The Manchester Arena bombing – Limitation periods and the boundaries of the Osman test

In Various Claimants v Security Service & others (2024) UKIPTrib 6 the Investigatory Powers Tribunal (Lord Justice Singh and Mr Justice Farbey) considered an application to extend time under section 7(5) of the Human Rights Act 1998 (mirrored in section 67(5) of the Regulation of Investigatory Powers Act 2000).

The claims arose out of the tragic events of 22^nd May 2017 when 22 innocent people were killed as a result of a suicide bombing at an Ariana Grande concert at the Manchester Arena. One of the bombers died and the other is serving a life sentence.

The claim was brought by 300 claimants affected by the bombings. Three lead claimants were identified. The first was the estate of a 17-year-old girl killed in the explosion. The second was a 14-year-old girl who sustained life-changing injuries. The third was the mother of a girl who was informed that her daughter had died and who suffered psychiatric injury as a result of witnessing the immediate aftermath of the atrocity.

The Respondents were government organisations responsible for monitoring and acting upon, amongst other matters, terrorist threats to the country.

In March 2023, Sir John Saunders concluded the Manchester Arena Inquiry and produced a detailed report. One of the key findings was as follows:

There was a significant missed opportunity to take action that might have prevented the Attack. It is not possible to reach any conclusion on the balance of probabilities or to any other evidential standard as to whether the Attack would have been prevented. However, there was a realistic possibility that actionable intelligence could have been obtained which might have led to actions preventing the Attack. The reasons for this significant missed opportunity included a failure by a Security Service Officer to act swiftly enough.

The report led the then Director General of MI5, Ken McCallum to issue a public apology in which he stated:

I deeply regret that such intelligence was not obtained. Gathering covert intelligence is difficult – but had we managed to seize the slim chance we had, those impacted might not have experienced such appalling loss and trauma. I am profoundly sorry that MI5 did not prevent the attack. …

MI5 exists to stop atrocities. To all those whose lives were forever changed on that awful night: I am so sorry that MI5 did not prevent the attack at the Manchester arena.

During the course of the coronial inquest into the deaths resulting from the bombing, Counsel to the Inquest had produced an explanatory note (dated 16^th January 2019) in which he expressed the provisional view that Article 2 of the European Convention was probably engaged. However, it was not until 29^th February 2024, some 5 years later, that the claims were issued in the tribunal. It was alleged that the Defendants had violated Articles 2 and/or 3 of the Convention and that the Claimants were each entitled to damages (as just satisfaction) together with their legal costs.

On 1^st August 2024, the Tribunal ordered a hearing to consider 2 preliminary issues:

• Whether the claims ought to have brought within the 1-year time limit under the HRA and the 2000 Act and, if not, whether time should be extended.

• Whether the test for a violation of articles 2 and 3 of the ECHR in this context is whether ‘the relevant authorities knew or ought to have known at the time of the existence of a real and immediate threat to life or of Article 3 ill-treatment from the criminal acts of a third party and failed to take measures within their powers, which judged reasonably, might have been expected to avoid that risk’?

Limitation

The Tribunal referred to guidance in earlier cases on the factors relevant to the discretion to extend time, but reminded itself that “ultimately the statutory question for the Tribunal is what would be “equitable having regard to all the circumstances”.

The Claimants contended that:

• No viable claim could have been brought until the Inquiry had published its report on 2^nd March 2023.

• There would be no prejudice to the Respondents because all relevant evidence had already been adduced before the inquiry and there would be a transcript of witness evidence made available. This would include closed evidence.

• Although there had been some delay between the publication of the report and the issue of the claims, the claimants had suffered trauma, including trauma induced by the publication of the report itself. Furthermore, parallel civil claims against other authorities had delayed matters.

• The subject matter was of supreme importance to the Claimants. Although there was a significant public interest in expedition, the sheer seriousness of the atrocity and the fact that it concerned past violation of human rights (as opposed to a public law claim to enforce an existing right) meant that expedition was less imperative.

The Respondents were formally ‘neutral’ on the issue of whether time should be extended but made submissions which all pointed against exercising the discretion:

• There was unjustifiable delay in bringing the claims.

• There would be some impact on the cogency of witness evidence.

• The Respondents’ conduct in respect of the proposed claim could not reasonably be criticised.

• It was highly unlikely that any further information would be brought to light by the Tribunal, given the comprehensive nature of the Inquiry.

• The continuation of the claims against the security services would divert their resources away from the core function of public protection.

• It was unlikely that any claimant could establish causation of pecuniary (i.e. financial) loss and therefore the value of any claims for non-pecuniary loss would be small. The findings of the inquiry, even taken at their highest, did not fulfil the ‘direct causal link’ required by the ECHR jurisprudence between the ‘missed opportunities’ and any losses alleged.

The Tribunal started by acknowledging the fundamental importance of the rights protected by Articles 2 and 3 and the horrendous impact of the terrorist attack on the Claimants and their families. However, it concluded that it would not be equitable to permit the claims to proceed.

The core of the tribunal’s reasoning focused on the Claimants’ procedural delays. Their solicitors had written to the security services on 9^th May 2018 referring to the limitation period under the HRA and asking for a ‘general amnesty’ under which time would be extended to a period of 3 years from the date of the incident. The Respondents replied, refusing the request and noting that if a claim was made outside the 1-year limit under the Act, they would carefully consider whether to raise a limitation argument at that stage. Some 2 years later, the Claimants’ solicitors wrote again, inviting the Respondents to agree a ‘standstill agreement’. Although an interim agreement was reached as a holding response, the parties did not agree to enter into a formal standstill agreement.

The Tribunal observed that whilst it was reasonable and understandable why the Claimants felt unable to issue even ‘holding claims’ before the publication of the report in March 2023, this only reinforced the need to act with particular expedition as soon as the report was published. In fact, instructions were not sent to leading counsel until 23^rd May 2023, counsel’s advice was not received until 13^th December 2023, and it was a further 3 months until the claim was brought. This represented a failure to give the issue of proceedings the priority it warranted.

The tribunal was also heavily influenced by the fact that the proceedings were unlikely to reveal any new evidence or serve any significant legal purpose. Referring to the observations of Lord Bingham in R (Greenfield) v Secretary of State for the Home Department (2005) 1 WLR 673, the primary purpose of claims under the HRA was to vindicate rights, not award compensation.  Although the inquiry had not formally included a declaration that the Claimants’ rights had been violated, it had led to a full public apology. A further declaration to the same effect, together with a modest award of damages, was insufficient reason to allow the claims to continue when balanced against the fact that it would inevitably divert resources from core responsibilities, including the ‘protection of the lives of people in this country by, for example, preventing future attacks’.

It is difficult to conceive of a more serious or tragic set of facts than those underlying this claim. The delay, if taken from the date of the public of the report in March 2023, was less than 12 months, there was no real evidential prejudice and respondents had adopted a neutral stance. These were all factors which might be expected, certainly in a claim to extend time under section 33 of the Limitation Act 1980, to weigh in favour of the Claimants. Yet the tribunal reached a clear conclusion that the discretion to extend time should not be exercised.

Two factors stand out.

The first is the criticism of the Claimants’ solicitors for failing to act with insufficient haste in light the Respondents’ refusal to agree to a limitation moratorium. The discretionary exercise under section 7(5) is not a punitive one, but the relative conduct of the parties will always be a relevant factor and will include the conduct of legal representatives. Since the issue of limitation had been identified between the parties at an early stage, it was not appropriate (to use the words of the trial judge endorsed by the Court of Appeal in RE v GE (2015) EWCA Civ 287 under section 33 of the Limitation Act 1980) to “put a cause of action onto a shelf with a view to taking it down again sometime later in the indeterminate future when you feel like using it.”. It may, in the circumstances, have been a more sensible approach to issue protective proceedings as soon as the March 2023 report was published and thereafter to request a stay of the claim to allow final investigations and case preparation to be completed (by analogy with section 1.6 of the Pre-Action Protocol for Personal Injury Claims).

The second factor is the tribunal’s reiteration that the primary purpose of claims under the Convention is the vindication of rights rather than provision of monetary compensation. This is particularly striking on the facts, given that some of the claimants were presumably seeking substantial awards of (at the very least) non-monetary damages for life-changing physical and psychiatric injuries. Although there may have been reason to doubt the claimants’ ability to recover associated financial loss flowing from the injury, the Courts have held that in claims involving the violation of bodily integrity and therefore personal injury, awards under the ECHR should not be lower than those made in tort: R (Alseran) v MOD (2019) QB 1251, paragraph 931.

In the final analysis, the Tribunal’s decision is perhaps best viewed as one firmly rooted in the ground of public policy. The policy which underlies the imposition of a 1-year time limit under the HRA in the first place is that “All such claims are, by definition, brought against public authorities, and there is no public interest in these being burdened by expensive, time-consuming and tardy claims brought years after the event” (Bedford v Bedfordshire County Council (2014) HRLR 33. This had particular resonance in a case where, notwithstanding the missed opportunities to prevent the Manchester bombing, the role of the Security Services in ensuring public protection remained as critical as ever.

The Osman Test

In light of the Tribunal’s finding on the limitation issue, it was not necessary for it to consider the application of the test for imposition of the operational duty under Article 2/3.

However, Counsel for the respondents made the interesting submission that, based on the Inquiry’s findings, the claim could not succeed without significantly extending the Osman test so that the concept of “knew or ought to have known at the time” encompassed “ought to have taken investigative action which might possibly have revealed additional information that could have led to the future identification of a real and immediate risk”.

Was this an accurate submission?

The starting point is to acknowledge that the Osman test clearly envisages that either actual or constructive knowledge is enough. Constructive knowledge is that which the public authority ought to have known if had acted reasonably.

Perhaps the leading authority on the meaning of ‘ought to have known’ is the Grand Chamber’s decision in Kurt v Austria (62903/15) which was concerned with the alleged failure of social services to prevent a father from killing his children following a long history of domestic abuse aimed at their mother. The Court dismissed the claim, finding that there was no real and immediate risk to the children (as distinct from the mother). However, in so doing it emphasised that risk-assessment, as a tool, was essential in identifying, and therefore acquiring knowledge of, the relevant risk. It followed that if a risk assessment was inadequate, the authority could not rely upon its ignorance of the risk if an adequate assessment would have identified it. In the context of domestic violence, factors which might be taken into account as part of the assessment included the perpetrator’s history of violent behaviour, any failure to comply with court injunctions, a trajectory of escalating behaviour, access to weapons and the victim’s pleas for assistance.

However, it seems that the submission made by the Security Services before the Tribunal was that there is a substantive difference between the situation where an ongoing assessment would have revealed an immediate (meaning ‘present and continuing’: Rabone v Pennine NHS Trust (2012) AC 72) risk, and one which merely triggers a potential train of enquiry which may or may not have identified this risk at an indeterminate point in the future. This would be stretching the definition of constructive knowledge because there are too many steps in the process between the initial inquiry and learning of the risk to conclude that the authority should have known about it, even if it was possible.

The closest that the European Courts have come to recognising the limitations of the ‘constructive’ knowledge element of the Osman tests are found in a pair of separate concurring or partly concurring judgments, one in Kurt itself and the other in the earlier case of Talpis v Italy (41237/14).

In Kurt, 8 judges delivered a single concurring opinion. Addressing the question of risk assessment, they stated as follows:

Even where an assessment of risks is carried out in advance, the prediction of concrete acts remains hard and uncertain. A judicial body such as the Court, when reviewing events that have already taken place, should not underestimate the difficulties involved in predicting the concrete nature, targets and timing of violent behaviour. Nor should the Court underestimate in this context the challenges facing the domestic authorities in the determination of the appropriate response in various concrete situations. In some circumstances, a protective measure designed to reduce the risks for potential victims might become a provoking factor for the perpetrator. Even with a high level of professional experience and care, to accurately predict what might happen, when and where, and to know how to ensure prevention with the means available, will at best be difficult. Judging a course of action, or its timing, is much easier with the benefit of hindsight than in the situation prevailing at the time of the original decision-making.

In Talpis, another domestic violence case, Judge Spano partially dissented from the majority, which had held that the authorities had failed, in breach of Article 2, to prevent a husband from murdering his son and assaulting his wife. It was alleged that the authorities had failed to act in response to a series of previous assaults. Judge Spano posed the follow specific question “can investigative passivity give rise to constructive knowledge?”. His answer, excising references to the particular facts in the case, was as follows:

…inactivity of the sort demonstrated in the present case, and the results thereof, do not of themselves create constructive knowledge such as to trigger an obligation under Article 2 (although it will usually, and in the present case does, give rise to an Article 3 violation in the domestic violence context). What is ultimately required is a set of facts rendering untenable the claim that the authorities did not know, or could not have known, of a real and immediate risk to life

…

Regardless of how the judgment frames it, the Osman test continues to apply in the same way here as in other contexts triggering the State’s Article 2 preventive obligation; the Court’s domestic violence case-law has continued to apply a strict Osman test without any alterations. Diluting the Osman standard, to take account of the nature of different types of fatal criminal offences between individuals, will simply impose an unrealistic burden on law enforcement authorities. Again, the law, even human rights law, has its limits.

…

Although it may be tempting to dilute legal concepts such as the Osman test when faced with heartrending facts and give solace to individuals in situations such as that of the applicant, there are reasons why the threshold under the Convention is set high, and, in my view, why it must continue to remain so.

The inevitable truth that ‘even human rights law has its limits’ lies at the heart of the Tribunal decision in the Manchester Arena case and is also clearly present in the other leading authorities in this area. It will be interesting to see whether further attempts are made to push the boundaries of the Osman test in the future.