I. Introduction
The UK Supreme Court noted in Smith v Chief Constable of Sussex Police that the approach to causation in claims based on Article 2 of the European Convention on Human Rights (ECHR) is “looser” than in negligence. While the but-for test is generally applied in negligence, in Article 2 claims “it appears sufficient generally to establish merely that [the claimant] lost a substantial chance” of avoiding harm.Footnote 1 The English courts have not always been comfortable with this divergence. In Re E. (A Child) Lady Hale stated that she was “troubled by the rejection of the ‘but for’ test” by the European Court of Human Rights (ECtHR).Footnote 2 In part this discomfort seems attributable to a lack of detailed exposition as to how the approaches to causation differ in negligence and human rights law. The lack of clarity as to what the “looser” approach consists of is unsatisfactory. Although the reference to causation in Smith was a relatively minor observation in the wider decision not to develop the duty of care owed by the police in negligence to a victim of crime, we cannot appreciate whether the difference in approach is justified without first understanding what the difference actually is. In response, one aim of this paper is to draw together key decisions on causation in the ECHR to identify what the “looser” approach entails and how it differs from the approach in negligence.
Divergences in legal doctrines are generally attributed to the fact that the two causes of action have different aims. In Van Colle, Lord Brown explained that
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… It is also why section 8(3) of the [Human Rights Act 1998] provides that no damages are to be awarded unless necessary for just satisfaction.Footnote 3
There has, however, been little analysis of whether those differing aims explain the differences in approach to causation. A second aim of this paper is therefore to address concerns about the divergences by considering why they exist, and whether they can be justified by the different objectives of negligence and human rights law and by the different nature of the rights being protected.
The paper begins, in Section II, by elucidating the causal requirements and the function of causation within each area of law. It is beyond the scope of this paper to address all of the ECHR rights, so the focus is on Articles 2 and 3, and, in the final section Article 8, which have useful parallels with negligence law. Section III then contemplates in greater depth the relationship between the nature of the rights being protected and the approaches adopted towards causation. It does this by comparing developments in causation in a jurisdiction that does not maintain the same separation between negligence and human rights law, South Africa. South African courts must “respect, protect, promote and fulfil the rights in the Bill of Rights” (s. 7(2) of the South African Constitution) which, it has been held, requires convergence between the law of delict and the Bill of Rights: “where the common law deviates from the spirit, purport and objects of the Bill of Rights the courts have an obligation to develop it by removing that deviation”.Footnote 4 This led the Constitutional Court in Lee v Minister for Correctional Services to adopt a flexible approach to the factual causation requirement in a negligence action against a state body in order to protect the claimant's constitutional right to human dignity.Footnote 5 This was a risk-based approach to causation, which has some parallels in the Fairchild exception in English law and is similarly difficult to reconcile with wider negligence principles,Footnote 6 yet is more difficult to circumscribe as an exceptional approach. In comparison, the separate development of negligence and human rights law in the UK permits differences of approach to causation which, it is argued, better reflect the objectives of each area of law. The final section of the paper returns its focus to the jurisprudence of the ECtHR and highlights areas where a “looser” approach shades into opacity, where the court's failure to articulate its reasoning as to causation obscures the function of the damages award. Where the right to private and family life is engaged by a failure to provide information about risks of harm, the court does not differentiate between damages that vindicate the right and damages that compensate any resulting injury. These issues are not inherent in the “looser approach” and ought to be resolved by the court. Additionally, by appreciating what the “looser approach” does entail and how it reflects the right being protected, it is suggested that the Article 8 right could potentially be developed in relation to failure to warn of risk in a healthcare context. Finally, where the right to life is engaged in a healthcare context it will be seen that there is a lack of clarity as to the causal requirements, and the court generally fails to identify whether those causal principles affect liability or merely the availability or quantum of damages. The result, it is argued, is that the vindicatory objective of liability is frustrated by an implicit damage requirement.
II. The Function of Causation in Negligence and Human Rights Actions
To bring a negligence action the claimant must generally establish that the defendant's negligence was a but-for cause of,Footnote 7 or materially contributed to,Footnote 8 the damage suffered. In contrast, to bring a human rights claim it is not necessary to show that what the defendant did resulted in damage, it is enough to show that what they did violated the relevant right. Causal concepts are relevant to establishing this violation in some contexts, but their role is much more limited as will be explored below. The limited relevance of causation of damage is also reflected in the remedies. While the remedy in negligence is compensatory damages, “just satisfaction” under Article 13 of the ECHR and section 8 of the Human Rights Act 1998 (HRA) allows for a range of remedies such as declarations, and damages are not available as of right. Where damages are sought for pecuniary or non-pecuniary loss, it must be shown that the violation of the relevant right caused this loss,Footnote 9 but this affects the remedy rather than liability itself.
The different function of causation in the two types of action reflects the different kinds of rights being protected. As Nolan has argued, damage is a constitutive element of the claim right in negligence, so that the wrong consists of negligently causing damage rather than simply of exposing others to unreasonable risks of interference with their bodily integrity, personal property etc with damage confined to a role as a condition of actionability.Footnote 10 The right is therefore a right against the other person that they not cause damage to the right-holder through a failure to take reasonable care (against the risk of that damage). Causation occupies a central role since it is the relation that connects the unreasonable risk with the damage suffered by the claimant.Footnote 11 In contrast, the violation of ECHR rights need not entail damage or loss. These fundamental rights can be violated by acts (or omissions) that have the potential to cause loss, so that the violation of the right itself is actionable.Footnote 12
Violation of the right to life need not entail death at all, for example the manner in which the police pursued the applicant in Makaratzis v Greece violated his right to life through “conduct which, by its very nature, put his life at risk, even though, in the event, he survived”.Footnote 13 As Wicks explains:
it is not, or at least not only, the sanctity of life that is protected under the ECHR's right to life but rather the necessary respect for all human life. The focus seems to be less upon life versus death than upon the protection of everyone from actions that put their life at risk and thus fail adequately to respect it.Footnote 14
The Article 2 right to life entails various duties: a negative duty on the state to refrain from interfering with the right to life, a positive duty on the state to secure the protection of the right to life against interference by non-state actors in certain circumstances, and a procedural duty to investigate interferences. Clearly a breach of the procedural obligation does not cause death, indeed the procedural obligation highlights that an important aspect of the right to life is having a finding of fact about the circumstances, and not just the causes, of death. The decision in R. (on the application of Tainton) v HM Senior Coroner for Preston and West Lancashire considers how causation should be addressed by a coroner when the deceased has died in circumstances where the possibility of an Article 2 violation cannot be excluded. This was a case where the deceased died of natural causes, oesophageal cancer, but there was a five month delay in diagnosis due to the substandard healthcare he received while in custody. The High Court concluded that the coroner was right not to put the issue to the jury because it would not have been safe for them to conclude, on the balance of probabilities, that the substandard medical care had causally contributed to the death. While recording a verdict of death from natural causes they ought, however, to have included in the Record of Inquest a brief narrative “of any admitted failings forming part of the circumstances in which the deceased came by his death, which are given in evidence before the coroner, even if, on the balance of probabilities, the jury cannot properly find them causative of the death”.Footnote 15 The court went on to state that this “was required in this case in order to discharge in full the obligation on the state imposed by article 2 of the ECHR”.Footnote 16 In this way, Article 2 demonstrates concern with respect for life rather than solely with protection for the sanctity of life.
The emphasis on risk rather than damage is evident in Osman v United Kingdom, where the ECtHR held that to bring an Article 2 claim against the police under the positive duty to safeguard, it must be shown that
the authorities knew or ought to have known at the time of the existence 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.Footnote 17
Wright correctly states that
This is a very much weaker test than the standard “but for” test, but it is consistent with the aims of the ECHR which are the promotion and protection of human rights standards and the rule of law, rather than compensation for damage dependent upon proof of loss.Footnote 18
In respect of Article 3 claims, where it is likewise not necessary for the victim to suffer physical or mental harm since the right is violated when the treatment of the victim amounts to inhuman or degrading treatment or punishment,Footnote 19 the same approach applies. This was explained in E. v United Kingdom, a case where the local authority failed to monitor a step-father after he was convicted of sexual abuse so failed to detect that he was abusing the children and take steps to protect them. The ECtHR said there
The test under article 3 however does not require it to be shown that “but for” the failing or omission of the public authority ill-treatment would not have happened. A failure to take reasonably available measures which could have had a real prospect of altering the outcome or mitigating the harm is sufficient to engage the responsibility of the state.Footnote 20
The significance of this risk-focused approach is evident from the case of Sarjantson v Chief Constable of Humberside Police.Footnote 21 The claimant in that case suffered serious injury when he was attacked by a group of young men. As the incident unfolded a number of calls were made to the police, by various people, reporting that the group were armed with baseball bats and attacking individuals, eventually including the claimant. An internal police investigation found that there had been an 11 minute delay before police officers were deployed to the scene of the incident. The claim was brought under Articles 2 and 3, arguing that there had been a breach of the positive duty to take measures to avert a real and immediate risk to life and a real and immediate risk of injury, respectively. One of the arguments made by the police was that they should not be held liable since causation was not established because even if there had been an immediate despatch of officers they would not have arrived until after the claimant had been assaulted. This argument did not succeed and the Court of Appeal was clear that for Article 2 and 3 claims such a causation inquiry is not necessary: “the fact that a response would have made no difference is not relevant to liability”.Footnote 22 They explained:
The duty to provide protection arose at the time when the first emergency call was made. At that time, it was impossible to know whether and, if so, how quickly an assault would take place …. As the court made clear at para 116 in Osman, it must be established that the police knew or ought to have known “at the time” of the existence of a real risk and immediate risk …. This implies that compliance with article 2 should not be determined with the benefit of hindsight. This is confirmed by the court saying at para 116 “… and that [the authorities] failed to take measures within the scope of their powers which, judged reasonably, might have been expected to avoid that risk”.Footnote 23
This is what Wright describes as an obligation “of means, not results”.Footnote 24 By focusing on the ex ante risk rather than the result, the element of moral luck is excluded from the determination of liability,Footnote 25 and the inquiry squarely addresses the question of whether the police showed adequate respect for the claimant's life. The court did clarify that causation is relevant to quantum: “A finding that a response would have made no difference may mean that there is no right to damages. But it is not relevant to liability”.Footnote 26
This is particularly important to note in the context of claims against the police where the courts have resisted attempts to impose a duty of care on the police in negligence for failure to prevent harm caused by a third party. Wright is critical of this refusal to develop the duty of care in light of the HRA, arguing that negligence cases like Hill and Smith,Footnote 27 “whether we like it or not…were cases about human rights; formalistic reasoning can cast dust in our eyes and prevent us from seeing the real issues”.Footnote 28 Tofaris and Steel also note that there are disadvantages to bringing a HRA claim such as the shorter limitation period and that damages are not available as of right.Footnote 29 It is not necessary to reject the argument that these cases were about human rights to accept the view that they were also cases about the private law rights protected by the tort of negligence. The case of Sarjantson highlights the potential benefits to the claimant of bringing a claim under Article 2/3 rather than in negligence. Even if the police had owed a duty of care to the claimant in Sarjantson in negligence, the claim would have failed for lack of causation. Instead, the claim under Article 2 was successful. While the claimant would not have recovered compensatory damages for the injury suffered, they were still entitled to a remedy.
The risk-based approach has troubled the English courts at times. In Re E. (A Child),Footnote 30 Lady Hale expressed concern that it was not necessary to satisfy the but-for test in order to establish an Article 3 claim. The claims in that case concerned a mother and daughter who, along with others attending the same school, had to walk to and from school for a period of a few months through an area of sectarian protest and disorder where they were subjected to abuse of a violent and intimidating nature,Footnote 31 and where the walkway established by police was delineated by armoured vehicles and police officers with riot shields. The protest was eventually resolved through ongoing efforts of the police and other bodies, and the claimant argued that the police should have taken steps sooner to actively remove the protestors. Ultimately the court found that given the delicate and volatile situation, the police had taken reasonable measures because acting sooner may have exacerbated the problem. Lady Hale stated, however, that she was “troubled by the rejection of the ‘but for’ test” in E. v UK,Footnote 32 and observed that “I do not think that it has been demonstrated that, had the police behaved at the outset in the way in which it is now said that they should have behaved, the children's experience would have been any better. Indeed, it could have been a great deal worse”.Footnote 33 In practice it may be difficult to distinguish clearly between measures that would have prevented the outcome and measures that “could have had a real prospect of altering the outcome”.Footnote 34 The key to understanding the difference is that the but-for test depends upon hindsight while the Osman/E. approach is prospective. Osman is a risk-based analysis so it is an ex ante inquiry into the impact measures could have been expected to have, assessed at the time the measures should have been taken. It seems to be a very contextualised, fact-specific approach, so rather than asking simply whether police action could have reduced the risk in Re E., the court was concerned with whether the specific steps it is argued that the police should have taken could have affected the outcome, including whether they could also have made the outcome worse. This is still, however, a forward-looking inquiry. As Steel has explained, the creation of risk in the absence of causation is insufficient to justify negligence liability since the idea of undoing the wrong cannot justify the obligation to compensate: in the world in which the defendant's wrong (risk creation) does not occur, the claimant is still injured.Footnote 35 Since human rights liability is not primarily concerned with compensation but with upholding respect for particular rights, the wrong can consist of risk creation irrespective of whether loss follows.
We must not ignore the fact that in Fairchild v Glenhaven Funeral Services, the House of Lords did adopt a risk-based approach to proof of causation in a negligence action.Footnote 36 The victim in Fairchild had been exposed to asbestos through the defendant's negligence and died from mesothelioma, a cancer that is caused by asbestos. A combination of factors meant that the claimant, his widow,Footnote 37 was unable to establish causation on the but-for test: in addition to the defendant's negligent asbestos exposure, the victim had been exposed to asbestos by a number of former employers over the course of his working life, and there was an evidentiary gap in that while it was known to medical science that asbestos does cause mesothelioma, it was not known how it does so: whether it is caused by one or many fibres, and at what stages in the development of the disease asbestos plays a role. The court adopted an exceptional approach to causation, holding that it was sufficient to prove that the defendant's negligence had materially increased the risk of harm. On this basis the defendant was held liable for the whole of the claimant's loss. It was subsequently held in Barker v Corus that liability should be apportioned among those who had exposed the claimant to asbestos,Footnote 38 and Lord Hoffmann rationalised this on the basis that the “damage” the defendant had been proved to have caused was not the disease but the risk of that disease,Footnote 39 but more recent decisions have held that it is still the mesothelioma that forms the gist of the negligence action.Footnote 40 This damage requirement is an important point of difference from the risk-based approach to HRA liability.Footnote 41 As we have seen, liability under Article 2 is imposed when the defendant fails adequately to respect the relevant right of the victim by exposing them to a risk of death. Death need not occur. For negligence liability to arise within the scope of the Fairchild exception the claimant must have developed mesothelioma.Footnote 42 This means that while the causation requirement has been significantly relaxed, it is still regarded by the courts as an essential ingredient of liability along with damage. This is not to suggest that the Fairchild exception can be easily reconciled with wider negligence principles, but it reflects the importance of damage and causation within the rights protected in negligence.
III. Causal Tensions between Negligence and Human Rights
Having clarified the differences of approach to causal issues in negligence and human rights law, this section considers the complications that arise when negligence law is developed as a mechanism for protecting human rights. It takes as a comparator South Africa, where the courts must develop the common law in order to protect, promote and fulfil the rights contained in the Bill of Rights. This will help further the argument that causal principles occupy different roles in negligence and HRA claims, reflecting the different nature of the rights being protected, such that in some instances it may be appropriate to hold the defendant liable in both branches of law.
A. The South African Experience: Lee
The decision of the South African Constitutional Court in Lee v Minister of Correctional Services provides an example of the strain that is placed on the causation requirement in negligence, delict in South Africa, when there is less clear separation between negligence and human rights law.Footnote 43 Faced with a problem of proof of factual causation, the majority “stretched” the concept of causation,Footnote 44 while the minority favoured developing the law to carve out an exception. The former solution challenges the internal coherence of the causation requirement, and both approaches strain the coherence of the law by pursuing the goal of protecting the constitutional right to human dignity within the law of delict.
The claimant in that case had contracted tuberculosis whilst in prison. He sought to recover damages in delict arguing that the state had failed to implement a reasonable system for the management of tuberculosis in the prison. Tuberculosis is an airborne communicable disease which spreads easily, especially in confined, poorly ventilated, overcrowded environments. The overcrowded, poor conditions in the prison were ideal for transmission of the disease. The negligence in this case consisted not of the individual treatment of the claimant, but of systemic failure to take preventive and precautionary measures against the spread of tuberculosis. The Supreme Court of Appeal had rejected Mr. Lee's claim since he had failed to establish that but-for the negligence of the prison authority he would not have developed tuberculosis. They considered that to succeed he needed to show what a reasonable system would have consisted of, and that such a system “would have altogether eliminated the risk of contagion”.Footnote 45 While this seems to apply a standard of proof more demanding than the balance of probabilities, this view was based on the difficulty of pinpointing the source of the claimant's infection and therefore the difficulty of showing that his individual case of tuberculosis would have been avoided unless it was possible to show that all cases would have been avoided.
Under section 7(2) of the South African Constitution, courts, as organs of the state, must “respect, protect, promote and fulfil the rights in the Bill of Rights”. The Constitutional Court explained in Carmichele that this requires convergence between the law of delict and the Bill of Rights: “where the common law deviates from the spirit, purport and objects of the Bill of Rights the courts have an obligation to develop it by removing that deviation”.Footnote 46 This is a more onerous obligation than UK courts face under s6 HRA, which provides that it is unlawful for a public authority, including a court, to act in a way which is incompatible with a Convention right. Wright notes that this is “an obligation to ‘respect’, but not to protect or fulfil the relevant standards”,Footnote 47 so the horizontal effect of the HRA is much more limited. The relevant right in Lee is the right to human dignity: “Everyone has inherent dignity and the right to have their dignity respected and protected”.Footnote 48 The court therefore framed the issue in the following way:
The complaint is that the unlawful detention and specific omissions violated the applicant's right to freedom and security of the person and the right to be detained under conditions consistent with human dignity, and to be provided with adequate accommodation, nutrition and medical treatment at state expense. The question is whether the causation aspect of the common law test for delictual liability was established and, if not, whether the common law needs to be developed to prevent an unjust outcome.Footnote 49
Just as there are various causes of action in English law to address the failings of state bodies including liability in tort law and under the HRA 1998, in South African law there are also a range of sources of damages including the law of delict, statutory compensation schemes, the Promotion of Administrative Justice Act 3 of 2000 (PAJA) which gives courts the power in judicial review actions to award compensation in exceptional cases, and “constitutional damages” which may constitute “appropriate relief” where the state breaches a requirement imposed by the Constitution.Footnote 50 Both the majority and minority approaches in Lee will be explored in more detail since arguably both threaten the coherence of delictual liability and a preferable solution would have been an award of constitutional damages.
The decision of the Supreme Court of Appeal was overturned by a 5:4 majority of the Constitutional Court who considered the law of delict already adopted a “flexible” approach to factual causation, which was satisfied on these facts, so there was no need to develop the law. They considered that this “flexibility” allowed them to accept proof of an increase in risk as sufficient in this case:
It seems to me that if a non-negligent system reduced the risk of general contagion, it follows – or at least there is nothing inevitable in logic or common sense to prevent the further inference being made – that specific individual contagion within a non-negligent system would be less likely than in a negligent system. It would be enough, I think, to satisfy probable factual causation where the evidence establishes that the plaintiff found himself in the kind of situation where the risk of contagion would have been reduced by proper systemic measures.Footnote 51
While there is much to be said for an approach that is qualitative and avoids recourse to a merely quantitative approach to proof such as the “doubles the risk” test which conflates the balance of probabilities standard of proof with what it is that is being proven,Footnote 52 this approach effectively substitutes the but-for test with a material increase in risk test. This is problematic because, unlike in the English decision in Fairchild,Footnote 53 there is no suggestion in the majority judgment that this is an exceptional approach with a limited scope of application. Indeed, the court expressly distanced itself from the need for exceptional approaches to causation that have been developed in other jurisdictions, stating that the need for such exceptional approaches arises out of inflexible application of the but-for test, so the flexible application of the but-for test in South African law eliminates the need for recourse to exceptions.Footnote 54 This means that it potentially applies to claims against non-state actors and to cases involving operational rather than systemic negligence, indeed subsequent case law discussed below reinforces this impression. The coherence of the causation requirement is disrupted by this solution since it is unclear when courts should insist on but-for causation and when the increase in risk test is appropriate. Furthermore, it fails to acknowledge that proof of factual causation is central to establishing interpersonal responsibility and justifying why the defendant should be liable to compensate the claimant.
The minority judgment highlights the constitutional aspect of the issue: “All this indicates that the common law but-for test for causation is an over-blunt and inadequate tool for securing constitutionally tailored justice in cases where prisoners have proved exposure to disease because of negligence on the part of the prison authorities, but cannot pinpoint the source of their injury”.Footnote 55 Their preferred approach was to explicitly develop the law of delict in order to satisfy the requirements of the Constitution. This task, they argued, was best done by the High Court so that they could undertake a “full assessment of the intricacies of a system of risk-based compensation”,Footnote 56 for example whether it should lead to proportionate damages, and what should be its scope of application. There are various defining features of the case that may assist in determining the scope of the exception. As concerns tuberculosis itself, it was noted to be a serious public health problem in South Africa,Footnote 57 and its aetiology entails problems of proof of causation since it can be spread by a single airborne mycobacterium and carriers can be contagious before showing symptoms, meaning it is extremely difficult, perhaps impossible, to pinpoint the source of infection for an individual.Footnote 58 On a more policy-based level, the claimant was part of “a vulnerable group to whom our system of constitutional protections owes particular solicitude”Footnote 59; the defendant was a state body; there was a public interest in liability: “The country's interest in the development of a sound system of incarceration, in which risk of exposure to pathogens is minimised as much as is reasonably possible, suggests there may be a need to develop the common law of causation”.Footnote 60 The question remains whether the exception can be rationally circumscribed; the following section argues that this remains problematic.
B. The Aftermath of Lee: Blurred Lines
By blurring the lines between delict and constitutionally protected rights, it becomes unclear where the lines ought to be drawn around the increase in risk approach to causation. Although the majority in Lee rationalised their approach to causation as relying on existing flexibility, the South African High Court in Nkala v Harmony Gold Mining has since expressed the view that the Constitutional Court in Lee “has expanded our perceptions of causation”.Footnote 61 This decision concerned a negligence class action brought by mineworkers against gold mining companies in respect of silicosis and TB contracted during their employment. A number of the parties have subsequently reached a settlement,Footnote 62 but the remaining parties may still proceed with their claims. Proof of causation is particularly problematic in the TB claims, but the High Court has held that it is clear from the “development of the law” in Lee that the claimants are not incapable of proving a causal link.Footnote 63 The Lee approach still imposes significant hurdles for these claimants,Footnote 64 but it is significant that the court was willing to adopt it in a negligence claim against a non-state defendant. The Constitutional Court in Mashongwa v PRASA considered further the effect of Lee, confirming the view that “it adopted an approach to causation premised on the flexibility that has always been recognised in the traditional approach”,Footnote 65 whilst clearly regarding it as being a wider approach than but-for causation, stating “where the traditional but-for test is adequate to establish a causal link it may not be necessary, as in the present case, to resort to the Lee test”.Footnote 66 This case is also noteworthy in that, while the defendant was a state body, the negligence claim did not arise out of a systemic failing but from a failure to close the doors of the train compartment in which the claimant was travelling, which meant that criminals who attacked the claimant were able to throw him from the moving train. These cases suggest that the Lee test will be raised, and potentially applied, when causation cannot be established on the but-for test, in claims against state and non-state defendants, in cases involving systemic or isolated failings, and involving omissions and positive creation of risk.
Price has suggested that Lee might rationally be limited to cases involving systemic failures, noting that Lee itself
involved a systemic failure to provide an obligatory government service, rather than any specified negligent act or omission. That is, it involved a failure by an organ of state adequately to perform a positive constitutional obligation to design and implement a reasonable system to protect a vulnerable class of people against a genuine risk to their life and personal security.Footnote 67
This is significant because, besides any scientific problems of proof, and beyond the status of the defendant as a state body, it means that when applying the but-for test and constructing the hypothetical world in which the defendant took reasonable care there are a range of possible systems that would satisfy the requirement of reasonableness. The scope of the exception could therefore be limited to cases of systemic failings where proof of causation is hampered by the range of satisfactory counterfactual systems possible.
This is arguably more relevant than the act/omission distinction. In a more recent decision in Oppelt v Dept of Health, the Constitutional Court emphasised that Lee arose in the context of omissions liability and held that “the ‘but-for’ test is not always the be-all and end-all of the causation enquiry when dealing with negligent omissions”.Footnote 68 The act/omission distinction does not adequately circumscribe the Lee test. Besides the much discussed blurred border between acts and omissions, the need to replace the defendant's conduct with hypothetical non-negligent conduct is just as much a part of the application of the but-for test to acts as it is to omissions. In contrast, where there has been a systemic failure to provide a service, the hypothetical non-negligent conduct is more difficult to insert into the causation inquiry because a range of possible systems would likely have been reasonable. In the healthcare context, the ECtHR requires a systemic or structural dysfunction in hospital services which results in a patient being deprived of access to life-saving emergency treatment.Footnote 69 The facts of Oppelt resemble those where the ECtHR has established liability under Article 2 since it did not involve negligent medical treatment but “unreasonable delays [which] justified the conclusion that the applicant was refused emergency medical treatment”,Footnote 70 so might come within the scope of Lee even if Lee were confined to cases of systemic negligence. The application of Lee in the context of medical negligence still stands in contrast to the English approach to medical negligence where courts have resisted a loss of chance or increase in risk approach to proof of causation so would surely be slow to adopt it simply because the negligence consisted of a systemic failing.Footnote 71
It may not, however, be rational to limit a risk-based approach to causation to cases involving systemic failings as the following case shows. In the UK case of Re E. discussed above, Lady Hale similarly observed that it was not obvious what the police could have done to protect the children from harm because alternative responses would have carried different risks.Footnote 72 Yet Re E. did not concern a systemic failing, but the response to a particular situation (albeit a situation extending over a few months); the uncertainty seems to flow from the discretion afforded to public bodies. The Osman test also applies not only to systemic failings but to operational failings, and in the recent decision in Commissioner of Police of the Metropolis v DSD a majority of the Supreme Court accepted that there could be a violation of the Article 3 right in cases of operational failure rather than failure to implement an adequate system of investigation.Footnote 73 While cases involving systemic failures do, therefore, give rise to difficulties in establishing causation, from the perspective of protecting human rights this may not be a rational limit. From the perspective of ensuring that the state body affords sufficient respect to the right to human dignity and maintains appropriate standards, the prospective risk-based approach to causation is coherent but should apply across the full spectrum of cases engaging that right rather than only those involving systemic failings, and need not attract an award of compensatory damages.
Given the difficulty of drawing coherent limits on the scope of the increase in risk approach within delict, whether one adopts the majority solution of viewing it as part of the existing flexible approach or the minority solution of explicitly developing an exceptional approach pursuant to their constitutional obligation, it would be preferable to maintain the integrity of delict and develop the award of constitutional damages instead. This would also enable courts to draw directly on the range of constitutional remedies which are able to more effectively tackle the conditions giving rise to a claim. This can be seen from another decision, highlighted by Keehn and Nevin,Footnote 74 concerning Pollsmoor prison, Sonke v Government of Republic of South Africa.Footnote 75 This case directly addressed the overcrowding and inhumane conditions of detention as a violation of detainees’ constitutional rights to health and human dignity. The remedy included an order requiring the Government to reduce overcrowding to 150% within six months (the population of the prison was previously around 225–250%), and to develop a comprehensive plan for addressing the deficiencies in the prison conditions. This remedy is better tailored to promoting the prisoners’ right to dignity than an award of damages in negligence.
One challenge, however, is that in light of the obligation to develop the common law, the objectives of delictual and constitutional awards are not so explicitly divergent as the separation between negligence and HRA liability in English law. In South Africa, Price explains that “these liability regimes potentially overlap” so that state bodies “may wrong individuals by acting unlawfully in these different ways, and therefore could in principle be held liable to pay damages on various grounds”.Footnote 76 That said, he goes on to explain that “an award of constitutional damages is correctly regarded as a subsidiary remedy, which should not be made where adequate statutory or common-law remedies are available”.Footnote 77 This stands in contrast to the position in English law where a claimant can be awarded damages under the HRA for breach of a Convention right in addition to recovering damages from other sources.Footnote 78 This is significant because in allowing for concurrent liability in negligence and the HRA, English law recognises that the two actions involve the interference with different kinds of rights, and that an adequate remedy may require an award of damages for both.
C. The Nature of the Rights in Negligence and the HRA
In Van Colle, Lord Brown explained that
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… It is also why section 8(3) of the [HRA] provides that no damages are to be awarded unless necessary for just satisfaction.Footnote 79
Lord Kerr elaborated on that in the Supreme Court decision in Commissioner of Police of the Metropolis v DSD which concerned the investigative obligation on the police under Article 3Footnote 80:
Laws L.J. said in para 68 of his judgment in the Court of Appeal, that the inquiry into compliance with the article 3 duty is “first and foremost concerned, not with the effect on the claimant, but with the overall nature of the investigative steps to be taken by the State”. I agree with that. The award of compensation is geared principally to the upholding of standards concerning the discharge of the state's duty to conduct proper investigations into criminal conduct which falls foul of article 3.Footnote 81
This meant that it was appropriate to award damages for the breach of Article 3 in addition to damages that were received from the perpetrator of the attacks on the claimants and from the CICA.Footnote 82 The vindicatory imperative underpins Varuhas's argument that damages awarded by UK courts in HRA claims should be assessed on the same principles as torts actionable per se where, for example, courts are willing to presume that certain losses, such as injury to feelings and distress, have been suffered once liability is established.Footnote 83 Mowbray also notes that the ECtHR has expressed a “hint, albeit limited in explanation, of the moral foundations underpinning its equitable conception of just satisfaction” in declining to make an award of damages to the victims in McCann and others v UK. Footnote 84 In that case, the court did not consider it “appropriate” to award damages where the victims had suffered a violation of their right to life in being killed by the police, but had been intending to plant a bomb at the time.Footnote 85 The aim of upholding standards rather than allocating responsibility for outcomes is thus reflected in numerous doctrines; causation is no exception.
The difference in aims was echoed by Lord Hughes in the same decision: “in substance, the Convention-based duty is not aimed at compensation but at upholding and vindicating minimum human rights standards. It is, substantially, to insist on performance of a public duty”.Footnote 86 But while he agreed with the majority on the outcome of the case in DSD, he did not agree that liability under Article 3 should be based on operational as opposed to systemic failures since this would impose an operational duty on the police that courts have strongly resisted imposing in negligence law: “one cannot both uphold the distinction and effectively eliminate it by employing a Convention claim to serve substantially the same purpose as an action in tort”.Footnote 87 Rather than drawing on human rights as a reason to develop the common law of negligence, this argument draws on the limits of negligence as a reason to constrain the development of human rights law. Lord Hughes argued:
True it is that the limitation period differs, but this will not remove the disadvantages to policing which were identified in the English [negligence] cases. It may be that there is a more relaxed approach to causation in a Convention-based claim, but that if anything only increases the prospect of such a claim becoming a substitute for a claim in tort. There is no doubt some difference of approach to the calculation of compensation, but the present case is a good illustration of the marginal, if not imperceptible, nature of the distinction in outcome.Footnote 88
As we have seen, the impact of the relaxed approach to causation is important but should not be overstated since, as noted in Sarjantson, compensatory damages will not be available where it is not proved that the violation of the right made a difference to the outcome. The HRA claim is unlikely to become a substitute for a negligence action if compensatory damages remain unavailable. While the difference in the aims and theoretical underpinnings of human rights and negligence actions has resulted in different doctrinal requirements for establishing liability, Lord Hughes’ concern is with the practical impact of liability which, in his view, raises the same public policy concerns about interfering with policing regardless of the legal basis of the duty.
The recent decisions in Michael Footnote 89 and Robinson Footnote 90 are important in this regard because the legal reasoning for establishing or denying a duty of care on the part of the police has shifted away from public policy arguments and more firmly onto a principled basis. The decision in Michael applies to the police the same principles that apply to other defendants in respect of omissions liability, rather than relying on public policy arguments as to why the police specifically should not be under a duty of care to prevent harm from being inflicted by a third party. In a Diceyan approach, the police are treated as a defendant like any other. If anything, the status of the police as a public authority arguably weighs in favour of treating them differently by imposing a duty on them in respect of omissions. Steel and Tofaris suggest:
it must be questioned how valuable the freedom of a public authority negligently to fail to take steps to assist an identified individual at serious risk of physical injury is. A private individual's freedom arguably has intrinsic value in so far as her having freedom to do various things contributes to her having an autonomous life. By contrast, the value of the state's freedom is purely instrumental: the state's freedom is valuable only in so far as it contributes to the fulfilment of its proper functions.Footnote 91
The obligations imposed under the ECHR arise specifically because the defendant is the state, so an argument in favour of limiting those obligations because the state is not subject to them in negligence, does not get us far when negligence law treats the state as a defendant like any other. In negligence the question is “why” impose an obligation on the police, in human rights law the question is “why not”. It does not make sense to “normalise” state responsibility, by aligning it with the negligence-based obligations of individuals, in an area of law that is premised on the exceptional status of the state.
McBride suggests that an explanation for the duties that may be imposed in human rights law but not in negligence lies in what he calls a “personalist view” based on the idea that “everyone should count for something with the state”.Footnote 92 He explains:
on the personalist view we do have a right, in the case where a public body knows we are in danger, that it try to save us from harm. This duty to try is a very different duty from a duty of care, which requires the public body not just to try to save us from harm but to make a good fist of the attempt.Footnote 93
This duty to try echoes the objective of human rights law, identified above, to secure respect for life and other rights and not solely to protect citizens from harm. Risk-based liability is essential to securing this respect, while the causation requirement in negligence reflects the interpersonal responsibility at stake there. It is not necessary to agree/disagree with Lord Hughes on the boundaries of this duty of means to accept both that the rights protected by each area of law, as well as the objectives of each area of law, differ in a way that not only justifies but requires a different approach to causation (as well as to other elements of liability not explored in this paper).
In 2003, Fairgrieve stated that the award of HRA damages “is perceived as a residual remedy”Footnote 94; this is surely no longer true. The actions complement each other by developing distinct principles that reflect the distinct rights at stake, and enabling concurrent protection of both rights where appropriate. The causal principles are a key point of difference, and we have seen that applying negligence-based causal principles in a human rights action would artificially truncate liability and fail to adequately protect the victim's human right, while encroachment of the risk-based human rights principles into negligence law is inconsistent with the interpersonal right and difficult to coherently circumscribe. Recognition of the distinct causes of action reflects the richness of the law rather than simply a failure of negligence law to develop in line with human rights law or vice versa.
IV. The ECHR Approach to Causation: Gaps and Potential Developments
Having identified the difference of approach to factual causation in negligence and the positive operational duty under Articles 2 and 3 of the ECHR, and explored the underlying reasons for this, the final section of the paper turns its focus more squarely to the jurisprudence of the ECtHR. It draws out a more fine-grained picture of the ECHR approach to causation. This will lead both to identification of issues that remain to be clarified and to insights into areas where human rights actions could be developed to protect rights more effectively than is possible in negligence.
A. Article 8 and Risks to Health
The analysis thus far has focused on the protection against risks to life and health under Articles 2 and 3. Article 8, which protects the right to private and family life, home and correspondence, also warrants comparison with negligence law because the ECtHR has developed this right to impose obligations on the state to provide individuals with information that would enable them to assess risks to their health. Although this paper criticises the court for a lack of clarity as to causal requirements, it will be suggested that it might be beneficial for the English courts to develop the Article 8 right more fully in the healthcare context.
The Strasbourg court recently considered the Article 8 right in the context of asbestos-related injury in Brincat and others v Malta.Footnote 95 The applicants had been exposed to asbestos as employees of Malta Drydocks Corporation (MDC), a state-owned enterprise. Mr. Dyer was not suffering any asbestos-related disease but was at an increased risk of such diseases developing in the future as a result of his exposure. The remaining applicants were suffering from pleural plaques and the increased risk of developing mesothelioma in future as a result of their asbestos exposure, and some were suffering from asbestosis, with Mr. Abela being confined to bed for years as a result of acute respiratory problems. The court found that the state had violated the applicants’ Article 8 right to private and family life by failing to provide access to essential information enabling them to assess the risk to their health and lives.
Once the violation was established, causation of the injury grounding the claim for damages for pecuniary and non-pecuniary loss was addressed fleetingly: “The Court has accepted the link between the medical conditions affecting the relevant applicants and their exposure to asbestos during the time they worked at MDC, and it thus discerns a causal link between the violation found and some of their claims in respect of pecuniary damage”.Footnote 96 Ultimately these applicants failed to substantiate their claims for pecuniary loss but the court considered that the mere finding of a violation of the Article 8 right was not a sufficient remedy and awarded non-pecuniary damages of 9,000 EUR to each applicant, apart from Mr. Dyer who received 1,000 EUR and Mr. Abela who received 12,000 EUR. The lack of explanation surrounding causation is characteristic of ECtHR jurisprudence but surprising to a tort lawyer.Footnote 97 The Government had argued, based on a 2009 factsheet produced by the National Cancer Institute, “Asbestos Exposure and Cancer Risk”, that the risks of developing asbestos-related disease depend on a variety of factors, including smoking,Footnote 98 and while the court noted that most of the applicants were non-smokers,Footnote 99 that implies that some of them were smokers but the court did not differentiate between the causal contribution made by asbestos to each. Sulyok therefore suggests that the court found causation to be established “not primarily on the basis of the expert evidence but on account of widely held views on the toxic nature of asbestos”.Footnote 100
One concern with the decision in Brincat and doubtless many other cases, is that the ECtHR does not specify what the non-pecuniary damages relate to. This means it is unclear whether the purpose of the award was solely to vindicate the interference with the right to private and family life, or whether it also compensated physical damage and other forms of non-pecuniary loss such as mental distress. It would be helpful for the court to distinguish the vindicatory and compensatory portions of the award. Each applicant's right to private life was violated by the failure to provide information about the risks to which they were exposed, so the vindicatory portion should be consistent across the applicants. So while it may at first seem surprising that Mr. Dyer, who was not suffering any asbestos-related illness, was awarded non-pecuniary damages, the damages correspond to the direct adverse effect on his private life. Damages are higher for those suffering physical illness, but the exposure and lack of information themselves interfere with the applicant's private life to a degree that requires non-pecuniary damages in order to vindicate that right. The real concern then is that the other applicants both received 9,000 EUR when the information about their conditions early in the judgment suggests that while they both suffered pleural plaques, they were not both suffering asbestosis. The House of Lords decision in Rothwell was clear that pleural plaques do not constitute damage for the purposes of negligence liability since they are asymptomatic and in most cases remain so, while asbestosis produces deleterious symptoms and constitutes damage.Footnote 101 Given that the court in Brincat has sought to differentiate between the applicants with the most and least harm, one might expect a more fine-grained differentiation between levels of award of damages. Varuhas has argued for a tort-based approach to HRA damages, drawing on the proposition that “our courts must be free to try to give a lead to Europe as well as to be led”,Footnote 102 and this seems like an area that would benefit from the more detailed assessments of quantum that take place in tort claims.
In Guerra v Italy,Footnote 103 the state was found to have violated the applicants’ Article 8 right where they lived near to a factory emitting toxic substances and there was a failure to provide them with information about the risks, both the ongoing risks and what to do in the event of an accident. The factory itself was not operated by the state so the basis of liability was a positive duty to take steps in relation to a risk created by a private individual, in other words an Article 8 equivalent of the positive duty arising under Article 2 in Osman. The court stated that
severe environmental pollution may affect individuals’ well being and prevent them from enjoying their homes in such a way as to affect their private and family life adversely. In the instant case the applicants waited, right up until the production of fertilisers ceased in 1994, for essential information that would have enabled them to assess the risks they and their families might run if they continued to live at Manfredonia, a town particularly exposed to danger in the event of an accident at the factory.Footnote 104
The applicants had not shown that they suffered pecuniary losses but were awarded 10,000,000 lire each for the non-pecuniary damage that they had “undoubtedly”Footnote 105 suffered (about £5,000).
Risk has a more limited role in establishing liability here than in Article 2 since the question is not whether the state failed to reduce the risk but merely, given the existence of the risk, whether the state failed to take steps to put information about it into the victim's hands. Wright notes that “it was not necessary for the victims to prove causation by showing that they would have moved away if they had known of the risk”.Footnote 106 Where the state has failed to provide information about physical risk, that risk must exist and must be a risk to the applicant in order to engage their right to private and family life. In Guerra the factors considered relevant by the court focus on the proximity between the factory and the town: the short distance (the applicants lived around 1km from the factory), the ongoing emission of toxic materials, that there had been an accident in the past which resulted in 150 people being hospitalised, and that the geographical position of the factory meant that emissions were often channelled towards the applicants’ town. The court explained that “the direct effect of the toxic emissions on the applicants’ right to respect for their private and family life means that Article 8 is applicable”.Footnote 107 In other words, Article 8 will be engaged when there is a direct risk to the applicant, and will be violated when the state has failed to provide information enabling the applicants to evaluate that risk for themselves, or even where it has failed to undertake a “genuine and procedurally fair environmental impact assessment” of an activity that presents risks to the environment which might impact on the personal lives of those living nearby.Footnote 108 The positive obligation on the state does not extend as far as taking steps to mitigate the risk, but simply involves the provision of information about the risk to those potentially affected by it. There is certainly no requirement to show that the applicants would have taken steps to avoid the risk if they had received adequate information, and again this highlights that the obligations created by the ECHR are obligations “of means, not ends”. What matters is putting the information in the applicants’ hands, not what result that ultimately leads to. This highlights that the concern of Article 8 is similarly wider than risks to life or to private life, and requires states to act in a manner that shows adequate respect for the right to private life.
This leads Wright to suggest an HRA action might be an alternative source of a remedy in cases of medical non-disclosure of risk where claimants face difficulties of proof of causation in negligence.Footnote 109 The decision in Vilnes v Norway lends weight to this suggestion.Footnote 110 The applicants in Vilnes were divers suffering health problems associated with the conditions under which they dived, and the court found a violation of Article 8 in the state's failure to ensure that their employers were transparent about the diving tables they used, allowing them instead to go undisclosed in the interests of competition between the companies conducting the dives. Transparency about the diving tables and their concerns for the divers’ health “constituted essential information that they needed to be able to assess the risk to their health and to give informed consent to the risks involved”.Footnote 111 This focus on the provision of information required for informed consent may assist claimants struggling to prove causation in negligence in cases of medical non-disclosure of risk. An Article 8 claim would relieve them from the need to prove that the failure to disclose the risk actually affected their decision to consent to the procedure, instead needing to show simply that the information should have been disclosed in order for them to give informed consent. This would go even further than the exceptional approach adopted in Chester v Afshar.Footnote 112 The defendant doctor in that case had negligently failed to disclose to the claimant patient a risk inherent in the proposed operation, which materialised when the operation was performed, resulting in injury. The claimant was unable to establish causation on orthodox principles because she may still have undergone the operation if she had been properly informed of the risks; although this would have been on a different date having sought a second opinion, the inherent risks would have been the same.Footnote 113 The House of Lords exceptionally allowed her claim to succeed, and Lord Steyn justified this on the basis that the patient's “right of autonomy and dignity can and ought to be vindicated by a narrow and modest departure from traditional causation principles”.Footnote 114 While this case might suggest the English courts see vindication of rights as an objective of negligence, at least in a healthcare context, it has not been applied more widely and recent decisions indicate a concern to tightly circumscribe Chester, insisting both that the patient suffer physical harm and that the harm be intimately connected to the duty to warn.Footnote 115 These requirements, which are consistent with wider negligence principles, act as a barrier to achieving vindication of patient rights. An Article 8 action would be a more suitable vehicle for achieving this aim, since causation of damage is not an element of liability. English courts ought, however, to continue to apply causal principles with the same rigour as in tort law if the victim seeks an award of damages in respect of physical harm that ensues so that this does not become a back-door route to achieving compensation. From a patient's perspective it may be difficult to explain the result that NHS patients would have a remedy where recipients of private medical care would not. The core concern is the patient's right to make an informed, autonomous choice, regardless of whether the defendant healthcare provider is the state. From that perspective an Article 8 claim draws an arbitrary distinction while a negligence action is an unsuitable vehicle for vindicating that right because of the causation and damage requirements. In any case it is unclear that the ECtHR would be inclined to adopt the Guerra approach in a case involving medical non-disclosure of risk since, as the final section now explores, the court's approach to the Article 2 right suggests a greater reluctance to impose human rights obligations in the healthcare context.
B. Article 2 and Healthcare Provision
In Lopes de Sousa Fernandes v Portugal the Grand Chamber of the ECtHR stated that errors in diagnosis leading to delay in treatment or delay in performing a medical intervention are not on a par with the denial of healthcare.Footnote 116 Liability may arise where there is a delay in the provision of any healthcare service, but where healthcare provision is made and performed negligently, there will not be liability unless there is a systemic deficiency in the regulatory framework. Throughout the case law on denial of healthcare provision there is a lack of clarity as to what the causal requirements are. It is also unclear whether the observations as to causation underpin the court's finding of liability or their decision as to the availability of damages.
In Anguelova v Bulgaria, where the applicant's son died in police custody several hours after his arrest from a skull fracture suffered prior to arrest, in recognising liability under Article 2 in respect of the delayed provision of medical care, the court stated that the medical report had found that the delay “contributed in a decisive manner to the fatal outcome”.Footnote 117 In contrast, in the more recent decision in Mustafayev v Azerbaijan concerning failure to provide medical care to a prisoner, the state objected that there was no link between the delayed transfer to hospital and the death of the applicant's son but the court responded that “the object of its examination is whether or not the domestic authorities fulfilled their duty to safeguard the life of the applicant's son by providing him with proper medical treatment in a timely manner”.Footnote 118 This difference was not explained by the court. Perhaps it is the case that where causation is present it will be noted, but is not required for liability to be established, although this does not explain why both cases resulted in similar awards of non-pecuniary damages.Footnote 119 The decision in Mustafayev suggests that, as in other contexts, the focus in cases of failure to provide medical care is on respect for the right to life through provision of appropriate medical treatment, rather than specifically whether the lack of provision actually resulted in death. It is also possible that the court's approach to causation is more relaxed in cases arising from failure to provide healthcare to a person in custody. White and Ovey observe that
the protection afforded by Article 2 would be of no value if a State could avoid international sanction by concealing the evidence of killings caused by its agents. Where an individual is known to have been taken into custody and subsequently disappears or is found dead, therefore, it is logical that a heavy burden should fall on the State to establish an innocent explanation.Footnote 120
In cases arising squarely in a healthcare context, where medical treatment is provided but is alleged to involve some systemic deficiency, it appears that there is a causal requirement but again this area is characterised by a lack of consistency and clarity. In Lopes de Sousa Fernandes, the court used the vague terminology of “a link”: “there must be a link between the dysfunction complained of and the harm which the patient sustained”.Footnote 121 The earlier Chamber decision had held that “Without wishing to speculate on the applicant's husband's prospects of survival if his meningitis had been diagnosed earlier” the hospital's negligence had deprived the applicant's husband of access to appropriate emergency care so found a breach of Article 2.Footnote 122 A “link” appears to be a very loose concept. Yet the Grand Chamber later went on to state that it had not been established that there was a structural or systemic failing, about which the authorities knew or ought to have known, and “and that such a deficiency contributed decisively to the death of the applicant's husband”.Footnote 123 This suggests that risk creation is not sufficient in this context, and that an actual causal link to the death is required for liability to arise.
Yet the meaning of a “causal link” in this context requires further elucidation. In Aydoğdu v Turkey, the court explicitly found that a causal link existed between the systemic failings and the death of the applicant's baby,Footnote 124 on the basis that the authorities had not done what could reasonably be expected of them to protect the baby's life from a real risk, whilst still stating that it was not necessary to speculate on the baby's chances of survival given proper treatment.Footnote 125 Similarly in Asiye Genç v Turkey, where a lack of neonatal care facilities meant that the applicant's baby was sent back and forth between hospitals with each refusing to admit him, the court held that
while it is inappropriate to speculate as to the baby's chances of survival had he received immediate treatment, the court notes that, in spite of the above risk, the staff members in question did not take the necessary measures to ensure that the patient would be properly cared for at the KTÜ Farabi public hospital before deciding to transfer him there.Footnote 126
This contrasts with negligence law where the patient's chances of survival are central to determining whether, on the balance of probabilities, she or he would have died even with proper treatment.Footnote 127
It appears that in cases involving systemic failings in the provision of healthcare, there is a causal requirement, in that the court generally insists that the systemic failings have “caused” the victim's death. Yet it is clear from the fact that the court generally declines to speculate as to the victim's chances of survival given proper treatment, that this is not the same as the causation requirement in negligence. Instead, it seems to indicate that in the healthcare context death is required. In contrast to, for example Makaratzis,Footnote 128 it is insufficient that the state's failings merely expose the victim to a risk to their life which they ultimately survive. But this “causal” requirement will be satisfied by proof that the state failed to take steps that could reasonably be expected of it to protect the victim from a risk to their life, and the fact that the victim died. This amounts to something akin to a Fairchild approach in that it must be shown that the defendant exposed the victim to a risk to their life and that harm materialised which was within the scope of that risk. By avoiding speculation as to the victim's chances of survival the court avoids suggestions that it is the loss of the chance rather than the death that is the relevant outcome being compensated by any award of damages. A possible explanation of the opacity as to causation lies in the earlier discussion in relation to the South African decision in Lee. There we saw that where the fault of the state consists of systemic failings, proof of causation is problematic because the state enjoys a margin of appreciation as to how to fulfil its obligations so it is not appropriate to posit a single alternative system for the purposes of the counterfactual analysis of causation. This may explain the court's refusal to speculate as to the victim's chance of survival in the healthcare cases. If this is the case, it would be preferable for the court to avoid causal language such as “contributed decisively to the death” and state explicitly that it suffices that the state's systemic failings increased the risk of death and that death within the scope of that risk occurred. Yet it is difficult to reconcile this damage requirement with the vindicatory objective of human rights liability.
V. Conclusion
The aims of this paper were twofold: to establish a clearer picture of what the ECHR/HRA approach to causation entails and how it differs from that in negligence, and to consider whether the differing objectives of liability that are oft-cited as a rationale for any divergences really are capable of explaining the different causal requirements. On the first point we now have a clearer picture of how causation is addressed in human rights claims and how this leads to different results from factually similar negligence claims. There is, however, a need for greater clarity from the ECtHR as to the causal requirements, if any, in the healthcare context and how this fulfils the aims of human rights liability. In particular, the court ought to clarify whether any causal requirements are relevant to liability or merely to damages, and when awarding non-pecuniary damages should delineate between those that vindicate the right that has been violated and those that compensate for consequential loss.
From this base of understanding of the law it has been possible to appreciate the function of the causation inquiry within the two areas of law and see that the different approaches can be explained in a coherent way by reference to the different rights protected and the different aims of liability. This has implications both for human rights law and negligence law. Where the ECtHR appears to have adopted a damage requirement for Article 2 actions in a healthcare context, this risks frustrating the objective of Article 2 which is to secure respect for life in addition to protecting the sanctity of life. Greater clarity is needed as to whether damage is strictly required for liability to arise in that context and how this is to be justified. Looking more widely, in the context of the negligence liability of public authorities, in particular the police, the insights acquired as to the function of causation within the different rights add weight to the case for separate development of negligence and human rights liability. This has been the clear position of the courts, and Nolan, among others, has argued that convergence between negligence and human rights law is not only unnecessary, but that it is also not desirable.Footnote 129 While we have seen that Wright is critical of “formalistic reasoning”, making the case that negligence actions against the police do raise human rights issues so that negligence principles should be developed to reflect the human rights aspects of the claims,Footnote 130 this paper has highlighted that elements such as the causation requirement are not simply formalistic but reflect the objective and rights at stake. Nolan's view is that the argument for convergence is based on the false assumption that negligence and human rights law serve the same purpose.Footnote 131 As their purposes actually differ, convergence would “weaken [the] structural underpinnings [of negligence] and cut across its core principles” so separate development is necessary to preserve the coherence of negligence law.Footnote 132 In elucidating not only the different approaches to causation but also how those approaches fulfil the different functions of the causal requirements which further reflect the objectives and rights in question, this paper adds weight to the case for separate development. If the duty of care owed by police in negligence were to be extended to cover situations like Osman and Sarjantson, this would result in incoherence. One possibility is that the orthodox rules of causation would continue to apply but would constitute an artificial barrier to protecting the right to life. The alternative is the relaxation of the rules of causation, as has occurred in South Africa in Lee, and courts would then be faced with the task of defining when this exceptional approach should apply. Instead, separate development reflects the richness of English law and maintains the coherence of both areas of law while allowing the distinct aims of each to be achieved, and causation is a key aspect of this.
