BAILII [Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

England and Wales Court of Appeal (Civil Division) Decisions


You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> K v Secretary of State for the Home Department [2002] EWCA Civ 775 (31 May 2002)
URL: http://www.bailii.org/ew/cases/EWCA/Civ/2002/775.html
Cite as: [2002] EWCA Civ 775

[New search] [View without highlighting] [Printable RTF version] [Help]


    his is a draft of the judgment to be handed down on Friday 31 May 2002 at 9.30 a.m. in Court No 63. It is confidential to Counsel and Solicitors, but the substance may be communicated to clients not more than one hour before the giving of judgment. The official version of the judgment will be available from the shorthand writers once it has been approved by the judge.

    Neutral Citation Number: [2002] EWCA Civ 775
    Case No: B3/2001/1064

    IN THE SUPREME COURT OF JUDICATURE
    COURT OF APPEAL (CIVIL DIVISION)
    ON APPEAL FROM THE HIGH COURT OF JUSTICE
    QUEEN’S BENCH DIVISION

    31 May 2002

    B e f o r e :

    The Honourable Mr Justice Holland)
    LORD JUSTICE SIMON BROWN
    (Vice-President of the Court of Appeal, Civil Division)
    LORD JUSTICE LAWS
    and
    LADY JUSTICE ARDEN

    ____________________

    Between:
    K
    Appellant
    - and -

    The Secretary of State for the Home Department
    Respondent

    ____________________

    Mr Peter Hughes QC and Mr Michael Bailey (instructed by H C L Hanne & Co) for the Appellant
    Mr Philip Sales and Mr Jason Coppel (instructed by The Treasury Solicitors) for the Respondent
    Hearing dates : 29 April 2002

    ____________________

    HTML VERSION OF DRAFT JUDGMENT
    ____________________

    Crown Copyright ©

      Lord Justice Laws:

      INTRODUCTORY

      1. This appeal is brought with permission granted by Judge LJ on 22nd June 2001 against the decision of Holland J by which on 3rd May 2001 it was ordered that the appellant’s claim form and particulars of claim be struck out as disclosing no reasonable cause of action. The appeal requires the court to revisit well-trodden territory. It involves consideration of the circumstances in which the common law will attach liability in negligence for damage which is not inflicted by the impleaded defendant but by a third party.

      THE FACTS

      2. Rashid Musa, a citizen of Kenya born in 1976, entered the United Kingdom on 26th June 1992 and was granted limited leave to remain until 12th October 1994. In July 1994 he was arrested and charged with the buggery of a fifteen year-old girl, and on 10th January 1995 was sentenced to a term of 18 months detention. He was released in May 1995. In July 1995 he was arrested again, for burglary of an occupied dwelling-house at night. For this offence he was sentenced to 18 months imprisonment on 1st December 1995. The court recommended his deportation pursuant to s.6 of the Immigration Act 1971, which it is unnecessary to set out. After he had served his sentence he was further detained under powers provided for in the Immigration Act for the purpose of his being deported in due course. A renewed application for leave to appeal against the recommendation for deportation was refused by the Criminal Division of this court on 19th November 1996. On 14th May 1997 the Secretary of State made a deportation order. But Musa was not then deported. Shortly after the making of the deportation order he applied for a writ of habeas corpus. Before that was determined, on 27th June 1997 the Secretary of State released him from detention. This was done under powers conferred by paragraph 2(3) of Schedule 3 to the Immigration Act 1971, which provides that a person detained pursuant to a recommendation for deportation shall continue to be detained when a deportation order is made against him “unless the Secretary of State directs otherwise”.

      3. On 6th February 1998 Musa raped the appellant. The offence was committed at knifepoint in the premises where she was employed as a cleaner. She was subjected to additional degrading indignities, the least of which was the robbery of some of her possessions. The next day Musa committed an anal rape of a boy on a train.

      4. We know virtually nothing about the Secretary of State’s decision to release this man on 27th June 1997. On 27th August 1999, he appeared before the Crown Court and was sentenced to life imprisonment by His Honour Judge Radford. On that occasion the judge enquired why Musa had been released into the community, and not been earlier deported. He was apparently told that the explanation was “lack of resources.” Mr Hughes QC who appeared for the appellant before us indicated his understanding, that this was a reference to the lapse of time since the recommendation for deportation of 1st December 1995, without deportation having been effected. That is very likely right; but it does not explain the distinct decision to release Musa into the community. In his sentencing remarks Judge Radford stated:

      “It is in my judgment utterly lamentable, with your criminal record and immigration history, that the Home Office did not act with the expedition that would have been hoped for, bearing in mind that a Crown Court judge had recommended your deportation as far back as the 1st December 1995.”

      5. At length the appellant proposed to bring proceedings in negligence against the Secretary of State for damages for all the harm, including psychological harm, she had suffered as a result of the rape. Her case in broad terms was that the Secretary of State released Musa into the community when he knew or should have known that he was a very dangerous individual, liable to commit crimes of sexual violence: he ought to have kept him in detention until he could be deported.

      6. In order to further her potential claim, the appellant applied for pre-action disclosure under CPR Part 31.16. The application was determined by Smith J on 28th July 2000. The judge stated (transcript at 3A) that the Secretary of State had decided not to oppose Musa’s application for a writ of habeas corpus. My impression is that Smith J understood that that decision gave rise to Musa’s release from detention, but it is by no means entirely clear. The judge declined to order pre-action disclosure in light of the single submission advanced before her by Mr Sales for the Secretary of State, viz. that it would not be desirable to order disclosure since the proposed action had no real prospect of success: if it were launched the Secretary of State would apply to strike it out and that application, said Mr Sales, was bound to succeed. After reviewing authority which had been cited to her (essentially the learning later cited to Holland J), Smith J held that the appellant would face “very considerable difficulty in taking her case past the strike out stage”, and “her prospects of success are such that it would not be desirable for me to impose upon the Secretary of State a requirement to disclose documents, which, in my view, he will probably never be under a duty to disclose.”

      7. On 3rd January 2001 the appellant issued these proceedings against the Secretary of State. Paragraphs 18 and 19 of the particulars of claim alleged that at the time of his release from detention (as well as thereafter) Musa “presented an obvious and serious danger to members of the public weaker than himself”, and that the Secretary of State knew or ought to have known as much. Paragraph 20 alleged that the harm suffered by the appellant was foreseeable, and paragraph 21 that it was fair, just and reasonable to impose liability on the Secretary of State. Then paragraph 22 was in these terms:

      “The claimant was proximate to the defendant in that, in particular, but not comprehensively until after disclosure,
      (a) In the premises Mr Musa was one of a finite and identifiable class of persons who presented such a danger,
      (b) Mr Musa was one of a finite and identifiable sub-class of persons within the class referred to in (a) above whom the defendant had to power to detain in a secure place pending deportation and to deport,
      (c) The defendant was entrusted by Parliament with that power to be exercised having regard inter alia to the safety of the public as a whole and in particular those members of the public including the claimant who were weaker than Mr Musa and vulnerable to sexual assaults.”

      Particulars of negligence were given in paragraph 24 as follows:

      “a. [The defendant] [f]ailed properly to consider the dangers to the public of releasing Mr Musa in circumstances where it [sic: viz. the defendant Secretary of State] knew or ought to have known the grave dangers presented by him.
      b. Alternatively to (a), in releasing him from detention, acted in a manner inconsistent with those considerations.
      c. Alternatively to (a), in failing to deport him, failed to act in a manner consistent with those considerations.
      d. Released Mr Musa when it was unreasonable to do so.
      e. Failed to re-detain him.
      f. Failed to deport him properly or at all.
      g. Failed to act under the Deportation Order.
      h. Exposed to and/or failed to protect the public at large, and the claimant in particular, from the serious and obvious dangers presented to them by Mr Musa.”

      There followed in paragraph 25 an allegation that these acts and omissions “were manifestly unreasonable to the point of being irrational”.

      THE JUDGMENT OF HOLLAND J

      8. As had been foreshadowed before Smith J, the Secretary of State applied to strike out the claim, and so the matter came before Holland J in April 2001. The judge accepted Mr Sales’ submission that the pleaded facts could not sustain the allegation of “proximity” between claimant and defendant, which had to be established if the defendant were to be found liable. He said (paragraph 9):

      “Any range of contemplated victims would be remote, extensive and indirect depending upon that third party, Musa, and his choice as to whether to obey the law or whether to commit some (and if so, what) crime involving a person or persons whose identity was wholly speculative… [paragraph 10]… Essentially, the claimant could not be identified as a potential ‘victim’ and consequently was outwith the proximity that is a key element of a duty of care situation.”

      And so Holland J struck out the claim.

      THE APPROACH TO BE TAKEN TO THE APPLICATION TO STRIKE OUT

      9. It is elementary that upon an application to strike out a claim on the footing that the claimant’s pleading discloses no cause of action, the court assumes (for the purposes of the application), that the allegations of fact made in the pleading are true. Mr Hughes QC for the appellant sought before us to make something of the fact, which I have recorded, that virtually nothing is known of the circumstances of the decision to release Musa from detention, not least since Smith J had declined to order pre-action disclosure by the Secretary of State. This was an empty complaint, precisely because the court will assume that the appellant’s allegations of fact are true. In this case that includes even the assertion in paragraph 25 of the particulars of claim that the acts and omissions said to constitute negligence “were manifestly unreasonable to the point of being irrational”. This allegation was necessary to the appellant’s cause of action, because in a case where, as here, a public authority’s exercise of a statutory discretionary power is impugned as having been in breach of a common law duty of care owed by the public authority to the claimant (there being on the statute’s true construction no right to damages for breach of statutory duty), it must be shown that the mode of the power’s exercise was irrational in the Wednesbury sense ([1948] 1 KB 223): see for example Stovin v Wise [1996] AC 923.

      10. I should make this further preliminary observation. The nature of the approach to be taken by the court to an application to strike out a claim such as this is interwoven with the court’s appraisal of the substantive law. Mr Hughes submits that the test (for the existence of a duty of care) of proximity is no more than an aspect of a more general criterion, namely whether it is “fair, just and reasonable” that liability should be imposed. I shall cite the relevant learning shortly. The argument is that if “fair, just and reasonable” is the touchstone, the existence or otherwise of the duty of care can hardly ever be ascertained without the facts being investigated, so that it will almost always be inappropriate to strike out the claim and thus deprive the claimant of a trial of the facts. However, whatever the relation between “proximity” and “fair, just and reasonable”, this argument has to be approached with a degree of caution, again because of the factual assumptions which the court will make in the claimant’s favour. I would accept that the application of an idea (or collection of ideas) as broad as “fair, just and reasonable” to any given set of circumstances may sometimes depend on nuances and details which will not appear in a pleading. But in the ordinary way, a competent pleading, which alleges all the relevant facts in accordance with the claimant’s instructions, either discloses a cause of action or it does not.

      THE LAW – GENERAL

      11. The conditions upon which the common law will impose a duty of care so as to give rise to a cause of action in negligence are by no means precisely fixed. This typifies the common law’s virtue, that claims are decided case by case so as to do justice in a myriad of different circumstances. This however has to be tempered by the influence of another virtue, which is the common law’s capacity to develop firm principle. That is virtuous because it affords a strong measure of legal certainty. The tension between these two virtues is a major characteristic of the common law’s method. It is important that we should not make too great a sacrifice of the principle of legal certainty on the altar of the common law’s flexibility, or vice versa.

      12. The evolution of the law of negligence following Donoghue v Stevenson [1932] AC 562 has seen a number of high water marks along the way, not least in the cases of Dorset Yacht Co Ltd v Home Office [1970] AC 1004 and Caparo Industries v Dickman [1990] 2 AC 605. In Caparo, after reviewing earlier authority, Lord Bridge of Harwich summarised at 616 the conditions required to give rise to liability in negligence (over and above proof of negligence itself) as (1) foreseeability of damage, (2) a relationship of ‘proximity’ or ‘neighbourhood’ between putative claimant and defendant, and (3) “that the situation should be one in which the court considers it fair, just and reasonable that the law should impose a duty of a given scope upon the one party for the benefit of the other”. He added:

      “But… the concepts of proximity and fairness embodied in these additional ingredients are not susceptible of any such precise definition as would be necessary to give them utility as practical tests, but amount in effect to little more than convenient labels to attach to the features of different specific situations which, on a detailed examination of all the circumstances, the law recognises pragmatically as giving rise to a duty of care of a given scope.”

      See also per Lord Oliver of Aylmerton at 633 (a passage I shall cite below at paragraph 23), and per Lord Nicholls of Birkenhead in Stovin v Wise [1996] AC 923, 932B-C (and also Lord Nicholls’ citation at 932E-G of Deane J in Sutherland Shire Council v Heyman (1985) 157 CLR 424, 498). I shall have to return to the loose and interlocking nature of these considerations of proximity and fairness in addressing Mr Hughes’ submission that the latter concept – “fair, just and reasonable” – has effectively swallowed up the narrower idea of proximity, so as all but to exclude a strike out in a case of this kind. But I must first address other considerations of principle.

      13. It is important to have in mind as a starting-point that it has never been the law that liability in negligence may be established merely upon proof of want of care by the defendant, the foreseeability of damage in consequence, and damage actually resulting. This proposition is perfectly elementary, and is repeatedly vouched in the learning. It is a truth implied by the very language used by Lord Atkin in his formulation of the ‘neighbour’ test in Donoghue’s case at 580. And it is, with great respect, a premise of the tri-partite approach given in Caparo (indeed it is expressly stated by Lord Oliver, in a passage I shall cite shortly). Against that background it may be seen that Dorset Yacht and Caparo are instances, respectively, of two situations which have called for particular attention by the common law courts. The first is where primary (as opposed to vicarious) liability in negligence is sought to be imposed upon a party for the consequences of the acts or omissions of another (Dorset Yacht). The second is where liability in negligence is sought to be imposed upon a party for having caused purely economic loss, as opposed to damage to the person or property (Caparo). The present case is manifestly an instance of the former situation, but it is instructive to observe the similarity in the courts’ approach to both.

      14. Where damage to someone’s person or property is, without the intervention of any other cause, occasioned by the act or omission of another, the injured party is all but certainly someone whom the putative defendant should have had in contemplation as “closely and directly affected” – to use Lord Atkin’s phrase in Donoghue – by what he may or may not do. Everyday examples are to hand in running-down cases on the roads, clinical negligence cases, and injuries at work caused by defective equipment. In such instances foreseeability, proximity and fairness will generally be established with little scope for argument. It is instructive to notice what was said by Lord Oliver in Caparo at 632C-G about

      “…that essential relationship between the appellants and the respondent to which, in any discussion of the ingredients of the tort of negligence, there is accorded the description ‘proximity’, for it is now clear from a series of decisions in this House that, at least so far as concerns the law of the United Kingdom, the duty of care in tort depends not solely upon the existence of the essential ingredient of the foreseeability of damage to the plaintiff but upon its coincidence with a further ingredient to which has been attached the label ‘proximity’ and which was described by Lord Atkin in the course of his speech in Donoghue v. Stevenson [1932] AC 562, 581 as:
      ‘such close and direct relations that the act complained of directly affects a person whom the person alleged to be bound to take care would know would be directly affected by his careless act.’
      It must be remembered, however, that Lord Atkin was using these words in the context of loss caused by physical damage where the existence of the nexus between the careless defendant and the injured plaintiff can rarely give rise to any difficulty. To adopt the words of Bingham L.J. in the instant case [1989] Q.B. 653, 686:
      ‘It is enough that the plaintiff chances to be (out of the whole world) the person with whom the defendant collided or who purchased the offending ginger beer.’”
      But the presence of this ‘nexus’ is by no means self-evident in cases of pure economic loss, or of harm (albeit physical harm) caused by a third party for which the defendant is sought to be held responsible. This is why these two classes of cases are problematic.

      15. In looking at these two problematic instances it seems to me that many, perhaps most, cases of pure economic loss arising for consideration in the law of negligence in fact constitute a sub-set of the class of cases of harm caused by a third party (someone other than the party sought to be held liable). Where A directly inflicts pure economic harm on B, without the agency of any third party, in the nature of things B’s claim is likely to lie in fraud, conversion, breach of trust, or breach of contract: such instances will usually involve deliberate and not merely careless action on A’s part. But where B suffers pure economic harm by reason of A’s negligence, the strong likelihood is that there will be another agency at work in the causation of the loss, such as factors affecting the valuation of property, of which it is said that A should have warned B. If you try and suppose an instance in which A’s negligence causes pure economic loss to B with no causative intervention by another agency, the examples which tend to come to mind are ones where there is in any event a ‘special’ or particular relationship between A and B: A is B’s accountant, solicitor or banker, and somehow manages to lose B’s money with no one else being involved.

      16. At all events I think we may say that the case where primary (as opposed to vicarious) liability in negligence is sought to be imposed upon a party for the consequences of a third cause or agency – whether the acts or omissions of another person distinctly, or something less specific like the movements of markets – is the paradigm instance of the class of cases where issues of proximity and fairness are acute, and not merely given, in the law of negligence. More concretely, in the common law’s present maturity we have, very broadly, arrived at a position in which liability for damage carelessly occasioned to another’s person or property is, as a matter not of legal principle but of pragmatic reality, the rule not the exception where there is no third agency which constitutes the immediate cause of the damage; but where there is such a third agency, liability is the exception not the rule. We need to look at the reasons for this with some little care.

      17. In the latter class of case, where a third agency is the immediate cause of the damage, the act or omission of the defendant is by definition at one remove (at least one, maybe more) from the outturn events of which the claimant comes to complain. A duty of care imposed on the defendant in those circumstances would be very open-ended. It might be said to expose him, in the well-known words of Cardozo CJ in Ultramares Corporation v Touche (1931) 174 NE 441, 444 (cited by Lord Bridge of Harwich in Caparo at 621D and 622F) to “liability in an indeterminate amount for an indeterminate time to an indeterminate class”. Ultramares, like Caparo, was an economic loss case; and one dimension in the reasoning in Caparo was that it would not be right, by means of the law of negligence, to confer on many members of the public the benefits of recourse against an expert adviser which in the ordinary way a man might enjoy only if he had paid for it. Such reasoning is no doubt particularly apt in relation to economic loss cases. But in my judgment Cardozo CJ’s dictum also possesses force for situations such as that arising in the present case; not least if, as I would have it (see paragraphs 15 and 16), there is something of an assimilation of the economic loss examples to the third agency cases (as I may call them).

      18. The question ‘Am I my brother’s keeper?’ in our law generally receives a restricted reply, as Lord Diplock said in Dorset Yacht at 1061D, citing at 1063E-F the observations of Dixon J in the High Court of Australia in Smith v Leurs 70 CLR 256, 262:

      “The general rule is that one man is under no duty of controlling another man to prevent his doing damage to a third. There are, however, special relations which are a source of a duty of this nature.”

      These observations are contained within a longer passage cited in the Privy Council in Yuen Kun Yeu [1988] AC 175, 193, but with great respect Dixon J’s reasoning there does not appear to provide a principle by reference to which this restricted position may be justified. In many instances, however, it seems to be accounted for by what is seen as a want of causation between the defendant’s carelessness and harm to the claimant. Thus the courts have held that, given the unpredictability of human conduct, the act of the party – uncontrolled by the defendant – who actually does the damage breaks the chain of causation between the defendant’s negligence and the claimant’s damage: it is, in the once popular Latinism, a novus actus interveniens. This approach is exemplified in the speech of Lord Reid in Dorset Yacht at 1027D-1028D. But it is clear that want of causation is by no means the whole explanation for the court’s restricted approach. Nor plainly could it be. If it were, third agency cases would not merely arise as exceptional instances of the application of the law of negligence. They would not arise at all; they would always be ruled out for want of causation. But if that is so, what is the basis – if any, aside from the dictum of Cardozo CJ – for restricting liability in third agency cases where on the facts the court accepts that the claimant has established causation?

      19. In addressing that question, I would say first that I think it important that such cases (other than instances of pure economic loss, so far as those may be assimilated to third agency cases) very often involve on their facts the acts or omissions of public authorities. In Dorset Yacht the defendant was of course the Home Office. In another landmark case, Hill v Chief Constable for West Yorkshire [1989] 1 AC 53, a police force was sought to be made liable for damages in negligence arising out of the death of one of the victims of the so-called Yorkshire Ripper. Indeed there is a raft of police cases, including the recent decision of this court in Cowan v Chief Constable for Avon & Somerset Constabulary [2001] EWCA Civ 1699 (14th November 2001), in which Keene LJ collects much of the learning. Other public authority defendants to third agency claims have included local authorities, as in X v Bedfordshire CC [1995] 2 AC 633, where (to oversimplify) the allegation was that abused or otherwise damaged children might or would have been spared their suffering if the council had acted on information it had and intervened appropriately.

      20. It is no coincidence that third agency cases so often involve public authority defendants. The acts or omissions of public authorities are especially likely to create situations in which third party agencies – whether they are the acts of persons or other causes – will affect members of the public, sometimes to their hurt, who may not be in the authority’s direct contemplation. This is simply a consequence of the very obvious fact that public authorities have public responsibilities in pursuit of which they will adopt policies, and exercise discretions, and decide on courses of action, across a range of human experience generally far bigger than the scope of the private individual’s influence on other people’s lives.

      21. Where the putative defendant is a public body, the existence and nature of the duties it owes to the public are frequently critical to the judgment that falls to be made, whether the public body should be held liable for damage caused by a third agency. The following passage from the speech of Lord Browne-Wilkinson in Barrett v Enfield London Borough Council [1999] 3 WLR 79, 85 (cited by Keene LJ at paragraph 24 in Cowan), a case to which I must refer further in addressing Mr Hughes’ submissions, represents what I think with respect has become a familiar approach:

      “In a wide range of cases public policy has led to the decision that the imposition of liability would not be fair and reasonable in the circumstances, e.g. some activities of financial regulators, building inspectors, ship surveyors, social workers dealing with sex abuse cases. In all these cases and many others the view has been taken that the proper performance of the defendant's primary functions for the benefit of society as a whole will be inhibited if they are required to look over their shoulder to avoid liability in negligence. In English law the decision as to whether it is fair, just and reasonable to impose a liability in negligence on a particular class of would-be defendants depends on weighing in the balance the total detriment to the public interest in all cases from holding such class liable in negligence as against the total loss to all would-be plaintiffs if they are not to have a cause of action in respect of the loss they have individually suffered.”
      Keene LJ also cites (Cowan, paragraph 31) this passage from the judgment of Beldam LJ in Ancell v McDermott [1993] 4 AER 355, 365e-f:
      “It is exceptional to find in the law a duty to control another’s actions to prevent harm to strangers and where they are found they arise from special relationships. When it is contended that such special relationship arises out of duties carried out in the performance of a public office, the court must have regard to the purpose and scope of the public duties, whether they are intended to benefit a particular section of the public, eg investors or depositors, and whether such persons could reasonably place reliance on the fulfilment of the duties.”

      22. The authorities dealing with the police as defendants to negligence actions, from Hill onwards, and of which I think Cowan is the most recent instance, demonstrate with particular clarity the force of the public interest factors which tend to inhibit the imposition of liability in third agency cases. In Costello v Chief Constable of Northumbria Police [1999] 1 AER 550 May LJ said this at 563f-g:

      “For public policy reasons, the police are under no general duty of care to members of the public for their activities for their activities in the investigation and suppression of crime (Hill’s case). But this is not an absolute blanket immunity and circumstances may exceptionally arise when the police assume a responsibility, giving rise to a duty of care to a particular member of the public (Hill’s case and Swinney’s case [sc. [1997] QB 465]). The public policy considerations which prevailed in Hill’s case may not always be the only relevant public policy considerations (Swinney’s case).
      Neither the police nor other public rescue services are under any general obligation, giving rise to a duty of care, to respond to emergency calls (Alexandrou’s case [sc. [1993] 4 AER 328]), nor, if they do respond, are they to be held liable for want of care in any attempt to prevent crime or effect a rescue. But if their own positive negligent intervention directly causes injury which would not otherwise have occurred or if it exacerbates injury or damage, there may be liability (the Capital & Counties Plc case [1997] QB 1004).”
      Keene LJ specifically adopted this reasoning in holding that on the facts in Cowan (which I need not set out) the police owed no duty of care. He stated (paragraph 41): “It is only if a particular responsibility towards an individual arises, establishing a sufficiently close relationship, that a duty of care may be owed to that individual”. And at paragraph 44 he stated:
      “In so far as matters of public policy come into consideration, particularly under the concept of what is just and reasonable, there is the well-established public interest in not fettering or influencing the police in operational matters by the ‘spectre of litigation’: see Ward LJ in Swinney, pp. 486H-487A. Their duties are difficult enough without that added complication which would arise from imposing a duty of care towards individual citizens to prevent crime.”

      23. Now I may draw the threads together. (1) The learning shows that in third agency cases the law in principle looks for a ‘special relationship’ between claimant and defendant before it will find a duty of care. (2) The reason for this rests, I think, on two propositions. The first is that it has never been the law that liability in negligence may be established merely upon proof of want of care by the defendant, the foreseeability of damage in consequence, and damage actually resulting; from this it follows that there must always be some further factor in the case to provide a nexus between claimant and defendant so as to found a duty of care. The second is that where there is no third agency, and the damage which the claimant suffers is directly occasioned by the defendant’s act or omission, such a nexus will generally be taken to be present without more: the facts must themselves have drawn the claimant and defendant together. Thus the driver must necessarily have in mind his fellow road-users as liable to be directly affected by his act or omission; so also the employer and his employees, the doctor and his patients. But there is no like necessity in a third agency case. Usually in such cases there will be no particular nexus; if there is to be a duty of care, the nexus will have to be found on the particular facts. Thus where it is found, it may reasonably be termed a ‘special relationship’. (3) These considerations explain the practical antithesis I sought to state in paragraph 16: in the common law’s present maturity we have, very broadly, arrived at a position in which liability for damage carelessly occasioned to another’s person or property is the rule not the exception where there is no third agency which is the immediate cause of the damage; but where there is such a third agency, liability is the exception not the rule. (4) A public authority owing duties to the public at large or a section of it is particularly likely to find itself exposed to situations in which it may be an instrument of damage through a third agency; and there may be important policy considerations which will militate against the finding of such a nexus or relationship as would generate a duty of care. (5) These observations sit with Cardozo CJ’s description of the effect of imposing a general duty of care for the negligent infliction of economic loss as “liability in an indeterminate amount for an indeterminate time to an indeterminate class”: many, perhaps most, instances of pure economic loss are in truth third agency cases, and in any event Cardozo CJ’s dictum is apt to show the wisdom, or some of the wisdom, inherent in the rule that carelessness, foreseeability, and damage are without more not enough to found a duty of care. (6) All of these points live under the aegis of the Caparo tri-partite formulation, and not least the loose and interlocking nature of the three elements of foreseeability, proximity, and fairness as explained by Lord Oliver at p. 633:

      “it is difficult to resist a conclusion that what have been treated as three separate requirements are, at least in most cases, in fact merely facets of the same thing, for in some cases the degree of foreseeability is such that it is from that alone that the requisite proximity can be deduced, whilst in others the absence of that essential relationship can most rationally be attributed simply to the court's view that it would not be fair and reasonable to hold the defendant responsible. ‘Proximity’ is, no doubt, a convenient expression so long as it is realised that it is no more than a label which embraces not a definable concept but merely a description of circumstances from which, pragmatically, the courts conclude that a duty of care exists.”

      THE PARTIES’ SUBMISSIONS

      24. Mr Hughes acknowledges, as he must, that the imposition of liability for negligence in a third agency case is exceptional. Upon analysis his argument (aside from what he has to say about the European Convention on Human Rights (‘ECHR’) to which I will refer separately) amounts in my judgment to two propositions as follows. (1) Given the decision of their Lordships’ House in Barrett, a negligence claim such as that made here ought not to be struck out upon a preliminary application, but should go to trial for all the facts to be found before a determination is made as to whether or not in the circumstances a duty of care exists. (2) Where the risk which the third agency – here, of course, the released man Musa – poses is to the knowledge of the putative defendant especially grave, that circumstance can of itself supply a nexus, or relationship, between claimant and defendant sufficient to create a duty of care.

      25. In Barrett the allegation was that the defendant local authority had failed to safeguard the plaintiff’s welfare during the years when as a child he had been in their care. The House of Lords declined to strike out the claim. Their Lordships’ speeches cover a great deal of legal ground, not least the circumstances in which the law will or will not impose a common law duty of care upon a public authority exercising functions under statutory provisions which on their proper construction do not impose duties such as will give rise, if they are violated, to a cause of action for damages in private law for breach of statutory duty; and also the impact of the Strasbourg court’s decision in Osman (2000) 29 EHRR 245. For present purposes it is with respect sufficient to cite this short passage from the speech of Lord Browne-Wilkinson at p. 83:

      “In my speech in the Bedfordshire case [1995] 2 AC 633, 740-741 with which the other members of the House agreed, I pointed out that unless it was possible to give a certain answer to the question whether the plaintiff's claim would succeed, the case was inappropriate for striking out. I further said that in an area of the law which was uncertain and developing (such as the circumstances in which a person can be held liable in negligence for the exercise of a statutory duty or power) it is not normally appropriate to strike out. In my judgment it is of great importance that such development should be on the basis of actual facts found at trial not on hypothetical facts assumed (possibly wrongly) to be true for the purpose of the strike out.”

      As it seems to me Barrett may be regarded as a case where, in terms of the Caparo tri-partite approach (however interdependent the three heads or criteria may be in principle), there was on the facts no argument as to foreseeability or proximity (the child as I have said was in the defendant’s care), and everything turned on what might be shown to be ‘fair and reasonable’. In such a case one may, with great deference, readily understand the court’s reluctance to decide the matter on a strike-out application. Much may depend on the flavour of the case, which will not necessarily emerge from the pleadings.

      26. But the case before us is quite different. The whole burden of Mr Sales’ submission for the respondent is that here there was no proximity: no nexus or special relationship between claimant and defendant. He submits, correctly in my judgment, that nothing in Barrett, nor indeed in the dicta in Caparo and elsewhere as to the interlocking nature of the three elements required to establish a duty of care, has undermined the need to find such a nexus. That is especially so in a third agency case where the nexus is not given, but has to be found.

      27. In relation to both of Mr Hughes’ propositions Mr Sales relies on the decision of this court in Palmer [2000] PNLR 87. In that case a man called Armstrong had been in the care of the defendant health authority. He was last admitted to hospital as an in-patient in June 1993. It was to be alleged that before that he had attempted suicide on five occasions. While he was in hospital in June 1993 he said he had sexual feelings towards children and that a child would be murdered after his discharge. He was discharged on 21st June 1993 but remained an out-patient. While in the care of the defendants he was variously diagnosed as suffering from personality disorder or psychopathic personality. On 30th June 1994 he abducted, sexually assaulted and murdered the claimant’s four year-old daughter.

      28. Stuart-Smith LJ in Palmer held that there was nothing in the reasoning in Barrett which required that the case before him should go to trial: see paragraphs 11-21 of his judgment. He placed emphasis (paragraph 13) on the fact, to which I have adverted, that there was in Barrett no issue as to proximity. Thorpe LJ agreed with him without offering separate reasoning of his own. Pill LJ also agreed; with respect I read his reasoned judgment as expressing some disquiet about the second issue in the case, namely whether a sufficient proximity or nexus was established on the facts. That issue invokes a parallel with Mr Hughes’ second submission in this case – that where the risk which the third agency poses is to the knowledge of the putative defendant especially grave, that circumstance can of itself supply a nexus between claimant and defendant sufficient to create a duty of care. I will come to that directly. On the first issue – what may be called the Barrett point – I can see no basis for adopting any different approach from that taken in Palmer, which of course binds us if it cannot be distinguished. Like Stuart-Smith LJ in Palmer, I consider that Barrett addresses a state of affairs wholly different from that presently facing the court. That suffices to refute Mr Hughes’ first proposition.

      29. In my judgment his second proposition – that where the risk which the third agency poses is to the knowledge of the putative defendant especially grave, that circumstance can of itself supply a nexus between claimant and defendant sufficient to create a duty of care – is contrary to principle. A defendant does not become the world’s insurer against the grave danger (where the danger is general) posed by a third agency, which he might control but does not, by virtue only of the fact that he appreciates that the danger exists. In Palmer Pill LJ said at 108D-E:

      “I see force in the submission that the question whether the identity of a victim is known ought not to determine whether the proximity test is passed. It is forcefully argued that the difference between the threat ‘I will kill X’ and the threat ‘I will kill the first bald-headed man I meet’ ought not to determine whether a duty is placed upon a defendant…”

      But this acknowledges, as of course it must, that the proximity test has to be passed; whether it is so or not depends on the facts pleaded and (if the case goes to trial) proved. I have already set out (paragraph 7) the particulars of proximity pleaded by the appellant in paragraph 22 of the particulars of claim. On their face, they allege no nexus between claimant and defendant: the nexus they allege is between the defendant and the third agency, Mr Musa. That is not enough.

      30. Moreover Mr Hughes’ argument, if right, would I think have most unorthodox consequences. If a public authority were to be held liable in negligence on facts like those pleaded here, with no true nexus shown between claimant and defendant (its absence being supposedly filled by the perceived gravity of the risk involved), the negligence or fault in the case would have to be supplied by proof that the authority had acted unreasonably in the public law sense: see paragraph 9 above. But if that were sufficient, without the added element of proximity, the result as it seems to me would be that the court would have in effect created a category of administrative tort sounding in damages. Our law, however, knows no such tort outside the confines of misfeasance in public office, or those circumstances in which breach by the State of an obligation arising under the law of the European Union may give rise to a right of compensation.

      31. Mr Hughes has a further argument based on the ECHR. He submits (skeleton argument, paragraph 7 page 13) that “the common law should be so interpreted and refined as to promote and safeguard the claimant’s rights under the Convention and ensure compliance by the UK with its Convention obligations.” He submits that the rape of his client amounted to inhuman or degrading treatment within ECHR Art. 3, and that accordingly the common law should accord her a remedy no less effective than what is required by Art. 13. Implicit in this is the claim that such a remedy can only be provided by according a duty of care whose breach would be justiciable by the courts, although as I understand it any direct reliance on the right of access to an independent and impartial court guaranteed by ECHR Art. 6(1) is disavowed given that the relevant events in this case took place before incorporation of the Convention into English law by the Human Rights Act 1998 on 2nd October 2000.

      32. I mean no disrespect to Mr Hughes’ argument if I deal with this aspect shortly. It is plain from Bromiley v UK (33747/96), in which the Strasbourg court declared the application inadmissible on 23rd November 1999, that an order striking out a negligence claim for want of proximity may be consistent with Art. 6. The facts in Bromiley were similar to those here. The claimant’s daughter had been murdered by a psychopath who had been released from prison on home leave. The court in Bromiley recalled what had been said in Osman v UK (2000) 29 EHRR 245 at paragraph 116:

      “In the opinion of the Court where there is an allegation that the authorities have violated their positive obligation to protect the right to life in the context of their… duty to prevent and suppress offences against the person…, it must be established to its satisfaction that the authorities knew or ought to have known at the time of a real and immediate risk to the life of an identified individual or individuals from the criminal acts of a third party and that they failed to take measures within the scope of their powers which, judged reasonably, might have been expected to avoid that risk.”

      CONCLUSION

      33. That passage seems to me to demonstrate a conformity between the law of the Convention and the domestic law of England upon the scope of liability in third agency cases where the claim relates to serious injury or death. There is nothing in the order striking out this claim which is repugnant to the Strasbourg jurisprudence, such as might require the common lawyers to think again. So far as the common law itself is concerned, given the approach to third agency cases which I have ventured to canvass in light of the authorities, no sufficient nexus between claimant and defendant is pleaded, and the claim was properly struck out.

      34. I would dismiss the appeal.

      Lady Justice Arden:

      35. I agree that this appeal must be dismissed for the reasons which Laws LJ has given in paragraphs 24 to 30 of his judgment. The need to show proximity (meaning closeness of a relevant kind between the appellant and the respondent) is a separate requirement of the duty of care, which cannot be established simply by showing the knowledge on the part of the respondent of a glaring danger to the public as a result of Mr Musa’s release. Accordingly, this case cannot in my judgment be distinguished from the decision of this Court in Palmer v Tees Health Authority [2000] PNLR 87.

      36. In the earlier part of his judgment Laws LJ has sought to provide a principled basis to the situations involving the agency of a third party or event in which the law does not impose a duty of care. These are observations requiring the most careful consideration and analysis. However, as they have not formed the basis of the argument in this case, I prefer not to express any view on them at this stage other than to record four reservations.

      37. First, I would leave for future consideration the question whether cases of pure economic loss can usefully be regarded as a sub-set of the cases where damage is directly inflicted by a third party. This seems to me to be potentially a topic requiring a detailed study of the case law for this purpose.

      38 . My second reservation requires more explanation.

      39. In Caparo plc v Dickman [1990] 2 AC 605, the House of Lords identified the three ingredients of a duty of care situation, namely foreseeability, proximity and that it is fair, just and reasonable to impose a duty of care, but held that those requirements was not capable of precise definition (see per Lord Bridge at 618A-C, per Lord Roskill at 628C-E and per Lord Oliver at 633A-D). Thus the House did not in that case adopt any general basis of liability in negligence. In this field, experience has shown that no single abstract formulation can determine the existence of a duty of care.

      40. Accordingly, the existence of a duty of care in any new situation not covered by authority falls to be established on a case by case basis. That is not to say that the law of negligence is not capable of further development. As Lord Macmillan said in Donoghue v Stevenson [1932] AC 562, at 619, “…the conception of legal responsibility may develop in adaptation to altering social conditions and standards….The categories of negligence are never closed”.

      41. I can now state my second reservation. In the light of the various considerations which I have set out above, it seems to me that any antithesis of the kind which Laws LJ draws in paragraph 16 of his judgment must, at least in this Court, remain, as Laws LJ indicates, at the level of empirical observation rather than form the basis of the development of a new legal precept.

      42. My third reservation is that, for my own part, I have not given the consideration which would need to be given to the extent to which the antithesis drawn by Laws LJ can be said to be valid under the law as it presently stands. I have in mind, for example, such matters as liability for psychiatric damage.

      43. My fourth reservation is that the antithesis drawn by Laws LJ merges the process of recognising a duty of care with that of limiting the consequences of a breach, whereas it may be more useful to analyse the separate considerations affecting each process.

      44. I return to the further issues arising in the instant appeal. Mr Hughes QC has made a number of submissions based on the European Convention of Human Rights. The appellant cannot rely on the Convention directly since the Human Rights Act 1998 was not in force at the relevant time. However, he relies on the Convention indirectly, arguing that the common law should recognise a duty of care in any case where a public authority would be in breach of a positive obligation under article 3 of the Convention to take measures so that a citizen is not exposed to risk of inhuman or degrading treatment.

      45. I accept, of course, that the common law is very porous and susceptible to development in accordance with many influences, including Convention values. I assume for the purpose of this submission that a finding of a breach of article 3 would follow if the pleaded allegations were proved. However, as a matter of Strasbourg jurisprudence, it is established that article 6 does not guarantee the content of civil rights under national law (see for example Z v United Kingdom [2002] 34 EHRR 97). Moreover, the European Court of Human Rights has specifically recognised in a number of cases that a holding that under English law there is no duty of care because proximity is lacking and therefore no remedy in negligence is not a violation of article 6: see for example TP and KM v UK [2002]34 EHRR 42. In addition article 13 of the Convention confers the right to an effective remedy if a Convention right has been violated, and in any case occurring after the commencement of the Human Rights Act 1998, section 7 of that Act is available. In those circumstances, I do not consider that there is a need for the common law to develop in the very significant way that would be necessary to accommodate this case (assuming for this purpose that such a course is open to the court).

      46. Since the material events occurred before the commencement of the Human Rights Act 1998, we need not decide the question whether the matters alleged amount to a breach of a positive obligation imposed on the state by virtue of article 3 of the Convention. This is a question on which the recent admissibility decision of the European Court of Human Rights (Third Section) in Ivison v United Kingdom (16 April 2002), sent to us after the hearing, may have a bearing. Consideration of that decision (which relates to articles 2,6,8 and 13), however, must await a future case.

      Lord Justice Simon Brown:

      47. I too agree that this appeal must be dismissed for the reasons given by Laws LJ in paragraphs 24 – 32 of his judgment. I prefer to express no view upon the other matters which he discusses, interesting and important though I recognise them to be.


BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII
URL: http://www.bailii.org/ew/cases/EWCA/Civ/2002/775.html