GRAND CHAMBER

CASE OF ARMANI DA SILVA v. THE UNITED KINGDOM

(Application no. 5878/08 )

JUDGMENT

STRASBOURG

30 March 2016

This judgment is final.

In the case of Armani Da Silva v. the United Kingdom,

The European Court of Human Rights, sitting as a Grand Chamber composed of:

Guido Raimondi, President , Dean Spielmann, Işıl Karakaş, Josep Casadevall, Luis López Guerra, Mark Villiger, Päivi Hirvelä, George Nicolaou, Ledi Bianku, Nebojša Vučinić, Vincent A. De Gaetano, Linos-Alexandre Sicilianos, Paul Mahoney, Krzysztof Wojtyczek, Dmitry Dedov, Branko Lubarda, Yonko Grozev, judges , and Lawrence Early, Jurisconsult ,

Having deliberated in private on 10 June 2015 and 20 January 2016,

Delivers the following judgment, which was adopted on the last-mentioned date:

PROCEDURE

1 . The case originated in an application (no. 5878/08) against the United Kingdom of Great Britain and Northern Ireland lodged with the Court under Article 34 of the Convention for the Protection of Human Rights and Fundamental Freedoms (“the Convention”) by a Brazilian national, Ms Patricia Armani Da Silva (“the applicant”), on 21 January 2008.

3 . The applicant complained that the decision not to prosecute any individuals following the fatal shooting of her cousin by police officers was in breach of the procedural aspect of Article 2 of the Convention, which required the authorities to conduct an effective investigation capable of leading to the establishment of the facts, a determination of whether the force used was or was not justified in the circumstances and of identifying and – if appropriate – punishing those responsible.

4 . On 28 September 2010 notice of the application was given to the Government.

5 . On 9 December 2014 a Chamber of the Fourth Section composed of Ineta Ziemele, Päivi Hirvelä, George Nicolaou, Ledi Bianku, Zdravka Kalaydjieva, Paul Mahoney and Krzysztof Wojtyczek, judges, and Françoise Elens-Passos , Section Registrar, relinquished jurisdiction in favour of the Grand Chamber, neither of the parties having objected to relinquishment (Article 30 of the Convention and Rule 72 of the Rules of Court).

6 . The composition of the Grand Chamber was determined according to the provisions of Article 26 §§ 4 and 5 of the Convention and Rule 24.

7 . The applicant and the Government each filed a memorial on the admissibility and merits of the application.

8 . In addition, third-party comments were received from the Equality and Human Rights Commission, which had been granted leave to intervene in the written procedure (Article 36 § 2 of the Convention and Rule 44 § 3).

9 . A hearing took place in public in the Human Rights Building, Strasbourg, on 10 June 2015 (Rule 59 § 3).

10 . There appeared before the Court:

(a) for the Government Mr P. McKell , Agent , Ms C. Montgomery QC, Counsel , Mr J. Edwards , Adviser ;

(b) for the applicant Mr H. Southey QC , Ms H. Hill QC , Counsel , Mr A. Straw , Ms H. Wistrich , Ms M. Willis Stewart , Advisers .

11 . The Court heard addresses by Ms Montgomery QC and Mr Southey QC as well as their replies to questions put by Judges Villiger, Pinto de Albuquerque (substitute judge), López Guerra and Spielmann.

THE FACTS

I. THE CIRCUMSTANCES OF THE CASE

A. Background

13 . On 7 July 2005 four suicide bombers detonated explosions on the London transport network. Three of the suicide bombers were on underground trains and one was on a bus. Fifty-six people, including the four suicide bombers, were killed in the attack and many more were injured.

14 . The Metropolitan Police Service (“the MPS”) initiated a major police investigation to establish the identities of persons involved in or otherwise connected with the explosions. Available intelligence indicated that terrorists were actively planning a further attack within a matter of days, and the threat level posed to the United Kingdom from international terrorism was raised from Level 3 to Level 1.

15 . On 21 July 2005, precisely two weeks after the first bombings, four explosive devices were discovered in rucksacks left on three underground trains and on one bus. As it was feared that the failed bombers would regroup the following morning and attempt to detonate further explosions, the MPS immediately launched an operation to find them (“Operation THESEUS 2”). This operation was led by Police Commander John McDowall as Gold Commander.

16 . At 4.20 a.m. on 22 July 2005, Commander McDowall was informed that intelligence had identified Hussain Osman as a suspect in the failed bombings of 21 July. Both Mr Osman and another suspect were thought to be living in an apartment at 21 Scotia Road, London.

B. Operation THESEUS 2

1. Commander McDowall ’ s strategy

17 . At 4.38 a.m. on 22 July 2005, Commander McDowall decided to mount surveillance operations at both Scotia Road and another London address. The overall aim of the operation at 21 Scotia Road was to establish whether the suspects were present in the apartment and to arrest them safely if they came out. Commander McDowall ’ s strategy for this operation was not recorded; however, it would appear to have been to oversee the premises at Scotia Road through covert surveillance, to follow persons leaving the premises until it was felt safe to challenge them, and then to stop them. In order to implement this strategy, his plan was that a surveillance team from SO12 (Special Branch) should be in attendance at Scotia Road. They were to be supported by a unit from SO19, the Specialist Crime & Operations branch of the MPS. The unit from SO19 consisted of highly trained Special Firearms Officers (“SFOs”) who were usually deployed on pre-planned operations. Although some surveillance officers were armed for their own protection and that of the public, their training did not enable them to be used as a resource to arrest armed suspects. SO19 would normally undertake this task, although armed officers from SO12 could be used for this purpose as a last resort.

18 . The Crown Prosecution Service (“the CPS”) Review Note later found that if Commander McDowall ’ s strategy had been followed (notably, had the team from SO19 been deployed in time to support the surveillance teams at Scotia Road), events would not have unfolded as they did.

2. The command structure

19 . Commander McDowall appointed Commander Cressida Dick as the Designated Senior Officer (“the DSO”) in charge, who was to be responsible for accomplishing the THESEUS 2 strategy safely. As such, she had responsibility for the operation at 21 Scotia Road on 22 July 2005. She was based in control room 1600, where she was supported by Trojan 80, an experienced SFO from SO19 who was acting as her tactical adviser.

20 . Detective Chief Inspector C (“DCI C”) was appointed as Silver Commander for the operation at Scotia Road. Although a Silver Commander would usually have ultimate responsibility for the management of an incident and deployment of firearms resources, on this occasion the DSO retained this responsibility and DCI C operated as the DSO ’ s ground commander. DCI C was supported by and accompanied on the ground by Trojan 84, who, like Trojan 80, was an experienced SFO from SO19 who was acting as a tactical adviser. Trojan 84 was in charge of the SFO team to be deployed and he was in direct contact with Trojan 80.

21 . Detective Superintendent Jon Boutcher (“DS Boutcher”), the Senior Investigating Officer for the investigation into the identity of the persons responsible for the bombings on 7 July 2005, was also appointed as a Silver Commander.

3. Implementation of Commander McDowall ’ s strategy

22 . At 5.00 a.m. on 22 July 2005, a surveillance team from SO12 was called out. No request was made at this stage for a unit from SO19.

23 . By 6.04 a.m. two surveillance teams from SO12 had been deployed to the Scotia Road address to oversee the premises and to follow anyone coming out of the apartments. 21 Scotia Road was accessed by the same doorway as 17 Scotia Road and the surveillance teams were stationed in an observation van which had a view of that doorway.

24 . SO13 (the Anti-Terrorist Branch of the MPS) deployed four officers to assist with any arrest and to gain intelligence. DS Boutcher was the link between the control room and SO13.

25 . At 6.50 a.m. Commander McDowall held a briefing during which the firearms strategy was outlined. Trojan 80 was present at the briefing together with the Silver Commanders for the surveillance operations at Scotia Road and the second London address. The DSO arrived at 7.15 a.m.; however, Commander McDowall spoke to her after the briefing to ensure she had all the information and assistance she needed.

26 . As they had not been called out earlier (see paragraph 22 above), SFOs from SO19 were allocated to the operation when they reported for duty. At 7.45 a.m. Trojan 84 briefed the SFOs. The briefing was not recorded, but he appears to have told the team that they “may be required to use unusual tactics because of the situation they were in and that they should think about this”. When asked for clarification, Trojan 84 added that, in relation to a critical shot, the instruction would come directly from the DSO. However, if they were deployed to intercept a subject and there was an opportunity to challenge but the subject was non-compliant, a critical shot could be taken. The CPS later found that this briefing “stoked the [SFOs ’ ] fears that they would meet suicide bombers and that they may have to shoot such people”.

27 . Following the briefing, the unit from SO19 travelled to a police station at Nightingale Lane, which was approximately two miles from Scotia Road. They stopped off for petrol on the way. Upon arrival they received a further briefing from DCI C, which commenced at 8.50 a.m. The briefing was not recorded, but it appears that DCI C confirmed the terrorists had the capacity to attach a device to themselves that would be difficult to detect. He described the individuals involved in the bombings as being “deadly and determined” and “up for it”. The CPS later criticised this briefing as unbalanced, as DCI C had failed to caution the SFOs that not everyone they would stop leaving Scotia Road would be a suicide bomber and that they should not overreact in the heat of the moment.

28 . The team from SO19 was not deployed on the ground until after 9.30 a.m.

4. Events leading to the death of Mr Jean Charles de Menezes

29 . Jean Charles de Menezes was a Brazilian national who lived at 17 Scotia Road. At 9.33 a.m. he left his apartment building through the common doorway in order to go to work. An officer in the surveillance van saw Mr de Menezes, described him and suggested “it would be worth someone else having a look”. However, as the unit from SO19 had not yet reached Scotia Road it was not possible to stop Mr de Menezes at this stage (as per the strategy outlined at paragraph 17 above). Instead, he was followed by the surveillance officers.

30 . On leaving Scotia Road, Mr de Menezes walked a short distance to a bus stop and got on a bus heading towards Brixton. The CCTV on the bus did not capture the entire journey due to vibrations but Mr de Menezes was recorded as being on the bus by 9.39 a.m. At this point the surveillance team described him as “a good possible likeness” to Hussain Osman. By 9.46 a.m. the description had changed to “not identical”.

31 . At 9.47 a.m. Mr de Menezes got off the bus. He was then seen using his mobile telephone before running back to the bus and reboarding.

32 . There are conflicting accounts of whether a positive identification was made of Mr de Menezes as the suspect at this stage. It appears from the Stockwell One Report of the Independent Police Complaints Commission (“the IPCC” – see paragraphs 45 - 71 below) that those on the ground had not been able to identify Mr de Menezes as Hussein Osman. The fact that the Surveillance Running Log refers to him at each entry as being an “U/I [unidentified] male” lends some support to this position. Nevertheless, those in control room 1600 appear to have believed that a positive identification of Hussein Osman had been made.

33 . At around the time that Mr de Menezes reboarded the bus, the unit from SO19 began to make its way towards Brixton. The SFO team leader later told the IPCC that he heard over the radio that “it was definitely our man and that he was nervous and twitchy”.

34 . At 9.59 a.m. the surveillance teams were asked to give a percentage indication of the likelihood that Mr de Menezes was the suspect, and they replied that it was “impossible [to do so] but thought that it was [the] suspect”.

35 . Mr de Menezes got off the bus at Stockwell and walked towards Stockwell underground station. There were several surveillance officers in the vicinity and their leader offered to stop Mr de Menezes before he entered the station. The DSO initially ordered that they perform the stop, having been informed that the unit from SO19 was not yet in a position to intervene. However, almost immediately thereafter she was informed that the unit was on hand. As a consequence, she countermanded her original order and instructed the SFOs to stop Mr de Menezes. By this time Mr de Menezes was already in the underground station. Trojan 84 relayed the order to the SFOs, informing them that “they want us to stop the subject getting on the tube”. The SFOs were told that they were going to Code Red, which meant that they were to have ultimate control of the situation and that an armed interception was imminent.

36 . The CCTV at the station shows Mr de Menezes entering the station at 10.03 a.m. wearing a thin denim jacket, a T-shirt and denim jeans, walking calmly and not carrying anything. He went down an escalator and onto a platform. There is no CCTV recording of the lower end of the escalator or of the platform: the relevant tapes, when seized by the MPS, were blank. The IPCC Stockwell One Report and the CPS later found that this was because a cable had been damaged during recent refurbishment work.

37 . At 10.05 a.m. a number of SFOs entered Stockwell underground station and ran down the escalators. At 10.06 a.m. they followed Mr de Menezes onto the platform. Eyewitness accounts as to what exactly happened next are conflicting, and some of the witnesses gave accounts which, it is now known, could not have been accurate. However, it would appear from the accounts quoted in the IPCC Stockwell One Report that Mr de Menezes went into the third coach of a stationary train and sat down; one of the surveillance officers shouted to the SFOs that Mr de Menezes was there; Mr de Menezes stood up, arms down; he was pushed back onto his seat and pinned down by two police officers; according to one witness his hand may have moved towards the left hand side of his trouser waistband; and two SFOs (Charlie 2 and Charlie 12) shot Mr de Menezes several times and killed him.

38 . Within days of the shooting, after it had become apparent that Mr de Menezes had not been involved in the attempted terror attacks on 21 July, the Commissioner of the Police of the Metropolis, the Prime Minister and the Secretary of State for Foreign and Commonwealth Affairs expressed their profound regret at his death. A representative of the MPS travelled to meet his family in Brazil and apologised directly to them on behalf of the police. An ex gratia payment was agreed upon to ensure that the family ’ s financial needs were met. They were encouraged to take independent legal advice from a solicitor in the United Kingdom, and they were advised that their legal costs in doing so would be met.

C. Post-death investigations

1. The initial investigations

39 . The Police Reform Act 2002 and the Police (Complaints and Misconduct) Regulations 2004 required a police shooting to be referred to the IPCC. However, following the shooting of Mr de Menezes, the Commissioner of the Police of the Metropolis wrote to the Home Office to inform it that he had decided not to refer the matter to the IPCC at that time.

41 . After the shooting, Charlie 2 and Charlie 12 had been taken to a police station. At 2.30 p.m., having taken legal advice, they indicated that they would not be making statements at that time. Instead, they made their statements together at approximately 2.00 p.m. the next day, after they had been told that Mr de Menezes was not connected to the attempted bombings. Some of the details they initially provided have since been either proved false or called into doubt. For example, they initially indicated that Mr de Menezes had been wearing a bulky jacket (CCTV footage showed him wearing a light denim jacket) and that the officers from SO19 had shouted “armed police” when they boarded the train (the IPCC considered such action to be “illogical” when confronting a possible suicide bomber).

42 . An officer from SO12 had seized the surveillance log at 12.35 p.m. on 22 July 2005. However, at 8.40 p.m. the same day, it was handed back to the officers from the unit. Around this time an amendment appears to have been made to an entry; the words “a split second view of his face. I believe it was [the suspect]” appear to have been altered to read “I believe it was NOT [the suspect]”.

43 . At 9.45 p.m. on 22 July 2005, the Anti-Terrorist Branch formally handed over control of the scene to the DPS as they were satisfied that Mr de Menezes was not connected to the attempted bombings.

44 . On 23 July 2005 a post-mortem examination took place and recorded the cause of death as “multiple gunshot wounds to the head. The cause of death is severe disruption to the brain”.

2. The first IPCC investigation and the IPCC Stockwell One Report

45 . On 25 July 2005 the DPS formally referred the investigation to the IPCC, whose investigation began on 27 July 2005 when the DPS provided it with the relevant material in its possession. Because of the seriousness of – and the public interest in – the matter, the IPCC determined that it would use its own staff to carry out the investigation. It was overseen by the Chair of the IPCC personally and the investigating team possessed all the powers and privileges of a police constable carrying out an investigation.

46 . The purpose of the investigation was to advise the CPS of any criminal offence that might have been committed; to provide it with the evidence necessary to come to a decision regarding any prosecution; to enable the “responsible authorities” of the officers concerned (the MPS and Metropolitan Police Authority, or “MPA”) to consider what disciplinary or other action they might need to take; to inform the Secretary of State for the Home Department of the circumstances of Mr de Menezes ’ death; and to assist the coroner in relation to any inquest.

47 . In particular, the investigation by the IPCC was to examine :

(a) the information that led to the surveillance of the apartments at Scotia Road;

(b) the command structure of the operation, including details of the numbers and types of specialist officers deployed and the tactics available to them;

(c) the qualification and training of those involved and their suitability to carry out their role;

(d) details of the briefing given to the officers involved and any description or photograph of any suspect made available;

(e) whether or not the operation was designated as a “KRATOS” operation (the national strategy for dealing with suspected suicide bombers which permitted the use of lethal force if absolutely necessary) and the policy, operational tactics and authority levels of “KRATOS”;

(f) the details of the mobile surveillance operation from Scotia Road to Stockwell underground station;

(g) the details of police action once Mr de Menezes had reached Stockwell underground station;

(h) whether or not the policy and operational authorities of “KRATOS” were followed and were effective; and

(i) whether “KRATOS” was compliant with Article 2 of the Convention.

48 . The IPCC was also to report on the actions and statements of the DPS from the time of the incident to the formal handover of the investigation to the IPCC to ensure that the IPCC investigation met its obligations under Article 2 of the Convention.

49 . During the course of the investigation, nearly 890 witness statements were taken from police, forensic experts and civilian witnesses, and more than 800 exhibits were collected. The family of Mr de Menezes, together with their legal representatives, were given regular detailed verbal briefings on the progress of the investigation and eventually on its conclusions.

50 . On 30 September 2005 the IPCC investigating team submitted a report to the IPCC indicating, inter alia , that certain officers might have committed criminal or disciplinary offences. The IPCC therefore wrote to the MPS and to the MPA regarding the officers concerned.

51 . On 19 January 2006 the IPCC Stockwell One Report was completed and submitted to the CPS. On 6 and 22 March 2006 the legal representatives of Mr de Menezes were briefed on the IPCC investigation and report. IPCC personnel also offered to travel to Brazil to brief any member of his family residing there. On 14 March 2006 the IPCC submitted its recommendations to the MPS, MPA, Her Majesty ’ s Inspector of Constabulary and to the Home Office.

(a) Summary of the Stockwell One Report ’ s conclusions

52 . The report considered all the witness statements and outlined in detail the events of 22 July 2005 and the investigative steps which followed the shooting. In particular it examined the actions and responsibility of the Commanders, their advisers and all the frontline SFOs and surveillance officers. While it accepted that the death of Mr de Menezes was not the result of any deliberate act designed to endanger the life of any innocent third party, it nevertheless concluded that

“20.01 [t]here can be no doubt that on the morning of 22 July 2005 a combination of circumstances between 0500 and 1006 led to the killing of an entirely innocent man”.

53 . With regard to this “combination of circumstances”, it identified a number of failings.

54 . Firstly, it criticised the briefings given by DCI C and Trojan 84.

“20.8 There is no doubt that the briefings provided by [DCI C and Trojan 84] included a comprehensive update on the intelligence including the links between 7 July and 21 July and the possibility that the firearms officers may have to confront one of the terrorists who had survived the suicide bombings the previous day. What the briefing for [SO19], and indeed the other teams, did not include was any rider about the circumstances in which the Operation KRATOS policy could be used. That policy was only one option available to the Metropolitan Police for dealing with suspected terrorists and suicide bombers. The [SO19] officers were not told that it should only be used as a matter of last resort when they were sure of the identity of the person in relation to whom the policy was to be applied. That should have been included in the briefing.”

55 . Secondly, it criticised the failure to implement Commander McDowall ’ s strategy by deploying the unit from SO19 to Scotia Road earlier.

“20.15 The management of the operation between 07:15hrs and 09:30hrs should have involved giving practical effect to the strategy devised by Commander MCDOWALL so that appropriate resources were in place at SCOTIA ROAD from the earliest possible time. Commander DICK was in charge of the operation following her briefing from Commander MCDOWALL. The policy, which is described at paragraphs 6.3 and 6.4, was, in essence, one of containment, stop and arrest. What occurred between 07:15hrs and 10:06hrs was a failure of that policy. Between 07:15hrs and 09:33hrs there was no adequate effort to put in place police resources at SCOTIA ROAD that would have enabled the Metropolitan Police to give effect to the policy. During those hours there was a series of briefings. None of the eight people who left the flats before Mr DE MENEZES left were stopped in accordance with the strategy and when he left he was simply followed while ineffective attempts were made during the course of half an hour to determine whether he was [the suspect]. If appropriate resources had been in place there would have been the opportunity to stop Mr DE MENEZES during the course of his five minute walk from SCOTIA ROAD before catching the bus in TULSE HILL.

...

20.32 Detective Chief Inspector C, the Silver Commander, was effectively the ground commander with responsibility for SO12, SO13 and [SO19] officers. However owing to the fact that he was still with SO13 and [SO19] at NIGHTINGALE LANE when Mr DE MENEZES left SCOTIA ROAD, and stationary at the T.A. Centre where DE MENEZES was identified as the suspect at BRIXTON, DCI C was always playing ‘ Catch up ’ in respect of the operation.

...

20.49 ... evidence from the CCTV at STOCKWELL underground station reveals that the [SO19] officers did not enter the station until two minutes after Mr DE MENEZES had passed though the ticket barriers.

20.50 While two minutes is a very short time period, the delay in [SO19] getting to the scene and the failure to get a positive identification had enabled a person, believed to be a possible suspect for attempting to detonate a bomb on the underground system the day before, to get on the same bus twice and enter an underground station.”

56 . Thirdly, the report was critical of the delay in handing the investigation to the IPCC.

“17.22 The pressures under which the Metropolitan Police were operating following the events of 7 July and 21 July are self-evident. However, the fact that the independent body established by an Act of Parliament to investigate complaints and serious incidents involving the police, and which has independently investigated every fatal police shooting since 1 April 2004, was now to be excluded from the scene, is a major concern for an independent investigation, and should never occur again.

17.23 The fact that there was such concern over the problems with the CCTV tapes at STOCKWELL and the fact that the hard drives on the train were missing highlights the problem. This issue could have been resolved a lot earlier had they been under the control of the IPCC.

...

17.25 The failure to allow the IPCC access has also been highlighted by the fact that the surveillance log 165330 has been altered.

...

17.33 Had the IPCC been involved at the commencement of the investigation, the surveillance log would not have been released for amendments to be made.”

57 . Nevertheless, the IPCC found that high vibrations had interfered with the recording of most of the bus journey, the hard drive on the train had not been replaced on the relevant day, and the recording equipment in the station had been broken during prior refurbishments. Consequently, it concluded that there was “no evidence of a cover-up to withhold this evidence from the investigation”.

58 . Likewise, two expert witnesses who examined the surveillance log could not agree either that it had been altered or, if it had been, who might have altered it.

(b) Prosecutions

59 . The report also identified a number of individuals whom the CPS might consider prosecuting.

(i) Charlie 2 and Charlie 12

60 . As to the shooting of Mr de Menezes after he had been tackled on the train, the IPCC noted as follows.

“20.71 The actions of Charlie 2 and Charlie 12 should be considered in light of the day ’ s events and those of the previous two weeks. At the briefing, they were supplied with a full briefing on the capabilities of the terror suspects. During the operation they had heard the man being followed was being identified as one of the suspects from the previous day ’ s attempted bombings. On arrival at STOCKWELL, [SO19] went to State Red, authorising a firearms intervention, following an order from the DSO to stop the man from entering the station and tube train.

20.72 They had seen ‘ Ivor ’ [a surveillance officer from SO12] point at the suspect, who they saw get off his seat. ‘ Ivor ’ then grabbed the man and forced him back to the seats. Both officers state they believed they had to act immediately to prevent loss of life to the people on the train.

...

20.74 Charlies 2 and 12 clearly believe[d] they were acting in self-defence, and had the right in law to use the force they did. The [CPS] may wish to consider whether the actions of Charlie 2 and Charlie 12 amount to murder in the context of their justification for the shooting of Mr DE MENEZES and having regard to the fact that there were explanations given for the shooting at that time which did not accord with the accounts given 36 hours later.

...

20.94 ... [ The CPS] ... may also wish to consider whether they were grossly negligent to come to the conclusion that they were confronting a suicide bomber. ”

(ii) The DSO

61 . With regard to the role of the DSO, the IPCC stated as follows.

“20.77 The order given by Commander DICK was to stop the suspect getting onto the underground station and subsequently the underground train. When interviewed she was asked to explain the word ‘ Stop ’ and her response was that ‘ Stop ’ is a common word in policing terms and it was meant as ‘ stop and detain ’ . This opinion is supported by DCI C and Trojan 80 and 84.

20.78 However, the way the order was received by [SO19] must be considered. Following a full briefing, many of the [SO19] officers have described that they believed that they would have to confront a suicide bomber. The [SO19] officers have stated that they believed the man being followed on the bus had been identified as one of the suspects for the failed bombings on 21 July 2005. They had been in a situation of trying to ‘ Catch up ’ with the surveillance team since their briefing had finished. And as they approached STOCKWELL underground station they hear that the suspect had entered the underground station and they received an order to stop him getting on the underground train. I do not believe that the use of the word ‘ Stop ’ can be related to normal policing duties. With the mind-set of the [SO19] officers believing that a suicide bomber had entered the underground station, to receive such an order to stop him from DSO cannot be related to normal duties. They had not had the benefit of a rider to their briefing of the sort to which I refer at paragraph 20.8. If they had received such a briefing they might have been more cautious in the way they approached and dealt with Mr DE MENEZES.

...

20.82 I [Senior Investigator J.D. Cummins] comment[s] at paragraph 20.47 on the consequences of the surveillance team having failed to adequately identify the person they were following. However, that team had spent thirty minutes following and staying with Mr DE MENEZES and attempting to identify him. That provided Commander DICK with a thirty minute opportunity to act in accordance with the operation strategy. There was no attempt to do so.

20.83 The SO12 officers who were following Mr DE MENEZES had been authorised to carry firearms for their personal protection and the protection of the public. In the context of the events of 7 July and 21 July when, respectively, there had been a successful detonation and an attempted detonation of bombs on buses it was a failure of the management of the operation to permit Mr DE MENEZES to get on the bus at TULSE HILL. If he had been a suicide bomber that event could have been catastrophic. Therefore the failure to use SO12 to stop him getting back on the bus in BRIXTON is an even more inexplicable failure to apply the strategy.

...

20.87 [The DSO] has endorsed that she was the person in command.

The [CPS] may wish to consider whether the manner in which this operation was commanded, the failures to have resources properly deployed and the absence of any other tactical options could be considered to be grossly negligent. ”

(iii ) “James”

62 . With regard to the “identification” of Mr de Menezes as the suspect, the IPCC noted:

“20.53 ... James [the head of the surveillance teams] did not communicate that some of his team thought that the subject was not [the suspect]. This information should have been fully communicated to [the DSO] as it may have influenced her decision-making. The [CPS] may wish to consider whether this negligence by ‘ James ’ ... satisfies the test for gross negligence.”

(iv) The other officers on the train

63 . As to any potential offence on the part of the eight officers on board the train:

“20.91 Given that they believed they were confronting a suicide bomber it is perhaps illogical that they would have challenged him prior to trying to detain him. The [CPS] may wish to consider whether any of the eight officers on the train who state they shouted or heard the words ‘ armed police ’ have conspired to pervert the course of justice. ”

(v) Trojan 80, Trojan 84 and DCI C

64 . As the IPCC did not consider that Trojan 80, Trojan 84 and DCI C had been in a position to influence the outcome of events, it was of the opinion that they could not be held responsible.

(vi ) The surveillance log

65 . In respect of the possible alteration of the surveillance log (see paragraphs 42 and 56 above), the IPCC did not find sufficient evidence against any individual to suggest that criminal proceedings might be appropriate.

(c) Operational recommendations

66 . The IPCC noted that, in the course of its investigation, grave concerns had been raised regarding the effectiveness of the police response on 22 July 2005. These concerns were not only that an entirely innocent member of the public had been killed in error but also that the police response might not have been adequate to stop a terrorist who was intent on causing harm. It therefore made a number of detailed operational recommendations.

68 . In respect of the surveillance operations, the IPCC expressed concern that the surveillance team, the SFOs, and those in command were not used to working together and were not sufficiently familiar with each other ’ s working practices; that two surveillance officers believed the person being followed was not the suspect and that this was not communicated to the DSO; and that the surveillance log had been altered.

70 . As regards the communications infrastructure, the IPCC was concerned that key briefings and strategic and tactical decisions were not recorded and, furthermore, that the command and control of the incident was inevitably lost when the unit from SO19 entered the underground. Concerns were also expressed that the existing Firearms Manual and the “KRATOS” policy were patently insufficient to deal with the current terrorist threat.

(d) Publication

71 . The IPCC Stockwell One Report was not made public until 8 November 2007, as publication was delayed pending the criminal trial of the Office of the Commissioner of the Police of the Metropolis (“the OCPM” – see paragraphs 100 - 01 below).

3. The second IPCC investigation and the IPCC Stockwell Two Report

72 . On 14 October 2005 the MPA referred a complaint to the IPCC regarding the MPS ’ s handling of public statements following the shooting of Mr de Menezes. The IPCC carried out a second investigation and the IPCC Stockwell Two Report was published on 2 August 2007. The contents of that report are not directly relevant to the complaint currently before the Court.

D. Disciplinary proceedings against the frontline and surveillance officers

73 . The IPCC had the power to recommend or direct the MPS to bring disciplinary proceedings against individuals. During the IPCC investigation, fifteen officers were served with notices under Regulation 9 of the Police (Conduct) Regulations 2004, informing them that they were being investigated and warning them that the investigation might result in disciplinary proceedings being brought against them.

74 . However, on 11 May 2007 the IPCC decided that no disciplinary action should be pursued against any of the eleven frontline and surveillance officers involved in the operation since there was no realistic prospect of any disciplinary charges being upheld. One surveillance officer received “words of advice” in connection with the alteration of the surveillance log.

75 . A decision concerning disciplinary charges against the two Commanders and their tactical advisers was postponed until after the prosecution of the OCPM (see paragraphs 100 - 01 below).

E. The first prosecutorial decision

1. The decision

76 . On receiving the IPCC Stockwell One Report, the CPS considered whether to bring prosecutions against any individual officers for murder, involuntary manslaughter by way of gross negligence (“gross negligence manslaughter”), misconduct in public office, forgery or attempting to pervert the course of justice. It also considered whether to prosecute the OCPM or any individual for offences under the Health and Safety at Work etc. Act 1974 (“the 1974 Act”). In deciding whether or not to prosecute, it first had to apply a threshold evidential test, namely, whether or not there was a realistic prospect of conviction, before asking whether or not prosecution would be in the public interest (see paragraph 163 below).

(a) The first decision letter

77 . By letter dated 17 July 2006, the CPS notified the deceased ’ s family that the Director of Public Prosecutions (“the DPP”) had decided to prosecute the OCPM, not in the Commissioner ’ s individual capacity but as an employer of police officers, for failing to provide for the health, safety and welfare of Mr de Menezes, contrary to sections 3 and 33 of the 1974 Act (see paragraphs 157 - 58 below). No individual was to be prosecuted in relation to the death as there was “insufficient evidence to provide a realistic prospect of conviction against any individual police officer”, that is, it was more likely than not that a jury would not convict.

78 . The decision letter, in so far as relevant, provided as follows.

“In the circumstances of this case, if the prosecution could prove that [the SFOs] were not acting in self defence (either of themselves or others) then they would be charged with murder. The order was given that Jean Charles was to be stopped from getting on the train. Although officers in the control room intended that Jean Charles should be arrested outside the station, the [SFO team] were not in place to make such an arrest, nor was this intention made explicit to the [SFOs] who were being sent down to the train. All the available evidence suggests that they believed that Jean Charles had been identified as a suicide bomber, that they had been directed to stop him from blowing up the train and that they had to shoot him to prevent that. ...

The burden would be on the prosecution to prove beyond reasonable doubt that these two officers did not honestly and genuinely believe that they were facing a lethal threat and so I looked to see if there was sufficient evidence to disprove that they had such an honest and genuine belief. Both officers stated that Jean Charles was wearing a ‘ bulky ’ jacket when they saw him but in fact Jean Charles was wearing a simple denim jacket. I therefore took this into account as it could indicate that the officers had lied. However even if I could prove that the officers had lied, rather than simply being mistaken, this alone would not be enough to commence a prosecution for murder as there could be other reasons for an officer to lie. I also considered their explanations of Jean Charles ’ s movements when they approached him, to see if there was evidence that they had fabricated those accounts to justify their actions. Both refer to Jean Charles getting up and advancing towards them with his hands down by his side before he was tackled by a surveillance officer and forced back into the seat. The [SFOs] then shot Jean Charles. I had to consider whether the prosecution could argue that the restraint meant that no bomb could be detonated and that the firearms officers ’ actions were unlawful. However I must bear in mind that this happened in a matter of seconds and there is some independent evidence that supports the officers ’ accounts that they feared Jean Charles might detonate a bomb. A witness sitting opposite Jean Charles said ‘ I got the impression that he was reaching to the left hand side of his trouser waistband. ’ ...

As I cannot prove the officers did not act in genuine self-defence, I cannot charge them with murder or any other offence of assault, including manslaughter.

There is some disagreement between officers and the members of the public as to whether any warning was given that armed police were approaching the train. In a situation such as this, where a warning to a suspected bomber could be fatal for officers and the public, no warning should be given. However some police officers say that they did hear a call of ‘ armed police ’ before the shooting and although passengers did hear officers shouting as they ran down the stairs, none of them heard the words ‘ armed police. ’ Both of the [SFOs] say that they shouted ‘ armed police ’ immediately before they fired but whether they did, and if so, whether it was intended as a warning to Jean Charles or to others in the carriage is unclear. There is no doubt that some police officers did shout something before any shots were fired .... Unless I could prove that officers had lied ... to mislead any investigation, I could not prosecute them for attempting to pervert the course of justice.

Next I carefully examined the roles of those police officers concerned in planning the surveillance and stop and those who carried it out. [T]here were a number of people involved and there is no doubt that messages were misinterpreted with tragic consequences. I have considered whether any errors or other conduct by individuals could be categorised as criminal. In this I have applied the law on gross negligence manslaughter, misconduct in public office and the [1974] Act. Even where I found that individuals had made mistakes, I found insufficient evidence that those mistakes were so bad that they could be described as criminal. As criminal proceedings are to be brought against the [OCPM], I cannot provide you with a detailed account of the conduct of those individuals, as that conduct will form part of the prosecution case.”

(b) The Review Notes

79 . More detailed reasons were provided in a fifty-page Review Note dated 9 March 2006 as well as in a Final Review Note of 9 July 2006.

(i) The IPCC investigation

80 . In respect of the investigation by the IPCC, the Review Note stated as follows.

“I am satisfied that the investigation has complied with Article 2 and the procedural requirements that flow from it. The IPCC is clearly independent of the Metropolitan Police and the investigation has not been limited to the actual shooting on the train but has examined the whole of the operation. I have had a number of discussions with senior investigators at the IPCC who have assisted me with any queries I have raised. I am therefore satisfied that I have sufficient material before me to reach a decision on the criminal liability of those officers involved in the operation that led to the death of Mr de Menezes and the Commissioner as corporation sole.”

81 . However, the Review Note drew attention to one particular evidential difficulty:

“Perhaps the most significant problem in understanding what occurred is that there is an almost complete absence of any worthwhile contemporaneous records and the accounts from the participants vary significantly on all the crucial aspects. It is at times impossible to say with any certainty what was said, by whom, to whom and when. There is also the issue that some accounts were made in the knowledge that something terrible had gone wrong.”

82 . With regard to the witness statements taken from the passengers on the train, the Review Note indicated that there were inevitable inconsistencies in their recollections of events with the consequence that “the accounts do not match either among themselves or with those of the police”. For example, some of the witnesses confused Mr de Menezes with “Ivor”, one of the surveillance officers.

(ii ) Charlie 2 and Charlie 12

83 . With regard to Charlie 2 and Charlie 12, the Review Note reiterated that there was insufficient evidence to persuade a jury that they did not genuinely believe they were acting in self-defence. It noted that, if they did hold a genuine belief, then the actions they took in shooting dead a “suicide bomber” would be reasonable and would not be unlawful.

(iii ) The DSO

84 . As for Commander Cressida Dick, the Review Note stated that there was no evidence against her to sustain a charge of murder, as she did not order any officer to open fire. The prosecutor was, however, satisfied that there was evidence her actions and direction and failure to plan fell below the standard of a reasonable officer in her position and, as such, a breach of the duty of care and causation could be shown. Nevertheless, he considered that there was “nowhere near enough” evidence to persuade a jury that her conduct was so bad as to justify a charge of gross negligence manslaughter. He also considered the possibility of prosecuting Commander Dick for offences under sections 7 and 33 of the 1974 Act, but, having applied the relevant criteria, found that the prosecution of her or of any of the other individual officers under these provisions would not be in accordance with Health and Safety Executive Policy.

(iv ) Trojan 84

85 . In the Review Note the prosecutor identified Trojan 84 as the officer most closely connected with the death of Mr de Menezes. In particular he had failed to dispatch firearms cover to Scotia Road, he gave the briefing that stoked the SFOs ’ fears that they would meet suicide bombers and that they might have to shoot them, and, finally, he should have known that, once the SFOs were away from the armed response vehicle and were to engage with a potential suicide bomber, the overwhelming likelihood was that they would shoot. However, he could not be prosecuted for murder as he did not direct the officers to fire and his actions were not “bad” enough to satisfy the test for gross negligence manslaughter.

(v) Trojan 80 and DCI C

86 . Likewise, the prosecutor considered that there was insufficient evidence to prosecute Trojan 80, DCI C or the surveillance teams for gross negligence manslaughter.

(vi ) Alteration of the surveillance log

87 . The Review Note considered the alleged alteration of the log (see paragraph 56 above), but found that it had been examined by two experts who did not agree to the required standard either that there had been alterations or, if there had been, who might have made them. Therefore, as it could not be proved that the relevant entry was a forgery, let alone who might have forged it, there was no basis for a prosecution for conspiracy to pervert the course of justice.

(vii ) Missing recordings

88 . It also indicated that there was no evidence the police, or anyone else, had tampered with the recording equipment on the bus, at the station or on the train. Although there were gaps in the recordings at all three locations, the IPCC investigation had revealed that high vibrations had interfered with the recording of most of the bus journey, the hard drive on the train had not been replaced on the relevant day, and the recording equipment in the station had been broken during prior refurbishments.

(viii ) The decision to prosecute the OCPM

89 . The Review Note explained in greater detail the decision to prosecute the OCPM. The prosecutor indicated:

“In my view, this operation was badly handled from the moment it passed from Commander [McDowall]. It resulted in an innocent man being shot dead in the most horrific manner. The Metropolitan Police were under tremendous pressure and were doing their best to protect the public from suicide bombers. These are factors that I take into account but these do not detract from the failure to carry out [Commander McDowall ’ s] strategy which would have best protected Mr de Menezes.”

90 . He continued:

“In my view, the lack of planning led to the death of de Menezes and, as such, constituted an offence under section 3 of the [1974 Act]. I believe that if such a charge is preferred, we can prove the case on the evidence already available but a decision not to prosecute individuals will enable the IPCC to seek further evidence to strengthen the case, from those individuals who are at present declining to.”

91 . The only defence open would be one of “reasonable practicability” and it was

“difficult to see how the police could argue the lack of reasonable practicability in ensuring the safety of [Mr de Menezes]. If this came to a contested trial, the police would probably have to call a number of officers ... who were interviewed as suspects. Their failures in the planning would then be highlighted.”

2. Judicial review of the first prosecutorial decision

92 . On 16 October 2006 the applicant sought leave to apply for judicial review of the decision not to prosecute any individual police officer for criminal offences, which she argued was incompatible with Article 2 of the Convention.

93 . In particular the applicant argued that the threshold evidential test in the Code for Crown Prosecutors (“the Code”), which prevented a prosecution unless a jury, properly directed in accordance with the law, was likely to convict (see paragraph 163 below), was not compatible with Article 2. She also submitted that Article 2 required the courts to undertake a more intensive review of a prosecutor ’ s decision than that provided for in R v. Director of Public Prosecutions, ex parte Manning ([2001] 1 QB 330) , in which the Divisional Court stated that it would accord great weight to the judgment of experienced prosecutors and, as such, a prosecutorial decision would be lawful if it was taken in accordance with the Code and was a decision reasonably open to the prosecutor on the material before him (see paragraph 165 below ).

94 . On 14 December 2006 a Divisional Court of the High Court granted permission to apply for judicial review but dismissed the substantive application.

95 . In relation to the compatibility of the Code with Article 2, the court found that this Court ’ s jurisprudence did not determine any particular evidential test to be applied when deciding whether or not there should be a prosecution. The test set out in the Code was therefore compatible with the obligation under Article 2 to put in place effective criminal-law provisions to deter the commission of offences against the person, backed up by law-enforcement machinery for the prevention, suppression and punishment of breaches of such provisions. Bringing prosecutions which were likely to fail, even if they could survive a dismissal application and a submission of no case to answer, would have profound consequences for all parties concerned . Furthermore, if the threshold was lowered in cases where lethal force was employed by State agents, it was likely that a significant proportion of prosecutions would fail because the evidence was lacking. If this were to happen, public confidence in both law-enforcement agencies and in the CPS would be undermined.

96 . The court also held that Article 2 did not require a change to the established position regarding judicial review of a decision not to prosecute. The “careful scrutiny” review required in Öneryıldız v. Turkey ( [GC], no. 48939/99, § 96, ECHR 2004 ‑ XII) was compatible both with the test outlined in the Manning case (see paragraph 165 below) and with the domestic courts ’ general approach to cases involving fundamental human rights.

97 . Thirdly, applying the Manning test, the court found that the decision of the CPS was in accordance with the Code and was one which was reasonably open to it. The decision was taken by a very senior and highly experienced prosecutor and it was reviewed by the head of the CPS and by independent counsel. It was lengthy, careful, thorough, clear and detailed, and the CPS had applied the correct test to each individual considered, namely, “whether there was sufficient evidence to provide a realistic prospect of conviction, or, in other words, whether a jury was more likely to convict than not to convict”.

98 . Although it was not necessary for the court to go so far, it also indicated that it saw “no reason to disagree with the decision”. Consequently, it concluded that the DPP ’ s decision was lawful and dismissed the applicant ’ s challenge to it. Leave to appeal to the House of Lords was refused by the court and, on 26 July 2007, by the House of Lords itself.

99 . On 22 January 2007 the court also rejected an application by the OCPM to have the charges under the 1974 Act dismissed.

3. The prosecution of the OCPM

100 . On 1 October 2007 the criminal trial of the OCPM commenced. A total of forty-seven witnesses were called to give evidence during the course of the trial, including Commander McDowall and Commander Dick. The prosecution argued that the OCPM was guilty of the following:

(a) Commander McDowall ’ s strategy had not been communicated adequately to the officers who took over the running of the operations on 22 July 2005, the surveillance officers or the SFOs;

(b) Commander McDowall ’ s strategy for controlling the premises was not adequately planned for or carried out;

(c) the control-room officers, the SFOs and the surveillance officers had a confused and inconsistent understanding of the strategy for Scotia Road;

(d) officers had not been deployed to stop and question persons emerging from the Scotia Road premises, including Mr de Menezes;

(e) the SFOs were not in attendance at Scotia Road when Mr de Menezes emerged from the doorway that was common to numbers 17 and 21;

(f) there was no contingency plan for dealing with persons who emerged from the apartment building before the firearms officers arrived;

(g) persons emerging from Scotia Road had not been stopped and questioned;

(h) a safe and appropriate area where those leaving Scotia Road could be stopped and questioned had not been identified;

(i) the briefings given to the SFOs were inaccurate, unbalanced, and provided the SFOs with inadequate and inaccurate information regarding the operation, including the operation at Scotia Road;

(j) the information concerning the identification of Mr de Menezes, his clothing, demeanour and likely level of threat, was not properly or accurately assessed or disseminated to officers and, in particular, to the SFOs;

(k) doubts concerning the correctness of the identification of Mr de Menezes as the suspect were not communicated to the control room;

(l) the control-room officers failed to satisfy themselves that a positive identification of Mr de Menezes as the suspect had been made by the surveillance officers;

(m) the SFOs had not been deployed at relevant locations in time to prevent Mr de Menezes from getting on the bus and entering Stockwell underground station;

(n) the SFOs failed to satisfy themselves that a positive identification of Mr de Menezes as the suspect had been made by the surveillance officers;

(o) effective steps were not taken to stop underground trains or buses so as to minimise the risk to the travelling public;

(p) Mr de Menezes was permitted to get on a bus twice and to enter Stockwell underground station despite being suspected of being a suicide bomber and despite having emerged from an address linked to a suspected suicide bomber;

(q) a clear and timely order that Mr de Menezes be stopped or arrested before he entered Stockwell underground station had not been given;

(r) accurate information had not been given to the DSO regarding the location of the SFOs when she was deciding whether the SFOs or officers from the Anti-Terrorist Branch should stop Mr de Menezes; and

(s) the risk inherent in effecting the arrest of Mr de Menezes by armed officers had not been minimised, whether in relation to the location, timing or manner of his arrest.

101 . On 1 November 2007 the jury returned a verdict, finding the OCPM guilty of breaching sections 3 and 33 of the 1974 Act (see paragraphs 157-58 below). The jury also attached a rider to its verdict to the effect that Commander Dick bore no “personal culpability” for the impugned events. This rider was endorsed by the trial judge. The OCPM was fined 175,000 pounds sterling (GBP) and ordered to pay costs of GBP 385,000.

F. Disciplinary proceedings against the two Commanders and their tactical advisers

102 . After the trial, the IPCC decided not to issue a recommendation for the senior officers to face disciplinary proceedings. In particular it had regard to the jury ’ s rider that no blame should be attached to Commander Dick, who was the most senior officer.

G. The inquest

103 . The inquest, which had been adjourned pending the trial of the OCPM, commenced on 22 October 2008. In the course of the inquest, seventy-one witnesses were called, including Commander McDowall, Commander Dick, Trojan 80, Trojan 84, Charlie 2 and Charlie 12. The family of Mr de Menezes were represented at the hearing at the State ’ s expense and were able to cross-examine witnesses and make submissions.

104 . On 24 November 2008 the coroner delivered a written ruling on what, if any, verdicts should be left to the jury. The options available to him were lawful killing, unlawful killing and an open verdict. However, the coroner was not permitted to leave a verdict to a jury if it fell foul of the test used to determine a submission of “no case to answer”, namely, if there was no evidence to support it or the evidence was so weak, vague or inconsistent with other evidence that, taken at its highest, a jury properly directed could not properly return that verdict (see paragraph 166 below).

105 . The coroner therefore considered the verdicts to be left to the jury separately as regards certain police officers.

1. The SFOs who shot Mr de Menezes (Charlie 2 and Charlie 12)

106 . The coroner found as follows.

“16. ... There is no doubt that the officers intended to kill Mr de Menezes when they fired. Therefore, if their contention that [they] were acting lawfully in defence of themselves or others could be disproved, they would have committed ... the offence of murder.

17. There is agreement between all Interested Persons as to what test I should apply in determining whether the officers acted lawfully in defence of themselves and others:

(i) Did the officer honestly and genuinely believe that it was necessary for him to use force in defence of himself and/or others? This is a question of subjective belief. Even if the belief was mistaken, and even if the mistake was unreasonable, the defence can still run. The reasonableness of the belief is only relevant in helping the jury to decide whether the belief was honestly held.

(ii) If the officer did hold the belief, did he use no more force than was reasonably necessary in the circumstances as he believed existed at the time? This is an objective test, but it is applied realistically. Where a person faces a threat, the Courts will not judge with too precise a measure the degree of force he uses... It is also significant for present purposes that a person under threat is not required to wait passively for the blow to fall. A pre-emptive strike can be justified by the circumstances.

...

18. The legal test is no different when the person facing the threat is a police officer or a soldier. However, as Waller LJ said in Bennett at paragraph 15, the tribunal is entitled to take account of the person ’ s training when applying the two limbs of the test to the facts of the given case. The same must apply to specific briefings as well as general training.”

107 . It was accepted by the parties that the SFOs honestly believed that the man in front of them in the carriage was Hussain Osman, the person who was strongly suspected of having attempted to explode a bomb on the underground the day before. However, the coroner rejected the submission on behalf of the de Menezes ’ family that the officers did not honestly believe that Mr de Menezes represented an imminent threat. He therefore found that the jury could not properly conclude to the criminal standard of proof that the two officers did not honestly believe that Mr de Menezes represented a mortal threat to those around them. In reaching that conclusion, he stated:

“27. If the officers honestly believed that Mr de Menezes represented a mortal threat to themselves and those around them, it could not be said that they used more force than was reasonably necessary ... An argument was made ... to the effect that [one of the officers] used excessive force because he fired too many times ... In my judgment, it has no merit. The events took place in a few seconds, and one cannot fairly say that some of the shots to the head constituted reasonable force and some did not. In any event, the officers had been trained to fire until the threat was neutralised.”

108 . The coroner therefore declined to leave to the jury the option of returning a verdict of unlawful killing in relation to the actions of Charlie 2 and Charlie 12.

2. The senior officers

109 . The coroner then considered whether the senior officers could safely be found to have committed manslaughter by reason of having caused death by gross negligence. It was accepted by all parties that this offence had to be proved against a particular officer; the failings of a number of persons could not be aggregated. Four elements had to be proved in order to establish that the offence had been committed: the defendant must have owed a duty of care to the victim, the defendant must have breached that duty, the breach must have caused the death (namely, made a more than minimal causal contribution to the death), and the breach must be characterised as “gross”.

110 . In relation to the duty of care, the coroner concluded that

“35. ... a police officer can owe a duty of care in directing other police officers to perform an armed interception. The content of the duty here would be to take reasonable care to ensure that such an interception took place in such a location and at such a time as to minimise, so far as reasonably practicable, the risk of unnecessary injury to the subject of the intervention, to the officers concerned and to others in the immediate vicinity. In this case the duty would not arise before the point at which firearms officers were ordered to move through with a view to performing an interception.”

(a) Commander McDowall

111 . In relation to Commander McDowall, there were three alleged breaches of a duty of care: that he should have set a strategic plan to ensure that suspects were stopped between leaving the premises and reaching the public transport system; that he did not ensure that the unit from SO19 was deployed sooner; and that he had failed to keep himself informed and to ensure that his orders were being followed. In respect of each of these allegations, the coroner did not accept that Commander McDowall had owed any duty of care to Mr de Menezes. However, even if a breach of duty could be established, the coroner did not accept that it had led to Mr de Menezes ’ death.

(b) The DSO

112 . There were three allegations against Commander Dick.

“54. ... First, ... that [she] failed to ensure that the block on Scotia Road was kept under careful surveillance control and that tactics were employed to ensure that all suspects could be identified and stopped before reaching a bus stop. As it happens, the nearest bus stop was on Upper Tulse Hill, only a few minutes ’ walk from the block. The first obstacle [to this] argument is the difficulty of constructing a positive duty of care at that stage to stop Mr de Menezes close to his home. In my judgment, no such duty could exist. Even if it could, I consider that it would not have been practicable to implement this as a fixed and inflexible tactical plan... In any event, the surveillance control was good: Mr de Menezes was kept continually under surveillance but the covert status of the operation near Scotia Road was maintained. The failure to stop him at an earlier stage was based on an inability of officers to say whether he was identifiable with the suspect. Therefore, his death was not caused by any failure of surveillance control at Scotia Road.

55. Secondly, it is alleged that [the DSO] failed to keep herself informed of where surveillance and firearms officers were as Mr de Menezes was travelling from Tulse Hill towards Stockwell. Again, I do not think that a police officer owes a duty to a person under surveillance to ensure that he is informed of the movements of other officers, at least before any intervention is immediately in prospect. If there were such a duty, it would only be to keep oneself reasonably well-informed, since it would not be practicable to keep note of the precise position of every officer and car. The thrust of the evidence is that [the DSO] did keep herself reasonably well-informed. She was aware, through the surveillance monitor in the control room, that surveillance officers were following Mr de Menezes and of what they were saying. In any event, as [counsel for the family of Mr de Menezes] accepts, nothing could have been done to stop Mr de Menezes between his getting on the bus at Tulse Hill and his alighting at Stockwell. [The DSO] had [the SFOs] at the proper holding point at the time she wanted to deploy them. In the minutes before she ordered the intervention, she was relying upon information from [her tactical adviser] as to the position and readiness of the [SFOs]. In my judgment, she was entitled to rely upon that information. In all those circumstances, any failure on her part to keep herself informed was not causative of the fatal events in the carriage.

56. Thirdly, it is submitted ... that [the DSO] failed to exercise proper judgment in her decisions in the last critical minutes, after Mr de Menezes left the bus at Stockwell. In my judgment, she probably did owe a duty of care to him at this stage in making decisions and giving directions for an armed stop. However, she cannot fairly be said to have breached that duty. When she became aware that the subject of surveillance had left the bus, she ordered the [SFOs] to perform an armed stop. Upon hearing that they were not in a position to make the stop, she instructed the surveillance officers to do so. That order cannot be characterised as negligent. If there were any slight delay in giving the order, that can probably be explained by the need to take thought before ordering a suspected suicide bomber to be stopped by officers who were not trained for such situations. Once she was told that the [SFOs] were in position, she countermanded the earlier order. It might be possible to say that she made the wrong decision at that point, given where Mr de Menezes was known to be, but these were fast-moving events and her decision cannot be described as negligent. [It was submitted] that using [SFOs] gave rise to a particular risk that lethal force would be used. However, there were obvious advantages to using officers who had the training and experience to perform armed interventions in a public place.”

(c) Trojan 80 (the DSO ’ s firearms tactical adviser)

113 . As to this officer, the coroner stated as follows.

“58. The first charge against [Trojan 80] is that, upon arriving at New Scotland Yard at around 6.00 a.m., he failed to take steps to expedite the despatch of [the unit from SO19] to the Scotia Road area. For the reasons already given, I do not consider that he would have owed a duty of care to Mr de Menezes in this regard. In any event, when he started work, all the critical decisions had been taken in relation to the [SFOs ’ ] deployments. It would probably not have been safe or sensible to try to expedite the deployments at that stage. As explained in paragraph 52 above, I do not think it can be established to the necessary standard of proof that any delay in deploying firearms teams was causally relevant to the death of Mr de Menezes.

59. The second allegation is that he failed to devise a tactical plan to ensure that any suspect coming out of the block was stopped before reaching a bus stop. This is, in essence, the same as one of the allegations made against [the DSO]. For the reasons I have given in paragraph 54, this argument fails at every stage.

60. The third point made in criticism of [Trojan 80] is that he failed to pass on to [the DSO] accurate information about the position of the [SFOs] in the minutes after it became apparent that Mr de Menezes was leaving the bus. However, [Trojan 80] was reliant for his information on the tactical adviser who was with the team on the ground, ‘ Trojan 84 ’ . That officer initially told [Trojan 80] that his team were ‘ not in contention ’ because they were behind the wrong bus. [Trojan 80] duly passed on that information. Even if it were incorrect, it is difficult to criticise him for passing it on.”

114 . If, contrary to all of the above, any of the allegations were made out, the coroner concluded that none approached the level of gross or criminal negligence.

(d) Conclusion

115 . In light of the above, the coroner decided not to leave the potential short-form verdict of unlawful killing to the jury in respect of the senior officers and instead left them to decide between a verdict of lawful killing and an open verdict.

3. Questions

116 . The coroner also included in his Ruling a list of proposed questions which would be left to the jury and which required responses of “yes”, “no”, or “cannot decide”. Having heard the parties ’ submissions, on 1 December 2008 he finalised the list of questions to include questions of fact concerning the events in the train carriage and questions concerning the factors which had contributed to Mr de Menezes ’ death. However, he refused to leave “open questions” to the jury inviting them to add any other factors which they regarded as causally relevant.

4. Judicial review of the coroner ’ s decision

117 . On 2 December 2008 Mr de Menezes ’ mother had sought leave to apply for judicial review of the coroner ’ s decision to exclude both the verdict of unlawful killing and certain narrative verdict questions. At the hearing, she pursued the second point only because, by that date, the coroner had started summing up and had already indicated the verdicts that were to be left to the jury.

118 . The claimant argued that the coroner had been obliged to ensure that the jury members were permitted to resolve the disputed factual issues at the heart of the case and were able properly to determine by what means and in what circumstances Mr de Menezes had come by his death. The question of how he came by his death went far beyond determining whether to return a verdict of lawful killing or an open verdict. The coroner ’ s approach had precluded the jury from commenting on whether or not they regarded any particular failings by the police as serious and, if so, how serious – and how important in terms of accountability – these failings were. As such, the jury ’ s findings were at best likely to beg more questions and at worst be confusing or meaningless. The claimant therefore wished to put additional narrative verdict questions to the jury once the coroner ’ s summing up was finished.

119 . On 3 December 2008 Silber J refused leave to apply for judicial review .

120 . Firstly, he found that the existing verdicts and questions satisfied the statutory obligation under section 11 of the Coroners Act 1988 and Rule 36 § 1 (b) of the Coroners Rules 1984 (see paragraphs 167 -68 below) to enable the jury members to ascertain by what means and in what circumstances Mr de Menezes had come by his death. Furthermore, the inquiry required by the coroner of the jury in this case was significantly more demanding than that sought from, and given by, the jury in both Bubbins v. the United Kingdom (no. 50196/99, ECHR 2005 ‑ II) and McCann and Others v. the United Kingdom (27 September 1995, Series A no. 324), and in those cases this Court had found that the procedural obligations under Article 2 of the Convention had been met.

121 . Secondly, the judge observed that the claimant had not pointed to any case decided domestically or in this Court which held that specific questions were required to be asked of a jury over and above asking “by what means and in what circumstances” the deceased had died.

122 . Thirdly, as the coroner had a discretion “to decide how best in the particular case to elicit the jury ’ s conclusion on the central issue or issues”, the judge considered that the only grounds for interfering with it would probably be Wednesbury grounds, namely, that the coroner ’ s decision was so unreasonable that no reasonable coroner would have done the same.

123 . Fourthly, the judge found that there was a risk that, if the jury members were required to answer the additional questions proposed by the claimant, they would be acting in contravention of Rule 36 § 2 of the Coroners Rules 1984 by expressing opinions on matters other than those on which they were entitled to comment and in particular by appearing to determine questions of criminal or civil liability.

124 . Fifthly, he considered that the proposed questions would expose the jury to a risk of making contradictory and conflicting findings.

125 . Sixthly, the judge found that the claimant had failed to show, even arguably, that there were strong grounds for disturbing the decision of the coroner.

126 . The claimant ’ s grounds relating to the short-form verdicts were adjourned generally with liberty to both parties to apply to restore. The claimant subsequently agreed that no further action would be taken in relation to these grounds because, inter alia , even if the judicial review was successful the only remedy for the family would be for the court to order a fresh inquest and the claimant did not “see any great benefit in re-hearing all the evidence to enable a different jury to come to a verdict, particularly bearing in mind the very high cost of holding such an inquest”.

5. Verdict

127 . On 12 December 2008 the jury returned an “open verdict”. In answering the questions left to them the jury found as follows:

(a) that Charlie 12 did not shout “armed police”;

(b) that, while Mr de Menezes did stand up before being grabbed in a bear hug by one of the surveillance officers, he did not move towards the SFOs;

(c) that the general difficulty in identifying the man under surveillance in the time available and the innocent behaviour of Mr de Menezes (which may have increased suspicion) were not contributory factors to his death;

(d) that the following were contributory factors to his death: the failure to obtain and provide to surveillance officers better photographic images of the failed bomber Hussain Osman; the fact that the views of the surveillance officers regarding the identification of the suspect were not accurately communicated to the command team and the SFOs; the failure by police to ensure that Mr de Menezes was stopped before he reached public transport; the fact that the position of the cars containing the SFOs was not accurately known by the command team as the SFO teams were approaching Stockwell underground station; the shortcomings in the communications system between various police teams on the ground; and a failure to conclude at the time that surveillance officers could have been used to carry out the stop on Mr de Menezes at Stockwell underground station; and

(e) it was not clear whether the pressure on police after the suicide attacks in July 2005 was a contributory factor to Mr de Menezes ’ death.

6. The coroner ’ s report

128 . After the verdict, the coroner delivered a report as required under Rule 43 of the Coroners Rules 1984. In the report he identified MPS systems and practices which gave rise to concern and the risk that other deaths might arise in the future. He further identified action which should be taken to prevent the occurrence or continuance of such circumstances or to eliminate or reduce the risk of death created by such circumstances. The coroner also reviewed material indicating which remedial steps had already been taken to develop police practice since the events of July 2005.

129 . In the report the coroner expressed concerns regarding the command structure employed by the police on 22 July 2005 and observed that the Association of Chief Police Officers (ACPO) manual on the police use of firearms and the command structure should be reviewed. He also made specific recommendations concerning the role of the DSO, who was responsible for ordering any intentional shot that might be required in anti-terrorist operations.

130 . The coroner also reported on the communication problems that the jury found had contributed to the death of Mr de Menezes. He recommended that changes should be made to the systems and methods of communication to ensure that there was better information available to enable accurate identification to be made and communicated and to ensure that appropriately trained police officers were available to deal with possible terrorist threats on the basis of as much up-to-date information as possible.

131 . Finally, the coroner made recommendations regarding the recording of briefing and control-room activity and recommended that the practice of police witnesses conferring before recording their accounts of events should cease.

H. The second prosecutorial decision

132 . Following the inquest, further meetings and exchanges of correspondence took place between the CPS and Mr de Menezes ’ family. On 26 March 2009 the family asked the DPP to review the decision not to prosecute in light of new evidence which had emerged at the inquest.

133 . On 8 April 2009 the DPP confirmed by letter that there remained insufficient evidence to prosecute any individual.

134 . Mr de Menezes ’ family did not apply for leave to seek judicial review of this decision, considering that there would be no prospect of success in light of the previous judicial-review action. The factual matrix had not significantly changed: the claim would have been on similar grounds to the previous claim for review and was therefore bound to fail.

I. Confirmation of decision not to recommend disciplinary proceedings

135 . By letter dated 2 October 2009, the Chairman of the IPCC rejected the family ’ s request to review its decision not to initiate disciplinary proceedings as no new evidence had emerged during the inquest to justify bringing disciplinary charges against any individual officer.

136 . He noted that the trial of the OCPM and the inquest had confirmed the conclusion of the IPCC that Mr de Menezes was killed because of mistakes that could and should have been avoided. Indeed, the trial of the OCPM, the coroner ’ s report, the IPCC recommendations, Her Majesty ’ s Inspectorate of Constabulary, the MPA and the MPS had all recognised the organisational failings that led to his death. Major efforts had been made to rectify these organisational failings and it was necessary to take them into account when judging the individual culpability of the officers concerned. Every independent judicial, prosecuting and disciplinary authority which had considered the conduct of the officers had concluded that individual criminal or disciplinary charges were not merited.

137 . In respect of Charlie 2 and Charlie 12, there was insufficient evidence to undermine their claim that they honestly believed they were dealing with a suicide bomber or to warrant proceedings based on the disciplinary offences of using excessive force or abusing authority. The officers had had at best five to ten seconds to assess whether to shoot to kill and given the overall scene of confusion, coupled with the stress of the circumstances, it was not possible to conclude that the mistakes which were made were deliberate or negligent.

138 . With regard to Commander McDowall, the IPCC concluded that it was not likely that any tribunal would find that failings which occurred after he set his strategy were due to negligence on his part.

139 . As for Commander Dick, the IPCC had regard to the criminal jury ’ s unambiguous conclusion that she had no personal culpability, especially as no evidence had emerged at the inquest which would cause a disciplinary tribunal to ignore this finding.

140 . The IPCC considered that there was no evidence that might cause a tribunal to accept the jury ’ s rider in respect of Commander Dick but not in respect of Trojan 80 or DCI C.

141 . In respect of “James”, the IPCC accepted that the degree of doubt as to the identity of Mr de Menezes was not communicated sufficiently clearly by the surveillance team. However, the IPCC concluded that this was the result of technical as well as personal shortcomings, the speed and stress of the circumstances and the lack of an unambiguous communications process.

J. A civil action for damages

142 . A civil action in damages was brought by the family of Mr de Menezes (including the applicant) against the Commissioner of the Police of the Metropolis. This was settled by way of mediation during the week of 16 November 2009. The settlement was on a confidential basis.

K. Operational changes implemented following the shooting of Mr de Menezes

143 . Following the death of Mr de Menezes, the MPS took a number of steps to improve its methods of command and control in counter-terrorism operations. In particular a common command model was introduced for planned firearms operations, a smaller team, or “cadre” of firearms commanders was formed, a new cadre of ACPO officers was created to deal with high-risk counter-terrorism operations and a new ACPO Firearms Manual was published.

144 . In addition, a Surveillance Command was formed to provide consistency of training, procedure and professional practice and to create a platform for increased interoperability (that is, how different units and personnel work together operationally) with other departments and national units. A structured rotation of teams between counter-terrorism and crime operations was implemented to familiarise personnel in both types of operation.

145 . Furthermore, a new counter-terrorism control room came into operation and steps were taken to clarify the roles and responsibilities of control-room staff and to provide them with high-quality training. Operational delivery of a new secure photo-imaging system for transmitting images of suspects and other data also took place. Audio-recording is now available in the control room, which is activated when dealing with any suicide-bombing threat, and a new and evolving covert airwave communication system has been introduced to ensure effective radio communications are available throughout the London underground system as well as above ground.

146 . Pursuant to ACPO guidance issued in October 2008, the practice of officers writing up their notes together after an incident has ceased in cases where police officers have discharged firearms. Internal instructions drawn up in consultation with the IPCC and DPS have since extended this change of practice to officers involved in other (non-shooting) death and serious-injury cases.

II. RELEVANT DOMESTIC LAW AND PRACTICE

A. Relevant criminal offences and defences

1. Murder

147 . The unlawful taking of life with intent to kill or cause really serious harm constitutes the common-law offence of murder, which is punishable by a mandatory sentence of life imprisonment.

2. Self-defence

(a) Common law

“It is both good law and good sense that a man who is attacked may defend himself. It is both good law and good sense that he may do, but only do, what is reasonably necessary.”

“If there has been an attack so that the defence is reasonably necessary, it will be recognised that a person defending himself cannot weigh to a nicety the exact measure of his necessary defensive action. If a jury thought that in a moment of unexpected anguish a person attacked had only done what he honestly and instinctively thought was necessary, that would be most potent evidence that only reasonable defensive action had been taken.”

“It is thus clear that the European Court of Human Rights has considered what English law requires for self-defence, and has not suggested that there is any incompatibility with Article 2. In truth, if any officer reasonably decides that he must use lethal force, it will inevitably be because it is absolutely necessary to do so. To kill when it is not absolutely necessary to do so is surely to act unreasonably. Thus, the reasonableness test does not in truth differ from the Article 2 test as applied in McCann [ and Others ]. There is no support for the submission that the court has with hindsight to decide whether there was in fact absolute necessity. That would be to ignore reality and to produce what the court in McCann [ and Others ] indicated was an inappropriate fetter upon the actions of the police which would be detrimental not only to their own lives but to the lives of others.”

(b) The Criminal Justice and Immigration Act 2008

“ Reasonable force for purposes of self-defence etc.

...

(3) The question whether the degree of force used by D was reasonable in the circumstances is to be decided by reference to the circumstances as D believed them to be, and subsections (4) to (8) also apply in connection with deciding that question.

(4) If D claims to have held a particular belief as regards the existence of any circumstances—

(a) the reasonableness or otherwise of that belief is relevant to the question whether D genuinely held it; but

(b) if it is determined that D did genuinely hold it, D is entitled to rely on it for the purposes of subsection (3), whether or not—

(i) it was mistaken, or

(ii) (if it was mistaken) the mistake was a reasonable one to have made.

(5) But subsection (4)(b) does not enable D to rely on any mistaken belief attributable to intoxication that was voluntarily induced.

(6) The degree of force used by D is not to be regarded as having been reasonable in the circumstances as D believed them to be if it was disproportionate in those circumstances.

(7) In deciding the question mentioned in subsection (3) the following considerations are to be taken into account (so far as relevant in the circumstances of the case)—

(a) that a person acting for a legitimate purpose may not be able to weigh to a nicety the exact measure of any necessary action; and

(b) that evidence of a person ’ s having only done what the person honestly and instinctively thought was necessary for a legitimate purpose constitutes strong evidence that only reasonable action was taken by that person for that purpose.

(8) Subsection (7) is not to be read as preventing other matters from being taken into account where they are relevant to deciding the question mentioned in subsection (3).”

3. Gross negligence manslaughter

156 . Any person causing death by gross negligence may be guilty of manslaughter. In R v. Adomako ([1995] 1 AC 171) the House of Lords stated that the offence of gross negligence manslaughter would be committed where the defendant was in breach of a duty of care owed to the victim; the breach of duty caused the death of the victim; and the breach of duty could be characterised as grossly negligent. In determining whether or not there had been gross negligence and whether this caused the death, it was not possible to aggregate the failures of various individuals.

4. Offences under the Health and Safety at Work etc. Act 1974

157 . Section 3(1) reads as follows:

“It shall be the duty of every employer to conduct his undertaking in such a way as to ensure, so far as is reasonably practicable, that persons not in his employment who may be affected thereby are not thereby exposed to risks to their health or safety.”

158 . Section 33(1)(a) provides that it is an offence for a person to fail to discharge a duty to which he is subject by virtue of, inter alia , section 3 of the 1974 Act.

B. Prosecutorial decisions

1. The CPS

159 . In 1986 the CPS was established as an independent body to prosecute criminal cases in accordance with the Code. Pursuant to sections 1 and 3 of the Prosecution of Offences Act 1985, the DPP is the head of the CPS and operates independently under the superintendence of the Attorney General. As a government minister, the Attorney General is accountable to Parliament for the work of the CPS.

160 . According to the Protocol between the Attorney General and the Prosecuting Departments (July 2009), other than in exceptional cases, decisions to prosecute are taken by prosecutors; the Attorney General will not seek to give a direction in an individual case save very exceptionally where necessary to safeguard national security. Moreover, it is a constitutional principle that in such exceptional cases the Attorney General acts independently of government, applying well-established prosecution principles of evidential sufficiency and public interest.

161 . The circumstances in which the CPS will pursue a prosecution are governed by the 1985 Act and the Code.

2. The Prosecution of Offences Act 1985

162 . Section 10 provides as follows.

“(1) The [DPP] shall issue a Code for Crown Prosecutors giving guidance on general principles to be applied by them ‑

(a) in determining, in any case ‑

(i) whether proceedings for an offence should be instituted or, where proceedings have been instituted, whether they should be discontinued; or

(ii) what charges should be preferred; and

(b) in considering, in any case, representations to be made by them to any magistrates ’ court about the mode of trial suitable for that case.

(2) The Director may from time to time make alterations in the Code.

...”

3. The Code for Crown Prosecutors

163 . The relevant sections read as follows.

“ 5. THE FULL CODE TEST

5.1 The Full Code Test has two stages. The first stage is consideration of the evidence. If the case does not pass the evidential stage it must not go ahead no matter how important or serious it may be. If the case does pass the evidential stage, Crown Prosecutors must proceed to the second stage and decide if a prosecution is needed in the public interest. The evidential and public interest stages are explained below.

THE EVIDENTIAL STAGE

5.2 Crown Prosecutors must be satisfied that there is enough evidence to provide a ‘ realistic prospect of conviction ’ against each defendant on each charge. They must consider what the defence case may be, and how that is likely to affect the prosecution case.

5.3 A realistic prospect of conviction is an objective test. It means that a jury or bench of magistrates or judge hearing a case alone, properly directed in accordance with the law, is more likely than not to convict the defendant of the charge alleged. This is a separate test from the one that the criminal courts themselves must apply. A court should only convict if satisfied so that it is sure of a defendant ’ s guilt.

5.4 When deciding whether there is enough evidence to prosecute, Crown Prosecutors must consider whether the evidence can be used and is reliable. ...

THE PUBLIC INTEREST STAGE

4.11 Accordingly, where there is sufficient evidence to justify a prosecution or to offer an out-of-court disposal, prosecutors must go on to consider whether a prosecution is required in the public interest.

4.12 A prosecution will usually take place unless the prosecutor is sure that there are public interest factors tending against prosecution which outweigh those tending in favour, or unless the prosecutor is satisfied that the public interest may be properly served, in the first instance, by offering the offender the opportunity to have the matter dealt with by an out-of-court disposal (see section 7). The more serious the offence or the offender ’ s record of criminal behaviour, the more likely it is that a prosecution will be required in the public interest.

4.13 Assessing the public interest is not simply a matter of adding up the number of factors on each side and seeing which side has the greater number. Each case must be considered on its own facts and on its own merits. Prosecutors must decide the importance of each public interest factor in the circumstances of each case and go on to make an overall assessment. It is quite possible that one factor alone may outweigh a number of other factors which tend in the opposite direction. Although there may be public interest factors tending against prosecution in a particular case, prosecutors should consider whether nonetheless a prosecution should go ahead and for those factors to be put to the court for consideration when sentence is passed. ...”

164 . An Explanatory Memorandum issued to prosecutors in 1994 provided that

“4.14 Crown Prosecutors should resist the temptation to define the evidential test as ‘ a 51% rule ’ . The CPS has always stated that weighing evidence (and the public interest) is not a precise science; it is therefore misleading to talk in terms of percentages – particularly to a single percentage point – because it implies that we can give individual pieces of evidence an exact weight and then add them up to reach a decision about prosecution. Crown Prosecutors should continue to avoid using any expressions which could convey the impression that the decision-making process is susceptible of very precise numerical definition. On the other hand, it is not unreasonable to talk of a conviction being ‘ more likely than not ’ .”

4. Judicial review of prosecutorial decisions

165 . In R v. Director of Public Prosecutions, ex parte Manning ([2001] 1 QB 330) Lord Bingham of Cornhill CJ, giving the judgment of the court, stated as follows.

“23. Authority makes clear that a decision by the [DPP] not to prosecute is susceptible to judicial review ... But, as the decided cases also make clear, the power of review is one to be sparingly exercised. The reasons for this are clear. The primary decision to prosecute or not to prosecute is entrusted by Parliament to the [DPP] as head of an independent, professional prosecuting service, answerable to the Attorney General in his role as guardian of the public interest, and to no-one else. It makes no difference that in practice the decision will ordinarily be taken by a senior member of the CPS, as it was here, and not by the [DPP] personally. In any borderline case the decision may be one of acute difficulty, since while a defendant whom a jury would be likely to convict should properly be brought to justice and tried, a defendant whom a jury would be likely to acquit should not be subjected to the trauma inherent in a criminal trial. If, in a case such as the present, the [DPP ’ s] provisional decision is not to prosecute, the decision will be subject to review by senior Treasury counsel who will exercise an independent professional judgment. The [DPP] and his officials ... will bring to their task of deciding whether to prosecute an experience and expertise which most courts called upon to review their decisions could not match. In most cases the decision will turn not on an analysis of the relevant legal principles but on the exercise of an informed judgment of how a case against a particular defendant, if brought, would be likely to fare in the context of a criminal trial before (in a serious case such as this) a jury. This exercise of judgment involves an assessment of the strength, by the end of the trial, of the evidence against the defendant and of the likely defences. It will often be impossible to stigmatise a judgment as wrong even if one disagrees with it. So the courts will not easily find that a decision not to prosecute is bad in law, on which basis alone the court is entitled to interfere. At the same time, the standard of review should not be set too high, since judicial review is the only means by which the citizen can seek redress against a decision not to prosecute and if the test were too exacting an effective remedy would be denied. ...”

C. The threshold evidential test for deciding whether to leave a case to the jury (“the Galbraith test”)

166 . In R v. Galbraith ([1981] 1 WLR 1039) it was held that a court could not stop a prosecution if there was “some evidence”, even if it was “of a tenuous character”, for example, because of inherent weaknesses or vagueness or because it was inconsistent with other evidence. Moreover, if the strength or weakness depended on the view to be taken of a witness ’ s reliability, or other matters which were generally within the province of the jury and where “on one possible view of the facts there is evidence on which the jury could properly come to the conclusion that the defendant is guilty”, then the judge should allow the matter to be tried by the jury.

D. Inquests

1. Statutory basis

167 . The law governing inquests is found in the Coroners Act 1988 and the Coroners Rules 1984. Section 11 of the Act provides that, at the end of an inquest, a coroner or jury must complete and sign an inquisition. Pursuant to section 11(5), an inquisition shall set out, so far as such particulars have been proved, who the deceased was and how, when and where the deceased came by his death. Neither the coroner nor the jury shall express any opinion on any other matters (Rule 36 § 2 (2)) and in particular “[n]o verdict shall be framed in such a way as to appear to determine any question of (a) criminal liability on the part of a named person, or (b) civil liability” (Rule 42).

168 . Section 16(7) provides:

“Where a coroner resumes an inquest which has been adjourned in compliance with subsection (1) above–

(a) the finding of the inquest as to the cause of death must not be inconsistent with the outcome of the relevant criminal proceedings;

...”

2. Relevant case-law

169 . In R (Middleton) v. West Somerset Coroner ([2004] 2 AC 182) the House of Lords considered the implications of Article 2 of the Convention on the interpretation of the Coroners Act and Rules. It concluded that an investigation should be capable of reaching a conclusion which resolved the central issues of fact in the case. Where a choice between “short-form” verdicts (unlawful killing, open verdict, lawful killing) was not capable of resolving those central issues, the inquest would not be Article 2 compliant. In such cases it might therefore be necessary for the judge or jury to return a narrative verdict, in order to be able to answer not only “by what means the deceased came by his death”, but also “in what circumstances”.

170 . The refusal by a coroner to leave a particular short-form verdict to a jury is governed by R v. HM Coroner for Exeter, ex parte Palmer (unreported, 10 December 1997); R v. Inner South London Coroner, ex parte Douglas-Williams ([1999] 1 All ER); and R (Bennett) v. HM Coroner for Inner South London ([2007] EWCA Civ 617).

171 . In Palmer the Court of Appeal stated that the coroner should not leave a verdict to a jury if it fell foul of the test used to determine a submission of “no case to answer” in criminal trials, namely that there was no evidence to support it or the evidence was so weak, vague or inconsistent with other evidence that, taken at its highest, a jury properly directed could not properly return that verdict (the Galbraith test). By contrast, if the strength or weakness of the evidence depended on the view to be taken of a witness ’ s reliability, then the verdict should be left to the jury.

172 . In Douglas-Williams (cited above) the Court of Appeal clarified the extent of the discretion of a coroner not to leave to the jury what was, on the evidence, a possible verdict. Lord Woolf MR stated at p. 348:

“If it appears there are circumstances which, in a particular situation, mean in the judgment of the coroner, acting reasonably and fairly, it is not in the interest of justice that a particular verdict should be left to the jury, he need not leave that verdict. He, for example, need not leave all possible verdicts just because there is technically evidence to support them. It is sufficient if he leaves those verdicts which realistically reflect the thrust of the evidence as a whole. To leave all possible verdicts could in some situations merely confuse and overburden the jury and if that is the coroner ’ s conclusion he cannot be criticised if he does not leave a particular verdict.”

173 . The Court of Appeal further clarified this in R (Bennett) (cited above). Waller LJ, giving the judgment of the court, considered that “there is some (if small) distinction between the position of a coroner deciding what verdict to leave to a jury after hearing all the evidence and that of a judge considering whether to stop a case after the conclusion of the prosecution case ”, that is, on a submission of no case to answer. At paragraph 30, he continued:

“ ... coroners should approach their decision as to what verdicts to leave on the basis that facts are for the jury, but they are entitled to consider the question whether it is safe to leave a particular verdict on the evidence to the jury i.e. to consider whether a verdict, if reached, would be perverse or unsafe and to refuse to leave such a verdict to the jury. ”

174 . A jury or coroner may only return a verdict of unlawful killing if satisfied beyond reasonable doubt that one or more persons unlawfully killed the deceased (see, inter alia , Bennett , and R (Sharman) v. HM Coroner for Inner North London [2005] EWHC 857 (Admin)).

III. RELEVANT COMPARATIVE LAW

A. Contracting States

175 . From the information available to the Court, it would appear that, leaving aside the question of private prosecutions, in at least twenty-five Contracting States the decision to prosecute is taken by a public prosecutor. That is the case in Albania, Armenia, Austria, Azerbaijan, Bulgaria, the Czech Republic, Estonia, Finland, Georgia, Hungary, Iceland, Ireland, Italy, Latvia, the Republic of Moldova, Montenegro, Poland, Portugal, Romania, Russia, Serbia, Sweden, Switzerland, Turkey and Ukraine. In a further twelve Contracting States, the prosecutorial decision is first taken by a public prosecutor before being put before a judge and/or a court. This is the position in Belgium, Cyprus, France, Germany, Lithuania, Luxembourg, the former Yugoslav Republic of Macedonia, Malta, Monaco, Slovakia, Slovenia and Spain.

176 . There is no uniform approach among Contracting States as to the threshold evidential test necessary to prosecute a case, although in at least twenty-four States a written threshold does exist. These States are Austria, Belgium, Bulgaria, the Czech Republic, Germany, Finland, France, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, the former Yugoslav Republic of Macedonia, the Republic of Moldova, Poland, Portugal, Romania, Russia, Serbia, Slovakia, Slovenia, Switzerland and Turkey.

177 . In one group of States, the threshold focuses on whether the elements of the offence have been made out; in the second group, it focuses on the chance of conviction by a court. However, these two categories are not watertight as it is not possible to say how prosecutors and judges apply the tests in practice. For example, a prosecutor applying a test based on the elements of the offence may also consider whether the strength or quality of evidence is sufficient for a conviction.

178 . In addition to the respondent State, at least four countries fall into the second group: Austria, Iceland, the former Yugoslav Republic of Macedonia and Portugal. In Austria, the test is “the likelihood of conviction before the court”; in Iceland, it is whether the evidence is “sufficient or probable for conviction”; in the former Yugoslav Republic of Macedonia the test is whether there is “enough evidence from which [the prosecutor] can expect a conviction”; and finally, in Portugal, it is whether there is a “reasonable possibility of imposing a penalty at trial.”

179 . In some States once the evidentiary threshold has been reached, the prosecutor must pursue the case. In Italy, for example, a decision to prosecute shall be taken if doubt as to the strength of the evidence could be rectified by new evidence presented at trial. In Germany, the principle of mandatory prosecution holds that “the public prosecution office shall be obliged to take action in relation to all prosecutable criminal offences, provided there are sufficient factual indications”.

180 . In other States, the evidentiary threshold allows the prosecutor to bring a case, but does not compel prosecution. The practice in Ireland, for example, as defined by the Guidelines for Public Prosecutors, is “the prosecutor approaches each case first by asking whether the evidence is sufficiently strong to justify prosecuting. If the answer to that question is ‘ no ’ , then a prosecution will not be pursued. If the answer is ‘ yes ’ , before deciding to prosecute the prosecutor will ask whether the public interest favours a prosecution or if there is any public interest reason not to prosecute.” In Cyprus, even if there is sufficient evidence to pursue a prosecution, there is no legal obligation to do so.

181 . The decision not to prosecute is susceptible of some form of judicial review or appeal to a court of law in at least eighteen Contracting States, namely Albania, Armenia, Austria, Azerbaijan, Belgium, France (albeit in limited circumstances), Ireland, Italy, Luxembourg, Malta, Monaco, Poland, Portugal, Russia, Spain, Switzerland, Turkey and Ukraine. In at least seven Contracting States, the decision of the prosecutor is normally contested before a hierarchical superior in the prosecution service with the final decision being susceptible of judicial review. These States include Bulgaria, Estonia, Germany, Lithuania, the Republic of Moldova, Romania and Slovakia. Finally, in at least twelve Contracting States there is no possibility of judicially reviewing the decision not to prosecute, although in some cases the decision may be contested to a hierarchical superior in the prosecution service. The States which do not permit judicial review include Cyprus, the Czech Republic, Finland, Georgia, Hungary, Iceland, Latvia, the former Yugoslav Republic of Macedonia, Montenegro, Serbia, Slovenia and Sweden.

B. Common-law countries

182 . In Australia, prosecutorial decisions are taken by the Office of the Director of Public Prosecutions which applies the Australian Prosecution Policy. The first criterion of this policy is that of evidential sufficiency, which is met if there is evidence sufficient to justify the institution or continuation of a prosecution. A prosecution should not be instituted or continued unless there is admissible, substantial and reliable evidence that a criminal offence known to the law has been committed by the alleged offender. The existence of a bare prima facie case is not sufficient to justify prosecution. Once it is established that there is a bare prima facie case, it is then necessary to give consideration to the prospects of conviction. A prosecution should not proceed if there is no reasonable prospect of a conviction being secured.

183 . In New Zealand, the Solicitor-General has published prosecution guidelines that draw extensively on the Australian Prosecution Policy, the CPS Code for Crown Prosecutors, and guidelines developed by the Public Prosecution Service for Northern Ireland and the Director of Public Prosecutions in the Republic of Ireland. The test is in two parts: the evidential test and the public-interest test. The evidential test is met if “the evidence which can be adduced in Court is sufficient to provide a reasonable prospect of conviction”.

184 . In Canada, the Public Prosecution Service Deskbook sets the standard in relation to decisions to prosecute. The first criteria is the evidential test, which requires Crown counsel objectively to assess the whole of the evidence likely to be available at trial, including any credible evidence that would favour the accused, to determine whether there is a reasonable prospect of conviction. A reasonable prospect of conviction requires that there be more than a bare prima facie case; however, it does not require a probability of conviction (that is, that a conviction is more likely than not).

185 . Finally, in the United States of America, the standard is whether there is “probable cause” to bring a prosecution, which means reasonable and objective grounds for belief in guilt.

THE LAW

I. ALLEGED VIOLATION OF ARTICLE 2 OF THE CONVENTION

187 . Article 2 provides as follows:

“1. Everyone ’ s right to life shall be protected by law. No one shall be deprived of his life intentionally save in the execution of a sentence of a court following his conviction of a crime for which this penalty is provided by law.

2. Deprivation of life shall not be regarded as inflicted in contravention of this Article when it results from the use of force which is no more than absolutely necessary:

(a) in defence of any person from unlawful violence;

(b) in order to effect a lawful arrest or to prevent the escape of a person lawfully detained;

(c) in action lawfully taken for the purpose of quelling a riot or insurrection.”

188 . The Government contested the applicant ’ s argument.

A. Admissibility

189 . The Court notes that this complaint is not manifestly ill-founded within the meaning of Article 35 § 3 (a) of the Convention. It further notes that it is not inadmissible on any other grounds. It must therefore be declared admissible.

B. Merits

1. The parties ’ submissions

(a) The applicant ’ s submissions

190 . The applicant did not complain that her cousin was killed by State agents in circumstances which breached Article 2 in its substantive aspect; consequently, she did not aver that his shooting was unlawful or that the conduct and planning of Operation THESEUS 2 was in breach of Article 2. Rather, her complaints fell solely under the procedural limb of Article 2 of the Convention and related solely to the fact that no individual police officer was prosecuted following the fatal shooting of Jean Charles de Menezes.

191 . More specifically , she argued that:

(a) the investigation into her cousin ’ s death fell short of the standard required by Article 2 of the Convention because the authorities were precluded from considering the reasonableness of the belief of Charlie 2 and Charlie 12 that the use of force was necessary; and

(b) the prosecutorial system in England and Wales prevented those responsible for the shooting from being held accountable and, as a consequence, the procedural requirement under Article 2 of the Convention had not been satisfied.

(i) The investigation

192 . In the applicant ’ s submission, the test for self-defence under domestic law was lower than the standard required by Article 2 of the Convention. Under the law of England and Wales, an officer who used lethal force in self-defence would have a defence if he honestly but mistakenly believed he was under imminent threat, even, so she argued, if that belief was wholly unreasonable. However, the test applied by the Court required an honest belief to be supported by “good reasons”. Therefore, if the honest belief was mistaken, the use of force could only be justified if the person had good reasons for believing it was necessary based on what was seen and known by him or her at the time.

193 . The applicant contended that, as the investigating authorities were applying a lower standard than that required by the Court, they were prevented from considering whether the use of force by Charlie 2 and Charlie 12 was or was not justified in the circumstances within the meaning of Article 2 of the Convention. In other words, the extent to which the domestic authorities were able to submit the actions of State agents to careful scrutiny was undermined, with the consequence that the State ’ s investigation was unable to secure accountability through a prosecution for a violation of Article 2 (see, for example, Vasil Sashov Petrov v. Bulgaria , no. 63106/00 , § 52, 10 June 2010).

194 . In the particular circumstances of the present case, the applicant contended that the authorities could not consider the reasonableness of the mistaken belief of Charlie 2 and Charlie 12 that Mr de Menezes posed a threat; nor were they required to analyse whether those officers had conducted both a careful assessment of the surrounding circumstances and an evaluation of the threat Mr de Menezes posed by his presence on the train.

(ii ) The prosecutorial decisions

195 . Although the applicant did not contend that there must always be a prosecution when there has been a death at the hands of a State agent, she submitted that there should be a prosecution where there was sufficient evidence to justify it. She argued that there had been sufficient evidence to justify the prosecution of a number of police officers involved in Operation THESEUS 2, but that flaws in the prosecutorial system in England and Wales had prevented the persons responsible for the death of her cousin from being held to account.

196 . The applicant did not submit that prosecutors in England and Wales were not adequately independent for the purposes of Article 2 of the Convention. However, relying on Maksimov v. Russia (no. 43233/02 , 18 March 2010), she criticised the fact that the prosecutor normally makes decisions without the benefit of oral testimony. She submitted that i n cases like the present, where honesty and credibility were decisive, it was vital that the prosecutor should be in a position to assess the demeanour of witnesses giving oral evidence.

197 . The applicant accepted that States were entitled to apply a threshold evidential test for permitting prosecutions to proceed, but contended that the threshold in England and Wales was too high. She accepted that “a realistic prospect of conviction” was used in some other States, particularly those with common-law legal systems, but she argued that in England and Wales this test had been interpreted to mean that a conviction should be more likely than not, that is, the chances of conviction were over fifty percent. She contended that the appropriate threshold should be the same as that used by the trial judge in deciding whether to allow a matter to be tried by a jury (the Galbraith test), namely, that there was “some evidence”, even if it was “of a tenuous character”, on which the jury could properly come to the conclusion that the defendant was guilty .

198 . Although the applicant accepted that this was the same test applied by the coroner in deciding what short-form verdicts to leave to the jury, she argued that the CPS was wholly independent of the coroner and had not been bound by his decision. It could not, therefore, be said that a prosecutor applying the same test would have come to the same conclusion. In any case, the applicant had sought permission to judicially review the coroner ’ s decision but, by the time the Administrative Court considered her claim, the jury had already been directed.

199 . In light of the absolute nature of Article 2, the applicant rejected any suggestion that there was a margin of appreciation in setting the threshold evidential test. However, even if there were, she submitted that the current threshold, which was substantially higher than the Galbraith test, was too high and therefore incompatible with Article 2 of the Convention. In particular she claimed that the threshold was set too high to maintain public confidence, to ensure adherence to the rule of law and to prevent any appearance of tolerance of or collusion in unlawful acts. Moreover, as the prosecutor could prevent a case from going to trial where there was sufficient evidence in the case for a jury properly to have convicted, there was a chance that life-endangering offences could go unpunished.

200 . More particularly, the applicant argued that there was a substantial chance that life-endangering offences in fact had gone unpunished in the present case because, had the threshold evidential test been lower, there would have been sufficient evidence to have led to a prosecution of a number of officers, including Charlie 2 and Charlie 12 for murder, and Commander McDowall, Commander Dick, Trojan 84, Trojan 80, DCI C and “James” for gross negligence manslaughter. The fact that the inquest jury, having heard oral testimony, returned an open verdict indicated that they were not satisfied that, at the time they fired, Charlie 2 and Charlie 12 honestly believed that Mr de Menezes represented an imminent, mortal danger.

201 . Relying on Enukidze and Girgvliani v. Georgia (no. 25091/07 , § 274, 26 April 2011), t he applicant submitted that the need to secure public confidence by ensuring accountability was particularly fundamental where a fatal shooting by a police officer was concerned and that confidence would be undermined by a perceived failure to prosecute public officials who were alleged to have violated Article 2 of the Convention. Consequently, it would be permissible to have a lower threshold for prosecutions for serious breaches of Convention rights by State agents than for other offences.

202 . The applicant further alleged that the level of scrutiny that the domestic courts applied to a decision not to prosecute was incompatible with Article 2 of the Convention. This was because, pursuant to the dicta in Manning , even if a court considering a claim for judicial review concluded that a prosecution was likely to succeed, it would only have to order such prosecution if there had been an error of law. Such approach was inconsistent with Article 2 of the Convention.

203 . In the alternative, the applicant argued that even if individual prosecutions were not required in the present case, the prosecution of the OCPM had not amounted to an adequate acknowledgment of responsibility on the part of the State, as the offence under the Health and Safety at Work etc. Act 1974 was established if there was a possibility of danger instead of actual danger, that is, since proof of actual harm was unnecessary to establish the offence, it was not necessary for the domestic court to determine whether any breach of duty in fact caused the death of Mr de Menezes. Consequently, despite the serious criticisms made in the IPCC report, and the verdict of the inquest jury, no individual or organisation had been held to account for Mr de Menezes ’ death.

204 . Although the applicant accepted that she could have brought a private prosecution, she argued that this would not have addressed her complaints because it was clear from the Court ’ s case-law that it was the State that had the responsibility for complying with Article 2.

205 . The applicant further submitted that disciplinary proceedings could not, by themselves, have complied with Article 2 of the Convention as they were essentially administrative proceedings intended to govern future employment. Where serious breaches of the Convention were concerned, effective protection had to be provided by the criminal law because the sanctions available were more punitive and had better deterrents than disciplinary proceedings. In a case such as the present, disciplinary proceedings could not have satisfied the procedural obligation under Article 2 of the Convention because there would have been a manifest disproportion between the gravity of the act and the punishment available.

206 . More particularly, the applicant contended that police disciplinary proceedings in the United Kingdom were often not sufficiently independent to satisfy the procedural limb of Article 2, as it was usually the Chief Officer of the officer ’ s own force who took all the key steps in the investigation, and the members of the panel which conducted the proceedings could also be from the same force. The proceedings were not conducted in public and at the time of Mr de Menezes ’ death police officers could avoid the disciplinary process by resigning.

(b) The Government ’ s submissions

(i ) The investigation

207 . The Government argued that the formulation of the law of self-defence in England and Wales struck an appropriate balance between permitting the use of force to prevent lethal attacks on the public and ensuring that any individuals who may be exposed to a real and immediate risk to life by any operational measures were protected. In doing so, it recognised that it was not for the courts, with the benefit of detached reflection, to substitute their own opinion for that of a police officer required to act in the heat of the moment.

208 . More particularly, the Government contended that the test of “absolute necessity” in Article 2 § 2 of the Convention ought to be assessed from the standpoint of the person wielding lethal force in self-defence without any requirement of reasonableness by reference to objectively established facts, that is, a person ought to be criminally liable for causing death only where he was aware that in the circumstances his conduct was not absolutely necessary. This was supported by the Court ’ s case-law, which provided that the use of force might be justified where it was based on an honest belief, perceived for good reasons, to be valid at the time ( McCann and Others v. the United Kingdom , 27 September 1995, § 200, Series A no. 324; Andronicou and Constantinou v. Cyprus , 9 October 1997, § 192, Reports of Judgments and Decisions 1997 ‑ VI; and Giuliani and Gaggio v. Italy [GC], no. 23458/02, § 178, ECHR 2011). The honest belief did not have to be shown to be reasonable by reference to objectively established facts, although the objective reasonableness of the belief would nevertheless be relevant in determining whether or not it was genuinely held. An honest belief could be held for good reasons even if, objectively, another person might consider the belief to be irrational or based on either a flawed premise or faulty perceptions.

209 . Finally, the Government argued that the applicant ’ s proposed change to the law could have far-reaching and counter-productive effects. In particular, if officers were liable to prosecution even when their use of force was legitimate based on their honest beliefs at the time, there could be a chilling effect on the willingness of officers to carry out essential duties where they might be required to act in the heat of the moment to avert a danger to life. Consequently, it could have a profoundly detrimental effect on their ability to act in defence of their own lives and the lives of others.

(ii ) The prosecutorial decisions

210 . The Government argued that as the investigative obligation was one of means and not result, Article 2 only required a prosecution where it was justified by the findings of the investigation. The effectiveness of the investigation could not therefore be assessed only by reference to whether it resulted in criminal or disciplinary proceedings against individuals. An effective investigation, carried out against an appropriate framework of criminal law, could lead to the conclusion that such proceedings would not be justified.

211 . Thus, the fact that no individual officer was prosecuted was not, properly viewed, a specific ground of complaint; the crucial question was why there were no individual prosecutions. In the present case, the reason was that none of the independent authorities who reviewed the case concluded that there was sufficient evidence to justify a prosecution for murder or manslaughter. All necessary measures had been taken to discharge the Article 2 duty and in such a case it was not for the Court to substitute its own assessment of the facts for those of the domestic authorities and courts ( Klaas v. Germany , 22 September 1993, § 29, Series A no. 269).

212 . With regard to prosecutorial decisions in England and Wales, the Government submitted that the CPS was an independent prosecution service. Although in certain other legal systems this function was carried out by judicial officers, such a system was not mandated by the Convention. On the contrary, Article 2 merely required that such decisions were taken independently on the basis of a thorough review of the evidence. Furthermore, the applicant was wrong to say that a decision was taken by the CPS without the benefit of hearing witnesses. In taking its decision, the CPS had the benefit of all the materials generated by the IPCC during its investigation, including witness statements, and it carried out a review of its decision following the inquest, during which all the key witnesses had given oral evidence.

213 . The Government further argued that the current threshold evidential test did not require prosecutors to be satisfied that there was a fifty percent or more prospect of conviction. The Explanatory Memorandum issued to prosecutors in 1994 made it clear that although it was not unreasonable to talk about a conviction being “more likely than not”, they should “resist the temptation to define the evidential test as ‘ a 51% rule ’ ” (see paragraph 164 above). This was because it was impossible to measure with arithmetical precision the probability or likelihood of a particular outcome in a criminal case as there were many variable factors and elements of complexity and uncertainty that defied accurate calculation.

214 . The correct test was whether or not there was a “realistic prospect of conviction” against each suspect on each charge; in other words, whether a reasonable and impartial court, properly directed and acting in accordance with the law, was more likely than not to convict the defendant of the charge(s) alleged. A “merits based” approach was therefore applied, in which the prosecutor essentially asked himself whether, on balance, the evidence was sufficient to merit a conviction taking into account what he or she knew about the defence case. In reaching a decision the prosecutor was required to undertake a thorough and conscientious review of the case and it was only when he or she considered on balance that the evidence was not sufficient to merit conviction that the case would not be prosecuted.

215 . The Galbraith test, on the other hand, was a very low threshold which would be met where there was “some evidence but it is of a tenuous character, for example, because of inherent weakness or vagueness or because it is inconsistent with other evidence”. In the Government ’ s view, if the Galbraith test were the threshold evidential test, prosecutions would have to be brought in cases where there was no realistic prospect of conviction and where the prosecutor considered that the case was unfounded.

216 . In any case, the Government noted that even if the threshold evidential test had been the Galbraith test, it would not necessarily have led to the prosecution of any individual officer. The Galbraith test was also the test used by coroners in deciding whether to leave a verdict to a jury at an inquest. Furthermore, the elements of the inquest verdict of unlawful killing were precisely the same as the elements of the crimes which the CPS had to consider in taking the decision on criminal charges. However, in the present case, after hearing all the relevant witnesses cross-examined at length, the coroner weighed the evidence against the Galbraith threshold and decided that it was not met.

217 . Moreover, the Government submitted that in England and Wales the threshold evidential test had been the subject of frequent and anxious consideration through public consultation and political scrutiny. Detailed reviews of the Code for Crown Prosecutors were carried out in 2003, 2010 and 2012 and during those reviews the threshold evidential test had not been subject to substantive criticism by the Equality and Human Rights Commission (EHRC) or any of the human rights organisations with an interest in criminal law. In the 2003 review the then Attorney General specifically considered whether a lower threshold should apply to deaths in custody but found little, if any, support for such an approach as it would be unfair – and inconsistent – to subject potential defendants in such cases to the burden of prosecution in the absence of a realistic prospect of conviction. Public confidence was maintained by prosecuting where the evidence justified it, and not prosecuting where it did not.

218 . In light of the fact that the threshold evidential test had been given careful and anxious scrutiny, the Government argued that it should be accorded a significant margin of appreciation in assessing the appropriate evidential thresholds for the initiation and continuation of criminal proceedings in all cases.

219 . This was particularly important in light of the primacy of the jury in the United Kingdom criminal justice system. Once a case was prosecuted, the trial judge could not remove it from the jury if the Galbraith test was satisfied, that is, if there was some evidence, however tenuous, on which the jury could properly come to the conclusion that the defendant was guilty. The trial judge could not, therefore, act as a filter for unmeritorious cases, and a higher threshold evidential test for bringing prosecutions was necessary to ensure that the emotional and financial costs of trial were not incurred simply because there was a bare possibility of conviction. In other words, it was particularly important to weed out weak cases at an early stage because cases which did go to trial were usually pursued right to the end.

220 . The Government also noted that significant procedural protections had been built into CPS practices in cases of police shootings or deaths in custody: the prosecutor had to write to the family of the victim to explain any decision; the family had to be offered a meeting with the prosecutor to explain the decision; all charging decisions had to be reviewed personally by the DPP and, if there was a decision not to proceed, if it was not plain beyond all doubt that there was no case to answer, advice had to be sought from senior independent counsel. In addition, since June 2013 the victim has had the right to request a review of the CPS decision, first by the local CPS office that made the decision, and then by means of an independent review by either the CPS Appeals and Review Unit or by the relevant CPS Chief Crown Prosecutor.

221 . Although the Government accepted that in practice prosecution of State agents for causing death were rare, they did not consider this to be a cause for concern. Fatalities caused by armed police officers did not normally require the prosecution of the officer. In England and Wales a rigorous approach to the use of firearms was adopted; in particular, firearms officers were subjected to a high level of screening, training, guidance and monitoring to ensure that they only discharged firearms when it was absolutely necessary to do so. This is evidenced by the statistics: from 2003/04 to 2012/13, the annual number of police operations in which the use of firearms was authorised ranged from 10,996 (in 2012/13) to 19,595 (in 2007/08). However, during the same period the annual number of incidents in which conventional firearms were discharged ranged from three (in 2006/07 and 2012/13) to nine (in 2005/06).

222 . The Government further argued that the remedy of judicial review was not intended to provide an appeal system on the merits of the prosecutorial decision. In this regard, the primary protection lay not with the judiciary but in the requirement that the initial decision be taken by an independent and qualified prosecutor exercising an impartial judgment based on a public and accessible policy, subject to scrutiny by the DPP. On an application for judicial review, the Administrative Court retained the power to intervene where a decision not to prosecute was based on an error of law or was otherwise irrational or procedurally flawed.

223 . Finally, the Government submitted that a real tension existed between the paradigm of criminal culpability based on individual responsibility and the increasing recognition of the potential for harm inherent in large-scale or complex activity where no one person was wholly to blame for what went wrong. Cases such as McCann and Others , cited above, indicated that it might be simplistic to attribute an Article 2 breach to the individuals who directly caused the death, especially in a case such as the present where the death resulted from failures in the overall system. In such cases, it would be inaccurate and unfair to ascribe blame to the individuals who happened to form the last link in the chain. It could also be dangerous, diverting attention away from the real problems in the system which could then go unremedied and create risks to life in future. The prosecution of the OCPM, on the other hand, enabled the issues of planning and execution to be directly addressed in the context of a criminal trial.

(c) The third-party intervener ’ s submissions

(i) The investigation

224 . The EHRC argued that the criminal-law provisions of England and Wales failed to ensure accountability for deaths occurring under the State ’ s responsibility. In particular, they submitted that the definition of self-defence in English law was drawn very widely, was partially subjective and was inconsistent with Article 2 of the Convention. The clear and constant case-law of the Court is that an “honest belief” must be founded on “good reason”; to permit State officials to escape punishment in criminal proceedings based on an honest but objectively unjustifiable belief was incompatible with the strict requirements of Article 2 of the Convention. The use of force should therefore be objectively justifiable, that is, law-enforcement officers should be required to make reasonable attempts to ascertain the true facts before using lethal force.

(ii ) The prosecutorial decisions

225 . The EHRC further submitted that the criminal-law provisions in England and Wales were inadequate because the threshold evidential test for bringing prosecutions was too high. Although the State should not be obliged to prosecute hopeless cases, it was arbitrary to set the test as high as it currently was and there could be no objection in principle if it were lower. Like the applicant, the EHRC also considered the Galbraith test to be a more appropriate threshold evidential test for bringing a prosecution.

226 . The EHRC argued that aligning the threshold evidential test for bringing a prosecution with that for leaving a case to the jury would not require the prosecution of every case of a potential violation of Article 2, no matter how weak the evidence. Hopeless or legally unmeritorious cases would not cross this threshold. However, a case which had an assessed forty-nine percent chance of conviction could not sensibly be described as one with little prospect of conviction. A criminal justice system which operated so as to preclude trial in circumstances where evidence existed upon which a properly directed jury could lawfully convict was not one which secured the full accountability required by Article 2.

227 . Moreover, the lowering of the threshold evidential test in Article 2 cases involving killing by State agents would not involve any irreversible prejudice. Every trial was thoroughly reviewed at the close of the prosecution case and the judge was duty bound to withdraw the case upon a successful submission of no case to answer, that is, if a properly directed jury on one view of the facts could not lawfully convict. If the threshold evidential test for bringing a prosecution were lowered, at worst some cases which would not be prosecuted under the existing test would be withdrawn by the judge at the close of the prosecution case. At best, some which would not have been brought to trial at all under the existing test might result in convictions of State agents for culpable homicide.

228 . In support of their submissions, the EHRC noted that between 1990 and 2014 there were fifty-five deaths caused by police shootings in England and Wales. However, since 1990 there has been no criminal conviction of an armed officer, even in those cases where an inquest jury recorded a verdict of unlawful killing. Indeed, between 1993 and 2005 there were thirty fatalities and only two prosecutions.

2. The Court ’ s assessment

(a) The procedural requirement in cases concerning the use of lethal force by State agents

(b) Application to the present case

241 . In the present case the applicant has not complained generally about the investigation, which was conducted by an independent body (the IPCC). In the course of the investigation, the IPCC secured the relevant physical and forensic evidence (more than 800 exhibits were retained), sought out the relevant witnesses (nearly 890 witness statements were taken), followed all obvious lines of enquiry and objectively analysed all the relevant evidence (see paragraph 49 above). Moreover, the deceased ’ s family were given regular detailed verbal briefings on the progress of the investigation and, together with their legal representatives, they were briefed on the IPCC ’ s conclusions (see paragraph 49 above). They were also fully briefed on the CPS ’ s conclusions (notably by means of a fifty-page Review Note and a follow-up final Review Note (see paragraphs 77, 79 and 133 above)), they were able to judicially review the decision not to prosecute, and they were represented at the inquest at the State ’ s expense, where they were able to cross-examine the seventy-one witnesses called and make representations.

(i) Adequacy of the investigation: were the authorities able properly to consider whether the use of force was justified?

( α ) The test applied by the Court

244 . T he test consistently applied by the Court in determining whether the use of lethal force was justified is set out in McCann and Others (cited above, § 200).

“[T]he use of force by agents of the State in pursuit of one of the aims delineated in paragraph 2 of Article 2 of the Convention may be justified under this provision where it is based on an honest belief which is perceived, for good reasons, to be valid at the time but which subsequently turns out to be mistaken. To hold otherwise would be to impose an unrealistic burden on the State and its law-enforcement personnel in the execution of their duty, perhaps to the detriment of their lives and those of others.”

245 . The Government have argued that the reasonableness of a belief in the necessity of lethal force should be determined subjectively. Although the applicant has accepted this, the third-party intervener has submitted that an honest belief should be assessed against an objective standard of reasonableness. It is, however, apparent both from the application of the stated test to the particular facts in McCann and Others itself and from the Court ’ s post McCann and Others case-law that the existence of “good reasons” should be determined subjectively. In a number of cases the Court has expressly stated that, as it is detached from the events in issue, it cannot substitute its own assessment of the situation for that of an officer who was required to react in the heat of the moment to avert an honestly perceived danger to his life or the lives of others; rather, it must consider the events from the viewpoint of the person(s) acting in self-defence at the time of those events (see, for example, Bubbins v. the United Kingdom , no. 20196/99, § 139, ECHR 2005-II, and Giuliani and Gaggio , cited above, §§ 179 and 188). Consequently, in those Article 2 cases in which the Court specifically addressed the question of whether a belief was perceived, for good reasons, to be valid at the time, it did not adopt the standpoint of a detached observer; instead, it attempted to put itself into the position of the person who used lethal force, both in determining whether that person had the requisite belief and in assessing the necessity of the degree of force used (see, for example, Makaratzis v. Greece [GC], no. 50385/99, §§ 65-66, ECHR 2004 ‑ XI; Oláh v. Hungary (dec.), no. 56558/00, 14 September 2004; and Giuliani and Gaggio , cited above, § 189).

( β ) Compatibility of the test applied in England and Wales

249 . In the present case the coroner described the test to be applied as follows (see paragraph 106 above).

“Did the officer honestly and genuinely believe that it was necessary for him to use force in defence of himself and/or others? This is a question of subjective belief. Even if the belief was mistaken, and even if the mistake was unreasonable, the defence can still run. The reasonableness of the belief is only relevant in helping the jury to decide whether the belief was honestly held.”

250 . Although the Court has previously considered the compatibility of this test with Article 2 of the Convention, those cases do not assist the Court in its consideration of the one at hand. It is true that i n Bennett v. the United Kingdom ((dec.), no. 5527/08, 7 December 2010) the Court expressly found that there was “no sufficiently great difference between the English definition of self-defence and the ‘ absolute necessity ’ test for which Article 2 provides”. However , t he issue in Bennett was whether the test applied by the coroner, namely that the use of lethal force should be “reasonably justified”, was compatible with the “absolute necessity” requirement in Article 2 of the Convention. The Court was not, therefore, called upon to consider the compatibility of domestic law with the requirement that an honest belief be perceived, for good reasons, to be valid at the time . The issue did arise in Caraher v. the United Kingdom ((dec.), no. 24520/94 , ECHR 2000 ‑ I), in which the Court found that the approach taken by the domestic judge was compatible with the principles established in McCann and Others . However, the approach of the domestic judge in Caraher differs somewhat from the one adopted in the present case, and the Government have accepted the latter to be an accurate reflection of domestic law.

251 . It is clear both from the parties ’ submissions and the domestic decisions in the present case that the focus of the test for self-defence in England and Wales is on whether there existed an honest and genuine belief that the use of force was necessary. The subjective reasonableness of that belief (or the existence of subjective good reasons for it) is principally relevant to the question of whether it was in fact honestly and genuinely held. Once that question has been addressed, the domestic authorities have to ask whether the force used was “absolutely necessary”. This question is essentially one of proportionality, which requires the authorities to again address the question of reasonableness, that is, whether the degree of force used was reasonable, having regard to what the person honestly and genuinely believed (see paragraphs 148 - 55 above).

252 . So formulated, it cannot be said that the test applied in England and Wales is significantly different from the standard applied by the Court in McCann and Others and in its post McCann and Others case-law (see paragraphs 244 ‑ 48 above). Bearing in mind that the Court has previously declined to find fault with a domestic legal framework purely on account of a difference in wording which can be overcome by the interpretation of the domestic courts (see Perk and Others v. Turkey , no. 50739/99, § 60, 28 March 2006, and Giuliani and Gaggio , cited above, §§ 214-15), it cannot be said that the definition of self-defence in England and Wales falls short of the standard required by Article 2 of the Convention.

253 . It is also clear that in the present case all the independent authorities considering the actions of Charlie 2 and Charlie 12 carefully examined the subjective reasonableness of their belief that Jean Charles de Menezes was a suicide bomber who might detonate a bomb at any second. In the Stockwell One Report the IPCC noted that the actions of Charlie 2 and Charlie 12 should be considered in light of the day ’ s events and those of the previous two weeks. In particular, it had regard to the SFO ’ s briefing, the positive identification of Mr de Menezes by the surveillance teams, the decision to go to State Red when the SFOs arrived at Stockwell, and the DSO ’ s order to “stop” Mr de Menezes (see paragraph 60 above).

254 . The CPS also had regard to the fact that the events at Stockwell “ happened in a matter of seconds” and there was “some independent evidence that supports the officers ’ accounts that they feared Jean Charles might detonate a bomb” (see paragraph 78 above). The CPS further noted that if Charlie 2 and Charlie 12 did genuinely believe that they were acting in self-defence, then the actions that they took in shooting Mr de Menezes dead would be reasonable and not unlawful (see paragraph 83 above).

255 . Similarly, the coroner made it clear that he had to consider the reasonableness of the belief of Charlie 2 and Charlie 12 that the use of force was necessary in order to decide whether or not it was honestly and genuinely held (see paragraph 106 above).

256 . Consequently, it cannot be said that the domestic authorities failed to consider, in a manner compatible with the requirements of Article 2 of the Convention, whether the use of force by Charlie 2 and Charlie 12 was justified in the circumstances.

(ii) Adequacy of the investigation: was it capable of identifying and – if appropriate – punishing those responsible?

258 . As noted at paragraph 241 above, there is nothing before the Court to suggest that in the present case the domestic authorities failed to secure the relevant physical or forensic evidence, or to seek out relevant witnesses or relevant information. Furthermore, the secured evidence was thoroughly analysed and assessed by the IPCC, an independent investigatory body which took witness statements from nearly 890 people and collected more than 800 exhibits; by the CPS; by a judge and jury during the criminal trial of the OCPM, at which forty-seven witnesses were called to give evidence; and by a coroner and jury during the inquest at which seventy-one witnesses were called (see paragraphs 45 - 71 , 77-101 and 103 - 27 ). The applicant has not sought to argue the contrary. Therefore, the sole issue before the Court is whether the decision not to prosecute individual officers, and to prosecute only the OCPM in its capacity as an employer of police officers, could itself constitute a procedural breach of Article 2 of the Convention.

259 . To date, the Court has not faulted a prosecutorial decision which flowed from an investigation which was in all other respects Article 2 compliant. In fact, it has shown deference to Contracting States both in organising their prosecutorial systems and in taking individual prosecutorial decisions. In Kolevi (cited above) the Court made it clear that

“[it] is not oblivious to the fact that a variety of State prosecution systems and divergent procedural rules for conducting criminal investigations may be compatible with the Convention, which does not contemplate any particular model in this respect ... Independence and impartiality in cases involving high-ranking prosecutors or other officials may be secured by different means, such as investigation and prosecution by a separate body outside the prosecution system, special guarantees for independent decision-making despite hierarchical dependence, public scrutiny, judicial control or other measures. It is not the Court ’ s task to determine which system best meets the requirements of the Convention. The system chosen by the member State concerned must however guarantee, in law and in practice, the investigation ’ s independence and objectivity in all circumstances and regardless of whether those involved are public figures.” ( § 208)

260 . Likewise, in Brecknell v. the United Kingdom (no. 32457/04, § 81, 27 November 2007), although the Court held that the initial investigative response lacked the requisite independence (and was therefore in breach of the procedural limb of Article 2) , it found no grounds on which to criticise a decision not to prosecute where it was not “apparent that any prosecution would have any prospect of success” and where it could not “impugn the authorities for any culpable disregard, discernible bad faith or lack of will”. In Brecknell the application was lodged nearly three decades after the death in issue; nevertheless, it clearly demonstrates the Court ’ s reluctance to interfere with a prosecutorial decision taken in good faith following an otherwise effective investigation .

261 . That being said, the Court has, on occasion, accepted that “institutional deficiencies” in the criminal justice or prosecutorial system may breach Article 2 of the Convention. In Kolevi (cited above, § 209) the Court found that such deficiencies in the prosecutorial system resulted in the absence of sufficient guarantees for an independent investigation into offences potentially committed by the Chief Public Prosecutor. In particular it found that the centralised structure of the prosecutorial system made it “practically impossible to conduct an independent investigation into circumstances implicating [the Chief Public Prosecutor]”. Although there was no such obstacle to an effective investigation in the present case, the applicant has argued that there were other obstacles preventing any meaningful prosecutions. If such obstacles existed, they could enable life-endangering offences to go unpunished and, as such, give rise to the appearance of State tolerance of – or collusion in – unlawful acts. Consequently, it will be necessary for the Court to consider each of the applicant ’ s submissions in turn in order to determine whether there were any “institutional deficiencies” giving rise to a procedural breach of Article 2 of the Convention.

( α ) The CPS

262 . In England and Wales the decision whether to prosecute is taken by a prosecutor in the CPS. The Government have asserted – and the applicant has not contested – that the CPS was independent for the purposes of Article 2 of the Convention. In serious cases such as the one at hand, the decision is taken by a senior prosecutor having first taken independent legal advice. The Court has never stated that the prosecutorial decision must be taken by a court (see, for example, Hugh Jordan , cited above, §§ 122-24, in which the Court did not take issue with the fact that the prosecutorial decision was taken by a public official). Indeed, in at least twenty-five Contracting States the decision to prosecute is taken by a public prosecutor (see paragraph 175 above). Consequently, the fact that the decision is taken by a public official is not problematic in and of itself, provided that there are sufficient guarantees of independence and objectivity.

263 . Furthermore, the Court does not consider that Maksimov (cited above) can be interpreted as authority for the proposition that prosecutors should hear oral testimony from witnesses before taking decisions. In that case, the prosecutorial decision had been taken without any independent investigatory body hearing oral testimony from important witnesses. The situation in the case at hand is quite different, as the IPCC, an independent investigatory body, had conducted a thorough investigation which included interviewing all relevant witnesses and the CPS had access to its findings in taking its prosecutorial decisions (see paragraph 80 above). Moreover, an examination and cross-examination of witnesses was conducted before the coroner at the inquest and he concluded that there was no evidence capable of being left to a jury that could establish unlawful killing in relation to any individual police officer (see paragraphs 103-27 above). Following the inquest, the CPS reviewed its original decision but concluded that there was still insufficient evidence to prosecute any individual (see paragraph 133 above). In such a case, there is nothing in the Court ’ s case-law to suggest that an independent prosecutor must also hear oral testimony before deciding whether or not to prosecute.

264 . Consequently, the Court does not consider that the applicant ’ s complaints concerning the role and organisation of the CPS disclose any “institutional deficiencies” which prevented the authorities from adequately securing the accountability of those responsible for the death of Mr de Menezes.

( β ) The threshold evidential test

266 . It is not in dispute that States should be permitted to have a threshold evidential test to prevent the financial and emotional costs of a trial being incurred where there are weak prospects of success. In Gürtekin and Others v. Cyprus ((dec.), nos. 60441/13 and 2 others, 11 March 2014) the Court implicitly recognised this:

“A prosecution, particularly on such a serious charge as involvement in mass unlawful killings, should never be embarked upon lightly as the impact on a defendant who comes under the weight of the criminal justice system is considerable, being held up to public obloquy, with all the attendant repercussions on reputation, private, family and professional life. Given the presumption of innocence enshrined in Article 6 § 2 of the Convention, it can never be assumed that a particular person is so tainted with suspicion that the standard of evidence to be applied is an irrelevance. Rumour and gossip are a dangerous basis on which to base any steps that can potentially devastate a person ’ s life.” (§ 27)

267 . Moreover, for the following reasons the Court considers that Contracting States should be accorded a certain margin of appreciation in setting that threshold.

268 . Firstly, in setting the threshold evidential test the domestic authorities are required to balance a number of competing interests, including those of the victims, the potential defendants and the public at large and those authorities are evidently better placed than the Court to make such an assessment. In this regard, it is clear that the threshold applied by prosecutors in England and Wales is not an arbitrary one. On the contrary, it has been the subject of frequent reviews, public consultations and political scrutiny. In particular, detailed reviews of the Code were carried out in 2003, 2010 and 2012. It is also a threshold that applies across the board, that is, in respect of all offences and by whomsoever they were potentially committed.

269 . Secondly, there is no uniform approach among Contracting States with regard to the threshold evidential test employed in their legal systems. A written threshold evidential test exists in at least twenty-four Contracting States (see paragraph 176 above). In principle, in twenty of those States the threshold test focuses on the sufficiency of evidence against the suspect; however, in practice it is impossible to state with any certainty that the prosecutorial decision-makers in those States do not also take into consideration the prospect of securing a conviction. In the four countries where the test expressly focuses on the prospect of conviction, the tests differ. In Austria, the test is “the likelihood of conviction”; in Iceland, the question is whether the evidence is “sufficient or probable for conviction”; in the former Yugoslav Republic of Macedonia, it is whether there exists “enough evidence from which the prosecutor can expect a conviction”; and in Portugal, “a reasonable possibility of imposing a penalty at trial” (see paragraph 178 above).

270 . In any event, the threshold evidential test has to be viewed in the context of the criminal justice system taken as a whole. While the threshold adopted in England and Wales may be higher than that adopted in certain other countries, this reflects the jury system that operates there. Once a prosecution has been brought, the judge must leave the case to the jury as long as there is “some evidence” on which a jury properly directed could convict, even if that evidence is “of a tenuous nature” (this being the so ‑ called Galbraith test – see paragraph 166 above). As weak or unmeritorious cases cannot be filtered out by the trial judge, the threshold evidential test for bringing a prosecution may have to be a more stringent one. In this regard, it is significant that other common-law countries appear to have adopted a similar threshold to the one applied by prosecutors in England and Wales (see paragraphs 182 - 85 above).

271 . In the circumstances, it cannot be said that the threshold evidential test in England and Wales was so high as to fall outside the State ’ s margin of appreciation. In Brecknell , a case concerning Northern Ireland, the Court indicated that Article 2 did not require States to prosecute cases where it was not apparent that prosecution would have any prospect of success (see paragraph 260 above). This is very similar to the test of “realistic prospect of conviction” applied in England and Wales and the fact that it has subsequently been interpreted by the domestic courts and authorities to mean that a conviction should be “more likely than not” does not, in the Court ’ s opinion, suffice to bring it outside the State ’ s margin of appreciation. In any case, it is impossible to state with any certainty that the test in England and Wales is higher than those employed in the four member States which also have a threshold focusing on the prospect of conviction (see paragraph 178 above).

272 . The applicant has suggested that the threshold should be lower in cases involving the use of lethal force by State agents. However, there is nothing in the Court ’ s case-law to support this proposition. Although G ürtekin did not concern unlawful killing by State agents, in that case the Court made it clear that the fact that a crime engaging Article 2 of the Convention was particularly “serious” (in that case, mass killings) was not a sufficient reason to prosecute individuals regardless of the strength of the evidence. On the contrary, it found that since the consequences of a prosecution on such serious charges would be particularly severe for any defendant, it should not be lightly embarked upon (see paragraph 266 above).

273 . The same considerations apply in cases concerning the use of lethal force by State agents. It is true that public confidence in both the law-enforcement agencies and the prosecution service could be undermined if State agents were not seen to be held accountable for the unjustifiable use of lethal force. However, such confidence would also be undermined if States were required to incur the financial and emotional costs of trial in the absence of any realistic prospect of conviction. The authorities of the respondent State are therefore entitled to take the view that public confidence in the prosecutorial system is best maintained by prosecuting where the evidence justifies it and not prosecuting where it does not (see paragraph 217 above).

274 . In any case, it is clear that the domestic authorities have given thorough consideration to lowering the threshold in cases engaging the responsibility of the State, but decided that it would be both unfair and inconsistent to place an increased burden on potential defendants in these cases. Nevertheless, they did ensure that a number of safeguards were built into the system in cases of police shootings and deaths in custody: the DPP personally reviews all charging decisions; in all cases other than the most straightforward a decision not to prosecute has to be reviewed by independent counsel; the prosecutor has to write to the family of the victim to explain his or her decision; and the family has to be offered a meeting with the prosecutor to explain the decision (see paragraph 220 above). While it is true that there are not frequent prosecutions for police killings in the United Kingdom (as submitted by the third-party intervener at paragraph 228 above), this can be explained by the extremely restrictive policy on the use of firearms by State agents (see paragraph 221 above). As the Government have pointed out, between 2003/04 and 2012/13, the annual number of police operations resulting in the discharge of weapons has always been in single figures, even though the annual number of operations in which the use of weapons has been authorised has ranged from ten thousand to twenty thousand (see paragraph 221 above).

275 . Furthermore, in the present case i t is by no means certain or even likely that individual police officers could have been prosecuted had the Galbraith test been the threshold evidential test for bringing a prosecution against them (that is, if there only had to be some evidence, even if it was of a tenuous character). Indeed, on the facts of the present case the contrary is so, given that the self-same test was used by the coroner, an independent judicial officer, in deciding whether to leave a verdict of unlawful killing to the jury and, after hearing seventy-one witnesses, he concluded that the test was not satisfied in relation to any of the individual police officers concerned (see paragraphs 103 - 27 above). That being said, even if individual prosecutions had been possible had the threshold evidential test been the Galbraith test, it would not follow that the threshold in England and Wales was so high as to be in breach of Article 2.

276 . In light of the above, the Court does not consider that the threshold evidential test applied in England and Wales constituted an “institutional deficiency” or failing in the prosecutorial system which precluded those responsible for the death of Mr de Menezes from being held accountable.

( γ ) Review of prosecutorial decisions

277 . As already noted, a decision not to prosecute is susceptible to judicial review in England and Wales but the power of review is to be sparingly exercised; the courts can only interfere if a prosecutorial decision is wrong in law (see paragraph 165 above).

278 . Nevertheless, the Court is not persuaded by the applicant ’ s argument that the scope of review is too narrow. In G ürtekin (cited above, § 28) the Court noted that the procedural obligation in Article 2 did not necessarily require that there should be judicial review of investigative decisions, although such reviews were doubtless a reassuring safeguard of accountability and transparency. The Court further noted that it was not its role to micro-manage the functioning of, and procedures applied in, criminal investigative and justice systems in Contracting States which may well vary in their approach and policies. Likewise, in Mustafa Tunç and Fecire Tunç (cited above, § 233), the Court held that while the intervention of a court or a judge enjoying sufficient statutory safeguards of independence was a supplementary element enabling the independence of the investigation as a whole to be guaranteed, it was not in itself an absolute requirement.

279 . According to the information available to the Court, the decision not to prosecute is susceptible to some form of judicial review or appeal to a court of law in at least twenty-five Contracting States and in these countries the standard of review varies considerably. In seven of these countries the decision must first be contested before a hierarchical superior in the prosecution service. In twelve countries, the decision of the prosecutor may only be contested before such a hierarchical superior (see paragraph 181 above). Consequently, it cannot be said that there is any uniform approach among member States with regard either to the availability of review or, if available, the scope of that review.

280 . In England and Wales there was, at the relevant time, a right to have prosecutorial decisions judicially reviewed by an independent court. In view of the fact that the prosecutorial decision in the case at hand was made by a senior independent prosecutor, having first taken independent legal advice, and the reasons for that decision were fully explained to the family of the deceased, the Court finds nothing in its case-law which would support the applicant ’ s assertion that Article 2 required the Administrative Court to have greater powers of review. In any event, the Court notes that in the present case the Administrative Court had regard to this Court ’ s case-law, in particular the requirement of “careful scrutiny” enunciated in Öneryıldız . Moreover, it did not simply find that the prosecutor ’ s decision had not been irrational; although not required to go so far, it expressly indicated that it agreed with the prosecutor ’ s conclusions (see paragraph 98 above).

281 . In light of the above, the Court does not consider that the scope of judicial review of prosecutorial decisions in England and Wales could be described as an “institutional deficiency” which impacted upon the ability of the domestic authorities to ensure that those responsible for the death of Mr de Menezes were held to account.

(δ) Partial conclusion

282 . Accordingly, having regard to the criminal proceedings as a whole, the applicant has not demonstrated that there existed any “institutional deficiencies” in the criminal justice or prosecutorial system which gave rise – or were capable of giving rise – to a procedural breach of Article 2 of the Convention on the facts of the instant case.

(iii) Overall conclusion on the applicant ’ s Article 2 complaint

283 . The facts of the present case are undoubtedly tragic and the frustration of Mr de Menezes ’ family at the absence of any individual prosecutions is understandable. However, it cannot be said that “any question of the authorities ’ responsibility for the death ... was left in abeyance” (compare, for example, Öneryıldız , cited above, § 116, in which there had been no recognition of the responsibility of the public officials for the death of the applicant ’ s relatives ). As soon as it was confirmed that Mr de Menezes had not been involved in the attempted attack on 21 July 2005, the MPS publicly accepted that he had been killed in error by SFOs. A representative of the MPS flew to Brazil to apologise to his family face-to-face and to make an ex gratia payment to cover their financial needs. They were further advised to seek independent legal advice and assured that any legal costs would be met by the MPS. The individual responsibility of the police officers involved and the institutional responsibility of the OCPM were subsequently considered in depth by the IPCC, the CPS, the criminal court and the coroner and jury during the inquest. Later, when his family brought a civil claim for damages, the MPS agreed to a settlement with an undisclosed sum being paid in compensation.

287 . In light of this conclusion, it is not necessary for the Court to consider the role of private prosecutions or disciplinary proceedings in fulfilling the State ’ s procedural obligations under Article 2 of the Convention.

288 . Accordingly, the Court finds that in the present case no violation of the procedural aspect of Article 2 of the Convention has been established.

II. ALLEGED VIOLATION OF ARTICLE 3 OF THE CONVENTION

289 . The applicant complained that the decision not to prosecute any individual for the death of her cousin also constituted a procedural violation of Article 3 of the Convention.

290 . In the absence of any evidence to suggest that Mr de Menezes was subjected to ill-treatment within the meaning of Article 3, the Court considers this complaint to be manifestly ill-founded within the meaning of Article 35 § 3 (a) of the Convention.

III. ALLEGED VIOLATION OF ARTICLE 13 OF THE CONVENTION READ TOGETHER WITH ARTICLES 2 AND/OR 3

292 . As the essence of the applicant ’ s complaint is that no individual was prosecuted for her cousin ’ s death, the Court considers that it more properly falls to be considered under the procedural aspect of Article 2 of the Convention.

293 . Accordingly, the Court also considers this complaint to be manifestly ill-founded within the meaning of Article 35 § 3 (a) of the Convention.

FOR THESE REASONS, THE COURT

1. Declares , unanimously, the complaint under Article 2 of the Convention admissible;

2. Declares , by a majority, the remainder of the application inadmissible;

3. Holds , by thirteen votes to four, that there has been no violation of the procedural limb of Article 2 of the Convention.

Done in English and in French, and delivered at a public hearing in the Human Rights Building, Strasbourg, on 30 March 2016.

Lawrence Early Guido Raimondi Jurisconsult President

In accordance with Article 45 § 2 of the Convention and Rule 74 § 2 of the Rules of Court, the following separate opinions are annexed to this judgment:

(a) dissenting opinion of Judges KarakaÅŸ,Wojtyczek and Dedov;

(b) dissenting opinion of Judge López Guerra.

G.RA. T.L.E.

JOINT DISSENTING OPINION OF JUDGES KARAKAÅž, WOJTYCZEK AND DEDOV

1. We respectfully disagree with the majority because in our view there has been a violation of Article 2 under its procedural limb in the instant case.

2. In the instant case the applicant has not complained of a violation of Article 2 under its substantive limb. In this respect the case has been settled between the parties with the payment of compensation to the victim ’ s family. We note, however, that had the case been brought under the substantive limb of Article 2, the Court would have had to find a violation of this provision. The substantive and procedural issues are so closely intertwined in the instant case that it is impossible to assess whether the respondent State has fulfilled its obligation under the procedural limb of Article 2 without taking into account the substantive dimension of the case.

3. In assessing compliance by the respondent State with its obligations, it is important to bear in mind the international standards on the use of force by the police. “Law enforcement officials may use force only when strictly necessary and to the extent required for the performance of their duty” (Article 3, Code of Conduct for Law Enforcement Officials, adopted by United Nations General Assembly Resolution 34/169 of 17 December 1979). “In general, firearms should not be used except when a suspected offender offers armed resistance or otherwise jeopardizes the lives of others and less extreme measures are not sufficient to restrain or apprehend the suspected offender” (Official commentary on Article 3 of the Code of Conduct). “Law enforcement officials shall not use firearms against persons except in self-defence or defence of others against the imminent threat of death or serious injury, to prevent the perpetration of a particularly serious crime involving grave threat to life, to arrest a person presenting such a danger and resisting their authority, or to prevent his or her escape, and only when less extreme means are insufficient to achieve these objectives. In any event, intentional lethal use of firearms may only be made when strictly unavoidable in order to protect life” (Principle 9, Basic Principles on the Use of Force and Firearms by Law Enforcement Officials, adopted by the Eighth United Nations Congress on the Prevention of Crime and the Treatment of Offenders, Havana, Cuba, 27 August to 7 September 1990).

For those reasons, if the police plan an operation which may require the use of firearms, they have the duty to act with the utmost care and in particular to meticulously check all the relevant information on which the operational plan is based. While planning their operations, the police also have the obligation to carefully assess the available alternatives and to choose the means which entail the least risk for human life and health.

4. Article 2 of the Convention, as interpreted by the Court, requires that States carry out an investigation of cases of alleged unlawful killing by State agents. According to the established case-law of the Court, an investigation must be capable of leading to the establishment of the facts and, where appropriate, the identification and punishment of those responsible. Substantive criminal law, by defining specific offences, indicates the exact purpose of an investigation into a person ’ s death and in particular determines the specific issues which have to be investigated. Therefore, the quality of the investigation depends first and foremost on the quality of substantive law. An investigation will be capable of leading to the identification and punishment of those responsible only if there are adequate provisions of substantive criminal law. Defective provisions of substantive law can render the investigation ineffective from the perspective of Article 2. In particular, if national provisions of criminal law concerning the use of force by the police do not comply with the Convention standards, the authorities will not be able to investigate whether the use of lethal force by the police was absolutely necessary under Article 2 of the Convention. The investigation may then focus on other issues of lesser importance from the viewpoint of the Convention.

5. The investigation in the instant case led to the conclusion that the members of the police force who killed Mr de Menezes had genuinely believed that he was about to detonate a bomb on the underground and that they were accordingly required to repel an imminent terrorist attack threatening the passengers. The killing of Mr de Menezes by members of the police can thus be characterised as an act committed in putative self ‑ defence. Force used in putative self-defence is never absolutely necessary.

Article 2 of the Convention requires that the substantive criminal law should ensure protection against excessive use of force by the police. This requirement of criminalisation does not mean that any use of force which is not absolutely necessary has to entail criminal liability. A person will bear criminal liability only if personal guilt can be shown. It would not be just to criminalise acts committed in putative self-defence if the factual error was justified in the specific circumstances and the person responsible could not be reproached for it. At the same time, in our view, Article 2 of the Convention requires the State to criminalise putative self-defence in so far as the factual error was not justified in the circumstances and the perpetrator may therefore legitimately be reproached for it. If acts of killing in putative self-defence based on an unjustified error are not properly criminalised and punished under domestic law, there is a serious danger that the police may use excessive force with lethal effect.

The Court has set the following standard which is relevant for assessing cases of putative self-defence.

“[T]he use of force by agents of the State in pursuit of one of the aims delineated in paragraph 2 of Article 2 of the Convention may be justified under this provision where it is based on an honest belief which is perceived, for good reasons , to be valid at the time but which subsequently turns out to be mistaken” (see McCann and Others v. the United Kingdom , 27 September 1995, § 200, Series A no. 324, emphasis added; see also Andronicou and Constantinou v. Cyprus , 9 October 1997, § 192, Reports of Judgments and Decisions 1997-VI; Brady v. the United Kingdom (dec.), no. 55151/00, 3 April 2001; Bubbins v. the United Kingdom , no. 50196/99, §§ 138-39, ECHR 2005-II; and Giuliani and Gaggio v. Italy [GC], no. 23458/02, §§ 178-79, ECHR 2011; see also the concurring opinion of Judge Pinto de Albuquerque in Trévalec v. Belgium , no. 30812/07, 14 June 2011).

Under this approach, national criminal law on putative self-defence is compliant with Article 2 if it provides for two cumulative conditions for exemption from criminal liability: a subjective one (an honest belief which subsequently turns out to be mistaken or, in other words, actual error as to factual circumstances) and an objective one (existence of good reasons for which the belief is perceived to be valid at the time or, in other words, the existence of objective grounds justifying the error). Acts committed in putative self-defence may be exempted from criminal liability if these two conditions are met jointly. However, in the instant case the majority seem to reinterpret the existing case-law by putting the emphasis on the subjective element and by diminishing the importance of the objective element. In our view, such an approach is not acceptable. It puts citizens ’ lives at risk in the context of police operations because acts committed by the police in putative self-defence as a result of gross negligence may become immune from criminal liability.

Furthermore, effective protection of the right to life under Article 2 of the Convention requires also that the substantive criminal law should ensure protection against gross negligence in the preparation and carrying out of police operations in which force is used.

6. The test applicable in English law for justified putative self-defence is a subjective one: “Did the officer honestly and genuinely believe that it was necessary for him to use force in defence of himself and/or others?” (see paragraph 249 of the present judgment). Therefore, the investigation in the instant case had to answer the question whether the police officers involved honestly and genuinely believed that Mr de Menezes was about to detonate a bomb. However, the crucial question which should have been investigated and answered in the instant case was whether the police officers ’ belief that a bomb was about to be detonated was justified in the circumstances. The investigation should have established whether the error of each officer involved, including those who directed the whole operation, was justified. Furthermore, the reasonableness of this belief should have been assessed in the context of the police ’ s duty to exercise the utmost care in preparing operations which may potentially entail the use of lethal force. Because of the content of the relevant substantive law, in the circumstances of the instant case the investigation did not focus on these crucial questions. Therefore, in the circumstances of the case, the use of force by the police officers concerned was not adequately investigated and the investigation was not able to lead to the punishment of those involved in using such force. More generally, under English law the investigation will not be adequate and will not always be able to lead to punishment in cases where police officers use lethal force in putative self-defence.

7. We also would like to draw attention to another important factual element in the instant case. The tragic events of the case took place within the context of a pre-planned police operation. It was the duty of the police to devise a realistic plan of action which made it possible to arrest the suspect without using lethal force. It appears that Mr de Menezes could and should have been arrested by the police just after leaving his home. It was the fact that the police officers waited until he entered the underground which caused the situation entailing a putative threat to the lives of a large number of people. In other words, the putative danger arose because of the delay in the reaction by the police. This factor is also of primary importance for establishing personal criminal liability on the part of the individuals involved. Even assuming that Mr de Menezes really was carrying a bomb, the delayed reaction of the police officers could not be considered absolutely necessary, because it appears that the suspect could have been apprehended much earlier. In our view, this aspect of the case has likewise not been properly investigated for the purpose of establishing criminal liability on the part of the individuals involved.

8. We note that in the instant case criminal proceedings were instituted against the police service. Criminal proceedings against legal entities may be helpful for establishing the facts. However, the Convention requires that criminal law should provide for the punishment of individuals and that an investigation should be able to lead to such punishment. Under the Convention, criminal liability of legal entities can never replace criminal liability of individuals. In the instant case, the criminal liability of the police service as such is not sufficient to satisfy the Convention criteria. Furthermore, gross negligence on the part of a legal entity always stems from the misconduct of specific individuals. It is difficult to understand that, in the instant case, the persons responsible for the negligence could not be prosecuted under English law.

9. In assessing the overall effectiveness of the investigation, regard should be had in our opinion to some important mistakes committed by the investigators at the very beginning. The IPCC expressed its concern regarding the delay in handing the scene and the investigation to it, and regarding the fact that Charlie 2 and Charlie 12 had been allowed to return to their own base, refresh themselves, confer and write up their notes together (see paragraph 69 of the present judgment). These mistakes might have affected the subsequent stages of the investigation.

It is worth recollecting here that the Court made the following assessment in Makbule Kaymaz v. Turkey (no. 651/10, § 141, 25 February 2014; translation from original French).

“The Court observes at the outset that ... the police officers implicated in the incident were not interviewed by the public prosecutor until 4 December 2004, more than ten days after the events. Furthermore, they were not kept apart after the incident and were called to give evidence in the context of the administrative investigation before the prosecuting authorities became involved. In this connection, the Court reiterates that in Bektaş and Özalp (cited above, § 65, seven days after the incident) and Ramsahai and Others (cited above, § 330, three days after the incident) it held that such delays not only created an appearance of collusion between the judicial authorities and the police, but could also lead the victims ’ relatives – and the public in general – to believe that members of the security forces operated in a vacuum and thus were not accountable to the judicial authorities for their actions. In the instant case, although there is no indication that the police officers in question colluded with each other or with their colleagues from the Mardin police, the mere fact that appropriate steps were not taken to reduce the risk of such collusion amounts to a significant shortcoming in the adequacy of the investigation (see Ramsahai and Others , cited above, § 330).”

10. According to the established case-law of the Court, an investigation carried out under Article 2 of the Convention should be prompt. This criterion has not been fulfilled in the instant case. The United Nations Human Rights Committee stated the following in its Concluding Observations, issued on 30 July 2008, on the report submitted by the United Kingdom under Article 40 of the International Covenant on Civil and Political Rights.

“10. The Committee is concerned at the slowness of the proceedings designed to establish responsibility for the killing of Jean Charles de Menezes and at the circumstances under which he was shot by police at Stockwell underground railway station (art.6).

The State party should ensure that the findings of the coroner ’ s inquest, due to begin in September 2008, are followed up vigorously, including on questions of individual responsibility, intelligence failures and police training. ” (UN Doc. CCPR/C/GBR/CO/6, paragraph 10, emphasis in the original.)

Many years elapsed after the events before such an investigation started. We cannot agree with the assessment of the majority that the criterion of promptness has been fulfilled in the instant case.

11. The applicant complained about the test for prosecution in cases of putative self-defence. Under English law, a prosecution will be brought only if a conviction is “more likely than not” (see paragraphs 164 and 265 of the present judgment). The majority refer in the reasoning to an interesting comparative-law report on this issue (see paragraphs 176 and 269 of the present judgment). We note in this connection that the analysis of comparative-law data leads to the conclusion that the test applied under English law for prosecution is clearly more stringent than in other States Parties to the Convention. Such a stringent test may prevent the prosecution and conviction of a person who has committed an offence if the prospects of success are not correctly assessed by the prosecutor. There is a serious risk that borderline cases will escape independent judicial assessment. As a result, certain acts involving excessive use of force by the police may be covered by a de facto immunity from prosecution. In our view, for the sake of efficient protection of the right to life, if there are serious doubts concerning the legitimacy of lethal force used by the police in actual or putative self-defence, the final decision on the question of criminal liability should be left to the courts.

12. We agree with the majority that in the instant case there was indeed an investigation which clarified many relevant aspects of the factual circumstances and triggered important reforms in the police. However, in our view, the combination of the different factors mentioned above led to a situation in which the death of an innocent person was not properly investigated in compliance with the Convention standards. The investigation carried out was not capable of leading to the establishment of individual criminal liability as required by the Convention.

13. Finally, we would like to note briefly that the majority considered that the complaint concerning the alleged violation of Article 13 of the Convention read together with Articles 2 and/or 3 was manifestly ill ‑ founded within the meaning of Article 35 § 3 (a) of the Convention. Judges KarakaÅŸ and Dedov voted with the majority on this point. Judge Wojtyczek considers that the reasoning on this issue is not persuasive. In his view, this part of the complaint should have been communicated to the parties and examined by the Court.

DISSENTING OPINION OF JUDGE LÓPEZ GUERRA

1. I disagree with the Grand Chamber judgment. I consider that in this case the United Kingdom authorities did not comply with the procedural requirement deriving from Article 2 of the Convention to conduct an adequate investigation into the responsibility of the State agents involved in the killing of Mr Jean Charles de Menezes.

The starting-point and basis for my disagreement with the Grand Chamber judgment is that no individual responsibility was derived from the established fact, acknowledged by the United Kingdom agencies, that there were very serious deficiencies in all aspects of the police operation resulting in Mr de Menezes ’ death.

The facts are adequately stated in the judgment, but I consider it essential to note several central aspects.

2. Suspects of previous bombings in the city of London were thought to be living in an apartment at 21 Scotia Road in London. In order to identify and arrest them safely, a strategy was devised consisting in following the persons leaving those premises in order to challenge and stop them. The officers in charge of the operation were to be supported by a group of highly trained special firearms officers (SFOs).

In this case, the entire operation went wrong. When Mr de Menezes (a Brazilian national living at 17 Scotia Road, with no connection to the bombings) left his apartment, the support unit from the Metropolitan Police Specialist Crime and Operations Branch (SO19) had not yet arrived. As a result, Mr de Menezes was not stopped. He was followed for over half an hour from Scotia Road to Stockwell underground station. The facts of the case indicate that during that time the surveillance team had not identified Mr de Menezes as a terrorist suspect. At Stockwell station, while in a stationary underground train, two members of the SO19 team shot him several times and killed him.

3. The deficiencies of the operation were extensively stated in a report by the Independent Police Complaints Commission (IPCC). The report concluded (see paragraph 66 of the present judgment) that in the course of the investigations grave concerns had been raised regarding the effectiveness of the police response, identifying a number of failings related to the different phases of the operation resulting in Mr de Menezes ’ death.

4. However, despite this detailed and extensive report, no individual responsibility for his death was ever established. This is particularly surprising, given that institutional criminal responsibility for Mr de Menezes ’ death was found in a court decision declaring that the Office of the Commissioner of the Police of the Metropolis (“the OCPM”) had contravened sections 3 and 33 of the Health and Safety at Work etc. Act 1974, for having exposed third parties to risks to their health and safety. Since the charges were directed against the OCPM as an institution, no responsibility was determined in those proceedings with respect to individuals.

5. Despite this finding of severe organisational deficiencies, other decisions provided a blanket exemption from any individual responsibility. The IPCC decided not to pursue disciplinary action against any of the eleven frontline or surveillance officers involved in the operation (see paragraphs 74 and 135 of the present judgment). Moreover, the IPCC did not issue any recommendation for the senior officers involved in the operation to face disciplinary proceedings.

All the individuals participating in the operation resulting in Mr de Menezes ’ death were not only free of any disciplinary liability; they were likewise exempt from criminal prosecution. On 17 July 2006 the Crown Prosecution Service (“the CPS”) decided that no individual was to be prosecuted in relation to the death of Mr de Menezes. That decision was confirmed by the Director of Public Prosecutions (“the DPP”) on 8 April 2009 (see paragraph 133 of the present judgment), since it deemed there was insufficient evidence to prosecute any individual.

6. It is difficult to understand how it is possible to establish that an institution (the OCPM) was criminally responsible (as adjudicated in a court of law) and, in spite of that, to exclude (as a consequence of the decisions of the IPCC and the CPS) all disciplinary liability and to preclude any effective investigation into the criminal responsibility of individual members of that institution.

7. In the light of the circumstances of this case, there is no justification for the United Kingdom ’ s failure to comply with its obligations deriving from the procedural dimension of Article 2 of the Convention, as consistently established in the Court ’ s case-law – that is, the obligation to conduct an effective investigation to establish the circumstances leading to intentional loss of life and to determine the possible punishment for those responsible for the death. The IPCC ’ s report acknowledging serious deficiencies in the police operation and the judgment finding the OCPM criminally responsible clearly provided a reasonable basis for investigating possible individual responsibilities for those organisational deficiencies, since organisations do not act independently of their members.

8. It cannot be concluded that the United Kingdom ’ s positive obligation was met merely because the authorities in charge of the initial investigation (the IPCC) and those that decided not to prosecute (the CPS) were deemed to be independent authorities for the purposes of Article 2 of the Convention (see paragraph 262 of the present judgment). Independence in itself is not enough to guarantee the existence of an effective investigation. In this case, what is missing are all of the other guarantees deriving from judicial proceedings in which evidence is publicly examined, with the intervention of all the affected parties, so that responsibilities may be ascertained accordingly. This was what the applicant sought when she asked the DPP to review the previous decision not to prosecute: to ensure the conduct of judicial proceedings with all the appropriate procedural guarantees, which go further than the independence of an administrative investigating body.

9. In other respects, the existence of a practice authorising the prosecution service to decline to bring criminal proceedings on the basis of the probability of achieving a guilty verdict (the so-called Manning test) is not in itself a sufficient reason to fail to determine responsibility, in judicial proceedings, for an intentional death. The question for our Court was not to rule whether the Manning test conformed to Convention requirements in the abstract, but rather whether the application of that test in this specific case represented a failure to comply with the procedural obligations under Article 2 of the Convention – that is, whether the CPS ’ s decision to refrain from bringing criminal proceedings against the individuals involved in the operation disregarded those obligations.

10. The Grand Chamber judgment accepts the reasonableness of the Manning test; but the Grand Chamber essentially takes into account its application to the two SFOs (Charlie 2 and Charlie 12) who killed Mr de Menezes. The responsibility of those two officers is not, however, the only or even the main question in the case, which relates to the responsibilities of all those involved in the police operation and the deficiencies in its planning and execution. Even admitting that there was a subjective perception of grave danger, justifying an honest belief in a situation of legitimate self-defence on the part of the two SFOs, the fundamental question remains concerning the responsibility of the other participants in the operation – that is, whether the SFOs ’ fatal subjective perception was the result of the previous actions or omissions of other individuals, and of the erroneous or deficient instructions they had received as a result of the mismanagement of a serious incident in which human lives were at stake, and in which the SFOs ’ briefing had indicated from the outset that “a critical shot could be taken” (see paragraph 26 of the present judgment).

11. In such circumstances, involving multiple subjects and actions at different levels, entailing considerable risk for human lives and findings of serious deficiencies, a complete investigation of possible individual responsibilities for those deficiencies should not have been precluded on the basis of a conjectural test applicable only to certain aspects of the police operation. In practice, the lack of such an investigation, with all the appropriate guarantees of adversarial and public proceedings, effectively granted immunity to those responsible for the serious and acknowledged errors resulting in Mr de Menezes ’ death.