(back to preceding text)

106.  The principles laid down in Hill v Chief Constable of West Yorkshire [1989] AC 53 were reviewed and affirmed by the House as recently as 2005 in Brooks v Commissioner of Police of the Metropolis [2005] UKHL 24, [2005] 1 WLR 1495 and I do not think that it is necessary to depart from them. One must acknowledge at once that the price of the certainty of the rule and the freedom from liability afforded to police officers is that some citizens who have very good reason to complain of the police handling of matters affecting them will not have a remedy in negligence (cf Lord Steyn’s observation in Brooks at para 31). One has to face this and decide whether in the wider public interest the law should allow that. I am of opinion, in agreement with Lord Hope (para 78) that in the interests of the wider community it is necessary that it should do so for the better performance of police work.

107.  One cannot escape feelings of some concern when one applies the broad general rule to the facts of Mr Smith’s case, which tests the principle severely. Mr Smith, on the assumed facts of the case, reported the lurid death threats made by Jeffrey on several occasions to the police, but for some reason for which no explanation has been put forward they declined to look at the messages containing the threats, make an entry in their notebooks, take a statement from Mr Smith or complete a crime form. Instead they commenced a rather slow procedure for tracing Jeffrey’s telephone calls, which inexplicably they continued to pursue even after Mr Smith had given them Jeffrey’s home address in the forms which he completed at their request and then given them the telephone numbers from which Jeffrey sent threatening text messages. He received similarly dismissive treatment from an inspector in Brighton, who again declined to look at the threatening messages or make a note of the meeting. One must recognise that police officers may quite properly be slow to engage themselves too closely in such domestic type matters, where they may suspect from experience the existence of a degree of hysteria or exaggeration on the part of either or both persons involved. Making all allowances for that factor, however, the inertia of the police was such, on the assumed facts, that if there is a duty of care in law it would be entirely wrong to strike out the claim.

108.  I am satisfied nevertheless that the reasons underlying the acceptance of the general rule that a duty of care is not imposed upon police officers in cases such as the present remain valid. Those reasons are summarised in para 76 of Lord Hope’s opinion, with which I agree, and I need not set them out again. The factor of paramount importance is to give the police sufficient freedom to exercise their judgment in pursuit of their objects in work in the public interest, without being trammelled by the need to devote excessive time and attention to complaints or being constantly under the shadow of threatened litigation. Over-reaction to complaints, resulting from defensive policing, is to be avoided just as much as failure to react with sufficient speed and effectiveness. That said, one must also express the hope that police officers will make good use of this freedom, with wisdom and discretion in judging the risks, investigating complaints and taking appropriate action to minimise or remove the risk of threats being carried out.

109.  It remains to be considered whether there are any exceptions to the generality of the rule. Lord Hope has referred in paragraph 79 to the existence of a duty of care in respect of operational matters. As he says, imposing liability in such cases does not compromise the public interest in the investigation and suppression of crime. I also agree with his view (para 78) that the test propounded by Lord Bingham, dependent on the production of apparently credible evidence of a specific and imminent threat to the life or physical safety of the complainant, would be difficult to operate and would tend to lead to a defensive approach to the carrying out of police work. I would not dissent from the view expressed by Lord Nicholls of Birkenhead in Brooks at para 6 that there might be exceptional cases where liability must be imposed. I would have reservations about agreeing with Lord Steyn’s adumbration in para 34 of Brooks of a category of cases of “outrageous negligence", for I entertain some doubt whether opprobrious epithets provide a satisfactory and workable definition of a legal concept. I should accordingly prefer to leave the ambit of such exceptions undefined at present.

110.  I therefore would allow both appeals, make the order proposed by Lord Bingham in Van Colle’s case and restore the order made by the judge in Smith’s case.

LORD BROWN OF EATON-UNDER-HEYWOOD

My Lords,

111.  Threats to witnesses are a real problem for the criminal justice system: in the Van Colle appeal the court below ([2007] 1 WLR 1821, para 11) noted Home Office/Association of Chief Police Officers statistics showing that 10% of crimes lead to incidents of intimidation. So too, more generally, high levels of violence within our society—domestic violence, gang violence, violence amongst young people and so forth—involve the police frequently learning in advance of grave threats to people’s safety. Sometimes the police are alerted to the risk by those threatened. But often not, and in these latter cases it is the police who warn those they believe to be at risk. A recent article in The Times (9 June 2008) stated that last year the police warned over a thousand people “that they were the subject of a murder plot”—"at serious risk of being killed by individuals with the resources to arrange their death”—these being (inappropriately) described as “so-called Osman warnings".

112.  These two appeals concern occasions when it was the eventual victims themselves who alerted the police to threats received and they raise in different contexts the possibility of civil actions against the police when, as here, the threats come to be carried out and the victim suffers death or serious injury. Van Colle is a Human Rights Act claim following the murder of the respondents’ son and it asserts that the police were in breach of article 2 of the Convention in failing to do all that could reasonably have been expected of them to avoid the risk to his life to which he had alerted them. Mr Smith claims in negligence at common law for the very serious injuries which he suffered and he must therefore establish that the police owed him a duty of care.

113.  I have had the advantage of reading in draft the opinions of my noble and learned friends Lord Bingham of Cornhill and Lord Hope of Craighead and, like others, gratefully adopt Lord Bingham’s exposition of the facts in both cases.

Van Colle

114.  I am in full agreement with everything said by Lord Bingham both as to the facts and the law regarding this case and there is little I wish to add.

115.  The test set by the European Court of Human Rights in Osman v United Kingdom (1998) 29 EHRR 245 and repeatedly since applied for establishing a violation of the positive obligation arising under article 2 to protect someone from a real and immediate risk to his life is clearly a stringent one which will not easily be satisfied. This is hardly surprising given, as the Osman judgment itself recognises (at para 116), “the difficulties involved in policing modern societies, the unpredictability of human conduct and the operational choices which must be made in terms of priorities and resources". It is, indeed, some indication of the stringency of the test that even on the comparatively extreme facts of Osman itself—rehearsed by Lord Bingham at para 56—the Strasbourg court found it not to be satisfied.

116.  No less clear is it that the Osman test is a constant, to be applied whatever the particular circumstances of the case. So much was decided by the House in In re Officer L [2007] 1 WLR 2135—see Lord Carswell’s leading opinion there at para 20. In so far as the Court of Appeal’s decision in R (A) v Lord Saville of Newdigate [2002] 1 WLR 1249, has been understood to the contrary (as by the court below), it should not be followed. The decision in that case was amply justified without resort to article 2, merely by the application of ordinary public law principles of fairness, reasonableness and proportionality.

117.  I agree entirely with Lord Bingham (at para 39) that the fact that Giles Van Colle was a witness was plainly relevant. But it was relevant only to the extent that realistically it increased the likelihood that Brougham would actually carry out his threat to kill (or seriously injure) him. Nothing in the facts here is comparable, for example, to the increased risk to life experienced by political journalists and others considered by the Strasbourg Court in Kiliç v Turkey (2000) 33 EHRR 1357, paras 65-68.

118.  I cannot hope to improve upon Lord Bingham’s factual analysis of the present case and respectfully concur with his conclusion that, tragic though Giles’ death was, this claim does not satisfy the Osman test—by a wider margin, indeed than the Osman claim which itself failed. I too would allow the Chief Constable’s appeal and enter final judgment in his favour.

Smith

119.  In what circumstances ought the police to be subject to civil liability at common law for injuries deliberately inflicted by third parties i.e. for crimes of violence? When, in short, should they in this type of case be held to owe a duty of care to the victim?

120.  That there are such cases is not in doubt. Swinney v Chief Constable of Northumbria Police Force [1997] QB 464 provides one example, the facts there suggesting that the police had assumed responsibility for the complainant informer’s safety (although his claim in the event failed at trial). Another example (again on the basis of assumption of responsibility) is Costello v Chief Constable of the Northumbria Police [1999] 1 All ER 550 where a police inspector was found liable to a woman police constable for injuries inflicted on her by a woman prisoner in a police station cell. The factual findings were stark. The inspector was present specifically to help the plaintiff if she needed it and himself acknowledged that he had a duty to do so, yet he stood by and did nothing when in fact she was attacked. By the same token it may well be that the family of the victim whose death underlay the criminal conviction in R v Dytham [1979] QB 722 (referred to by Lord Bingham at para 53) would nowadays have a common law claim against the police. The constable in that case was, after all, convicted not merely of non-feasance but of deliberate failure and wilful neglect.

121.  For my part I would be disposed to accept that a common law duty of care would be found to exist also in circumstances such as arose in Edwards v United Kingdom (2002) 35 EHRR 487 (in fact a successful article 2 application in Strasbourg) where the applicants’ son was killed by his cell-mate whilst on remand. Had such a death occurred in a police cell, it would be difficult to argue that the police had not assumed a clear responsibility for the safety of those in their custody.

122.  All these cases, however, are of a different character from that now before your Lordships and the question arising here is whether a duty of care should similarly be found to exist when the police, without having assumed any particular responsibility towards the eventual victim, are engaged rather in discharging their more general duty of combating and investigating crime.

123.  Generally speaking, it is accepted that in the discharge of this function the police owe no legal duty of care to individuals affected. Lord Steyn in Brooks v Commissioner of Police of the Metropolis [2005] 1 WLR 1495 more than once referred to this as the “core principle in Hill’s case” (Hill v Chief Constable of West Yorkshire [1989] AC 53). Can this “core principle", however, accommodate Lord Bingham’s “liability principle": the proposition that “if a member of the public (A) furnishes a police officer (B) with apparently credible evidence that a third party whose identity and whereabouts are known presents a specific and imminent threat to his life or physical safety, B owes A a duty to take reasonable steps to assess such threat and, if appropriate, take reasonable steps to prevent it being executed"?

124.  For my part I would acknowledge that the facts of the present case are vastly different from those either in Hill or in Brooks and that it is easier to contemplate civil liability here than in either of those cases. The particular reasons for rejecting the contended for duty of care in each of them were compelling: in Hill because it was surely unthinkable that the conduct of the entire Yorkshire Ripper investigation should be subject to minute examination; in Brooks because the main complaint was of psychiatric injury caused by the police’s cavalier treatment of the claimant when they were first called to Stephen Lawrence’s horrific murder, unsure at that stage whether he was a suspect or a witness, and intent above all on finding out who were the killers.

125.  I recognise too that the facts of the present case are really very strong—as, indeed, I clearly recall having regarded the facts in Osman when originally that case was before the Court of Appeal in 1992: [1993] 4 All ER 344.

126.  But all that said, is it really possible to find a satisfactory basis upon which to distinguish this class of case (it would, of course, have to be distinguished on a class basis since, if a duty of care were found to exist, claims a good deal less meritorious than this one would inevitably be brought) from all the many other situations in which the Hill principle would clearly apply?

127.  Not without hesitation—for Lord Bingham’s opinion is undoubtedly persuasive and it is tempting to provide redress in as meritorious a case as this— I conclude not.

128.  In the first place, it seems to me difficult to limit “the liability principle” in the way my Lord does. Why, logically, should the principle apply only to threats to “life or physical safety” and not also to property? Paragraphs 54 and 55 of my Lord’s opinion, indeed, invite that very question. I do not question for a moment the highly authoritative statements made by the House in Glasbrook Brothers Ltd v Glamorgan County Council [1925] AC 270 as to the “absolute and unconditional obligation” on the police to protect property, as well as persons, from criminal injury. But nothing said there suggests that a breach of this obligation would give rise to civil liability for damages and, indeed, the authors of those statements would surely have been astonished by any such suggestion.

129.  Secondly, as Lord Hope asks at para 77, who is to judge whether the evidence is both (a) “apparently credible” and also (b) of “a specific and imminent threat"? These, of course, are critical preliminary questions since “the liability principle” would determine whether or not a duty of care arises in the first place.

130.  Thirdly, although it is tempting to regard the police’s function from which this duty is said to spring as something different from that of “investigating crime” (the function which prompted the Hill principle)— as, perhaps, the exercise of a protective function—this really would be a false dichotomy. Not only are the threats themselves criminal offences which, like other reported crimes, fall to be investigated but, as I indicated earlier, the police have to investigate many other risks of physical violence which come to their attention.

131.  Fourthly, some at least of the public policy considerations which weighed with the House in Hill and Brooks to my mind weigh also in the present factual context. I would emphasise two in particular.

132.  First, concern that the imposition of the liability principle upon the police would induce in them a detrimentally defensive frame of mind. So far from doubting whether this would in fact be so, it seems to me inevitable. If liability could arise in this context (but not, of course, with regard to the police’s many other tasks in investigating and combating crime) the police would be likely to treat these particular reported threats with especial caution at the expense of the many other threats to life, limb and property of which they come to learn through their own and others’ endeavours. They would be likely to devote more time and resources to their investigation and to take more active steps to combat them. They would be likely to arrest and charge more of those reportedly making the threats and would be more likely in these cases to refuse or oppose bail, leaving it to the courts to take the responsibility of deciding whether those accused of making such threats should remain at liberty. The police are inevitably faced in these cases with a conflict of interest between the person threatened and the maker of the threat. If the police would be liable in damages to the former for not taking sufficiently strong action but not to the latter for acting too strongly, the police, subconsciously or not, would be inclined to err on the side of over-reaction. I would regard this precisely as inducing in them a detrimentally defensive frame of mind. Similarly with regard to their likely increased focus on these reported threats at the expense of other police work.

133.  The second public policy consideration which I would emphasise in the present context is the desirability of safeguarding the police from legal proceedings which, meritorious or otherwise, would involve them in a great deal of time, trouble and expense more usefully devoted to their principal function of combating crime. This was a point made by Lord Keith in Hill and is of a rather different character from that made by Lord Steyn in para 30 of his opinion in Brooks—see para 51 of Lord Bingham’s opinion. In respectful disagreement with my Lord, I would indeed regard actions pursuant to the liability principle as diverting police resources away from their primary function. Not perhaps in every case but sometimes certainly, the contesting of these actions would require lengthy consideration to be given to the deployment of resources and to the nature and extent of competing tasks and priorities.

134.  Just such policy considerations as these (the conflicts of interest involved and the desirability of limiting litigation against those concerned to act in the interests of the wider community) informed the judgments of the House, not only in Hill and Brooks but also (of the majority) in D v East Berkshire Community Health NHS Trust [2005] 2 AC 373.

135.  True it is that in Brooks both Lord Nicholls of Birkenhead and Lord Steyn contemplated the possibility of exceptional cases on the margin of the Hill principle which might compel a different result. If, say, the police were clearly to have assumed specific responsibility for a threatened person’s safety—if, for example, they had assured him that he should leave the matter entirely to them and so could cease employing bodyguards or taking other protective measures himself—then one might readily find a duty of care to arise. That, however, is plainly not this case. There is nothing exceptional here unless it be said that this case appears exceptionally meritorious on its own particular facts—plainly not in itself a sufficient basis upon which to exclude a whole class of cases from the Hill principle. That said, the apparent strength of this case might well have brought it within the Osman principle so as to make a Human Rights Act claim here irresistible.

136.  That brings me to the final argument advanced for finding a common law duty of care in these cases. It is suggested that the common law should now be developed to reflect the Strasbourg jurisprudence about the positive obligation arising under articles 2 and 3 of the Convention. This, indeed, it was which so clearly influenced the Court of Appeal in the present case. For my part, however, I would reject the argument. To the extent that articles 2 and 3 of the Convention and sections 7 and 8 of the Human Rights Act already provide for claims to be brought in these cases, it is quite simply unnecessary now to develop the common law to provide a parallel cause of action—although it might have been otherwise had the Osman line of authority become established before the Human Rights Act came into force. And to the extent that the proposed development of the common law would go further than Strasbourg jurisprudence—as undoubtedly it would under the liability principle—this would seem to me undesirable and to give insufficient weight to the public policy considerations already referred to which to my mind militate against the creation of civil liability in these cases. Indeed it is plainly because of Strasbourg’s own recognition of these public policy considerations (see the passage from para 116 of Osman cited at para 115 above), that the Osman test itself has been so narrowly drawn.

137.  True it is that the possibility of a Human Rights Act claim now to some extent weakens the value of the Hill principle insofar as that is intended to safeguard the police from the diversion of resources involved in having to contest civil litigation. That, however, is no good reason for mirroring the Osman principle by the introduction of a common law duty of care in this very limited class of case, still less for weakening the value of the Hill principle yet further by creating a wider duty of care.

138.  There is this too to be said as to why, certainly in the present context, your Lordships should not feel tempted to develop the common law “in harmony with” Convention rights (as Rimer LJ put it below). As Lord Bingham pointed out in R (Greenfield) v Secretary of State for the Home Department [2005] 1 WLR 673, Convention claims have very different objectives from civil actions. Where civil actions are designed essentially to compensate claimants for their losses, Convention claims are intended rather to uphold minimum human rights standards and to vindicate those rights. That is why time limits are markedly shorter—the one year (albeit extendable) limitation period under section 7(5) of the Human Rights Act comparable to the one year permitted for defamation claims intended, analogously, to vindicate a claimant’s reputation. It is also why section 8(3) of the Act provides that no damages are to be awarded unless necessary for just satisfaction. It also seems to me to explain why a looser approach to causation is adopted under the Convention than in English tort law. Whereas the latter requires the claimant to establish on the balance of probabilities that, but for the defendant’s negligence, he would not have suffered his claimed loss—and so establish under Lord Bingham’s proposed liability principle that appropriate police action would probably have kept the victim safe—under the Convention it appears sufficient generally to establish merely that he lost a substantial chance of this.

139.  Clearly the violation of a fundamental right is a very serious thing and, happily, since the Human Rights Act, it gives rise to a cause of action in domestic law. I see no sound reason, however, for matching this with a common law claim also. That to my mind would neither add to the vindication of the right nor be likely to deter the police from the action or inaction which risks violating it in the first place. Such deterrence must lie rather in the police’s own disciplinary sanctions (as, indeed, were applied in Van Colle) and, in a wholly exceptional case like R v Dytham, in criminal liability. Rather I am satisfied that the wider public interest is best served by maintaining the full width of the Hill principle. There is, of course, in these cases (as in D v East Berkshire) always a price to be paid by individuals denied for public policy reasons (as not being “fair, just and reasonable” within the Caparo principle—Caparo Industries plc v Dickman [1990] 2 AC 605) a civil claim in the interests of the community as a whole. At least in the present context the state makes some payment under the Criminal Injuries Compensation Act (albeit nowadays a tariff sum far short of common law damages). But nonetheless one sympathises with Mr Smith for paying the price as well as for his severe injuries.

140.  As a final comment I add only this. In common, I think, with all your Lordships, I regards this issue as plainly one which the House should decide one way or the other on the pleaded facts. Either a duty of care arises on these facts or it does not. No useful purpose would be served by allowing the action to go to trial for facts to be found and then for further consideration to be given to the applicable law. In my judgment the Court of Appeal erred in this respect as well as in supposing that the claim might ultimately succeed.

141.  For these reasons, together with those given by Lord Hope, I too would allow the Chief Constable’s appeal, set aside the Court of Appeal’s order and restore the order of the judge below by which the respondent’s claim was struck out.