England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
ADMINISTRATIVE COURT
IN THE MATTER OF CLAIMS FOR JUDICIAL REVIEW

		Royal Courts of Justice Strand, London, WC2A 2LL
		19th December 2014

Lord Justice Richards :

1. These two linked applications for judicial review challenge the lawfulness of guidance relating to the post-incident management of investigations into deaths that follow the use of force by police officers. One claimant is the father of Rafal Delezuch, a young man who suffered a cardiac arrest and died in hospital in August 2012 soon after he had been forcibly restrained and detained by police officers. The other claimant is the mother of Mark Duggan, who died in August 2011 in circumstances that have been the subject of extensive publicity: the vehicle in which he was travelling was stopped by police officers, one of whom shot him twice, fatally injuring him.
2. In each case the main argument is that in failing to require the immediate separation of the officers who either used force or witnessed its use, the guidance is unlawful, because of the risk of deliberate collusion or innocent contamination of evidence if officers have an opportunity to confer with one another before they make statements for the purposes of the investigation.
3. In each case permission to apply for judicial review was originally refused on the papers by a judge of the Administrative Court and at an oral renewal by the Divisional Court. At the hearing of an application for permission to appeal, a different constitution of this court granted permission to apply for judicial review and directed that the matter be retained in the Court of Appeal for substantive consideration (see [2014] EWCA Civ 388). That is the route by which the substantive applications for judicial review came to be listed before us for the equivalent of a first instance hearing.
4. In the Delezuch case the original target of challenge was directions issued by the Chief Constable of Leicestershire Constabulary to his officers and applied at the date of Rafal Delezuch's death. Those directions were based on guidance issued by the Association of Chief Police Officers ("ACPO") in 2009. The ACPO guidance was subject thereafter to a number of revisions before it was replaced in October 2013 by guidance issued by the College of Policing as part of the College's "Armed Policing Authorised Professional Practice" manual. An amended version of the College of Policing guidance was produced in September 2014 ("the 2014 guidance"). Mr Boyle, counsel for the Chief Constable, informed us on instructions during the hearing that the Chief Constable had now adopted the 2014 guidance. In those circumstances, and without objection from the Chief Constable, we permitted an amendment to the claim so as to focus the challenge onto the 2014 guidance. Whilst the guidance relates in terms to armed deployments, the case has proceeded on the basis that the procedures in it are applied to other situations involving death following the use of force by officers.
5. In the Duggan case the original target of challenge was the ACPO guidance in force at the date of Mark Duggan's death, on the basis that the guidance defined national standards and was intended to be followed by police forces. In this case, too, the focus of attention has now switched to the 2014 guidance issued by the College of Policing.
6. There is no useful purpose to be served by dwelling on the details of the earlier versions of the guidance. The important question is whether the current guidance is lawful. There is an issue as to whether the claimants have a sufficient interest in maintaining a challenge to the current guidance. As to that, I am satisfied that each of them had a sufficient interest in challenging the guidance applicable at the time of the investigation into their own son's death; and since they contend that the matters of complaint have not been remedied by subsequent changes to the guidance, they retain a sufficient interest in challenging the guidance in its current form. The defendants acknowledge the value of a ruling by the court on the issues now before it and their objections on grounds of standing were not pursued with any vigour.
7. At the heart of each claim is the obligation under article 2 of the European Convention on Human Rights to conduct an independent and effective investigation where a death has arisen in circumstances where force was used by state actors. There are, however, significant differences in the legal basis on which the claims are advanced. The contention in the Delezuch claim is that the policy applied by the Chief Constable (i.e. now the 2014 guidance) creates an unacceptable risk that the investigation into a death will fail to meet the procedural requirements of article 2 and is therefore unlawful. Whilst that way of putting the case is also adopted in the Duggan claim in relation to the 2014 guidance, a separate basis on which that claim is advanced is that ACPO has acted in breach of article 2 by adopting guidance that limits the ability of the Independent Police Complaints Commission ("the IPCC"), the independent investigator required by article 2, to set its own parameters to an investigation. As explained below, the IPCC itself has now issued draft statutory guidance, which has been the subject of consultation but not of any final decision, the substance of which is materially different from the 2014 guidance and is broadly in line with the result for which the claimants contend.

Investigations: the domestic regulatory framework

8. The IPCC was created by the Police Reform Act 2002 ("the 2002 Act") to conduct investigations under Part 2 of the 2002 Act into complaints against the police, conduct matters and death or serious injury matters. A death or serious injury matter ("DSI") is defined by section 12(2A) as "any circumstances (other than those which are or have been the subject of a complaint or which amount to a conduct matter) – (a) in or in consequence of which a person has died or has sustained serious injury; and (b) in relation to which the requirements of either subsection (2B) or subsection (2C) are satisfied". The relevant requirements are those in subsection (2C), namely that at or before the time of the death or serious injury the person had contact (of whatever kind, and whether direct or indirect) with a person serving with the police who was acting in the execution of his duties, and there is an indication that the contact may have caused (whether directly or indirectly) or contributed to the death or serious injury.
9. Section 15 imposes various duties on local policing bodies and chief officers of police. They include, by subsection (4), a duty on the chief officer of every police force to provide the IPCC and every member of the IPCC's staff with all such assistance as the IPCC or that member of staff may reasonably require for the purposes of, or in connection with, the carrying out of any such investigation.
10. Part 2A of Schedule 3 makes provision for the handling of DSI matters. Within Part 2A, paragraph 14A provides that where a DSI matter comes to the attention of the police authority or chief officer who is the appropriate authority in relation to that matter, it shall be the duty of the appropriate authority to record that matter. Paragraph 14B imposes duties to obtain and preserve evidence, including the following:

"(2) Where:

(a) a chief officer becomes aware of a DSI matter, and

(b) the relevant officer in relation to that matter is a person under his direction and control,

it shall be his duty to take all such steps as appear to him to be appropriate for the purposes of Part 2 of this Act for obtaining and preserving evidence relating to that matter.

(3) The chief officer's duty under sub-paragraph (2) must be performed as soon as practicable after he becomes aware of the matter in question.

(4) After that, he shall be under a duty, until he is satisfied that it is no longer necessary to do so, to continue to take the steps from time to time appearing to him to be appropriate for the purposes of Part 2 of this Act for obtaining and preserving evidence relating to the matter.

…

(6) It shall be the duty of the chief officer to take all such specific steps for obtaining or preserving evidence relating to any DSI matter as he may be directed to take for the purposes of this paragraph by the police authority maintaining his force or by the Commission."

By paragraph 14C it is the duty of the appropriate authority to refer every DSI matter to the IPCC, and by paragraph 14D it is the duty of the IPCC to determine whether or not it is necessary for the matter to be investigated.

The 2014 guidance

11. The 2014 guidance is a lengthy document. It contains sections under the headings of post-deployment procedures, post-incident procedures, immediate post-incident command considerations, management at the scene, post-incident management, welfare considerations, article 2 ECHR and the duty to investigate, providing accounts, security and welfare of officers, documentation and disclosure, and post-incident responsibilities. The following summary picks out passages of particular relevance to the issues before us.
12. In relation to post-deployment procedures the guidance states that the procedures are designed to ensure that all armed deployments are conducted in a manner which, inter alia, "ensures the integrity of the legal process in respect of police action, persons arrested or evidential material seized".
13. As to post-incident procedures, the initial action required is that the force control room or tactical firearms officer (if appointed) must be informed immediately and arrangements should also be made to ensure that the strategic firearms commander and appropriate ACPO officer on duty or on call are informed as soon as possible. The overall responsibility for post incident procedures should rest with an ACPO officer or senior officer (who has not been involved in the operational phase). This officer is responsible for initiating the post-incident investigation (including informing the IPCC) and post-incident management.
14. Immediate post-incident command considerations include a requirement on the tactical firearms commander initially to establish what has taken place and to take action as appropriate to ensure the adequate deployment of resources to deal with the situation, continuity of command of any ongoing crime-in-action, integrity of process in relation to securing best evidence, notification of senior command and independent investigative authorities, and consideration of community impact. That leads into the section on management at the scene, which includes matters such as making sure that the scene is safe and is secured.
15. The section on post-incident management deals first with "principal officers", the absence of a clear definition of which is one of the lesser matters of complaint before us. It reads:

"Principal officers

During a post incident investigation, the IIA [Independent Investigative Authority] will, at an early stage, wish to identify the principal officers directly related to the decision to use force.

As the investigation unfolds, others involved in the operation, whose actions or decisions were involved in informing or making critical decisions, may be regarded as principal officers. In the initial stages all actions taken by, and in respect of, principal officers in relation to securing evidence, discussion undertaken and notes made must be documented."

16. The section goes on to deal with the role of the post incident manager ("PIM"). In order to be assessed to be occupationally competent to perform the role of a PIM, an officer has to have completed satisfactorily a course of instruction based on the post-incident management module in the National Police Firearms Training Curriculum. The PIM, in conjunction with other relevant senior officers, should make an early decision on which officers will be considered principal officers. His responsibilities normally commence following the return of the principal officers to a police station or other area where the post-incident procedures will take place. His role at that stage is to facilitate the investigation, ensure integrity of process and ensure that the principal officers' needs are addressed in a manner which does not compromise the investigative process.
17. The section on welfare considerations states that the welfare needs of officers should be addressed throughout the post-incident procedure and that considerations will include medical assistance, securing weapons and equipment, provision of refreshments, making phone calls to immediate family members or partners regarding wellbeing and possible retention on duty, and showering and appropriate change of clothes (subject to forensic issues).
18. The section on article 2 and the duty to investigate summarises the basic legal requirements in a manner to which no objection is taken.
19. The section on providing accounts takes one to the heart of the present case. It states that there will be a requirement for those involved in an incident to provide relevant information in a number of formats and that it may be helpful to consider the provision of information and accounts chronologically as a staged process, as follows:

i) Stage one – situation report. This is likely to be by radio communication from an officer at the scene of the incident, giving sufficient information to enable the tactical firearms commander to manage the ongoing incident and assist officers to discharge their post-incident responsibilities.

ii) Stage two – PIM basic facts. The PIM is responsible for establishing the basic facts of what happened, where possible from a source other than a principal officer. The facts obtained should be sufficient to confirm which officers were at the scene, to describe in brief the roles of those at the scene, and to confirm who discharged their weapons. Special provision is made for the situation where the only person able to give the basic facts has been or may be identified as a principal officer.

iii) Stage three – personal initial accounts. The guidance describes this stage as follows: "Subject to legal and medical advice, officers should provide a personal initial account of the incident before going off duty. Each officer's initial account should consist only of their individual recollection of events and should be written, signed and dated. Detailed accounts will be made later. The purpose of the personal initial account is to record their role, what they believe to be the essential facts and should, where relevant, outline the honestly held belief that resulted in their use of force. The same guidance relating to conferring applies to both personal accounts and detailed accounts."

iv) Stage four – detailed accounts, statements and interviews. The guidance states the following in relation to this stage: "Detailed accounts should not normally be obtained immediately. They can be left until the officers involved in the shooting are better able to articulate their experience in a coherent format. This is usually after at least forty-eight hours. The detailed account should include, if relevant, why the witness considered the use of force and discharge of firearms to be absolutely necessary."

20. It is stated that the particular circumstances of the incident may make it unnecessary or inappropriate to include all of the stages. Where a principal officer is unfit to make a personal initial account, stage three will not be appropriate. Where a principal officer is suspected of a criminal offence or misconduct, stages three and four will not apply.
21. After referring to those four stages, the guidance contains important sections on conferring, integrity, process initiation and management, and separation, which it is appropriate to set out in full:

"Conferring

As a matter of general practice, officers should not confer with others before making their accounts (whether initial or subsequent accounts). There may, however, be a need [to] speak to one another following the discharge of a police firearm in order to resolve an ongoing operational or safety matter. The important issue is to individually record what their honestly held belief of the situation was at the time force was used. There should, therefore, be no need for an officer to confer with others about what was in their mind at the time force was used. If, however, in a particular case a need to confer on other issues does arise, then, in order to ensure transparency and maintain public confidence, where some discussion has taken place, officers must document the fact that this has taken place, highlighting:

time, date and place where conferring took place;

the issues discussed;

with whom;

the reasons for such discussion.

Integrity

The College of Policing (2014) Code of Ethics is a code of practice for the principles and standards of professional behaviour for the policing profession of England and Wales. The Code of Ethics provides a positive obligation on police officers and the police service to ensure that all activity, including that which relates to the recording of accounts, is transparent and capable of withstanding scrutiny. Where an officer has concerns that the integrity of the process is not being maintained, they must immediately draw this to the attention of the person in charge of the post incident process and ensure that this is documented.

Process initiation and management

An officer of ACPO rank is ultimately responsible for the initiation and management of the post incident procedure. This will include the appointment of an officer to supervise the preparation of officers' evidence and ensure compliance with the guidance on conferring prior to their accounts being recorded.

The officer appointed should, where practicable, be of substantive inspector rank (or equivalent) and be an accredited PIM. They must be appointed in addition to the PIM and not have held an operational or investigatory role in the originating incident. Their responsibility for ensuring compliance with guidance begins when the principal officers arrive at the post incident suite. It includes supervising those principal officers while personal initial accounts and detailed accounts are made (unless such accounts are provided during an interview or are recorded).

The appointed officer will operate under the direction of the PIM and may, where appropriate, support the PIM in discharging their responsibilities related to the provision of accounts.

The appointed officer must fully record their observations and actions. The following support may be appropriate:

establishing principal officers' legal status

ensuring access to legal representation/staff association or DLS where necessary or requested

explaining the conferring guidance contained in APP [Authorised Professional Practice] on armed policing to principal officers

ensuring that any conferring that takes place is compliant with APP and documented by those conferring

preventing any inappropriate conferring (eg, related to an individual's honestly held belief at the time force was used)

ensuring that any reference material used by the officers is secured and handed to investigators against a receipt/ exhibit number

documenting the process by which accounts are provided (including interruptions, breaks and abstractions)

ensuring that all accounts are time stamped by those making them as soon as practicable after completion.

Where access to the post incident suite is requested by the independent investigatory authority or initial investigating officer, such access should be provided. This access should be managed so as not to adversely impact on the post incident process or the welfare of principal officers, and will not relate to medical examinations or consultation with legal or staff association representatives.

Separation

Principal officers and key police witnesses should not be separated as a matter of routine. Separation should be considered where:

it is both necessary and practicable, and

where there are no implications for the safety of the public or officers.

A decision to separate officers should be made having considered whether separation is necessary to prevent the officers from conferring. This may be required, for example, where there are reasonable grounds to suspect that a criminal offence has been committed or where there is evidence that an officer has breached the standards of professional behaviour set out in the Code of Ethics. It would also be appropriate to consider separation where there is evidence that Authorised Professional Practice (APP) is not being followed or complied with.

…

A decision whether to separate principal officers or key police witnesses should be made by the PIM or responsible ACPO officer, as appropriate.

Where appropriate safeguards are in place to ensure that APP is adhered to, it may be decided that it is unnecessary to separate principal officers. Under these circumstances a written record should be made of this decision, the rationale supporting it, and all measures that have been put in place to prevent conferring.

Where a command decision is made to separate principal officers, or a request is made to do so, the rationale for this decision or request should also be recorded."

22. The guidance goes on to refer to the effects that a traumatic or life-threatening encounter may have on an officer's ability immediately after the incident to recall what may be important detail, and to state that where, over time, officers recall further information this should be recorded in a further account. It also states that there may be circumstances where it is necessary for officers to provide more detailed information at an early stage.

The IPCC's draft statutory guidance

23. The IPCC has power under section 22 of the 2002 Act to issue statutory guidance to local policing bodies and chief officers, subject to prior consultation and to the approval of the Secretary of State. It has not yet issued actual guidance but it has issued a consultation draft. Its "Draft statutory guidance to the police service on achieving best evidence in death or serious injury matters", issued on 5 March 2014, includes the following relevant passages:

"Identification and handling of key police witnesses

16. The police must immediately identify the key policing witnesses to the death or serious injury.

17. For the purpose of this guidance a policing witness is a:

police officer;

special constable under the direction and control of a chief officer;

… [etc.]

Key policing witnesses

18. A key policing witness is anyone from the above categories who has witnessed, or claims to have witnessed, visually or otherwise, all or part of a death or serious injury, or events closely connected with it. Such individuals must be treated as key policing witnesses at the outset.

Separation and prohibition on conferring

19. Any conferring between witnesses has the potential to undermine the integrity of their evidence, and to damage public confidence in the investigation. As a result, non-police witnesses are routinely warned not to discuss the incident in question either before or after they have given their accounts. The same should apply to policing witnesses.

20. Once the key policing witnesses have been identified:

20.1 They should be instructed not to speak (or otherwise communicate) about the incident with each other, or any other potential witnesses, both before and after they have given their accounts.

20.2 If it is necessary for key policing witnesses to discuss the incident with each other to avert a real and immediate risk to life, the extent to which such discussion has taken place, the justification for doing so and the content of that conversation, must be recorded as soon as possible.

20.3 From the moment it is operationally safe to do so, they should be kept separate until after their detailed individual factual account ('DIFA') is obtained.

Detailed Individual Factual Accounts (DIFA)

21. All key policing witnesses will be expected to assist in the investigation into the death or serious injury by providing a full and detailed account at the earliest opportunity. Nothing in this guidance affects their right to be provided with support by other people not involved in the incident, including the right to obtain legal advice and the right to refuse to provide any statement. However, a decision not to provide an account when asked to do so will be noted and will be taken into account in the investigation and in any subsequent proceedings.

22. A DIFA will be required from each policing witness before s/he goes off duty ….

23. The objective of the DIFA is to obtain a clear picture of all facts which may be relevant to the death or serious injury.

24. It is accepted that no one can be compelled to give an account and that consideration must be given to the welfare of those who have been directly involved in a serious incident. However, the public rightly expects that those who witness a death or serious injury, or incidents relating to it, whilst acting in a professional capacity, should co-operate fully with an investigation, offering up all relevant information in a prompt and open manner. Failure to do so damages not only the effectiveness of the investigation but also the public's confidence in the police service."

24. The consultation period closed on 27 May 2014. We are told that the IPCC received 29 detailed responses to the consultation. The IPCC was due to meet in early November to review the responses in detail and to consider whether there should be any amendments to the draft statutory guidance. It is then due to meet again in February 2015 to consider a final version of the guidance.

The relevant procedural requirements of article 2

25. The general principles governing the investigation required by article 2 are conveniently summarised in the judgment of the Grand Chamber of the European Court of Human Rights ("ECtHR") in Nachova v Bulgaria (2006) 42 EHRR 43:

"110. The obligation to protect the right to life under Art.2 of the Convention, read in conjunction with the state's general duty under Art.1 of the Convention to 'secure to everyone within [its] jurisdiction the rights and freedoms defined in [the] Convention', requires by implication that there should be some form of effective official investigation when individuals have been killed as a result of the use of force. The essential purpose of such an investigation is to secure the effective implementation of the domestic laws safeguarding the right to life and, in those cases involving state agents or bodies, to ensure their accountability for deaths occurring under their responsibility.

…

112. For an investigation into alleged unlawful killing by state agents to be effective, the persons responsible for and carrying out the investigation must be independent and impartial, in law and in practice.

113. The investigation must also be effective in the sense that it is capable of leading to a determination of whether the force used was or was not justified in the circumstances and to the identification and punishment of those responsible. The authorities must have taken the reasonable steps available to them to secure the evidence concerning the incident, including, inter alia, eye witness testimony and forensic evidence. The investigation's conclusions must be based on thorough, objective and impartial analysis of all relevant elements and must apply a standard comparable to the 'no more than absolutely necessary' standard required by Art.2(2) of the Convention. Any deficiency in the investigation which undermines its capability of establishing the circumstances of the case or the person responsible is liable to fall foul of the required measure of effectiveness."

26. The point is made on behalf of the claimants that, although the investigative obligation under article 2 may be discharged in various ways, including an inquest or a criminal prosecution, an IPCC investigation is central to establishing the facts in cases such as those under consideration here. It is stressed that in such cases knowledge of the true circumstances may lie mainly or solely with the police.
27. The claimants place particular weight on the decision of the ECtHR in Ramsahai v Netherlands (2008) 46 EHRR 43. In that case the deceased, Moravia Ramsahai, had made off with a scooter in a robbery involving the use of a firearm. He was stopped shortly afterwards and resisted arrest, during the course of which he was shot and killed by the police. Following a police inquiry, the public prosecutor decided that the shooting had been an act of self-defence and for that reason did not institute criminal proceedings against the police officer who fired the shot. The issues before the ECtHR included the compatibility of the investigation with article 2.
28. At paragraphs 324 et seq., the Court clarified the scope and content of the requirement of effectiveness. It stated that in order to be effective in the relevant sense an investigation into a death that engages the responsibility of the state must firstly be adequate; that is, it must be capable of leading to the identification and punishment of those responsible. This is an obligation of means, not of result. The authorities must have taken the reasonable steps available to them to secure the evidence concerning the incident. Secondly, the persons responsible for the investigation and carrying it out must be independent of those implicated in the events, which means not only a lack of hierarchical or institutional connection but also a practical independence: "What is at stake here is nothing less than public confidence in the state's monopoly on the use of force" (paragraph 325).
29. In relation to the adequacy of the particular investigation, the applicants pointed out that several forensic examinations which one would normally expect in a case such as this had not been carried out, and that the two officers directly involved in the incident (Officers Brons and Bultstra) had not been questioned until several days after the fatal shooting, during which they had had the opportunity to discuss the incident with others and with each other. The Court upheld most of those criticisms:

"327. It is true that no attempt was made to establish the trajectory of the bullet. It may be questioned whether this could have been determined on the basis of the information available ….

328. However, the Court considers that the other failings pointed out by the applicants impaired the adequacy of the investigation ….

329. The failure to test the hands of the two officers for gunshot residue and to stage a reconstruction of the incident, as well as the apparent absence of any examination of their weapons … or ammunition and the lack of an adequate pictorial record of the trauma caused to Moravia Ramsahai's body by the bullet … have not been explained.

330. What is more, Officers Brons and Bultstra were not kept separated after the incident and were not questioned until nearly three days later …. Although, as already noted, there is no evidence that they colluded with each other or with their colleagues on the Amsterdam/Amstelland police force, 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.

331. These lacunae in the investigation are all the more regrettable in that there were no witnesses who saw the fatal shot fired from close by, except for Officers Brons and Bultstra themselves. The Court has already drawn attention to the inconsistencies between their statements to the effect that the fatal shot was fired by Officer Brons and those of Officers Braam and Van Daal, who both stated that they had heard Officer Bultstra report that he had fired and call for an ambulance ….

332. There has accordingly been a violation of art.2 of the Convention in that the investigation into the circumstances surrounding the death of Moravia Ramsahai was inadequate."

30. The Court went on to consider the independence of the police investigation. It pointed out that 15½ hours passed from the time of the death until the State Criminal Investigation Department became involved. The Court considered the delay, for which no explanation had been given, to be unacceptable. Both before and after the Department became involved, essential parts of the investigation were carried out by the force to which Officers Brons and Bultstra belonged. The Court stated that "whilst it is true that to oblige the local police to remain passive until independent investigators arrive may result in the loss or destruction of important evidence, the Government have not pointed to any special circumstances that necessitated immediate action by the local police force in the present case going beyond the securing of the area in question" (paragraph 338). The Court found a violation of article 2 in that the police investigation had not been sufficiently independent.
31. In R (Saunders) v Independent Police Complaints Commission [2008] EWHC 2371 (Admin), [2009] PTSR 1192, Underhill J had to consider the decision in Ramsahai in the context of a challenge to the lawfulness of an investigation by the IPCC in circumstances where no steps had been taken to prevent the police officers involved in a fatal shooting from speaking to one another before they gave their first accounts, or to prevent them from collaborating in producing the notebook entries or statements which constituted those accounts, and where they had in fact so collaborated. That conduct was expressly permitted by the ACPO guidance in force at the time of the incidents, though the guidance was in the course of revision at the time of the judgment. The judge said this:

"38. In my view the judgment in the Ramsahai case … demonstrates that in the case of a fatal shooting by police officers the state may be held to have violated article 2 if, in the course of the investigation required by the article, adequate steps were not taken to prevent the police officers directly concerned from conferring before producing their first accounts of the incident; and that that is so even if it cannot be shown that they did in fact confer. I accept that the opportunity which was given to Officers Brons and Bultstra to 'collude' was only one of three reasons which were held, cumulatively, to give rise to a breach. But I can see no principled reason why a vitiating factor of this kind needs to be supported by other factors. I also accept that the court explicitly referred to the risk only of 'collusion' rather than of innocent contamination. But the risks of collusion and of innocent contamination are both alike products of the opportunity to confer, and in cases where contamination does occur it will often be difficult to know whether that was deliberate or innocent. Both are capable of prejudicing an effective investigation, and the measures aimed at preventing the one would also protect against the other. While the court was, for obvious reasons, most exercised by the risk of collusion I very much doubt that it regarded the risks of innocent contamination as being of no concern.

39. It follows that if the circumstances of either of these cases were in due course to be considered by the court it might very well find that a breach of article 2 had occurred ….

40. I am not, however, prepared to say that the mere fact that there was collaboration in the production of witness statements in these two cases means that a breach of article 2 has been definitively established. Decisions of the European Court of Human Rights on the facts of a particular case ought not to be treated as a binding precedent, even in a case where the material facts appear to be similar. The only authoritative parts of a judgment are the statements of principle which it expounds. In my view the relevant statements of principle emerging from the Ramsahai case are that there must in every case of a killing by state agents be an effective investigation, and that in order to be effective such an investigation must be both independent and 'adequate'. The case also establishes that an investigation may be inadequate, and therefore ineffective, if 'appropriate steps' are not taken to 'reduce' the risk of collusion (see the Ramsahai case, para 330): I do not myself regard that as a statement of principle so much as an application of the underlying principles which I have identified. But, even if I am wrong about that, the principle in question is far from absolute in its formulation and involves the need to make judgments as to what steps are 'appropriate' and to what extent it is possible to 'reduce' the risks: those are precisely the kinds of judgment which ACPO is having to make in formulating its revised guidance …."

32. Later in his judgment, at paragraph 65, the judge stressed that he was not saying that collaboration in note-taking complied with article 2. On the contrary, he believed that a practice of permitting principal officers to collaborate generally in giving their first accounts was highly vulnerable to challenge under article 2.
33. Matters have moved on substantially since the judgment in Saunders. I have set out already the provisions of the 2014 guidance that prohibit conferring between officers as a matter of general practice and that limit the opportunities for collaboration about their accounts. The question for us to decide is whether the existing guidance nevertheless gives rise to an unacceptable risk of breach of article 2 or is otherwise unlawful on the grounds advanced by the claimants.

The case for the claimants

34. Ms Kaufmann QC, on behalf of Mr Delezuch, starts from the principle that a policy which creates an unacceptable risk of illegality in its application is itself unlawful. In support of that principle she cites R (Refugee Legal Centre) v Secretary of State for the Home Department [2004] EWCA Civ 1481, [2005] 1 WLR 2219, as explained and followed in R (Tabbakh) v Staffordshire and West Midlands Probation Trust [2014] EWCA Civ 827. The actual issue in each of those cases was a rather different one, namely whether the system established by the relevant policy was inherently unfair. I need not dwell on the authorities, however, because the principle on which Ms Kaufmann relies is not itself in dispute. (Ms Kaufmann disavows any reliance on the test of "significant risk" in R (Munjaz) v Mersey Care NHS Trust [2005] UKHL 58, [2006] 2 AC 148, a case concerning risk of ill-treatment contrary to article 3. Whilst Mr Southey QC, for Mrs Duggan, does rely on Munjaz, I think it sufficient for present purposes to refer to paragraph 43 of my judgment in Tabbakh where I expressed the view that the observations in Munjaz were concerned specifically with the risk of article 3 ill-treatment and were not intended to be an expression of any wider principle.)
35. The first main ground on which Ms Kaufmann submits that the 2014 guidance creates an unacceptable risk that an investigation will not comply with the article 2 requirement of effectiveness concerns the adequacy of the investigation. The guidance applies in practice to all investigations into a DSI matter and will cover cases where the witness evidence of the officers involved in the incident is of critical importance. It is essential to prevent the contamination of that evidence if the integrity of the process is not to be undermined. An instruction not to confer may be an adequate means of reducing or eliminating the risk of innocent contamination but will not affect the risk of deliberate collusion. In order to eliminate the risk of such collusion it is essential that, save where it is not practicable for operational or other reasons (including considerations of safety), the officers involved are separated immediately after the incident and are kept separate until they have given their detailed accounts to the IPCC. The guidance does not go far enough in that direction. It provides that principal officers should not be separated as a matter of routine; and although it goes on to state that separation should be considered where "it is both necessary and practicable", and in particular after considering whether it is necessary to prevent the officers from conferring, it does not do enough to ensure that separation occurs where article 2 requires it. The failure to keep the officers separate in Ramsahai was a reason why the investigation was inadequate and ineffective even though there was no evidence of collusion. A requirement that all officers who are involved in or are witnesses to an incident involving the fatal use of force by the police shall be separated immediately and shall remain separate until they have all given statements has been accepted in Los Angeles, pursuant to a consent decree in federal proceedings; and Ms Kaufmann submits that if Los Angeles can do it, so can police forces in the United Kingdom.
36. A further concern voiced by Ms Kaufmann, though not articulated in her pleaded case, relates to the distinction between officers' initial accounts and detailed accounts. She complains that the guidance leads to a very limited amount of detail in initial accounts (the initial witness statements in the Duggan case are cited as a practical example); less than would be expected to be included in the ordinary way in officers' notebook entries. She submits that officers should be required to give as much detail as possible at an early stage. What makes this all the more important is that, to the extent that the guidance contains measures aimed at reducing the risk of collusion, in particular through the role of the officer appointed as supervisor, they bite in practice only on the making of first accounts and they leave officers free to put their heads together after going off duty and before giving their detailed accounts. A further concern is that the provisions relating to supervision apply only to principal officers, not to witness officers (where the risk of collusion is as great). Other detailed criticisms of the guidance include the absence of a clear definition of "principal officers" and the lack of clarity as to when officers are permitted to confer with one another.
37. The second main ground on which Ms Kaufmann submits that the 2014 guidance creates an unacceptable risk that an investigation will not comply with the article 2 requirement of effectiveness concerns the independence of the investigation. She submits that the guidance is deficient in failing to provide that first accounts (which, as already mentioned, should in her submission be detailed accounts) are not taken until the IPCC investigators have arrived on the scene and are present to supervise them.
38. Thus, the position advocated by Ms Kaufmann can be seen to be broadly in line with the substance of the IPCC's draft statutory guidance.
39. Mr Southey, on behalf of Mrs Duggan, adopts Ms Kaufmann's submissions (subject to his reliance on the test of "significant risk" in Munjaz to which I have already referred). He also expresses the same substantive points in terms of the guidance containing a "misdirection" that effective investigations do not generally require officers to be separated. He cites the decision of the ECtHR in Özcan and Others v Turkey (Application no. 18893/05, judgment of 20 April 2010), paragraph 67, as an example of a delay in questioning giving rise to a risk of collusion. By reference to a report of the Assistant Coroner appointed to conduct the inquest into the death of Mark Duggan ("Report to prevent future deaths", dated 29 May 2014), he refers to features of the Duggan case as illustrating the deficiencies of the 2014 guidance. For example, the key officers travelled back to the police station together; officers were warned against conferring but no step was taken to prevent them from doing so and it was left for them to decide what to confer about; the first accounts were bland and uninformative; officers then went home and it was three days before they re-gathered to compile statements. Those events took place in 2011 when earlier ACPO guidance applied but it is submitted that the deficiencies have not been remedied by the 2014 guidance.
40. Mr Southey also advances an alternative case based on the position of the IPCC as the independent investigative authority. He refers to an historical tension between the IPCC and ACPO in relation to the guidance. It is illustrated by a passage from the IPCC's response to Mrs Duggan's letter before claim. Having referred to the power of IPCC to issue case-specific directions not to confer, the letter states:

"The existence and content of the ACPO guidance, to which the IPCC's representations had substantially contributed even though they did not fully reflect the IPCC's views, … is an inevitable part of the background. The IPCC cannot ignore the fact that if it were to give binding directions in every DSI case as a matter of policy to reflect its own views as fully as possible, then (to the extent that the ACPO guidance did not fully accept the IPCC's views) the directions would be inconsistent with the ACPO guidance and would effectively countermand it.

The IPCC's view remains that countermanding the ACPO guidance, after the process by which that guidance had been formulated, would be more likely to prejudice than to assist an effective investigation …."

41. Mr Southey submits that the tension will remain even if the IPCC adopts its draft statutory guidance. He refers to a passage in the third witness statement of Deputy Chief Constable Simon Chesterman on behalf of ACPO:

"24. In light of the revised APP [the College of Policing's Authorised Professional Practice, including the 2014 guidance], I do not consider that it is necessary or desirable for the IPCC to introduce its proposed Statutory Guidance. The APP will remain the 'go to' document for firearms officers and commanders since it deals with matters far broader than post incident procedures. Having an alternative IPCC document dealing with post incident procedures would, in my view, be both unhelpful and confusing. Moreover, where there is a difference between the proposed IPCC Statutory Guidance and the re-drafted APP, it is my view that the guidance in the APP should be preferred …."

42. Mr Southey refers to what was said in Ramsahai and in other cases about the need for an investigation to be independent and the importance of this for public confidence. He points to the role given to the IPCC by section 10 of the 2002 Act and to judicial statements to the effect that it is for the independent investigator to decide how to proceed (see e.g. R (L (A Patient)) v Secretary of State for Justice [2008] UKHL 68, [2009] 1 AC 588, paragraph 76). He submits that if the independent body entrusted with the task is saying that because of the ACPO guidance it cannot undertake the investigation in the way it would wish, its independence is being undermined and there is a breach of article 2. For this purpose he equates the 2014 guidance with ACPO guidance even though the 2014 guidance is issued by the College of Policing.

The case for the defendants

43. The submissions made by Mr Beggs QC, on behalf of ACPO, extend beyond ACPO's narrow position as defendant to the Duggan claim and encompass a defence of the 2014 guidance even though it is not strictly an ACPO document. Mr Beggs accepts that conferring between officers has the potential to undermine public confidence in a post-incident investigation, which is why the guidance includes a direction against conferring. He submits, however, that the overwhelming majority of officers will obey the direction and that, although there is always a risk of deliberate collusion by a small number of officers, a general policy should not be formulated by reference to exceptional cases. The 2014 guidance is appropriate and flexible and does not give rise to an unacceptable risk that an investigation will fail to meet the article 2 requirement of effectiveness. There are places where the drafting could be improved (e.g. as regard the definition of "principal officers", to require greater detail in initial accounts at stage three, and to clarify the circumstances in which separation may be considered necessary) but those deficiencies do not render the guidance unlawful.
44. Mr Beggs puts forward several objections to the suggested requirement of immediate separation of all principal officers. First, such an approach would be intrusive and stigmatising. Mr Chesterman, drawing on his extensive experience, expresses the view in paragraph 25 of his third witness statement that "requiring their separation from colleagues after a traumatic shooting regardless of the circumstances and effectively treating them as suspects (without, I observe, the protections afforded to criminal suspects) would be deeply unpalatable to many firearms officers and police officers considering applying for firearms duties". He expresses the view that unless there are reasonable grounds to suspect otherwise, officers are witnesses and should be treated as such. Mr Beggs reinforces those observations by pointing to the likely effect on morale if officers have to be chaperoned from the scene and kept in separate rooms at the police station, presumably without mobile phones and/or subject to constant invigilation; all this on top of the traumatising effect that an incident involving firearms can have on the officers concerned.
45. The second objection to separation is one of practicability. There are many instances where officers cannot reasonably be separated and may need to confer, for reasons that include safety, scene preservation, exhibit management and hot pursuit. There may be a large number of officers involved: for example, there were no fewer than 70 in the Saunders case, of whom 13 were initially regarded as principal officers, and numbers easily expand in terrorist situations. It would often be impracticable to chaperone officers individually to the police station and keep them in separate post-incident suites there. The claimants have allowed for exceptions for operational reasons, but the exceptions would become the norm.
46. Thirdly, it is submitted that routine separation is unnecessary and disproportionate having regard to other checks and balances. The 2014 guidance deprecates conferring and this is underpinned by the disciplinary system, the supervisory structures (PIMs etc) and the discretion to separate with cause. Such concerns as remain will largely fall away if the use of body cameras by firearms officers, currently under trial, becomes established: provision can be made in the guidance for body cameras to remain switched on from the start of the incident until the point when officers enter the post-incident suite and are subject to supervision.
47. Fourthly, it is submitted that there is no evidence that inappropriate conferring is a widespread problem. The vast majority of police officers are trained and decent professionals who will heed the warning against conferring. Separation is not required in their case. If, in the exceptional case, officers are intent on dishonest collusion, they are likely in any event to find an early opportunity to do so before any separation could be effected.
48. As to the claimants' contention that officers should be required to give an early detailed account before they go off duty, Mr Beggs submits by reference to Mr Chesterman's second witness statement that the four-stage sequence contained in the 2014 guidance is based on years of experience and extensive academic research and consultation as to how to obtain best evidence, and that there is no proper basis for interfering with the provision in the guidance that a detailed account should not be obtained for at least 48 hours. A balance must be struck between the objectives of achieving best evidence, safeguarding the welfare of officers, and transparency and public confidence. The claimant's proposal promotes the last objective over the others. Officers should not be required to provide detailed accounts when they are tired and traumatised. It is standard practice for officers to write a basic account in their notebook or incident report book before going off duty. That is the equivalent of the initial account at stage three. It can be used for the preparation of a detailed account at a later stage. As to the suggestion that the initial accounts should be given in the presence of an IPCC investigator, it is to be borne in mind that this is not an interview: officers write their accounts down for themselves (a process which, under the 2014 guidance, will be supervised by the appointed officer). They should be free to write them as soon as they wish and should not be required to wait for the arrival of an IPCC investigator, which can take several hours.
49. As regards the separate case advanced on behalf of Mrs Duggan, Mr Beggs submits that the ACPO has done nothing to impede the discharge of the IPCC's functions. It took the IPCC ten years to produce draft statutory guidance. ACPO is not to be criticised for having produced guidance of its own in the meantime. The IPCC will consider the representations received in response to its consultation exercise and the court should not seek to predict the outcome of that exercise. The fact that ACPO guidance is regarded as the "go to" guidance does not mean that chief officers will be encouraged to ignore any statutory guidance that the IPCC may issue.
50. Finally, Mr Beggs has a number of technical points to the effect that ACPO is not the proper defendant to the Duggan claim (where he says that the correct defendant is the Commissioner of Police for the Metropolis) and that the court should decline to hear any challenge to the ACPO guidance, which is now academic; but he accepts that the court should consider the merits of the challenge to the 2014 guidance and that ACPO is in a position to defend that guidance.
51. For the Chief Constable in the Delezuch claim, Mr Boyle adopts and supports the submissions made on behalf of ACPO, with some additional detail to show the practical problems that would arise for policing in the Chief Constable's area if a general requirement of separation were introduced.
52. The IPCC, appearing through Mr Glasson QC, has made no submissions on the substantive issues in the case but was represented at the hearing in order to update the court with regard to the draft statutory guidance.

Discussion

53. I will consider first the argument that the 2014 guidance is unlawful because of the risk of breach of article 2.
54. It should be stressed that the principle on which the claimants rely is that a policy is unlawful if it creates an unacceptable risk that individual decisions or action taken in application of the policy will be unlawful – in this case, that investigations carried out in accordance with the 2014 guidance will be in breach of the procedural requirements of article 2. The principle is not that any risk of an unlawful outcome must be avoided. Nor, as I have said at para 34 above, is the test one of significant risk. The Refugee Legal Centre and Tabbakh cases cited at para 34 above were both concerned with whether there was an unacceptable risk of procedural unfairness, effectively equated with a risk of unfairness inherent in the system established by the policies under challenge. In that context the court in Tabbakh expressed the view that the relevant threshold was a high one, but it rejected a general test formulated in terms of "a very high risk or an inevitability" of unlawful action and it did not lay down any alternative general test (see paras 47-49 of my judgment in that case). As the authorities stand, whether a risk is unacceptable is a matter of judgment for the court in the particular circumstances of the case.
55. I accept that, in a case of death following the use of force by police officers, a failure to separate the police officers who used or witnessed the use of force may impair the adequacy of the investigation because of the risk of collusion (a term which I will use for the sake of simplicity to cover both dishonest collusion and innocent contamination of police evidence). Whether it does so to any material extent will, however, depend on the circumstances, including the other safeguards in place; and whether the investigation as a whole is adequate for the purposes of article 2 will depend on an overall assessment of relevant factors, of which the risk of collusion is only one. The ECtHR in Ramsahai (paras 27-29 above) did not hold that a failure to keep officers separate will necessarily render an investigation inadequate. The Court did say that the failure to take appropriate steps to reduce the risk of collusion amounted to a significant shortcoming in the adequacy of the investigation in question, but that was only one of several factors that led to the finding that the investigation was inadequate and in breach of article 2. I agree generally with the observations of Underhill J in Saunders as to the effect of the judgment in Ramsahai (see para 31 above).
56. The approach taken in the IPCC's draft guidance (para 23 above) has much to commend it from the point of view of article 2. The combined effect of the provisions relating to identification of the key policing witnesses (para 18 of the draft guidance) and those relating to the separation of such witnesses and the prohibition on conferring (para 20) would minimise the risk of collusion. The risk would not be eliminated altogether: para 20 provides for the separation of the key policing witnesses "[f]rom the moment it is operationally safe to do so", which would still leave open the possibility of their discussing their evidence during the period when they were together for reasons of operational safety. That residual risk, however, could not realistically be avoided.
57. The fact that the IPCC's draft guidance has much to commend it from the point of view of article 2 does not mean that the 2014 guidance, in particular by taking a different approach towards separation of officers, creates an unacceptable risk of breach of article 2. Nor does it mean that the IPCC should adopt its draft guidance. The decision on that is entirely a matter for the IPCC, in the light of the responses to its consultation exercise and the considerations, both practical and of principle, raised by those responses.
58. The 2014 guidance is on any view an imperfect document. It has been accepted on behalf of ACPO that there are places where its drafting could be improved (see para 43 above). Despite those drafting deficiencies, however, if properly applied it goes a long way towards reducing the risk of collusion:

i) It provides that as a matter of general practice officers should not confer with others before making their accounts (whether initial or subsequent accounts). It allows for exceptions, specifically where there is a need for officers to speak to one another to resolve an ongoing operational or safety matter, but it requires the fact of any discussion to be documented with appropriate particulars. I accept the submission on behalf of ACPO that the overwhelming majority of officers can be expected to obey the direction against conferring, both as a matter of ordinary professional integrity and because of the disciplinary sanctions available for breach. The opportunities for conferring are, moreover, substantially limited in practice by the further safeguards referred to below.

ii) The regime established for the supervision of principal officers following their arrival at the post-incident suite at the police station is aimed at ensuring not only that officers are reminded of the direction against conferring but also that inappropriate conferring does not take place in practice. Supervision is the responsibility of a senior officer appointed for the purpose and the process must be documented. The regime is plainly not watertight. It is also deficient in applying on the face of it only to "principal officers" (itself an expression not clearly defined) and not to other "key police witnesses" (referred to as a distinct group in the section of the guidance on separation). It does, however, provide a reasonable safeguard against things happening in the post-incident suite that might impair the adequacy of the investigation.

59. The question of separation of officers needs to be considered against that background. The guidance provides that principal officers and key police witnesses should not be separated as a matter of routine but that separation should be "considered" where it is both necessary and practicable and where there are no implications for the safety of the public or officers. It goes on to state that a decision to separate officers should be made "having considered whether separation is necessary to prevent the officers from conferring" and that this may be required "for example" where there are reasonable grounds to suspect that a criminal offence has been committed or where there is evidence than an officer has breached the standards of professional behaviour. Those provisions lack clarity and precision: this is one of the areas where it has been accepted on behalf of ACPO that the drafting could be improved. On the claimants' case, separation is always "necessary to prevent the officers from conferring", but that is plainly not what is contemplated by the provisions read as a whole. I shall proceed on the basis that separation of officers is an exceptional measure and that the normal position is that separation will not be required.
60. In taking forward the question of separation, it is helpful to break things down into stages:

i) Even if separation were required, it could not reasonably take effect until, at the earliest, it was operationally safe to separate the officers concerned. What this means in practice is that the risk of collusion at the scene cannot be avoided. The direction against conferring, about which all officers should be aware, goes as far as is practicable to prevent collusion at this stage.

ii) Once it was operationally safe to do so, it would be possible in principle to separate officers at the scene and during their conveyance back to the police station. It seems to me, however, that any general rule requiring such separation would still have to be subject to an exception on grounds of practicability (as distinct from operational safety), given the additional police resources that would be needed to render separation effective at this stage. The submission on behalf of ACPO that exceptions for operational reasons would in practice become the norm may be realistic.

iii) Any rule requiring separation on arrival at the police station would also have to be subject to a practicability exception, both for reasons of physical space (a separate room would be needed for each officer concerned) and because of the need for invigilation to render separation effective. It seems to me, however, that the regime of supervision in the post-incident suite, considered above, provides a reasonable safeguard against collusion without any general need for separation at this stage.

iv) The process of supervision in the post-incident suite applies, as I understand it, both to initial and to detailed accounts. The guidance allows, however, for a potentially lengthy interval (48 hours or more) between the two, during which an officer will go off duty and will not be subject to such supervision. An unsupervised interval of that nature carries with it an inherent risk of collusion; but that risk could not sensibly be removed by a requirement of separation, since any such requirement would in practice be unenforceable. Whether it is appropriate to retain the approach of initial account followed by detailed account, rather than aim for a single "detailed individual factual account" as envisaged in the IPCC's draft guidance, is a matter of considerable controversy and in my view is not one that can sensibly be resolved in these proceedings. But given the existing approach as set out in the 2014 guidance, it does seem to me to be important to ensure that the initial account contains greater detail than in some of the examples we were shown from the Duggan case. The less that is said in the initial account, the greater is the scope for contamination of the detailed account by reason of intervening collusion. This is another area where it was acknowledged on behalf of ACPO that the drafting of the guidance could be improved.

61. Where does that all lead? The risk of collusion cannot be avoided altogether. The 2014 guidance goes a long way towards reducing the risk but there are two areas of particular concern. One is the absence of a requirement of separation at the scene (from the moment when separation is operationally safe) and while the officers are conveyed back to the police station. The other is the unsupervised interval between initial and detailed accounts. In relation to the former, however, there are strong reasons to doubt the practicability of separation in the generality of cases. In relation to the latter, there could be no effective requirement of separation and the problem can and should be addressed instead by ensuring that initial accounts contain a reasonable amount of detail (without collapsing the distinction between initial and detailed accounts).
62. Overall, the 2014 guidance leaves open a greater risk of collusion than would be left open by the IPCC draft guidance, thereby creating a greater risk that an investigation carried out in accordance with the guidance would fail to meet the procedural requirements of article 2. But in the light of the safeguards that the guidance does provide, and bearing in mind that the adequacy of an investigation for the purposes of article 2 would have to be assessed by reference to all the features of that investigation, I take the view that the risk of breach of article 2 to which the guidance itself gives rise is a relatively low one. I do not consider it to be an unacceptable risk, such as would justify a finding that the guidance itself was unlawful. Accordingly, I would dismiss the main ground of claim advanced on behalf of the two claimants. In so far as Mr Southey sought to express the same points in terms of the guidance containing a misdirection that effective investigations do not generally require officers to be separated, I would reject the submission for essentially the same reasons.
63. I can deal with the other grounds much more briefly. First, I do not accept Ms Kaufmann's submission that the guidance breaches the article 2 requirement of independence in failing to provide that first accounts are not taken until the IPCC investigators are present to supervise them. It can take several hours for an IPCC investigator to arrive and I see no reason why an officer should be prevented from making his initial account before then. The guidance ensures that it will be made in controlled conditions under the supervision of a senior officer. In my judgment the procedure laid down could not realistically be said to prejudice the independence of the investigation.
64. As to Mr Southey's case that ACPO has acted in breach of article 2 by issuing guidance that prevents the IPCC from undertaking an investigation in the way it would wish or that otherwise undermines the role of the IPCC as the independent investigative authority, I confess to having great difficulty in understanding the argument at all, even leaving aside the point that the 2014 guidance is not strictly ACPO guidance. In any event I accept Mr Beggs's submission that ACPO has done nothing to impede the discharge of the IPCC's functions. The IPCC has asserted its independence in issuing draft guidance that differs materially from the 2014 guidance. If it proceeds to issue statutory guidance, whether in that or in revised form, chief officers will have to take due account of it and consideration will no doubt have to be given to whether the 2014 guidance should be brought into line with it. All that, however, is for the future. As to the past and present, I do not see how ACPO's stance in issuing guidance or supporting the 2014 guidance could conceivably be said to prejudice the independence of investigations carried out by the IPCC.

Conclusion

65. For those reasons I would dismiss both applications for judicial review.

Lord Justice Tomlinson :

66. I agree.

Lord Justice Moore-Bick :

67. I also agree.