ON 27 JUNE 2002
 UKHL 28
LORD BINGHAM OF CORNHILL
1. For nearly half a century, legal aid provided out of public funds was the main source of funding for those of modest means who sought to make or (less frequently) defend claims in the civil courts and who needed professional help to do so. By this means access to the courts was made available to many who would otherwise, for want of means, have been denied it. But as time passed the defects of the legal aid regime established under the Legal Aid and Advice Act 1949 and later statutes became more and more apparent. While the scheme served the poorest well, it left many with means above a low ceiling in an unsatisfactory position, too well off to qualify for legal aid but too badly off to contemplate incurring the costs of contested litigation. There was no access to the courts for them. Moreover, the effective immunity against adverse costs orders enjoyed by legally-aided claimants was always recognised to place an unfair burden on a privately-funded defendant resisting a legally-aided claim, since he would be liable for both sides’ costs if he lost and his own even if he won. Most seriously of all, the cost to the public purse of providing civil legal aid had risen sharply, without however showing an increase in the number of cases funded or evidence that legal aid was directed to cases which most clearly justified the expenditure of public money.
2. Recognition of these defects underpinned the Access to Justice Act 1999 which, building on the Courts and Legal Services Act 1990, introduced a new regime for funding litigation, and in particular personal injury litigation with which alone this opinion is concerned. My noble and learned friend Lord Scott of Foscote makes full reference to these Acts and the relevant subordinate legislation made under them in his opinion, which I have been privileged to read in draft, and I gratefully adopt his account which I need not repeat. The 1999 Act and the accompanying regulations had (so far as relevant for present purposes) three aims. One aim was to contain the rising cost of legal aid to public funds and enable existing expenditure to be refocused on causes with the greatest need to be funded at public expense, whether because of their intrinsic importance or because of the difficulty of funding them otherwise than out of public funds or for both those reasons. A second aim was to improve access to the courts for members of the public with meritorious claims. It was appreciated that the risk of incurring substantial liabilities in costs is a powerful disincentive to all but the very rich from becoming involved in litigation, and it was therefore hoped that the new arrangements would enable claimants to protect themselves against liability for paying costs either to those acting for them or (if they chose) to those on the other side. A third aim was to discourage weak claims and enable successful defendants to recover their costs in actions brought against them by indigent claimants. Pursuant to the first of these aims publicly-funded assistance was withdrawn from run-of-the-mill personal injury claimants. The main instruments upon which it was intended that claimants should rely to achieve the second and third of the aims are described by my noble and learned friend: they are conditional fee agreements and insurance cover obtained after the event giving rise to the claim.
3. At the time when the 1999 Act was enacted and brought into effect, new Civil Procedure Rules were also in the course of being implemented. The objects underlying these rules were not new, but the rules gave a sharply increased emphasis to the need for expedition in the conduct of legal proceedings, to the need for simplicity and to the need to avoid unnecessary and disproportionate costs. To achieve these ends new and detailed procedures were devised to moderate the traditional adversarial approach to the making and defending of claims. There was inevitably a bedding-down period during which both judges and practitioners adjusted to the practical implications of the new procedural regime to which they were required to give effect.
4. If the objects underlying the new procedural regime were not new, those underlying the new funding regime were. Arrangements which had until relatively recently been professionally improper were to become the norm. It was however evident that the success of the new funding regime was threatened by two contingencies which, had they occurred, could have proved fatal. One was that lawyers, in particular solicitors, would decline to act on a conditional fee basis. To counter that risk the maximum permissible uplift, on the first introduction of conditional fees in 1995, had been fixed, despite very strong opposition, at 100% and this high level of permissible uplift was retained. It was no doubt felt, rightly as events have proved, that if solicitors were permitted in some cases to earn, as the reward for success, double the fee otherwise receivable, they would be tempted into the market. The other contingency was that no accessible market would develop in after the event insurance. There was at the outset very little knowledge and experience of whether or how such a market would develop.
5. Even if these contingencies did not occur, the new funding regime was obviously open to abuse in a number of ways. One possible abuse was that lawyers would be willing to act for claimants on a conditional fee basis but would charge excessive fees for their basic costs, knowing that their own client would not have to pay them and that the burden would in all probability fall on the defendant or his liability insurers. With this expectation the claimant’s lawyers would have no incentive to moderate their charges. Another possible abuse was that lawyers would be willing to act for claimants on a conditional fee basis but would contract for a success uplift grossly disproportionate to any fair assessment of the risks of failure in the litigation, again knowing that the burden of paying this uplifted fee would never fall on their client but would be borne by the defendant or his insurers. A third possible abuse was that claimants, although able to obtain after the event insurance, would be able to do so only at an unreasonably high price, the after the event insurers having no incentive to moderate a premium which would be paid by the defendant or his insurers and which might be grossly disproportionate to the risk which the insurer was underwriting. Under the new regime, a claimant who makes appropriate arrangements can litigate without any risk of ever having personally to pay costs either to those acting for him or to the other side and without any risk of ever having to pay an after the event insurance premium whatever the outcome: the practical result is to transfer the entire cost of funding this kind of litigation to the liability insurers of unsuccessful defendants (and defendants who settle the claims made against them) and thus, indirectly, to the wider public who pay premiums to insure themselves against liability to pay compensation for causing personal injury.
6. The front-line responsibility for making the new funding regime work fairly and effectively and in accordance with the objects both of the 1999 Act and the new Civil Procedure Rules lay with lawyers agreeing to act under conditional fee agreements and insurers offering after the event insurance cover. The role of watchdog would be exercised, in the first instance, by district judges and costs judges, on whose judgment and insight in assessing recoverable costs much would depend. If they were too restrictive in the level of success fees or after the event insurance premiums which they allowed, lawyers and clients might be deterred from acting or proceeding on this basis and the objects of the new regime would be defeated. If they were too generous and too uncritical, excessive fees and premiums might be allowed and an unfair and disproportionate burden placed on defendants and their liability insurers, thereby undermining one of the key objects of the Civil Procedure Rules. The difficult task entrusted to district judges and costs judges called for a clear understanding of the object of both the 1999 Act and the Civil Procedure Rules, an understanding how the new funding regime was developing in practice and an alert willingness to make appropriate orders if and when signs of abuse appeared. However carefully and attentively district judges and costs judges applied themselves to their task, it was inevitable that occasions would arise when judges misdirected themselves and made erroneous orders, and given the number of orders made by different judges in different parts of the country disparities in practice would be likely to arise. The responsibility for curbing errors and giving guidance to district judges and costs judges on the exercise of their powers in this context, of correcting erroneous orders and of seeking to harmonise practice between various courts rests with circuit judges and then, importantly, with the Court of Appeal.
7. There is obvious force in the appellant’s contention that even a 20% success uplift provided a generous level of reward for Mr Callery’s solicitors given the minuscule risk of failure which his claim apparently presented, that it would have been reasonable to await a reply from Mr Gray or his liability insurers before obtaining after the event insurance cover and that the premium charged for such cover, on the facts of the case as they then appeared, was unreasonable and disproportionate. There are nonetheless two reasons which lead me to the conclusion that the House should not intervene.
8. The first is that the responsibility for monitoring and controlling the developing practice in a field such as this lies with the Court of Appeal and not the House, which should ordinarily be slow to intervene. The House cannot respond to changes in practice with the speed and sensitivity of the Court of Appeal, before which a number of cases are likely over time to come. Although this is a final and not an interlocutory appeal, there is in my view some analogy between appeals on matters of practice and interlocutory appeals, of which Lord Diplock in Birkett v James  AC 297 at 317 observed that only very exceptionally are appeals upon such matters allowed to come before the House.
9. Even if this were one of those exceptional cases, however, I would decline to intervene because, as the Court of Appeal repeatedly stressed, the present issues arise at a very early stage in the practical development of the new funding regime, when reliable factual material is sparse, market experience is meagre and trends are hard to discern. I would draw attention, in the first judgment of the Court of Appeal reported at  1 WLR 2112, to passages in paragraphs 64, 99(v), 103, 105 and 116. In the second judgment of the Court of Appeal reported at  1 WLR 2142 similar points are made in paragraphs 14, 15, 17 and 69. In the report of Master O’Hare annexed to the second judgment of the Court of Appeal, relevant passages appear in paragraphs 19 and 23. The Court of Appeal made plain that it was not purporting to lay down rules applicable for all time but was giving provisional guidance to be reviewed in the light of increased knowledge and developing experience.
10. For these reasons (and those given by my noble and learned friends Lord Nicholls of Birkenhead and Lord Hope of Craighead) I would dismiss this appeal. In doing so I would not wish to discount either the risk of abuse or the need to check any practices which may undermine the fairness of the new funding regime. This should operate so as to promote access to justice but not so as to confer disproportionate benefits on legal practitioners or after the event insurers or impose unfair burdens on defendants or their insurers. I feel sure that district and costs judges, circuit judges and in the last resort the Court of Appeal can be relied on to maintain a fair and publicly beneficial balance between competing interests.
LORD NICHOLLS OF BIRKENHEAD
11. I have had the advantage of reading in draft the speeches of my noble and learned friends Lord Bingham of Cornhill and Lord Hope of Craighead. For the reasons they give I too would dismiss this appeal. I wish to emphasise only one point.
12. All legal procedures are open to abuse. A theme running through much of the appellant’s case in your Lordships’ House was that the new funding arrangements for personal injuries claims in road traffic cases are unduly open to abuse and are being abused. Costs are being incurred unnecessarily and at an excessive level. The underlying problem, it was said, is that claimants now operate in a costs-free and risk-free zone.
13. In short, this result comes about as follows. By entering into a conditional fee agreement at the outset, a claimant achieves the position that his solicitor’s charges will never be payable by him or at his expense. If his claim is successful the fees, including the amount of the uplift, will be payable by the defendant’s liability insurers. If his claim is unsuccessful, nothing will be due from him to his solicitor under the agreement. Likewise with the premium payable for after the event insurance: if the claim is successful, the premium will be payable by the other side’s liability insurers. If the claim is unsuccessful, nothing will be payable by the claimant when, as frequently happens, the policy provides that no premium will be payable in that event.
14. The consequence, it was said, of these arrangements, hugely attractive to claimants, is that claimants are entering into conditional fee agreements, and after the event insurance, at an inappropriately early stage. They have every incentive to do so, and no financial interest in doing otherwise. Moreover, in entering into conditional fee agreements and insurance arrangements they have no financial interest in keeping down their solicitors’ fees or the amount of the uplift or the amount of the policy premiums. Further, they have no financial incentive to accept reasonable offers or payments into court: come what may, their solicitors’ bills will be met by others. So will the other side’s legal costs.
15. As a result, it was said, the new arrangements, as they are currently working, are unbalanced and unfairly prejudicial to liability insurers and the general body of motorists whose insurance policy premiums provide the money with which liability insurers meet these personal injuries claims and costs. The appellant urged your Lordships to promulgate guidelines to reduce the scope for abuse in this situation.
16. My Lords, I agree that for the two reasons given by Lord Bingham your Lordships should decline this invitation. Plainly, however, the criticisms outlined above give cause for serious concern. It is imperative that these aspects of the new funding system should be watched closely as the system develops and matures.
17. The Court of Appeal is traditionally and rightly responsible for supervising the administration of civil procedure. This is an area in which your Lordships have in the past seldom intervened and, it must be said, the few exceptions to this policy of self-restraint have usually tended to confirm the wisdom of the general practice. In addition, the issues in this case arise out of an interaction between the new Civil Procedure Rules and the new conditional fee system for funding personal injury litigation. Both are in a relatively early stage of development. So I would be very reluctant to differ from a Court of Appeal which included Lord Woolf CJ, the architect of the new system, Lord Phillips MR, the Head of Civil Justice and Brooke LJ, one of the foremost experts on civil procedure. I am not satisfied that their decisions were wrong and therefore would dismiss the appeals.
18. That said, I am bound to add that I feel considerable unease about the present state of the law. In this respect I do not think that I am alone. There seems to be widespread recognition among those involved in personal injury litigation that costs, particularly in relation to small claims, are getting out of hand. They are excessive in relation to the amounts at stake (contrary to the principle of proportionality), some elements (such as after the event insurance premiums) lack transparency and, perhaps in consequence, too much time, money and court resources are spent in disputes over costs. The view of the Court of Appeal, at any rate as expressed in the judgments under appeal, is that things will settle down when costs judges acquire greater experience of applying the new rules to the new system of litigation funding. But I must express some doubt as to whether the questions which arise in these appeals are capable of solution by the traditional method of adjudication by costs judges, subject to guidance from the Court of Appeal. It may be that they are not justiciable and require a legislative solution.
19. Callery v Gray was a typical straightforward personal injury claim. On 2 April 2000 Mr Callery was a passenger in a car driven by Mr Wilson, which was involved in a collision in Ormskirk with a car driven by Mr Gray, insured by the Norwich Union. He instructed Messrs Amelans, solicitors who specialise entirely in personal injury litigation and process such claims on an industrial scale. On 28 April 2000 he signed a conditional fee agreement (CFA) which provided for a success fee of 60%. On 4 May 2000 he took out an after the event (ATE) insurance policy with Temple Legal Protection Ltd (“Temple”) for a premium of £367.50 inclusive of VAT. On the same day Amelans wrote a standard letter of claim to Mr Gray, which he passed on to his insurers. On 19 May 2000 Norwich Union wrote back admitting liability. A medical report were obtained and on 12 July 2000 Amelans made a Part 36 offer to accept £3,010 and costs. On 24 July 2000 the Norwich Union made a counter-offer of £1,200. On instructions from Mr Callery, Amelans telephoned Norwich Union and agreed to accept £1,500 and reasonable costs. This was confirmed on 7 August 2000.
20. There followed a dispute over costs. Amelans submitted a bill for £4,709.35. In proceedings commenced by Amelans under CPR 44.12A, District Judge Wallace summarily assessed their costs at £1,941 (including VAT) and ordered them to pay £285 as costs of the assessment. The assessment included a 40% success fee and allowed the Temple insurance premium as a disbursement. Norwich Union obtained permission to appeal on the question of whether it was reasonable to incur the cost of ATE insurance before sending the letter before action. At the hearing before Judge Edwards QC it also argued that a 40% success fee was far too high. The judge dismissed the appeal on both points. The Court of Appeal upheld the judge on the insurance point but reduced the success fee to 20%.
21. There are two points in relation to the success fee. The first is whether it was reasonable to fix it without waiting to find out whether the claim would be admitted. The second is whether in any event a 20% fee was too high for a claim which was as certain of success as anything in litigation can be (the Court of Appeal described as a “very, very low risk case”:( 1 WLR 2112, 2140).
22. Three arguments were given for fixing the success fee at once. The first was that it was a necessary part of a conditional fee agreement and that it was natural for client and solicitors to want to agree at the first opportunity upon the terms of engagement. The client wants to be sure from the beginning that whatever happened he will not have to pay any costs and the solicitor wants to be sure that any work he did will be covered by the agreement and recoverable (in the event of success) from the defendant. The Court of Appeal recorded (at p 2132, para 90) the claimants’ argument:
23. I am sure that giving such an assurance is an important selling point (it is stressed on the Amelans web site) and perhaps under the present rules an immediate conditional fee agreement is the only practical way of achieving it. In a large-scale study undertaken in 1998 on behalf of the Legal Aid Board Research Unit (Report of the Case Profiling Study Personal Injury Litigation in Practice) Mr Pascoe Pleasence noted (at p. 19) that “it was common for clients to have their initial advice in a free consultation with a solicitor. Often firms used free first consultations as a marketing tool…” It may therefore be that if a solicitor had to wait until he received an answer to his letter before action before fixing the success fee, he would be willing for marketing purposes to take the risk of not being able to recover from anyone the relatively trivial costs already incurred. On the other hand, it may need a change to the indemnity principle to provide him with the additional incentive of being able to recover those costs from the defendant if the claim succeeds. These are empirical questions on which it is difficult for judges to form a view.
24. The second argument was that by agreeing to a success fee at the first meeting, the client so to speak insures himself against having to pay a higher one later if his case turns out to be more difficult than at first appeared. (This is very similar to the argument for an early ATE insurance, which I shall come to later). At first sight, therefore, one could say that agreeing an immediate success fee is no more than economically rational behaviour on the part of any client and that the fee should therefore be recoverable as an expense reasonably incurred.
25. The difficulty is that while, in principle, it may be rational to agree a success fee at the earliest moment, it is extremely difficult to say whether the actual “premium” paid by the client was reasonable or not. This is because the client does not pay the “premium”, whether the success fee is agreed at an earlier or later stage. The transaction therefore lacks the features of a normal insurance, in which the transaction takes place against the background of an insurance market in which the economically rational client or his broker will choose the cheapest insurance suited to his needs. Since the client will in no event be paying the success fee out of his pocket or his damages, he is not concerned with economic rationality. He has no interest in what the fee is. The only persons who have such an interest are the solicitor on the one hand and the liability insurer who will be called upon to pay it on the other. And their interest centres entirely upon whether the agreed success fee will or will not exceed what the costs judge is willing to allow.
26. Amelans in fact assessed the success fee by reference to a “matrix” under which points were allocated between 0 (no success fee) and 20 (100%). The effect of the matrix was that virtually no personal injury cases could score less than 7 (35%) and Mr Callery’s scored another six because there were no witnesses (2) Amelans were funding the case (2) and it was expected to take over 6 months to settle (1). No doubt some kind of point system like this is essential in a firm in which large numbers of claims are processed. But it can hardly be regarded as a rational calculation and I do not think that the judge took much notice. What in fact determines the success fee solicitors charge is what costs judges have been willing to allow in more or less comparable cases, the fee being set at the level regarded as optimistic but hopefully not so optimistic as to provoke the liability insurers into contesting the amount. I shall in due course come back to the question of whether this is a sensible system.
27. The third argument is that assessing risk at the earliest stage is logical because the success fee should be related, not to the prospects of success in a particular case, on which it might be reasonable first to find out more before making the assessment, but on the prospective success rate over the whole range of cases which the solicitor undertakes. The argument for the Norwich Union, on the other hand, was that the success fee should reflect the risk in the particular case. If it was thought to be 90%, the success fee should be 11%. If it was evenly balanced, it should be 100%. If the prospects were less than even, an economically rational solicitor, faced with a statutory maximum success fee of 100%, should not undertake the case at all.
28. The respondents on the other hand said that lawyers needed to be compensated for undertaking risks at all. If they undertook two cases with even prospects of success, they might statistically be expected to win one and lose one. But there was a 25% chance they might lose both and they were entitled to compensate for this by charging a higher success fee on cases almost certain to win. If they could not do so, they would have to decline cases which appeared less than sure fire winners and this would reduce access to justice.
29. The Court of Appeal accepted this argument, which it described as a “global” approach to fixing success fees. It said, at p 2132, para 93:
30. It was therefore, said the Court of Appeal, an “inevitable consequence of Government policy that unsuccessful defendants should be subjected to an additional costs burden”: p 2133, para 95. As between the individual parties, it might appear unjust to saddle defendants with the cost of (high) success fees without giving them a fair chance to identify cases in which success is assured, but the alternative would be to impose liability for the same overall costs in the form of higher success fees payable by defendants who (perhaps reasonably) contested liability and lost. Turning to the question of what success fee could be regarded as reasonable in a simple road accident claim, the court said in its view 20% was the maximum which a costs judge should allow in a “modest and straightforward claim” which had no special features to suggest that the claim might not be sound: see p 2135, paras 102-104. It acknowledged that this view was based on very limited data and might have to be revised in the light of experience.
| 31. My Lords, the Court of Appeal gave this question the most anxious consideration and, as I have said, it has unrivalled knowledge of the problem. But I rather doubt whether, at any rate in their judicial capacities, they had the material on which to make a decision. The liability insurers who supported these appeals accept that the policy of the Access to Justice Act 1999 was to transfer the burden of funding unsuccessful motor accident claims from public funds to the motor liability insurers and thence to the motoring public by way of increased premiums. That is a perfectly rational social and economic policy and is not contested. The real questions are, first, whether the level of success fees chargeable by lawyers gives the motoring public reasonable value for the money it has to spend on funding litigation and, secondly, whether assessment by costs judges (with guidance from the Court of Appeal) is the best way to ensure that such value for money is obtained.
32. The reasoning of the Court of Appeal in para 93 of its judgment, which I have quoted above, assumes that the present cost of motor accident litigation is a fixed sum which must be paid by liability insurers one way or the other. But that is the very question in issue. And the reason why it is disputed is that, in the circumstances in which such litigation is funded, market forces are insufficient to keep costs within reasonable limits. As I have already said, solicitors offering motor accident personal injury CFAs have no incentive to compete on the success fees they charge. So the next question is whether a decision of a costs judge, or the Court of Appeal on appeal from a costs judge, is the best way of compensating for the absence of price competition in the market. The traditional function of the costs judge, or taxing master, as he used to be called, was to decide what fees were reasonable by reference to his experience of the general level of fees being charged for comparable work. But this approach only makes sense if the general level of fees is itself directly or indirectly determined by market forces. Otherwise the exercise becomes circular and costs judges will be deciding what is reasonable according to general levels which costs judges themselves have determined. In such circumstances there is no restraint upon a ratchet effect whereby the highest success fees obtainable from a costs judge are relied upon in subsequent assessments.
33. The matter becomes even more difficult when a solicitor “carrying on litigation business on a large scale” is entitled, as the Court of Appeal have said (at p 2131, para 83) to fix success fees 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.” The costs judge has simply no way of knowing whether the solicitor is carrying on business on a large enough scale to justify such an approach, still less what level of success fees would give him a “reasonable return overall.” Such matters are traditionally outside the consideration of costs judges. As your Lordships’ Appellate Committee said in The Clerk of the Parliaments’ Reference Regarding Criminal Legal Aid Taxation (1st Report Session 1997-1998) at para 30:
34. I rather doubt whether difficulty is likely to be removed merely by the passage of time. All that costs judges will learn is what other costs judges are allowing. Solicitors will charge whatever is currently allowed and exert upward pressure to be able to charge more. But that will not tell anyone whether the fees paid to the solicitors represent reasonable value for money.
35. As my noble and learned friend Lord Scott of Foscote has observed, the criteria prescribed by the Civil Procedure Rules for determining whether costs are reasonable are framed entirely by reference to the facts of the particular case. Once one invokes a global approach designed to produce a reasonable overall return for solicitors, one moves away from the judicial function of the costs judge and into the territory of legislative or administrative decision.
36. It seems to me likely (although reliable statistical evidence is not presently available) that the costs which can reasonably be expended in a small personal injury claim conducted in accordance with the protocol and which settles within the protocol period are unlikely in the great majority of cases to vary much from the statistical mean. The present case – letter before action, acknowledgement of liability, medical report and a relatively brief negotiation – must be fairly typical. It involves standardised legal services at a fairly low level. This would suggest that these costs – both the basic costs and the success fee – should be fixed by the rules which apply in all but exceptional circumstances. I understand that such a system is under consideration by the Civil Justice Council. A legislative decision to fix costs at levels calculated to provide adequate access to justice in the most economical way seems to me a more rational approach than to leave the matter to individual costs judges. If it is considered the most appropriate way to secure value for money when the expenditure is borne by the public as a whole (as for example, in the fixing of graduated fees for criminal legal aid) it should be no less appropriate when the expenditure is borne by a section of the public, namely the motorists. Not only would this be more likely to keep the actual costs within reasonable levels but it would also greatly reduce the cost of disputes over costs. We were told that no less than 150,000 cases awaited the outcome of your Lordships’ decision in this case.
37. In their submissions to the Court of Appeal, the liability insurers said that what they wanted most was certainty. If the consequence of your Lordships’ decision was that motor insurance premiums had to be raised, so be it. But they wanted to be able to make the calculations with reasonable certainty about what the motoring public would have to pay. This would not be the case if everything turned upon a case by case assessment by the costs judge.
38. My Lords, the arguments for legislative determination of allowable costs apply even more strongly to the other two questions in these appeals, namely whether it was reasonable to take out ATE insurance before it was known whether the claim would be resisted and whether £375 was a reasonable premium in the circumstances.
39. The arguments put forward to justify taking out an ATE insurance policy before sending the letter before action are much the same as those for fixing the success fee at the same time. The client is immediately assured that in no circumstances will he be liable for the defendant’s costs. He may pay a higher premium than would be individually justifiable if it was known that the claim was admitted, but he insures himself against the possibility that a later premium might be much higher or even unobtainable. And ATE insurers say that they cannot obtain a reasonable premium income unless everyone takes out insurance when they first instruct solicitors. This was the principle upon which insurers such as Temple, in this case, delegated to solicitors the authority to issue policies.
40. Again, as it seems to me, the Court of Appeal accepted these arguments. They said (at p 2128, para 67) that –
41. Perhaps that is true. I am certainly not in a position to say that it was wrong. But neither, in my respectful opinion, was the Court of Appeal. Of course it is true that, other things being equal, premiums will rise if fewer people take out ATE insurance. And it is true that when ATE insurance made its first appearance, at a time when it could not be recovered from the defendant and claimants would take out policies only if they were seriously concerned about losing the case, underwriters burnt their fingers. But premiums were then also very low compared to current rates. Now, premiums are much higher and ATE insurers insist upon all claimants taking out policies. Whether the latter is necessary to keep ATE insurers in business at current premium rates is an open question.
42. Furthermore, it is a question which costs judges are quite unable to answer. When the Court of Appeal asked for the report of Master O’Hare on the question of whether the Temple premium in this case was reasonable, he said  1 WLR 2142, 2163, para 20:
43. That seems to me obviously right. ATE insurers do not compete for claimants, still less do they compete on premiums charged. They compete for solicitors who will sell or recommend their product. And they compete by offering solicitors the most profitable arrangements to enable them to attract profitable work. There is only one restraining force on the premium charged and that is how much the costs judge will allow on an assessment against the liability insurer.
44. Again, the costs judge has absolutely no criteria to enable him to decide whether any given premium is reasonable. On the contrary, the likelihood is that whatever costs judges are prepared to allow will constitute the benchmark around which ATE insurers will tacitly collude in fixing their premiums. In its submissions to Master O’Hare, Temple said that the court “should not arrogate to itself the functions of a financial regulator of the insurance industry”: see  1 WLR 2142, 2164, para 22. I am sure that is right, because the costs judge is wholly unequipped to perform that function. But that does not mean that some form of financial regulation is not necessary. Such regulation is normally considered necessary in those parts of the economy in which market forces are insufficient to produce an efficient use of resources. And that seems to me to be the position in ATE insurance, in which the premiums are not paid either by the claimants who take out the insurance or by the solicitors who advise or require them to do so.
45. My Lords, these are my reservations about the exercise which your Lordships are asked to undertake. But, given that your Lordships have to say whether, on the materials before you, the Court of Appeal were right or wrong, I would dismiss the appeals.
LORD HOPE OF CRAIGHEAD
46. This appeal raises important questions about the operation of a new costs regime to assist members of the public who wish to claim damages for personal injury. The facts have been fully described by my noble and learned friend Lord Scott of Foscote, and I gratefully adopt his account of them.
47. The previous scheme was based on the provision of legal aid out of public funds. But it had become clear that it was no longer possible for the government to fund legal aid in civil cases in England and Wales across the board at the required level. So the government decided to withdraw legal aid from certain categories of proceedings including actions of damages for personal injury, and to replace it with a new system based on conditional fee and litigation funding agreements. The legislation which was required to give effect to this was enacted as Part II of the Access to Justice Act 1999. An essential part of the new arrangement was the government’s decision that claimants should not be at risk of being left out of pocket by reason of the additional liabilities which they would have to incur under this scheme. These included the cost of taking out an insurance policy against the risk of incurring a liability to the other party in the proceedings for costs.
48. The new scheme was the subject of a consultation paper entitled Conditional Fees: Sharing the Risks of Litigation which was issued by the Lord Chancellor’s Department in September 1999. It was based upon two recent developments in the private funding of litigation. The first was the introduction of conditional fee agreements (“CFAs”) under section 58 of the Courts and Legal Services Act 1990. Initially there had been concerns that these agreements would be abused and that this part of the new system would be unworkable. But experience of its operation showed that this was not so. In Thai Trading Co (A Firm) v Taylor  QB 781, 790F Millett LJ said that fears that lawyers might be tempted by such an arrangement to act improperly were exaggerated, and that there was a countervailing public policy in making justice readily accessible to persons of modest means. The second was the concept of after the event insurance (“ATE”). This was a new kind of insurance policy which had been developed by the Law Society in conjunction with insurance brokers as a means of protecting claimants against their liability for the other side’s costs.
49. These developments had gone a long way towards removing the barrier to justice for prospective litigants. But they still left one problem untouched. In its response to consultation the government said that it recognised that a major barrier to justice had always been the risk that litigants would have to pay the other side’s costs as well as their own if they lost the case. The policy which it decided to pursue was to ensure that the expense of shifting the claimant’s liability in costs, whether to the solicitor under a CFA or to the insurer for the premium payable under the ATE insurance policy, was met by the losing party where damages were claimed for personal injury. As the consultation paper pointed out, the purpose of the legislation was to provide the framework within which the new system was to operate. The detail was to be informed by rules of court and by practice directions. There was to be a division of responsibility between the courts and the legislature.
50. The appellant seeks to challenge the operation of two powers which have been given by the legislature under the new scheme to the courts. The first is the power to include in an award of costs in favour of the claimant an order for payment by the losing party of the amount of the success fee payable by the claimant under a CFA entered into with his legal representative. Section 58A(7) of the Courts and Legal Services Act 1990, inserted by section 27(1) of the Access to Justice Act 1999, provides:
The second is the power of the court to include in a costs order the amount of the insurance premium for a claimant’s ATE insurance policy against the risk of incurring a liability to pay the other parties’ costs in the event of his having to raise proceedings to recover damages. Section 29 of the Access to Justice Act 1999 provides:
51. Three broad questions are raised by these new powers. The first is whether it is right in principle that a claimant’s liability for his legal representative’s success fee and the cost of his insurance premium should be recoverable from the losing party. The second is whether it should ordinarily be considered to be reasonable for the purposes of any costs order for the claimant to enter into the CFA, and to take out the ATE policy at the same time, on the occasion when he first consults his legal representative. The third is whether the cost of insuring against a cost liability that cannot be passed on to the opposing party should be included in the recoverable amount of the insurance premium. In addition the appellant has challenged the decision of the Court of Appeal not to interfere with the judge’s conclusion that a success fee of 20% was appropriate in this case.
52. Answers to these questions were provided by the Court of Appeal in the two judgments which are before your Lordships in this appeal. In Callery v Gray  1 WLR 2112 (Lord Woolf CJ, Lord Phillips of Worth Matravers MR and Brooke LJ) the court held that, where a reasonable uplift is agreed at the outset as part of a CFA and an ATE insurance is taken out at that stage at a reasonable premium, the costs of both the uplift and the premium 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: p 2134F, para 100. The court said that it had reached this conclusion after taking account of the legislative policy and various practical considerations which were established by the evidence: p 2133F, para 99. Its conclusion on the issue as to the reasonableness of the uplift by way of a success fee was that, where a CFA is agreed at the outset in straightforward cases such as this one, 20% is the maximum uplift that can reasonably be agreed.
53. The Court of Appeal dealt in a separate judgment with the reasonableness of the ATE insurance premium: Callery v Gray (No 2)  1 WLR 2142 (Lord Phillips of Worth Matravers MR and Brooke LJ). It held that the jurisdiction which has been given to the courts by section 29 of the 1999 Act to order the defendant to pay the cost of the insurance premium extends to the cost of insurance against the risk of incurring a costs liability that cannot be passed on to the opposing party: p 2157B-D; paras 59-60. The court said that it believed that this interpretation of section 29 gave the words the meaning that would be attributed to them by the reasonable litigant, and that it also gave them a meaning that accorded with the legislative intention and with the overall scheme for the funding of legal costs.
54. I accept that the new regime is not beyond criticism. It is plain that it has transferred the burden of meeting the cost of access to justice in these cases from the taxpayer through the medium of the legal aid fund. Instead a different kind of taxation is involved. The burden of meeting the cost of access to justice now falls on liability insurers. It thus falls indirectly on their policy holders, who are likely to have to face increased premiums. Furthermore, unless the new regime is controlled very carefully, its effect may be to benefit ATE insurance providers unreasonably and to place a burden on liability insurers which is disproportionate. It may lead to a culture of incurring additional costs which lacks any incentive on claimants to keep costs down. The circumstances of the present case add force to these criticisms. It must have been obvious from the outset that this was a claim which was almost certain to succeed on liability and that there was almost nothing to discuss on the quantum of damages. The claimant’s exposure to risk was, at the worst, minimal. And, as my noble and learned friend Lord Scott has pointed out, the premium which Temple charged for the insurance cover was not based on an individual assessment of the risk but was a uniform premium. The costs incurred in this case by way of the success fee and the ATE premium do indeed appear at first sight to be wholly out of line with what the case required.
55. But it would be hard to imagine judges who were better qualified to examine these issues than the Lord Chief Justice, the Master of Rolls – who has particular responsibilities in this field as the Head of Civil Justice in England and Wales – and Brooke LJ. They are in close and regular contact with all relevant aspects of civil practice. They heard representations from a number of bodies who have a direct interest in these issues. They also had the advantage in this case of having a report on the reasonableness of the premiums charged for ATE insurance from Master O’Hare. He has an unrivalled expertise in the matter of costs. The Court of Appeal could not have been better informed. I think that their judgment is entitled to the greatest respect in a matter of this kind. I would be very slow to differ from the conclusions which they have reached.
56. In Girvan v Inverness Farmers Dairy 1998 SC (HL) 1, 21D-G I said that the Court of Session is far better placed that this House can ever be to assess what changes could appropriately be made in procedure and practice relating to the conduct of civil jury trials in that court. I drew attention to the process of consultation that is available to it through its statutory rules council, to its ability to keep its rules under regular review and to the advantages which it can bring to bear of speed and flexibility. I suggested that the proper approach for this House to take was to leave it to the Court of Session to decide what changes, if any, should be made to its own rules. This case is concerned with the role of the court in administering the new regime in a way that seeks to do justice between the parties within the limits that have been set by the legislative policy. For similar reasons to those which I gave in Girvan I consider that the responsibility for dealing with these issues lies pre-eminently with the Court of Appeal and not with this House.
57. Two other factors seem to me to favour the decisions which the Court of Appeal reached in this case. The first is the fact that it was common ground that not enough is yet known about the likely effect on CFAs of different levels of success fees: Callery v Gray  1 WLR 2112, 2128A, para 64. The appellant did not suggest to your Lordships that the position was otherwise. The second is the state of the evidence about the effect on the market for ATE insurance if your Lordships were to uphold the appellant’s argument that the taking out of these policies should be delayed until proceedings have been issued. The Court of Appeal held that there was 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 and cover may indeed not be available at all in such circumstances: p 2134E-F, para 99(ix).
58. Ample support for the finding about the effect on ATE insurance is to be found in witness statements provided by Christopher Ward and Christopher Wait. Mr Ward is the managing director of Abbey Legal Protection Ltd, a specialist legal expenses underwriting agency providing both before the event and after the event legal expenses insurance. Mr Wait is an underwriting director of Temple Legal Protection Ltd, which specialises in the same field. Mr Temple QC for the appellant said that these statements were self-serving and that little weight should be attached to them. But there was no contrary evidence.
59. I note also that in para 27 of his report to the Court of Appeal Master O’Hare acknowledged the risks to which ATE providers would be exposed if increasing premiums and the need to make inquiries were to encourage more claimants to dispense with insurance altogether: Callery v Gray (No 2)  1 WLR 2142, 2165D. In preparing his report he considered written and oral submissions from many parties: see para 2. He also drew attention in para 16 to the advantages which flow from the insurers’ practice of giving delegated authority to solicitors to offer ATE cover to their clients. As he explained, it is characteristic of these policies that they do not differentiate between cases which are strong and those which are borderline. This all-in approach reduces the administrative cost of risk assessment, and there is the further point that risk assessment at the outset of proceedings may well be imprecise or unreliable. His findings support the Court of Appeal’s observation in Callery v Gray  1 WLR 2112, 2128H that it was hardly surprising that delegated authority arrangements would only work successfully if the solicitor did not “cherry-pick” by taking out ATE insurance only in risky cases. It is plain that a finding that the cost of taking out ATE insurance at the outset was in principle irrecoverable from the losing party would have a profound effect on the cost and availability of this form of insurance cover.
60. These and the other practical considerations relied on by the Court of Appeal seem to me to point clearly in favour of allowing the ATE insurance policy to be taken out at the outset when the claimant first consults his legal representative. The extent to which the cost of doing so will be recoverable from the losing party will require to be watched very carefully. But a sufficient safeguard lies in the fact that it will continue to be subject to the supervision of the court on the basis described in Callery v Gray (No 2)  1 WLR 2142, 2157D, para 61:
61. I think that it is also worth recording that the appellant’s counsel did not challenge the conclusions of the Court of Appeal in Callery v Gray (No 2)  1 WLR 2142, 2157C-D, para 60 as to how section 29 of the 1999 Act ought to be interpreted. This was an issue of law which would have been open for further consideration by your Lordships if it had been suggested that the section was not capable of bearing the meaning which the Court of Appeal said could and should be given to it. But Mr Birts QC accepted that it had reached a sound conclusion on this point, and for my part I think that he was right to do so. The remaining issues seem to me to be essentially issues of practice. They are best dealt with by penalising unreasonable behaviour on a case by case basis as the circumstances require.
62. For these reasons, and for those given by my noble and learned friend Lord Bingham of Cornhill whose speech I have had the advantage of reading in draft and with which I am in entire agreement, I would dismiss the appeal.
LORD SCOTT OF FOSCOTE
| 63. The civil courts of this country have power to order a party to litigation to pay to the other party his (the latter’s) costs of the litigation. This power presumably owes its origin to the inherent power of the court to regulate the proceedings before it, but in modern times the power has been a statutory one, embodied in rules of court made under rule-making powers conferred by statute. The power to make costs orders has always been, and remains, a discretionary one. But the discretion is circumscribed by guidance laid down under rules of court and by principles established by judicial precedent.
64. At the risk of some over-simplification, the issues that may arise in relation to costs orders fall into three groups. First, there are issues as to which, if any, of the litigants should have a costs order made in his favour and against whom the order should be made. Secondly, there may be issues as to the extent of the receiving party’s costs that the paying party should be ordered to pay. And, thirdly, there may be issues of quantification. The sum that, under the order, the paying party must pay to the receiving party will have to be quantified. A process of assessment must be carried out.
65. For example, a successful claimant in a road traffic accident case would expect to have an order for costs made in his favour against the negligent defendant. But if the engagement by the claimant of a particular expert witness appeared to have been unnecessary, the court might exclude the fee paid to the expert witness from the costs payable under the order. And if the engagement of the expert witness appeared reasonable but the fee paid to the expert appeared to be excessive, the court might reduce to a reasonable amount the sum in respect of the fee payable under the costs order. These basic principles are very well known and understood and hardly need repeating, but, in my opinion, need to be kept in mind in considering some of the issues before the House on this appeal.
66. The appeal relates to costs issues arising out of a very simple and commonplace type of claim. The difficulties that have led to this appeal are attributable to recent changes in the costs regime, which reflect changes in the means by which the bringing of actions can be funded. In order to deal with these issues it is necessary to set out the facts which have given rise to them and the statutory provisions that bear upon them.
67. On 2 April 2000 Mr Callery, respondent before your Lordships, was a passenger in a car which was struck side-on by a vehicle driven by the appellant, Mr Gray. Mr Callery sustained minor injuries. He instructed solicitors, Messrs Amelans, to pursue a claim for damages against Mr Gray. Amelans are a firm who deal with a large number of road traffic accident claims. They are specialists in this field. I have described Mr Callery’s claim as commonplace. It was. There are thousands of such claims every year. They rarely reach the courts.
The funding arrangements
68. Under the Access to Justice Act 1999, legal aid to support road traffic accident claims such as Mr Callery’s was no longer available. Instead the use of conditional fee agreements (CFAs) was authorised. A feature of a CFA is that, if the claim fails, the lawyer, whether solicitor or barrister, who has given his services under the CFA, receives no remuneration but, if the claim succeeds, the lawyer becomes entitled not only to the normal fee for his services but also to a success fee, calculated as a percentage of the normal fee. (see the Conditional Fee Agreements Order 1998 (SI 1998/1860). The percentage uplift is required by the 1998 order to be specified in the CFA and must not exceed 100 per cent of the normal fee.
69. The logic of the success fee is that its size will reflect the risk the lawyer is incurring in taking on the case. The more difficult the case and the less clear that the outcome will be a successful one, the higher the percentage uplift that can be justified; and, of course, vice versa. A CFA protects the claimant who enters into it from having to remunerate his lawyers if the claim fails. A typical CFA does not, however, protect the claimant from having to reimburse his lawyers for their disbursements in pursuing the claim. These disbursements might, for example, include the fees of experts whose opinion on some issue or other had been sought. If the claim succeeds the claimant will have the expectation that the burden of paying for his solicitor’s reasonable disbursements will fall on the defendant. If the claim fails, he will have to meet the disbursements himself.
70. Nor does a CFA protect the claimant from the risk that if litigation is commenced he may find himself ordered to pay the costs, or some part of the costs, of the defendant. But it is important to notice that this risk cannot arise unless litigation is commenced.
71. In order to protect himself against the risk that he may find himself liable to pay the costs, or some of the costs, of the defendant, a claimant can take out after-the-event (“ATE”) insurance.
72. In the present case Mr Callery entered into a CFA with Amelans at the same time as instructing them to pursue his damages claim. The CFA was dated 28 April 2000. It said that if Mr Callery succeeded in his claim for damages he would pay Amelans’ basic charges and a success fee. It said that, win or lose, Mr Callery would have to pay their disbursements. It set the success fee at 60 per cent of Amelans’ basic charges but provided also that in the event of an assessment of costs by the court
73. The effect of a provision in these terms would be, in practice, and save in exceptional circumstances, to relieve the claimant of his contractual liability to pay any excess of the success fee above the amount that, on the assessment of costs, the court had decided the defendant should pay.
74. On 4 May 2000 Mr Callery took out an ATE insurance policy with Temple Legal Protection Ltd (“Temple”). This was the date on which Amelans’ first letter to Mr Gray, informing him of the claim, was sent. The policy was taken out before Mr Callery or his legal advisers, Amelans, could have come to the conclusion that litigation to pursue the claim would be necessary or even likely. The premium for the policy was £350 plus £17.50 tax, a total of £367.50. Under the policy Temple agreed to indemnify Mr Callery, up to a limit of £100,000,
(i) for any sum in respect of Mr Gray’s costs that might be ordered by the court to be paid by Mr Callery; and
(ii) for disbursements paid out by Amelans in the event that the litigation came to an end without the disbursements becoming payable by Mr Gray.
In this description of the insurance cover provided under the policy I have attempted to describe the broad scope of the cover rather than set out its exact details. The exact details are not relevant to any issue before the House.
75. A noteworthy feature of the policy is that the premium of £350.00 would not become payable until the conclusion of the legal proceedings in respect of which the policy had been taken out and, if the amount of the premium were challenged by Mr Gray in any cost assessment process, the recoverable amount of the premium would be reduced to the amount payable by Mr Gray under the assessment.
The progress of Mr Callery’s claim
76. On 4 May 2000 (the day on which the ATE policy was taken out) Amelans wrote to Mr Gray giving notice of their client’s damages claim and enclosing a copy of the letter for his insurers. The letter asked for an acknowledgement within 21 days from Mr Gray or his insurers.
77. On 19 May 2000, Mr Gray’s insurers, CGU Insurance, replied. Their reply said that they were able “to admit liability as to negligence but not to causation”. Following an exchange of offer and counter-offer, the parties reached agreement on 2 August 2000 that Mr Callery should receive damages of £1500. Amelans wrote on 7 August 2000 to CGU confirming “acceptance of your offer in the sum of £1500 in respect of our client’s damages” and that “acceptance of the offer is subject to payment of our client’s reasonable costs and disbursements …. “. However, the parties were unable to agree the amount of the “reasonable costs and disbursements”. There were two main issues.
78. First, CGU took the view that a 60 per cent success fee for Amelans was unreasonably high. It was always certain, they thought, that Mr Callery would succeed in recovering damages. He had been a passenger in a vehicle struck side-on by Mr Gray’s vehicle. He had been injured in the accident. He was bound to recover something. So this was not a case in which there was any risk that Amelans, in pursuing the claim on Mr Callery’s behalf, would be working for nothing. It followed that any success fee should be very low.
79. Second, CGU contended that the £350.00 premium was an item of expenditure that had not been reasonably incurred. The likelihood of litigation being necessary was always very remote. At the least Mr Callery should have waited until a response to Amelans’ letter of 4 May 2000 had been received before taking out a policy to protect himself against the costs consequences of litigation that was unlikely ever to materialise.
80. In these circumstances, there being agreement as to the amount of the damages and agreement that Mr Callery would be paid his reasonable costs and disbursements but no agreement as to the amount of a reasonable success fee or whether the £350.00 premium represented a reasonable disbursement, Mr Callery commenced “costs only” proceedings (see CPR 44.12A) in order to have the reasonable costs and disbursements to which he was entitled under the settlement agreement quantified by the court.
The relevant statutory provisions
81. The statutory provisions relevant to the recovery under a costs order of a CFA success fee or of the premium payable for ATE insurance cover have been comprehensively set out in paragraphs 24 to 39 of the judgment of the Court of Appeal handed down by Lord Woolf CJ, sitting with Lord Phillips of Worth Matravers MR and Brooke LJ, on 17 July 2001:  1 WLR 2112, 2118-2123. This was the first of the two Court of Appeal judgments from which the appeal to the House is brought. It is unnecessary for me to do more than refer to the critical provisions.
82. Section 58(A)(6) of the Courts and Legal Services Act 1990, as substituted by section 27 of the Access to Justice Act 1999 provides that:
Section 29 of the 1999 Act provides:
83. There is, therefore, no doubt but that a costs order can require the paying party to pay a sum in respect of a success fee for which the receiving party has become contractually liable and to pay a sum in respect of the premium payable for an ATE insurance policy that the receiving party has taken out.
84. In the Civil Procedure Rules, Rule 43.2 provides that:
and that “additional liability” includes the success fee under a CFA and the premium payable for an ATE insurance policy.
85. Rule 44.4 sets out the basis upon which costs are to be assessed:
86. Rule 44.5 sets out the factors the court must take into account in assessing costs:
87. These factors, as one would expect, concentrate on the circumstances and conduct of the parties in relation to the claim in question and on the nature of the particular claim. They are case specific factors.
88. The Costs Practice Direction that supplements the Rules relating to costs contains further guidance on the assessment, for costs purposes, of an additional liability. Section 11 of the Costs Practice Direction contains the following relevant provisions:
89. Rule 44.12A and Section 17 of the Costs Practice Direction relate to “costs only” proceedings. “Costs only” proceedings may be commenced where the parties, before the commencement of litigation, have agreed in writing on all issues in the case, including which party is to pay the costs, but have been unable to agree the amount of the costs (see CPR 44.12A). In effect, the court is allowing its costs quantification procedures to be used to support an out-of-court settlement under which one party is to pay the reasonable costs and disbursements of another party. It is implicit in such a settlement agreement (unless the contrary is expressly stated) that the costs to be paid should be quantified on the standard basis in accordance with the Rules and Practice Directions that would have been applicable if there had been a court order for the payment of the costs.
90. Paragraph 17.8(2) of the Costs Practice Direction makes clear that in assessing the reasonableness of an additional liability, whether a success fee or an ATE premium, the likelihood of litigation having to be commenced is an important factor. The paragraph provides that:
The course taken by the “costs only” proceedings
91. On 12 September 2000 Mr Callery commenced proceedings under CPR Part 8 asking the court to assess his reasonable costs of and occasioned by his damages claim against Mr Gray. The particulars of claim pleaded the agreement that had been reached that Mr Gray would pay £1500 damages and Mr Callery’s reasonable costs.
92. On 14 September an ex parte order was made in the Macclesfield County Court ordering, inter alia, that Mr Callery’s costs be assessed.
93. The assessment hearing took place before District Judge Wallace on 7 November 2000. The District Judge assessed Amelans’ basic charges, reduced the success fee from 60 per cent to 40 per cent of the basic charges and allowed the ATE insurance premium of £350 plus tax as a disbursement. The assessment allowed a total of £1008, plus VAT, for fees and £617.50 for disbursements.
94. Mr Gray (prompted by CGU) appealed to the circuit judge. The appeal was heard by Judge Edwards in the Chester County Court on 29 January 2001. Two points were taken on the appeal, one relating to Amelans’ success fee under the CFA, the other relating to the ATE insurance premium. They are the two points to which I have already referred. The appeal was dismissed on both points.
95. CGU took the view that important points of principle were at stake, with implications for personal injury litigants and their insurers generally, and prompted Mr Gray to apply to the Court of Appeal for permission to bring a second appeal to that court. On 16 March 2001 Lady Justice Hale granted permission on condition that Mr Gray, if successful, would not seeks costs from Mr Callery unless his liability to pay were covered in full by the terms of the ATE policy.
96. Further evidence was, with the permission of the Court of Appeal, introduced at the hearing of the appeal and written and oral intervention by the Law Society, the Association of British Insurers, the After Event Insurers Group Forum, the Association of Personal Injury Lawyers and the Forum of Insurance Lawyers was authorised. The hearing of the appeal was directed to be conjoined with the hearing of the appeal in Russell v Pal Pak Corrugated Ltd under which similar success fee issues arose.
97. The Court of Appeal heard the appeal on 5, 6 and 7 June 2001. On 17 July 2001 the court handed down the judgment of the court on the success fee issue, reducing Amelans’ success fee from the 40 per cent fixed by the district judge to 20 per cent (and dismissing the appeal in Russell v Pal Pak)  1 WLR 2112. In addition the court ruled that an ATE insurance premium could in principle be recovered as part of a claimant’s costs in “costs only” proceedings. But the court formed the view that, on the evidence available, it was not possible to come to a conclusion as to the reasonableness in amount of the £350.00 premium. So the court postponed judgment on the issue relating to the amount of the ATE insurance premium pending an inquiry by a costs judge, Master O’Hare. A separate judgment on that issue would await Master O’Hare’s report.
98. It is important to be clear that the court’s further reservation of judgment regarding the ATE insurance premium related only to the reasonableness of the amount of the premium. In principle the premium was held to be recoverable notwithstanding that the policy had been taken out before the commencement of litigation for the purpose of recovering damages had become necessary, or even likely. As it was put at p 2134, para 100 of the judgment:
99. I will return later to examine the reasons why, in relation to the ATE insurance premium, the court came to this conclusion.
100. Master O’Hare’s report was dated 23 July 2001. The purpose of his report was expressed to be
101. Master O’Hare had received written submissions from a number of interested parties but nonetheless concluded that:
And, at para 23:
102. On 31 July 2001 Lord Phillips of Worth Matravers, sitting with Brooke LJ, handed down the judgment of the court dealing with the reasonableness of the amount of Mr Callery’s ATE premium  1 WLR 2142. The Court concluded that “the amount of the premium does not strike us as manifestly disproportionate to the risk” (p 2159, para 70).
103. So the appeal on the insurance premium point was dismissed. The appeal on the success fee point had been allowed by the reduction of the percentage uplift from 40 per cent to 20 per cent.
The section 29 jurisdiction issue
104. One of the contentions argued on behalf of Mr Gray, the appellant, before the Court of Appeal was that section 29 of the 1999 Act was not applicable to a case where a settlement out of court had been reached and all that was left was the quantification of the claimant’s costs and disbursements to be paid by the defendant. The point was based on the language of section 29. The “proceedings” referred to in the section were the substantive proceedings for recovery of damages. Absent the section, ATE insurance premiums would not have been recoverable under a costs order. So, in the absence of any substantive proceedings for the recovery of damages, or any order for costs made in such proceedings, the court, it was said, had no jurisdiction to include an ATE insurance premium in the recoverable costs. This argument was advanced in the Court of Appeal but was rejected. The argument has not been raised again before your Lordships and I mention it only in order to express my respectful agreement with the Court of Appeal’s rejection of it. It was a term of the settlement agreement that Mr Callery would be paid his reasonable costs and expenses. It was plainly implicit that these would be standard basis costs and that the amount, unless agreed, would be assessed by a costs judge in accordance with the relevant rules and practice directions. These rules and practice directions allow for a sum in respect of an ATE premium that has been reasonably incurred to be included in the recoverable costs. So in quantifying Mr Callery’s recoverable costs on that basis the district judge was acting in accordance with the agreement the parties had made. There is no jurisdictional reason why a costs judge should not assess the costs to which a party has become contractually entitled (see Gomba Holdings (UK) Ltd v Minories Finance Ltd (No. 2)  Ch 171, 188). No issue of jurisdiction, in my opinion arises.
The issues before the House
105. The appellant’s written case identifies ten issues—
“(1) Was the insurance premium recoverable given the stage when the policy was taken out? (prematurity).
(2) Was the amount of the premium reasonable having regard to the characteristics of the instant case?
(3) Was the amount of the premium reasonable having regard to prematurity?
(4) Was the Court of Appeal wrong not to resolve doubt – if any – as to the amount of the recoverable premium in favour of the paying party?
(5) Should the defendant be liable for any part of the premium which covers adverse costs orders and all aspects of own costs, however incurred?
(6) The success fee (“uplift”): is the full uplift recoverable given the stage when the CFA was effected?
(7) Was the 20 per cent uplift a reasonable maximum for simple accident cases?
(8) Was the Court of Appeal’s assessment of 20 per cent uplift for the instant case wrong?
| (9) Should an uplift be assessed by reference to the risk in the particular case, or on a more general, and if so what, basis?
(10) Should uplift levels be conditioned, and if so how, by public interest considerations?”
Issues (1) to (5) relate to the ATE premium; issues (6) to (10) relate to the success fee.
The ATE premium
Issue (1) – The ATE policy and prematurity
106. The prematurity issue was dealt with at  1 WLR 2112, 2130-2134, in paragraphs 80 to 100 of the first Court of Appeal judgment in relation both to the time at which a claimant could reasonably enter into a CFA and to the time at which he could reasonably take out ATE insurance. The Court of Appeal’s approach to the prematurity issue is expressed at p 2132, para 91 of the judgment:
107. I agree with the Court of Appeal’s proposition that it is reasonable for a claimant to enter into a CFA with his solicitor at their first meeting and before the defendant’s reaction to the claim is known. It was not contended before your Lordships that Mr Callery instructed Amelans prematurely. It was plainly reasonable for him to instruct solicitors to pursue his claim and for them then to notify Mr Gray of his claim. And it was reasonable for him, having instructed Amelans, immediately to enter into a CFA with regard to their fees. After all, the fees clock begins ticking as soon as a solicitor is instructed. The issues for your Lordships regarding the success fee relate, in my opinion, not to the time when the CFA was entered into but to the size of the success fee and the manner in which its reasonableness should be assessed.
108. It is otherwise with an ATE policy. The main issue, indeed I think the only real issue before the House, is whether it is reasonable, in a cost assessment context, for a claimant to take out an ATE policy at a time when litigation is highly unlikely. The purpose of an ATE policy is to protect the claimant against adverse costs orders. But the risk of such orders cannot arise unless and until the claimant commences litigation to pursue his damages claim. The CFA is needed as soon as he instructs his solicitors. But the ATE policy is not needed unless there is going to be litigation.
109. The circuit judge, Judge Edwards, interpreted CGU’s reply of 19 May 2000 to Amelans’ letter of 4 May as a denial of liability. This led him to conclude that it had been reasonable for Mr Callery to take out ATE insurance cover at an early stage “because a contest was looming”. The Court of Appeal, rightly in my opinion, did not accept that view of the letter of 19 May. They commented, at  1 WLR 2112, 2139-2140,in para 131 of the first judgment:
110. I would respectfully agree with this analysis of the letter of 19 May and how matters stood after it had been received. But the content of the letter was not the reason why the ATE policy was taken out. The policy had been taken out well before Amelans’ receipt of the 19 May letter. Indeed, it was taken out before either Mr Gray or CGU could have received the letter of 4 May notifying them of the claim.
111. It is apparent, therefore, that the Court of Appeal’s conclusion that it had been reasonable for Mr Callery to take out the ATE insurance policy on 4 May immediately on instructing Amelans was not based on any perception that, on what was then known, the policy was a reasonable precaution to be put in place in order to protect Mr Callery against the possibility of an adverse costs order in litigation. No one in Amelans was contemplating that litigation might be needed in order to obtain the payment of a satisfactory sum of damages for Mr Callery; it was not completely impossible that litigation might be needed, but it was very unlikely indeed. On what then was the Court of Appeal’s conclusion based? It seems to me to have been based on the evidence placed before the court about the ATE insurance market and the Court of Appeal’s concern that unless premium recovery under costs orders were allowed in such commonplace, minimal risk cases as Mr Callery’s, the market in ATE insurance policies might wither. The Court of Appeal was told that if claimants in road traffic accident cases postponed taking out cover until a litigation contest seemed likely, the premiums charged would rise steeply, putting at risk the ability of those who needed the cover in risky or speculative litigation to obtain it and thereby impeding their access to justice. The Court referred, in particular, to evidence from Mr Ward, managing director of Abbey Legal Protection, whose firm ran the Law Society’s Accident Line Protect conditional fee insurance scheme. Under this scheme, nominated solicitors were given delegated authority to offer clients ATE insurance cover once agreement on a CFA had been reached. But, in order to ensure that the principle of “the many paying for the few” was observed, solicitors who wished to offer the Accident Line Protect cover to their clients had to do so at the time the CFA was entered into and before notifying the proposed defendant, or his insurers, of the claim (see p 2129, para 70 of the first Court of Appeal judgment commenting on Mr Ward’s evidence). As to the policy of Temple, Mr Callery’s insurers, paragraph 74 of the first judgment said this:
112. It may be noticed how, in the ATE insurance context, the “many pay for the few” insurance principle has become distorted. The principle means that the “many”, all those who take out policies and pay premiums but many of whom do not suffer an insured risk, build up a premium fund by means of which the “few”, who likewise take out policies and pay premiums but who do suffer an insured risk, can have their insured losses met. In the ATE insurance field, however, the “many”, ie the defendants who are to be called upon to meet the premiums that build up the premium fund from which the insured losses of the few can be met, do not take out policies at all. The premium fund is not built up for their benefit or protection. It has been suggested that defendants, as well as claimants, benefit from ATE insurance because, in the event of costs orders in favour of a defendant being made, the claimant’s insurance cover will provide a means by which this liability can be met. In view of the number of exclusions likely to be found in a typical ATE policy (see paragraphs 1, 2, 7 and 8 of the exclusions clause in Mr Callery’s policy), I doubt whether this will be of much practical comfort to defendants. More importantly, the policy will come into play only in the event of litigation and in a case, such as the present, where the prospect of litigation was always highly remote, the suggested benefit to the defendant seems to me theoretical and illusory.
113. Your Lordships cannot know what will be the consequence for the ATE insurance market if premiums payable under ATE policies that have been taken out where there is no real risk of litigation, and therefore no real risk of the occurrence of the event insured against, are ruled to be irrecoverable from defendants. I would for my part be prepared to accept that, if recovery of premiums is restricted to those premiums paid in respect of policies taken out where there is at least a fair likelihood that litigation in pursuit of the damages claim will be needed, the size of premiums will rise. But I would certainly not be prepared to accept that cover will be unavailable. In any event, however, it is, in my opinion, contrary to principle for the reasonableness of the premature taking out of ATE insurance to be judged by reference to arguments about the impact on the ATE insurance market if recovery of premiums in commonplace cases such as Mr Callery’s is not allowed.
114. The correct approach for costs assessment purposes to the question whether an item of expenditure by the receiving party has been reasonably incurred is to look at the circumstances of the particular case. The question whether the paying party should be required to meet a particular item of expenditure is a case specific question. It is not a question to which the macro economics of the ATE insurance market has any relevance. If the expenditure was not reasonably required for the purposes of the claim, it would, in my opinion, be contrary to long-established costs recovery principles to require the paying party to pay it.
115. Section 29 of the 1999 Act allows recovery of ATE insurance premiums: “…. the costs payable to him may … include costs in respect of the premium ….” (my emphasis). The section does not mandate the inclusion of costs in respect of the premium. And whether a particular ATE insurance premium, or part of it, should be included in the costs ordered to be paid in a particular case should be tested and answered by reference to the same principles of reasonableness that apply under the rules and practice directions to all other items of expenditure.
116. There is, I think, no dispute that the likelihood of litigation being necessary in order to pursue Mr Callery’s damages claim was always very remote. Amelans, with their experience in the field, knew this. They offered ATE insurance to their client on 4 May, before notifying the defendant of the claim and obtaining his reaction to it, not on the footing that their client would be likely to need it, but because it had become their practice to do so. The offer was no doubt made on the footing that it would not be he who would have to pay the premium.
117. The argument advanced by the respondent, both in the Court of Appeal and before your Lordships, is, in effect, that defendants to claims which are bound to succeed and where litigation to pursue the claims is highly unlikely, must bear the cost of ATE insurance cover that is not needed in order to enable ATE insurance cover in cases where it is needed to be offered by insurers at lower rates than would otherwise be commercially possible. The clear principles on which costs recovery is based, both before and after the 1999 Act, are in my opinion hostile to this argument. In my opinion, the Court of Appeal fell into error in accepting it and in adopting an approach to the ATE premium that was not case specific but was based on the evidence about the ATE insurance market to which I have referred.
118. Submissions made to the House on behalf of the Lord Chancellor’s Department have contended that “access to justice would be restricted if claimants could not insure against liability for costs from the point they instructed a solicitor” (para 13 of the written case). This, in my opinion, misses the point. There is nothing to prevent claimants from taking out ATE policies as soon as they instruct a solicitor. The question is whether this is a reasonable step to take vis-à-vis the defendant who they expect to bear the burden of paying the premium. If there is no real risk of litigation or of adverse costs orders, the absence of an ATE policy will not discourage access to justice for the claimant in question. If, despite the absence of any such real risk, the claimant wants to take out the policy, he can do so but cannot then reasonably expect the defendant to pay for it. To describe this state of affairs as restricting access to justice seems to me unwarranted.
119. Paragraph 13 of the Lord Chancellor’s case continues by referring to:
120. The paragraph does not say that it is the Department’s policy objective that claimants would ordinarily be able to recover ATE premiums even though there was never any likelihood that litigation in which adverse costs orders might be made would be necessary. If that is or was the Department’s policy, I would expect your Lordships to be unimpressed by it. In any event, it is Parliament’s policy underlying section 29, not the Department’s policy, that the House should take account of and, as to that, there is nothing to lead one to suppose that expenditure on ATE premiums was, for costs purposes, to be approached differently from any other item of potentially recoverable expenditure.
121. Moreover, if general policy considerations are to be brought into account, they tend strongly, in my opinion, against allowing recovery of this ATE premium. One of the main purposes of Lord Woolf’s civil justice reforms was to reduce the cost and the complexity of pursuing legal remedies and an important step in that direction was the introduction of pre-action protocols. These set out steps which parties are intended to take before the commencement of litigation so as to confine litigation, as a means of dispute resolution, to cases where the dispute cannot be resolved without it. Under pre-action protocols the claimant must write a letter to the proposed defendant giving the outline of the claim and an opportunity to the defendant to respond. Various key documents, eg a medical report in accident cases, should be supplied to the defendant. This correspondence enables the parties to ascertain what, if anything, is in issue between them and whether, or on what issues, litigation will be necessary. If litigation is shown to be unnecessary, because the defendant admits the claim and the parties can agree terms of settlement, the expense of unnecessary litigation is, by the parties’ observance of the pre-action protocol requirements, avoided. And if a claimant fails to observe the requirements of the relevant pre-action protocol and commences litigation prematurely, the court may, even if his claim succeeds, deprive him of costs or order him to pay the defendant’s costs (see paras 2.2 and 2.3 of the Protocol Practice Direction: Civil Procedure, 2000, vol 1, C1-002, p 1303).
122. In the present case, Amelans’ letter of 4 May 2000 corresponded with the relevant pre-action protocol requirements. So did CGU’s response on 19 May. So did their ensuing correspondence which led, by early August, to a settlement agreement. Litigation was not necessary to pursue Mr Callery’s claim. The pre-action protocol had served its purpose. In these circumstances the proposition that the defendant should have to pay by way of costs for an item of Mr Callery’s expenditure that related exclusively to the litigation that the parties had succeeded in avoiding seems to me inconsistent with the purpose of the pre-action protocol. If during the pre-action protocol period the claimant incurs expenditure for the purpose of litigation that, in the event, never happens, there are, in my opinion, very sound policy reasons, consistent with those that underlie the civil justice reforms, for leaving the claimant to bear the cost of what, in the event, is useless expenditure.
123. For all these reasons I would give the answer “No” to the question posed in issue (1) and allow the appeal to that extent.
Issues (2), (3) and (4)
124. Issues (2) and (3) are fact dependent reasonableness issues. Your Lordships should not, in my opinion, second guess the courts below on issues such as these unless their decisions are plainly unreasonable or unless some evident error has been made. In the present case, the Court of Appeal, in the second judgment,  1 WLR 2142, 2159, para 70, concluded that the amount of the premium was not manifestly disproportionate to the risk being under taken and was reasonable. But in reaching this conclusion the Court of Appeal was proceeding under a misapprehension as to the basis on which the premium had been calculated. The court thought that Temple had fixed the amount of the premium, £350, having regard to the facts of this particular case. Counsel are agreed that this was an error. The £350 premium was in fact a uniform premium charged by Temple for ATE insurance cover in respect of every claim which carried a prospect of success of better than 50 per cent (see para 22 of the second judgment). This was a “block rating” case, not an individual assessment case.
125. The factual basis, therefore, on which the Court of Appeal concluded that the £350 premium was reasonable in amount was erroneous. The Court of Appeal did not address the question whether in a case, like the present, where the percentage prospect of success is at the least in the mid 90s, it was reasonable to charge a premium calculated on a block basis for all cases with a percentage prospect of success of better than 50. Your Lordships do not know what answer the Court of Appeal would have given. It is at this point that issue (4) becomes relevant.
126. The injunction in CPR 44.4 (2)(b) that requires the court to resolve doubts about reasonableness or proportionality in favour of the paying party does not mean that in every case where a decision as to reasonableness or as to proportionality is a difficult one to reach, with something to be said on each side, the paying party must win. The costs judge must endeavour to reach a conclusion on the issue, taking account of all the circumstances and of the factors that the rules and practice directions require to be taken into account. If, having done so, the costs judge remains uncertain whether the item of costs under review is reasonable and proportionate, then the paying party should win.
127. In the present case, where a block rate calculated for use in cases with a prospect of success of above 50 per cent, has been applied to a case with a prospect of success of well over 90 per cent, there cannot, in my opinion, be other than doubt as to the reasonableness of the charge. It might be over-simplistic to conclude that a charge reasonable for a case with a prospect of success of, say, 70 per cent, must be too high for a case with a prospect of success of, say, 95 per cent. But at least there must be doubt. So CPR 44.4 (2)(b) comes into play and requires that the doubt be resolved in favour of the paying party. The £350 premium, and the £17.50 tax, should, in my opinion, have been disallowed on this ground as well.
128. This issue relates to adverse costs orders made against a claimant where (a) the claimant has failed to beat a CPR Part 36 offer; or (b) the claimant has lost on some discrete issue and been ordered to pay the costs of that issue; or (c) there has been some other adverse exercise of costs discretion in the defendant’s favour. It is contended that recovery of the part of an ATE premium that relates to adverse costs orders of the sort described ought not to be allowed. It is put almost as a point of public policy. The burden of adverse costs orders that reflect the court’s disapprobation of some aspect of the manner in which a claimant has conducted his case ought not to be shifted to the defendant.
129. Counsel for the appellant accepts that a premium, or more likely part of a premium, paid to protect a claimant against the risk of liability under these adverse costs orders is within the scope of section 29 but submits that it cannot be reasonable to require the defendant, who has presumably been unsuccessful on the main issue in the case and who has had a normal order for costs made against him, to reimburse the claimant for that premium, or that part.
130. If it is reasonable for a claimant to take out ATE insurance cover in order to protect himself against incurring a costs liability in litigation, whether in respect of the other side’s costs or his own costs, I can, for my part see no reason why any distinction should be drawn between the types of adverse costs orders that might be made or the circumstances in which the orders might be made. If the claimant becomes entitled to costs and the expenditure has been reasonably incurred and is reasonable in amount, the expenditure, in my opinion, ought in principle to be included in the costs to be paid.
131. In my view, for the reasons already expressed, the expenditure on the ATE premium was not reasonably incurred. If, on the other hand, it was reasonably incurred, and is reasonable in amount, I do not think there is any principle that could justify subtracting from the recoverable premium some part of it notionally attributable to the risk of the adverse costs orders described.
The Success Fee
132. The full contractual uplift of 60 per cent was reduced by the district judge to 40 per cent and then further reduced by the Court of Appeal to 20 per cent. The full uplift was held to be not recoverable. There has been no cross appeal on the point and this issue does not, therefore, arise.
Issues (7) and (8)
133. These issues are fact dependent reasonableness issues. Your Lordships should not, in my opinion, interfere with the decision of the Court of Appeal that 20 per cent represented a reasonable uplift. The appellant has submitted that the 20 per cent was unduly high for “a very, very low risk case”. I think this may be so, but it is not, in my opinion, sufficiently out of line to justify your Lordships in interfering. It is reasonable, for the reasons I have given, for a would-be claimant to enter into a CFA with his solicitors as soon as he instructs them. The amount of the success fee must be stated in the CFA and, therefore, fixed before it is possible for the solicitors to know for certain what part, if any, of their client’s story the defendant will accept.
134. I am in no doubt but that a success fee should be assessed by reference to the risk in the particular case. As I have said already, a costs assessment process should be case specific. I agree with the appellant that the costs rules and practice directions reflect a case specific approach. If the risk of a claim failing is minimal then, in my opinion, the success fee should be correspondingly low.
135. This is too general a question to enable a useful answer to be given. I would simply repeat that, in my opinion, the reasonableness of a success fee is a case specific issue.
|136. I would allow the appeal to the extent of disallowing recovery by the respondent of the £350 premium and £17.50 tax.