WHAT is the proper inter-relationship between tort liability (at common law) and claims for breach of the ECHR under the Human Rights Act 1998, ss. 6–8 (HRA)? The question has considerable importance. Sharp divergence exists over public authority liability. The European Court of Human Rights in Strasbourg has recognised positive protective duties requiring state intervention to protect individuals against harm from other private individuals. Whereas the common law still denies a general positive duty of care upon public authorities to protect members of the public (e.g. Michael v Chief Constable of South Wales [2015] UKSC 2, [2015] A.C. 1732). Thus claims for breach of the ECHR may often succeed when a negligence claim would not. The House of Lords held decisively that the systems run along parallel channels and do not mix; that is, the existence of positive duties under the ECHR does not require analogous expansion of common law liability: Van Colle v Chief Constable of Hertfordshire [2008] UKHL 50, [2009] A.C. 225.
Does the inverse corollary hold? It might seem inevitable that the common law's restrictive stance towards positive duties can have no inhibitory effect upon the ECHR's more extensive duties. This would seem to follow from the “separate channels” approach. And that is precisely what the majority held in DSD v Metropolitan Police Commissioner [2018] UKSC 11, [2018] 2 W.L.R. 895. But the different reasoning of Lord Hughes (concurring in the result) shows that the question is not straightforward.
For Lord Hughes, the common law's firm hostility to positive duties should limit the reach of the ECHR, at least when the Convention's meaning is not settled. Domestic courts are, of course, bound to apply the ECHR in claims brought under the HRA, and the Strasbourg Court's decisions about the Convention's meaning are an important – frequently determinative – factor in such cases. HRA, s. 2, may provide merely that UK courts “must take into account” judgments of the Strasbourg Court. But, as is well known, the courts have adopted a realistic self-denying ordinance under which they are normally, de facto, bound to follow clear and settled Strasbourg caselaw (e.g. Lord Bingham's classic statement in R. (Ullah) v Special Adjudicator [2004] UKHL 26, [2004] 2 A.C. 323, at [20]). The courts are only too aware that failing to follow such caselaw is likely to lead to “reversal” in the European Court of Human Rights. On rare occasions domestic courts have (as permitted by HRA, s. 2) consciously departed from Strasbourg's position in the hope of making the European Court think again on an important question (see e.g. R. v Horncastle [2009] UKSC 14, [2010] 2 A.C. 373 and Horncastle v United Kingdom (2015) 60 E.H.R.R. 31).
The majority in DSD held the relevant Strasbourg caselaw to be clear and consistent. Lord Hughes and Lord Mance disagreed. So the overt point of difference in DSD concerned the state of the Strasbourg authorities. Yet a deeper disagreement is discernible. If the common law's long-standing refusal to recognise liability for public authority nonfeasance is well-founded as a matter of principle, public policy or both, should that immunity be comprehensively outflanked by a liberal approach to ECHR claims? If the liability channel is rightly watertight at common law, should the sluices be opened to a flood of claims in the ECHR channel?
In DSD the Supreme Court unanimously upheld the award of compensation for breach of Article 3 ECHR (which prohibits, inter alia, “inhuman or degrading treatment”). The defendant police force had not itself subjected the claimants to such treatment. But the police had (through admitted serial incompetence) failed to investigate properly the claimants’ rapes by the notorious and prolific “black cab rapist” John Worboys.
Does Article 3 impose a duty to investigate inhuman or degrading treatment (which rape clearly is) at the hands of private individuals? The Strasbourg Court said so in MC v Bulgaria (2003) 40 E.H.R.R. 20. That decision has often been followed since in Strasbourg. For the majority in DSD, this was “clear and constant jurisprudence” that UK courts ought to follow (in the absence of “special circumstances” justifying Horncastle-style defiance of Strasbourg): see at [44], per Lord Kerr, quoting R. (Alconbury Developments Ltd.) v Secretary of State for the Environment, Transport and the Regions [2001] UKHL 23, [2003] 2 A.C. 295, at [26], per Lord Slynn.
Lord Hughes disagreed. On its core textual meaning, Article 3 prohibits torture and cognate mistreatment by the state. This has been extended by two Strasbourg “glosses”. First, the state has a duty to investigate alleged inhuman treatment by its own officers (to buttress effective enforcement of the main prohibition). The second judicial gloss requires positive state action to protect individuals from inhuman treatment by other individuals. For Lord Hughes, the crucial decision in MC v Bulgaria was a gloss upon a gloss, and a doubtful one. The Strasbourg Court had failed to consider in MC whether the rationale of the investigative duty in state-actor cases requires investigations into inhuman treatment from any source (public or private). The Court had failed to consider the objections to this double extension of Article 3: that it would transform the ECHR into a system for “monitor[ing] every act of enforcement or policing of the varied domestic legal requirements” relating to every violent crime (“a regrettably unavoidable feature of life”): at [109]. Nor had this objection been considered in the extensive caselaw uncritically applying MC v Bulgaria.
Similarly, Lord Mance lamented at [142] Strasbourg's “unfortunately, not unprecedented” failure to uphold basic standards of caselaw technique. He criticised the extension of “a solidly rationalised principle … to situations to which the rationale does not apply, without overt recognition of the extension, without formulating any fresh rationale and relying on supposed authority which does not actually support the extension” (ibid.). Lord Mance too thought that MC v Bulgaria provided but slender support for the weight of caselaw subsequently piled upon it. In particular because of the Strasbourg Court's failure to consider the serious “implications for policing” of the doctrine in MC v Bulgaria – in notable contrast with the extensive English discussions of the possible negative effects of tort liability upon police efficacy (prominent examples of such discussion being in Van Colle and Michael, above).
Despite such doubts, Lord Hughes did not deny the existence of an investigatory duty. But at [127] he would have confined it to “a proper structure of legal and policing provision” to investigate and punish “reports of past violence”; that is, Article 3 did not require “assessment of whether [a particular] investigation was careless or made mistakes which ought not to have been made”. On the facts though, Lord Hughes held at [140] that the failings of the defendant police in DSD had comprised “plain structural errors”.
To some extent the majority shared Lord Hughes's concerns. Lord Kerr stressed at [29] that not every mistake in police investigations of violent crime would amount to a compensable breach of Article 3. There would have to be “egregious and significant” or “conspicuous or substantial errors in investigation”. Lord Neuberger took a similar approach. At [96] he rejected Lord Hughes's approach because of the difficulty of defining “systemic” failings. But only “serious defects” in investigation would ground liability. As Lord Neuberger recalled, the Strasbourg Court had itself recognised in Osman v United Kingdom [1999] 1 F.L.R. 193, at [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”, and the consequent need to interpret the ECHR “in a way which does not impose an impossible or disproportionate burden on the [police] authorities”. (Compare Lord Hughes at [127]: this important passage from Osman has only occasionally been cited in Strasbourg caselaw.)
The resulting high threshold for breach of Article 3 still contrasts with the common law. The effect of holding that there is no duty of care is no liability however grave or gross the negligence in question. Lord Bingham's advocacy of a serious breach threshold in his dissent in Van Colle (above) was deliberately rejected. The Law Commission's suggestion of a similar test for public authority liability was withdrawn after government opposition (cf. Law Com. C.P. 187 (2008)).
The majority in DSD viewed any outstanding divergence with equanimity. Divergence was inherent in the Van Colle approach where Lord Brown had said at [138] that negligence and HRA claims have different purposes. More specifically, Lord Kerr held that the reasons for denying common law liability (most recently in Michael, above) had no application to ECHR claims. The common law concept of “proximity” had no equivalent in the Strasbourg jurisprudence: at [69]. Nor could the “fair, just and reasonable” element in duties of care “easily be accommodated in Convention jurisprudence”: at [70].
Whether or not these doctrinal points are convincing (do different concepts play a similar limiting role in the European caselaw?), the majority's main reason for denying that the common law and Article 3 must be consistent was scepticism about the predicted deleterious effects of liability on police efficiency. Lord Kerr said at [71] that there was no evidence for this claim (made by Lord Hughes at [132]) – ignoring the fact that it had also grounded the pivotal decision in Van Colle. At [97], Lord Neuberger thought that it was just as plausible that liability would improve the quality of police investigations than hamper it. Maybe so. But does that justify the divergent approaches in the Strasbourg and English caselaw as easily as Lord Neuberger believed? This is not divergence based on principle, but merely on different factual assumptions. Is such inconsistency tolerable? When domestic law is (for good or ill) so firmly settled on the point in question, there is good sense in interpreting the ECHR to limit the inconsistency, where possible. As Lord Hughes said at [134], while the common law cannot determine the meaning of an international human rights treaty, still when the same “delicate balance” between police operational independence and judicial control must be struck under the Convention, the hard-fought position at common law should at least be relevant. Especially, one might add, when the Strasbourg jurisprudence has only erratically considered the implications of over-broad liability for police efficacy.
The “separate channels” approach may eventually be reconsidered (and a final systemic choice made between broader and narrower public authority liability). But then it is effectively a compromise enabling the common law and ECHR to proceed by their own lights. Compromise frequently endures.
