Maughan, R (On the Application Of) v Senior Coroner for Oxfordshire

The question raised by this claim is whether a coroner or a coroner’s jury, after hearing the evidence at an inquest into a death, may lawfully record a conclusion to the effect that the deceased committed suicide reached on the balance of probabilities; or whether such a conclusion is only permissible if it has been proved to the criminal standard of proof (i.e. so that the coroner or jury is sure that the deceased did an act which was intended to and did cause his or her own death).

At approximately 5.20am on 11 July 2016 the claimant’s brother, James Maughan, who was in custody at HMP Bullingdon, was found hanging in his prison cell. Ambulance staff confirmed his death around an hour later.

An inquest into James Maughan’s death was held by the defendant, HM Senior Coroner for Oxfordshire, with a jury. The inquest was heard over four days in October 2017. After the close of the evidence, the coroner accepted that there was insufficient evidence upon which the jury could be sure that the deceased intended to kill himself. He took the view that in these circumstances the jury could not be permitted to consider a ‘short-form’ conclusion of suicide. However, he invited the jury to record a narrative conclusion which answered five questions (provided to the jury in writing). Questions 3 to 5 were:

“3.

Did James Maughan deliberately place a ligature around his neck and suspend himself from the bedframe?

4.

Are you able to determine if it is more likely than not that he intended the outcome to be fatal, or for example, if it is likely that he intended to be found and rescued? If you are unable to determine his intention, please say so.

5.

Were there any errors or omissions on the 10-11 July in the provision of care on the part of HMP Bullingdon/prison staff which caused or contributed to James Maughan’s death?”

The coroner also directed the jury to add to question 4 and to consider whether the deceased was unable to form a specific intent to take his own life through mental illness.

The questions for the jury were accompanied by written instructions, one of which was:

“The standard of proof you should apply when considering these questions is the balance of probabilities. In reaching your conclusions, you therefore have to be satisfied it is probable (more likely than not) that something did or did not happen.”

The jury’s narrative statement, written on the record of inquest, included the following findings:

“We believe James deliberately tied a ligature made of sheets around his neck and suspended himself from the bedframe.

James Maughan had a history of mental health challenges and on the night of 10 July 2016, James was visibly agitated. We find that on the balance of probabilities, it is more likely than not that James intended to fatally hang himself that night.

… neither formally opening an ACCT, nor increased vigilance generally would have likely prevented James’ death, given what we believe was James’ intent to end his life. …”

In this claim for judicial review, the claimant contends that the jury’s conclusion was unlawful, as it amounted to a verdict (or, as it is now called, a “conclusion”) of suicide reached on the balance of probabilities. It is said that the coroner erred in law in instructing the jury to apply the civil standard of proof when considering whether James Maughan intended to kill himself and that the law is clear that a conclusion of suicide, whether recorded in short form or as part of a narrative statement, may only be returned on the criminal standard of proof.

Ms Hewitt, who represents the defendant coroner, points out that his directions to the jury were in accordance with express guidance given in The Coroner Bench Book (June 2015) and, arguably, in the Chief Coroner’s Guidance No. 17: “Conclusions: Short-Form and Narrative”. While suggesting reasons why this guidance is arguably correct, Ms Hewitt made it clear that the coroner takes a neutral stance on whether his directions and the jury’s conclusion were lawful.

None of the interested parties has taken any active part in the proceedings.

Before considering the claimant’s case in more detail, we will describe the purpose of a coroner’s investigation and refer to the guidance which the coroner was following.

As set out in section 5(1) of the Coroners and Justice Act 2009, the purpose of a coroner’s investigation into a person’s death is to ascertain: (a) who the deceased was; (b) how, when and where the deceased came by his or her death; and (c) certain formal particulars required by the Births and Deaths Registration Act 1953 to be registered concerning the death. Historically, the task of ascertaining “how” the deceased came by his or her death has been understood narrowly as meaning “by what means”. However, in R (Middleton) v West Somerset Coroner [2004] UKHL 10; [2004] 2 AC 182, the House of Lords held that, where necessary to comply with the state’s obligations under the European Convention on Human Rights, the purpose of the investigation extends to ascertaining “in what circumstances” the deceased came by his or her death. This is now expressly provided for by section 5(2) of the 2009 Act.

The specific obligation with which the House of Lords was concerned in the Middleton case arises under article 2 of the Convention, which protects the right to life. The European Court of Human Rights has interpreted article 2 as imposing on contracting states not only substantive obligations to protect life, but also a procedural obligation to hold an effective investigation into any death where it appears that one or other of the state’s substantive obligations has been, or may have been, violated and that agents of the state are, or may be, in some way implicated (see para 3 of the Middleton case and the cases there cited). One context in which such a procedural obligation arises is where a person dies while in state custody – as happened in the Middleton case itself. Such cases, where the death was a violent or unnatural one or the cause of death is unknown, are also one of the categories of case in which an inquest must be held with a jury: see section 7(2)(a) of the 2009 Act.

Section 10 of the 2009 Act requires the coroner (if there is no jury) or the jury (if there is one), after hearing the evidence at an inquest into a death, to make a determination as to the questions mentioned in section 5. Pursuant to rule 34 of the Coroners (Inquests) Rules 2013, such a determination is to be made using Form 2 in the Schedule to those Rules. As prescribed by Form 2, the record of the inquest must contain the statutory determination as to how, when and where (and, if applicable, in what circumstances) the deceased came by his or her death and also the conclusion of the coroner or jury as to the death. The notes to the form state that one of nine listed “short-form” conclusions may be adopted but also that, as an alternative or in addition to one of the listed short-form conclusions, the coroner or jury may make a brief “narrative” conclusion. One of the listed short-form conclusions is “suicide”.

Note (iii) to Form 2 states:

“The standard of proof required for the short-form conclusions of ‘unlawful killing’ and ‘suicide’ is the criminal standard of proof. For all other short-form conclusions and a narrative statement the standard of proof is the civil standard of proof.”

The first Chief Coroner, Sir Peter Thornton QC, issued detailed guidance to coroners on various matters. This guidance has no legal force but is intended to assist coroners with the law and their legal duties.

Guidance No. 17, which was issued on 30 January 2015 and revised on 14 January 2016, deals with the use of short-form and narrative conclusions. The Guidance suggests (para 26) that:

“Wherever possible coroners should conclude with a short-form conclusion. This has the advantage of being simple, accessible for bereaved families and public alike, and also clear for statistical purposes.”

The Guidance notes (at para 47), however, referring to the Middleton case, that where article 2 of the Convention is engaged:

“Frequently a narrative conclusion will be required in order to satisfy the procedural obligation under article 2, including, for example, a conclusion on the events leading up to the death or on relevant procedures connected with the death.”

In cases where a narrative conclusion is used, the Guidance emphasises the need for brevity and that narrative conclusions are not to be confused with findings of fact (paras 35-36).

The Guidance states (at para 56), citing note (iii) of Form 2, that the standard of proof required for the short-form conclusions of “unlawful killing” and “suicide” is the criminal standard of proof and that, for all other short-form conclusions and a narrative conclusion, the standard of proof is the civil standard of proof. According to a footnote:

“There is an ongoing discussion as to whether suicide should be proved to the criminal or civil standard. The Ministry of Justice are considering the alternatives.”

Counsel for the coroner in this case contacted the Chief’s Coroner’s office to find out whether there is any such ongoing discussion or consideration and was told that, so far as the Chief’s Coroner’s office is aware, there is no active review of this issue currently being undertaken by the Ministry of Justice.

In discussing the conclusion of suicide, the Guidance indicates (para 62) that where suicide is not found the coroner should explain why, for example:

“Looking at the two elements which must be proved to the higher standard of proof before a conclusion of suicide can be recorded, I am satisfied that [the deceased] took his own life, but I am not satisfied that he intended to do so. I cannot be sure about it. It is in my judgment more likely than not that he had that intention, but on the evidence looked at as a whole I cannot rule out that this was a terrible accident. For those reasons my conclusion is not suicide or accident but an open conclusion.”

It seems clear that this example is given as a possible explanation that the coroner could give for reaching an open conclusion rather than one of suicide. As such, it would not form part of the conclusion recorded on the record of inquest. However, at para 73 of the Guidance it is suggested that, as in this example, an open conclusion “can have extra words appended to it by way of explanation.”

The Coroner Bench Book likewise has no legal force but was issued by the Chief Coroner in June 2015 to provide coroners with a guide to their use of words in court. It includes some “specimen examples” of words that may be used in summing up for a jury certain types of case. One of these examples is for summing up in a prison death where the deceased has died by hanging. The Bench Book suggests the following form of words as to how a jury could be directed to consider a narrative conclusion in such a case:

“28.

The third conclusion is in the form of a short narrative and is appropriate where either you cannot decide AB’s state of mind or you find that his mental condition caused him to be incapable of forming an intention or where, on balance, you find he intended to take his own life but you cannot be sure about it.

29.

You could say something like: AB deliberately chose to suspend himself by a belt

a.

but the evidence does not fully explain whether or not he intended that the outcome be fatal OR

b.

but he was not capable of forming an intention that the outcome be fatal OR

c.

and, on balance, he intended that the outcome be fatal.

What words you use is entirely a matter for you.

30.

As with misadventure, you must be satisfied on a balance of probabilities that the act of suspension was deliberate and also that either AB could not form an intention as to the consequences or you cannot determine his state of mind one way or the other or, on balance, he intended that the outcome be fatal.”

The coroner’s written questions and instructions for the jury in the present case were consistent with this suggested approach.

The claimant’s central argument, ably advanced on his behalf by Mr Bunting, starts from the premise that it is settled law that a conclusion of suicide at an inquest can only be returned if suicide is proved to the criminal standard of proof – that is to say, so that the coroner or jury is sure that the deceased did an act which caused his own death with the intention that the act would have that fatal consequence. It is inconsistent with that legal rule, Mr Bunting submitted, to record such a finding in a narrative conclusion when it has not been proved to the criminal standard but only to the civil standard of the balance of probabilities. That is what the jury did in this case when, having found that James Maughan deliberately tied a ligature around his neck and suspended himself from the bedframe, the jury recorded in its narrative conclusion that “on the balance of probabilities it is more likely than not that James intended to fatally hang himself that night.” It follows, Mr Bunting submitted, that the jury’s conclusion was unlawful.

Ms Hewitt for the coroner, while acknowledging that the law on this issue may be ripe for reconsideration, felt constrained not positively to dispute that the criminal standard of proof is required for a short-form conclusion of suicide, as this was the view taken by the coroner in the present case. She suggested, however, that the narrative conclusion of the jury was permissible as it simply recorded facts which, by definition, fell short of a conclusion of suicide because the required standard of proof had not been met.

Ms Hewitt further submitted that, at least arguably, it was necessary for the jury to consider whether it was more likely than not that the deceased intended to kill himself in order to fulfil the investigative duty under article 2 of the European Convention on Human Rights. As part of that duty, one of the matters to be investigated was whether there were any errors or omissions on the part of the prison staff which caused or contributed to James Maughan’s death. This was one of the questions which the coroner asked the jury to answer (see para 3 above). The test for causation is now accepted to be whether, on the balance of probabilities, the conduct in question more than minimally, negligibly or trivially contributed to death: see R (Tainton) v HM Senior Coroner for Preston and West Lancashire [2016] EWHC 1396 (Admin); [2016] 4 WLR 157, para 41. To determine whether or not there was any such causative error or omission in this case, therefore, the jury had to consider whether on the evidence it was more probable than not that the deceased was intent on ending his life such that, for example, increased vigilance by prison staff would not have prevented his death. In this case the jury made a finding to that effect (see para 5 above).

Mr Bunting responded that article 2 does not mandate the standard of proof to be applied in ascertaining the relevant facts for the purpose of investigating a death. Accordingly, it does not prevent domestic law from requiring a suicidal intent to be proved to the criminal standard of proof. He further submitted that a coroner can, by framing appropriate questions for the jury, ensure that the jury returns a conclusion of suicide only if the jury is sure that the deceased intended to die while enabling the jury to take account of such an intention established only on the balance of probabilities when considering whether any error or omission caused or contributed to the death.

We see force in the point made by Ms Hewitt that, in order to determine the causative relevance of any acts or omissions on the part of state agents, it may be necessary for a coroner or jury to make a finding on the balance of probabilities as to whether the deceased intended to take his own life. If such a finding is made and is important, the coroner or jury must be entitled to record it in a narrative conclusion. Yet it appears illogical to conclude (on the balance of probabilities) that the deceased intended to end his life in the context of deciding whether his death could have been prevented whilst at the same time concluding that he did not intend to end his life (because the coroner or jury is not sure of that fact) for the purpose of deciding whether he committed suicide.

Nevertheless, if the premise of the claimant’s argument is correct, the conclusion is in our view irrefutable. A narrative conclusion to the effect that on the balance of probabilities the deceased did a deliberate act which caused his own death intending the outcome to be fatal clearly amounts to a conclusion that the deceased committed suicide whether or not the word “suicide” is used. It is sophistry to say that such a conclusion is not one of suicide because the required standard of proof has not been met. The standard of proof even if referred to in the record of inquest, as it was in this case, is not itself part of the substantive conclusion adopted by the coroner or jury. It is simply a statement of the evidential test which must be met in order to reach a particular conclusion. If the standard of proof required to determine that the deceased committed suicide is the criminal standard and the necessary facts have been proved only on the balance of probabilities, this does not mean that a conclusion which records those facts is not one of suicide. It means that the coroner or jury cannot lawfully reach that conclusion.

One of the first lessons in logic, however, is that accepting the validity of an argument always leaves open two possibilities: one is to accept the conclusion of the argument; but the other is to reject the premise. It is therefore necessary to examine whether it is indeed the law that a conclusion of suicide at an inquest may only be reached if the requisite act and intention have been proved to the criminal standard of proof. Before looking at the authorities which are said to establish that proposition, it is salutary to consider the question in terms of legal principle.

As a general rule, the standard to which a putative state of affairs must be proved before it may be found as a fact by a court or tribunal depends not on the nature of the putative fact but on the nature of the proceedings. Thus, in criminal proceedings the prosecution must prove the charge beyond reasonable doubt. But in civil proceedings a lower standard of proof (on the balance of probabilities) applies. It is clearly established that this is the applicable standard in civil proceedings even in relation to an allegation of criminal conduct.

The underlying reason why a particularly high standard of proof is required in criminal proceedings is that a criminal conviction has serious consequences for the accused, which may include loss of liberty. For that reason the standard of proof is weighted in favour of the accused to reflect the policy that it is better to let the crime of a guilty person go unpunished than to condemn an innocent person. In civil proceedings, which are generally concerned with determining the rights of parties as between each other, there is no equivalent policy reason for weighting the fact-finding exercise in favour of or against one or other party. Instead, in order to cater for those cases in which the evidence is inadequate to enable any positive finding to be made, it is sufficient and expedient simply to have a rule which requires the party who advances a case to prove that the facts relied on to support it are more likely than not to be true.

Of course, decisions in civil proceedings can also have serious consequences for parties involved. For example, a finding in a family court of sexual abuse by a father may result in children being taken into care and could lead to a criminal prosecution being brought against the father, quite apart from causing devastating reputational damage. A large award of damages may cause a defendant financial ruin. An injunction may impose an onerous restriction on a person’s freedom of action. An order for possession may result in a family losing their home. Such consequences may by any practical measure be much more serious for the party affected than, say, a criminal conviction for a minor road traffic offence. The common law, however, has rejected an approach of applying a variable standard of proof. Instead, in the interests of simplicity, consistency and uniformity, a single standard of proof is applied in all civil cases, just as a single (though higher) standard of proof is applied in all criminal cases. The only exceptions are proceedings, such as proceedings for committal for contempt of court, which, although classified as civil, are functionally equivalent to criminal proceedings having regard to the possibility that a person may be sent to prison.

For a long time it was said to be the law that, the more serious the allegation made in civil proceedings, the more cogent the evidence required to prove it. This could be understood – and was sometimes expressly said – to mean that the civil standard of proof is flexible and is to be applied with greater or less strictness depending on the gravity of the subject matter. However, in Re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563 the House of Lords rejected this approach. Lord Nicholls explained the notion that more cogent evidence is required to prove a more serious allegation on the basis that the nature of the allegation may affect its inherent probability. He said (at 586):

“When assessing the probabilities the court will have in mind as a factor, to whatever extent is appropriate in the particular case, that the more serious the allegation the less likely it is that the event occurred and, hence, the stronger should be the evidence before the court concludes that the allegation is established on the balance of probability. Fraud is usually less likely than negligence. Deliberate physical injury is usually less likely than accidental physical injury. A step-father is usually less likely to have repeatedly raped and had non-consensual oral sex with his underage stepdaughter than on some occasion to have lost his temper and slapped her. Built into the preponderance of probability standard is a generous degree of flexibility in respect of the seriousness of an allegation.

Although the result is much the same, this does not mean that where a serious allegation is in issue the standard of proof required is higher. It means only that the inherent probability or improbability of an event is itself a matter to be taken into account when weighing the probabilities and deciding whether, on balance, the event occurred.”

The decision in Re H did not completely clarify the law. The passage quoted above appeared to assume – or could be read as assuming – that there is a necessary connection between the seriousness of an allegation and its inherent probability. Lord Nicholls did not confront the question of what approach should be adopted in circumstances where the serious nature or consequences of an allegation do not make it less likely that the event occurred. That question was subsequently addressed by the House of Lords in Re B (Children) (Care Proceedings: Standard of Proof)(CAFCASS intervening) [2008] UKHL 35; [2009] 1 AC 11. In Re B, at para 15, Lord Hoffmann explained what Lord Nicholls had said in Re H in the passage quoted above as follows:

“Lord Nicholls was not laying down any rule of law. There is only one rule of law, namely that the occurrence of the fact in issue must be proved to have been more probable than not. Common sense, not law, requires that in deciding this question, regard should be had, to whatever extent appropriate, to inherent probabilities. If a child alleges sexual abuse by a parent, it is common sense to start with the assumption that most parents do not abuse their children. But this assumption may be swiftly dispelled by other compelling evidence of the relationship between parent and child or parent and other children. It would be absurd to suggest that the tribunal must in all cases assume that serious conduct is unlikely to have occurred. In many cases, the other evidence will show that it was all too likely.”

Baroness Hale, who gave the other reasoned judgment, was likewise concerned (at para 70) to “announce loud and clear” that the applicable standard of proof was the simple balance of probabilities, neither more nor less, and that:

“Neither the seriousness of the allegation nor the seriousness of the consequences should make any difference to the standard of proof to be applied in determining the facts. The inherent probabilities are simply something to be taken into account, where relevant, in deciding where the truth lies.”

The Supreme Court has expressly confirmed that this is the approach which must be applied in a civil case in deciding whether a person committed suicide: see Braganza v BP Shipping Ltd [2015] UKSC 17; [2015] 1 WLR 1661, paras 33-35.

It is therefore now clearly established, first, that there is no flexible or variable standard of proof in civil proceedings, only that of the balance of probabilities; and, second, that the significance of the seriousness of the allegation is contingent on the facts of the particular case.

As explained in Jervis on Coroners (13^th Edn, 2014), paras 1-28 to 1-30, coroners once played a role in the criminal justice system in England and Wales. Until 1977, it was the duty of a coroner’s jury in any case where they found that the deceased had died through murder, manslaughter or infanticide, to state in the verdict the name of the person or persons considered to have committed the offence or to have been accessories to it. In such a case the record of the verdict (then known as the “inquisition”) would have the same effect as a bill of indictment committing the person or persons named to trial at what were originally the Courts of Assize and later the Crown Court. This power to commit originated before there were police forces or a fully established judicial system in England and Wales. It resembled, and was functionally equivalent to, that of the Grand Jury in criminal cases. The power was abolished by the Criminal Law Act 1977, section 56, which provided:

“At a coroner's inquest touching the death of a person who came by his death by murder, manslaughter or infanticide, the purpose of the proceedings shall not include the finding of any person guilty of the murder, manslaughter or infanticide; and accordingly a coroner's inquisition shall in no case charge a person with any of those offences.”

The principle established by that provision is now embodied in section 10(2) of the 2009 Act, which states that a determination at an inquest “may not be framed in such a way as to appear to determine any question of (a) criminal liability on the part of a named person …”

The practice where an inquest concludes with a conclusion of unlawful killing is now set out in an agreement between the Crown Prosecution Service, the National Police Chiefs’ Council, the Chief Coroner and the Coroners’ Society of England and Wales. This provides for the Crown Prosecution Service to be notified of the conclusion and to consider the significance for any charging decision or criminal investigation of any new evidence or information revealed in the coroner’s proceedings. It does not seem to us that the existence of this agreement makes the position in coroner’s proceedings materially different from the situation which exists in civil proceedings whereby, if there is evidence of criminal conduct, the court can refer the matter to the Crown Prosecution Service or to the Attorney-General.

The fundamental distinction between an inquest and a criminal proceeding was explained by Lord Lane LCJ in R v South London Coroner, ex parte Thompson, (1982) 126 SJ 625; The Times, 9 July 1982, when he said:

“… it should not be forgotten that an inquest is a fact-finding exercise and not a method of apportioning guilt. The procedure and rules of evidence which are suitable for one are unsuitable for the other. In an inquest it should never be forgotten that there are no parties, there is no indictment, there is no prosecution, there is no defence, there is no trial, simply an attempt to establish facts. It is an inquisitorial process, a process of investigation quite unlike a criminal trial where the prosecutor accuses and the accused defends, the judge holding the balance or the ring, whichever metaphor one chooses to use.”

Lord Lane went on to say: “The function of an inquest is to seek out and record as many of the facts concerning the death as public interest requires.”

Given the nature and function of a modern inquest, it seems to us that there is today no relationship or analogy between coroner’s proceedings and criminal proceedings which can in principle justify applying in coroner’s proceedings the criminal standard of proof.

There are also significant differences between coroner’s proceedings and civil proceedings. In particular, there are no parties to coroner’s proceedings, only interested persons. The process is inquisitorial rather than adversarial. An inquest is, as mentioned, a fact-finding inquiry and a conclusion reached at an inquest has no direct effect on legal rights. Section 10(2) of the 2009 Act goes further in precluding any determination of civil liability than it does in relation to criminal liability. Thus, whereas a conclusion at an inquest must not be framed “in such a way as to appear to determine any question of criminal liability on the part of a named person” (emphasis added), the prohibition on appearing to determine any question of civil liability is unqualified.

These differences, in our view, make it, if anything, less rather than more appropriate to apply in coroner’s proceedings a standard of proof higher than the civil standard. In circumstances where the function of an inquest is to determine the relevant facts concerning the death as accurately and completely as possible without determining even any question of civil liability, we can see no justification in principle for weighting the fact-finding exercise against any particular conclusion and requiring proof to any higher standard than the balance of probabilities. That is so even if the facts found disclose the commission of a criminal offence. Given that in civil proceedings the standard of proof of criminal conduct remains the ordinary civil standard, we can see no principled reason for adopting a different approach in coroner’s proceedings. The position is a fortiori where the conclusion under consideration is one of suicide as, although it was once a crime, suicide has not been a crime for over 50 years since that rule of law was abrogated by section 1 of the Suicide Act 1961.

On behalf of the claimant, Mr Bunting submitted that a finding of suicide is a serious matter. He quoted the description of suicide given in a case that we will come to later as “a drastic action which often leaves in its wake serious social, economic and other consequences.” He emphasised that, still today, suicide is a sensitive topic which in some communities may carry a stigma and is regarded by some religious teaching as a sin. That includes the teaching of the Roman Catholic Church. Mr Bunting sought permission to rely on a witness statement from a Deacon in the Church which confirms this. We give such permission although it is a matter of which we could in any event take notice. We were further informed that the claimant in this case and his family are devout Roman Catholics, which means that the conclusion at the inquest of James Maughan which was tantamount to one of suicide has had a particularly detrimental impact on them.

We cannot but sympathise with the distress felt by the claimant and other members of the deceased’s family. But such sympathy cannot alter the principles of law which the court must apply. We have noted already that the fact that a particular finding made in civil proceedings will potentially have serious social, economic or other consequences for one or more persons involved or interested in the proceedings is not a consideration which in law affects the standard to which the relevant facts must be proved. We have also explained why we can see no principled basis for taking a different approach in coroner’s proceedings.

The judicial duty of a coroner is to establish how the death occurred and to do so without fear or favour. The fact that the deceased’s family will be distressed by a particular conclusion cannot be a reason to alter the standard of proof required in order to reach that conclusion. In any case, the topic of suicide, sensitive as it is, is one on which there is a wide range of attitudes. In R (Evandro Lagos) v HM Coroner for the City of London [2013] EWHC 423 (Admin), a case we will consider later, it is apparent that the claimant positively wanted a determination that his wife had committed suicide rather than an open conclusion because he saw this as necessary to recognise her autonomy and dignity as a human being. It is not for the law in this area to adopt one conception of human dignity in preference to another. Still less would it be right to allow the attitudes or wishes of family members, however strongly they may be felt, to shape the fact-finding approach to be taken by a coroner or jury.

Unless bound by authority to decide otherwise, our clear view would therefore be that a conclusion of suicide, whether expressed as a narrative statement or in short-form, is required to be proved to the civil, and not the criminal, standard of proof.

Mr Bunting made a submission that the court is bound to decide otherwise by note (iii) in Form 2 contained in the Schedule to the Coroners (Inquests) Rules 2013. This note, quoted at para 13 above, states that the standard of proof required for the short-form conclusions of “unlawful killing” and “suicide” is the criminal standard of proof. Mr Bunting submitted that this requirement has thereby been given statutory force so that a court is bound to apply it.

We observe that, if this submission were correct, it would defeat the claimant’s case because note (iii) also says that, for a narrative statement, the standard of proof is the civil standard of proof. If the note has statutory authority, it follows that the coroner’s direction to the jury in the present case to apply the balance of probabilities when making their narrative statement – including for the purpose of deciding whether the deceased intended to kill himself – was correct.

But in our view the submission is not correct. We accept that the power under section 45 of the 2009 Act to make coroners’ rules is sufficiently broad to enable a rule to be made stipulating the standard of proof to be applied in coroner’s proceedings. But if the intention had been to make such a rule, the appropriate place to do so would be in the body of the rules, and not in a prescribed form. Form 2, as is clear from its subject matter, is simply a form which must be used to record the determination which the coroner or jury has made. Its function is not to enact rules about how evidence given at an inquest must be approached. In our view, the reasonable interpretation of note (iii) is simply as stating, for the assistance of those using the form, what the law with regard to the standard of proof is understood to be, and not as legislating what the law shall be.

We therefore reject the suggestion that note (iii) has set in statutory stone the applicable standard of proof.

We turn to consider the position at common law. Although counsel for the claimant submitted that decades (if not centuries) of case law have established that a verdict of suicide at an inquest can only be returned on the criminal standard of proof, the authorities cited to us do not bear this out.

There is a line of cases in which courts have emphasised that suicide should never be presumed. The earliest example to which we were referred was Southall v Cheshire County Mews Co Ltd (1912) 5 BWCC 251. This was in fact a workman’s compensation case in which a county court judge had found that an accident at work had caused a workman to commit suicide and on that basis had awarded his dependants a sum of money. The man had been discovered drowned in a canal some 400 yards from his home but with no evidence to show how he came there. The Court of Appeal held that there was insufficient evidence to justify the judge’s finding. Cozens-Hardy MR said:

“Suicide is never to be presumed. In the next place there is no evidence of suicidal tendency. … The judge seems to have thought it was more likely that the man committed suicide than anything else. The judge is not entitled to act upon a surmise of that nature.”

In R v Huntbach, ex parte Lockley [1944] KB 606, the Court of Appeal applied this authority to coroner’s proceedings in holding that there was no evidence on which a verdict of suicide could properly have been returned at an inquest. Viscount Caldecote CJ observed (at 608):

“Passages in [the coroner’s] affidavit illustrate how much he was basing his verdict on guesswork instead of on particulars which had been proved in evidence. The conclusions stated in those passages are nothing more than the expression of a theory which is attractive in its probability. It is not probability, however, which determines verdicts, but proved facts, and, if facts which justify a specific verdict are not proved at an inquest, there is no alternative but to return an open verdict.”

Croom-Johnson J (at 610), agreeing, described the error made by the coroner as being that “he went beyond his power to draw inferences and filled in gaps in the evidence which was before him.”

To similar effect, in Re Davis [1968] 1 QB 72, 82, Sellers LJ said:

“Suicide is not to be presumed. It must be affirmatively proved to justify the finding.”

In that case, however, the Court of Appeal declined to quash a coroner’s verdict of suicide in a case where the deceased had jumped from a second-floor window because they considered that any coroner on a reconsideration of the cause of death would probably come to the same conclusion.

In R v City of London Coroner, ex parte Barber [1975] 1 WLR 1310 the deceased, after going to the pub with his wife for lunch and consuming around seven pints of beer, went up to the roof above their flat, as he often did. From the roof he fell three storeys to his death in the street below. There were railings which, although they could easily be climbed over, had the appearance of being an effective means of preventing an accidental fall. The coroner recorded a verdict of suicide but it was quashed by a Divisional Court. Lord Widgery CJ said (at 1313):

“If a person dies a violent death, the possibility of suicide may be there for all to see, but it must not be presumed merely because it seems on the face of it to be a likely explanation. Suicide must be proved by evidence, and if it is not proved by evidence, it is the duty of the coroner not to find suicide but to find an open verdict.

I approach this case, applying a stringent test, and asking myself whether on the evidence which was given in this case any reasonable coroner could have reached the conclusion that the proper answer was suicide. I take the view that no reasonable coroner properly understanding the obligation to prove suicide could have found suicide in this case. There is, as I see it, no single fact which definitely points to the deliberate taking of this man's life and every possibility that the matter was an accident and no more.”

This passage was quoted and the principle applied in R v Essex Coroner, ex parte Hopper, unreported, 13 May 1988. In that case a young man was found dead at his home, slumped in a chair, having evidently been killed by a shotgun which had discharged very close to his head. There was a pile of ironing half done and the iron had been left on. The deceased was an apparently happy person and there was no evidence of any depression or state of mind which would make suicide in the least likely. However, the gun could only be fired if the safety catch was off and had a trigger which could not have led to a discharge of the weapon by ordinary accidental means, such as dropping the gun. The coroner concluded that the deceased must have committed suicide. A Divisional Court, applying ex parteBarber, quashed the verdict on the ground that the coroner “did not adequately exclude, nor could he adequately have excluded, the possibility that the death had been caused by some unexplained accident.”

In none of these cases is there to be found any statement let alone decision to the effect that suicide must be proved to the criminal standard of proof. The essential point made in the judgments is that it is wrong to conclude that a person committed suicide simply because other explanations of the death appear improbable or more improbable, and that a conclusion of suicide is only justified if it is proved by evidence. This does not mean such a conclusion cannot properly be inferred from proven facts in an appropriate case. For example, in ex parte Hopper Parker LJ noted counsel’s concession that if a man is found dead in his car with a pipe connecting the exhaust of the car into the interior of the car, it would be perfectly proper for a coroner to conclude that there was no possible explanation other than suicide. But a finding that the deceased did a deliberate act which caused his death with the intention that the act would have that consequence will not be justified if the possibility cannot be excluded that the death was caused by some unexplained accident (as in ex parte Hopper) or if the deceased may only have intended to cause himself harm not resulting in death.

This reasoning is no more than an illustration of a more general point about the standard of proof. To describe the standard of proof in civil proceedings as the “balance of probabilities” is capable of misleading if this is taken to mean that just because one theory appears more probable, or less improbable, than any other it must be accepted. Such an approach overlooks the fact that proof depends not only on the balance of evidence but on its sufficiency or weight. Evidence can be balanced because it is conflicting – different pieces of evidence favour different hypotheses – or because there simply is not enough of it to reach a concluded view. John Maynard Keynes explained the importance of evidential weight when he wrote in “A Treatise on Probability” (1921) p78:

“As the relevant evidence at our disposal increases, the magnitude of the probability of the argument may either decrease or increase, according as the new knowledge strengthens the unfavourable or the favourable evidence; but something seems to have increased in either case – we have a more substantial basis upon which to rest our conclusion. I express this by saying that an accession of new evidence increases the weight of an argument. New evidence will sometimes decrease the probability of an argument, but it will always increase its ‘weight’.”

The significance of this point for the law of evidence was explained in Rhesa Shipping Co SA v Edmunds (The ‘Popi M’) [1985] 1 WLR 948, where the House of Lords held that the trial judge was wrong, in a case where the evidence was incomplete and not all the relevant facts were known, to regard himself as compelled to choose between two theories, both of which he regarded as extremely improbable, or one of which he regarded as extremely improbable and the other as virtually impossible. The proper approach in such a case is for the judge:

“to say simply that the evidence leaves him in doubt whether the event occurred or not, and that the party on whom the burden of proving that the event occurred lies has therefore failed to discharge such a burden.”

See [1985] 1 WLR 948, 956. The reference in this passage to the judge being left in doubt is not to be understood as suggesting that proof beyond reasonable doubt (i.e. to the criminal standard) is required in a civil case. It simply means that, in order to find a hypothesis proved, there must be sufficient evidence to enable the judge (or other fact-finder) adequately to exclude all other possibilities.

A further notion that no doubt underlies the maxim that suicide must never be presumed is that suicide is an inherently improbable cause of death. People seldom choose to end their own lives, at least if they are in ordinary mental health. In several of the cases mentioned, therefore, where there was no evidence that suggested any history of depression or risk of suicide, correspondingly cogent evidence was needed before the coroner or jury could properly conclude that this was the cause of death.

As we have indicated, none of the cases so far cited held or contained any statement of opinion that a conclusion of suicide at an inquest may only be reached if the elements of suicide are proved to the criminal standard. However, in R v West London Coroner, ex parte Gray [1988] QB 467 a Divisional Court considered that this was the implication of this line of cases and, in particular, of ex parte Barber. Ex parte Gray was a case in which a coroner’s jury had returned a verdict of unlawful killing. The Divisional Court quashed the verdict on the grounds that the coroner had misdirected the jury in a number of respects. One of the errors made by the coroner was held to be that he had failed to direct the jury that they should apply the criminal standard of proof.

In discussing this issue (at 477), Watkins LJ (with whom Roch J agreed) observed that there was a lack of direct authority on the point. He then referred to the cases which have emphasised that suicide is not to be presumed. He noted that suicide was no longer a crime but observed that “it is still a drastic action which often leaves in its wake serious social, economic and other consequences.” Watkins LJ then cited ex parte Barber. After quoting the passage from the judgment of Lord Widgery CJ which we have quoted at para 53 above, he said:

“It will be noted that Lord Widgery CJ alluded to the stringent test, but without reference to what may be called the conventional standards of proof. I cannot believe, however, that he was regarding proof of suicide as other than beyond a reasonable doubt. I so hold that that was and remains the standard. It is unthinkable, in my estimation, that anything less will do. So it is in respect of a criminal offence. I regard as equally unthinkable, if not more so, that a jury should find the commission, although not identifying the offender, of a criminal offence had been committed, without being satisfied beyond a reasonable doubt.”

Although expressed in terms of a holding, the remarks made about proof of suicide in this passage were, in our view, in truth only dicta since no possibility that the deceased might have committed suicide arose for determination in ex parte Gray. Mr Bunting submitted that the remarks were a necessary step in the court’s reasoning to the conclusion that proof to the criminal standard was required for a verdict of unlawful killing and were therefore part of the ratio decidendi. We are not persuaded by this submission, as the essential reason for holding that proof to the criminal standard was required to return a verdict of unlawful killing appears to have been that such a verdict involved a finding that a criminal offence had been committed. But even if the remarks about proof of suicide were part of the ratio of the case, we are satisfied that they are wrong.

In the first place, the remarks seem to us to have been based on a misreading of Lord Widgery CJ’s judgment in ex parte Barber. As already indicated, that case was not concerned with the standard of proof at all but with the different point that suicide must not be presumed and must be proved by evidence. When Lord Widgery CJ referred to applying “a stringent test”, we think it reasonably clear that he was not referring to the test for proving suicide, as Watkins LJ appears to have thought, but to the standard of review of the coroner’s decision. That is to say, the “stringent test” was the test which Lord Widgery CJ then proceeded to apply of asking “whether on the evidence which was given in this case any reasonable coroner could have reached the conclusion that the proper answer was suicide.”

Secondly and more fundamentally, no reference was made in the judgments in ex parte Gray to the rule of law that, where a question is raised in civil proceedings as to whether a criminal offence has been committed, the standard of proof applicable is the civil, and not the criminal, standard. It appears from the law report that the decision of the Court of Appeal in Hornal v Neuberger Products Ltd [1957] 1 QB 247, which established that proposition, was cited in argument. But that authority is not mentioned in the judgments, which make no attempt to explain why a different principle should apply in relation to a coroner’s inquest. This omission is particularly striking, as Watkins LJ quoted (at 473) the passage from the judgment of Lord Lane CJ in ex parte Thompson which we have quoted at para 37 above, contrasting an inquest with a criminal trial and emphasising that an inquest “is a fact-finding exercise and not a method of apportioning guilt.”

A Divisional Court hearing a claim for judicial review is not bound by a decision of another Divisional Court in the same way as the Court of Appeal is bound by its own earlier decisions. The same rule applies as for a single judge sitting at first instance in the High Court when another decision of similar status is cited – namely, that the court should follow the earlier decision as a matter of judicial comity unless convinced that the decision is wrong: see R v Manchester Coroner, ex parte Tal [1985] QB 67, 80-1. Even if, therefore, contrary to our view, ex parte Gray should be taken as a decision that the standard of proof of suicide in coroner’s proceedings is the criminal standard, we are not bound by it in circumstances where we are convinced, for the reasons given, that it is mistaken.

We are reinforced in this conclusion by the fact that the view expressed in ex parte Gray about the standard of proof of suicide was disapproved by the Court of Appeal in Braganza v BP Shipping Ltd [2013] EWCA Civ 230; [2013] 2 Lloyd’s Rep 351, paras 15-16. The Supreme Court agreed with the Court of Appeal on this point: Braganza v BP Shipping Ltd [2015] UKSC 17; [2015] 1 WLR 1661, paras 33-35. We also note that the Ontario Court of Appeal declined to follow ex parte Gray on this issue, holding that the civil standard of proof applies to a finding of suicide at an inquest: see Re Beckon (1992) 93 DLR (4^th) 161. The same view has been taken in New Zealand: see Re Sutherland (Deceased) [1994] 2 NZLR 242, 251.

Mr Bunting cited two cases decided since ex parte Gray in which the standard of proof for a conclusion of suicide at an inquest has been assumed or held to be the criminal standard.

The first in time was R (Jenkins) v HM Coroner for Bridgend and Glamorgan Valleys [2012] EWHC 3175 (Admin). In that case a coroner’s jury returned a verdict of suicide which was quashed by a Divisional Court. The brief facts were that the deceased, a young man of 23, was killed by a train which hit him when he was sitting slumped forward on a railway track in the early hours of the morning. There was evidence that he had been drinking heavily and had got very drunk. Witnesses who had seen and spoken to the deceased in the period leading up to his death had detected no sign that he was in a low mood, let alone a suicidal state of mind. The coroner directed the jury that, to return a verdict of suicide, they would have to be sure that the deceased did a deliberate act with the intention of taking his own life; not just satisfied on the balance of probabilities. No challenge was made to the correctness of that direction, which was treated in the Divisional Court as an accurate statement of the law. The basis for doing so appears to have been the following words attributed (at para 19) to Parker LJ in the case of ex parte Hopper, reviewed at para 54 above:

“It was possible that the jury could not conclude, so that they were sure, that the deceased had committed suicide. Nor, on a balance of probability, that his death was an accident. In which case they could return an open verdict.”

The curiosity about this purported quotation is that nowhere in the judgment of Parker LJ in ex parte Hopper are these words (or any words to similar effect) actually used.

In any event, the assumption that a verdict of suicide could only be returned if the jury was sure that the deceased had committed suicide was not necessary to the court’s decision in the Jenkins case. The reason why the verdict was quashed was essentially that the coroner had failed to direct the jury about how they should approach circumstantial evidence. Pitchford LJ said (para 29):

“There was no emphasis upon the essential direction that before the jury could return a verdict of suicide, they had to be sure that every other alternative had been excluded by the evidence; that in a circumstantial case, such as the present, it was not permissible to fill in gaps in the evidence; that there was an important difference between speculation and the drawing of an inference which excluded all other reasonable possibilities.”

This case, in our view, is properly understood as a further illustration of the principle that suicide is not to be presumed and of the distinction between drawing an inference supported by the evidence, which is permissible, and filling in gaps in the evidence, which is not. It is not an authority which decides that suicide must be proved to the criminal standard of proof.

The last case relied on by Mr Bunting is such an authority. In R (Evandro Lagos) v HM Coroner for the City of London [2013] EWHC 423 (Admin) the claimant applied for judicial review of an open verdict recorded at an inquest into the death of his wife, contending that the coroner should have returned a verdict of suicide. So far as the judgment records, the claimant, who had no legal representation, did not expressly raise any issue about the applicable standard of proof. But he alleged that the verdict was irrational in the light of the circumstances of his wife’s death and the evidence of her suicidal state of mind and behaviour in the period leading up to her death. He also argued that a verdict of suicide would have acknowledged and respected the way in which she chose to end her life, and thus accorded her the dignity to which she was entitled under article 1 of the Universal Declaration of Human Rights.

Lang J dismissed the claim. In doing so, she held – principally on the authority of ex parte Gray – that the coroner had applied the correct legal test in directing himself that a verdict of suicide could not be returned because he could not be sure that the deceased had intended to kill herself. After reviewing the case law, Lang J summarised the approach of the courts to suicide verdicts as reflecting “(a) the fact that a finding of suicide is a serious matter which can cause serious distress and stigma, and other adverse consequences; and (b) the complexities of human psychology which can cause people to harm themselves seriously or to put themselves in very dangerous positions without the clear intention to end their lives” (para 37).

We think it important to emphasise that Lang J, sitting as a single judge, was more narrowly constrained by the doctrine of precedent to adhere to the view taken in ex parte Gray about the standard of proof required for a conclusion of suicide than are we in the present case, sitting as a Divisional Court. In R v Manchester Coroner, ex parte Tal [1985] QB 67, 81, Robert Goff LJ (giving the judgment of the court) said that it was “difficult to imagine” that a single judge exercising judicial review jurisdiction would ever depart from a decision of a Divisional Court. We have explained, however, why we are convinced that the view, robustly expressed as it was in ex parte Gray, that suicide must be proved to the criminal standard is incorrect in law and why we can and should decline to follow it. The same reasoning applies to the decision in the Lagos case. It is also relevant to note that – perhaps because the claimant was unrepresented and the court may not have had the benefit of the same high quality of assistance as we have had from counsel in the present case – no reference was made in the Lagos case to the settled rule that in civil proceedings the standard of proof of criminal conduct (or any other serious allegation) is the ordinary civil standard. There was accordingly no consideration of what possible justification there could be for applying a different rule in coroner’s proceedings.

As for the two policy considerations identified by the judge as underlying the approach of the courts, we have explained earlier why, although we recognise that a finding of suicide is a serious matter which can cause serious consequences, this is not a consideration which can in principle or consistently with the approach of the law in civil proceedings affect the legal standard of proof. We also recognise that the complexities of human psychology may make the requisite intention difficult to prove, even to the ordinary civil standard. But the fact that the human mind is often hard to fathom cannot be a reason for imposing a higher than normal standard of proof.

We are led to the conclusion that the Lagos case was wrongly decided.

In summary, we are unable to accept the claimant’s contention that a conclusion of suicide at an inquest requires proof to the criminal standard. We are satisfied that the authorities relied on to support that contention either on analysis do not support it or do not correctly state the law. We consider the true position to be that the standard of proof required for a conclusion of suicide, whether recorded in short-form or as a narrative statement, is the balance of probabilities, bearing in mind that such a conclusion should only be reached if there is sufficient evidence to justify it.

It follows that there was nothing wrong with the coroner’s directions to the jury in this case and that the jury’s conclusion was lawful. The claim must therefore be dismissed.