this procedural shortcoming that CPR 44.12A was added to the Rules on 14 July 2000, the purpose of which was to enable inevitable disputes as to the amount of costs to be resolved without interfering with the general settlement of a dispute.
51. In addition, the wording of section 29 can be explained much more simply: Parliament was seeking to restrict the recovery of premiums to those relating to ATE, as opposed to BTE, insurance. The wording of the section achieves this end.
52. For these reasons, we are unable to accept the interpretation of section 29 advanced by Mr Nice. Judge Edwards dealt with this issue of interpretation as follows:

"I must look at the Act itself and I think that what this section means is that when a costs order is made in favour of the party who has taken out an insurance policy against the risk of incurring a liability in those proceedings, that must mean that if he has taken out an insurance policy against incurring a liability for costs of the other side, in due course should he be compelled actually to pursue the matter in that way, then when it comes to assessment of costs simply under Part 8 that must be regarded as part and parcel of the contemplated proceedings against which the insurance policy was taken out, and that therefore such a premium is in fact recoverable under the wording of section 29. I think that gives a sensible purposive construction to the section, but I would not presume to try to give a purposive construction to a section if its plain meaning, to my mind, was to the contrary, but I do not believe that the plain meaning is to the contrary. It is clearly to be anticipated that insurance will be taken out before proceedings have started. It is always possible for the matter then to be compromised before the proceedings actually are started, and it would be quite wrong to interpret this section as shutting out a Claimant from recovering that premium in such circumstances."

53. This analysis treats the Part 8 costs-only proceedings as being included within the phrase "in any proceedings" in section 29. It does not, however, provide a satisfactory answer to Mr Birts's point that the only proceedings before the Court at that stage are the Part 8 costs-only proceedings, and that the insurance premium claimed does not relate to an insurance policy taken out against the risk of liability in those costs proceedings.
54. Our conclusions in respect of this issue of interpretation have been assisted by the argument advanced by Mr Drabble QC, on behalf of the Law Society. The conclusions are as follows:

(i) When the 1999 Act was enacted the only circumstances in which a Claimant could obtain a costs order from the Court was by seeking such an order in the substantive action in relation to which the costs had been incurred.

(ii) Where an action is commenced and a costs order is then obtained, the costs awarded will include costs reasonably incurred before the action started, such as costs incurred in complying with a pre-action protocol.

(iii) Section 29 of the Access to Justice Act enables the Claimant to include in such costs the premium for an insurance policy against liability for costs in the substantive proceedings, even where that policy was taken out in contemplation of those proceedings before they were commenced.

(iv) The "proceedings" referred to in section 29 are therefore proceedings advancing a claim for substantive relief.

(v) CPR 44.12A is a new procedure introduced to enable 'pre-action' costs to be recovered where an action has been settled before substantive proceedings have been commenced. One object of this Rule is to facilitate the settlement of proceedings where there is agreement upon all issues save the assessment of the pre-action costs incurred.

(vi) CPR 44.12A states that it sets out a procedure under which the Court can make a costs order where:

"(a) the parties to a dispute have reached an agreement on all issues (including which party is to pay the costs) which is made or confirmed in writing; but

(b) they have failed to agree the amount of those costs";

(vii) The meaning to be accorded to "the costs" and "those costs" in the Rule is 'the costs which would have been recoverable in the proceedings had the proceedings been commenced'. There is no other meaning that can sensibly be given to these words.

(viii) By reason of section 29 of the 1999 Act, such costs may include the costs of an ATE insurance premium taken out in contemplation of the commencement of substantive proceedings.

55. It follows that Judge Edwards had jurisdiction to include the ATE insurance premium in his award of costs by virtue of:

(i) the agreement between the parties that the Appellant would pay the Respondent's reasonable costs and disbursements.

(ii) section 29 of the 1999 Act which permits costs recoverable in proceedings to include an ATE insurance premium.

(iii) CPR 44.12A which provides a simple procedure facilitating the assessment and recovery of the costs that would be recoverable if substantive proceedings were commenced, without the need to initiate such proceedings.

56. Before turning to deal with the remaining issues we propose to refer to some of the evidence that we received.

7. The CFAs in these two cases

57. In both the Callery and the Russell appeals, the claims were in fact settled during the protocol period, but after the CFA had been entered into and ATE insurance obtained. We refer to both CFAs as examples of the terms such agreements are likely to contain. In the Callery case, the CFA was dated 28th April 2000. In the Russell case, it was dated 11th August 2000. In both cases the CFAs referred to the Law Society conditions, and they also both covered claims for damages for personal injuries suffered on the respective dates, any appeal by an opponent, any appeal made against an interim order during the proceedings and any proceedings taken to enforce a judgment order or agreement. The CFAs did not cover counter-claims or an appeal made against a final judgment or order.
58. In the Callery case, the Claimant had to pay disbursements "whatever happens". The Russell agreement provided that "if you lose, you do not pay our charges, but we may require you to pay our disbursements". Both CFAs provided that if the agreement was terminated before the Claimant won or lost, the Claimant would have to pay basic charges.
59. The Law Society conditions gave the Claimants the right to end the agreement at any time. They also allowed the solicitors to terminate the CFA if they believed the Claimant was unlikely to win, in which case only disbursements would be payable, or if the Claimants rejected their opinion about settling with their opponent. In both cases there were some conflicts between the terms of the CFA and the Law Society conditions. These are not significant for present purposes.

8. The development of success fees in CFAs

60. Before turning to the two main issues, we must first say something about the development of success fees in CFAs and ATE insurance policies in recent years. So far as success fees are concerned, the Law Society told us that following the enactment of section 58 of the Courts and Legal Services Act 1990, the then Lord Chancellor had originally proposed that the maximum permitted level of success fee should be 10% or 20%. He was in due course persuaded to change his mind, and in the 1995 Order the maximum permitted fee was prescribed to be 100%.
61. Apparently the Law Society successfully argued that, even if it was accepted that the public interest did not favour facilitating litigation where the prospects of success were less than 50%, it should surely be possible for litigation to be conducted on a CFA when the prospects of success were 50% or a little better. This, it said, required provision for a maximum uplift of 100%, in order to enable lawyers to earn enough in successful cases to compensate for the fact that they would receive no fees at all in unsuccessful cases. The maximum permitted fee level of 100% was retained in both the Conditional Fees Agreements Order 1998 and in Article 4 of the Conditional Fee Agreements Order 2000, which replaced the 1998 order with effect from 1st April 2000.
62. The percentage by which the ordinary fee is enhanced to reflect the fact that payment is conditional upon success is commonly referred to as "the uplift". It has hitherto been generally understood that a CFA must provide for a single uplift that will be payable regardless of the circumstances in which the success is achieved. Thus, in the case of the type of claim with which these appeals are concerned, the uplift has to reflect the fact that (i) it is likely that the claim will be settled swiftly without the need for litigation but (ii) the possibility must exist that the claim will result in contested litigation with no certainty of a successful outcome.
63. It was on this understanding of the need for a single all-purpose uplift that APIL provided a series of worked examples designed to demonstrate that uplifts should be set at a relatively high level from the outset. They took an imaginary cohort of 150 personal injury cases, and postulated that 50 of these would be rejected after the solicitor had devoted time on each which would ordinarily be chargeable at £150. 92 of the remainder would be settled, two abandoned, and six fought out at trial. The tables show variations within this scenario (more of the six cases lost than won, and varying stages at which the claim was settled). It was suggested on these figures that if three of the contested cases in this scenario were won, the "standard" success fee based on a 95% success rate would have to be 28.88% if the lawyer was to receive the same total fees for his work on all these cases as he would have received under the previous regime. If some of the 92 cases settled late, so that more work would be done on these successful cases, the level of success fee needed to "break even" would be reduced to 22.64%. This illustration is necessarily stylised, and the figures do not allow any increment to cover the risk that the particular "book" of the particular lawyer deviates from the overall statistical norm.
64. We were also shown research studies whose authors had endeavoured to create a larger database of information about what is actually happening in the personal injuries market. It was common ground, however, that not enough is yet known about the likely effect of different levels of success fees. APIL, for instance, told us that they:

"… have been concerned about the lack of hard research and knowledge as to success rates. Acquiring such knowledge has proved difficult and we are far from confident that the patterns of success achieved thus far are not distorted by: (1) research into the performance of CFAs run alongside publicly funded cases; (2) defendants and insurers' attitudes to personal injury litigation carried over from a publicly funded era; (3) an initial over-caution on the part of some and over-adventurousness on the part of other personal injury lawyers as they come to terms with CFAs as the dominant funding mechanism."

9. ATE insurance

65. We have already observed (see para 15 above) that ATE insurance can take a number of quite distinct forms. The major distinction is between the ATE insurers who provide litigation costs insurance cover for personal injury related claims directly through solicitors or through claims management companies and those who insure non-personal injury or commercial claims. There is also a distinction between ATE cover that is provided only in respect of the "other side's costs" and that provided "for both sides' costs". ATE cover can also be provided for an individual claimant or in standard form by solicitors under delegated authority. As we stated in paragraph 15, the only form of ATE insurance to which this judgment relates is insurance providing cover against the other side's costs. We do not deal with the question whether ATE cover against other risks falls within section 29.
66. Mr Ward (see para 16 above) explained to us the problems his agency had faced in the past as a result of what it described as "adverse selection". It has always operated the Law Society's scheme on a delegated authority basis. This allows solicitors, within the limits of their authority, to decide whether to offer ATE cover to their clients and to run the personal injury case as they think fit, once they have agreed to take a case under a CFA. This is believed to be attractive to solicitors because it enables them to retain their independence of judgment. It also allows clients to know that the insurer is not controlling their litigation. In addition, delegated authority arrangements minimise cost in a volume market by avoiding individual risk assessment by the insurer, or by some third party on its behalf, in every case.
67. It is hardly surprising that delegated authority arrangements will only work successfully if the solicitor does not "cherry pick" by taking out ATE insurance only in risky cases. It is a basic principle of insurance that the many pay for the few, and we can well understand academic comment to the effect that:

"For a firm to cherry-pick dead certs and run them without paying for insurance must alter the underwriting assumptions of the insurer and increase premiums."

(See the Law Society Gazette publication "Litigation Funding" (Issue 2, 1999) by Professor John Peysner.)

68. In 2000 a Lloyds' syndicate took over the provision of cover under the Accident Line Protect scheme. We were told that this syndicate was only willing to participate in providing this cover because the recoverability of premiums permitted by section 29 of the 1999 Act obviated the incentive to indulge in adverse selection. Mr Ward explained to us how the situation had developed.
69. In order to preserve the principle of the "many paying for the few", his agency had made it obligatory from the outset for solicitors to issue an ATE policy at the same time as a client signed a CFA. In practice, many firms delayed taking either of these steps until after proceedings had been started and after their client had decided not to accept a payment made into court.
70. As a consequence the benefits of Accident Protect cover was then restricted for the most part to those cases where proceedings had not been commenced when the insurance was taken out. Solicitors, however, still delayed taking out the cover until proceedings were about to be commenced. This limited the cover to the riskier cases, and the whole system of delegated authority was therefore put in jeopardy. For this reason the advent of the pre-action protocol was very welcome. It enabled a new principle to be adopted, whereby the ATE insurance had to be taken out in conjunction with a CFA before the letter of claim was despatched.
71. All these changes were predicated by the insurers' desire to spread the risk as much as possible, so that the cost of the cover could be kept within reasonable limits.
72. We were also told that different insurance schemes offered different coverage, and that the coverage in a particular case might depend on the process by which an insurer selected the risks it was willing to cover. Mr Ward's experience, however, had led him to the belief that it was in everybody's interests that cover in appropriate cases should be taken out at the earliest opportunity if the level of premiums was to be affordable. The adoption of this approach has enabled cover to be offered to clients for pre-proceedings disbursements, such as the cost of medical and other expert reports, in the event that there are no subsequent proceedings and the claim has to be abandoned.
73. This evidence was supplemented by the evidence of Mr Christopher Wait, the underwriting director of another company which provides underwriting and claims management functions for certain Lloyd's syndicates in relation to both types of legal expenses insurance. They sell BTE insurance through insurance brokers and ATE insurance mainly through solicitors. Their two directors have a long history of experience in legal expenses underwriting and claims management.
74. Mr Wait told us that most of the ATE insurance schemes available on the market today can only provide insurance if cover is in place before the initial letter of claim is sent. Again, this practice follows the basic insurance principle that "the many pay for the few". He is of the clear opinion that if premiums are not recoverable from the losing party when cases are settled before proceedings are issued, the result will be that many individuals and businesses will find it difficult to seek a legal remedy with effective insurance cover in place, and that their access to justice will be frustrated.
75. Much of his evidence was devoted to explaining the problems that would arise if it were obligatory to delay taking out ATE insurance until after proceedings were issued. Mr Birts made it clear that it was not part of his clients' case to say that the inception of cover had to be delayed in this way.
76. We were provided with worked examples of the manner in which the premium was bound to rise in the event that recovery of the premium was only permissible in those cases in which proceedings were issued. We were told by representatives of claimants that liability insurers had raised what they accepted were legitimate concerns regarding the nature and price of some of the premiums charged for ATE insurance. APIL, a body of about 5,000 members whose interest in personal injury work is largely claimant-orientated, stated that it accepted that if premiums were inflated to include fees to claims managers disguised as commissions for the sale of insurance or to cover costs of advertising their services, they should be recoverable exclusive of such objectionable elements.
77. We also received submissions both orally and in writing from the ATE Grouping, which includes the companies whose directors had provided individual witness statements. These bodies are all involved with the provision of ATE insurance. They are currently involved in a negotiation process sponsored by the Law Society and the ABI in which the parties are seeking to reach an agreed industry standard as to the recoverability of ATE premiums and success fees in the new costs regime.
78. They told us that the ATE market was mainly divided between those who were in the business of providing litigation costs insurance cover for personal injury related claims through solicitors or claims managers and those who provided cover for non-personal injury or commercial claims. On this appeal we are concerned only with the first of these categories. Similarly, we are concerned with ATE cover linked to a CFA, as opposed to such cover provided on a "stand alone" basis. This grouping supported the points made by Mr Wait and Mr Ward in their witness statements.
79. We now turn to deal with the remaining issues.

10. The prematurity issue

80. The information placed before us shows that, in cases such as those with which these appeals are concerned, many solicitors adopt a similar approach to insurers when deciding on what terms to act for clients under CFAs. The uplift in their fees which they demand in respect of this category of business is set at a level designed to produce additional income on the cases which succeed which is adequate to compensate them for the cases which lose and thus earn them no fee. The cases which win have to subsidise the cases which lose. However, while this may be true of many, it is not clear how many. Lawyers are accustomed to assessing a risk without the need to carry out the actuarial calculations which insurers would consider appropriate.
81. A different approach was adopted in the submissions made by FOIL. FOIL did not consider that the need to provide a fund to compensate a Claimant's lawyer for the cases he/she lost on a CFA should play any part in the process of setting a success fee in any given case. It robustly argued that:

"To be consistent with the public interest in reducing the cost of dispute resolution and with the objectives of the CPR, the level of success fee recoverable against the Defendant in any particular case should be assessed by reference to the risks of losing that case."

82. Different circumstances will call for different approaches. Insurers set premium rates designed to balance their books and achieve a reasonable profit. There are some classes of risk which are so remote that a uniform approach may be adopted to the assessment of premium - e.g. household insurance. In other classes, premiums may vary according to specific features of the risk - e.g. age of driver or power of car in the case of motor insurance - but again the premiums will reflect claims experience and be designed to produce a positive return overall. In some cases risks will have peculiar features which require individual assessment, but again the approach is the same: to achieve a balanced book and a reasonable return overall.
83. The solicitor carrying on litigation business on a large scale may have regard to similar considerations. He may seek to ensure that the uplifts agreed result in a reasonable return overall, having regard to his experience of the work done and the likelihood of success or failure of the particular class of litigation. This will not mean that he does not consider the merits of the particular case, where he is aware of facts which call for individual assessment. But there may be categories of claim that have, so far as he is aware, sufficient common characteristics to justify a standard approach to determining uplift.
84. We are in this case concerned with such a category of claim: claims for the consequences of a motor accident where, on the claimant's account of the accident, the solicitor reasonably concludes that the claim has every prospect of an early settlement as to both liability and quantum. At that stage the risk assessment that results in the determination of the uplift is likely to turn, not on peculiar features of the instant case - for there will be none - but on his experience that in a small minority of such cases, when the claim is pursued some unforeseen circumstance results in the ultimate failure or abandonment of the claim.
85. These comments are not, of course, directed to solicitors who choose, as they reasonably may, to defer agreeing a CFA until they know more about the claim than they have learned from the claimant. Nor are they apposite in the case of a solicitor who does not specialise in litigation, but who on occasion conducts a piece of litigation for a client. Such a solicitor is likely to decide on the uplift by asking himself what reward he requires to induce him to take the risk that he may not recover his fees from the case in question.
86. The vital issue in relation to uplift that is raised on these appeals is whether the courts should allow recovery of uplift where this is agreed at the initial stage on the basis described above, or whether it should require all solicitors to defer agreeing uplift until the defendant's response to the claim is known, so that the risk of failure can be assessed on an individual basis. The latter approach would result in a high uplift being justified in a minority of cases, but a small, if any, uplift in the majority where those acting for defendants will make it plain that liability will not be contested.
87. It is the Appellants' argument that the cost of (i) taking out an ATE insurance policy and (ii) the uplift of a success fee should only be recoverable where sufficient information is available to form a reasonable prognosis of what will be the risk involved in a claim. The appellants further argue that a claimant cannot reasonably incur these liabilities until the reaction of the defendant to a claim is known and the merits of any defence raised considered. At that point, so the appellants argue, it will be apparent whether there is a risk that the claim may fail which makes it reasonable to enter into a CFA and take out ATE insurance.
88. If it is reasonable to take these steps, the Appellants argue that the claimant will then be in a position to do what the law requires, namely to assess the appropriate uplift and insurance premium having regard to an informed appraisal of the extent of the risk that the claim may fail.
89. Thus Mr Birts, on behalf of the Appellant in the Callery appeal understandably urges that the time to enter into an ATE insurance policy is at the end of the protocol period, i.e. three months from the notification of the claim; and for his part Mr King, on behalf of the Appellant in the Russell appeal argues that no CFA should be entered into before that time. Entering into funding arrangements at this stage has obvious advantages for a defendant. It gives him the opportunity to settle the case without incurring liability for additional costs. This is a matter of importance, bearing in mind that over 90 percent of cases can be expected to settle and may well settle in the protocol period. In addition, the exchange of information which is central to the protocol will enable the claimants and their lawyers to assess the risk more accurately.
90. The Respondents contend that, in cases such as those before the Court, i.e. modest claims in respect of a road traffic accident, where liability is unlikely to be in issue and the question of damages is unlikely to create complexities, it is reasonable for a claimant to take out ATE insurance and enter into a CFA at the stage that the claimant first instructs a solicitor to pursue his claim. At that stage the claimant will be concerned that, by giving instructions to the solicitor, he is not exposing himself to liability for costs. The solicitor for his part will be anxious to offer the claimant services on terms that, whatever the outcome, he will not find himself liable for costs.
91. In these circumstances, we consider that, from the viewpoint of both the claimant and his solicitor, it will normally be reasonable for a CFA to be concluded and ATE cover taken out on the occasion that the claimant first instructs his solicitors. What we have to decide is whether, having regard to the statutory provisions, (i) the cost of the success fee and (ii) the ATE premium, when incurred at that early stage, can be recovered.
92. In considering this issue we think it right to bear in mind the purposes of the new regime. The first purpose is to facilitate access to justice on the part of those who cannot afford the costs of litigation. The second purpose is to reduce the burden of legal aid in relation to certain categories of case where it was previously available.
93. Including success fees in recoverable costs has the general effect of shifting from the legal aid fund to defendants, or their insurers, the costs incurred by litigants whose claims fail. In the first instance the claimants' solicitors shoulder the risks in relation to these costs, in exchange for uplift. But the fact that the uplift in successful cases is transferred to the unsuccessful defendants results, if one takes a global view, in the burden of unsuccessful claimants' costs being born by unsuccessful defendants.
94. Permitting ATE insurance premiums to be recovered as costs has the effect of shifting to unsuccessful defendants the costs which the insurers will have to pay to successful defendants. Under the old regime successful defendants would not normally recover their costs where claims were legally aided. Thus, in bearing the burden of meeting ATE insurance premiums, defendants in general are paying for cover that will ensure that their costs are paid if they succeed.
95. When seeking to do justice between the parties we have to accept that it is an inevitable consequence of Government policy that unsuccessful defendants should be subjected to an additional costs burden. We also have to bear in mind that the new regime tends to remove from claimants the incentive to control costs. The shelter afforded to the claimant by a CFA and by ATE cover means that he will not be over-concerned at the costs that are being incurred, or even at the size of the ATE premium. In these circumstances, the role of the Court in administering the new regime is particularly important.
96. The scheme of the legislation and the regulations contemplates that both the ATE insurance premium and the amount of uplift will reflect an assessment of the risk that the claim may fail, having regard to the circumstances that are known, or should reasonably be known, to the legal representative at the time that the relevant agreements are entered into. We do not consider, however, that this makes it mandatory for the claimant to delay entering into a CFA or taking out ATE insurance in order to enable his legal representative to acquire a greater knowledge of the circumstances of the case than that provided to him by the claimant.
97. In the type of claim with which these appeals are concerned, the circumstances of the case will often lead the legal representative to assess the risk of failure, not only on the basis of particular features of the case, but on his general experience that claims which appear to have every prospect of success nonetheless occasionally founder as a result of matters which are unforeseen. We consider that this approach is in principle compatible with the legislative scheme.
98. The Appellants contend, however, that it is unjust to saddle defendants with the costs of the ATE insurance premiums and success fees without giving them a fair chance to identify those cases where liability and quantum is not disputed so that success is assured.
99. We see the force of this submission, but we have concluded that, at least in the circumstances of the two appeals, the prejudice to defendants is not as clear as is suggested and that it is outweighed by the legislative policy and by a number of practical considerations. Thus:

i) If the new regime is to achieve its object, the legal costs of claimants whose claims fail should fall to be borne by unsuccessful defendants in the manner described in para 93 above. On these appeals the Court has to decide whether to permit liability for success fees to be apportioned in relatively small amounts among many unsuccessful defendants, or to insist on an approach under which they will be borne in much larger amounts by those unsuccessful defendants who persist in contesting liability.

ii) If the latter alternative is adopted, the defendants who contest liability will not share liability for costs in a manner which is equitable. Where there is a strong defence which it is reasonable to advance, a larger uplift will be appropriate than where a defendant unreasonably persists in contesting liability despite the fact that the defence is weak. Thus the more reasonable the conduct of the defendant, the larger the uplift that he will have to pay if his defence fails.

iii) In relation to claims arising out of road accidents, where defendants will be insured, the same insurers will often be sharing the costs involved, whether in the form of many uniform small uplifts or fewer large uplifts.

iv) So far as insurance premiums are concerned, these will produce cover which benefits the defendants, for they will ensure that costs are awarded against unsuccessful claimants and that such awards are satisfied.

v) Defendant interests, with the assistance of the Court, should be able to restrict uplifts and insurance premiums to amounts which are reasonable having regard to overall requirements of the scheme. In saying this we are contemplating a position where there will be adequate data to enable informed judgment of the amount of uplift and the size of insurance premium that are reasonable in circumstances such as those before the Court. We are well aware that that position has not yet been reached and that, on these appeals, we are faced with doing our best on very sketchy data. We have had particular regard to the fact that the representations and evidence submitted after the hearing have not been tested or analysed in the course of oral argument.

vi) Claimants naturally want to know at the outset that a satisfactory arrangement to cover the costs of litigation has been made which provides sufficient protection for them, no matter what the outcome.

vii) Claimants incur liabilities for costs to their legal advisers as soon as they give them instructions. Once a defendant starts to incur costs in complying with a protocol, the claimant will be exposed to liability for those costs if proceedings are commenced.

viii) Solicitors and claims managers are anxious to be able to offer legal services on terms that the claimant will not be required to pay costs in any circumstances. This will assist access to justice.

ix) There is the overwhelming evidence from those engaged in the provision of ATE insurance that unless the policy is taken out before it is known whether a defendant is going to contest liability, the premium is going to rise substantially. Indeed the evidence suggests that cover may not be available in such circumstances.

100. For these reasons we have concluded that where, at the outset, a reasonable uplift is agreed and ATE insurance at a reasonable premium is taken out, the costs of each are recoverable from the defendant in the event that the claim succeeds, or is settled on terms that the defendant pay the claimant's costs.

11. The reasonableness issue

101. There has not yet been any authoritative guidance from the higher courts as to the level of success fee which would be considered reasonable on an assessment of costs in litigation supported by a CFA. The editors of the current edition of Cook on Costs have endeavoured at pp 467-468 to give some help to the profession, based on the propositions that there will be a single success fee throughout the life of a CFA, and that a solicitor is entitled to cover his/her prospective losses in unsuccessful cases by the success fee income earned in successful cases. The Claimant's solicitors in the Callery appeal, for their part, created an in-house matrix, with points being awarded for different features of a case on a sliding scale. This matrix produced what seemed to us to be a surprisingly high success fee for a fairly straightforward passenger claim, but the matrix provides a useful illustration of what some claimants' solicitors are doing at the moment in the absence of guidance from the higher courts.
102. It should be recognised that any general guidance that we provide is given in the context of the type of claims which are the subject of this appeal, that is to say, modest and straightforward claims for compensation for personal injuries resulting from traffic accidents (see also paragraph 85 above). However, even within this limited area, as APIL recognises,

"the Court is faced with a difficult balancing exercise in setting guidelines for a new regime where there is little experience or published data to rely upon. Allowing success fees to be set too high compared to the risk being run will lead to inflation of fees paid to lawyers by the public who pay insurance premiums. But allowing them to be fixed too low compared to the risk being run will lead to lawyers only being able to take on the most certain cases and a denial of access to justice to some of the most vulnerable people in society".

103. There is some statistical support for a success rate in respect of claims of the type with which we are concerned of up to 98%. However, at this stage of the court's experience of funding arrangements it is not possible to be precise as to what is the correct percentage. We do not consider that it can ever be said that a case is without risk. In this category of litigation, the prospects of some success on liability is increased because of the ability of a court to make a reduced award on account of contributory negligence. It is, however, impossible to foresee all the circumstances in which a straightforward claim can become one with a material degree of risk. In the case of a claim by a passenger, for example, the risk will be small. However, the fact that a Claimant contends that his or her driving was perfect whilst that of the proposed Defendant was atrocious provides no guarantee that, if the case is contested, this is what the Judge will decide. In the circumstances we think that it is reasonable to proceed on the premise that at least 90% of such claims will settle without the need for proceedings, or will succeed after proceedings have been commenced.
104. After careful consideration and having reflected on the reasoning in the judgments below in the two appeals, we have concluded that, where a CFA is agreed at the outset in such cases, 20% is the maximum uplift that can reasonably be agreed. In reaching this conclusion, we have been particularly assisted by the reasoning placed before us by APIL. We wish to emphasise two matters in respect of this conclusion. The first is that it assumes that there is no special feature that raises apprehension that the claim may not prove to be sound. Where there is such a feature, the appropriate uplift will be higher, but it may not be reasonable to attempt to assess that uplift until further information about the defendant's response is to hand.
105. The second matter is that our conclusion is based on very limited data. In particular, it is too early to see what effect the new costs regime is having on the rate of settlements, and this judgment may itself affect that rate. It will be desirable to review our conclusion once sufficient data is available to enable a fully informed assessment of the position.
106. In concluding this portion of our judgment, we wish to draw attention to an alternative type of success fee, which we consider that it is open to the solicitor and the client to agree at the outset of proceedings. We can describe this as a "two-stage" success fee.
107. A success fee can be agreed which assumes the case will not settle, at least until after the end of the protocol period, if at all, but which is subject to a rebate if it does in fact settle before the end of that period. Thus, by way of example, the uplift might be agreed at 100%, subject to a reduction to 5% should the claim settle before the end of the protocol period.
108. The logic behind a two-stage success fee is that, in calculating the success fee, it can properly be assumed that if, notwithstanding the compliance with the protocol, the other party is not prepared to settle, or not prepared to settle upon reasonable terms, there is a serious defence. By the end of the protocol period, both parties should have decided upon their positions. If they are prepared to settle, they should make an offer setting out their position clearly and providing the level of costs protection which they determine is appropriate.
109. A further advantage of a two-stage success fee would be the knowledge that if a claim was not settled, the full success fee would be payable. This knowledge would encourage rigorous consideration of the merits of the claim during the protocol period and therefore accord with the intent of the CPR.
110. If a claim is settled before the end of the protocol period, it would be reasonable that there should still be a success fee payable since:

i) The lawyers are entitled to be compensated for accepting a retainer on a no-fee-no-win basis with the inevitable risk that this involves, however small this risk may appear in many cases.

ii) An appropriate success fee would contribute towards those cases where no fees are payable because they end unsuccessfully.

111. A two-stage success fee would have the advantage that the uplift would more nearly reflect the risks of the individual case, so that where a claimant's solicitor had to pursue legal proceedings, this would be in the knowledge that, although a significant risk of failure existed, the reward of success would be that much the greater. Where, on the other hand, the claim settled as a consequence of an offer by the defendant, he or his insurer would have the satisfaction of knowing that he had ensured that the success fee would be reduced to a modest proportion of the costs.
112. We have considered the risk that a two-stage success fee would encourage claimant's solicitors to take claims past the protocol stage in order to benefit from the higher uplift. Such conduct would, however, be prevented by a defendant who was prepared to settle by making a formal settlement offer, putting the claimant at risk as to costs.
113. In its second written representations the Law Society points out that it considered the possibility of providing for variations in success fees when first developing its standard letters. It decided not to do so, however, because of the complications that these would create for clients.
114. In the Callery appeal, the Appellants provided useful submissions on this subject. They put forward a choice of models. They also dealt with the legality of two-stage success fees. We consider there is no need to consider the question of the legality of a two-stage success fee as we see no difficulty in having a single success fee calculated by reference to an upper level and a reduced level in specified circumstances.
115. A two-stage success fee of the type we propose, agreed at the outset, would be likely to be agreed before the merits of the individual claim were apparent. Thus, the uplift would be unlikely to reflect precisely the likelihood of failure of any individual claim that did not settle. The determination of the reasonable figures for the full uplift and the rebated uplift would have to be based on overall claims experience, with the proportion of contested cases which succeed, and the costs earned from such cases, being particularly significant. While the exercise involved in determining a reasonable two-stage fee would be more complex, we suggest that, once the necessary data is available, consideration will need to be given to the question whether, where fees are agreed at the outset, the requirement to act reasonably mandates the agreement of a two-stage success fee.
116. Whilst as a result of the evidence put before us we have felt able to form an assessment as to the reasonableness of success fees, we do not feel able to form any conclusion as to the reasonableness of premiums charged for ATE insurance. For that reason, we have directed an enquiry before Master O'Hare, Costs Judge, and we will provide a separate judgment on this question when we have received his report.
117. We now turn to the facts of the two appeals.

12. The Callery appeal: (i) The facts

118. On 29 January 2001, His Hon. Judge Edwards at the Chester County Court dismissed an appeal by the Defendant against an order of District Judge Wallace made in the Macclesfield County Court on 7 November 2000. The District Judge gave his judgment in costs only proceedings instituted by the Claimant.
119. On the 2 April 2000, the Claimant had been a passenger in a vehicle which was struck side-on by a vehicle driven by the Defendant. He consulted Messrs Amelans and instructed them to claim damages under a conditional fee agreement. A success fee of 60% was agreed. Amelans deal with a large number of similar cases. In calculating the success fee, Amelans had included 20% for the delays which they were likely to incur in recovering their fees. This element was appropriately conceded before the District Judge not to be recoverable (see Rule 44.3B(1), for which see para 29 above).
120. On 4 May 2000, an ATE insurance premium of £350 plus insurance premium tax was paid. On the same day, Amelans sent a letter before action to the Defendant personally, claiming damages and asking him to pass a copy of the letter to his insurers.
121. On 19 May 2000, the CGU Insurance Company responded to Amelans in the following terms:

"We are able to admit liability as to negligence but not as to causation… please note that we would have no objection to yourselves instructing one of three proposed medical experts. We look forward to receiving confirmation as to whom has been appointed and request sight of the instruction letter sent to them."

122. A medical expert was agreed and a report obtained in July. On 12 July Messrs Amelans, on behalf of the Claimant, made a formal offer of £3,010, together with costs and disbursements, in full and final settlement. On 24 July, a counter offer of £1,200 was made by the Defendant's insurers. On 7 August, the Claimant's solicitors wrote to the Defendant's insurers confirming an agreed settlement of £1,500 plus payment of reasonable costs and disbursements. Subsequently there was a claim for costs of £4,709.35 (which included the 60% success fee and the insurance premium). The Defendant's insurers suggested that the amount claimed was unreasonable and offered a total of £1,877.87.
123. Costs-only proceedings pursuant to CPR 44.12A were issued. The sum of £4,709.35 was claimed.
124. The District Judge summarily assessed the costs at £1,008 +VAT and the disbursements at £617.50, making a total of £1,940.83. He accepted that the insurance premium was payable and that a success fee of 40% was reasonable.

13. The Callery appeal: (ii) The judgment appealed

125. We have already seen (see para 52 above) that on the appeal to HHJ Edwards, the Judge rejected the argument of the Defendant under section 29 of the 1999 Act. The Judge also considered that it was entirely reasonable to have taken out insurance at a very early stage and before it was known whether the matter would be contested because:

i) This would lead to substantial reduction in the costs of the insurance.

ii) There was a potential argument that there could have been a tacit agreement relating to the costs of the pre-action protocol itself. It was conceivable that insurance cover could be needed against these costs. As soon as the claimant needed to be covered he had to take out insurance and needed to notify the other side to ensure that the premium was recoverable.

iii) He could see nothing wrong with insurance being taken out before there was an exchange of letters.

126. The Judge accepted that the Appellant's insurers' letter of 19 May 2000 was a denial of liability, on the ground that a breach of duty alone is not sufficient to establish liability for negligence, and the consequences of sending out a letter that had not been checked by solicitors fell upon the insurers. In view of this finding, the Judge felt the Appellant could not complain. He regarded the letter of the 19 May as validating the decision to make a funding arrangement. He did, however, suggest that in the future, it might be more prudent to wait for the result of the initial exchange of letters. If there was a denial of liability or no reply, then insurance ought to be taken out.
127. As to the reasonableness of the success fee, the Judge noted that the statutory regime clearly envisaged a connection between the size of the fee and the risk of failure presented by a particular case. A further factor could be the need to incur substantial costs up front in an appropriate case, such as a medical negligence claim. He expressed substantial agreement with the analysis contained in Cook On Costs. His conclusions can be summarised as follows:

i) Assuming the chances of success in a given case are 50%, the solicitors will wish to ensure that they recover sufficient costs from a defendant in the event of success to fund another "50% case" which fails.

ii) Accordingly the statutory maximum 100% uplift was appropriate. It followed from this that a solicitor ought not to enter into conditional fee agreements where he considered that the prospect of success was less than 50%.

iii) He had reservations as to whether there could be a personal injury case where there was no risk at all. Even where there was an admission of liability at the outset, he thought that this could subsequently be revoked or put in a different context. There were also the procedural dangers in connection with a Part 36 Offer.

iv) He accepted the Claimant's contentions that there was no case which was entirely risk free and that a Judge should only interfere with the assessment of a District Judge if this was grossly out of place.

128. The Judge concluded that the Respondent had a 75% chance of success. His reasoning for coming to that figure appears to be the underlying litigation risk of 10%, on top of which he made a small additional allowance. Using the Cook approach, the Judge found he would have arrived at a success fee of 33¹/₃% taking into account that the vast majority of costs would be incurred against the background of a denial of liability until a settlement was reached. Having regard to these conclusions, the District Judge had not been manifestly wrong and the appeal failed.
129. The Judge, in addition, made the following points which are worth repeating:

i) Parties should try to avoid litigating costs which require detailed assessment. Where a case was settled, claimants should limit their costs to a reasonable sum and defendants should not try to insist on unjustifiably low amounts.

ii) Claimants in these types of personal injury cases who claimed unreasonable costs placed a great strain on the resources of the courts as this led to a series of detailed assessments. In this connection, the Judge noted that the court below had approximately halved the figure claimed by the Claimant's solicitors already.

iii) As to insurance premiums, the Judge noted that it would be wiser to await the results of the first round of correspondence in a matter before taking out a policy. As this was a discretionary area, the Judge could not guarantee that the premium for every policy taken out prior to a denial of liability would be recoverable.

iv) The CFA was made at a time when the risk was first assessed. This should have an influence on what was recoverable at the end of the proceedings. The Judge did, however, acknowledge that an alteration of the success fee might be warranted after certain stages of the proceedings.

130. In view of his conclusions, he dismissed the appeal, but he indicated that he would have given permission to appeal to the Court of Appeal if he had had power to do so.

14. The Callery appeal: (iii) Our conclusions

131. We have already made it clear that we agree with the Judge's conclusions as to the issue of jurisdiction. We also agree with a substantial amount of the Judge's reasoning and his general remarks. In particular, we agree that on a costs assessment, the assessing Judge has a broad discretion with which an appeal court should only interfere if it is satisfied that the Judge is clearly wrong. However, we would make the following comments in relation to the Judge's approach:

i) While the letter of 19 May is technically a denial of liability, whether this is the case is a red herring in relation to the present appeal. In view of that letter, the Claimant's solicitors would not be justified in continuing to investigate and prepare the case on liability. Their sole concern after the letter was to establish the extent of the damages. Therefore all that was required of the Claimant's solicitors was that they should proceed in accordance with the protocol to obtain a medical report and investigate the question of damages. From a practical point of view, this was, as Mr Birts contends on behalf of the Appellant, a very, very low risk case.

ii) We note the Judge's comment that it would be a wise precaution to await the first round of correspondence before taking out a policy or entering into a CFA. However, we would not categorise it as unreasonable not to do so for reasons we have explained. It is only necessary to look at the relationship between the overall costs in this case and the amount recovered to see what are the consequences of adopting the Claimant's approach. Regrettable though this may be, if the uplift is reasonable, the scale of costs are a consequence of the present method of funding the costs of litigation.

iii) On an assessment, the court is not concerned with the question of the effect on proportionality of the uplift and the insurance premium (para 11.9 of the Practice Direction on Costs: see para 33 above) nor can the assessment be made with the benefit of hindsight (para 11.7 of the Practice Direction: ibid). However, subject to this, the approach of the Claimant must accord with the general principles set out in Part 1 of the CPR.

132. On the information then available to the party, it must act reasonably. If a CFA in a very straightforward case is entered into at the outset, it is going to be, and should be, scrutinised critically by the court on an assessment if the uplift is other than modest.
133. For the reasons that we have given earlier in this judgment, we consider that a success fee involving a 40% uplift for a case of this type was too high. We would allow the appeal to the extent of reducing the uplift to 20%. The parties will have to calculate the exact amount. We note, however, that the costs of this claim, which was readily settled, will still exceed the agreed damages.

15. The Russell appeal: (i) The facts

134. This appeal is an appeal against the decision of His Honour Judge Marshall Evans made on 25 January 2001 at the Liverpool County Court in costs-only proceedings. A success fee of 30% was claimed, and in the exercise of his discretion, the Judge reduced it to 20%. The Claimant was involved in a road traffic accident on 5 July 2000. His case was that the Appellant's vehicle reversed into his car when it was stationary.
135. He consulted E. Rex Makin & Company ("Makin") in order to claim damages for personal injuries. That firm deals with a substantial number of personal injury claims arising out of road traffic accidents in the course of its practice. A letter of claim was sent on 20 July 2000. On 31 July 2000, the Appellant's insurers, the Royal and Sun Alliance, replied stating that the matter had been referred to a claims handler who would be in contact shortly. It was the practice of these insurers to send a claims handler to visit this firm of solicitors to discuss cases. It was said that this led to cases settling swiftly and efficiently.
136. On 2 August 2000, notice was given to the insurers that the matter was proceeding by way of a CFA incorporating a success fee and that an insurance policy was pending. Contrary to the terms of that letter, the CFA was not in fact signed until 11 August 2000. The uplift was fixed at 30%. A medical report, dated 7 August 2000, was obtained by Makin which indicated that the Claimant had suffered whiplash injuries typical of this type of accident. The report indicated that he was in good health apart from another whiplash injury he had suffered a year earlier.
137. On 7 September 2000, the meeting with the claims handler took place. An offer was made of £1,450 plus reasonable costs. This offer was subsequently accepted. Costs were quantified at £420 +VAT, to which were added £70 in respect of the medical report, a 30% uplift and an insurance premium in the sum of £495. The costs were not paid. Costs-only proceedings were issued. The District Judge referred the proceedings to the Circuit Judge for determination because of the importance of the issue it raised. The claim for the insurance premium was abandoned prior to the hearing before the Judge.

16. The Russell appeal: (ii) The decision of the Judge and our conclusion

138. The argument before the Judge on behalf of the Appellant was similar to the argument that Mr King advanced before us, namely, that a CFA should not in the circumstances have been issued at the stage that it was. The Judge found this a difficult question, but he concluded that it was not so obvious that a CFA was patently unnecessary. There were very few "no-risk" cases in litigation. The Judge decided that there was nothing objectionable to a CFA being entered into at the outset, with the uplift being fixed at that time. He rejected a submission that it was far too early to assess what success fee was appropriate. He concluded that as at 11 August, 20% was a more appropriate uplift.
139. In view of our earlier conclusions, this is an appeal which on the facts of the case cannot succeed unless it is established that, as a matter of principle, a CFA could not have been entered into when it was. In our judgment, as we have made clear, there is no such principle. It is also clear from what we said in relation to the previous appeal that we should not interfere with the figure of 20% which the Judge decided was reasonable. We therefore dismiss this appeal.
ORDER:
140. Callery & Gray (Appeal allowed in part)
141. Russell & Pal Pak (Appeal dismissed)
142. Quantum of premium reserved pending report from costs Judge.
143. Respondents to have ¾ costs in Callery v Gray.
144. Payment on account £100,000 within 14 days.
145. No order as to costs in Russell v Pal Pak.
146. Appellant's applications for leave to appeal refused.
(Order does not form part of approved Judgment)