Arthur J.S Hall and Co. v. Simons and Barratt v. Ansell and Others v. Scholfield Roberts and Hill [2000] UKHL 38 (20 July 2000)

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just £1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	United Kingdom House of Lords Decisions
You are here: BAILII >> Databases >> United Kingdom House of Lords Decisions >> Arthur J.S Hall and Co. v. Simons and Barratt v. Ansell and Others v. Scholfield Roberts and Hill [2000] UKHL 38 (20 July 2000) URL: http://www.bailii.org/uk/cases/UKHL/2000/38.html Cite as: [2000] UKHL 38, [2000] Fam Law 806, [2002] 1 AC 615, [2000] 2 FLR 545, [2000] 2 FCR 673, [2002] AC 615, [2000] 3 All ER 673, [2000] 3 WLR 543, [2000] ECC 487, [2000] BLR 407

Arthur J.S Hall and Co. v. Simons and Barratt v. Ansell and Others v. Scholfield Roberts and Hill [2000] UKHL 38; [2000] 3 All ER 673; [2000] 3 WLR 543; [2000] 2 FLR 545; [2000] 3 All ER 673; [2000] Fam Law 806 (20th July, 2000)

HOUSE OF LORDS

Lord Browne-Wilkinson Lord Steyn Lord Hoffmann Lord Hope of Craighead Lord Hutton Lord Hobhouse of Woodborough Lord Millett

OPINIONS OF THE LORDS OF APPEAL FOR JUDGMENT

IN THE CAUSE

There are three appeals before the House from orders of the Court of Appeal in a building case and in two cases involving family proceedings. Clients raised claims in negligence against firms of solicitors. In response the solicitors relied on the immunity of advocates from suits in negligence. In all three cases judges at first instance ruled that the claims against the solicitors were unsustainable. The circumstances of these cases and the disposals are set out in the judgment of the Court of Appeal given by Lord Bingham of Cornhill, L.C.J.: Arthur J.S. Hall & Co. (a firm) v. Simons [1999] 3 WLR 873. In effect the Court of Appeal ruled in all three cases presently before the House that the claims were wrongly struck out. The solicitors now appeal. The results of the appeals are of great importance to the parties. But transcending the importance of the specific issues arising on the appeals there are two fundamental general questions namely:

The position in Scotland was not the subject matter of argument on these appeals.

These questions before the House affect both branches of the legal profession. Your Lordships have had the benefit of careful arguments from three sides. First, by counsel for the appellant solicitors who were supported by the Solicitors Indemnity Fund. Secondly, by counsel for the Bar Council who was given leave to intervene and played a particularly helpful part in the appeal. Thirdly, by counsel for the individual litigants who put forward the contrary argument. Having studied the detailed written arguments and heard the oral arguments of counsel for the appellants, the intervenors, and the respondents, your Lordships are now in as good a position to form a judgment on the principal issues as is achievable.

It is necessary to explain the scheme of my opinion. There is a direct link between the two general questions. How the law deals with the problem of re-litigation of matters already decided, as identified in the Hunter case, is an important aspect of any re-consideration of the immunity of advocates. It will be necessary to examine the two issues together. Secondly, although the cases before the House involve actions against solicitors and not against barristers, the reality is that the immunity of barristers is of longer standing and underpinned to some extent by arguments not available to solicitors. It will therefore be convenient first to concentrate by and large on the position in regard to barristers and then to consider whether the conclusions arrived at also apply to solicitors.

For more than two centuries barristers have enjoyed an immunity from actions in negligence. The reasons for this immunity were various. It included the dignity of the Bar, the "cab rank" principle, the assumption that barristers may not sue for their fees, the undesirability of relitigating cases decided or settled, and the duty of a barrister to the court: Roxburgh, "Rondel v. Worsley: The Historical Background" (1968) 84 L.Q.R. 178; and Roxburgh, "Rondel v. Worsley: Immunity of the Bar" (1968) 84 L.Q.R. 513. In 1967 when the House decided Rondel v. Worsley the dignity of the Bar was no longer regarded as a reason which justified conferring an immunity on advocates whilst withholding it from all other professional men. In Hedley Byrne & Co. Ltd. v. Heller & Partners Ltd. [1964] AC 465 the rule was established that irrespective of contract, if someone possessed of a special skill undertakes to apply that skill for the assistance of another person who relies upon such skill, a duty of care will arise: at pp. 502-503. The fact that the barrister did not enter into a contract with his solicitor or client ceased to be a ground of justification for the immunity. Nevertheless, in a unanimous decision the House in Rondel v. Worsley [1969] 1 AC 191 upheld the ancient immunity on considerations of "public policy [which are] not immutable:" at p. 227B, per Lord Reid. It is worth recalling that in that case the appellant had obtained the services of the respondent to defend him on a dock brief, and alleged that the respondent had been negligent in the conduct of his defence. It is undoubtedly right, as counsel for the solicitors submitted and nobody disputed, that the principal ground of the decision is the overriding duty of a barrister to the court. The House thought that the existence of liability in negligence, and indeed the very possibility of making assertions of liability against a barrister, might tend to undermine the willingness of barristers to carry out their duties to the court. Lord Morris of Borth-y-Gest encapsulated the core idea by saying (at p. 251D): "It would be a retrograde development if an advocate were under pressure unwarrantably to subordinate his duty to the court to his duty to the client." Other members of the Appellate Committee expressed similar views: see p. 231E, per Lord Reid; pp. 272B-273F, per Lord Pearce; pp. 283E-283G, per Lord Upjohn; and p. 293E, per Lord Pearson. This factor is the pivot on which in 1967 the existence of the immunity hinged. But for it the case would probably have been decided differently. There were however supporting reasons. Perhaps the most important of these was the undesirability of relitigating issues already decided: see p. 230B-F, per Lord Reid and pp. 249A-250B, per Lord Morris of Borth-y-Gest. Another factor to which some weight was attached was the "cab rank" rule, which imposed (and still imposes) upon barristers, but not solicitors, the obligation to accept instructions from anyone who wishes to engage their services in an area of the law in which they practised. In the year after Rondel v. Worsley was decided Sir Ronald Roxburgh (formerly Mr. Justice Roxburgh) said that "the pressures for putting barristers on the same footing as other professional men . . . are already strong, and may grow stronger:" 84 L.Q.R. 513, 527.

Eleven years later in Saif Ali v. Sydney Smith Mitchell & Co. [1980] AC 198 the House revisited this topic. On this occasion the immunity established in Rondel v. Worsley was not challenged and was not directly in issue. The existence of the debate on the merits of the immunity was not re-opened. The terrain of the debate centred on the scope of the immunity. Except for Lord Diplock, the members of the House accepted the rationale of Rondel v. Worsley, which Lord Wilberforce said, at p. 213C, was that "barristers . . . have a special status, just as a trial has a special character: some immunity is necessary in the public interest, even if, in some rare cases, an individual may suffer loss." About a barrister's overriding duty to the court Lord Diplock observed, at p. 220C-E:

Lord Diplock did, however, think that the immunity could be justified on two other grounds. The first is the analogy of the general immunity from civil liability which attaches to all persons in respect of the participation in proceedings before a court of justice, namely judges, court officials, witnesses, parties, counsel and solicitors alike: p. 222A-C: The second was the public interest in not permitting decisions to be challenged by collateral proceedings: pp. 222D-223D. There matters rested for a time.

The next development was the introduction by statute of a power enabling the court to make wasted costs orders against legal practitioners: see section 51 of the Supreme Court Act 1981 as substituted by section 4 of the Courts and Legal Services Act 1990. Not surprisingly barristers are occasionally guilty of wholly unjustifiable conduct which occasions a waste of expenditure. The Bar argued that because of the immunity of barristers no such orders ought in principle to be made against barristers. The Court of Appeal ruled to the contrary: Ridehalgh v. Horsefield [1994] Ch 205. And that decision was accepted by the Bar. It operates satisfactorily. It has not been detrimental to the functioning of the court system or indeed the interests of the Bar.

As Roxburgh predicted in 1968 the pressure for a re-examination of Rondel v. Worsley mounted. There has been considerable academic criticism of the immunity. In a detailed and balanced discussion Peter Cane (Tort Law and Economic Interests, 2nd ed. (1996), pp. 233-238) found that, even taken together, the justifications adduced for the immunity do not support it strongly: see also similar effect Jonathan Hill, "Litigation and Negligence: A Comparative Study," (1986), 6 Oxford J.L.S. 183, 184-186. In an area where one is bound to a considerable extent to rely on intuitive judgments, the criticism of the immunity by two outstanding practising barristers is significant. In Advocates, 1992, pp. 197-206. Mr. David Pannick examined the case for and against the immunity in detail. While accepting that there is some substance in some of the arguments for an immunity, he found that on balance the immunity is not justified. He added, at p. 206: "This issue will not go away. English law will, in the future, have more to say on this topic." Recently, Sir Sydney Kentridge Q.C. expressed the view, making use of his experience as an advocate in South Africa and in England, that the "gloomy speculations" on which the immunity of barristers in England is based are wide off the mark: see Tortious Liability of Statutory Bodies, ed. Basil Markesinis and others, (1999), Foreword, p. ix. But even more important are the observations in the present case by Lord Bingham of Cornhill C.J., Morritt L.J. and Waller L.J. They clearly considered that, while the principle against collateral challenge as enunciated in the Hunter case ought to be maintained, nevertheless there was a substantial case for the sceptical re-examination of the immunity of barristers.

It is now possible to take stock of the arguments for and against the immunity. I will examine the relevant matters in turn. First, there is the ethical "cab rank" principle. It provides that barristers may not pick and choose their clients. It binds barristers but not solicitor advocates. It cannot therefore account for the immunity of solicitor advocates. It is a matter of judgment what weight should be placed on the "cab rank" rule as a justification for the immunity. It is a valuable professional rule. But its impact on the administration of justice in England is not great. In real life a barrister has a clerk whose enthusiasm for the unwanted brief may not be great, and he is free to raise the fee within limits. It is not likely that the rule often obliges barristers to undertake work which they would not otherwise accept. When it does occur, and vexatious claims result, it will usually be possible to dispose of such claims summarily. In any event, the "cab rank" rule cannot justify depriving all clients of a remedy for negligence causing them grievous financial loss. It is "a very price to pay for protection from what must, in practice, be the very small risk of being subjected to vexations litigation (which is, anyway, unlikely to get very far):" Cane, at p. 236. Secondly, there is the analogy of the immunities enjoyed by those who participate in court proceedings: compare however Cane's observation about the strength of the case for removing the immunity from paid expert witnesses: at p. 237. Those immunities are founded on the public policy which seeks to encourage freedom of speech in court so that the court will have full information about the issues in the case. For these reasons they prevent legal actions based on what is said in court. As Pannick has pointed out this has little, if anything, to do with the alleged legal policy which requires immunity from actions for negligent acts: ibid, at p. 202. If the latter immunity has merit it must rest on other grounds. Whilst this factor seemed at first to have some attractiveness, it has on analysis no or virtually no weight at all.

The third factor is the public policy against re-litigating a decision of a court of competent jurisdiction. This factor cannot support an immunity extending to cases where there was no verdict by the jury or decision by the court. It cannot arguably justify the immunity in its present width. The major question arises in regard to criminal trials which have resulted in a verdict by a jury or a decision by the court. Prosecuting counsel owes no duty of care to a defendant: Elguzouli-Daf v. Commissioner of Police of the Metropolis [1995] QB 335. The position of defence counsel must however be considered. Unless debarred from doing so, defendants convicted after a full and fair trial who failed to appeal successfully, will from time to time attempt to challenge their convictions by suing advocates who appeared for them. This is the paradigm of an abusive challenge. It is a principal focus of the principle in Hunter v. Chief Constable of the West Midlands Police [1982] AC 529. Public policy requires a defendant, who seeks to challenge his conviction, to do so directly by seeking to appeal his conviction. In this regard the creation of the Criminal Cases Review Commission was a notable step forward. Recently in Reg. v. Secretary of State for the Home Department, Ex parte Simms [1999] 3 WLR 328, 338, there was uncontroverted evidence before the House that the Commission is seriously under-resourced and under-funded. Incoming cases apparently have to wait two years before they are assigned to a case worker. This is a depressing picture. The answer is that the functioning of the Commission must be improved. But I have no doubt that the principle underlying the Hunter case must be maintained as a matter of high public policy. In the Hunter case the House did not, however, "lay down an inflexible rule to be applied willy-nilly to all cases which might arguably be said to be within it:" Smith v. Linskills [1996] 1 W.L.R. 763, 769C-F per Sir Thomas Bingham, M.R. (now Lord Bingham of Cornhill) It is, however, prima facie an abuse to initiate a collateral civil challenge to a criminal conviction. Ordinarily therefore a collateral civil challenge to a criminal conviction will be struck out as an abuse of process. On the other hand, if the convicted person has succeeded in having his conviction set aside on any ground, an action against a barrister in negligence will no longer be barred by the particular public policy identified in the Hunter case. But, in such a case the civil action in negligence against the barrister may nevertheless be struck out as unsustainable under the new flexible Civil Procedure rules, 1999; rules 3.4(2)(a) and 24.2. If the Hunter case is interpreted and applied in this way, the principal force of the fear of oblique challenges to criminal convictions disappears. Relying on my experience of the criminal justice system as a presiding judge on the Northern Circuit and as a member of the Court of Appeal (Criminal Division), I do not share intuitive judgments that the public policy against re-litigation still requires the immunity to be maintained in criminal cases. That leaves collateral challenges to civil decisions. The principles of res judicata, issue estoppel and abuse of process as understood in private law should be adequate to cope with this risk. It would not ordinarily be necessary to rely on the Hunter principle in the civil context but I would accept that the policy underlying it should still stand guard against unforeseen gaps. In my judgment a barrister's immunity is not needed to deal with collateral attacks on criminal and civil decisions. The public interest is satisfactorily protected by independent principles and powers of the court.

The critical factor is, however, the duty of a barrister to the court. It also applies to every person who exercises rights of audience before any court, or who exercises rights to conduct litigation before a court: see sections 27(2A) and 28(2A) of the Courts and Legal Services Act 1990 as inserted by section 42 of the Access to Justice Act 1999. It is essential that nothing should be done which might undermine the overriding duty of an advocate to the court. The question is however whether the immunity is needed to ensure that barristers will respect their duty to the court. The view of the House in 1967 was that assertions of negligence would tend to erode this duty. In the world of today there are substantial grounds for questioning this ground of public policy. In 1967 the House considered that for reasons of public policy barristers must be accorded a special status. Nowadays a comparison with other professionals is important. Thus doctors have duties not only to their patients but also to an ethical code. Doctors are sometimes faced with a tension between these duties. Concrete examples of such conflicting duties are given by Ian Kennedy, Treat Me Right; Essays in Medical Law and Ethics, (1988). A topical instance is the case where an Aids infected patient asks a consultant not to reveal his condition to the patient's wife, general practitioner and other healthcare officials. Such decisions may easily be as difficult as those facing barristers. And nobody argues that doctors should have an immunity from suits in negligence.

Comparative experience may throw some light on the question whether in the public interest such an immunity of advocates is truly necessary. In 1967 no comparative material was placed before the House. Lord Reid did, however, mention other countries where public policy points in a different direction: [1969] 1 AC 191, 228E. In the present case we have had the benefit of a substantial comparative review. The High Court of Australia followed Rondel v. Worsley: Gianarelli v. Wraith (1988) 165 C.L.R. 543; see also Boland v. Yates Property Corporation Pty. Ltd. (1999) 74 A.L.J.R. 209. In New Zealand the Court of Appeal has taken a similar course: Rees v. Sinclair [1974] 1 N.Z.L.R. 180. It is a matter of significance that the High Court of Australia and the Court of Appeal of New Zealand came to the conclusion that a barristers immunity from actions in negligence is required by public policy considerations in those countries. On the other hand, in countries in the European Union advocates have no immunity. It is true that there is a difference in that the control of a civilian judge over the proceedings is greater than is customarily exercised by a judge in England: see R.O. Graef, Judicial Activism in Civil Proceedings. A comparison between English and German Civil Procedural Approaches, (1996), passim. But with the advent of the Woolf reforms this difference is reduced to some extent in civil cases: see The Civil Procedure Rules, 1999, Part 1, Para. 1.1 (The over-riding objective). On the other hand, I accept that in the field of criminal procedure the role of a judge in England is far more passive than in European Union countries: see Van Den Wyngaert and others, Criminal Procedure Systems in the European Community (1993), passim. I am also willing to accept that, although an advocate in a civilian system owes a duty to the court, it is less extensive than in England. For example, in Germany there is apparently no duty to refer the court to adverse authorities as in England. Despite these differences the fact that the absence of an immunity has apparently caused no practical difficulties in other countries in the European Union is of some significance: Tortious Liability of Statutory Bodies, ed. B.S. Markesinis and others, (1999), p. 80. In the United States prosecutors have an immunity. In a few states the immunity is extended to public defenders. But otherwise lawyers have no immunity from suits of negligence by their clients: Ferri v. Ackermann (1979) 444 U.S. 193. While the differences between the legal system of the U.S.A and our own must be taken into account, the United States position cannot be altogether ignored. In Canada an advocate had no immunity from an action in negligence before Rondel v. Worsley was decided. In 1979 the question was re-examined in great detail as a result of the decision of the House of Lords in Rondel v. Worsley: see Demarco v. Ungaro (1979) 95 D.L.R. (3rd) 385. In Canada trial lawyers owe a duty to the court. After a detailed and careful review the court found there was no evidence that the work of Canadian courts was hampered in any way by counsel's fear of civil liability. The Demarco case has been consistently followed by Canadian courts: see Karpenko v. Paroian, Courey, Cohen & Houston (1980), 30 O.R. (2d) 776 (H.C.); Pelky v. Hudson Bay Insurance Co. (1981), 35 O.R. (2d) 97 (H.C.); Garrant v. Moskal, [1985] 2 W.W.R. 80 (Sask. Q.B.), affirmed [1985] 6 W.W.R. 31 (Sask. C.A.); Hodge & Son v. Monoghan (1985), 51 Nfld. & P.E.I.R. 173 (Nfld. T.D.) I regard the Canadian empirically tested experience as the most relevant. It tends to demonstrate that the fears that the possibility of actions in negligence against barristers would tend to undermine the public interest are unnecessarily pessimistic.

There would be benefits to be gained from the ending of immunity. First, and most importantly, it will bring to an end an anomalous exception to the basic premise that there should be a remedy for a wrong. There is no reason to fear a flood of negligence suits against barristers. The mere doing of his duty to the court by the advocate to the detriment of his client could never be called negligent. Indeed if the advocate's conduct was bona fide dictated by his perception of his duty to the court there would be no possibility of the court holding him to be negligent. Moreover, when such claims are made courts will take into account the difficult decisions faced daily by barristers working in demanding situations to tight timetables. In this context the observations of Sir Thomas Bingham M.R. (now Lord Bingham of Cornhill) in Ridehalgh v. Horsefield [1994] Ch 205 are instructive. Dealing with the circumstances in which a wasted costs order against a barrister might be appropriate he observed, at p. 236:

For broadly similar reasons it will not be easy to establish negligence against a barrister. The courts can be trusted to differentiate between errors of judgment and true negligence. In any event, a plaintiff who claims that poor advocacy resulted in an unfavourable outcome will face the very great obstacle of showing that a better standard of advocacy would have resulted in a more favourable outcome. Unmeritorious claims against barristers will be struck out. The new Civil Procedure Rules, 1999, have made it easier to dispose summarily of such claims: rules 3.4(2)(a) and 24.2. The only argument that remains is that the fear of unfounded actions might have a negative effect on the conduct of advocates. This is a most flimsy foundation, unsupported by empirical evidence, for the immunity. Secondly, it must be borne in mind that one of the functions of tort law is to set external standards of behaviour for the benefit of the public. And it would be right to say that while standards at the Bar are generally high, in some respects there is room for improvement. An exposure of isolated acts of incompetence at the Bar will strengthen rather than weaken the legal system. Thirdly, and most importantly, public confidence in the legal system is not enhanced by the existence of the immunity. The appearance is created that the law singles out its own for protection no matter how flagrant the breach of the barrister. The world has changed since 1967. The practice of law has become more commercialised: barristers may now advertise. They may now enter into contracts for legal services with their professional clients. They are now obliged to carry insurance. On the other hand, today we live in a consumerist society in which people have a much greater awareness of their rights. If they have suffered a wrong as a result of the provision of negligent professional services, they expect to have the right to claim redress. It tends to erode confidence in the legal system if advocates, alone among professional men, are immune from liability for negligence. It is also noteworthy that there is no obligation on the barrister (or for that matter the solicitor advocate) to inform a client at the inception of the relationship that he is not liable in negligence, and in practice the client is never so informed. Given that the resort to litigation is often one of the most important decisions in the life of the client, it has to be said that this is not a satisfactory position. Moreover, conduct covered by the immunity is beyond the remit of the Legal Services Ombudsman: section 22(7)(b) of the Court's and Legal Services Act 1990. In combination these factors reinforce the already strong case for ending the immunity.

My Lords, one is intensely aware that Rondel v. Worsley [1969] 1 AC 191 was a carefully reasoned and unanimous decision of the House. On the other hand, it is now clear that when the balance is struck between competing factors it is no longer in the public interest that the immunity in favour of barristers should remain. I am far from saying that Rondel v. Worsley was wrongly decided. But on the information now available and developments since Rondel v. Worsley I am satisfied that in today's world that decision no longer correctly reflects public policy. The basis of the immunity of barristers has gone. And exactly the same reasoning applies to solicitor advocates. There are differences between the two branches of the profession but not of a character to differentiate materially between them in respect of the issue before the House. I would treat them in the same way.

That brings me to the argument that the ending of the immunity, if it is to be undertaken, is a matter for Parliament. This argument is founded on section 62 of the Courts and Legal Services Act 1990. It reads as follows:

The background to this provision is, of course, the judicially created immunity of barristers, which in 1967 was held by the House to be founded on public policy. And it will be recollected that Lord Reid observed that public policy is not immutable. Against this background the meaning of section 62 is clear. It provides that solicitor advocates will have the same immunity as barristers have. In other words, the immunity of solicitors will follow the fortunes of the immunity of barristers, or track it. Section 62 did not either expressly or by implication give Parliamentary endorsement to the immunity of barristers. In these circumstances the argument that it is beyond the power of the House of Lords, which created the immunity spelt out in Rondel v. Worsley, to reverse that decision in changed circumstances involving a different balance of policy considerations is not right. Should the House as a matter of discretion leave it to Parliament? This issue is more finely balanced. It would certainly be the easy route for the House to say "let us leave it to Parliament." On balance my view is that it would be an abdication of our responsibilities with the unfortunate consequence of plunging both branches of the legal profession in England into a state of uncertainty over a prolonged period. That would be a disservice to the public interest. On the other hand, if the decision is made to end the immunity now, both branches of the profession will know where they stand. They ought to find it relatively easy to amend their rules where necessary and to adjust their already existing insurance arrangements insofar as that may be necessary.

My Lords, the cards are now heavily stacked against maintaining the immunity of advocates. I would rule that there is no longer any such immunity in criminal and civil cases. In doing so I am quite confident that the legal profession does not need the immunity.

So far as the Hunter case involves a separate question before the House I would refer to my discussion of this topic under the heading of Immunity of Barristers.

Given the conclusion that the immunity no longer exists, it follows that the appeals must fail. I would dismiss the three appeals.

I have had the advantage of reading the speeches of my noble and learned friends Lord Steyn and Lord Hoffmann. I agree with them and for the reasons they give, I would dismiss these appeals. However, since the point at issue is important and your Lordships' views are not unanimous, I will state shortly my views on the point on which your Lordships are divided.

Let me initially consider the points on which your Lordships are all agreed. First that, given the changes in society and in the law that have taken place since the decision in Rondel v. Worsley [1969] 1 AC 191, it is appropriate to review the public policy decision that advocates enjoy immunity from liability for the negligent conduct of a case in court. Second, that the propriety of maintaining such immunity depends upon the balance between, on the one hand, the normal right of an individual to be compensated for a legal wrong done to him and, on the other, the advantages which accrue to the public interest from such immunity. Third, that in relation to claims for immunity for an advocate in civil proceedings, such balance no longer shows sufficient public benefit as to justify the maintenance of the immunity of the advocate.

The point on which your Lordships are divided is whether the same rules should apply whether the negligence alleged against the advocate relates to his conduct of a civil action or to a criminal prosecution. Are there, as some of your Lordships think, special reasons which require the immunity of the advocate in a criminal trial to be maintained? Of the four main grounds relied upon as justifying the immunity, only one seems to me to be capable of justifying the immunity, namely that to allow an action for negligence against the advocate for his conduct in earlier litigation is necessarily going to involve the risk that different conclusions on issues decided in the first case will be reached in the later case. In the context of civil proceedings (i.e. where the advocate is sought to be made liable for his conduct of a civil action) although such conflicting decisions are undesirable, they are far from unknown. But in the context of criminal proceedings (i.e. when the advocate's negligence occurred in the course of a criminal trial) the decision is far more difficult. In the overwhelming majority of cases, the action in negligence will not be capable of succeeding unless the verdict of guilty in the original trial is held to have been incorrect; if the complainant was in any event guilty of the alleged crime, the negligence of his advocate, even if proved, would not have been shown to be causative of any loss. Therefore, if there is to be a successful action for negligence in criminal matters, so long as the plaintiff's criminal conviction stands there will be two conflicting decisions of the court, one (reached by judge and jury on the criminal burden of proof) saying that he is guilty, the other (reached by a judge alone on balance of probability) that he is not guilty. My Lords, I would find such conflicting decisions quite unacceptable. If a man has been found guilty of a crime in a criminal trial, for all the purposes of society he is guilty unless and until his conviction is set aside on appeal. Therefore, if the removal of the advocate's immunity in criminal cases would produce these conflicting decisions, I would have no doubt that the public interest demanded that the advocate's immunity be preserved.

But in my judgment the law has already provided a solution where later proceedings are brought which directly or indirectly challenge the correctness of a criminal conviction. Hunter v. Chief Constable of the West Midlands Police [1982] AC 529 establishes that the court can strike out as an abuse of process the second action in which the plaintiff seeks to re-litigate issues decided against him in earlier proceedings if such re-litigation would be manifestly unfair to the defendant or would bring the administration of justice into disrepute. In view of the more restrictive rules of res judicata and issue estoppel it is not clear to me how far the Hunter case goes where the challenge is to an earlier decision in a civil case. But in my judgment where the later civil action must, in order to succeed, establish that a subsisting conviction is wrong, in the overwhelming majority of cases to permit the action to continue would bring the administration of justice into disrepute. Save in truly exceptional circumstances, the only permissible challenge to a criminal conviction is by way of appeal.

It follows that, in the ordinary case, an action claiming that an advocate has been negligent in criminal proceedings will be struck out as an abuse of process so long as the criminal conviction stands. Only if the conviction has been set aside will such an action be normally maintainable. In these circumstances there is no need to preserve an advocate's immunity for his conduct of a criminal case since, in my judgment, the number of cases in which negligence actions are brought after a conviction is quashed is likely to be small and actions in which the conviction has not been quashed will be struck out as an abuse of process.

For these reasons, and the much fuller reasons given by Lord Steyn and Lord Hoffmann, I would dismiss these appeals.

In these appeals three clients are suing their solicitors for negligence. In the first, Mr. Simons says that his solicitors negligently allowed him to become involved in lengthy and expensive litigation when they should have advised him to settle. In the second, Mr. Barratt says that his solicitors negligently advised him to settle his divorced wife's claim for a share of the matrimonial home on disadvantageous terms. In the third, Mrs. Harris has a similar complaint about the terms upon which her solicitors advised her to settle her claim for maintenance against her ex-husband. None of these allegations has been investigated. The solicitors may or may not have a complete answer to them. But they say that even if they were negligent, they cannot be sued. They claim immunity under a modern version of an ancient rule of common law which prevented barristers from being sued for negligence.

The old rule for barristers survived until 1967. The way in which it was usually explained was that barristers, unlike solicitors, had no contract with their clients. They could not sue for their fees. And in the absence of a contract there could be no liability. But that reason was undermined when the House of Lords decided in Hedley Byrne & Co. Ltd. v. Heller & Partners Ltd. [1964] AC 465 that, even without a contract, a person who negligently performed professional or other duties which he had undertaken could be sued in tort. So the whole question was re-examined by the House in Rondel v. Worsley [1969] 1 AC 191. What emerged was a different rule of immunity, in some respects wider and in others narrower, not based upon any technicalities but upon what the House perceived as the public interest in the administration of justice.

The new rule was narrower because, although their Lordships were not unanimous about its precise limits, they agreed that it should in general terms be confined to acts concerned with the conduct of litigation. None of them thought that it could apply to non-contentious work. Barristers had previously been immune from liability for anything. On the other hand, the new rule was wider in that it also applied to solicitors.

Most of the speeches in Rondel v. Worsley [1969] 1 AC 191 were devoted to explaining why the new immunity was necessary. The old cases had not relied solely upon the technicalities of contract. The rule was also said to be an expression of public policy. But Lord Reid said, at p. 227B-C, that public policy was "not immutable" and that because "doubts appear to have arisen in many quarters whether that rule is justifiable in present day conditions in this country" it was proper to "re-examine the whole matter." The grounds upon which their Lordships considered that public policy required a modified immunity may be summarised under four heads: divided loyalty, the cab rank, the witness analogy and collateral challenge.

Lawyers conducting litigation owe a divided loyalty. They have a duty to their clients, but they may not win by whatever means. They also owe a duty to the court and the administration of justice. They may not mislead the court or allow the judge to take what they know to be a bad point in their favour. They must cite all relevant law, whether for or against their case. They may not make imputations of dishonesty unless they have been given the information to support them. They should not waste time on irrelevancies even if the client thinks that they are important. Sometimes the performance of these duties to the court may annoy the client. So, it was said, the possibility of a claim for negligence might inhibit the lawyer from acting in accordance with his overriding duty to the court. That would be prejudicial to the administration of justice.

It is a valuable professional ethic of the English bar that a barrister may not refuse to act for a client on the ground that he disapproves of him or his case. Every barrister not otherwise engaged is available for hire by any client willing and able to pay the appropriate fee. This rule protects barristers against being criticised for giving their services to a client with a bad reputation and enables unpopular causes to obtain representation in court. It was said that barristers would be less inclined to honour this professional obligation if they suspected that the client was the sort of person who would, if he lost his case, turn on his barrister and sue for negligence. This consideration was said to apply with particular force to the criminal bar, where the unsuccessful client, like Mr. Rondel, was likely to have leisure to ponder the way his trial had been conducted and access to legal aid if he could persuade another lawyer that he had an arguable case.

No one can be sued in defamation for anything said in court. The rule confers an absolute immunity which protects witnesses, lawyers and the judge. The administration of justice requires that participants in court proceedings should be able to speak freely without being inhibited by the fear of being sued, even unsuccessfully, for what they say. The immunity has also been extended to statements made out of court in the course of preparing evidence to be given in court. So it is said that a similar immunity against proceedings for negligence is necessary to enable advocates to conduct the litigation properly.

If a client could sue his lawyer for negligence in conducting his litigation, he would have to prove not only that the lawyer had been negligent but also that his negligence had an adverse effect upon the outcome. This would usually mean proving that he would have won a case which he lost. But this gives rise to the possibility of apparently conflicting judgments which could bring the administration of justice into disrepute. A client is convicted and sent to prison. His appeal is dismissed. In prison, he sues his lawyer for negligence. The lawyer's defence is that he was not negligent but that, in any case, the client has suffered no injustice because whatever the lawyer did would not have secured an acquittal. In seeking to establish the latter point, the lawyer may or may not be able to re-assemble the witnesses who gave evidence for the prosecution. The question of whether the client should have been acquitted is then tried on evidence which is bound in some respects to be different, before a different tribunal and in the absence of the prosecution. The civil court finds, on a balance of probability, that the lawyer was negligent and that if he had conducted the defence with reasonable skill, the client would have been acquitted. Or perhaps that he would have had a 50% chance of being acquitted. Damages are awarded. But what happens then? Does the client remain in prison, despite the fact that a judge has said there was an even chance that he would have been acquitted? Should he be released, notwithstanding that the prosecution has had no opportunity to say that his conviction was correct? Should it be referred back to the Court of Appeal and what happens if the Court of Appeal, on the material before it, takes a different view from the civil judge? The public would not understand what was happening. So it was said that to allow clients to sue for negligence would allow a "collateral challenge" to a previous decision of another court. Even though the parties were different, this would be contrary to the public interest.

Eleven years later, after Rondel v. Worsley [1969] 1 AC 191, the House of Lords in Saif Ali v. Sydney Mitchell & Co. [1980] AC 198 had to consider the limits of the immunity. There was no challenge to the decision itself or the core immunity for the conduct of litigation in court. The question was the extent to which that immunity cast its shadow upon acts done out of court. In the particular case, it was a barrister's failure to advise joining additional parties before the limitation period had expired. The test for the out of court immunity adopted by the majority of the House was whether the work was so "intimately connected" with the conduct of the case in court as to amount to a decision as to how it would be conducted at the hearing. By this test, the barrister's conduct fell outside the immunity.

Although the immunity itself was not under challenge in Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, Lord Diplock considered that the need to delimit its scope required a reconsideration of its rationale. He was unimpressed by the divided loyalty argument which had been in the forefront of the reasoning in Rondel v. Worsley [1969] 1 AC 191. He thought no better of the cab rank rule. But he considered that the analogy with witness immunity and the collateral challenge argument were sufficient to justify a limited immunity.

In the cases now under appeal, the Court of Appeal was of course bound by Rondel v. Worsley [1969] 1 AC 191 and Saif Ali v. Sydney Mitchell & Co. [1980] AC 198. It decided that in all three cases the alleged negligence of the solicitors was not within the scope of the immunity as extended to out of court work. Their advice was not intimately connected with the way in which the case, if it had not settled, would have been conducted in court. But before your Lordships, the respondent clients have made a root and branch attack on the immunity. They say that it should be altogether abolished. Over 30 years have passed since Rondel v. Worsley [1969] 1 AC 191; public policy, as Lord Reid said at the time, is not immutable, and there have been great changes in the law of negligence, the functioning of the legal profession, the administration of justice and public perceptions. They say that it is once again time to re-examine the whole matter. My Lords, I agree. In reconsidering these questions, I have been greatly assisted by a wealth of writing on the subject by judges, practitioners and academics, in the United Kingdom and overseas. I hope that I will not be thought ungrateful if do not encumber this speech with citations. The question of what the public interest now requires depends upon the strength of the arguments rather than the weight of authority.

My Lords, my point of departure is that in general English law provides a remedy in damages for a person who has suffered injury as a result of professional negligence. The landmark cases by which this principle was developed are Hedley Byrne & Co. Ltd. v. Heller & Partners Ltd. [1964] AC 465, to which I have already referred, and Henderson v. Merrett Syndicates Ltd. [1995] 2 AC 145. It follows that any exception which denies such a remedy requires a sound justification. Otherwise your Lordships would fail to observe the fundamental principle of justice which requires that people should be treated equally and like cases treated alike.

In considering whether such a justification still exists, your Lordships cannot ignore the fact that you are yourselves members of the legal profession. Members of other professions, and the public in general, are bound to view with some scepticism the claims of lawyers that the public interest requires them to have a special immunity from liability for negligence. If your Lordships are convinced that there are compelling arguments for such an immunity, you should not of course be deterred from saying so by fear of unfounded accusations of collective self-interest. But those arguments need to be strong enough to convince a fair-minded member of the public. They cannot be based merely upon intuitions. This is a case in which what Professor Peter Cane has described as an "empathy heuristic" will not do. (See Oxford Essays in Jurisprudence, ed. J. Morder 4th series (2000) , p. 56, footnote 35.)

My Lords, there is apt to be a certain amount of confusion about the exact nature of the divided loyalty argument. There are two distinct versions in circulation but they are not always recognised to be different.

The first argument is that the possibility of being sued for negligence would actually inhibit the lawyer, consciously or unconsciously, from giving his duty to the court priority to his duty to his client, or, as Lord Diplock preferred to put it in Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, 219H from observing the rules. This argument involves a prediction about the way in which the removal of the immunity would affect the way in which lawyers behave in court. It claims that their behaviour would change in a way which was contrary to the public interest in the administration of justice. This was the argument advanced by Mr. Sumption to your Lordships on behalf of the defendant solicitors. He said that if there was no immunity, lawyers would in marginal cases prefer the interests of their clients to the interests of justice. It is an argument which in view of the eminence of its proponents in Rondel v. Worsley [1969] 1 AC 191 and elsewhere must be taken seriously I shall in due course return to it.

The second version of the argument is that the divided loyalty is a special factor that makes the conduct of litigation a very difficult art. It is easy to commit what appear in retrospect to have been errors of judgment. Even if there is no real danger that a court would hold such errors to have been negligent, the advocate would be exposed to vexatious claims by difficult clients. The argument is pressed most strongly in connection with advocacy in criminal proceedings, where the clients are said to be more than usually likely to be vexatious. Your Lordships will observe that this version of the argument does not depend upon the proposition that lawyers will be deterred from observing the rules or their duty to the court. It is advanced as a good argument even if your Lordships think that there are no sufficient grounds for the prediction which Mr. Sumption invites you to make. It is rather an argument that the imposition of liability would be unfair. The efforts of lawyers in good faith to comply with their public duties should not leave them open to vexatious claims by dissatisfied clients. This is the argument which my noble and learned friend Lord Hutton calls the "second strand" of the divided loyalty argument. As he puts it, "it is not right that a person performing an important public duty by taking part in a [criminal] trial should be vexed by an unmeritorious action . . . ." I shall deal with this argument, which I propose to call the "vexation argument," before returning to the one advanced by Mr. Sumption.

My Lords, I do not think that the vexation argument, taken by itself, has any validity. It is true that the conduct of litigation is a difficult art and that one of the reasons why it sometimes requires delicate judgment is the advocate's duty to the court. But there are many professional activities which require delicate judgment and advocacy is not the only one which may involve a divided loyalty. The doctor, for example, owes a duty to the individual patient. But he also owes a duty to his other patients which may prevent him from giving one patient the treatment or resources he would ideally prefer. We do not say that they should have immunity merely because they do a difficult job in which it is easy to make a bona fide error of judgment. And although the criminal advocate is engaged in an activity of great public importance, I do not think it would be right to claim that he is in this respect unique among professional men. The fact is that the advocate, like other professional men, undertakes a duty to his client to conduct his case, subject to the rules and ethics of his profession, with proper skill and care. No other participant in the trial undertakes such a duty.

There is some overlap between the vexatious claims argument and the witness analogy, to which I shall come in due course. Essentially it depends upon the same reasoning as Fry L.J. used in the famous passage in Munster v. Lamb (1883) 11 QBD 588, 607 in defence of the absolute privilege of witnesses giving evidence in court:

But this argument depends upon the assumption that there is a powerful public interest which makes this degree of protection necessary. In the case of witnesses, it is the assumption that they would otherwise be less willing to come forward and tell the truth in court. In other words, that their behaviour would be affected in a particular way which was contrary to the interests of the administration of justice. It is not simply the general proposition that people doing their best in a difficult job should not be exposed to vexatious claims. This argument could apply to many people besides lawyers. So in my opinion it is only the first version of the divided loyalty argument which can have any prospect of success. The second is in principle misconceived.

Before returning to Mr. Sumption's divided loyalty argument, I should say that in my opinion one should not exaggerate the bogey of vexatious claims. As I have said, every other profession has to put up with them. A practitioner who is properly insured can usually expect such claims to be handled by solicitors instructed by the underwriters. And there have been recent developments in the civil justice system designed to reduce the incidence of vexatious claims.

The first is the new Civil Procedure Rules. Under the old rules, a defendant faced with what appeared to be a bad claim had a very heavy burden to satisfy the court that it was "frivolous and vexatious" and ought to be struck out. Now rule 24.2 provides that the court may give summary judgment in favour of a defendant if it considers that "the claimant has no real prospect of succeeding on the claim." The defendant may file written evidence in support of his application. In Swain v. Hillman The Times, 4 November 1999; Court of Appeal (Civil Division) Transcript No. 1732 of 1999 Lord Woolf M.R. encouraged judges to make use of this:

Of course the summary power has its limits. The court should not "conduct a mini-trial" when there are issues which should be considered at a full one. But it should enable the courts to deal summarily with truly vexatious proceedings. It should also be remembered that a lawyer defendant has the advantage that the power of summary dismissal is in the hands of lawyers. I do not suggest that they would be inclined to favour their own profession. The opposite is more likely to be the case. But they would understand what the case was about. They would be operating in their own field of expertise, not faced with the allegations of professional negligence in another discipline which they did not have the knowledge or experience to recognise as groundless. So in this respect lawyers faced with vexatious claims are in an advantageous position.

The second important change has been made by the Access to Justice Act 1999, which came into force on 1 April 2000. Civil legal aid has been abolished and replaced by legal services funded by the Legal Services Commission as part of the Community Legal Service. The Act altogether excludes legal help in relation to "allegations of negligently caused injury, death or damage to property...:" see paragraph 1(a) of Schedule 2. Although an action for damages for loss caused by negligent advocacy or related services may not strictly fall within these categories, it is clear that it will not be easy to obtain legal representation for such actions. The Lord Chancellor has approved a Funding Code prepared by the Commission under section 8 of the Act which indicates that they would not come very high on the Community Legal Service's scale of priorities. Paragraph 5.7.1 of the code provides that if "the nature of the case is suitable for a [conditional fee agreement], and the client is likely to be able to avail himself or herself of a [conditional fee agreement], full representation will be refused." Actions for damages for the negligent conduct of litigation would seem, by analogy with paragraph 1(a) of Schedule 2, to be suitable for conditional fee agreements. Furthermore, under paragraph 5.7.3, full representation in a claim for damages will be refused unless certain cost benefit criteria are satisfied. For example, if the chances of success are good (60% - 80%), the likely damages must exceed the likely costs by a ratio of 2:1. If the prospects are less than 50%, representation will be refused.

It will therefore be much more difficult than it has been in the past to obtain legal help for negligence actions which have little prospect of success. The public funding of cases like Rondel v. Worsley [1969] 1 AC 191, the very paradigm of a hopeless claim by a disgruntled criminal defendant, is unlikely to be repeated. The alternative will be a conditional fee agreement, which would require satisfying another lawyer that the claim had sufficient prospects to make it worth his while to take it on at his own risk as to costs. Once again, as a lawyer, he will be able to recognise a vexatious claim when he sees one.

After this digression, I return to Mr. Sumption's divided loyalty argument. I have no doubt that the advocate's duty to the court is extremely important in the English system of justice. The reasons are eloquently stated by their Lordships in Rondel v. Worsley [1969] 1 AC 191 and I do not think that the passage of more than 30 years has diminished their force. The substantial orality of the English system of trial and appellate procedure means that the judges rely heavily upon the advocates appearing before them for a fair presentation of the facts and adequate instruction in the law. They trust the lawyers who appear before them; the lawyers trust each other to behave according to the rules, and that trust is seldom misplaced. The question is whether removing the immunity would have a significant adverse effect upon this state of affairs.

To assess the likelihood, I think that one should start by considering the incentives which advocates presently have to comply with their duty and those which might tempt them to ignore it. The first consideration is that most advocates are honest conscientious people who need no other incentive to comply with the ethics of their profession. Then there is the wish to enjoy a good reputation among one's peers and the judiciary. There can be few professions which operate in so bright a glare of publicity as that of the advocate. Everything is done in public before a discerning audience. Serious lapses seldom pass unnoticed. And in the background lie the disciplinary powers of the judges and the professional bodies. Whereas in 1967 it might have been said that the concept of the duty to the court was somewhat undefined and that much was left to the discretion of the advocate, who might interpret his obligations in the way which suited him best, today both branches of the profession are governed by detailed codes of conduct.

Looking at the other side of the coin, what pressures might induce the advocate to disregard his duty to the court in favour of pleasing the client? Perhaps the wish not to cause dissatisfaction which might make the client reluctant to pay. Or the wish to obtain more instructions from the same client. But among these pressures, I would not put high on the list the prospect of an action for negligence. It cannot possibly be negligent to act in accordance with one's duty to the court and it is hard to imagine anyone who would plead such conduct as a cause of action. So when the advocate decides that he ought to tell the judge about some authority which is contrary to his case, I do not think it would for a moment occur to him that he might be sued for negligence. I think it is of some significance that the situation in which the interests of the client and the duty to justice are most likely to come into conflict is in the preparation of the list of documents for discovery. The lawyer advising on discovery is obliged to insist that he disclose relevant documents adverse to his case which are not protected by privilege. But solicitors who undertake no advocacy usually perform this task and it has never been thought to be protected by immunity.

Mr. Sumption did not really suggest that any conscious calculation would take place. What he said was that it would lead to defensive lawyering, rather as liability for professional negligence is said to lead to defensive medicine. The advocate would take every possible point when otherwise he might have been willing to shorten the proceedings by conceding that some were really non-starters. But prolixity is a recognised problem even with the immunity in place. Lawyers want to do as much as they honestly can for their client and occasionally more. The tendency to overkill is not inhibited by the system under which they are conventionally paid, which is reasonable remuneration for work reasonably done. So the problem has to be contained in other ways. The disapproval of the court is a traditional curb on prolixity. But it has not been enough. Other mechanisms have had to be put into place. The new Civil Procedure Rules have given judges a battery of powers to keep the resources expended on a case proportionate to the its value and importance.

An important innovation for the purpose of restraining unnecessary expenditure on costs has been the extension in 1990 of the power of the court to make wasted costs orders. The implications of this jurisdiction are in my view so relevant to the present argument that the subject deserves a section of its own.

The judgment of the Court of Appeal in Ridehalgh v. Horsefield [1994] Ch 205, 226-231 contains a history of the wasted costs jurisdiction. Briefly stated, the court had jurisdiction before 1990 to order solicitors to pay costs wasted by their clients or other parties by reason of their misconduct, default or serious negligence. The jurisdiction did not apply to barristers. But section 4 of the Courts and Legal Services Act 1990 conferred power to make rules under which the court could order any legal representative to pay costs wasted by any party as a result of "any improper, unreasonable or negligent act or omission" on their part. Rules to this effect came into force on 1 October 1991: R.S.C., Ord. 62, r. 11. Sections 111 and 112 of the Act conferred similar powers on judges and magistrates in criminal proceedings.

For present purposes, the significance of this development is that it made advocates, both barristers and solicitors, liable for negligence in the conduct of litigation. It is true that it was a limited form of liability because it was restricted to the payment of wasted costs. It did not extend to any other loss which their negligence might have caused to their clients or other parties. But the costs of modern litigation can amount to a good deal of money. Furthermore, the possibility that the negligent conduct of litigation may lead to a wasted costs order being visited upon the advocate by summary process, before the very judge hearing the case, is likely to be more present to the mind of an advocate than the prospect of an action for negligence at some time in the future. If, therefore, the possibility of being held liable in negligence is calculated to have an adverse effect on the behaviour of advocates in court, one might expect this to have followed, at least in some degree, from the introduction of wasted costs orders.

Such was certainly the submission of counsel for both the Law Society and the Bar Council to the Court of Appeal in Ridehalgh v. Horsefield [1994] Ch 205. The Courts and Legal Services Act 1990 had extended rights of audience in the superior courts to solicitors and section 62 recognised that they should in that capacity have whatever immunities were enjoyed by barristers:

The two professional bodies argued that any liability for wasted costs orders should be subject to the immunity recognised in section 62. Their counsel were not however agreed on how the divided loyalty of the advocate would be affected. Mr. Matheson Q.C. for the Law Society said, at p. 213E, that it would "affect the willingness of legal representatives fearlessly to represent their clients' interests." Mr. Rupert Jackson Q.C., for the Bar Council, advanced, at pp. 217-218, the Rondel v. Worsley [1969] 1 AC 191 argument that it would affect the ability of the barrister "to be able to perform his duty to the court fearlessly and independently." Either version of the argument would have made a sizeable hole in the new jurisdiction, particularly in its application to barristers in criminal proceedings. The Court of Appeal rejected it. Since then, many wasted costs orders have been made as a result of the negligent conduct of legal proceedings.

My Lords, I accept that the liability of a negligent advocate to a wasted costs order is not the same as a liability to pay general damages. But the experience of the wasted costs jurisdiction is the only empirical evidence we have available in this country to test the proposition that such liability will have an adverse effect upon the way advocates perform their duty to the court. There is no doubt that the jurisdiction has given rise to problems, particularly in exercising it with both fairness and economy. But I have found no suggestion that it has changed standards of advocacy for the worse. On the contrary. In Fletamentos Maritimos S.A. v. Effjohn International BV (unreported) 10 December 1997; Court of Appeal (Civil Division) Transcript No. 2115 of 1997 the Court of Appeal made a wasted costs order against a firm of solicitors who had instructed counsel to made a hopeless application for leave to appeal. Simon Brown L.J. ended his judgment by saying:

Mr. Sumption (for the solicitors) and Mr. Peter Scott, for the Bar Council, say that one cannot draw any useful conclusions from other legal systems in which no immunity exists. Legal cultures differ. The court procedures of Europe and the United States, for example, lack the predominantly oral character of litigation in the United Kingdom. In Australia and New Zealand, where procedures are most similar, Rondel v. Worsley [1969] 1 AC 191 is followed. In general I accept this, but I cannot refrain from drawing attention to the experience in Canada. It appears that in that country no immunity was claimed for lawyers before Rondel v. Worsley [1969] 1 AC 191. Then, in Demarco v. Ungaro (1979) 95 D.L.R. (3d) 385, a firm of barristers and solicitors at Niagara Falls, Ontario found themselves sued by a former client for negligence in the conduct of a case in which he had been ordered to pay $6,000 and costs. They argued that as long as the immunity in England was based on the absence of a contract with a barrister, it could obviously have no application in Canada. Lawyers there contracted with their clients. But now that the House of Lords in Rondel v. Worsley [1969] 1 AC 191 had reissued the immunity with a newly minted rationale, there was no reason why the arguments of public policy should not also pass current in Canada. Krever J. examined that case and Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, as well as the few Canadian cases on the subject and explained the differences between the Canadian and English legal professions. But I do not think it would be unfair to summarise the pith of the judgment on the divided loyalty argument as being that Canada had got on perfectly well without an immunity for over a hundred years and there was no reason to think that it needed to be introduced in order to encourage lawyers to perform their duties to the court. He said, at p. 406:

Although a decision at first instance in Ontario, the careful and fully reasoned decision of Krever J. appears to have been treated as settling the law in Canada. It has not since been challenged.

My noble and learned friend Lord Hope of Craighead considers that although in civil proceedings the possibility that the removal of the immunity may have an adverse effect upon the conduct of advocates is not strong enough to justify its retention, there is a sufficiently strong likelihood that it will have this effect in criminal cases. Counsel will be tempted "to pursue every conceivable point, good or bad . . . ." This must be an intuitive prediction, because there is in the nature of things no way of proving it now. I would not regard the current efflorescence of human rights points in Scottish criminal proceedings, notwithstanding the existence of the immunity, as any indication as to whether removal of the immunity would aggravate matters. This is an area in which cause and effect is not easy to establish. And of course, I acknowledge that my noble and learned friend's experience is far greater than mine. Indeed, it could hardly be less. But I am comforted by the fact that others with considerable experience of criminal proceedings do not have the same forebodings. In the end, I do not think that such intuitions are a sound basis upon which to proceed.

The argument for immunity in criminal proceedings depends heavily upon the image of litigants like Mr. Rondel, occupying their prison time with devising vexatious proceedings against their counsel which are then launched at public expense. But it must be remembered, first, that the abuse of process doctrine, which I shall discuss later, is likely to eliminate almost all such plaintiffs who have not succeeded in having their convictions set aside; and secondly, for the reasons which I have explained, that vexatious actions are less likely to be publicly funded and more likely to be struck out than they were in 1967. My noble and learned friend Lord Hutton chooses his example carefully when he says that "few members of the public would have been critical of Mr. Worsley being granted immunity." I quite agree that the case against him should have been struck out. But that is because it was hopeless. It would be easy to imagine other facts in which the public would react very badly to a grant of immunity.

My noble and learned friend Lord Hobhouse of Woodborough has a rather narrower point. He places emphasis not so much upon the way the advocate may conduct the criminal trial but upon the appellate process. He says that the advocate may be less inclined to assist the Court of Appeal with a full explanation of what went wrong at the trial if he thinks that a successful appeal would open the way to an action against him for negligence. In most appeals, no such assistance will be required. All the material will be on the record. But I accept that there are some cases in which it may be necessary to inquire of the advocate as to matters such as the instructions he received or why some witnesses were not called. Again it seems to me that the prediction of a change in the behaviour of the advocates is based upon intuition and even if the intuition is more soundly based, the class of cases involved is so narrow that it cannot justify a total immunity from actions for negligence in the conduct of all criminal cases.

This argument is that a barrister, who is obliged to accept any client, would be unfairly exposed to vexatious actions by clients whom any sensible lawyer with freedom of action would have refused to act for. It is, in the nature of things, intuitive, incapable of empirical verification, and I do not believe it has any real substance. The clients in question will presumably have already found solicitors to represent them without any professional compulsion. There may be many reasons why a barrister, free to choose, would prefer not to act for a client, such as the fact that he is particularly tiresome or disgusting, but I doubt whether fear of a vexatious action is a prominent consideration. In any case, for reasons which I have explained, I think that vexatious actions are an occupational hazard of professional men and that we are improving our ways of dealing with them. If the prospect of their being brought against lawyers serves as an incentive to improve those procedures even more, so much the better for everyone. I should mention that Lord Diplock in Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, 221 dismissed the cab rank argument for much the same reasons.

This argument starts from the well-established rule that a witness is absolutely immune from liability for anything which he says in court. So is the judge, counsel and the parties. They cannot be sued for libel, malicious falsehood or conspiring to give false evidence: Marrinan v. Vibart [1963] 1 Q.B. 528. The policy of this rule is to encourage persons who take part in court proceedings to express themselves freely. The interests of justice require that they should not feel inhibited by the thought that they might be sued for something they say. And, as Fry L.J. explained in the passage which I have already cited from Munster v. Lamb 11 QBD 588, 607 this policy is regarded as so important that it requires not merely qualified privilege but absolute immunity.

The application of the analogy to the negligence of lawyers involves generalising the policy of the witness immunity and expressing it, as Lord Diplock did in Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, 222A, as a "general immunity from civil liability which attaches to all persons in respect of their participation in proceedings before a court of justice." Stated at this level of generality, it includes immunity for advocates from liability for anything that they may do. The rationale is said to be to "ensure that trials are conducted without avoidable stress and tensions of alarm and fear in those who have a part to play in them."

My Lords, with all respect to Lord Diplock, it seems to me that to generalise the witness immunity in this way is illegitimate and dangerous. In the High Court of Australia in Mann v. O'Neill (1997) 71 A.L.J.R. 903, 912 McHugh J. spoke of the perils of extending the witness immunity by analogy. There is, he said, a temptation:

What is the rationale of the witness immunity? In Taylor v. Director of the Serious Fraud Office [1999] 2 A.C.177, 215C, I said that the policy of the immunity was "to encourage freedom of expression" and that was why it was limited to cases in which "the alleged statement constitutes the cause of action." My noble and learned friend Lord Hope of Craighead explained, at p. 219, that the immunity did not, for example, protect a witness against an action for malicious prosecution based on what he had said to the police because "it is the malicious abuse of process, not the making of the statement, which provides the cause of action." In other words, the immunity is based upon a perception that witnesses would otherwise be less inclined to come forward and tell the truth. They would behave differently in a way which was inimical to the interests of justice.

It is not sufficient, therefore, to explain any immunity relating to court proceedings by saying that the people involved should be free from "avoidable stress and tensions." That merely suggests that everyone would find litigation more agreeable if no awkward consequences could follow from anything which the participants did. It is another version of the vexation argument, which I have already rejected. It is necessary to go further and explain why the public interest requires that a particular participant should be free from the stress created by the possibility that he might be sued. How would he otherwise behave differently in a way which was contrary to the public interest?

If one asks the question in this way, as I think one must, then it becomes apparent that Lord Diplock was inconsistent in rejecting the divided loyalty argument and the cab rank argument but accepting the witness analogy. It involves, as Lord Diplock himself would have put it, a petitio principii. The witness rule depends upon the proposition that without it, witnesses would be more reluctant to assist the court. To establish the analogy, it is necessary to point to some similar effect on the behaviour of lawyers. But Lord Diplock rejected the only two candidates put forward for likely changes in behaviour and offered no others. The proposition that absence of immunity would have an effect contrary to the public interest was assumed without argument.

Mr. Scott invited your Lordships to apply by analogy the decision of the Court of Appeal in Stanton v. Callaghan [2000] 1 Q.B. 75, in which it was held that an expert witness could not be sued for agreeing to a joint experts' statement in terms which the client thought detrimental to his interests. He said that this was an example of a general immunity for acts done in the course of litigation. But that seems to me to fall squarely within the traditional witness immunity. The alleged cause of action was a statement of the evidence which the witness proposed to give to the court. A witness owes no duty of care to anyone in respect of the evidence he gives to the court. His only duty is to tell the truth. There seems to me no analogy with the position of a lawyer who owes a duty of care to his client.

Nor is there in my opinion any analogy with the position of the judge. The judge owes no duty of care to either of the parties. He has only a public duty to administer justice in accordance with his oath. The fact that the advocate is the only person involved in the trial process who is liable to be sued for negligence is because he is the only person who has undertaken a duty of care to his client.

This argument also has a number of strands which need to be examined separately.

It may be very difficult to arrive at a conclusion about what would have happened in earlier proceedings if in some respect they had been conducted differently. In Smith v. Linskills [1996] 1 W.L.R. 763, 773 Sir Thomas Bingham M.R. spoke of:

Of course this is true. But, in principle, evidential difficulties have never been regarded as a reason for declining jurisdiction. The plaintiff has to prove that the lawyer's negligence caused him loss. The burden of proof is upon him. His case may have become so weak with the passage of time that it has to be struck out. But that is no reason for giving lawyers immunity from suit even in cases in which there is no difficulty about proving that their negligence caused loss to the plaintiff. This has to be done in cases which fall outside the immunity. For example, in Kitchen v. Royal Air Force Association [1958] 1 W.L.R. 563 a firm of solicitors were negligent in failing to issue a writ before the limitation period expired. Lloyd-Jacob J. had to decide in 1957 what would have been the plaintiff's chances of success in an action which should have been brought before 1946 to establish that her husband's death by electrocution in 1945 had been caused by the negligence of the West Kent Electricity Co. when it installed a control box in 1940. The Court of Appeal upheld his estimate of the value of her claim.

Then it is said that while it is difficult enough to decide what would have happened at a trial which did not in fact take place (as in Kitchen v. Royal Air Force Association [1958] 1 W.L.R. 563), it may become positively invidious to decide how a judge who actually heard the case would have reacted if the advocate had advanced a different argument or called different evidence. Some judges are more receptive to certain kinds of points than others. I think that this is an imaginary problem. Whatever may have been the foibles of the judge who heard the case, it cannot be assumed that he would have behaved irrationally. If he did, it would have been corrected on appeal. Obviously one has to take into account the findings that the judge made on the case as it was actually presented. For example, if he did not believe anything which the plaintiff said, it may be difficult to show that a different line of argument would have persuaded him to find in his favour. But I do not see how it is relevant for the purposes of the hypothetical exercise to have regard to the judge's idiosyncrasies. It must be assumed that he would have behaved judicially.

The most substantial argument is that it may be contrary to the public interest for a court to retry a case which has been decided by another court. In Rondel v. Worsley [1969] 1 AC 191, 251 Lord Morris of Borth-y-Gest said that it would be:

In Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, 222-223, Lord Diplock developed this point in a passage which should be quoted at length:

It may be said that this passage is combining two arguments: the one based upon evidential difficulty, which is not, as I have said, a general reason for refusing to try a case, and the argument that conflicting decisions may bring the administration of justice into disrepute. But I think that Lord Diplock is saying that the fallibility of any conclusion about whether the earlier case would have been decided differently reinforces the public interest rule about avoiding conflicting decisions. This is obviously an argument entitled to great respect.

But actions for negligence against lawyers are not the only cases which give rise to a possibility of the same issue being tried twice. The law has to deal with the problem in numerous other contexts. So, before examining the strength of the collateral challenge argument as a reason for maintaining the immunity of lawyers, it is necessary to consider how the law deals with collateral challenge in general.

The law discourages relitigation of the same issues except by means of an appeal. The Latin maxims often quoted are nemo debet bis vexari pro una et eadem causa and interest rei publicae ut finis sit litium. They are usually mentioned in tandem but it is important to notice that the policies they state are not quite the same. The first is concerned with the interests of the defendant: a person should not be troubled twice for the same reason. This policy has generated the rules which prevent relitigation when the parties are the same: autrefois acquit, res judicata and issue estoppel. The second policy is wider: it is concerned with the interests of the state. There is a general public interest in the same issue not being litigated over again. The second policy can be used to justify the extension of the rules of issue estoppel to cases in which the parties are not the same but the circumstances are such as to bring the case within the spirit of the rules. I shall give two examples. In Reichel v. Magrath (1889) 14 App. Cas. 665 Mr. Reichel, the vicar of Sparsholt, resigned. The bishop of Oxford accepted his resignation. Then the vicar changed his mind. He brought an action against the Bishop and the Queen's College, Oxford, which had the right of presentation, for a declaration that his resignation had been void. The judge held that it had been valid and that the living was vacant. His decision was affirmed on appeal. The college appointed its Provost, Dr. Magrath, as the new vicar. Mr. Reichel refused to move out of the vicarage. Dr. Magrath brought an action for possession. Mr. Reichel pleaded in defence that his resignation had been void and he was still the vicar. The court struck out the defence as an "abuse of the process of the court." Although the parties were different, the case was within the spirit of the issue estoppel rule. Dr. Magrath was claiming through the college, which had been a party to the earlier litigation.

In Ashmore v. British Coal Corporation [1990] 2 Q.B. 338 Ms Ashmore worked in the canteen of a coal mine in Nottingham. She complained to an industrial tribunal that she was paid less than men were being paid for similar work, contrary to the Equal Pay Act 1970. Over 1500 other women employees of the corporation made similar complaints. The industrial tribunal decided to hear 14 sample cases, 6 selected by the employees and 8 by the employers, to lay down general principles according to which the others could be decided. Ms Ashmore was aware of these arrangements. The tribunal decided all the cases adversely to the applicants on grounds which were equally applicable to Ms Ashmore's application. She then asked for a separate hearing of her case. The Court of Appeal decided that it should be struck out as an abuse of the process of the court. Ms Ashworth had not been a party to the sample proceedings but the sensible procedure there adopted would be undermined if all other members of the group were entitled to demand a separate hearing.

The leading case on the application of the power to dismiss proceedings on this ground as an abuse of the process of the court is Hunter v. Chief Constable of the West Midlands Police [1982] AC 529. It concerned the trial of the six men convicted of an I.R.A. bombing in Birmingham in 1974. The defendants claimed that the police had beaten them to extract confessions. The trial judge held a voir dire and decided that the prosecution had proved beyond reasonable doubt that they had not been beaten. They were convicted. They applied for leave to appeal, but not on the ground that the confessions had been wrongly admitted. Leave to appeal was refused. In prison, the accused commenced proceedings against the policemen for assault, alleging the same beatings as had been alleged at the criminal trial. The House of Lords decided that it was an abuse of the process of the court to attempt to relitigate the same issue and that the actions should be struck out.

Criminal proceedings are in my opinion in a special category because although they are technically litigation between the Crown and the defendant, the Crown prosecutes on behalf of society as a whole. In the United States, the prosecutor is designated "The People." So a conviction has some of the quality of a judgment in rem, which should be binding in favour of everyone. As Lord Diplock pointed out in Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, 223, this policy is reflected in section 13 of the Civil Evidence Act 1968, which provides that in an action for libel or slander, proof of the plaintiff's conviction is conclusive evidence that he committed the offence of which he was convicted.

But one should not exaggerate this argument. The policy reasons which justify making the conviction conclusive evidence in a defamation action do not necessarily apply to other actions. I said that a conviction has some of the quality of a judgment in rem but, as a matter of law, it remains a judgment between the Crown and the accused and that is often the right way to consider it. The Court of Appeal is generally thought to have taken the technicalities of the matter much too far when it decided in Hollington v. F. Hewthorn & Co. Ltd. [1943] 1 K.B. 587 that in civil proceedings a conviction was res inter alios acta and no evidence whatever that the accused had committed the offence. But when Parliament reversed this rule in section 11(1) of the Civil Evidence Act 1968, it did not say that the conviction should be conclusive evidence, so that the issue could not be relitigated. It said only that the conviction was admissible evidence for proving that he committed the offence.

Hunter v. Chief Constable of the West Midlands Police [1982] AC 529 shows that, superimposed upon the rules of issue estoppel and the Civil Evidence Act 1968, the courts have a power to strike out attempts to relitigate issues between different parties as an abuse of the process of the court. But the power is used only in cases in which justice and public policy demand it. Lord Diplock began his speech, at p. 536, by saying that the case concerned:

I too would not wish to be taken as saying anything to confine the power within categories. But I agree with the principles upon which Lord Diplock said that the power should be exercised: in cases in which relitigation of an issue previously decided would be "manifestly unfair" to a party or would bring the administration of justice into disrepute. It is true that Lord Diplock said later in his speech, at p. 541, that the abuse of process exemplified by the facts of the case was:

But I do not think that he meant that every case falling within this description was an abuse of process or even that there was a presumption to this effect which required the plaintiff to bring himself within some exception. That would be to adopt a scheme of categorisation which Lord Diplock deplored. As I shall explain, I think it is possible to make some generalisations about criminal proceedings. But each case depends upon an application of the fundamental principles. I think that Ralph Gibson L.J. was right when, after quoting this passage, he said in Walpole v. Partridge & Wilson [1994] Q.B. 106, 116 that Hunter's case decides "not that the initiation of such proceedings is necessarily an abuse of process but that it may be."

My Lords, the discussion in the last sections shows, first, that not all relitigation of the same issue will be manifestly unfair to a party or bring the administration of justice in to disrepute, and secondly, that when relitigation is for one or other of these reasons an abuse, the court has power to strike it out. This makes it very difficult to use the possibility of relitigation as a reason for giving lawyers immunity against all actions for negligence in the conduct of litigation, whether such proceedings would be an abuse of process or not. It is burning down the house to roast the pig; using a broad-spectrum remedy when a more specific remedy without side effects can handle the problem equally well.

Cases in which actions for negligence have been brought against solicitors without immunity illustrate this point. Walpole v. Partridge & Wilson [1994] Q.B. 106 is one. The plaintiff was convicted before the magistrates of a statutory offence by preventing a veterinary officer from inspecting his pigs. His appeal to the Crown Court was dismissed. He issued proceedings against his solicitors for negligence, claiming that he had wanted to appeal by way of case stated and had arguable grounds for success on a point of law, but that they had negligently failed to lodge an appeal. The solicitors applied for the action to be struck out as an abuse of process on the ground that it would involve trying the question of whether the Crown Court had been wrong in law. In a closely reasoned and admirable judgment, Ralph Gibson L.J. decided that the claim would not be manifestly unfair to the solicitors or bring the administration of justice into disrepute. On the contrary, the denial of a remedy was more likely to do so.

It is easy to imagine a similar case in which the alleged negligence would have been within the immunity: failure on the part of counsel, for example, to take an obvious point of law in the Crown Court. (Compare Atwell v. Michael Perry & Co [1998] 4 All E.R. 65.) In such a case the consequence of the immunity would be to deny a remedy for negligence although the collateral challenge argument had no application.

My Lords, I have now considered all the arguments relied upon in Rondel v. Worsley [1969] 1 AC 191. In the conditions of today, they no longer carry the degree of conviction which would in my opinion be necessary to sustain the immunity. The empirical evidence to support the divided loyalty and cab rank arguments is lacking; the witness analogy is based upon mistaken reasoning and the collateral attack argument deals with a real problem in the wrong way. I do not say that Rondel v. Worsley [1969] 1 AC 191 was wrongly decided at the time. The world was different then. But, as Lord Reid said then, public policy is not immutable and your Lordships must consider the arguments afresh.

Mr. Sumption and Mr. Scott said that even if your Lordships thought that the immunity could no longer be justified, you should not, in your judicial capacity, alter the law. It was something which Parliament had considered fairly recently, during the passage of the Courts and Legal Services Act 1990. A legislative decision had been taken not to abolish the immunity. For the judges now to do so would be to trespass upon a competence which had been assumed by the sovereign legislature.

My Lords, I acknowledge the need for sensitivity on the part of the judges in entering into areas of law which are properly matters for democratic decision. Recently in Southwark London Borough Council v. Mills [1999] 3 WLR 939, 944, I said:

But, my Lords, there has been no statement of legislative policy on the immunity for lawyers. Section 62(1) of the Courts and Legal Services Act 1990, which I have already quoted, was careful not to endorse the immunity. It merely said that whatever immunity barristers had should also extend to solicitors. It is true that during the debate in committee in the House of Lords Lord Allen of Abbeydale moved an amendment to abolish the immunity which he afterwards withdrew (Hansard (H.L. Debates), 5 February1990, cols. 570-578). A similar amendment was moved but not voted on in Standing Committee D in the House of Commons (Hansard (H.C. Debates), 7 June 1990, cols. 325-340). It seems to me, however, that the government merely accepted what the judges had said in Rondel v. Worsley [1969] 1 AC 191 at face value. It may be that even as recently as10 years ago they were right to do so. A number of the changes to which I have referred earlier in this speech were a result of the Act of 1990 itself (such as wasted costs orders) and later developments in civil procedure and the public funding of litigation. So I do not think that your Lordships would be intervening in matters which should be left to Parliament. The judges created the immunity and the judges should say that the grounds for maintaining it no longer exist. Cessante ratione legis, cessat lex ipsa.

If there is to be no immunity, there will be more cases in which it becomes necessary to examine the limits of the Hunter doctrine of abuse of process. As I have said, the basic principles were clearly stated in that case. The House of Lords made it clear that the remedy should remain flexible and I cannot imagine that Parliament, if it legislated upon the subject of the immunity, would wish to give any more precise guidance as to how the abuse of process remedy should be used. It is peculiarly a matter of judicial application to the facts of each case. For the purposes of the present appeals, I therefore need say no more than that I agree with the Court of Appeal that the doctrine does not apply to any of them. If, as must for present purposes be assumed, the allegations made by the plaintiffs are correct, there seems to me nothing manifestly unfair to the solicitors in having to answer for them. Nor do I think it would bring the administration of justice into disrepute if the plaintiffs were allowed to claim that they would have got better terms if their solicitors had advised and acted for them with reasonable care. Although the two matrimonial cases involved approval of the settlement by a judge, that approval was given on the basis of the information put before him and, even more important, upon the basis that the parties, duly advised by solicitors, had agreed to the order. The judge was entitled to give weight to the fact that the parties themselves agreed that the order would make reasonable provision for both of them. The plaintiffs claim that if the judge had been given different information and if they had not been advised to agree to the order, they would have done better. That does not seem to me to involve any attack upon the judicial process.

I do not think, however, that I can entirely agree with the Court of Appeal's view that the question of whether a collateral challenge is an abuse of process depends upon the "weight" to be given to the judgement and that there is a scale of weighting according to the amount of judicial input, with a consent order at one end and a judgment after hearing full evidence at the other. I agree that, as a practical matter, it is very difficult to prove that a case which was lost after a full hearing would have been won if it had been conducted differently. It may be easier to prove that, with better advice, a more favourable settlement would have been achieved. But this goes to the question of whether, in the words of C.P.R., r. 24.2, the plaintiff has "a real prospect of succeeding on the claim." The Hunter question, on the other hand, is whether allowing even a successful action to be brought, would be manifestly unfair or bring the administration of justice into disrepute. In my view, there will be cases (such as conviction on a plea of guilty) in which the Hunter principle may be engaged although there has been virtually no judicial input at all. The Court of Appeal accepted this. On the other hand, I can see no objection on grounds of public interest to a claim that a civil case was lost because of the negligence of the advocate, merely because the case went to full trial. In such a case the plaintiff accepts that the decision is res judicata and binding upon him. He claims however that if the right arguments had been used or evidence called, it would have been decided differently. This may be extremely hard to prove in terms of both negligence and causation, but I see no reason why, if the plaintiff has a real prospect of success, he should not be allowed the attempt.

There is, I think, a relevant difference between criminal proceedings and civil proceedings. In civil proceedings, the maxim nemo debet bis vexari pro una et eadem causa applies very strongly. Fresh evidence is admissible on appeal only subject to strict conditions. Even if a decision is based upon a view of the law which is subsequently expressly overruled by a higher court, the judgment itself remains res judicata and cannot be set aside: see In re Waring (No. 2) [1948] Ch 221. An issue estoppel created by earlier litigation is binding subject to narrow exceptions: see Arnold v. National Westminster Bank Plc. [1991] 2 A.C. 93. But the scope for re-examination in criminal proceedings is much wider. Fresh evidence is more readily admitted. A conviction may be set aside as unsafe and unsatisfactory when the accused appears to have been prejudiced by "flagrantly incompetent advocacy:" see Reg. v. Clinton [1993] 1 W.L.R. 1181. After appeal, the case may be referred to the Court of Appeal (if the conviction was on indictment) or to the Crown Court (if the trial was summary) by the Criminal Cases Review Commission: see Part II of the Criminal Appeal Act 1995.

It follows that in my opinion it would ordinarily be an abuse of process for a civil court to be asked to decide that a subsisting conviction was wrong. This applies to a conviction on a plea of guilty as well as after a trial. The resulting conflict of judgments is likely to bring the administration of justice into disrepute. The arguments of Lord Diplock in the long passage which I have quoted from Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, 222-223 are compelling. The proper procedure is to appeal, or if the right of appeal has been exhausted, to apply to the Criminal Cases Review Commission under section 14 of the Act of 1995. I say it will ordinarily be an abuse because there are bound to be exceptional cases in which the issue can be tried without a risk that the conflict of judgments would bring the administration of justice into disrepute. Walpole v. Partridge & Wilson [1994] Q.B. 106 was such a case.

Once the conviction has been set aside, there can be no public policy objection to an action for negligence against the legal advisers. There can be no conflict of judgments and the only contrary arguments which remain are those of divided loyalty, vexation and the cab rank, all of which I have already rejected. Acton v. Graham Pearce & Co. [1997] 3 All E.R. 909 is a good example of such an action in a case which lay outside the immunity and illustrates the point that bringing such a claim is not in itself an abuse of process. While it is true that there is a power for the Crown to pay compensation to the person wrongly convicted, there is no reason why public funds should be used to pay the accused compensation for loss caused by the negligence of the lawyers who were paid to defend him.

On the other hand, in civil (including matrimonial) cases, it will seldom be possible to say that an action for negligence against a legal adviser or representative would bring the administration of justice into dispute. Whether the original decision was right or wrong is usually a matter of concern only to the parties and has no wider implications. There is no public interest objection to a subsequent finding that, but for the negligence of his lawyers, the losing party would have won. But here again there may be exceptions. The action for negligence may be an abuse of process on the ground that it is manifestly unfair to someone else. Take, for example, the case of a defendant who publishes a serious defamation which he attempts unsuccessfully to justify. Should he be able to sue his lawyers and claim that if the case had been conducted differently, the allegation would have been proved to be true? It seems to me unfair to the plaintiff in the defamation action that any court should be allowed to come to such a conclusion in proceedings to which he is not a party. On the other hand, I think it is equally unfair that he should have to join as a party and rebut the allegation for a second time. A man's reputation is not only a matter between him and the other party. It represents his relationship with the world. So it may be that in such circumstances, an action for negligence would be an abuse of the process of the court.

I would suspect that, having regard to the power of the court to strike out actions which have no real prospect of success, the Hunter doctrine is unlikely in this context to be invoked very often. In my opinion, the first step in any application to strike out an action alleging negligence in the conduct of a previous action must be to ask whether it has a real prospect of success. Hopeless cases like Rondel v. Worsley [1969] 1 AC 191 are not a suitable vehicle for deciding important points of public policy.

My Lords, I have said nothing about whether the immunity, if preserved, would be contrary to article 6 of the European Convention on Human Rights. The question does not arise. Nor have I said anything about the distinction between those acts of lawyers which are "intimately connected" with the conduct of litigation and those which are not. The Court of Appeal, being bound by Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, struggled with this distinction. Mr. Sumption's submissions as to why they were wrong served only to convince me that the distinction is very difficult to apply with any degree of consistency. That is perhaps another reason why the immunity should be altogether abolished. I would therefore dismiss the appeals.

The events with which these three appeals are concerned took place in 1991, when the parties on one side of the case ("the clients") were all engaged in civil litigation for the purposes of which they had appointed the other party to act as their solicitors. Mr. Simons, who is a building contractor, was in dispute with the owner of a building about the work which he had carried out for the owner under a building contract. The proceedings were settled on 19 August 1991, which was the day before the trial of his action was due to start. Mr. Barratt was in dispute with his wife after their marriage had broken down. Her claim for ancillary relief was settled on 5 September 1991 when the judge approved a minute of order lodged by his solicitors and directed that it should stand as the court's order made by consent. Mrs. Harris was also engaged in matrimonial proceedings following the breakdown of her marriage. In her case a consent order was made by the judge on 22 November 1991 following advice which she received from counsel outside the court on the day of the ancillary relief hearing.

In each case the clients are dissatisfied with the outcome of their litigation and in particular with the terms of settlement. They have alleged that the solicitors were negligent in regard to things which they did or omitted to do outside the courtroom. The essence of the case made by Mr. Simons against his solicitors is that they should have advised him at the outset that he should settle on the terms which he was ultimately forced to accept after much unnecessary delay and expenditure or that they should have prepared for trial so that he could pursue his case with unimpaired prospects of success. Mr. Barratt's case is that his solicitors failed at any stage to obtain or advise the obtaining of a valuation of the family home which was eventually sold for much less than it had been assumed to be worth when they were negotiating the terms of settlement, that they lodged with the court a minute of order which inaccurately recorded the valuation of the property and that they failed to advise him that the settlement should provide for the parties to receive percentage interests in the property rather than that his wife should receive a guaranteed sum when it was sold. Mrs. Harris alleges that her solicitors failed to brief competent counsel, to inform themselves properly of the facts and take proper instructions prior to the settlement and that they gave incorrect advice about the possibility of setting aside a consent order. The solicitors in each case claim that they are immune from suit in regard to the allegedly negligent conduct.

All three cases were listed and heard together in the Court of Appeal, as was a fourth case with which your Lordships are not now concerned. At the outset of their judgment the Court of Appeal (Lord Bingham C.J., Morritt and Waller L.JJ.) said that the following questions arose ([1999] 3 WLR 873, 881D-E): to what extent and in what circumstances does a lawyer's immunity from suit in relation to the allegedly negligent conduct of a case in court protect him against claims for allegedly negligent acts and omissions which take place out of court? Does a lawyer, if not otherwise immune from a claim in negligence by a client, become so when the court approves a consent order in any proceedings, but particularly in matrimonial proceedings in relation to ancillary relief? Is it in such circumstances an abuse of the process of the court to claim damages against a lawyer for alleged negligence leading to the making of a consent order?

The primary sources on which the Court of Appeal drew as to the advocate's immunity were the decisions of the House in Rondel v. Worsley [1969] 1 AC 191 and Saif Ali v. Sydney Mitchell & Co. [1980] AC 198. After setting out four propositions which it derived from them, the court made these observations, at p. 882F-H:

Now that the three remaining appeals have reached this House the opportunity has been taken to undertake the more radical challenge to the authority of Rondel v. Worsley [1969] 1 A.C.191 which was not undertaken in the Saif Ali case [1980] AC 198. It is therefore open to your Lordships to dispose of them on grounds which were not available to the Court of Appeal.

I wish to say, however, before turning to this wider and more general argument, that I consider that the grounds which the Court of Appeal gave for its decision in each case were entirely sound, sufficient and satisfactory and that I would have dismissed each of the appeals for the same reasons irrespective of the view that was taken about what the Court of Appeal has described as the core forensic immunity. In Mr. Simons's case this is because the acts and omissions of which he complains were done or not done, as the case may be, when the solicitors were acting otherwise than as advocates. Even if they had been acting in the relevant respects as advocates, none of the allegations against them satisfy the "intimate connection" test described by McCarthy P. in Rees v. Sinclair [1974] 1 N.Z.L.R. 180,187: see the Court of Appeal's judgment [1999] 3 WLR 873, 908E-G. In Mr. Barratt's case the solicitors were not acting as advocates in relation to any alleged act of negligence, nor was their conduct said to be negligent in an area where the solicitors could say that they were acting where public policy as the rationale for immunity had any impact: p. 911F. In Mrs. Harris's case her solicitors were not acting in any way as advocates in the respects in which they were alleged to be negligent, nor is there any public policy rationale for which immunity in their case could be said to be justified: p. 920G-921A. In short, I would regard the argument in each case for extending the immunity to the solicitors when they were negotiating the terms of settlement as entirely without merit on the existing state of the authorities. On this view it is unnecessary to examine the fundamental question whether the core forensic immunity can now - or, to put the question more accurately if it is to provide a ground for our decision in these three cases, could in 1991 - still be justified on grounds of public policy. Nevertheless I agree that your Lordships should accept the opportunity for reviewing the fundamental question, for the following reasons.

The first reason is that, as Lord Reid recognised in Rondel v. Worsley [1969] 1 AC 191, 227, public policy is not immutable. Lord Wilberforce was making the same point when he said in Roy v. Prior [1971] A.C. 470, 480F that immunities conferred by the law in respect of legal proceedings need always to be checked against a broad view of the public interest. Doubts have once again arisen as to whether the existing rule is justified in present day conditions in this country, so it is proper to re-examine the whole matter now. The second reason is that there is now a greater appreciation of the importance which has to be attached in this context to the principles of human rights law, especially in view of the imminence of the coming into force of the Human Rights Act 1998. The period which has to elapse before that Act comes into force in October 2000 is now very short. I think that it is appropriate in this case to anticipate that event by taking account of the relevant provisions of the European Convention on Human Rights and Fundamental Freedoms ("the Convention") and the jurisprudence of the European Court of Human Rights in our determination of the question whether, and if so to what extent, the core forensic immunity can still be justified. The third reason is that, while I would not regard it as necessary in order to dispose of these appeals for your Lordships to say that any change as regards the immunity rule should operate retrospectively, I consider it to be a legitimate exercise of your Lordships' judicial function to declare prospectively whether or not the immunity - which is a judge-made rule - is to be available in the future and, if so, in what circumstances.

I believe that none of your Lordships would wish to go so far as to hold that Rondel v. Worsley [1969] 1 AC 191 was wrongly decided and that it should be overruled. The issue is whether the decision which was reached in that case can now be justified. It seems to me to be preferable that we should address this issue by examining the circumstances relevant to this issue as we find them today, and that we should express our decision so that it applies only to the future - not to a period in the past as well, the commencement of which would be very difficult at this stage to identify.

Any immunity from suit is a derogation from a person's fundamental right of access to the court which has to be justified. This principle is found both in the common law and in the jurisprudence of the European Court of Human Rights. For the common law position it is sufficient to note the following observations. In Rondel v. Worsley [1969] 1 AC 191, 228 Lord Reid said:

In Rees v. Sinclair [1974] 1 N.Z.L.R. 180, 187 McCarthy P. said that the protection of the immunity should not be given any wider application than is absolutely necessary in the interests of the administration of justice. In the Saif Ali case [1980] AC 198, 214H Lord Wilberforce said that in fixing the boundary of immunity from an action, which depends on public policy, account must be taken of the principle that a wrong should not be without a remedy. As Kirby J. said in Boland v. Yates Property Corporation Pty. Ltd. (1999) 74 A.L.J.R. 209, 236, 238, 239, paras.129, 137 and 140, an immunity from liability at law is a derogation from the normal accountability for wrong-doing which is an ordinary feature of the rule of law and fundamental civil rights.

In the field of human rights law the individual's right of access to the court for the determination of his civil rights is to be found in article 6 (1) of the Convention. In Golder v. United Kingdom (1975) 1 EHRR 524, 535-536 paragraph 35 the European Court of Human Rights said:

In Fayed v. United Kingdom (1994) 18 EHRR 393, 429-430, para. 65, in a passage which was approved in Tinnelly Sons Ltd. v. United Kingdom (1998) 27 E.H.R.R. 249, 271, para. 74, the court said:

It is clear from the passage which I have quoted from Lord Reid's speech in Rondel v. Worsley [1969] 1 AC 191, 228 that under the common law the presumption is strongly in favour of the right of the individual to a remedy. Any immunity from suit must therefore be clearly justifiable. In terms of human rights law it will only be justifiable if it is designed to pursue a legitimate aim and then only if it satisfies the test of proportionality. If the restriction which the immunity imposes on the right of the individual is disproportionate to the aim sought to be achieved on grounds of public policy it will be incompatible with the right secured to the individual by article 6(1) of the Convention. Although the common law and the human rights law tests are expressed in different language, they are both directed to the same essential point of principle that an immunity from suit is a derogation from a fundamental right which requires to be justified.

I wish at the outset to summarise the main points with which I intend to deal in order to explain the position which I would adopt on the question of the immunity. I shall use the expression "the core immunity" to describe the immunity which attaches to the advocate, when engaged in conduct performed in court, from claims by his client for negligence. I am conscious of the fact that, if the immunity is to continue, the scope of its application may need to be defined more carefully in due course.

a. The sole basis for retaining the core immunity is the public interest in the administration of justice.

b. The public interest in the administration of justice is at its most compelling in the field of criminal justice.

c. The risks to the efficient administration of our system of criminal justice which would result from the removal of the core immunity greatly outweigh the benefits.

d. The principle in Hunter v. Chief Constable of the West Midlands Police [1982] AC 529 which treats collateral challenge as an abuse of process is not a satisfactory substitute in the field of criminal justice for the core immunity.e. The risks to the efficient administration of justice are significantly less in the field of civil justice, so in that field the retention of the core immunity of the advocate from claims by his client for negligence is no longer justified.

If, as I believe, your Lordships do not wish to go so far as to say that Rondel v. Worsley [1969] 1 AC 191 was wrongly decided, it is appropriate to take note of some the events that have happened since then - and especially since the date of the decision in the Saif Ali case [1980] AC 198 - which may throw light on the view that ought now to be taken as to the justification for the immunity on grounds of public policy.

The question whether the core immunity was in the public interest was considered by the 1979 Royal Commission on Legal Services. In its final report the Royal Commission concluded (Cmnd. 7648, vol. 1, p. 333, para. 24.7):

Legislation consistent with this conclusion, and with the decision in the Saif Ali case, was introduced under the Supply of Goods and Services Act 1982. Section 13 of that Act implies a term of reasonable skill and care into contracts for the supply of a service where the supplier is acting in the course of a business. But the Supply of Services (Exclusion of Implied Terms) Order 1982 (S.I. 1982 No. 1771), made under section 12(4) of the Act, provides that that section shall not apply to:

When the Conservative government came to power in 1989 the practices of the legal profession again came under close scrutiny. The aim was to bring to an end restrictive practices, such as those relating to rights of audience, that could no longer be justified. This resulted in the Courts and Legal Services Act 1990. That Act was preceded in 1989 by both a Green Paper The Work and Organisation of the Legal Profession (Cm. 570) and a White Paper entitled Legal Services: A Framework for the Future (Cm. 740) in which the view was expressed that the core immunity was justified in the public interest. The Green Paper stated in paragraph 62:

During the progress of the Bill attempts were made in both Houses to abolish the immunity (Hansard (H.L. Debates), 5 February 1990, cols 570-578); (H.C. Debates, Standing Committee D), 7 June 1990, cols. 325-340), but proposed amendments to that effect were withdrawn after debate. The Lord Chancellor said that the Government believed the immunity rule to be an appropriate one, and he emphasised that it had "placed it in the forefront of consultation right from the start" (Hansard (H.L. Debates), 5 February 1990, col. 576). In the result what is now section 62 of the Act of 1990, which extended the immunity to a person who is not a barrister but is lawfully providing legal services in any proceedings, was enacted against the background of the existing rule, which it did not alter. A further opportunity arose in Parliament to abolish the immunity when parts of the Courts and Legal Services Act 1990 were amended by the Access to Justice Act 1999. It was not suggested in either House that the existing immunity was no longer in the public interest and should be abolished.

The fact that Parliament has not seen fit to abolish the core immunity does not, of course, mean that your Lordships should feel inhibited from taking that initiative. The position which Parliament has adopted is consistent with the view that the question whether the immunity should be retained is pre-eminently a matter for the judges. But the heart of the matter is whether the immunity is in the public interest. It is true, as my noble and learned friend Lord Steyn has pointed out, that a number of distinguished commentators including Sir Sydney Kentridge Q.C. and David Pannick Q.C. have expressed views to the effect that it cannot be justified. But it is notorious that views as to what is in the public interest may vary widely from one person to another, and that they are heavily dependent upon each person's background, focus of attention and experience. The judicial task is to gather the evidence from all the sources that are available and, having done so, to assess the weight of that evidence.

For my part, I would be inclined to attach considerable weight to that fact that neither the 1979 Royal Commission nor the consultation exercise which preceded the enactment of the Courts and Legal Services Act 1990 revealed that there is widespread dissatisfaction among members of the public with the core immunity. I would also be inclined, even now, to attach weight to the observations of the judges in Rondel v. Worsley [1969] 1 AC 191 and the Saif Ali case [1980] A.C. 189 and, more recently, in Giannarelli v. Wraith (1988) 165 C.L.R. 543 in the High Court of Australia with particular reference to the public interest in the efficient administration of criminal justice. Another factor to which I would attach some importance is the marked lack of litigation directed to this issue in this country. The list which is provided in the Court of Appeal's judgment of the decided cases in which lawyers have been held entitled to avail themselves of the protection afforded by the immunity contains only one case in which the complaint related to the conduct of the trial: Bateman v. Owen White [1996] 1 P.N.L.R. 1 (failure to object to inadmissible evidence). The present cases, as I mentioned above, do not involve a challenge to the core immunity. They are concerned with the limits of its application. These factors suggest to me that the arguments for the abolition of the immunity are more finely balanced than some commentators have suggested, and that the case for abolition requires to be approached with caution and with careful regard to all the relevant factors.

My noble and learned friends Lord Steyn and Lord Hoffmann have analysed the arguments for the immunity under four headings: (1) the cab rank rule, (2) the analogy of the immunity of others who participate in court proceedings, (3) re-litigation or collateral challenge and (4) divided loyalty or the duty of the advocate to the court. I am content to accept this analysis of the various reasons which have been advanced to support the immunity on grounds of public policy. But I would approach each of them in a different way, by asking myself in each case what bearing each of these arguments has on the administration of our systems of criminal justice. I think that it is also worth bearing in mind that these arguments are not of equal weight. As my noble and learned friend Lord Steyn has said, the critical factor is the duty of the advocate to the court. He has used the word "barrister," but I think that we are all agreed that the position of advocates in Scotland and of solicitor advocates in all three jurisdictions is the same in this respect as that of barristers and I shall use the word "advocate" to embrace all of them.

I do not wish to say much about the cab rank rule. Its value as a rule of professional conduct should not be underestimated, but its significance in daily practice is not great and the extending of the rights of audience of solicitor advocates who are not bound by the same rule has reduced such importance as it may once have had in the context of discussions about advocates' immunity. I do not think that there is any sound basis for thinking that removal of the immunity would have the effect of depriving those who were in need of the services of advocates in criminal cases of the prospect of obtaining their services. The independent bars have a long and honourable tradition in the field of criminal justice that no accused person who wishes the services of an advocate will be left without representation. This is a public duty which advocates perform without regard to such private considerations as personal gain or personal inconvenience.

I think that there is a little more, but not much, to be said for the analogy with the immunity of others who participate in the proceedings which take place in court. At best it is only an analogy. It is a make-weight argument. Its significance lies in the fact that the other immunities exist because they also can be justified on grounds of public policy. They are illustrations of the fundamental point that it is in the public interest that those who are called upon to give evidence in court or who have to perform duties there should be enabled to do so without the risk of being sued for defamation or for negligence. As Mason C.J. said in Giannarelli v. Wraith, 165 C.L.R. 543, 557 the exception in favour of counsel is in conformity with the privilege which the law has always conferred on those engaged in the adminstration of justice, whether as judge, juror, witness, party, counsel or solicitor in respect of what they say in court. In an appropriate case the public interest will prevail over the private interest. But each of these immunities needs to be justified, and this can be done only on grounds which are relevant to the public interest in the efficient and impartial administration of justice.

This brings me to the two remaining arguments. In Giannarelli v. Wraith, at p. 555 Mason C.J. said that, of the various public policy factors, they were the only two which warranted serious examination.

The first of these two remaining arguments is the impact on the administration of justice of allowing court decisions to become the subject of collateral attack by means of actions raised against advocates by their clients for negligence. It is generally recognised that it is undesirable that collateral attacks of this kind should be permitted. The problem is that doubt will be cast on the soundness of the original decision, which may have been affirmed on appeal, if the later decision is in conflict with it. This problem is particularly acute in the field of criminal justice, as public confidence in the administration of justice is likely to be shaken if a judge in a civil case were to hold that a person whose conviction has been upheld on appeal would not have been convicted but for his advocate's negligence. He would have a remedy in damages but no remedy against the conviction. It is undesirable that a civil action should be treated as an avenue of appeal outside the system which Parliament has laid down for appeals in criminal cases. It is also undesirable that the same issue should be litigated time and again, and there is a strong public interest in the principle of finality.

On the other hand there are other ways of preventing challenges to convictions by collateral means and of ensuring that, if convictions are to be challenged, this must be done by means of an appeal to a criminal appeal court. In Hunter v. Chief Constable of the West Midlands [1982] AC 529 it was held that it was an abuse of the process of the court for a party to seek to litigate the same issue as that which had been the subject of a criminal trial. The power of the court to strike out a civil action on the ground that it is an abuse of process has not yet been recognised in Scotland. But in Law Hospital N.H.S. Trust v. Lord Advocate, 1996 SC 301 it was held that the Court of Session could not sit as a court of review over decisions of the High Court of Justiciary as these two courts had exclusive jurisdiction in regard to all matters falling within their own spheres. On this ground a civil case which was brought in Scotland to challenge a criminal conviction would be dismissed as incompetent.

There remains the argument based on the advocate's duty to the court or, as it has been put, the issue of divided loyalty. But in order to appreciate the force of this argument it is necessary to appreciate the extent of that duty and the extent to which the efficiency of our systems of criminal justice depends on it. The advocate's duty to the court is not just that he must not mislead the court, that he must ensure that the facts are presented fairly and that he must draw the attention of the court to the relevant authorities even if they are against him. It extends to the whole way in which the client's case is presented, so that time is not wasted and the court is able to focus on the issues as efficiently and economically as possible. He must refuse to put questions demanded by his client which he considers unnecessary or irrelevant, and he must refuse to take false points however much his client may insist that he should do so. For him to do these things contrary to his own independent judgement would be likely to impede and delay the administration of justice.

This point was made with equal force by Lord Morris of Borth-y-Gest in the House of Lords in the same case: [1969] 1 AC 191, 251:

He went on to say, at p. 251, that in his view in respect of criminal cases the public advantages of the immunity outweighed the disadvantages overwhelmingly. Lord Upjohn said, at p. 284A, that if the threat of an action was there counsel would be quite unable to give his whole impartial, unfettered and, above all, uninhibited consideration to the case, and that without that the administration of justice would be gravely hampered. Mason C.J. enlarged upon the same point in this passage of his judgment in Giannarelli v. Wraith 165 C.L.R. 543, 556:

In Boland v. Yates Property Corporation Pty. Ltd. 74 A.L.J.R. 209, 241, para. 148, Kirby J. observed that it might be more appropriate to recognise further restrictions on the availability of proceedings against a practitioner in respect of the conduct of criminal rather than civil proceedings.

I consider that the risk is as real today as it was in 1967 in this country and it was in 1988 in Australia that, if advocates in criminal cases were to be exposed to the risk of being held liable in negligence, the existence of that risk would influence the exercise by them of their independent judgment in order to avoid the possibility of being sued. The temptation, in order to avoid that possibility, would be to pursue every conceivable point, good or bad, in examination, cross-examination and in argument in meticulous detail to ensure that no argument was left untouched and no stone was left uncovered. The exercise of independent judgment would be subordinated to the instincts of the litigant in person who insists on pursing every point and putting every question without any regard to the interests of the court and to the interests of the administration of justice generally. As for the objection that to accord advocates an immunity on this ground which is not available to other professionals, the answer to it is as true today as it always was. The exercise by other professionals of their duty to their clients or to their patients may require them to face up to difficult decisions of a moral or ethical nature. But they do not have to perform these duties in the courtroom, where the exercise of an independent judgment by the advocate as to what to do and what not to do is essential to the public interest in the efficient administration of justice.

It may be said that recent reforms to the system of civil justice in England and Wales have greatly reduced the risk of disruption to the administration of justice by the taking of unnecessary points and the development of unhelpful and time-wasting arguments by advocates. As my noble and learned friend Lord Hoffmann has pointed out, the new Civil Procedure Rules have given the judges a battery of powers to keep the resources which the court expends on a case proportionate to its value and importance. The jurisdiction of the courts in England and Wales to make wasted costs orders has been extended to barristers in both civil and criminal cases where costs have been wasted by reason of any improper, unreasonable or negligent act or omission on their part: Court and Legal Services Act 1990, sections 4, 111 and 112.

But the opportunities for judicial intervention in the management of cases are significantly greater in civil cases than they are in criminal cases, where the liberty of the subject is at issue and everything depends on the accused having a fair trial. The system of pre-trial written pleading in civil cases in which both sides are required by the rules to participate assists the process of preliminary case management. In a criminal case written pleadings are largely absent. As the burden of proof throughout is on the prosecutor, very little is required of the accused by way of notice of the case which he wishes to present in his defence. It is much more difficult for the judge to determine when the boundary is reached between that which is necessary for a fair presentation of the defence and unnecessary questioning or time wasting. The power of the judge to make a wasted costs order in a criminal case in regard to the conduct of the case in court by the advocate will need to be exercised with great care once the Human Rights Act 1998 comes into force. It is one thing to penalise the advocate for wasting costs by failing to appear for the trial or for negligent conduct which leads to days being wasted or to the trial being aborted because he is dismissed by his client because of his conduct in the course of it. It is quite another to penalise him in this way for putting what the judge may regard as unnecessary questions or advancing what he may regard as unnecessary arguments. It would be unwise to make any assumptions at this stage as to its effectiveness as a means of reducing the risk of time-wasting by advocates in criminal trials as a result of the loss of immunity.

It is worth stressing in this connection the relevance to this issue of the coming into force of the Human Rights Act 1998. Article 6 of the Convention requires that the accused must receive a fair trial by an independent and impartial tribunal. It also requires that he is entitled to a fair and public hearing within a reasonable time. Both courts and prosecutors will require to observe these requirements. The efficiency of the criminal justice system will be severely tested, and the knock-on effects of delays as one trial follows on another should not be underestimated.

If one wishes to find some empirical evidence about the effects which the coming into force of the Act will have on the conduct of criminal trials in England and Wales it is to be found in Scotland, where compatibility with the Convention rights has been required of all acts of the Scottish Executive, including those of all those prosecuting under the authority of the Lord Advocate since 10 May 1999: Scotland Act 1998, section 57(2). It is no exaggeration to say that the whole climate within which the criminal process is being conducted has been transformed by the requirement of compatibility, especially with regard to the provisions of article 6 of the Convention. Any alleged incompatibility may be raised in any court or tribunal as a devolution issue. Almost without exception the many devolution issues which have been raised since the Scotland Act 1998 came into force relate to the conduct of criminal proceedings. Many of them have been raised by way of preliminary objections, with the inevitable result that delays have occurred in the conduct of criminal trials and substantial additional burdens have been placed on the appeal court. It is likely that similar consequences will be felt in England and Wales when the Act comes into force here. It would be unwise to do anything that might increase this burden unless this was clearly necessary in the public interest.

I would hold therefore that the core immunity pursues a legitimate aim in the field of criminal justice, which is to secure the efficient administration of justice in the criminal courts.

I have already described the risks to the administration of justice. As against that there is the principle that wherever there is a wrong there should be a remedy. How significant is the risk that accused are being deprived of a remedy by the existence of the immunity? Is the effect of the core immunity proportionate to the aim sought to be achieved by it?

The courts have been careful to point out that advocacy is a difficult art and that no advocate is to be regarded as having been negligent just because he has made an error of judgment during the conduct of the case in court. It may be said that the risk of their being subjected to findings of professional negligence is small and that they are adequately protected by the fact that the judges will not hesitate to strike out vexatious actions. But it seems to me that the relevant conclusion to be drawn from these considerations is that the quantity of unsatisfied claims is unlikely to be large.

Some guidance can also be obtained from the experience of the criminal appeal courts in both England and Scotland following the decisions in Reg. v. Clinton [1993] 1 W.L.R. 1181 and Anderson v. H.M. Advocate, 1996 S.C. 29 which established the carefully defined circumstances in which these courts will uphold an appeal based on allegations of negligence in the conduct of the trial by the appellant's advocate. The point that the advocate has been negligent is not infrequently taken but is rarely successful. It is also worth noting, as I said when delivering the opinion of the court in Anderson v. H.M. Advocate, at p. 45A, that difficult questions of professional practice may arise where allegations of this kind are made against counsel or a solicitor. My noble and learned friend Lord Hobhouse of Woodborough has drawn attention to the way in which this problem is currently dealt with in the Court of Appeal in England, and to the fact that to introduce into this scheme of criminal justice a principle that the defendant should be free to sue his advocate in negligence will significantly alter the relationships involved and make the achievement of justice more difficult. Experience in Scotland since the decision in the Anderson case has been that the allegation that the advocate has been negligent has been introduced in a considerable number of cases, sometimes as a last resort after an attempt has been made to introduce fresh evidence. The introduction of this ground causes delay in the disposal of the appeal, as the conflict of interest to which it gives rise renders a change in representation inevitable and the comments of those originally instructed must be obtained. This is because it was held in Anderson that, while it is essential that those against whom the allegations are made should be given a fair opportunity to respond to them, fairness also dictates that they should be under no obligation to do so at the stage when the matter is before the criminal appeal court. Exposure of the advocate to a liability in damages as well as to the existing procedures for professional discipline would be likely to increase the difficulty which the court has already experienced in the conduct of this procedure, which tends to prolong appeals to no good purpose and deprives it of the direct assistance of those originally instructed in the case.

How is one to balance the possibility that a small number of defendants in criminal trials are being denied a remedy against the benefits of maintaining the immunity in the public interest? This involves an assessment of the risks to which all those involved in criminal proceedings would be subjected if advocates were to feel bound to protect themselves in the way I have suggested. The time taken up by this activity would be likely to prolong trials to the inconvenience of members of the public such as jurors and witnesses. The ordeal to which vulnerable witnesses, especially those in rape and sexual abuse cases, are exposed could be extended. Judges in criminal cases are well aware of the difficulty of controlling a line of questioning as they are conscious of the fact that to intervene too frequently or too firmly may provide a ground of appeal in the event of a conviction. The combination of advocates in criminal trials erring on the side of caution in their own interest and of judges erring on the side of caution in the interests of a fair trial would be likely to impede rather than enhance the efficient administration of criminal justice.

On the other side of the balance there are the various mechanisms that are available in the field of criminal justice to prevent a miscarriage of justice if the effect of the advocate's negligence was to deprive the client of his right to a fair trial. Compensation for miscarriages of justice is available out of public funds in the circumstances provided for by section 133 of the Criminal Justice Act 1988, and in other cases ex gratia payments may be made. The advocate is also subject to the disciplinary procedures of his professional body should his conduct in court give rise to legitimate grounds for complaint by his client or at the instance of the trial judge. Your Lordships have not been shown any evidence that might suggest that those who rely on the services of advocates in criminal cases are placed at a significant disadvantage by the existence of the core immunity. On the contrary the removal of the core immunity from advocates in criminal cases would expose them to a significant risk of being harassed by the threat of litigation at the instance of clients who may well be devious, vindictive and unscrupulous but for whom they have felt bound to act in order that they may receive a fair trial.

For these reasons I do not think that the existence of the core immunity in the field of criminal justice is disproportionate to the aims that are sought to be achieved by it.

The present cases demonstrate that there are grounds for concern that the boundaries of the core immunity are at risk of being enlarged, in civil cases, beyond the limits that require to be set to it in the public interest. But, having examined the careful summary of the decided cases since Rondel v. Worsley [1969] 1 AC 191 which is set out in paragraphs 29-31 of the Court of Appeal's judgment [1999] 3 WLR 873, 892B- 897G, I have concluded that there is no evidence that the core immunity is exposed to the same risk in criminal cases. Furthermore the Court of Appeal were careful to say in paragraph 41 of their judgment, at p. 901E, that it was not open to them to question the existence of the core forensic immunity upheld in Rondel v. Worsley [1969] 1 AC 191, nor to doubt the limited extension recognised in the Saif Ali case [1980] AC 198. They recognised that it was plain from the tenor of the majority speeches in the Saif Ali case that any extension beyond the core immunity must be rigorously scrutinised and clearly justified by considerations of public policy; see also paragraph 48(6) at p. 904C-D where the same point is made. But there is no indication in the judgment that the core immunity itself was being called into question. While these observations can be taken to indicate that in their view there was a case for a re-examination of the immunity, I do not read them as amounting to an invitation to your Lordships to abolish entirely the core immunity. A critical re-examination need not go that far. A redefinition of the core immunity so that it is strictly confined within its proper limits may be a satisfactory alternative. Abolition should not be resorted to unless it is plain that it is clearly the only practicable alternative.

It is also worth noting that in two recent cases in Scotland involving allegations of negligence against a solicitor and an advocate following the settlement of a civil case on terms which the client regarded as unsatisfactory the opportunity to plead the immunity was not taken: Crooks v. Lawford Kidd & Co., 1999 G.W.D. 14-651; Crooks v. Haddow, 2000 G.W.D. 10-367. I have not detected any signs, other than the arguments which were advanced by the defendants in the present cases, that the core immunity in criminal cases would be likely to be pressed beyond the limits which can properly be set for it on grounds of public policy. I am not aware of any cases in Scotland where the application of the core immunity in criminal cases has given rise to concern on this ground.

I have already mentioned the cases from Australia and New Zealand in which on grounds of public policy in those countries the decisions in Rondel v. Worsley [1969] 1 AC 191 and the Saif Ali case [1980] AC 198 have been followed and applied. The question is whether any useful guidance can be gained from the position in other jurisdictions, notably the United States, other countries within Europe and Canada. My immediate response to it is to note Lord Reid's observation in Rondel v. Worsley, at p. 228E, that he did not know enough about conditions in any other country apart from England and Scotland to express any opinion as to what public policy there may require.

In regard to the United States it is necessary to distinguish between prosecuting and defence attorneys and between the position in federal law and that in each state. It has long been recognised that judges and prosecuting attorneys should be protected by immunity in relation to their conduct of legal proceedings. In Imbler v. Pachtman (1976) 424 U.S. 409 the Supreme court held that a state prosecutor had absolute immunity for the initiation and pursuit of a criminal prosecution, including the presentation of the state's case at a trial. On the other hand, in Ferri v. Ackermann (1979) 444 U.S. 193, the court held that the federal law of judicial immunity which protected prosecutors and grand jurors did not extend to the defence attorney, since he owed nothing more than a general duty to the public and was required to serve the undivided interests of his client. But the court also held in that case that each state had the right to determine for itself the extent and scope of any immunity acting on the basis of empirical data available to the state. Counsel for the Bar Council have drawn your Lordships' attention to the fact that some states have fashioned rules of immunity for the benefit of public defenders in criminal cases in view of the disruption and costs which would flow from the burden of defending civil claims, from which an analogy may be drawn as to the considerations of public policy which favour of immunity for advocates who provide services in this country under criminal legal aid - bearing in mind the existence of the cab rank rule and the constraints on legal aid fees in criminal cases. While Connecticut (Spring v. Constantino (1975) 362 A.2d 871) and Pennsylvania (Reese v. Danforth (1979) 406 A.2d 735) have not adopted such an immunity, the more recent trend in other states has been to uphold legislation granting immunity to public defenders: e.g. Nevada (Morgano v. Smith (1994) 879 P.2d 735); Delaware (Browne v. Robb (1990) 583 A.2d 949); Vermont (Bradshaw v. Joseph (1995) 666 A.2d 1175); and New Mexico (Coyazo v. State of New Mexico (1995) 897 P.2d 234).

The position in continental Europe is that advocates who undertake criminal cases in those countries do not have the benefit of immunity. But the role and duties of the advocate in those countries differ in significant respects from those of advocates under our systems of criminal justice. Many of the functions of the advocate under our systems of identifying and investigating the facts are performed by the judge in those countries, who does have immunity so long as he is exercising judicial functions in good faith. In that respect there is no inconsistency with the availability of the core immunity under our systems to the defence and prosecution advocate. Beyond that, the much wider scope which is accorded to the judicial function under the continental systems makes it very difficult to draw any useful comparisons.

The position in Canada is quite different. There never was a rule of immunity at common law in that country, and when the matter came up for review in the light of Rondel v. Worsley [1969] 1 AC 191 in Demarco v. Ungaro (1979) 95 D.L.R. (3d) 385 the court declined to introduce such a rule. There is no evidence that its absence has given rise to difficulty, perhaps because it was made clear that the court would be slow to conclude that a decision made by a lawyer in the conduct of the case was negligence rather than a mere error of judgment.

My noble and learned friend Lord Steyn has said that he would regard the Canadian experience as the most relevant but I do not see, with great respect, why that should be so. I should have thought that the Australian and New Zealand experience was the more relevant, as their jurisprudence is more closely modelled on that of our own jurisdictions and the way in which law is practised there is closer to the way law is practised here than it is in Canada. I also think that the distinction which has been drawn in the United States by the Supreme Court between the position of the prosecutor and that of the defence attorney is worth noting in our own jurisdiction. Whatever may be said about the position of defence advocates, it is plainly essential to the administration of justice that prosecuting advocates should continue to be protected by the absolute immunity from action in respect of their conduct of the prosecution case.

The conclusion which I would draw from the comparative material is that, taken as a whole, it does not suggest that we would be falling into a serious error if we were to hold on grounds of public policy that the core immunity against claims by their clients for negligence should continue to be available to advocates in criminal cases.

The Court of Appeal said [1999] 3 WLR 873, 900B-C that it seemed to them that the first question to be asked on any application to strike out or dismiss a claim for damages against lawyers based on their allegedly negligent conduct of earlier proceedings was whether the claim represented an abusive collateral challenge to an earlier decision of the court, that if it did represent such a challenge it should be dismissed or struck out and that this principle applied to claims against lawyers whether or not they were acting as advocates. But it was suggested in the argument in this case that the principle was itself a sufficient protection against unmeritorious claims and that for this reason the core immunity can now be discarded as unnecessary.

I am not persuaded that the principle which was applied in Hunter v. Chief Constable of the West Midlands Police [1982] AC 529 provides the protection which is needed to serve the public interest in the field of criminal justice. I accept that all cases which can be treated as amounting to a collateral challenge to a subsisting conviction will be dismissed or struck out on this ground. But the pattern of the protection is incomplete. There are various events which may arise in the course of a criminal trial, such as things done or not done which may cause delay or continued detention in custody, which may operate to the client's disadvantage irrespective of the question whether he is in the end of the day acquitted or convicted or, if he is convicted, the conviction is set aside. Then there is the problem about what happens if the conviction is set aside on appeal. The appeal may have been taken on grounds other than that the advocate was negligent because the high standard which is needed to set aside a conviction on that ground cannot be satisfied. But once the conviction has been set aside the way will be clear for allegations which would not satisfy that standard to be made because the client's action can no longer be dismissed or struck out as an abuse of process. It should not be forgotten that the setting aside of the conviction does not of itself mean that the client no longer has a claim in damages: see Acton v. Graham Pearce & Co. [1997] 3 All E.R. 909 He may have been detained in custody, or lost his job or suffered in other ways for which he may wish to be compensated.

A further problem about the Hunter case is that on its own facts it was directed to a different issue than that which will arise where the client seeks to recover damages from his advocate on the ground that his conduct of his defence was negligent. It was possible without much difficulty to say that the allegations which were made in that case were simply a repetition of allegations which had been made and disposed of in the course of the trial. But the position of the advocate is different. The question whether his conduct of the defence was negligent is something which arises outwith the trial process. There may be cases where it can be said that the question whether the conviction was attributable to the advocate's negligence is designed simply to cast doubt on the conviction. If so, it will fall within the category of a collateral attack. But I am not satisfied that that will be so in all cases. The Hunter principle, if it is applied too widely to deny the client a remedy in damages, seems to me to be vulnerable to attack on the ground that it is inconsistent with the client's fundamental right of access to a court for the determination of his civil rights. The justification for the core immunity rests upon factors which are directly related to the role of the advocate and his duties to the public and to the court in the interests of the administration of justice. The range of considerations which may lead to the conclusion, in the exercise of the court's discretion, that there is an abuse of process are much more loosely defined and are thus likely to be more difficult to justify if challenged on the ground that they are inconsistent with the client's rights under the Convention.

I would therefore hold that the Hunter principle does not provide a sound basis for discarding the core immunity in criminal cases.

As I have already indicated in my discussion of the position as it affects the system of criminal justice, the public policy considerations are significantly different in civil cases. I do not think that this is to be attributed simply to the changes which have taken place as a result of the introduction of the Civil Procedure Rules. The whole atmosphere in a civil case is different, as so many of the decisions as to what is to be done in the courtroom are taken out of court when the pressures and constraints which affect proceedings in court are absent and there is time to think and to assess the implications of what is being done or not done. It is also much easier for the judge in a civil case to exercise control over the proceedings than it is for a judge in a criminal trial. The risks to the administration of justice which would flow from the removal of the immunity of the advocate against claims by his client for negligence are far less obvious, and the continuation of the immunity is for this reason that much more difficulty to justify.

A further reason for regarding the core immunity in the civil field as no longer justifiable is the difficulty of finding a satisfactory way of defining the limits of that immunity. The test which was identified by McCarthy P. in Rees v. Sinclair [1974] 1 N.Z.L.R. 180 is whether the particular work on which the advocate was engaged was so intimately connected with the conduct of the case in court that it can fairly be said to be a preliminary decision affecting the way the case was to be conducted when it came to a hearing. But experience has shown that it is not an easy test to apply in regard to civil proceedings, especially in regard to allegations made about negligence in agreeing the terms of settlement: see, e.g., Kelley v. Corston [1998] Q.B. 686. It has not proved possible to devise a satisfactory alternative test for use in the field of civil justice, bearing in mind the overriding need to ensure that the protection given must not be any wider than is absolutely necessary.

I have come to the conclusion therefore that, while the core immunity may still be said to have a legitimate aim in civil cases, its application in this field is now vulnerable to attack on the ground that it is disproportionate. It is a derogation from the right of access to the court which is no longer clearly justifiable on the grounds of public interest. But here again I would stress the point which I have already mentioned several times, that the immunity to which I refer is the advocate's immunity against claims by his client for negligence. I would retain the immunity of the advocate against claims for negligence by third parties. For example, it is desirable that it should be retained where the position of the advocate in a civil case is analogous to that of the prosecutor - as where he is representing a professional body in disciplinary proceedings which have been brought against one of its members. The tort of malicious prosecution is a sufficient protection for the individual if the proceedings have been brought against him without reasonable and probable cause: see Martin v. Watson [1996] 1 A.C. 74; Taylor v. Director of the Serious Fraud Office [1999] 2 AC 177.

I do not think that it would be appropriate to bring to an end the application of the core immunity to work done by advocates in civil cases without saying something about the duty which the advocate owes both to his client, to the public and to the court. A proper understanding of the nature and scope of these duties will help to distinguish between claims which are unmeritorious and those where the advocate may properly be held liable in damages for negligence.

In Batchelor v. Pattison and Mackersy (1876) 3 R. 914, 918 Lord President Inglis, in a passage which was quoted by Lord Morris of Borth-y-gest in Rondel v. Worsley [1969] 1 AC 191, 241 and which laid down the foundations for the rules relating to the professional practice of advocates in Scotland, said:

There are a number of points in this passage which require either explanation or closer analysis when it is being applied to the position of the advocate today, and plainly it requires to be modified in its application to advocates such as the solicitor advocate who enter into contracts with their client. The case was one in which the client had sued both his solicitor and his advocate in the sheriff court for damages for loss and damage which he claimed to have sustained due to what he averred was their negligent conduct of the proceedings on his behalf in a civil action and their disregard of his instructions. His action was dismissed in the sheriff court on the ground that his averments were irrelevant. He then appealed to the Court of Session, where he appeared on his own behalf. It is plain from the judgment that the court was satisfied that there was no substance in the allegations of negligence. The real issue in the case was whether counsel was obliged to obey every instruction of his client or whether, as the court held, the conduct of the case was in the hands of counsel who was entitled to decide what was to be done for the benefit and advantage of his client in the exercise of his own judgment.

For present purposes it is unnecessary to dwell on those sentences in which the Lord President was explaining the basis of the cab rank rule. As for the proposition in the opening sentence that an advocate on undertaking the conduct of a civil case takes on himself an office, this terminology is no longer in keeping with the modern view of his position, which - especially in the light of the decision in Hedley Byrne & Co. Ltd. v. Heller & Partners Ltd. [1964] AC 465 - places a greater emphasis on the duty owed by the advocate to the client.

But it remains the case that duty which the advocate undertakes to his client when he accepts the client's instructions is one in which both the court and the public have an interest. While the advocate owes a duty to his client, he is also under a duty to assist the administration of justice. The measure of his duty to his client is that which applies in every case where a departure from ordinary professional practice is alleged. His duty in the conduct of his professional duties is to do that which an advocate of ordinary skill would have done if he had been acting with ordinary care. On the other hand his duty to the court and to the public requires that he must be free, in the conduct of his client's case at all times, to exercise his independent judgment as to what is required to serve the interests of justice. He is not bound by the wishes of his client in that respect, and the mere fact that he has declined to do what his client wishes will not expose him to any kind of liability. In the exercise of that judgment it is no longer enough for him to say that he has acted in good faith. That rule is derived from the civil law relating to the obligations arising from a contract of mandate which is gratuitous: see Stair, Institutions of the Law of Scotland, 1, 12, 10. He must also exercise that judgment with the care which an advocate of ordinary skill would take in the circumstances. It cannot be stressed too strongly that a mere error of judgment on his part will not expose him to liability for negligence.

I would hold that it is in the public interest that the core immunity of the advocate against claims by his client for negligence should be retained in criminal cases. I would however hold that it can no longer be justified in civil cases. But I consider that this is a change in the law which should take effect only from the date when your Lordships deliver the judgment in this case. I also would dismiss these appeals. But I would do so for the same reasons as those given by the Court of Appeal, and not on the ground that by 1991 it was already clear that the core immunity did not extend to work done by advocates in civil cases.

Two principal issues have been debated in the three appeals before the House. One issue is whether immunity should continue to be granted to an advocate against an action for negligence in respect of his conduct of a case in the course of a trial and in respect of pre-trial work intimately connected with the conduct of the case in court as held in Rondel v. Worsley [1969] 1 AC 191 and further considered in Saif Ali v. Sydney Mitchell & Co. [1980] AC 198. The second issue is the scope of the principle barring a collateral attack on an earlier judgment and the extent of the doctrine stated in Hunter v. Chief Constable of the West Midlands Police [1982] AC 529. I have had the advantage of reading in draft the speech of my noble and learned friend Lord Hoffmann and on the second issue, viewed as a matter separate and distinct from the immunity given to an advocate, I am in agreement with the views expressed by him, and I propose to confine my observations to the issue of the advocate's immunity.

The immunity recognised by the judgments of their lordships in Rondel v. Worsley was grounded upon considerations of public policy. But the primary requirement of public policy, as has been observed in many authorities, is that a person who has sustained loss by the negligence of another who owes him a duty of care should recover damages against the latter. This primary requirement was stated as follows by Lord Simon of Glaisdale in Arenson v. Arenson [1977] A.C. 405, 419C:

When this House in Rondel v. Worsley considered the long established immunity of advocates after the rule could no longer be supported on the ground that the advocate could not be sued because he had no contract with his client, Lord Reid observed at [1969] 1 AC 191, 228C:

The House held that the public interest required the existing rule of immunity to be retained. A number of reasons were given for this decision which have been fully set out in the judgment of my noble and learned friend Lord Hoffmann, but I consider that the essential grounds for the decision were those stated by Lord Wilberforce in Saif Ali v. Sydney Mitchell & Co. [1980] AC 198, 212E:

In Rondel v. Worsley, at p. 227C, Lord Reid observed that public policy is not immutable and that the rule of immunity required consideration in present day conditions in this country. Therefore, like all your lordships, I consider that it is right for this House to reconsider the immunity in the light of modern conditions and having regard to modern perceptions. Nevertheless, I do not think that conditions have changed so greatly in the thirty or more years which have passed since the judgments in Rondel v. Worsley and in the twenty years which have passed since the judgements in Saif Ali v. Sydney Mitchell & Co. that the views of the eminent judges in those cases can be completely discounted as relating to conditions and circumstances which were markedly different from those which exist today. I would be slow to dismiss the opinions of the members of the appellate committee in the former case that counsel could be subconsciously influenced to deviate from his duty to the court by the concern that he might be sued in negligence by his client - particularly as this view was also taken by Mason C.J. in the High Court of Australia in Giannarelli v. Wraith [1988] 165 C.L.R. 543, 557.

However, notwithstanding the weight of the argument which can be advanced for preserving the immunity of advocates, I have come to the conclusion for two main reasons that in assessing the public interest the retention of the immunity in respect of civil proceedings is no longer clearly justifiable and that therefore the immunity should no longer be retained. The first reason relates to public perception. The principle is now clearly established that where a person relies on a member of a profession to give him advice or otherwise to exercise his professional skills on his behalf, the professional man should carry out his professional task with reasonable care and if he fails to do so and in consequence the person who engages him or consults him suffers loss, he should be able to recover damages. This principle accords with what members of society now expect and consider to be just and fair, and I think that it is difficult to expect that reasonable members of society would accept it as fair that the law should grant immunity to lawyers when they conduct a civil case negligently, when such immunity is not granted to other professional men, such as surgeons, who have to make difficult decisions in stressful conditions. I consider that there is much force in the observation of Krever J. in the Ontario High Court of Justice in Demarco v. Ungaro (1979) 95 D.L.R. (3d) 385, 405 in relation to immunity in civil proceedings:

The second reason which leads me to the conclusion that the immunity should no longer be retained in civil proceedings relates to the difficulty which arises in drawing a distinction between that part of the work of an advocate which is entitled to immunity and that part of his work which is not. The work which fell to be considered in Rondel v. Worsley was the advocate's conduct of the case in court, and the claim to immunity was upheld in relation to such work. But their lordships also expressed the opinion that some work done in preparation for a trial was also entitled to immunity. Referring to these expressions of opinion in Saif Ali v. Sydney Mitchell & Co. Lord Wilberforce said at, [1980] AC 198, 214D:

In that latter case, where the alleged negligence by counsel occurred at an early stage before trial when counsel was instructed to settle a draft writ and statement of claim, the House was concerned to define more precisely the circumstance in which immunity did not apply to pre-trial work and it did so by adopting the test stated in the New Zealand decision of Rees v. Sinclair [1974] 1 N.Z.L.R. 180 and holding that the protection only applies where a particular work was so intimately connected with the conduct of the cause in court that it could fairly be said to be a preliminary decision affecting the way that the cause was to be conducted when it came to a hearing.

However this test has proved difficult to apply in practice and has given rise to considerable uncertainty, and I am in respectful agreement with the observation of Kirby J. in the High Court of Australia in Boland v. Yates Property Corporation Plc. Ltd. (1999) 74 A.L.J.R. 209, 238, para. 137:

Because of the difficulty of drawing a clear line to fix the boundaries of the immunity and because in civil proceedings the error which is alleged to constitute negligence, even though committed in court, will often be attributable to a decision taken, as Lord Diplock put it in Saif Ali, in the relative tranquillity of barristers' chambers and not in the hurly-burly of the trial, I consider that when this is linked to the public perception to which I have referred, the balance falls in favour of removing the immunity in civil matters.

However I am of opinion that the public interest requires a different result when consideration is given to the immunity of counsel who defend persons charged with criminal offences. As I have stated, I am in respectful agreement with the opinion of my noble and learned friend Lord Hoffmann that the principle stated in Hunter should ordinarily prevent a convicted person from suing his counsel for negligence unless and until his conviction is quashed on appeal. Therefore the issue of immunity arises in relation to an action brought against defence counsel by a person who has been convicted of a criminal offence but whose conviction has subsequently been quashed or (because the Hunter principle would probably not apply) by a person like the plaintiff Rondel who does not claim that the alleged negligence has led to a wrongful conviction. In respect of actions brought by such persons I am of opinion, applying Lord Reid's test, that the abolition of the rule would probably be attended by such disadvantage to the public interest as to make its retention clearly justifiable.

It has been recognised that the argument for retention of the immunity is stronger in criminal cases than in civil cases. In Rondel v. Worsley Lord Morris of Borth-y-Gest stated at [1969] 1 AC 191, 251G:

In Boland v. Yates Property Corporation Pty. Ltd. 74 A.L.J.R. 209, 241, para. 148 Kirby J. stated:

It is the duty of counsel who carry on a criminal practice to defend persons charged with criminal offences. The performance of this duty is of fundamental importance to the proper administration of the criminal law. Many defendants in criminal cases are highly unscrupulous and disreputable persons and I consider that some of them would be ready to sue their counsel if they knew that it was open to them to do so. I consider that the observations of Lord Pearce in Rondel v. Worsley at [1969] 1 AC 191, 275 are still valid today and apply with particular force to persons charged with criminal offences:

On the occasions when a conviction is quashed on appeal, there will often be no valid ground for alleging that the conduct of defence counsel amounted to negligence. If an error has been made in the course of the trial it may have been made by the trial judge in his ruling on a point of law or on the admissibility of evidence or in his summing up to the jury. In such circumstances I consider that it would be contrary to the public interest to remove the existing immunity from the advocate (including the solicitor advocate) of the defendant whose conviction has been quashed. In relation to the advocate in a criminal case I consider that the argument that he should not be vexed by an action for negligence is a strong one and that the countervailing arguments which I think, on balance, prevail in respect of an action for the negligent conduct of civil proceedings, do not prevail where the allegation relates to the conduct of a criminal trial.

There is no suggestion that the clearly established immunity of a judge in respect of an action for negligence brought against him for his conduct of a trial, whether criminal or civil, should be abrogated; that rule is essential for the proper administration of justice and immunity against action is expressly given to the judges of the European Court of Justice.

The argument that the public interest requires that counsel appearing in a criminal trial, like a judge, should not be vexed by unmeritorious actions for negligence (even though this necessarily means that meritorious claims, which I think would be relatively few, would be struck out) consists, in my opinion, of two strands and not one. One strand is that a judge or counsel must be protected because otherwise he may be consciously or subconsciously influenced to deviate from his duty by fear of being sued by a litigant. But a second strand is that it is not right that a person performing an important public duty by taking part in a trial should be vexed by an unmeritorious action and that such an action should be summarily struck out. In the authorities which discuss this matter emphasis is placed on the first strand, but I think it is clear that the authorities also recognise the second strand. The first strand is referred to in the judgments in Munster v. Lamb (1883) 11 QBD 588 but I think that the second strand is implicit in the judgment of Brett M.R., at p. 604:

In Sutcliffe v. Thrackrah [1974] A.C. 727, 736B Lord Reid, when considering the judicial functions of arbitrators, refers specifically to the two strands:

I think that in In re McC. (A Minor) [1985] A.C. 528, 541A Lord Bridge of Harwich had in mind the second strand when he said:

The American Supreme Court has also recognised the two strands in relation to judges and prosecutors. In Imbler v. Pachtman 424 U.S. 409; 422-424 Powell J. states:

In the United States the federal law of immunity has not been extended to defence counsel, although the laws of some states do grant immunity to public defenders.

I respectfully differ from the view of my noble and learned friend Lord Hoffmann that the second strand of the argument that counsel, like a judge, should be protected from vexatious actions is derived from the concept of "divided loyalty" or from the concept that the conduct of litigation is "a difficult art." In my opinion the argument flows from the recognition by the law that those discharging important public duties in the administration of justice should be protected from harassment by disgruntled persons who have been tried before a criminal court. A judge is given protection against an action for negligence although he has no divided loyalty, and he is not given immunity because judging is a difficult art. A judge is given immunity because the law considers that it is in the public interest that he should not be harassed by vexatious litigation. The law does not give immunity to a surgeon who performs very difficult and important work for the benefit of the public. But the reason for this difference is that the administration of criminal justice gives rise to problems and difficulties of the nature described by Lord Pearce in Rondel v. Worsley at [1969] 1 AC 191, 275 which differ from those which arise in the practice of surgery. In my opinion counsel, like a judge, is also entitled to protection in the performance of his public duty to defend persons charged with criminal offences.

There is, of course, an obvious distinction between a judge and defence counsel in that the judge owes a duty to the community to ensure that justice is done in a trial which he conducts and he does not owe a special duty of care to the defendant of the same nature as that of defence counsel who is instructed to appear on behalf of the defendant to represent his interests. There is also a similarity between defence counsel and a surgeon in that each owes a duty of care to his particular client or patient. But in my opinion these considerations are outweighed by the consideration that in representing his client counsel is performing an important public duty which is essential for the proper administration of justice.

It is now the position under the new Civil Procedure Rules that an action which has no real prospect of success can be summarily dismissed more easily than in the past. But this procedure does not give as effective protection against the harassment and vexation of blameless counsel as does immunity; it does not enable the action against counsel to be stopped at once, which is what Brett M.R. thought requisite in Munster v. Lamb at, 11 QBD 588, 605.

Therefore in my opinion the arguments against retaining immunity to protect counsel in criminal proceedings against vexatious actions are markedly weaker than those advanced against retaining immunity for the conduct of civil proceedings. The matter can only be viewed as one of perception, but my own perception would be that counsel who defend in criminal proceedings are at greater risk of harassment from vexatious actions than counsel who appear in civil proceedings because the unpleasant, unreasonable and disreputable persons, to whom Lord Pearce refers, are more likely to be defendants in criminal cases than parties in civil cases. Moreover, for this reason, I think that public perception would be more disposed to accept that it is reasonable and not a ground for criticism to protect counsel from actions by a person who has been charged with a criminal offence as opposed to a person who is a party to a civil dispute. For example, I think that few members of the public would have been critical of Mr. Worsley being granted immunity in order to protect him from being vexed by the action alleging that he had been guilty of negligence for failing to cross-examine to establish that the victim's injuries had been caused by biting or by the use of the accused's hands and not with a knife. There will, no doubt, be some cases in which there has been serious negligence by counsel representing an accused person and where members of the public would feel strongly that the accused person should be able to recover damages, but for the reasons which I have given I consider that it is less harmful to the public interest that such a person should not recover than that in other cases (which I think would be larger in number) blameless counsel should be harassed by vexatious actions.

I consider that the continuation of the immunity of defence counsel appearing in criminal cases would not constitute a breach of article 6(1) of the European Convention on Human Rights. In Fayed v. United Kingdom (1994) 18 EHRR 393, 429, para. 65, the European Court of Human Rights, quoting from Lithgow v. United Kingdom (1986) 8 EHRR 329, 393, para. 194 (stated the relevant principles as follows:

In my opinion the granting of immunity to defence counsel in criminal proceedings is in conformity with these principles. The immunity is in pursuit of the legitimate aim of advancing the administration of justice and of protecting from vexation and harassment those who perform the public duty of defending accused persons so that a criminal court will come to a just decision. The immunity is also proportionate to that aim as it is no wider than is strictly necessary to facilitate the proper administration of justice. Article 6 would clearly not prohibit the domestic law from granting absolute immunity to judges and, for the reasons which I have sought to state, defence counsel is entitled to the same protection.

Therefore I am of opinion that the public interest requires that the immunity of an advocate in respect of his conduct of a criminal case in court and in respect of pre-trial work intimately connected with the conduct of the case in court should continue, notwithstanding the difficulty of drawing a clear line in respect of pre-trial work.

As the present appeals relate to claims for immunity in civil proceedings I consider for the reasons which I have given that they should be dismissed.

All of your Lordships are in favour of dismissing the appeals; the solicitors are not entitled to the immunity which they claim in the present cases. Your Lordships agree that on any view the immunity claimed in these cases falls outside the recognised immunity afforded to advocates. The Court of Appeal arrived at the right conclusion. Further, all your Lordships would be prepared to arrive at the same conclusion on the basis that there is no longer an adequate justification for continuing to recognise a general immunity for advocates engaged in civil litigation.

But that is the limit of the unanimity. Some of your Lordships would be prepared to declare that the immunity should also no longer be recognised for advocates engaged in criminal litigation. Other of your Lordships, among whom I number myself, would not be prepared to take that step on the present appeals. These cases, unlike Rondel v. Worsley [1969] 1 AC 191 (but like Saif Ali v. Sydney Mitchell & Co. [1980] AC 198), do not concern criminal litigation and your Lordships have not heard any argument upon the distinctions that might, still less, should, be made between civil and criminal litigation beyond the generalised discussion arising from the case of Hunter v. Chief Constable of the West Midlands Police [1982] AC 529. That there is room for a difference of opinion on this point cannot be doubted. Further, it is clear that it is not necessary for this difference to be resolved for the purpose of deciding the present appeals. In my judgment, that resolution will have to await a case in which it does arise for decision.

Therefore, it is with the intention of assisting and informing the argument which I consider will have to take place in a later case that I enter upon this subject. Since the question of public policy is based not upon some higher moral imperative but upon a pragmatic assessment of what is justifiable in our society, that assessment may change as circumstances change. The answer that I would give today is not necessarily the same as that which I would give at a later date. I can give two examples of why that might be so. First, lessons may be learnt from the abrogation of the advocacy immunity in civil litigation which will better inform the consideration of the immunity in criminal litigation and the consequences, favourable or adverse, which would follow from its being abrogated as well. Secondly, a new regime of legal representation by quasi-public defenders operating under strict monetary limits is proposed for criminal litigation and it is possible that such a change will so alter the role of the defending advocate as to favour (or even necessitate) unrestricted civil liabilities along the American pattern.

Since the passing of s. 62 of the Courts and Legal Services Act 1990, nothing now turns upon the distinction between solicitors and barristers. This parity has been reinforced by s. 42 of the Access to Justice Act 1999 confirming the paramount duty to the court owed by all those exercising a right of audience. It is accepted that the current immunity (if any) is an advocacy immunity attaching to an advocate exercising his or her rights of audience. It is not a general litigation immunity. The appellants, the solicitors, sought to rely upon the formulation drawn by the House of Lords in Saif Ali v. Sydney Mitchell & Co. from the New Zealand case Rees v. Sinclair [1974] 1 NZLR 180 that the immunity covers what is done in court and preparatory work which is "intimately connected" with the conduct of the case in court. Counsel for the Bar Council argued for a narrower formulation being an immunity confined to conduct in the face of the court but covering any allegation concerning conduct out of court designed simply to evade that immunity.

It is also accepted that any immunity must be justified as being necessary in the public interest, otherwise it cannot survive. Before the 1960s it was thought that a contract was essential to the existence of a duty of care to avoid economic loss and that a barrister did not by accepting instructions enter any contractual or other legal relationship with his lay or professional client. There was simply a mutual absence of legal liability which required no justification. Rondel v. Worsley & Co. for the first time had to consider whether any immunity was justified and if so its extent. Various justifications for a limited immunity were accepted in that case as justified. The extent of the immunity has been revisited in Saif Ali v. Sydney Mitchell. There is no dispute as to the criterion to be applied: the dispute is as to the result.

Counsel for the Bar Council submitted that the rule was in truth a statement that no duty of care existed within the 'immune' area, apparently as an application of the public policy third leg of the 'Wilberforce' test. I do not accept that submission. What is in issue is a true immunity. But in any event, the submitted exclusion of a duty of care was based upon the same criterion as the immunity. Its relevance was to the human rights aspect of the debate. If it were a question of a blanket public policy limitation on the scope of the duty of care, the case of Osman v. United Kingdom [1999] 1 FLR 193 would be directly in point whereas if it is a question of an immunity the criteria laid down in the case of Ashingdane v. United Kingdom (1985) 7 EHRR 528 would govern. These criteria are similar to and no more rigorous than those to be applied under English law to justify the immunity: the immunity must "pursue a legitimate aim" and there must be "a reasonable relationship of proportionality between the means employed and the aim sought to be achieved". (paragraph 57.)

It is of the nature of a rule the continued existence of which has to be justified by the public interest that the balance of public interest may change. A decision such as Rondel v. Worsley is therefore open to review, not because it was wrong when it was decided, but because circumstances have changed since 1967 and it is appropriate that the rule should be reviewed and, if no longer justified, changed or abrogated. It is not a question of whether to over-rule previous authority but of declaring the law in current conditions.

However, the role of Parliament must also be taken into account. Parliament is the primary guardian of the public interest. In most areas of public policy, Parliament will be the sole arbiter and the courts should not allow themselves to trespass into them. But in the present appeals the relevant area is the system of justice and the administration of justice in the courts. In this area the judges have a legitimate competence to declare where the public interest in the achievement of justice lies and what is likely to be the impact of one rule or another upon the administration of justice.

It is also the case that Parliament has quite specifically refrained from intervening in this matter. S. 62 of the Act of 1990 disclosed no disapproval of the existence of an immunity for barristers and others performing a similar function; indeed, it could be argued that s. 62 assumes that there is such an immunity and that it will continue in being. Other statutes, such as the Access to Justice Act 1999, have likewise refrained from abrogating or qualifying the immunity even though such a provision would have been well within the purview of the statute. There are other statutory provisions to which I will refer in the course of this speech which are relevant to the consideration of the broader policy of the legislature and therefore to the existence of the immunity and which should accordingly be taken into account before reaching a conclusion. The leading role of Parliament must be recognised and any decision at which your Lordships were to arrive would have to be one which is consistent with the guidance to be gained from the acts of the Legislature.

Inevitably, Rondel v. Worsley deployed a number of reasons for recognising an immunity. These were commented on by Lord Diplock in Saif Ali. Some are more apt than others and they have already been rehearsed and criticised by several of your Lordships. However it is necessary to analyse some of them further. Some factors which seemed important 30 years ago have ceased to be so now and others which received only a passing reference then can now be seen to be essential to making the right evaluation of where the public interest lies. Likewise, in conducting now a re-examination of the cogency of the various factors, it is necessary to set them in the appropriate current context. The observations which follow are not exhaustive and are merely designed to make some of the points which I consider need to be made.

The advocate, independently of any immunity, has certain protections. The standard of care to be applied in negligence actions against an advocate is the same as that applicable to any other skilled professional who has to work in an environment where decisions and exercises of judgment have to be made in often difficult and time constrained circumstances. It requires a plaintiff to show that the error was one which no reasonably competent member of the relevant profession would have made. This is an important element of protection against unjustified liabilities. Similarly, there now exist improved procedures to enable obviously unsustainable claims to be brought to a conclusion at an early stage of any litigation. The availability of these protective features and their value in discouraging and limiting unmeritorious litigation is relevant when questioning the need for any immunity. The position was not the same in 1967.

I consider that it is not an argument that the immunity is needed to protect advocates against excessive liabilities. There is no evidence that any liabilities to which advocates would be subjected if not immune would be unsustainable or disproportionate. They are in this respect in the same position as any other professional. Such risks are insurable and advocates are now professionally required to carry liability insurance. There is no evidence that satisfactory insurance is not available. Indeed, the aspects of legal practice most obviously liable to give rise to large claims fall outside the scope of any immunity being contended for or, at the least, are likely to do so.

But, in any event, no case is being made - nor can it be made - that lawyers should as a profession be given any special protection. The immunity, if any, must exist for the benefit of the public not the lawyers. Thus, the element of protection only comes in collaterally and consequentially. The immunity, if upheld, would have the effect of protecting advocates from being harassed by unmeritorious claims: the justification would, on this basis, be that to require them to be subjected to such harassment and to have to guard against the risk of it would have a deleterious effect upon the administration of justice. (Munster v. Lamb 11 QBD 588; Roy v. Prior [1971] AC 470) It is the exposure to the risk which does the damage. It inevitably distorts professional practices and professional judgments, likewise the distribution of resources, and, where, as is the case with the practice of advocacy, the existing system is on the whole working well, this distortion will be adverse and will not assist the general good. A comparison of benefit (to the individual litigant) and detriment (to the public as a whole including litigants as a class) has to be made and a balance struck. This is not to devalue the rights of the individual but to recognise that in any communal society such a balance has to be struck. For others involved in the justice system the balance is judged to favour immunity. The question is whether the same judgment should be made for advocates as well.

Before leaving this aspect of protecting the practitioner, there is a difference between the solicitor's profession and that of the barrister which has in the past been of major relevance and is still not irrelevant. A solicitor who feels uncertain about his position can always take the advice of counsel and, provided that the counsel chosen was competent and the advice not manifestly wrong, that will protect the solicitor. The barrister has no equivalent protection, nor in practice does the advocate. The fact that solicitors have in the past successfully operated in a no immunity environment must be evaluated in this context before the same assumption is made for the advocate.

The argument based upon owing a paramount duty to the court (reinforced by s. 42 of the 1999 Act) is of only limited impact and needs further analysis. The relevant argument has to be based upon a conflict of duty. If the duty owed by the advocate to the court is no more than a duplication of his duty to his client, the existence of the duty presents no problem for the advocate: he must simply do his duty. (I will have to come back to other consequences of this later.) However where there is a conflict of duty he may have to make choices which are contrary to the wishes of his client. A threat by a client to sue the advocate may put the advocate in a difficult position particularly where the extent of his duty to the court and precisely what it entails may be itself a matter of judgment or disagreement. Thus the potential for a conflict of duty is a relevant, but far from dominant, factor in the assessment of the need for an immunity.

I am not impressed by the counter-argument that other professional men also owe duties which may conflict with the wishes of their client or patient. Typically these are ethical duties or obligations not to breach the criminal law. Such constraints upon conduct are of a character common to virtually all citizens. They do not as such raise the same potential problem as the conflicts faced by an advocate. The impressive counter-argument is that competent advocates are well able to cope with such conflicts and are confident that, where they adopt a particular view of their duty to the court in good faith, their judgment will be upheld by the court.

There is no evidence that the lack of immunity where it exists causes difficulties with the discharge of the lawyer's duty to the court. The most striking example of this is the duty in civil litigation to give discovery of all material unprivileged documents to the opposing side. Such documents include those of which the only relevance is that they damage the disclosing party's case or support the other side's case. It is contrary to the client's interest that the other side should see them yet it is the solicitor's task and duty to disclose them. Solicitors have for over a century performed this task without immunity from being sued by their clients. (However, as I warned in the previous section of this speech, it is an oversimplification to extrapolate from the position of the solicitor to a dismissal of any problem for the advocate.) Any threat of corruption of the lawyer comes not from the fear of being sued but rather the wish not to lose a valuable client by being over-zealous. (cf. the position of an auditor.) In general the client appears to understand that he is employing the solicitor to perform his, the client's, duty and is content that he should.

This also illustrates the further point that there are two types of duty involved. There are those which are equally duties of the client (eg see CPR 1.3) and there are those which are solely duties of the lawyer and personal to him. The advocate's duties which are relevant under this head come into the latter category and require the advocate to be prepared, in relation to the court, at times to stand apart from his client.

This is a duty accepted by the independent bar. No one shall be left without representation. It is often taken for granted and derided and regrettably not all barristers observe it even though such failure involves a breach of their professional code. It is in fact a fundamental and essential part of a liberal legal system. Even the most unpopular and antisocial are entitled to legal representation and to the protection of proper legal procedures. The ECHR confirms such right. It is also vital to the independence of the advocate since it negates the identification of the advocate with the cause of his client and therefore assists to provide him with protection against governmental or popular victimisation.

The principle is important and should not be devalued. But the relevant question is whether it provides a justification for the immunity. In my judgment it is properly taken into account as a factor since it restricts the freedom of action of the advocate and casts light upon the true nature of his role. (In the procedure of criminal courts, it goes hand in hand with the restrictions upon the ability of the defence advocate to withdraw during the trial.) But it does not in itself justify an immunity. The medical profession would normally accept an ethical obligation to provide medical care without discrimination without seeking any immunity in return. Historically the adoption of a common calling has carried both an obligation to accept all custom and an absolute liability. A common carrier had to accept and carry goods entrusted to him and was absolutely liable for their loss or damage subject to only very narrow exceptions.

This is, or should be, at the centre of this debate and is in my judgment the critical factor which must be evaluated. How does the role of the advocate and any immunity relate to the trial and appeal process? It is the fact that different answers are to be given to this question for the civil process and the criminal process that leads to the conclusion that for one the immunity may no longer be justified but for the other it should be retained.

The trial is where the advocate finally exercises his right of audience and practises his advocacy. It is a process which is unique in that it is conducted before the court or judge. It is under the direct supervision and control of the court or judge. The advocate is subject to a discipline judicially imposed. It is normally conducted in public. The purpose of a trial is to achieve finality and lead to a decisive adjudication.

Any decision reached at the trial is subject to appeal. The appeal is the process provided by the legal system for the rectifying of errors or mishaps which have occurred during the trial. The appeal process itself represents a working out of the policy of the law for qualifying the finality of the trial and incorporates appropriate safeguards. It is upon the appeal process more than upon the trial process that any system of civil fault-based remedies against advocates would encroach. The place for criticising the outcome of the trial and remedying any miscarriage of justice should in principle be the appeal court, not another trial where the advocate is the defendant.

A feature of the trial is that in the public interest all those directly taking part are given civil immunity for their participation. The relevant sanction is either being held in contempt of court or being prosecuted under the criminal law. Thus the court, judge and jury, and the witnesses including expert witnesses are granted civil immunity. This is not just privilege for the purposes of the law of defamation but is a true immunity: Roy v. Prior [1971] AC 470, especially per Lord Morris at pp. 477-8. This rule exists in the interests of the trial process, ie in the public interest. Under Rondel v. Worsley and Saif Ali the advocates have a similar immunity.

It is illuminating to consider the conceptual basis in the trial process for the witness immunity. It is that the witness, although called by a party, is giving evidence to the court. The witness's duty is to tell the truth to the court regardless of the interests of the party who has called him or who is asking him questions. This same scheme is spelled out in the new Civil Procedure Rules regarding expert witnesses. An expert witness is in a special position similar to that of the advocate. He is selected and paid by the party instructing him. Part of his duties include advising the party instructing him. If that advice is negligently given the expert, like the lawyer, is liable. But once the expert becomes engaged on providing expert evidence for use in court (CPR 35.2; Stanton v. Callaghan [2000] 1 QB 75) his relationship to the court becomes paramount as set out in the CPR and he enjoys the civil immunity attributable to that function.

If the advocate is to be treated differently, he alone of these participants in the trial will be being held civilly liable for what he does and does not say in court. This anomaly will require justification. The anomaly is not without further significance in that, if the advocate is to be held civilly liable for some adverse outcome of the trial, he will have to bear the whole loss even though other participants may have been equally, or more seriously, at fault. From the point of view of the aggrieved party, if some fault can be found with the performance of the advocate, he recovers in full from the lawyer; but, if only other participants were at fault, he recovers nothing at all. It is necessary to be very cautious before correcting one perceived anomaly by creating another.

A further feature of the trial process is its finality (subject to appeal). Some judgments establish a status in private or public law, others do no more than establish a liability, or non-liability between one individual and another. There are developed rules governing those who are bound by judgments and under what circumstances they can be challenged. A civil judgment itself creates rights which are distinct from, and in which may merge, rights which existed before. It is thus important to consider the relationship between the original trial which has given rise to the client's complaint and the subsequent litigation between the client and his advocate. Does the subsequent litigation challenge or affirm the outcome of the previous trial? If it affirms it, no problem arises. If on the other hand, the substance of the later litigation is to challenge the outcome of the previous trial, then a question of finality can arise. It may be a challenge to the status of the previous decision. This is a point to which I will have to return and is a cardinal point of distinction between the criminal and civil process.

This in turn ties in with the consideration of the interest of the client which the law of tort, if available, would serve to protect. The law of negligence exists to provide monetary compensation for losses capable of being valued in monetary terms. Where the loss suffered by the client is financial, the remedy is appropriate and effective. Where the complaint has a different character, as for example that the client has been convicted of a crime which he says he did not commit, an action in tort does not remedy that grievance and can at most provide a solatium or some means of visiting punishment upon the advocate alleged to have failed to secure an acquittal. Such a complaint also has the necessary character of challenging the conviction; it involves saying that an innocent man has been wrongly convicted.

To permit actions which involve a re-examination of a trial that has already occurred and a judgment already given inevitably must trespass on the finality of that trial and judgment and the appeal procedure and involve some duplication of the previous process. Accordingly such permission requires justification.

Another point which emerges from this discussion is that the oft resorted to analogy with the medical profession and its lack of immunity breaks down. The advocate's conduct is already public and within the purview of the judicial system both at the trial and on appeal. It is not necessary to permit negligence actions to be started in order to achieve this judicial control; nor is it necessary in order bring the advocate's conduct into the public domain.

Finally, in connection with the litigation process, one of the remedies it provides to the dissatisfied client is the ability to challenge the fees and expenses charged by the lawyer to the client. It is possible for the client to procure that those charges are disallowed or reduced on taxation. It is not necessary for him to bring an action for damages to achieve this result. Similarly the court has the power to make wasted costs orders against a litigator or advocate which consequentially benefit the litigants. (Ridehalgh v. Horsefield [1994] Ch 205)

The ability to stay or strike out an action as an abuse of the procedure of the court is a long standing remedy, an inherent power of the court, and is reflected in the CPR and their predecessors. Its essence is the use of civil litigation for an improper purpose, ie without a legitimate purpose. Where a client is seeking to recover damages from his former advocate for some breach of duty, this is clearly a proper purpose if the advocate is not immune. It is important to stress this at the outset as it has been submitted by the respondents that abuse of process provides a satisfactory solution to any problems arising from denying the existence of the immunity. It is not a substitute for the immunity. It is rather one of the existing features of the law, like the standard of care applied in professional negligence cases, against which to test the necessity of having the immunity. Another point to stress at the outset is that 'collateral attack' only comes into the picture when it discloses an abuse of process. It is a distinct concept and challenging a previous decision does not necessarily connote an abuse of process.

Rondel v. Worsley was a case where the claim could in any event have been struck out as disclosing no reasonable cause of action. Rondel did not suggest that his advocate caused him to be convicted; his grievance was that the advocate had not pursued sufficiently forcibly his allegation that he had used his hands and teeth to inflict the relevant injuries. It was not a case where there was any attack, collateral or direct, upon the jury's verdict. Neither that principle nor the decision in Hunter v. Chief Constable of the West Midlands Police would have caused the action to be halted.

A similar point is to be made in relation to Hunter. In that case the plaintiffs had been convicted of a terrorist offence substantially on the basis of their admissions to the police. At the criminal trial they contended that the confessions were involuntary as they had been beaten by the police. The trial judge, after a voir dire, rejected their evidence and preferred that of the police. The jury convicted them. Subsequently they sued the police for assault. They were trying to relitigate in a civil court the same issue as had been in dispute at the criminal trial and had been decided against them beyond reasonable doubt. It was a case of a collateral attack both on the trial judge's finding and upon the verdict of the jury. The courts and your Lordships' House held that the civil action was an abuse of process and should be struck out. It was not however an action against their lawyers. If they had had a bona fide complaint against their lawyers and had sued them, there would have been no reason why, subject to the immunity point and presenting a reasonable case on breach of duty, their action should not have gone ahead. The immunity point and the abuse of process point are distinct and separate. They do not serve the same purpose.

The 'collateral attack' point is a species (or 'sub-set') of abuse of process. There is no general rule preventing a party inviting a court to arrive at a decision inconsistent with that arrived at in another case. The law of estoppel per rem judicatem (and issue estoppel) define when a party is entitled to do this. Generally there must be an identification of the parties in the instant case with those in the previous case and there are exceptions. So far as questions of law are concerned, absent a decision specifically binding upon the relevant litigant, the doctrine of precedent governs when an earlier legal decision may be challenged in a later case.

A party is not in general bound by a previous decision unless he has been a party or privy to it or has been expressly or implicitly covered by some order for the marshalling of litigation. (Ashmore v. British Coal Corporation [1990] 2 QB 338) This overlaps with the concept of vexation where the same person is faced with successive actions making the same allegations which have already been fully investigated in a previous case in which the later claimant had an opportunity to take part. This reasoning does not apply to an action against a lawyer alleging that he has mishandled a previous case.

The case of Hunter is not apt or adequate to deal with cases brought by aggrieved clients against advocates alleged to have been negligent.

(1) The immunity of the advocate, if it is to be upheld, must be justified as necessary in the public interest.

(2) Rondel v. Worsley represented the assessment of where the public interest lay at the time it was decided in 1967.

(3) Parliament has not sought to abolish the immunity and has implicitly left it to the courts to consider whether the immunity should survive.

(4) Statutes have however not been silent upon relevant aspects of the public interest and such guidance must be respected and followed.

(5) There is a balance to be struck. There are factors to be placed on either side of the scales.

(6) The most important factors are the assessment of the role of the advocate in the court process and whether the interest of the client would be appropriately protected by the tort remedy.

(8) The abuse of process tool is no more than a relevant part of the existing law and does not address the same question as the immunity and does not provide a substitute for it.

(9) I consider that the balance of the public interest needs to be examined separately for the civil and the criminal process. The Civil Process:

The civil process includes most of the factors to which I and others have referred. The question is how potent they are and whether they still suffice to justify the immunity of the advocate in civil litigation. My Lords, in agreement with your Lordships, I consider that they do not.

The character of civil litigation is that it involves the assertion by one party that the other has infringed his rights; he seeks a remedy, normally a monetary remedy but sometimes a remedy of declaration of right or specific implement. The court, therefore, has essentially to make a decision between two conflicting parties and determining their respective rights inter se. It is primarily the provision by the state of a service similar to the provision of arbitration services. The public interest does not normally come into it save in so far as the provision of a system of civil dispute resolution and the enforcement of civil rights is a necessary part of a society governed by the rule of law not by superior force.

It is a system of relative justice. It exists in economic terms. The plaintiff complains that he has suffered loss and damage; he claims that the defendant should be required to pay monetary damages to compensate him; the remedy is a redistribution of wealth between the parties. Or he may assert a property right and ask that the court should assist him enforce it against the defendant. If something goes wrong in the litigation, the court does not simply ask whether the party directly affected will suffer an injustice if not assisted by the court, eg by having his time for doing some act extended, or by being allowed to amend his case. It asks whether assisting one party will cause an injustice to the other. Where the mishap has resulted from some act or omission of a party's lawyer, that party may be left to his remedy against his own lawyer rather than to allow the mishap to prejudice the other party. If all potential for a liability of the lawyer to his client is excluded, this will make it more difficult to do justice between the plaintiff and the defendant not less difficult.

The same applies on an appeal. The primary concern is not the fairness of the trial but its outcome; can the appellant show that he not the respondent was entitled to succeed? Complaints by an appellant against his own advocate will rarely advance his case because they will not normally impinge upon the case of the respondent. New evidence is only admitted under very restricted circumstances: Ladd v. Marshall [1954] 1 WLR 1489. The reasoning is that the unsuccessful party is not entitled to deprive the other of his judgment without showing cogent reasons as against that other.

If a party has suffered a loss, either by being held liable to the other party or through failing to recover from the other, that financial loss represents the starting point of the claim of the client against his lawyer and the remedy he claims from his lawyer, an award of damages expressed in monetary terms is the appropriate remedy for the wrong complained of. The dominant relationship is that of the lawyer and his client. The introduction of conditional or contingent fees gives the lawyer a financial interest in the litigation which only serves further to emphasise the commercial character of the relationship and the commercial enterprise in which they have joined.

A successful claim against the lawyer does not attack the position of the other party to the original litigation. It affirms that outcome of the original litigation as having established conclusively, as between those parties, their rights inter se. The client alleges that that outcome was caused by the failure of his lawyer to provide the stipulated service. This not different in kind to a client saying that the adverse tax treatment of a transaction was caused by the negligent advice or drafting of the lawyer he employed. It will not be cured by an appeal in the litigation.

In the preceding paragraphs I have simply referred to the client's lawyer because what I have said is equally true of both the in-court advocate and out-of-court litigator. It assists the doing of justice between plaintiff and defendant in civil litigation that the client's rights against his lawyers of any kind be preserved in full and the economic remedy is the right remedy. The appeal process is not apt to provide the remedy

One of the problems of any immunity is determining its boundaries. In civil litigation, defining the boundaries of what constitutes advocacy and would therefore qualify for the advocacy immunity is a serious problem not capable of satisfactory solution. The position has been made more difficult by the CPR. There is not a single moment of confrontation. The exercise of advocacy extends over a series of processes of which the trial is only one and the advocacy may be conducted as much in writing as orally. Counsel for the appellants signally failed to provide a satisfactory definition or categorisation of the functions to which, in civil procedure, the immunity would attach. This is a telling argument against the recognition of an immunity for advocates for civil procedure and has assisted to convince me that the immunity is not necessary or appropriate. In civil litigation the immunity is anomalous and the arguments in its favour, although they exist, do not suffice to justify its continued existence.

Even though the criminal process is formally adversarial, it is of a fundamentally different character to the civil process. Its purpose and function are different. It is to enforce the criminal law. The criminal law and the criminal justice system exists in the interests of society as a whole. It has a directly social function. It is concerned to see that the guilty are convicted and punished and those not proved to be guilty are acquitted. Anyone not proved to be guilty is to be presumed to be not guilty. It is of fundamental importance that the process by which the defendant is proved guilty shall have been fair and it is the public duty of all those concerned in the criminal justice system to see that this is the case. This is the public interest in the system.

The criminal trial does not exist to protect private interests. It exists as part of the enforcement of the criminal law in the public interest. Those who take part in the trial do so as a public duty whether in exchange for remuneration or the payment of expenses. The purpose of all is, or should be, to see justice done and to play their appropriate part in achieving that end. The proceedings are conducted in public under judicial control. The position of the advocates is the same as that of the other participants. The prosecuting advocate has a duty to see that the prosecution case is, on behalf of the Crown, presented effectively and fairly. That of the defending advocate is to see that the defendant has a fair trial, that the prosecution case is properly probed and tested both in fact and in law and that his factual and legal defences are properly placed before the court supported by the available evidence and arguments. The same applies to criminal appeals: the purpose and the roles of the participants are the same.

It follows from these fundamentals that the salient features of this procedure exist to serve the public interest, not to serve any private interest. The defendant is entitled to skilled professional representation and, if he cannot provide it for himself, it will be provided for him at public expense, as happens in virtually all cases. It is likewise necessary that the advocate having the task of representing the defendant shall be independent and fearless. If he is not he will not be equipped to discharge the public duty entrusted to him to see that the defendant has a fair trial and that he is not convicted unless proved guilty. The advocate is performing a public function in the public interest. It is his public duty to protect the interests of his client. The criminal justice system depends upon his doing so skilfully and independently.

The other participants have a similar public duty to perform their role. They take part in the trial as a public duty. All must be concerned to see that the defendant has a fair trial. Thus the judge and the prosecuting counsel will join in seeing that errors of fact or law are not made. It is the judge's duty to direct the jury on defences available on the evidence and to exclude inadmissible or unfair evidence. It is the duty of both counsel to draw the judge's attention to any errors he may have made. All witnesses are under a duty impartially to assist the court and give honest evidence. If the defence advocate is to be exposed to a civil liability in respect of his discharge of his public duty and the role he has to perform in the criminal trial process, he will be unique among the participants. All the others are in the public interest immune; the same logic applies to the defence advocate whose role derives from the same public interest and is just as important to the public interest as that of the other participants. As previously observed, if he alone is to be subjected to civil liability, he will be unable to obtain a contribution from any other participant although they may be equally blameworthy for what went wrong. The scheme is that the participants are subject to the jurisdiction of the court and the court has appropriate disciplinary powers to control the proceedings and the conduct of the participants. In cases of serious misconduct, it is the criminal law which intervenes not the civil law.

The appellate procedure follows the same logic. The only question on an appeal against conviction is whether the conviction was unsafe. If it was, then the appeal must be allowed: in all other cases the appeal must be dismissed: the Criminal Appeal Act 1968 section 2 (as amended). A whole variety of factors may affect the safety of a conviction - error of law, the admission of evidence which ought to have been excluded, some unfairness in the trial or the summing-up, relevant evidence not adduced at the trial. The powers of the Court of Appeal to admit fresh evidence or extend the time for appeal are wide; they are not constrained by the consideration of the interests of any other person. They are to be exercised whenever it would serve the interests of justice. Pleas, admissions and concessions can where it is just to do so be withdrawn.

The Court of Appeal will also listen to criticisms of the conduct of the defence and give effect to them when they have merit. It is hard to visualise a case where the criticism would (in the absence of immunity) be sufficiently substantial to justify a claim against the advocate but not give a ground of appeal which the Court of Appeal would have to evaluate. Similarly, when, at a later time, new factors arise which justify the reconsideration of the safety of the conviction, the case can be referred back to the Court of Appeal under the Criminal Appeal Act 1995, s. 9. The duty of the advocates appearing before the Court of Appeal are the same as at the trial, the achievement of a just outcome. Their role is adversarial but their duty is not partisan.

The prosecuting advocate is not in practice subjected to any consideration of personal liability for his conduct of the case. (Indeed, a general non-liability in negligence of the Crown Prosecution Service has been upheld by the Court of Appeal on policy grounds: Elguzouli-Daf v. Commissioner of Police of the Metropolis [1995] QB 335.) If he has to revisit what occurred at the trial, it will be solely to provide further assistance to an appellate court or other similarly placed body. The defending advocate will normally conduct any appeal from a conviction (or sentence). He will do so in the same interest as before, the interests of justice. If some question arises about his conduct of the trial, this will probably make it inappropriate that he represent, or continue to represent, the defendant on the appeal. But he will remain under a duty to assist the Court of Appeal. Normally the defendant will waive his privilege and a full and frank written account of what occurred and the reasons for it will be given by the advocate to the Court of Appeal. It will readily be appreciated that to introduce into this scheme of criminal justice a principle that the defendant should be free to sue his advocate for damages in negligence will significantly alter the relationships involved and make more, not less, difficult the achievement of justice within the criminal justice system which is its purpose and is also the public interest.

My Lords, I make no apology for emphasising the position on criminal appeals: the reason why the question of immunity arises is because of the argument that a defendant who has been the victim of a miscarriage of justice should have a remedy. On any view the primary remedy must be the criminal appeal. Therefore the primary inquiry must be how the abrogation of the immunity would affect the effectiveness of the Court of Appeal in rectifying such miscarriages. If its existence facilitates such rectification, that is a very strong argument indeed in justification of the immunity. (Contrast the position in the civil justice system where the position is the reverse.) To displace this justification needs some significant counter-argument. However, the evaluation of the other available arguments support rather than undermine the justification for the immunity.

The legitimate interest of the citizen charged with a criminal offence is that he should have a fair trial and only be convicted if his guilt has been proved. It is not an economic interest. His interest like his potential liability under the criminal law stems from his membership of the society to which he belongs - his citizenship. If the charge against him has not been proved, he should be acquitted. If he has been wrongly convicted, his appeal against conviction should be allowed. If he has been wrongly or excessively sentenced, his punishment should be remitted or reduced. His only remedy lies within the criminal justice system. This is appropriate. The civil courts do not have any part to play in such matters. The relevance of what the advocate does during the criminal trial is to the issues at that trial, not the remoter economic consequences of the outcome of that trial.

Any involvement of the citizen in the criminal justice system may have adverse consequences. There are adverse consequences for witnesses which they in the public interest have to accept. There are certainly adverse consequences for those suspected of or charged with criminal offences. They may be held in custody. They normally have to attend their trial. They may be arrested and subjected to interviews or searches or tests which would otherwise be an infringement of their civil liberties. They may be acquitted after a long and traumatic trial. They may be convicted but have their conviction overturned on appeal. Thus they will to a greater or lesser extent suffer disadvantage and loss including loss of liberty and reputation.

Provided that the relevant persons have acted in good faith, the citizen has to accept this as part of the price he pays for living in the community and enjoying the protection of the criminal law. A defendant who is detained in custody but acquitted at his trial receives no compensation for his loss of liberty or for having had serious allegations made against him. The same applies if he is convicted and sentenced at his trial but has his conviction quashed on appeal. He too receives no compensation. Those who have paid for their own defence have no assurance that they will necessarily be awarded costs.

An unsafe or wrong conviction may have occurred for any of a number of reasons. Someone may be to blame or there may have been no fault on anybody's part. It may arise from something that happened at the trial, eg erroneous expert evidence, or outside court, eg undiscovered evidence. There may have been some defect in the conduct of the trial like the failure of the judge or counsel to anticipate a restatement of the law by an appellate court. There is no need to proliferate examples; the diverse and various possibilities will be well within the experience of any one actively engaged in the criminal justice system. It will also be readily appreciated that some of these factors may be apparent at the conclusion of the trial; others may only come to light much later.

The payment of monetary compensation is something upon which Parliament has spoken. The statutory policy is set out in s. 133 of the Criminal Justice Act 1988 (as amended) . This provides, under the heading "Compensation for miscarriages of justice":

"(1) Subject to subsection (2) below, when a person has been convicted of a criminal offence and when subsequently his conviction has been reversed or he has been pardoned on the ground that a new or newly discovered fact shows beyond reasonable doubt that there has been a miscarriage of justice, the Secretary of State shall pay compensation for the miscarriage of justice to the person who has suffered punishment as a result of such conviction or, if he is dead, to his personal representatives, unless the non-disclosure of the unknown fact was wholly or partly attributable to the person convicted.

(2) No payment of compensation under this section shall be made unless an application for such compensation has been made to the Secretary of State.

(3) The question whether there is a right to compensation under this section shall be determined by the Secretary of State.

(4) If the Secretary of State determines that there is a right to such compensation, the amount of the compensation shall be assessed by an assessor appointed by the Secretary of State.

(4A) In assessing so much of any compensation payable under this section to or in respect of a person as is attributable to suffering, harm to reputation or similar damage, the assessor shall have regard in particular to—

The statute distinguishes between those factors which come to light in time to be considered on a normal first appeal to the Court of Appeal (no compensation) and those which only come to light later (potential compensation). Similarly it distinguishes between new (or newly discovered) facts and errors of law or other non-factual matters. There is a statutory policy, reflected also in the way in which the Home Secretary exercises his discretion, which strikes a balance between those encounters with the criminal justice system which the state should compensate and those which it should not. The discretionary element is similar to that contained in the criminal injuries compensation scheme. Those who have encounters with criminal activity are not all equally meritorious. The policy of the Legislature (and Executive) is not to provide indiscriminate compensation for erroneous convictions. To do so would be unacceptable in a liberal democratic society. My Lords, we should respect that assessment of the public interest and the needs of our society.

To provide a tort based liability to pay compensation in respect of the role of only one of the participants in the criminal justice system would not only destroy this balance but also produce a capricious distribution of compensation between ultimately acquitted defendants. If a defendant could say that a (I stress, a) cause of his conviction was the fault of his advocate, he would recover full civil damages; if it was the fault of anyone else involved in the trial, he could not recover anything unless he came within the scope of s. 133. From the defendant's point of view, it would be an arbitrary lottery and produce anomalies between one defendant and another. As a matter of statutory policy, it would provide a route by which the statutory limitations and safeguards built into s. 133 could be avoided. From the point of view of the administration of justice it would expose the professional advocate to a risk of litigation which would handicap him in performing his duty under the criminal justice system and disinterestedly assisting, particularly at the appellate level, in the correction of errors and remedying miscarriages of justice. To argue for a higher need for a supposed redistributive justice to enable the defendant to recover civil damages from his advocate, begs the question where the greater justice lies in relation to criminal litigation as well as the question whether such a need is indeed higher than the need to facilitate as far as possible the rectification of miscarriages of justice within the criminal justice system.

In summary, there are essential differences between the civil and criminal justice systems. In the civil justice system, the nature of the advocate's role in the whole process, the nature of the subject matter, the legitimate interest of the client, the appropriateness of the tort remedy and the absence of clear or sufficient justification all militate against the recognition of an advocate immunity. It is not necessary: in certain respects it is counterproductive.

In the criminal justice system, the position is the reverse of this. The advocate's role, the purpose of the criminal process, the legitimate interest of the client, the inappropriateness of the tort remedy, the fact that it would handicap the achievement of justice, the fact that it would create anomalies and conflict with the statutory policy for the payment of compensation for miscarriages of justice, all demonstrate the justification for the immunity in the public interest and, indeed, the interests of defendants as a class.

To put it at its lowest, strong arguments exist for making a distinction between the civil and criminal justice systems and the respective need for advocate immunity within them. Because these appeals did not raise this question it was not specifically examined either orally or in written submissions before your Lordships or before any lower court. In my judgment there would be significant consequences of what would be a radically new approach to the administration of criminal justice and (without prejudging the outcome) these potential consequences call for a focused evaluation with the assistance of judgments of lower courts.

One of the consequences of the limited issues raised by these appeals has been that your Lordships have not heard argument upon the definition of what would be the scope of some limited immunity applying to criminal advocacy only. The questions of definition are certainly not of the same order as the problems which would exist for the civil advocacy immunity. It is clear that the same difficulties of delimitation do not exist in the criminal justice system as in the civil justice system. The distinction between civil and criminal proceedings is already well established and used but a view would have to be taken about judicial review proceedings relating to the criminal courts. As regards what comes under the heading of advocacy, there is a clear point of focus being the trial at which the guilt of the defendant is sought to be established. There are existing authorities (eg Somasundaram v. M. Julius Melchior & Co. [1988] 1 WLR 1394 and Acton v. Graham Pearce & Co. [1997] 3 AER 909) which consider the scope of the immunity in the criminal justice system. Unlike in the civil system, the questions of delimitation are not such as to provide a reason for rejecting the immunity in the criminal system. But it is right that any necessary refinement and redefinition, whether by your Lordships or the Court of Appeal, should only result from a properly informed and considered argument directed to those points. The hearing of the present appeals has not been such an occasion.

Finally, I should refer to the suggestion that the Hunter principle (sic) provides an adequate answer to any problem arising from the absence of an immunity in relation to criminal advocacy and therefore renders the immunity unnecessary and disproportionate. As I have explained already the Hunter argument does not address the relevant question or relate to the justification for the immunity in the criminal justice system. It is simply irrelevant and fails to understand the justification for the immunity. The immunity exists and should be maintained because it serves the public interest by making a significant contribution to the working of the criminal justice system and not because it provides protection to lawyers.

The suggestion has been developed into the formulation of a rule that would be a novel rule of public policy: that no civil action in negligence for breach of professional duty can be brought against an advocate in respect of the conviction of his client unless the conviction had first been set aside by an appellate court. That this would be a novel rule cannot be disputed. It would create an anomalous judge-made bar to a negligence action which does not at present exist. The relevant concepts for the law of negligence are causation foreseeability and mitigation. It would need to be assimilated with the statutory law governing the limitation of actions in a way that it is probable that only Parliament should carry out (with or without the assistance of the Law Commission).

Hunter was a wholly exceptional case which had nothing to do with advocate liability. In Hunter there was an abuse of the civil process by using it for the improper purpose of mounting a collateral attack on an adverse criminal decision. But a client suing his lawyer would argue that it was proper for him use the civil process for the purpose of recovering compensation from his lawyer for breach of duty; indeed that is the only way in which he could enforce the civil obligation to pay such compensation under the law of tort. Provided that the action was not wholly without merit and was bona fide brought for the stated purpose and there was no immunity upon which the lawyer was entitled to rely, the lawyer would have difficulty in sustaining an argument that the action was an abuse of process. To challenge in later litigation an earlier non-binding decision between different parties is not in itself abusive, provided there are grounds for doing so. So far as questions of law are concerned, the doctrine of precedent contemplates this. So far as questions of fact are concerned, each court has to try and decide questions of fact upon the evidence adduced before it. Judicial comity and common sense take care of most of these situations in practice but the law does tolerate the possibility of apparently inconsistent decisions. The element of vexation is an aspect of abuse, the use of litigation for an improper purpose, trying to have repeated bites at the same cherry. The objectionable element is not the risk of inconsistency.

The suggested new rule would give a status in the civil law to a criminal conviction which at present it does not have. Under the rule in Hollington v. F. Hewthorn & Co. Ltd. [1943] KB 587, the decision of a criminal court was not evidence of the truth of the facts upon which it was based. This principle applied to any decision of another court or tribunal which did not come within the principles of res judicata as between the parties to the later action. Parliament modified this rule in relation to criminal convictions but it has not gone to the length proposed by the suggested new rule. Under section 11 of the Civil Evidence Act 1968 the person concerned is only to "be taken to have committed that offence [of which he was convicted] unless the contrary is proved". In other words, the conviction is not conclusive: cf section 13 relating to defamation actions. The relevant person (or anyone else with an interest in doing so) is at liberty to prove that he did not commit the crime of which he was convicted. The suggested new rule would have, either expressly or by implication, to contradict this provision. If the existing law is to be changed in this way, it would again be a matter for Parliament and the Law Commission.

The Hunter 'solution' is not a solution and provides no argument for not continuing to recognise the existing advocate immunity in the criminal justice system.

Accordingly, my Lords, I would dismiss the appeals. The claims disclose causes of action against the appellants. The appellants are not entitled to an immunity in respect of the claims made against them in these actions.

I have had the advantage of reading in draft the speeches of my noble and learned friends Lord Steyn and Lord Hoffmann, with which I am in full agreement.

I understand that all your Lordships would abolish the advocate's immunity in civil proceedings, but that some of you would retain it in criminal cases. I readily acknowledge that the case for abolition is stronger in civil litigation, and given my lack of experience of the criminal justice system I have given anxious consideration to the views of those of your Lordships who would retain the immunity in criminal proceedings. I have, however, come to the conclusion that such a partial retention of the immunity should not be supported.

My reasons for this conclusion are twofold. In the first place, I think that to make the existence of the immunity depend on whether the proceedings in question are civil or criminal would be to draw the line in the wrong place. There is a wide variety of cases tried before the magistrates which are for all practical purposes civil in character, and in which the retention of the immunity would be anomalous, but which are commenced by information or summons and which are classified as criminal proceedings. Conversely disciplinary proceedings before professional bodies are classified as civil proceedings but are criminal or quasi-criminal in character. Here the abolition of the immunity would be anomalous but its retention difficult to justify.

In the second place, even if the immunity were retained only in criminal cases tried on indictment, in which the liberty of the subject is at stake (and which is probably the kind of case your Lordships primarily have in mind), it is difficult to believe that the distinction would commend itself to the public. It would mean that a party would have a remedy if the incompetence of his counsel deprived him of compensation for (say) breach of contract or unfair dismissal, but not if it led to his imprisonment for a crime he did not commit and the consequent and uncompensated loss of his job. I think that the public would at best regard such a result as incomprehensible and at worst greet it with derision. The more thoughtful members of the public might well consider that we had got it the wrong way round.

These considerations persuade me that we ought not to retain the immunity in criminal proceedings in the absence of compelling reasons to do so. I acknowledge that there is a particularly high public interest in the efficient administration of criminal justice, that the need to ensure that the accused has a fair trial makes it difficult for the judge to intervene, and that both judge and defence counsel are likely to err on the side of caution. But that is the position today, despite the existence of the immunity. I have some scepticism in accepting the proposition that its removal will make matters significantly worse, and I observe that two of your Lordships with experience of criminal trials do not think that it will.

In my opinion the defending advocate in a criminal trial will retain formidable safeguards against vexatious attack even if he no longer enjoys a formal immunity from suit. His former client will not be allowed to challenge the correctness of the conviction unless and until it is set aside, and a claim which does not challenge the correctness of the conviction, like that in Rondel v. Worsley [1969] 1 AC 191 itself, should normally be struck out as an abuse of the process of the court. The withdrawal of legal aid combined with the new powers of the court to strike out hopeless claims even though they plead a good cause of action should make the great majority of unmeritorious claims still-born. But if the immunity from suit is retained for the moment in criminal cases alone, then sooner or later a case is bound to arise in which the House will be called on to reconsider the question. It will be a bad case involving a clear miscarriage of justice, for otherwise the immunity will not be engaged. It will be a case in which the accused was plainly innocent but was wrongly convicted and served a term of imprisonment as a result of the gross incompetence of his counsel. The conviction will have been quashed on appeal, perhaps accompanied by severe criticism from the court of the conduct of the counsel who was responsible. And by the time the civil claim reaches the House, the public will have become accustomed to read of cases where advocates have been successfully sued for incompetence in the course of civil proceedings even though far less than their client's liberty was at stake. Moreover, the Human Rights Act 1998 will be in force, and the House will have to reconsider the question in terms of article 6 of the European Convention of Human Rights.

I would grasp the nettle now. I believe that the general public would find the proposed distinction indefensible. In the absence of compelling reasons to support it based on more than instinct or intuition, of which I can find none, I find it hard to disagree. I also think that it is difficult to defend a blanket professional immunity in terms of the European Convention on Human Rights. I would dismiss these appeals and declare that the advocate has no immunity from suit in relation to his conduct of proceedings whether civil or criminal.

United Kingdom House of Lords Decisions

Arthur J.S Hall and Co. v. Simons and Barratt v. Ansell and Others v. Scholfield Roberts and Hill [2000] UKHL 38; [2000] 3 All ER 673; [2000] 3 WLR 543; [2000] 2 FLR 545; [2000] 3 All ER 673; [2000] Fam Law 806 (20th July, 2000)

HOUSE OF LORDS

OPINIONS OF THE LORDS OF APPEAL FOR JUDGMENT

IN THE CAUSE