Alexander Lewis-Ranwell v G4S Health Services (UK) Ltd & Ors

The Claimant in this case, who is the Respondent to this appeal, is aged 32. He was diagnosed with schizophrenia in his mid-twenties and had spells in psychiatric intensive care in 2016 and 2017. On 10 February 2019, in the course of a serious psychotic episode, he attacked and killed three elderly men in their homes in Exeter in the delusional belief that they were paedophiles. He was charged with murder but following a trial in Exeter Crown Court he was found not guilty by reason of insanity: as explained more fully below, that meant that because of his mental illness he did not know at the time of the killings that what he was doing was wrong. He was ordered to be detained in Broadmoor Hospital pursuant to a hospital order with restrictions under sections 37 and 41 of the Mental Health Act 1983.

In the two days before the killings the Claimant had twice been arrested by the Devon and Cornwall Police (“the Police”) and detained for some time at Barnstaple Police Station before being eventually released. In short:

During both periods of detention the Claimant behaved violently and erratically and was apparently mentally very unwell. He was seen or spoken to by mental health professionals employed by G4S Health Services (UK) Ltd (“G4S”) and Devon Partnership NHS Trust (“the Trust”). A face to face assessment by the mental health nurse employed by the Liaison and Diversion Service of the NHS Trust was discussed but did not take place. The need for a Mental Health Act Assessment was discussed with an Approved Mental Health Professional employed by the Council but was not arranged.

On 4 February 2020 the Claimant commenced proceedings in the High Court against G4S, the Police, the Trust and the Council. In broad terms it is his case that it should have been obvious to all concerned during both detentions that if he were released there was a real risk that he would injure other people, and that the necessary steps should have been taken to keep him in detention until it was safe for him to be released. The claims are advanced in negligence and under section 7 of the Human Rights Act 1998. The heads of damage pleaded in the Particulars of Claim are for personal injury, loss of liberty, loss of reputation, and “pecuniary losses”. The Claimant also seeks an indemnity in respect of any claims brought against him “as a consequence of his violence towards others on 9-11 February 2019”.

On 20 July 2021 the Council issued an application for the claim against it to be struck out, and similar applications were subsequently made by G4S and the Trust, though not by the Police. Those three Defendants are the Appellants in this appeal. In each case the ground for the application was, broadly speaking, that the Defendants were entitled to rely on “the illegality defence” – that is, the rule that the Court will not entertain a claim which is founded on a claimant’s own unlawful act – because the claim was based on the consequences of the Claimant’s three unlawful homicides. The illegality defence is often described as depending on “the ex turpi causa principle” (or “rule”), referring to the maxim ex turpi causa non oritur actio; and I will in this judgment use both labels indifferently.

The applications were heard at Exeter Crown Court by Garnham J on 9 and 10 March 2022. By that time the Appellants had accepted that the applications could only be pursued as regards the claim in negligence and that the claim under the 1998 Act would continue in any event. (Footnote: ¹)

By a judgment handed down on 20 May 2022 Garnham J dismissed the applications. The essence of his reasoning is that because the verdict of “not guilty by reason of insanity” meant that the Claimant did not know that what he was doing was wrong his conduct did not have the necessary element of “turpitude”. As he put it at para. 135:

“The Defendants can show that the death of the three men was the result of deliberate acts of the Claimant. But it is not sufficient to exclude liability that the immediate cause of the damage was the deliberate act of the claimant. The defendants must point to a turpitudinous act, an act of knowing wrongfulness. That means they must show that the claimant was guilty of criminal or quasi criminal acts, acts that engage the public interest. They have failed to do so.”

This is an appeal against that decision. G4S has been represented by Mr Gurion Taussig; the Trust by Ms Judith Ayling KC and Mr James Goudkamp; and the Council by Mr Andrew Warnock KC and Mr Jack Harding. The Claimant has been represented by Ms Selena Plowden KC and Mr Christopher Johnson. The representation was the same before Garnham J, except that Ms Ayling and Mr Warnock appeared without juniors. Mr Taussig did not advance any oral submissions but, in addition to relying on his skeleton argument, adopted the submissions of Mr Warnock and Ms Ayling.

The question whether the illegality defence operates in a case where the claimant was insane at the time that he or she did the unlawful act is not the subject of any binding authority. In Clunis v Camden and Islington Health Authority [1998] QB 978 this Court held that a mentally ill person who had been convicted of manslaughter by reason of diminished responsibility was barred by the ex turpi causa principle from bringing a claim against his doctors for negligent treatment which was said to have caused or contributed to his committing the offence; and that decision has since been upheld by the House of Lords in Gray v Thames Trains Ltd [2009] UKHL 33, [2009] 1 AC 1339, and by the Supreme Court in Henderson v Dorset Healthcare University NHS Foundation Trust [2020] UKSC 43, [2021] AC 563. However, the reasoning in those decisions, though clearly relevant to this case, is not determinative because diminished responsibility is not the same as insanity. The issue has been directly considered in some U.S. and Commonwealth cases, and also in a recent decision of the High Court, Traylor v Kent & Medway NHS Social Care Partnership Trust [2022] EWHC 260 (QB), [2022] 4 WLR 35.

In those circumstances it will be necessary to consider the case-law – both the cases identified above and the general authorities on the illegality defence – in some detail. But before I embark on that exercise I need first to summarise the law on the distinct though related question of the liability, both criminal and tortious, of persons who commit unlawful acts as a result of mental illness. Accordingly the structure of this judgment is:

Since the issue is one of legal principle, I do not believe that it is necessary to set out more of the facts than appears at paras. 1-3 above: those interested can find a fuller account at paras. 7-35 of Garnham J’s judgment. I should emphasise that since this is a strike-out application the Claimant’s allegations against the various Defendants have not yet been determined. But it is easy to see how it is a cause for concern that he should have twice been released when he was on any view mentally very unwell and had on the second occasion already committed a serious assault on a stranger. In the course of his criminal trial the jury sent a note to the Judge in the following terms:

“We the Jury have been concerned at the state of psychiatric health service provision in our county of Devon. Can we be reassured that the failings in care for [the Claimant] will be appropriately addressed following this trial.”

PRELIMINARY: LIABILITY OF THE MENTALLY ILL

Section 2 (1) of the Trial of Lunatics Act 1883, as originally enacted, provided that:

“Where in any indictment or information any act or omission is charged against any person as an offence, and it is given in evidence on the trial of such a person for that offence that he was insane, so as not to be responsible, according to law, for his actions at the time when the act was done or the omission made, then, if it appears to the jury before whom such person is tried that he did the act or made the omission charged, but was insane as aforesaid at the time when he did or made the same, the jury shall return a special verdict to the effect that the accused was guilty of the act or omission charged against him, but was insane as aforesaid at the time when he did the act or made the omission.”

A special verdict in that form was generally referred to as a verdict of “guilty but insane”.

The form of the special verdict provided for by the 1883 Act was amended by section 1 of the Criminal Procedure (Insanity) Act 1964 so as to read “that the accused is not guilty by reason of insanity”. The term “insanity” and its cognates have rather fallen out of ordinary use since then, at least in a medical context, but I will in this judgment avoid the temptation to use more modern terminology because to depart from the language of the statute would risk inaccuracy and confusion.

The meaning of “insanity” for the purpose of the 1883 Act as amended is still governed by the opinion of the judges (Maule J dissenting) in M’Naghten’s Case (1843) 10 Cl & F 200 [8 ER 718], at p. 210, namely that:

“… the jurors ought to be told … that to establish a defence on the ground of insanity, it must be proved that, at the time of the committing of the act, the party accused was labouring under such a defect of reason, from disease of the mind, as not to know the nature and quality of the act he was doing; or, if he did know it, that he did not know he was doing what was wrong.”

It will be noted that there are two alternative elements in the definition – (a) that the defendant did not know what he or she was doing or (b) that they did not know that it was wrong.  The nature of the two elements is fully and helpfully analysed at paras. 34-49 of the judgment of the Divisional Court, given by Irwin LJ, in Loake v Director of Public Prosecutions [2017] EWHC 2855 (Admin), [2018] QB 998, but I need not set the passage out here. The finding of insanity in the Claimant’s case was evidently on the second basis: that is, he must have known that he was killing (or at least seriously injuring) his victims – as Garnham J put it, he was acting deliberately – but because of his delusions he did not know that what he was doing was wrong.  We were also referred to the recent decision in R v Keal [2022] EWCA Crim 341, [2022] 4 WLR 41, which considers in more detail what is involved in a finding that the defendant did not know that what they were doing was wrong; but it does not add anything relevant for our purposes.

Where a special verdict has been returned in a murder case and the court has the power to make a hospital order, the Court is obliged to make such an order combined with a restriction order. The effect of a combined order is shortly explained at para. 27 of the judgment in Henderson (with a cross-reference to the fuller explanation given by Mustill LJ in R v Birch (1989) 90 Cr App R 78); but in broad terms it is that the defendant will be committed to a special hospital for an indeterminate, and typically prolonged, period.

There was some discussion before us about how the special verdict should be characterised. In Felstead v The King [1914] AC 534 the House of Lords made clear, in the context of an issue about whether the defendant could appeal from the special verdict, that it was indeed “a verdict of acquittal of the accused”; and that characterisation is of course reinforced by the change in the statutory language. But that does not mean that the act itself can be described as lawful, and the Appellants submit that the Claimant’s acts in the present case can properly be described as criminal, albeit that he has no criminal responsibility for them.

In support of that argument Mr Warnock pointed out that in cases of the present kind both the actus reus and the mens rea for murder are present – see para. 41 in Loake, where Irwin LJ says:

“If a man intentionally kills his wife because of his deluded belief that he is under threat from a representative of Satan and has received a divine order to slay, and that it is lawful to comply with divine orders, then he possesses the mens rea for murder but is not guilty of murder because he does not know that what he is doing is unlawful.”

He also referred us to Attorney-General’s Reference (no. 3 of 1998) [2000] QB 401 in which it was held that, for the purposes of obtaining a special verdict under the 1883 Act the Crown had only to prove “that the defendant has caused a certain event or that responsibility is to be attributed to him for the existence of a certain state of affairs, which is forbidden by criminal law ...” (p. 411F).

For a similar purpose Ms Ayling referred us to passages from the academic literature which make the point in various ways that verdicts of insanity or diminished responsibility do not imply that the act is in any way lawful. For example, a passage in Professor Gardner’s Offences and Defences (Oxford 2007) reads (pp. 88-89):

“It is important to distinguish between actions that are not wrong at all, and actions that are wrong but are justified; between actions that are wrong but justified and those that are wrong and unjustified but excused; and between unjustified but excused wrongs and those which are neither justified nor excused but are nevertheless not punishable. The last category includes, but is broader than, the category of wrongs which are committed without responsibility [because the defendant was insane or his/her responsibility was diminished].”

And in his article Diminished Mental Capacity as a Criminal Law Defence (1974) 37 MLR 264, Professor Fingarette says (pp. 270-271):

“An insane person who kills someone is doing what the criminal law is intended to deter men from doing. His defect with regard to criminal law is grave, notwithstanding the misleading ‘Not Guilty’ verdict.  The point here is not that he is guilty, nor that he is innocent but that his radical incompetence is such that he should not be morally judged at all.”

I have to say that I did not find this part of the Appellants’ submissions helpful. It is no doubt a reasonable use of language to say that in a special verdict case there has been a criminal act (and likewise that the killer is doing something which the criminal law is intended to deter) even though no-one is criminally responsible for it; and Professor Gardner’s analysis is highly persuasive for the purposes for which he deploys it. But the question which we have to decide is what the policy of the law is, or should be, as regards the recovery of damages for the consequences of an act of that kind; and for that purpose these characterisations do not advance the argument. I will in this judgment generally use the neutral term “unlawful act”.

Importantly also, because it features in the cases of Clunis, Gray and Henderson referred to above, section 2 of the Homicide Act 1957 introduces a partial defence of diminished responsibility in cases of homicide where, in short, the defendant is not insane according to the M’Naghten definition but where his or her ability to understand the nature of their conduct, or form a rational judgment or exercise self-control, is “substantially impaired” as a result of a recognised medical condition. Where that is proved, a defendant who would otherwise have been guilty of murder is liable to be convicted of manslaughter. In such a case the Court will typically make an order under section 37 of the 1983 Act but may or may not make an order under section 41.

Finally, I should note that cases of insanity, where the defendant has done the unlawful act but without the knowledge that it is wrong, are distinct from cases of “automatism”, where he or she has not consciously acted at all. That will be the case, for example, where a driver loses control of their car and kills someone as a result of a stroke or a hypoglycaemic episode. I mention automatism only because it, or something very like it, is identified as the only mental state giving rise to a defence to claims in tort: see para. 27 below.

The liability of the mentally ill in tort is more extensive than their liability in crime. This is clear from two decisions – Morriss v Marsden [1952] 1 All ER 925and Dunnage v Randall [2015] EWCA Civ 673, [2016] QB 639 – which I consider in turn.

In Morriss v Marsden the defendant attacked the plaintiff while in a seriously psychotic state. He was sued for assault and battery. Stable J found (see pp. 926-927):

“… that the defendant was not in a condition of automatism or trance at the time of the attack on the plaintiff, but that his mind directed the blows he struck, and that at the material time he was a catatonic schizophrenic and a certifiable lunatic who knew the nature and quality of his act, but whose incapacity of reason arising from the disease of his mind was of so grave a character that he did not know that what he was doing was wrong.”

The defendant submitted that it followed from that finding, as it would in the case of criminal liability (Footnote: ²), that he was not liable. Stable J rejected that submission. He accepted that it was necessary to prove that the defendant had acted voluntarily (Footnote: ³) but not that it was necessary that he should have appreciated that what he was doing was wrong. At p. 927H he said:

“... I accept the view that an intention – i.e., a voluntary act, the mind prompting and directing the act which is relied on, as in this case, as the tortious act – must be averred and proved. For example, I think that, if a person in a condition of complete automatism inflicted grievous injury, that would not be actionable.”

After discussing, obiter, whether the position might be different where (unlike in the case before him) the nature of the tort required some mental element, he continued at p. 928 E-F:

“The next matter to consider is whether, granted that the defendant knew the nature and quality of his act, it is a defence in this action that, owing to mental infirmity, he was incapable of knowing that his act was wrong. If the basis of liability be that it depends, not on the injury to the victim, but on the culpability of the wrongdoer, there is considerable force in the argument that it is, but I have come to the conclusion that knowledge of wrongdoing is an immaterial averment, and that, where there is the capacity to know the nature and quality of the act, that is sufficient although the mind directing the hand that did the wrong was diseased.”

Accordingly he gave judgment for the plaintiff.

Stable J’s observation that the basis of the defendant’s liability was not his culpability but the injury to the victim reflects the principle that tort law is generally concerned with the compensation of a person who has suffered loss and not with the punishment of the person who has caused the loss, and that for that reason it is just that the person who has suffered the loss should recover from the person who caused it even if they cannot be regarded as culpable. This principle goes back at least to Weaver v Ward, (1616) Hob 134, 80 ER 284, where the Court of Common Pleas said that although there was no felony

“… if a lunatick kill a man, or the like, because felony must be done animo felonico, yet in trespass, which tends only to give damages according to hurt or loss, it is not so; and therefore if a lunatick hurt a man, he shall be answerable in trespass”. (Footnote: ⁴)

Morriss v Marsden was concerned with trespass. In Dunnage v Randall this Court (Arden, Rafferty and Vos LJJ) had to consider the question of liability of a mentally ill person in the context of negligence. In that case a visitor to the claimant’s home, who was suffering from paranoid schizophrenia, set fire to himself and died. The claimant was badly burned when trying to intervene: there was no finding that the visitor was intending to injure him. He brought proceedings against the visitor’s estate for negligence.

All three members of the Court delivered substantive judgments. There are some differences of approach between them but each expresses agreement with the others, and it seems clear that the essential reasoning is the same. In brief, they regarded the case as indistinguishable from Morriss v Marsden (see per Arden LJ at para. 146). All that was necessary was that the defendant’s mind, however diseased or deluded, should direct their hand: in practice that means that the defendant will only escape liability in cases of automatism (see per Arden LJ at paras. 146-147 and Vos LJ at paras. 132-133).

Both Arden and Vos LJJ drew support for their conclusion from the decision of the House of Lords in Corr v IBC Vehicles Ltd [2008] UKHL 13, [2008] AC 884. In that case an employee who had suffered a serious accident at work as a result of the employer’s negligence developed a severe depressive illness and committed suicide. His widow brought proceedings against the employer both as administratrix of his estate and under the Fatal Accidents Act 1976. The main issues in the House of Lords were whether the duty of care owed by the employer extended to the risk of suicide and whether the employee’s suicide broke the chain of causation. But there was also an issue as to whether the damages could be reduced for contributory negligence. It was held that there was insufficient evidence to justify any such reduction; but Lord Scott, Lord Mance and Lord Neuberger all took the view that in principle a deduction might have been made, essentially because although the deceased’s depression had “impaired” his mental state it had not necessarily wholly “overborne his personal autonomy” (Lord Neuberger’s terminology) or reduced him to an “automaton” (Lord Scott’s).

We were also referred to Williams v Williams [1964] AC 698, in which the House of Lords held that in matrimonial proceedings a husband’s conduct towards his wife amounted to cruelty notwithstanding that it was the result of paranoid schizophrenia. The majority found that the application of the M’Naghten rules in this context was inappropriate because the court was concerned not with the respondent’s culpability but with the effect of their conduct on the petitioner. This is effectively the same approach as is taken as regards liability in tort.

THE ILLEGALITY DEFENCE: REVIEW OF THE AUTHORITIES

The principles underlying the illegality defence generally and the correct approach to its application were authoritatively restated by the Supreme Court in Patel v Mirza [2016] UKSC 42, [2017] AC 467. At this stage I need only say that the majority rejected a so-called “rules-based approach” (and more particularly the “reliance-based approach” in Tinsley v Milligan [1994] 1 AC 340) in favour of a more flexible approach based on an assessment of the competing public policy considerations and proportionality factors in the particular case. The leading judgment for the majority was given by Lord Toulson. He summarised his conclusion at para. 120, as follows:

“The essential rationale of the illegality doctrine is that it would be contrary to the public interest to enforce a claim if to do so would be harmful to the integrity of the legal system (or, possibly, certain aspects of public morality, the boundaries of which have never been made entirely clear and which do not arise for consideration in this case). In assessing whether the public interest would be harmed in that way, it is necessary (a) to consider the underlying purpose of the prohibition which has been transgressed and whether that purpose will be enhanced by denial of the claim, (b) to consider any other relevant public policy on which the denial of the claim may have an impact and (c) to consider whether denial of the claim would be a proportionate response to the illegality, bearing in mind that punishment is a matter for the criminal courts.”

The three questions identified at (a)-(c) in that passage had been described at para. 101 as “the trio of necessary considerations”.

However, in its more recent decision in Henderson the Supreme Court emphasised that Patel v Mirza did not represent “year zero” and that earlier decisions addressing particular types of situation where the illegality defence was invoked remained of precedential value unless they were shown to be incompatible with the approach set out in it: seepara. 77 of the judgment of Lord Hamblen (with which the other members of the Court agreed). Thus the appropriate lens through which to examine the issue in this appeal is not primarily the general principles enunciated in Patel but rather the comparatively limited number of cases, in this jurisdiction and other common law countries, which are specifically concerned with the case where the relevant illegal act has been done by a claimant who was mentally ill. I consider those cases at paras. 40-82 below.

Having said that, there is one aspect of the general law of illegality which, as will appear, is relevant to the analysis. There are a number of authorities which hold that the ex turpi causa principle has no application where the claimant was unaware of the unlawfulness in question. Two older cases in particular – Adamson v Jarvis (1827) 4 Bing 66, 130 ER 693, and Burrowsv Rhodes [1899] 1 QB 816 – are regularly cited, but we were also referred to the judgment of Lord Sumption in Les Laboratoires Servierv Apotex Inc [2014] UKSC 55, [2015] 1 AC 430 and to some observations by Lord Toulson in Patel v Mirza. I should briefly review those cases.

In Adamson v Jarvisan auctioneer had in good faith sold goods on behalf of a dishonest vendor to whom they did not in fact belong and was obliged to pay their value to the true owner. He brought proceedings against the vendor for reimbursement. One of the issues was whether his claim was barred by the fact that he had participated in the dishonest sale. The Court of Common Pleas held that it was not. Best CJ said, at p. 73:

“… [F]rom reason, justice, and sound policy, the rule that wrongdoers cannot have redress or contribution against each other is confined to cases where the person seeking redress must be presumed to have known that he was doing an unlawful act.”

The “reason, justice and sound policy” behind that requirement is self-evident. The old-fashioned term “turpitude” connotes moral culpability; and it is not normally just that a person should be treated as having behaved culpably unless they were aware (or “must be presumed” to have been aware) that what they were doing was wrong.

In Burrowsv Rhodes the plaintiff was injured while participating in the Jameson Raid. He brought proceedings against its promoters. It was objected that he could not recover because the Raid was unlawful. The Queen’s Bench Divisional Court rejected that argument. Kennedy J said that no claim for damages could be founded on an act if it is “manifestly unlawful, or the doer of it knows it to be unlawful” but that that was not the case where it “was not at the time apparently unlawful, and was done in honest ignorance of the particular circumstances which constituted its unlawfulness”: see at pp. 828-829.

I should add that there are many other decisions, particularly in the context of claims under an insurance policy or the like, in which the ex turpi causa rule has been formulated by the court in terms that refer explicitly to knowing wrongdoing. Examples are Hardyv Motor Insurers’ Bureau [1964] 2 QB 745 (p. 760), where Lord Denning MR refers to the conduct necessary to attract the defence as “wilful and culpable”, and Gray v Barr [1971] 2 QB 554, where he uses the same phrase (p. 568H).

In Les Laboratoires Servier the issue was whether the ex turpi causa defence could apply where the claim was based on a civil wrong. Lord Sumption, with whom the majority agreed, held that it could not. Although that is a long way from the issue with which we are concerned in this case, we were referred to two passages in Lord Sumption’s judgment.

The first is at para. 25, where he says (I omit some parts in the interests of brevity):

“The ex turpi causa principle is concerned with claims founded on acts which are contrary to the public law of the state and engage the public interest. The paradigm case is, as I have said, a criminal act. In addition, it is concerned with a limited category of acts which, while not necessarily criminal, can conveniently be described as ‘quasi-criminal’ because they engage the public interest in the same way. … [T]his additional category of non-criminal acts giving rise to the defence includes cases of dishonesty or corruption …; some anomalous categories of misconduct, such as prostitution, which without itself being criminal are contrary to public policy and involve criminal liability on the part of secondary parties; and the infringement of statutory rules enacted for the protection of the public interest and attracting civil sanctions of a penal character … .”

The Appellants sought to apply the terminology of “engagement of the public interest” and “quasi-criminality” to cases like the present where a criminal act is done albeit that the actor had no criminal responsibility. But Lord Sumption was not concerned in this passage with the issue of culpability. He was deploying those terms in order to make clear that the availability of the defence was limited to acts which were crimes, or which engaged the public interest in the same way as a crime, and did not therefore extend to civil wrongs. In fact all his examples of “quasi-criminality” involve conscious wrongdoing.

Lord Sumption does, however, address the question of unwitting wrongdoing in para. 29 of his judgment, where he considers “exceptional cases where even criminal and quasi-criminal acts will not constitute turpitude for the purposes of the illegality defence”. One such class of case is where the acts in question are trivial and would not normally be regarded as morally culpable. More pertinently for our purposes, he observes that where the relevant unlawful act is an offence of strict liability “the fact that the claimant was not aware of the facts making his conduct unlawful may provide a reason for holding that it is not turpitude at all” because it lacked the necessary “moral culpability”. That is in line with the older cases referred to above, and in fact he cites Burrows v Rhodes.

As for Patel v Mirza, the Appellants point out that in para. 93 of his judgment Lord Toulson set out a list of factors suggested by Professor Burrows as potentially relevant if a policy-based approach were adopted, and that these included, as (b), “whether the party seeking enforcement knew of, or intended, the conduct”. At para. 107 Lord Toulson described that list as “helpful” but continued:

“I would not attempt to lay down a prescriptive or definitive list because of the infinite possible variety of cases. Potentially relevant factors include the seriousness of the conduct, its centrality to the contract, whether it was intentional and whether there was marked disparity in the parties’ respective culpability [emphasis supplied].”

The Appellants argue that that shows that knowledge or intentionality of wrongdoing on the part of the claimant is not to be regarded as a prerequisite for the operation of the illegality defence but simply as a factor to be taken into account. That is correct as far as it goes, but Lord Toulson himself emphasised that he was speaking at a level of extreme generality. What he says is not inconsistent with the usual position being in accordance with the cases identified above, and the reference to intentionality may well be directed principally to cases of the kind referred to by Lord Sumption in Les Laboratoires Servier. In any event, it is important to bear in mind the point made at para. 31 above: we are not starting from a clean slate, and, as will appear, the question of the need for knowledge of wrongdoing, and thus for culpability, are expressly considered in the context of the mental illness cases that I consider below.

As noted above, the issue whether a person who has done an unlawful act – and, more specifically, who has committed homicide – while mentally ill and has suffered loss as a result can recover damages against others who have some causative responsibility for the act has been directly considered in the English case-law in Clunis, Gray and Henderson. But we were also referred to some U.S. and Commonwealth authorities. I will consider the cases in chronological order (subject to a couple of exceptions).

In Beresford v Royal Insurance Co Ltd [1937] 2 KB 197 the estate of a man who committed suicide – which was then a criminal offence – was held to be precluded from recovering under a policy of insurance. The jury at the trial of the action found that the deceased was sane at the time that he killed himself. In the course of giving the judgment of the Court of Appeal Lord Wright MR said, at pp 210-211:

“The question … is whether the felonious suicide of the assured is a bar to the present action. If the assured had taken his life while insane, the fact would not have constituted a defence. The act of an insane person is not in law his act – Felstead v. the King.”

The same point was not made expressly in the House of Lords ([1938] AC 536), but at p. 594 Lord Atkin set out the jury’s finding that the deceased was sane and went on to frame one of the issues in terms of “intentional suicide by a man of sound mind, which I will call sane suicide”.

In Boruschewitz v Kirts 554 NE 2d 1112 (Ill 1990) the plaintiff, who had been receiving treatment for mental illness, killed two people. She was charged with murder and pleaded “guilty but mentally ill”. She was “sentenced to incarceration”. She brought a claim in negligence against the doctor and the health centre responsible for her treatment. It seems clear that the losses for which she was claiming were the consequence of the killings. The claim was initially dismissed, but the decision was overturned on appeal by the Appellate Court of Illinois. The defendants argued that it was a principle of public policy that “a person should not be allowed to maintain an action if, in order to establish his cause of action, he must rely on his own violation of the criminal or penal laws or on his own immoral acts”: that is in substance the ex turpi causa principle.

The plaintiff’s answer was that the principle in question had no application in her case because she was insane at the time that she committed the killings. That was inconsistent with her plea of guilty but mentally ill, but the Court held that her plea was not conclusive. It continued (p. 1114):

“Plaintiff has alleged in her complaint that she was insane, and we must accept this allegation as true. An insane person is not held to be responsible for his acts. ... Plaintiff is allowed an opportunity to rebut the prima facie case and prove that she was criminally insane. In other words, she should be allowed to demonstrate that she did not commit an intentional act and thus was not guilty of a crime. ... Defendants claim that even if [the plaintiff] is not criminally responsible for her actions, she has still committed an immoral or wrongful act. We reject this argument. Society cannot hold people who are insane to the same moral standards as people who are sane. Additionally, the term ‘wrongful’ must also take on a different meaning in the context of an allegation of insanity.”

The Appellate Court referred to two decisions in other U.S. state jurisdictions in which claims on apparently similar facts had been summarily dismissed – Cole v Taylor 301 NW 2d 766 (Iowa) and Glazier v Lee 171 Mich App 216, 429 NW 2d 857 (Michigan); but it distinguished them on the basis that in both cases it appeared that the plaintiff had pleaded guilty, in the one case to manslaughter and in the other to murder, and had not raised a defence of insanity.

In Lingle v Berrien County 206 Mich App 528 (1994) the plaintiff had killed a man while insane. He was prosecuted and found “not guilty by reason of insanity”. He sued the county mental health centre for damages for negligent treatment. It appears that the losses claimed were all consequences of the killing. The claim was summarily dismissed, and that decision was upheld by the Michigan Court of Appeals. The relevant part of the report reads only:

“Although plaintiff Larry J. Lingle was found not guilty by reason of insanity in the shooting death of Robert Tollaksen the trial court did not err in granting defendants motion for summary disposition. A plaintiff cannot benefit from a cause of action founded upon an immoral or illegal act.”

The Court cited its own earlier decision in Glazier v Lee. It is not clear whether the plaintiff advanced the distinction which was applied in Boruschewitz.

In Traylor, to which I refer below, Johnson J refers to the decision of the Court of Appeals of Indiana in Rimert v Mortell (1997) NE 2d 867. That decision was not cited to us, but the Court apparently said, at pp. 874–875:

“A prohibition against imposing liability for one’s own criminal acts to another through a civil action is simply not justified when a plaintiff is not responsible for the act or acts in question. Thus, if in this case [the claimant] had been found not guilty by reason of insanity, he would bear no criminal responsibility for his acts and his subsequent civil action for recovery … could not be barred by the public policy expressed above.”

Finally, it is convenient to mention here, outside the strict chronological order, the only other U.S. authority to which we were referred. In Bruscato v O’Brien, 307 Ga App 452 (705 SE 2d 275) (2010), the plaintiff, who was mentally ill and had a history of violent behaviour, had killed his mother. He brought proceedings in Georgia against his psychiatrist alleging that he had done so as a result of negligent treatment. At first instance the claim was summarily dismissed on the basis that public policy would not permit him to benefit from his own wrongdoing – i.e. the illegality defence, albeit somewhat differently formulated. The Supreme Court of Georgia upheld a decision of the Court of Appeals overturning that decision. At the time of the decision the plaintiff, who was detained in a mental hospital, had not been tried in connection with the killing. The Supreme Court said:

“In this case, a question of fact remains whether Bruscato knowingly committed a wrongful act. There is no question that Bruscato killed his mother. That much has been admitted. There is considerable question, however, regarding Bruscato’s sanity and competency at the time the wrongful act was committed. At present, although Bruscato has been indicted, he has not been tried for the murder of his mother, and no jury has found him guilty of the crime. Therefore, as the Court of Appeals found: ‘[B]ecause no court has entered a judgment finding Bruscato legally responsible for his mother’s murder and because the issue of his mental competence at the time of the crime has been disputed, a jury issue exists as to whether Bruscato had the requisite mental capacity to commit murder.’ As a result, at this moment in time, it cannot be said that, should Bruscato’s claim against O’Brien be successful, he might profit from knowingly [original emphasis] committing a wrongful act. Concomitantly, then, O’Brien’s motion for summary judgment based on such an argument cannot succeed.The foreign cases relied upon by O’Brien, including Cole v. Taylor, 301 NW 2d 766 (Iowa 1981), do not alter this outcome, as those cases involved defendants who had already been convicted of the crimes forming the basis of their psychiatric malpractice claims.”

In a footnote the Court described Lingle v Berrien County as “only a short per curiam decision which contains no reasoning and lacks any persuasive authority”.

In Clunis the plaintiff killed a stranger, a Mr Zito, while suffering from a schizoaffective disorder. He was charged with murder but the prosecution accepted a plea of guilty to manslaughter on the grounds of diminished responsibility. He had been receiving treatment for mental illness from the defendant authority. He brought proceedings against it alleging that it had negligently failed to assess his condition in the period immediately before the killing, and that if it had done so he would have been detained and would not have committed the offence. The authority applied to have the claim struck out on two bases, one of which was that it was “based substantially, if not entirely, upon his own illegal act which amounted to the crime of manslaughter: ex turpi causa non oritur actio” (p. 985 F-G).

The application was dismissed at first instance, but the Court of Appeal (Beldam and Potter LJJ and Bracewell J) allowed the authority’s appeal. As regards the argument based on ex turpi causa, Beldam LJ, giving the judgment of the Court, pointed out that the plaintiff’s claim fell within the terms of the rule to the extent that it arose out of his commission of a criminal offence. But he referred toAdamson v Jarvis and Burrowsv Rhodes as authority for the proposition that its operation was restricted to “cases in which the person seeking redress must be presumed to have known that he was doing an unlawful act” (see p. 987C). Applying that approach to the facts of the case, he said, at p. 989 E-F:

“In the present case the plaintiff has been convicted of a serious criminal offence. In such a case public policy would in our judgment preclude the court from entertaining the plaintiff's claim unless it could be said that he did not know the nature and quality of his act or that what he was doing was wrong. The offence of murder was reduced to one of manslaughter by reason of the plaintiff’s mental disorder but his mental state did not justify a verdict of not guilty by reason of insanity. Consequently, though his responsibility for killing Mr. Zito is diminished, he must be taken to have known what he was doing and that it was wrong. A plea of diminished responsibility accepts that the accused’s mental responsibility is substantially impaired but it does not remove liability for his criminal act. [Emphases supplied]”

He goes on to consider the decisions of Woolf J in Meahv McCreamer [1985] 1 All ER 367 and Meah v McCreamer(no. 2) [1986] 1 All ER 943, and to refer to Gray v Barr (above). He concludes the judgment, at p. 990 D-E:

“In the present case we consider the defendant has made out its plea that the plaintiff’s claim is essentially based on his illegal act of manslaughter; he must be taken to have known what he was doing and that it was wrong, notwithstanding that the degree of his culpability was reduced by reason of mental disorder. The court ought not to allow itself to be made an instrument to enforce obligations alleged to arise out of the plaintiff's own criminal act and we would therefore allow the appeal on this ground.”

Prima facie it should follow from the reasoning of the Court of Appeal in Clunis, and specifically from the words which I have emphasised in the first of the two passages quoted above, that if the plaintiff had been found not guilty by reason of insanity the illegality defence would have failed, because the necessary ingredient of knowledge of wrongdoing would have been missing. (Footnote: ⁵) However, that cannot be regarded as part of the ratio: see para. 85 below.

Hunter Area Health Service v Presland [2005] NSWCA 33, (2005) 63 NSWLR 22, is a decision of the Court of Appeal of New South Wales (Spigelman CJ, Sheller and Santow JJA). The plaintiff (the respondent to the appeal) had killed his brother’s fiancée, a Ms Laws, while in an acute psychotic state. He was tried for murder but acquitted on the basis that his illness “so affected his capacity to reason that he did not know that what he was doing was wrong” (see per Sheller JA at para. 251); and he was committed to a psychiatric hospital as a “forensic patient”. That verdict is evidently substantially the same as a verdict of not guilty by reason of insanity in English law and governed by the same principles. Only a few hours before he killed Ms Laws the plaintiff had been discharged from a psychiatric hospital to which he had been taken by the police following an episode of bizarre and violent behaviour. He brought proceedings for negligence against the body responsible for the hospital and the consultant who had discharged him, a Dr Nazarian, on the basis that he was evidently seriously psychotic and that he should have been detained.

The plaintiff’s claim was upheld at first instance, and he was awarded damages in the sum of $369,300, comprising loss of income together with general damages of $225,000. The latter figure included compensation for the impact on the plaintiff of his detention on remand and in psychiatric hospital: the trial judge took into account the fact that if Dr Nazarian had acted properly he would have been detained for some time in any event, albeit that the period would have been shorter and the conditions less restrictive.

The Court of Appeal (Spigelman CJ dissenting) allowed the defendants’ appeal. All three justices treated the essential issue as being, as Spigelman CJ put it at para. 4 of his judgment, “the scope of the duty of care, particularly whether it extends to encompass the effects of unlawful conduct”, rather than whether what would otherwise have been a breach of duty was defeated by a supervening defence of illegality. As I understand it, this reflects the fact that binding Australian authority meant that the ex turpi causa principle as traditionally understood had no application. However, all three treated the underlying question as being one of policy, so that essentially the same considerations were in play as if the Court had been concerned directly with the illegality defence. (Footnote: ⁶)

All three judgments are very lengthy. I need to summarise each of them, but I will do so as shortly as possible.

The essence of Spigelman CJ’s reasons for his dissenting decision that the claim came within the scope of the defendants’ duty is that the plaintiff’s insanity meant that he was not morally culpable for his action. At paras. 42-77 of his judgment he carries out an exhaustive review of English, Australian and U.S. authority on the defence of illegality. He gives particular attention to Boruschewitz and Clunis,because of their closeness on the facts to the case before the court, but he also cites a number of other authorities to the effect that the defence only applies where the claimant knows that he or she is acting unlawfully: the English authorities cited are not only those cited in Clunis – Adamson v Jarvis, Burrows v Rhodes and Gray v Barr– but also Beresford and Hardy. At para. 78 he says:

“The significance of moral culpability in determining the weight to be given to unlawful conduct is clearly established on the authorities. Where, as here, a person has been held not to be criminally responsible for his or her actions on the grounds of insanity, the common law should not deny that person the right to a remedy as a plaintiff. In such a context the unlawfulness of the conduct is not entitled to weight in a multifactorial analysis.”

His conclusion on this part of the case, at para. 95, reads:

“Finally, I observe, how a society treats it citizens who suffer from mental illness, particularly the criminally insane, is often a test of its fairness. It is never easy to be fair where an innocent person has suffered as Ms Laws, and those who grieve her loss, clearly have. The law must, however, insist on protecting the rights of people, even if they are unpopular. Mr Presland was the instrument by which Ms Laws died. However, by reason of his insanity, his acts were not such that his right to receive proper medical treatment should effectively be taken away without compensation.”

Sheller JA’s reasoning on this part of the appeal appears at paras. 285-300 of his judgment. I have not found the sequence of reasoning in that passage entirely easy to follow, but the key stages appear to be as follows.

At para. 292 he observes:

“In general terms, two considerations may stand in the way of the plaintiff’s success in the present case. Although he was acquitted on grounds of mental illness, his act was and remained an unlawful act. His was not justifiable homicide but an unlawful homicide for which he was not criminally responsible. ... Adams J recognised that the plaintiff’s acts were deliberate acts of killing. His acquittal on the grounds of mental illness proceeded in the language of Dixon J in [The King v Porter (1933) 55 CLR 182] on the supposition that the plaintiff knew he was killing, knew how he was killing and knew why he was killing, but that he was quite incapable of appreciating the wrongness of the act. Although the plaintiff was acquitted and hence held not criminally responsible for the murder, what followed for which he now seeks compensation was the statutory response to the reason for his acquittal …. He was accordingly detained in strict custody in a psychiatric hospital. He claims damages for the consequence of that detention.”

As I understand it, the two “considerations” are (1) that the killing of Ms Laws was an unlawful act deliberately done by the plaintiff, albeit without knowledge of its wrongness, and (2) that his detention was the consequence of that act required by the law. At para. 294 he quotes an observation of Gleeson CJ to the effect that “where there is a problem as to the existence and measure of legal responsibility, it is useful to begin by identifying the nature of the harm suffered by a plaintiff”. He identifies the harm in question in para. 295 – in short, that the lawful detention which was the consequence of his killing of Ms Laws – and observes that “public policy must loom large in a court’s consideration of whether the plaintiff be compensated for the harm so suffered”.

The main public policy consideration which he goes on to consider is the risk that, if a claim lay in such a case, those responsible for making decisions about the care and treatment of mentally ill patients might be encouraged “to return to paternalistic practices, such as involuntary commitment, to protect themselves against possible medical malpractice liability” – in short, defensive medicine (see paras. 296-297). But he was also concerned by a different question. At para. 299 he refers to “the general rule … that one man is under no duty of controlling another man to prevent his doing damage to a third”, but notes that there are exceptions where there is some special relationship, such as that of parent and child. He continues:

“299.

… If responsibility is limited to a particular period of time, in this case six hours, or to harm done to persons with some relationship to the attacker, in this case the fiancée of the plaintiff’s brother, or otherwise, where is the line to be drawn either in the case of a claim by the attacker for the consequences of his attack (the present case) or a claim by the victim or the victim’s representatives?

300.

If, in the present case, instead of killing Ms Laws the plaintiff had come upon Dr Nazarian that night and killed or injured him, Dr Nazarian’s estate or Dr Nazarian would by parity of reason, have been liable to compensate the plaintiff for the consequences of his detention as a result of the unlawful killing of or assault upon Dr Nazarian. In this case, identification of the nature of the harm suffered by the plaintiff points as a matter of commonsense against the existence of a legal responsibility in the defendants for that harm. In my opinion, the verdict and judgment in favour of the plaintiff must be set aside.”

I understand the last two sentences of para. 300 to relate to the entirety of the reasoning from para. 285 onwards and not only to the point being made at the beginning of the paragraph and in para. 299. The essential point being made in that particular passage is that difficult questions would arise if the victim of the claimant’s unlawful act were the very person who was responsible for not detaining him. The implication, though it is not spelt out, is that it would be anomalous if he could recover damages in such a case; and that casts doubt on the existence of a duty more generally.

Turning to Santow JA’s reasoning, I have likewise not found it entirely easy to follow its detailed structure. He summarises his overall conclusion at the start of his judgment, from paras. 309-319. At para. 312 he says:

“… [T]o paraphrase what was said by McHugh J in Cole v South Tweed Heads Rugby (2004) 78 ALJR 933 at [46] ‘some minds may instinctively recoil’ at the idea that a hospital authority and psychiatrist, however careless, may be liable for the loss of liberty lawfully suffered by a forensic patient, following his killing of another while insane, itself an unlawful act, but without criminal consequence. Such an instinctive recoil is no substitute for the objective application of tort principle, as McHugh J there points out. But that reaction may nonetheless be a reflection of more considered community values, not to be stigmatised as based merely on prejudice or emotion.”

He goes on to acknowledge the “fundamental moral principle” that a person is not to be blamed for what he has done if he cannot help doing it, but he continues, at para. 313:

“But it does not follow that such a person, as distinct from his victim, should be compensated for the lawful consequences to him that followed the hospital authorities’ initial failure to detain him for treatment. This is for two possible reasons.  The first is grounded in legal policy and the second relates to what would have happened if the supposed duty had been performed.”

As to the first of those reasons, at para. 315 he says:

“While here the respondent was, by reason of insanity, judged incapable of acting with the necessary intent, his act of homicide was an unlawful act, hardly to be described as constituting reasonable action. Without in any way relying on the ex turpi causa maxim, I ultimately conclude that it would be unjust for the common law to allow the respondent a remedy for the non-physical injuries he has suffered in these circumstances. I here differ respectfully from Spigelman CJ’s conclusion to the contrary … . I do not base my conclusion on any moral culpability on the part of the respondent. Rather I base it on what I conceive legal policy, ultimately based on community values, would consider just in such a case.”

As to the second, at paras. 316-319 he says that there is real uncertainty as to whether, if the plaintiff had been detained when he should have been, he would not still have been released in due course and gone on to kill someone else, and he says that that uncertainty was itself a reason why the scope of the duty of care should not extend to the consequences of the killing which he in fact committed.

I need not follow through the detailed development of that reasoning in the following paragraphs. I will only quote a passage from para. 383, where Santow JA is dealing with what he describes (see para. 380) as “the normative aspects of causation”. He says:

“While I agree that the unlawfulness of an act of homicide committed while insane should not be an automatic bar to recovery, I respectfully disagree that it has no weight when it comes to determining what consequences of such actions should give rise to civil liability in negligence. Legal policy treats insanity as an ‘excuse’, though not justification, for what remains an unlawful act … . (Footnote: ⁷) But considerations of coherence as well as difficulties of causation lead it in my view to draw the line at permitting recovery by the person who committed that act for the non-physical consequences of his later detention in a mental hospital even though but for the hospital’s failure to detain, Ms Laws would not have been killed. … ”

His overall conclusion at para. 388 reads:

“I am of the view that it would be unjust to render the appellants as defendants legally responsible for a non-physical injury suffered by the respondent from deprivation of his liberty, when traced back to his unlawful but not criminal conduct. This is because that homicidal conduct is excused but not justified by the law on the ground of the respondent plaintiff’s insanity.  That conduct nonetheless constituted wholly unreasonable action on the respondent’s part, lacking moral culpability only by reason of his insanity. Such a normative conclusion is reinforced by the Act’s (Footnote: ⁸) emphasis on serious physical harm when none eventuated to the respondent. The law in consequence should be reluctant to visit civil liability, more especially in such a novel area and for non-physical consequences. Civil liability attaching to a failure to restrain risks promoting a bias towards detention, when the statutory scheme calls for an impartial exercise of discretion compulsorily to detain, taken only when fully justified, if not as a last resort. Therefore to introduce civil liability, which logically must also apply to decisions to restrain, is likely to induce a detrimentally defensive frame of mind on the part of the decision-maker in either context, so undermining coherence of the statutory scheme.”

As I read it, the principal thread in Santow JA’s reasoning is that it is repugnant to “community values” that the plaintiff should recover damages for the lawful consequences of his own unlawful act in killing Ms Laws, notwithstanding that the act was done while he was insane. But he also refers to “considerations of coherence as well as difficulties of causation”. The difficulties of causation appear to be those identified at paras. 316-319, summarised above. The reference to coherence appears to be to the risk that allowing a claim such as that advanced by the plaintiff would act as a discouragement to release and thus be contrary to the objectives of the Act – again, defensive medicine.

I should note, finally, that the Court unanimously agreed that the trial judge’s assessment of general damages should be reduced to $100,000. I am not sure that Spigelman CJ’s reasons for the reduction were the same as Sheller JA’s (with which Santow JA agreed); but I need not pursue the point.

In the New Zealand case of Ellis v Counties Manukau District Health Board [2006] NZHC 826, [2007] 1 NZLR 196, the plaintiff killed his father in the course of a psychotic episode. He was subsequently acquitted by reason of insanity but was detained in a mental hospital. For some time before the killing he had been exhibiting psychotic symptoms and had been assessed by doctors employed by the defendant Board. He brought proceedings for negligence and breach of statutory duty on the basis that if he had been properly assessed he would have been detained and the killing would never have occurred. The defendant applied to strike the action out on the basis that there was no duty of care to prevent damage of the kind claimed for. Potter J struck the action out on that basis. As in Presland, however, his reasoning involved him considering the ex turpi causa rule. He agreed that the rule had no application, stating:

“Mr Ellis was held not guilty by reason of insanity of killing his father. He has been acquitted of criminal responsibility for that act. This is not a factor which should be taken into account.”

In both Gray and Henderson the claimants pleaded guilty to manslaughter on the basis of diminished responsibility and were ordered to be detained in a hospital under sections 37 and 41 of the 1983 Act. In Gray the claimant killed a man in a road rage incident while suffering from PTSD as a result of injuries sustained in a train crash, and the defendants were the companies whose negligence had caused the crash. In Henderson the claimant, who had a history of paranoid schizophrenia, killed her mother during a psychotic episode, and the defendant was the NHS Trust responsible for the claimant’s psychiatric care: it was her case that if she had been properly treated she would not have committed the killing. It has not been suggested that that difference in the nature of the causative link in the two cases is relevant in this context.

The House of Lords in Gray held that the claimant’s claims were barred as a matter of public policy and upheld the decision in Clunis. On the face of it that decision applied equally to the claim in Henderson, and the claimant’s case was dismissed both at first instance and in this Court. But in the Supreme Court the claimant argued that Gray could be distinguished, alternatively that it was inconsistent with Patel v Mirza and that the Court should depart from it. Both arguments were rejected and the dismissal of the claim was accordingly upheld.

The principal opinions in Gray were delivered by Lord Hoffmann and Lord Rodger, with whom the other members of the Committee agreed. I take them in turn.

Lord Hoffmann’s starting-point is that the rule which excluded the claimant’s claim had both a wider and a narrower form, identified at para. 32 of his opinion as follows:

“The wider and simpler version is that … you cannot recover for damage which is the consequence of your own criminal act. In its narrower form, it is that you cannot recover for damage which is the consequence of a sentence imposed upon you for a criminal act. I make this distinction between the wider and narrower version of the rule because there is a particular justification for the narrower rule which does not necessarily apply to the wider version.”

As regards the narrower rule, Lord Hoffmann identifies the rationale as being based on “inconsistency”. He refers to Clunis and to the decision of this Court in Worrall v British Railways Board (unrep, 29.4.99). In that case the Court struck out a claim for loss of earnings by a plaintiff who had been imprisoned for serious offences which he blamed on a change of personality following an accident at work for which his employer was liable. Mummery LJ said:

“It would be inconsistent with his criminal conviction to attribute to the negligent defendant in this action any legal responsibility for the financial consequences of crimes which he has been found guilty of having deliberately committed.”

He goes on to quote a statement from Law Commission paper no. 160 (2001) The Illegality Defence in Tort  that “it would be quite inconsistent to imprison or detain someone on the grounds that he was responsible for a serious offence and then to compensate him for the detention”. He refers also to the decisions of the New South Wales Court of Appeal in State Rail Authority of New South Wales v Wiegold (1991) 25 NSWLR 500 and of the Supreme Court of Canada in British Columbia v Zastowny [2008] 1 SCR 27, which involved claims of essentially the same character. At para. 39 he quotes from the judgment of Rothstein J in Zastowny as follows:

“Zastowny’s wage loss while incarcerated is occasioned by the illegal acts for which he was convicted and sentenced to serve time. In my view, therefore, the ex turpi doctrine bars Zastowny from recovering damages for time spent in prison because such an award would introduce an inconsistency in the fabric of law. This is because such an award would be, as McLachlin J. described in Hall v. Hebert [1993] 2 SCR 159, 178, ‘giving with one hand what it takes away with the other’.”

As regards the wider version of the rule, Lord Hoffmann says, at para. 51, that it differs from the narrower version in at least two respects:

“[F]irst, it cannot, as it seems to me, be justified on the grounds of inconsistency in the same way as the narrower rule. Instead, the wider rule has to be justified on the ground that it is offensive to public notions of the fair distribution of resources that a claimant should be compensated (usually out of public funds) for the consequences of his own criminal conduct. Secondly, the wider rule may raise problems of causation which cannot arise in connection with the narrower rule. The sentence of the court is plainly a consequence of the criminality for which the claimant was responsible. But other forms of damage may give rise to questions about whether they can properly be said to have been caused by his criminal conduct.”

The reference to “public notions of the fair distribution of resources” has been glossed in Henderson: see para. 58 (3) of Lord Hamblen’s judgment quoted at para. 72 below.

Lord Rodger’s approach was differently structured but essentially to the same effect. After reviewing the reasoning in Clunis, Wiegold and Zastowny, he says, at para. 69:

“This line of authority, with which I respectfully agree, shows that a civil court will not award damages to compensate a claimant for an injury or disadvantage which the criminal courts of the same jurisdiction have imposed on him by way of punishment for a criminal act for which he was responsible. That principle can indeed be analysed in terms of the ex turpi causa rule since the plaintiff cannot even begin to mount his claim without founding on his own criminal activity.”

I should note that he quotes rather more fully from Zastowny than Lord Hoffmann, including a statement that “[p]reserving the integrity of the justice system by preventing inconsistency in the law is a matter of judicial policy that underlies the ex turpi doctrine”: that formulation too derives from the influential judgment of McLachlin J in Hall v Hebert[1993] 2 SCR 159.

At para. 58 of his judgment in Henderson Lord Hamblen, with whom the other members of the Court agreed, made the following observations about Gray:

“(1)

Both the narrow claim and the wide claim failed on the grounds of public policy.

(2)

All judges considered that the relevant policy in connection with the narrow claim was the need to avoid inconsistency so as to maintain the integrity of the legal system: ‘the consistency principle’.

(3)

Lord Hoffmann did not consider that this applied to the wide claim but held that a related policy did, namely that ‘it is offensive to public notions of the fair distribution of resources that a claimant should be compensated (usually out of public funds) for the consequences of his own criminal conduct’ (para 51). I understand this to mean that allowing a claimant to be compensated for the consequences of his own criminal conduct risks bringing the law into disrepute and diminishing respect for it. It is an outcome of which public opinion would be likely to disapprove and would thereby undermine public confidence in the law: ‘the public confidence principle’.

(4)

The public confidence principle is also applicable to the narrow claim. It is related to the consistency principle since one of the reasons that the public would be likely to disapprove of the outcome is the inconsistency which it involves between the criminal law and the civil law.

(5)

Although Lord Rodger appeared to consider that the consistency principle did not apply to the wide claim, the policy reasons he gives for rejecting the claim reflect that principle. The reason that a person cannot ‘attribute … to others’ acts for which he has been found criminally responsible, or ‘seek rebate’ of the consequences of those acts, is that it would be inconsistent with that finding of criminal responsibility. If a person has been found criminally responsible for certain acts it would be inconsistent for the civil courts to absolve that person of such responsibility and to attribute responsibility for those same acts to someone else.

(6)

Whilst the consistency principle more obviously applies to the narrow claim, on analysis it applies to the wide claim as well. In relation to the narrow claim the inconsistency is with both the criminal court’s finding of responsibility and the sentence it has imposed. In relation to the wide claim it is with the former only.”

Since in neither case was there any question of the claimant being insane, or therefore of their not knowing that what they were doing was wrong, it was unnecessary for either the House of Lords in Gray or the Supreme Court in Henderson directly to consider the application of the illegality defence in such a case. However, Lord Hoffmann did refer to it at para. 42 of his opinion in Gray, where he said:

“It should be noticed that in Hunter Area Health Service v Presland[2005] NSWCA 33; (2005) 63 NSWLR 22 the New South Wales Court of Appeal (again by a majority: Sheller and Santow JJA, Spigelman CJ dissenting) went even further and applied the rule when the plaintiff, who had been negligently discharged from a psychiatric hospital, was acquitted of murdering a woman six hours later on the ground of mental illness but ordered to be detained in strict custody as a mental patient. There are dicta (for example, in the passage I have quoted from Clunis’s case [i.e. the passage from p. 989 quoted at para. 49 above]) which suggest that the rule does not apply when the plaintiff, by reason of insanity, is not responsible for his actions. But the majority regarded compensation even in such a case as contrary to public policy. Sheller JA made the pertinent observation (at para 300) that if the rule did not apply and the plaintiff had killed the negligent psychiatrist who discharged him, the latter’s estate would have been liable to pay the plaintiff compensation for his consequent detention. This case, which Sheller JA (at para 294) described as ‘unusual if not unique’ raises an interesting question about the limits of the rule which it is not necessary to decide for the purposes of this appeal.”

It is clear, therefore, that Lord Hoffmann did not regard the application of the illegality defence in such a case as having been decided by necessary implication in Clunis. Although Presland was cited in Henderson Lord Hamblen makes no reference to it.

The structure of Lord Hamblen’s judgment in Henderson is that, having analysed the reasoning in Gray and Patel v Mirza, he goes on to address three issues: (1) whether Gray could be distinguished (paras. 79-86); (2) whether Gray should be departed from (and Clunis over-ruled) because it was inconsistent with Patel v Mirza (paras. 87-145); and (3) whether all the heads of loss claimed were irrecoverable (paras. 146-149). We are only concerned with issues (1) and (2).

As to issue (1), the claimant’s case that Gray should be distinguished was based on the proposition that on the particular facts of their respective cases Mr Gray had “significant personal responsibility” for the killing whereas she had none: he had been suffering from PTSD but unlike her was not psychotic. Lord Hamblen rejects that contention, on the basis that, even if the premise were correct, the crucial consideration for the majority in Gray “was the fact that the claimant had been found to be criminally responsible, not the degree of personal responsibility which that reflected” (para. 83).

As to issue (2), Lord Hamblen addresses the claimant’s case of incompatibility with Patel v Mirza under three headings, which I take in turn.

Heading (i) is “whether the reasoning in Gray cannot stand with the approach to illegality adopted by the Supreme Court in Patel” (paras. 89-96). Lord Hamblen holds that the reasoning in Gray is consistent with the approach adopted by the majority in Patel v Mirza. That point is irrelevant for our purposes, and I need not summarise his reasoning.

Heading (ii) is “whether it should be held that Gray does not apply where the claimant has no significant personal responsibility for the criminal act and/or there is no penal element in the sentence imposed” (paras. 97-112). Lord Hamblen answers that question in the negative. His essential point is that the claimant had been convicted of a criminal offence which necessarily involved both “blame” and “responsibility” (see in particular paras. 109 and 112). In that context, he repeats and evidently endorses the reasoning in Clunis and Gray, saying (at para. 105):

“As explained above, the key consideration as far as the majority in Gray were concerned was that the claimant had been found to be criminally responsible for his acts. That he had been convicted of manslaughter on the grounds of diminished responsibility meant that responsibility for his criminal acts was diminished, but it was not removed. It was not an insanity case and so, as Beldam LJ pointed out in Clunis (at p 989): ‘he must be taken to have known what he was doing and that it was wrong’.”

As I have already noted at para. 50 above, the apparent implication is that if Henderson had been an insanity case, so that the claimant had no knowledge of what he was doing and no criminal responsibility, the illegality defence would not have applied; but, as I also say, an implication of that kind cannot constitute binding authority.

Heading (iii) is “whether the application of the trio of considerations approach set out in Patel leads to a different outcome”. Lord Hamblen addresses those considerations in turn.

The first consideration is “the underlying purpose of the prohibition which has been transgressed and whether that purpose will be enhanced by denial of the claim”. As to that, Lord Hamblen says:

“125.

As explained above, this stage involves identification of policy reasons which support denial of the claim. Considering first general policy considerations rather than the purpose of the prohibition, for the reasons explained in Gray, the consistency principle is engaged in this case. There is a need to avoid inconsistency so as to maintain the integrity of the legal system. Whilst that most obviously applies to the narrower rule, it also applies to the wider rule. As Patel makes clear, this is a central and very weighty public policy consideration.

126.

For the reasons given by Lord Hoffmann in Gray, the public confidence principle is also engaged. Again, this applies to both the narrower and the wider rule.

127.

In the present case, the gravity of the wrongdoing heightens the significance of the public confidence considerations, as does the issue of proper allocation of resources. NHS funding is an issue of significant public interest and importance and, if recovery is permitted, funds will be taken from the NHS budget to compensate the appellant for the consequences of her criminal conviction for unlawful killing.

128.

This is also a case in which there is a very close connection between the claim and the illegality, thereby highlighting and emphasising the inconsistencies in the law which would be raised were the claim to succeed. The appellant’s crime was the immediate and, on any view, an effective cause of all heads of loss claimed. Indeed, applying Lord Hoffmann’s approach to causation in Gray, with which Lord Rodger and Lord Scott agreed, it was the sole effective cause of such loss.

129.

In relation to the underlying purpose of the prohibition transgressed, an important purpose is to deter unlawful killing thereby providing protection to the public. As far as the public is concerned there could be no more important right to be protected than the right to life. It is clearly in the public interest that everything possible is done to enhance protection of that fundamental right. There is also a public interest in the public condemnation of unlawful killing and the punishment of those who behave in that way.

130.

On behalf of the appellant it is submitted that it is absurd to suppose that a person suffering from diminished responsibility will be deterred from killing by the prospect of not being able to recover compensation for any loss suffered as a result of committing the offence. Indeed, more generally it is submitted that a person who is not deterred by a criminal sanction is unlikely to be deterred by being deprived of a right to compensation.

131.

There is force in these points, but the question should not be considered solely at the granular level of diminished responsibility manslaughter cases. Looking at the matter more broadly there may well be some deterrent effect in a clear rule that unlawful killing never pays and any such effect is important given the fundamental importance of the right to life. To have such a rule also supports the public interest in public condemnation and due punishment.”

The second consideration is “any other relevant public policy on which the denial of the claim may have an impact”. The claimant had suggested four public policies which weighed in favour of recovery: I need not set them out. Lord Hamblen examines these at paras. 133-136. At para. 137 he acknowledges that at least some of them have some force but he says they do not begin to outweigh the considerations supporting denial of the claim.

The third consideration is “whether denial of the claim would be a proportionate response to the illegality, bearing in mind that punishment is for the criminal courts”. Paras. 138-143 read:

“138.

It is not suggested that there were factors relevant to proportionality aside from the four factors identified by Lord Toulson at para 107 of his judgment in Patel, namely: (i) the seriousness of the conduct; (ii) the centrality of the conduct to the transaction; (iii) whether the conduct was intentional; and (iv) whether there was a marked disparity in the parties’ respective wrongdoing.

139.

As to the seriousness of the conduct, this was a very serious offence. It involved culpable homicide committed with murderous intent. As was acknowledged on behalf of the appellant, unlawful killing is the most serious conduct imaginable. The appellant knew what she was doing and that it was legally and morally wrong.

140.

As to the centrality of the conduct to the transaction, the offending is central to all heads of loss claimed and, as held in Gray, is the effective cause of such loss.

141.

As to whether the conduct was intentional, there was intent to kill or to do grievous bodily harm. Whilst there may have been no significant personal responsibility, there was nevertheless murderous intent.

142.

As to whether there was a marked disparity in the parties’ respective wrongdoing, the appellant was convicted of culpable homicide. Whilst she may not bear a significant degree of responsibility for what she did, she knew what she was doing and that it was morally and legally wrong. The respondent has admitted negligence in the appellant’s treatment. It is not the case, however, that the respondent’s staff did nothing in response to the appellant’s mental health relapse.

143.

In all the circumstances I do not consider that denial of the claim would be disproportionate. It would be a proportionate response to the illegality, bearing in mind that punishment is for the criminal court. The same would apply to the materially similar facts of Gray, even more clearly in so far as the offending in that case involved significant personal responsibility. The fact that proportionality was not specifically addressed in Gray does not therefore undermine the approach taken or the decision reached in that case.”

In Traylor the claimant suffered a severe psychotic episode during which he injured his daughter and was himself shot by police officers. He was prosecuted for attempted murder but found not guilty by reason of insanity. He brought proceedings for negligence against the NHS Trust from whom he had been receiving treatment prior to the injury. Johnson J dismissed the claim on the basis that although the Trust had been negligent its breaches of duty had not caused the losses in respect of which he claimed. However he went on to hold, albeit obiter, that if the claim had otherwise succeeded the Trust would not have been able to rely on the illegality defence.

The relevant part of his judgment is at paras. 105-119. His reasoning can be sufficiently summarised as follows:

It will be apparent from that survey that, as I have said, there is no binding authority on the question whether the illegality defence applies where the unlawful act in question was done when the claimant was insane – in Beldam LJ’s phrase, an “insanity case”. As far as English appellate authority is concerned, it is fair to say that the implication of the reasoning in Clunis, and in Henderson insofar as it adopts that reasoning, might be thought to be that the defence should not apply in an insanity case: see paras. 50 and 78 above. But it does not follow that those decisions are authoritative since they were not themselves concerned with an insanity case. We were reminded of Lord Halsbury’s statement in Quinn v Leathem [1901] AC 496, at p. 506, that

“… a case is only an authority for what it actually decides. I entirely deny that it can be quoted for a proposition that may seem to follow logically from it.”

It is also fair to observe that in Gray Lord Hoffmann clearly did not regard Clunis, as endorsed by him, as necessarily meaning that the defence would not apply in an insanity case: see para. 73 above. In Beresford Lord Wright’s statement that a person who committed suicide when insane would not have been debarred from claiming is explicit; but, again, that case concerned a claim under a life insurance policy, and what he said cannot be treated as necessarily applying to a claim of the kind with which we are concerned.

As regards other jurisdictions, the cases go both ways. The fullest discussion is in Presland, which holds that public policy would preclude recovery in an insanity case. But its persuasive authority is reduced by the fact that it is only a majority decision and by the rather different reasoning adopted by the two members of the majority. The preponderance of U.S. authority favours recovery, but the persuasive weight of Boruschewitz, Rimert and Bruscato is also limited since the issue is one of policy, and public policy in this country would not necessarily be the same as in the United States. (Footnote: ¹⁰)

Having said that, I believe that the general tendency of the authorities can fairly be said to be against the illegality defence applying in an insanity case. There is a consistent focus on the simple proposition that it is wrong – or, in Best CJ’s wordsin Adamson v Jarvis, contrary to “reason, justice, and sound policy” – to treat an otherwise good claim as barred because it depends on an unlawful act done by a claimant who had no knowledge that he or she was acting unlawfully and therefore is not morally culpable; and that proposition is applied in several cases where the relevant lack of knowledge was the result of mental illness.

THE JUDGMENT OF GARNHAM J

Without intending any disrespect to Garnham J’s careful judgment, I need not, since the issue which we have to decide is one of pure law, summarise his reasoning in any detail. After a thorough review of the authorities which I have considered above, at para. 127 he summarises the principles which he believed emerged from them in ten propositions. At paras. 128-142 he applies those principles to the facts of the case. His essential reasoning is that the illegality defence only applies where the claimant knew that he was acting unlawfully, and that the special verdict conclusively established that that was not the case (see paras. 129 and 131). The distinction between the present case and Henderson was “not arbitrary but fundamental, turning as it does on the presence or absence of criminal responsibility” (para. 134). I have already quoted his conclusion on this aspect at para. 135 (see para. 7 above). He addresses various other particular arguments advanced by the Defendants. I need not quote the passages in question.

DISCUSSION AND CONCLUSION

Each of the three Appellants pleads a number of grounds of appeal, although there is considerable overlap between them. I do not believe that the most helpful course is to go through each in turn. Instead I will address directly the issue of law raised by this appeal, namely whether the Claimant’s claims are barred by the illegality defence, while making sure that in doing so I cover the various pleaded grounds.

In considering this issue we must proceed on the basis that unless the illegality defence applies the Claimant has a good claim for negligence against the Appellants: if he did not, the issue would not arise. That means that we must proceed on the basis that it was evident during at least his second period of detention that he was seriously mentally ill and liable to commit violence against others (having already assaulted one man and behaved violently in the police station); that the medical professionals negligently failed to take the necessary steps to ensure that he was detained and thus prevented from attacking anyone else; that the killings which he went on to commit immediately following his release were a direct and foreseeable consequence of that negligence; and that he has in consequence suffered loss of the various kinds claimed. (Footnote: ¹¹) In short, he has suffered a serious injury (in the technical sense) from the failures of others who owed him a duty to protect him from harm of the kind that he in fact suffered. If we are using the language of public policy, it is clearly the policy of the law – reflected in the law of negligence – that people in that position should be compensated for the loss which they have suffered. The question is whether that public policy is outweighed by the other considerations of public policy relied on by the Appellants.

The public policies on which the Appellants rely are described in various ways in their skeleton arguments and oral submissions, but it is most convenient to start with the two broad headings considered in Gray and described by Lord Hamblen in Henderson as “the consistency principle” and “the public confidence principle” (see para. 58 (2) and (3) of his judgment, quoted at para. 72 above).

The inconsistency considered in Gray and Henderson is between, on the one hand, treating the claimant’s conduct as criminal and, on the other, allowing them to claim damages for the consequences of that conduct. But before us the Appellants also argued that allowing the Claimant to recover compensation would be inconsistent with the approach taken by the civil law. I take the two kinds of inconsistency in turn.

The Claimant’s case is straightforward. The verdict of not guilty by reason of insanity was an acquittal. Accordingly the law has not treated him as criminally responsible for his actions, and there is no inconsistency in allowing him to recover for the loss that he has suffered in consequence of them.

The Appellants’ primary answer to that case is that although the Claimant might not be criminally liable he had nevertheless committed a criminal act, in the sense that he had committed the actus reus of murder, and with the necessary mens rea (see para. 18 above); and that that was enough to engage the public interest and constitute the necessary “turpitude”. This is the same point that Sheller JA makes in Presland, where he emphasises that the plaintiff’s killing of Ms Laws remained an unlawful act notwithstanding his insanity: see para. 57 above.

I would accept the Claimant’s case on this issue. It is in my view sufficiently clear from Lord Hoffmann’s opinion in Gray that the inconsistency with which he was concerned was based on the claimant’s criminal liability, or responsibility, for the act in question and not simply on the fact of his having done what could be characterised as a criminal act. That is the implication both of a number of the passages which he quoted with evident approval as set out at para. 69 above – from the Law Commission (“responsible for a serious offence”) and from the judgments of Mummery LJ in Worrall (“inconsistent with his criminal conviction”) and Rothstein J in Zastowny (“the illegal acts for which he was convicted”). The same goes for Lord Rodger: see his conclusion quoted at para. 71 that a claimant could not recover compensation for an injury or disadvantage imposed “by way ofpunishment for a criminal act for which he is responsible”. It would, I suppose, be possible to account for that language on the basis that the claimants had in fact been found criminally liable, so that it was unnecessary to distinguish between the criminal act and the claimant’s responsibility for it; but I believe that the focus on the latter properly reflects the policy in play. It is only if it has imposed a criminal sanction on the claimant that, in McLachlin J’s words in Hall v Hebert, the state would be “giving with one hand what it takes away with the other” – that is, to spell it out, compensating the claimant for the consequences of the penalty which it has itself imposed.

That approach also seems to me to accord with the fundamental justice of the matter. At a superficial level you could still say that it was inconsistent to allow a person to recover for the consequences of an unlawful act which they have done. But at a more fundamental level the criminal law is concerned not with acts as such but with personal responsibility for those acts, and a difference in treatment based on differences in personal responsibility cannot be said – again, to quote McLachlin J – to undermine “the integrity of the justice system”. This reflects the basic perception reflected in the authorities which I refer to at paras. 33-38 above based on the requirement of moral culpability.

The Appellants make the point that the orders under sections 37 and 41 of the 1983 Act under which the Claimant was detained are the same as those often made in the case of defendants who are convicted of manslaughter by reason of diminished responsibility and in fact made in both Mr Gray’s and Ms Henderson’s cases. On that basis it is said to be inconsistent for the illegality defence to be available in the one case but not the other. I do not accept that argument. Quite apart from the fact that the making of those orders is mandatory in the case of a verdict of not guilty by reason of insanity but a matter of discretion in a diminished responsibility case, the fact that identical orders may be made does not mean that the nature of the claimant’s responsibility is the same: it only means that in both cases there is a need both for treatment and rehabilitation and for the protection of the public.

The Appellants also emphasised the point made by the majority in Presland that it was inconsistent for the claimant to claim damages for detention pursuant to an order lawfully made by the Court. I do not see this as a distinct consideration. If the detention has no punitive element – which in the case of a verdict of not guilty by reason of insanity it does not – its lawfulness is an irrelevance. If, say, a patient had been negligently prescribed a drug which caused a psychotic state and had to be sectioned it would be no answer to a claim against the prescriber for damages for the loss of liberty and distress caused by their detention to say that the detention was lawful.

I should emphasise that I am dealing here only with the inconsistency principle. The Appellants’ argument that the Claimant should not be entitled to recover compensation for the consequences of his criminal act (albeit one for which he had no criminal responsibility) can still be deployed in the context of the public confidence principle, and I consider it in that context below.

As identified at paras. 23-27 above, insanity is no defence to an action in tort. If the estates, or the dependants, of the Claimant’s victims chose to sue him for damages for their deaths he would be liable notwithstanding that he acted while insane: the law would treat him as responsible for his acts as long as, in Stable J’s words in Morriss v Marsden, his mind directed his hand. The Appellants contend that it would be incoherent if the law took a different approach to his responsibility for his acts in the context of a claim brought by him for damages against a third party.

Although I have for convenience included this point under the heading of the consistency principle, it is a rather different kind of inconsistency from the inconsistency with the criminal law identified in Gray – “incoherence” might be a better label. But I need not trouble with that taxonomic question because I do not believe that there is any incoherence about the different approaches. The question of the liability of the Claimant to his victims for the injury which he caused them is self-evidently different from the question of the liability of the Appellants for the loss which they have caused him. In the former case justice requires that the interest of the victim in receiving compensation comes before any question of moral culpability: see para. 25 above. In the latter it is the Claimant who is the victim of wrongdoing and the question whether he should nevertheless be denied recovery because his loss was the result of a criminal act has to be considered in that quite different context. Again, I am not saying that it has to be answered in his favour, only that to allow recovery would not be inconsistent with the rule that his insanity does not preclude his liability to his victims.

In Henderson Lord Hamblen defined “the public confidence principle” as being that “allowing a claimant to be compensated for the consequences of his own criminal conduct risks bringing the law into disrepute and diminishing respect for it” because that is “an outcome of which public opinion would be likely to disapprove” (see para. 58 (3) quoted at para. 72 above).

In my view it is this principle which is at the heart of this appeal, as it was for Santow JA in Presland, and I have not found it easy to decide whether it should operate in this case. I do not doubt that it would – at least as a first reaction – stick in the throats of many people that someone who has unlawfully killed three innocent strangers should receive compensation for the loss of liberty which is a consequence of those killings, however insane he was and however negligent his treatment had been. To the extent that that reaction reflects, in Santow JA’s language, “considered community values”, we should be very slow to disregard it: the law ought so far as possible to give effect to such values.

However, I have come to the conclusion that, although that first reaction is entirely understandable, the values of our society are not reflected by debarring a claimant from seeking compensation in this kind of case. It is necessary, as Santow JA accepted, to go beyond “instinctive recoil” and to consider what justice truly requires in a situation which most humane and fair-minded people would recognise as far from straightforward. Taking that approach, although of course those who are killed or injured must always be treated as the primary victims, it is fair to recognise that the killer also may be a victim if they were suffering from serious mental illness and were let down by those responsible for their care. I rather suspect that some such view underlies the observations of the jury at the Claimant’s trial which I quote at para. 11 above. But, whether it does or not, I believe that the considered view of right-thinking people would be that someone who was indeed insane should not be debarred from compensation for the consequences of their doing an unlawful act which they did not know was wrong and for which they therefore had no moral culpability. As we have seen, the law does not generally apply the illegality defence where the claimant does not know that what they are doing is wrong and has no moral culpability; and in my view that reflects ordinary and comprehensible principles of fairness. I do not believe that it is rational, or would accord with community values, that the position should be different where the claimant’s lack of knowledge or culpability was the result of insanity. In short, I would align myself with the approach taken by Spigelman CJ at para. 95 of his judgment in Presland: see para. 55 above.

Having reached that conclusion as a matter of principle, I think it is reinforced by the consideration that, as I believe, it reflects the tendency of the authorities and the implications of the reasoning in Clunis, and thus also of Gray and Henderson in which that reasoning was approved.

In support of their contention that to allow claimants to recover in circumstances such as those of the present case would bring the law into disrepute, the Appellants relied on two apparent anomalies about the rights of the victims of the relevant unlawful acts which would arise if the illegality defence did not operate. One is general, the other more specific. I take them in turn.

The general anomaly relied on is that whereas claimants would be entitled to claim compensation from their doctors for what they had lost as a result of not being prevented from committing their unlawful acts, the victims of those acts (or their estates or dependants) would have no claim against the doctors. Mr Warnock said that that was the inevitable result of the principle recently restated by the Supreme Court in Robinson v Chief Constable of West Yorkshire Police [2018] UKSC 4, [2018] AC 736, that “the common law imposes no liability to protect persons against harm caused by third parties, in the absence of a recognised exception such as a voluntary assumption of responsibility”: see para. 69 (1) of the judgment of Lord Reed. Ms Ayling differed from Mr Warnock inasmuch as she accepted that there might be particular circumstances in which the duty would be owed, because of Lord Reed’s reference to the existence of recognised exceptions to the rule; but she agreed with him to the extent that the victims would be unable to recover in the generality of cases.

I am prepared to assume that at least in the generality of cases victims in a situation such as the present would have no right to recover against the authorities whose negligence had allowed the attack to take place. But I do not accept that that gives rise to an anomaly. Victims may not have a right to compensation against the doctors, but they have a straightforward claim against their assailant, whose insanity would be no defence to a civil claim for assault. (Footnote: ¹²) It is true that, unlike a doctor or health authority, the assailant may not be in a position to meet a substantial award of damages. However, as we have seen, one of the heads of damage claimed by the Claimant in this case is an indemnity against any liability to his victims. I can see no reason why that would not be an admissible head of claim; and, if it is, it would afford a route by which victims could be assured of payment of any damages that they were awarded (Footnote: ¹³). However, Ms Ayling did not accept that a claim for such an indemnity would lie, though she did not advance any developed reason for that position. In the absence of full argument I am not prepared definitively to decide the point. But even if the claimant were not entitled to such an indemnity, the fact that they might not be able to meet any award of damages to the victim does not seem to me to be a principled reason for denying them recovery for their own loss.

The more specific anomaly arises only in a case where the victim of the claimant’s unlawful act is also the defendant, as in the example given by Sheller JA where a mentally ill patient attacks the negligent doctor: see para. 58 above. Can it be right that the claimant could in such a case sue their victim? Lord Hoffmann in Gray described this concern as “pertinent” (see para. 73 above). Such cases will no doubt only occur rarely, but they are not wholly implausible. In addition to Sheller JA’s example, there could be cases where the claimant’s mental illness was the result of the negligence of a family member whom they subsequently attack (they might, say, have been the passenger in a car which crashed due to the careless driving of a parent and have suffered a brain injury).

I fully accept that it is seems unjust that someone who has suffered unlawful injury at the hands of another can be required to pay damages to them for the consequences that they have suffered as a result of inflicting that injury. Of course the victim would have a cross-claim, but even if that exceeded the value of the claimant’s claim, so that there was no net liability, their net recovery would necessarily be less than the full compensation for their loss. The position would be worse still if the claimant, as in this case, claimed an indemnity against any such liability: that would on the face of it reduce the victim’s recovery to nil while still leaving them liable for the claimant’s loss. (It is true that they might be insured against their liability to the claimant – in my two examples, both the doctor and the driver would almost certainly be insured – but that ought not to affect the position in principle.)

I do not, however, believe that the problems that would arise in that scenario are a reason for barring a claim in the typical case where, as here, the defendant is not a victim of the claimant’s unlawful act. I ought not to seek to determine in advance how the Court would address such a situation; but since we are concerned with questions of public policy, it would have the tools to produce a just outcome.

The Appellants relied on some other public policy considerations referred to in the part of Lord Hamblen’s judgment in Henderson where he is considering the compatibility of Gray with Patel v Mirza. It is worth repeating that an important part of the message of Henderson is that a full Patel v Mirza analysis is not necessary where the courts have already determined how the relevant public policies play out in a particular kind of case. In my view Gray has identified the key considerations in a case where the unlawful act was done by a claimant who was mentally ill, and that is why I have given priority to the public policy considerations considered above. But of course the present case is not identical to Gray, and I accordingly accept that the analysis cannot stop there.

The further public policy considerations in question are identified by Lord Hamblen at paras. 125-129 of his judgment, which I have set out in full at para. 80 above. They can be summarised as (a) the impact on NHS funding of allowing a claim of the present kind (para. 126); and (b) deterring unlawful killing and providing protection to the public, there being no more important right to protect than the right to life (para. 129). (I note in passing that the concern about defensive medicine which weighed with Sheller and Santow JJA in Presland was not relied on in Henderson or before us.)

As to the impact on NHS funding, since Lord Hamblen identified this as a relevant public policy consideration in the context of potential claims by persons convicted of manslaughter it must equally be a relevant consideration in the case of claims by persons who commit homicide while insane. There is no way of quantifying the amounts concerned. In the context of the NHS budget generally, and indeed the NHS budget for meeting negligence claims, they must be tiny, since cases of the present kind must be very rare; but the same would be true for cases of the kind under consideration in Henderson, and the point is evidently one of principle.

As to deterring unlawful killing, it is hard to see that the denial of compensation to claimants who commit homicide while insane is likely in practice to have any deterrent effect. In Henderson Lord Hamblen acknowledged the force of that point but said that it was wrong to consider the question solely at that granular level and that, viewed more broadly, there might be some deterrent effect in a clear rule that unlawful killing never pays and it would also support “the public interest in public condemnation and due punishment” (paras. 130-131). I am not sure that his final point applies in the context of persons who are acquitted of homicide by reason of insanity, because in their case it would not seem that condemnation or punishment are appropriate; but the point about a universal rule having a general deterrent effect would in principle apply to insanity as well as diminished responsibility cases.

Broadly, therefore, these public policy considerations would appear to be in play. But the question is whether it is proportionate to treat them as outweighing the public interest in claimants in insanity cases receiving due compensation for the wrong that they have suffered. I do not believe that it is. The balance is quite different from in the diminished responsibility cases because the claimant has no moral culpability. That point is clearly made if one looks at how Lord Hamblen struck the balance at paras. 138-143 in Henderson. In those paragraphs he emphasises the importance of the fact that the claimant knew that what she was doing was legally and morally wrong: see paras. 139 and 142. In the absence of that element, and where, essentially for that reason, the consistency and public confidence principles are, as I would hold, not engaged, I do not believe that either the impact on NHS resources or the general deterrent effect of a rule against recovery could justify the denial of the claim in these proceedings.

Finally, I should note a point made by Ms Ayling that there is no sharp distinction between a finding of diminished responsibility and a finding of insanity: the distinction is one of degree only. That may be so, but the criminal law proceeds on the basis that the distinction is nevertheless real and that in any given case it will be possible to say on which side of the M’Naghten line the defendant falls. That being the case, there is nothing irrational about the application of the illegality defence depending on the selfsame distinction. If I had any unease about this aspect, it would, rather, be about the possibility that in some cases the distinction may reflect not a finding by a court but a forensic choice by the defendant or their advisers. Pleas of not guilty by reason of insanity are in practice rare; and there must be cases where a defendant tenders, and the Crown accepts, a plea of manslaughter by reason of diminished responsibility where the facts might arguably have justified a special verdict (Henderson may be an example). But if that results in the illegality defence being unavailable in some cases where it might have been available if the defendant had made a different choice I do not think that can affect the decision in principle which we have to make.

For those reasons I would hold that the illegality defence advanced by the Appellants is unavailable as a matter of law, and I would accordingly dismiss the appeal.

Andrews LJ would allow the appeal. I see the force of the points which she makes, and, as I have already said, I do not regard the question as an easy one. But I hope it is clear from the foregoing reasoning why I have in the end come to the conclusion that I have, and which I respectfully maintain.

Since the President is of the same view as I am, the action will now proceed against the Appellants, subject to any further appeal, as regards the claim in negligence as well as the claim under the 1998 Act. There are pleaded issues not only about whether any of the Defendants, and if so which, were negligent but also about causation, contributory negligence and quantum. None of those are in any way affected by our decision.

I have had the advantage of reading in draft the judgment of Lord Justice Underhill. There is nothing I can usefully add to his masterly analysis of the relevant case law, both domestic and foreign. I agree with him that, to the extent that it is possible to discern it, the general tenor of the authorities in which the matter has been considered seems to be against allowing the illegality defence in a case where the claimant satisfies the M’Naghten test for insanity. However, in most of them the issue was not fully argued, as it has been before us, and in Presland (where it was fully argued) the decision went the other way, albeit with a strong dissenting judgment.

The denial of the defence was a result which I initially found to be more attractive than the result for which the Appellants contended, which would potentially leave a person who has been wronged without a remedy, or at least severely circumscribe the heads of loss for which he could claim. Yet on further reflection, and with the greatest respect, I find myself unable to agree with my Lord and my Lady that a lack of knowledge or understanding by a person who intentionally takes the life of another human being that what he was doing was wrong is a sound and principled basis for allowing that person to make a claim in negligence against someone for putting them in a position which enabled them to commit an act which was both deliberate and tortious.

I agree with Underhill LJ that in an era where there is much greater understanding of mental health issues, it is fair to recognise that, as well as the primary victims, the killer also may be a victim, if they were suffering from serious mental illness and were let down by those responsible for their care. However, I am not persuaded that an absence of the state of knowledge of wrongdoing, which would afford the mentally ill perpetrator of a deliberate fatal assault a complete defence to criminal liability for murder or manslaughter, justifies drawing a bright line between the present case and similarly tragic cases such as Clunis, Gray and Henderson.

There are all kinds of reasons why a defendant suffering from a serious mental illness who faces a charge of murder might prefer to opt for running the partial defence of diminished responsibility rather than pleading insanity, even though it may be open to them to do so. The most obvious of these is the prospect of indefinite incarceration in a secure mental health unit. Moreover, it is not difficult to conceive of examples of situations where a person who is guilty of the criminal offences of murder or manslaughter, or causing death by careless driving, might be regarded by the public as less blameworthy for the death than a person in the position of the Claimant, who intended to kill his victims. Yet such a person would be precluded by their conviction from making a claim of this nature even if they were seriously mentally unwell at the time.

For me, the starting point in the analysis is that the killings were unlawful acts, therefore any claim against the Appellant would involve pleading and relying upon acts which were illegal and to which civil liability attaches. As Santow JA observed in Presland at [383], legal policy treats insanity as an excuse for homicide, but it does not treat it as a justification. The Claimant would be liable in tort for battery, because the fatal assaults were deliberate rather than accidental, notwithstanding that because of the state of his mental health he did not know that what he was doing was wrong: see Morriss v Marsden and Dunnage v Randall. Indeed, were that not so, there would be nothing for which he could seek an indemnity from the Appellants. He therefore bears legal responsibility for his unlawful actions, notwithstanding that he is excused from criminal liability and may not be regarded as morally culpable for them.

That is equally true of tortfeasors who are sane. If someone does a deliberate act which amounts to an assault, and which injures another, even if they had no intention of harming that other person, let alone of causing them death or serious injury, civil liability will flow from that deliberate act. So if someone deliberately pushes someone sideways in order to get out of a confined space, because they are feeling claustrophobic, or having a panic attack, and that person falls over and hits their head, with fatal consequences, the person who did the pushing would be regarded by many as morally blameless, or at least as bearing no greater moral culpability for the death than someone who deliberately pushes someone under a train in the delusional belief that they are the Devil. But as Stable J recognised in Morriss v Marsden, the absence of moral blame is irrelevant in that context. The law of tort is concerned with compensating the victim rather than with punishing the wrongdoer.

Given that liability in tort attaches to the deliberate wrongful act, then even if the claim in negligence in the present case were held to be actionable, the Claimant might face formidable obstacles when it came to establishing causation. At one time I thought that this might provide a more satisfactory answer than reliance on the illegality principle. However, on reflection it seems to me that, as the Supreme Court recognised in Henderson, there are wider principles of public policy at play.

Of course, as Underhill LJ has rightly pointed out, the question of the liability of the Claimant to his victims’ estates for the injury which he caused them is different from the question of the liability of the Appellants for any loss which they have caused him. The question with which we are concerned is whether the Claimant should be denied recovery for loss that he has suffered by reason of the Appellants’ alleged negligence, in consequence of the deliberate commission by him of acts which are undoubtedly unlawful, but for which he bore no criminal responsibility because he did not know that they were wrong. However, unlike Underhill LJ, I do perceive a lack of coherence between on the one hand, making the Claimant liable in tort to pay compensation to his victims or their estates, and, on the other, permitting him to avoid the consequences of such liability, by passing responsibility for his actions to someone else, on the basis that he would not have committed those intentional and tortious acts had it not been for the Appellants’ negligence.

Taking the example I have given above of the person who deliberately pushes someone out of the way, causing them to lose their balance and hit their head with fatal consequences, suppose that person has been negligently prescribed medication with a known side-effect of causing panic attacks. Would they be able to make a claim against the GP or NHS trust in negligence seeking a full indemnity in respect of the compensation they had to pay the deceased’s estate? I suggest that the answer would be no, because they deliberately did something that the law recognises to be an actionable wrong, (irrespective of why they did it, or what their mental state was at the time) and that legal wrong would be an essential ingredient of their claim for compensation. I very much doubt whether the answer would depend on whether, if prosecuted, they would be found guilty of manslaughter or some other criminal offence, or on whether they intended to hurt the person they pushed.

If the correct analysis of where the policy lines are to be drawn depends on the deliberate nature of the tortious act, as I consider it does, then the Claimant’s behaviour falls as much on the wrong side of the line as the behaviour of the hypothetical panic-stricken pusher. Both might be regarded as morally blameless but neither can rely on their own deliberate tortious act as a necessary ingredient of a claim against a third party. Denial of the claim cannot be regarded as disproportionate in circumstances where such a claim could not be made by a sane individual who was at least as morally blameless. Of course, that would not prevent the Claimant from claiming in negligence for any loss that did not require him to rely on his own tortious acts.

There is no justification for deciding this issue in favour of the Claimant because the deceased’s estates are unlikely to have a direct claim against the Appellants, whose pockets (or those of their insurers) are likely to be deeper than those of the Claimant himself.

I also part company with my Lord in his interpretation of what Lord Hamblen was saying in Henderson, and in particular with his view that the fact that the defendant knew that what she was doing was morally and legally wrong was central to Lord Hamblen’s reasoning.

In that case, the Supreme Court was not concerned with unlawful acts that did not attract criminal responsibility; the fact that a serious crime had been committed was sufficient to dispose of the appeal, irrespective of the degree of personal responsibility of the offender, see in particular Lord Hamblen’s analysis at paragraph 112. The focus was very much on the nature of the act itself. Indeed Lord Hamblen refers to the fact that by her guilty plea the appellant in that case accepted that she possessed the mental prerequisites of criminal responsibility for murder, namely an intention to kill or to cause grievous bodily harm (as the Claimant did in the present case). He then says that in that case, her psychiatrists also agreed that she knew that what she was doing was wrong. That appears to me to be something he is treating as an additional factor rather than the essential factor behind the policy.

In paragraph 119, when discussing the consistency principle and the public confidence principle, Lord Hamblen said this:

“..whilst preventing someone from profiting from his own wrong is not the rationale of the illegality defence, it is a relevant policy consideration, which is linked to the need for consistency and coherence in the law. For one branch of the law to enable a person to profit from behaviour which another branch of the law treats as criminal or otherwise unlawful would tend to produce inconsistency and disharmony in the law and so cause damage to the integrity of the legal system”. [Emphasis added].

He went on to say in paragraph 120 that the closer the connection between the claim and the illegal act, the greater and more obvious may be the inconsistency and consequent risk of harm to the integrity of the legal system. In this case the deliberate unlawful act is central to the claim against the Appellants. There is a well-established distinction between criminal responsibility and civil/tortious responsibility in cases such as the present. However, a failure to apply the illegality defence in a case such as this would, in practical terms, enable the Claimant’s mental health to enable him to place the legal responsibility for deliberately taking the lives of three people at someone else’s door.

It seems to me that all the public policy considerations identified by Lord Hamblen in Henderson as supporting denial of the claim are equally present here. The unlawful acts were of the same nature and gravity as the offence in Henderson. Funds would be taken from the NHS budget to compensate the Claimant for the consequences of his deliberate conduct in killing three people, even though there was and could be no criminal conviction. The unlawful acts are the immediate and/or effective cause of the main heads of loss claimed. Whilst Lord Hamblen saw the force of the argument that it was absurd to suppose that a person suffering from diminished responsibility for a killing will be deterred from killing by the prospect of not being able to recover compensation for any loss suffered as a result of committing the offence, he thought that there “may well be some deterrent effect in a clear rule that unlawful killing never pays and any such effect is important given the fundamental importance of the right to life.” I respectfully agree.

I have not reached this conclusion lightly. However it does seem to me that there is nothing disproportionate about precluding someone who intended to kill, and did so, from bringing a claim in negligence in reliance on that deliberate and unlawful act, and that the policy rule preventing such claims from being made should not rest on nice distinctions between having little or no personal responsibility for the killing because of the state of the claimant’s mental health at the time. For those reasons, I would have allowed this appeal.

For the reasons given by Lord Justice Underhill with which I entirely agree, I would dismiss this appeal. In view of the comprehensive nature of his judgment, I add only a few words of my own.

The criminal law draws a clear and principled distinction between homicide cases where responsibility is diminished, and those where it is extinguished. Where such matters are in issue, a scrupulous exercise is undertaken to distinguish between them. This was the task of the jury at the trial of Alexander Lewis-Ranwell, the Claimant in this appeal.

The Claimant had killed three elderly men on 10 February 2019 by bludgeoning them to death with a hammer in their own homes. He had also injured two other individuals with a saw. The Claimant was 27 at the time of these events. He was charged with three counts of murder, and with two counts of wounding with intent contrary to section 18 of the Offences Against the Person Act 1861. The Claimant raised the defence of insanity. The ultimate issue for the jury at his trial was whether he was guilty of manslaughter on the grounds of diminished responsibility or not guilty by reason of insanity (a special verdict).

If on a charge of murder, an accused wishes to rely on the partial defence of diminished responsibility and so be convicted of manslaughter not murder, it is for him to establish on the balance of probabilities, that he was suffering from an abnormality of mental functioning, which arose from a recognised medical condition,and which substantially impaired his ability to (1) understand the nature of his conduct; (2) to form a rational judgment or (3) to exercise self-control (section 2(1) and (1A) and (2) of the Homicide Act 1957, as amended). An abnormality of mental functioning will provide an explanation for the accused's conduct only if it causes, or is a significant contributory factor in causing, the accused to carry out that conduct (section 2(1B) of the 1957 Act).

As Lord Hughes pointed out in R v Golds [2016] UKSC 61; [2016] 1 WLR 5231, at [48] it is an important part of the Crown’s function where the charge is murder and a case of diminished responsibility is advanced, to assess the expert evidence and its relationship to any dispute of fact and quite a large proportion of verdicts of manslaughter on the grounds of diminished responsibility arise because the Crown accepts as it is entitled to do, that the correct verdict is guilty of manslaughter. Such a plea can be accepted by the court without a jury having to return the verdict: R v Cox, [1968] 1 WLR 308 subject to the approval of the court, which will consider whether it is fair and in the interests of justice: Criminal Procedure Rules PD 9.3.

The insanity defence is set out in M’Naghten’s case. The M’Naghten rules were laid down by the House of Lords in 1843 in the response given by Tindal CJ, on behalf of all the other judges except for Maule J:

“Jurors ought to be told in all cases that every man is to be presumed to be sane, and to possess a sufficient degree of reason to be responsible for his crimes until the contrary be proved to their satisfaction and to establish a defence on the ground of insanity, it must be clearly proved that, at the time of the committing of the act, the party accused was labouring under such a defect of reason, from disease of the mind, as not to know the nature and quality of the act he was doing; or, if he did know it, that he did not know what he was doing was wrong.”

Where the issue is one of insanity, the prosecution is required to prove that the defendant committed the actus reus of the crime charged. (Footnote: ¹⁴) The burden of proving insanity lies on the defendant, on the balance of probabilities. The prosecution cannot accept a plea of insanity: R v Crown Court at Maidstone, ex p. London Borough of Harrow [2000] 1 Cr. App. R. 117, DC. Instead, a verdict must be reached by the court: by a jury in the Crown Court pronouncing the special verdict, or by an acquittal in the youth or magistrates' court. A jury cannot return a special verdict except on the written or oral evidence of two or more registered practitioners at least one of whom is duly approved by the Secretary of State as having special experience in the diagnosis or treatment of mental disorder (sections 1 and 2 of the Criminal Procedure (Insanity and Unfitness to Plead) Act 1991; and section 54(1) Mental Health Act 1983).

The Claimant’s trial took place in November 2019 before May J and a jury and lasted two weeks. There was no dispute that the Claimant had killed the three men, or injured the two other individuals, and the charges remained ones of murder. (Footnote: ¹⁵) The only issue as I have said was whether he was guilty of manslaughter on the grounds of diminished responsibility or was not guilty by reason of insanity. What determined that issue was the Claimant’s mental state at the time of the killing. The single question left to the jury by the trial judge was in these terms: “Are we satisfied that it is more likely than not that when he killed…Mr Lewis-Randall did not know that what he was doing was against the law?” The trial judge went on to direct the jury that if their answer to that question was yes, the jury would return a verdict of not guilty by reason of insanity. If their answer to the question was no, their verdict would be not guilty of murder, but guilty of manslaughter.

The three experienced psychiatrists who gave evidence for the defence at the trial, had each independently concluded after their examination of the Claimant, and with the benefit of their long experience of psychotic illness, that the Claimant was suffering from severe paranoid schizophrenia when he killed the three men and injured two others; and that the likelihood was that the Claimant would not have known that what he was doing was wrong. The prosecution invited the jury to reject their evidence on knowledge of legality. As the trial judge reminded the jury, the issue was one for them to decide. By their verdict, reached after more than six hours deliberation, the jury unanimously concluded that on the three counts of murder, the Claimant was not guilty by reason of insanity. The jury sent the trial judge a note before delivering their verdicts. This said: “We, the Jury, have been concerned at the state of psychiatric health service provision in our county of Devon. Can we be reassured that the failings in care for ALR will be appropriately addressed following this trial?”

Following the return of the special verdict, the trial judge did not pass any sentence. A person found not guilty by reason of insanity is not culpable because of their mental disorder; they have not been convicted of any crime and cannot be sentenced. They are, as Lord Bingham described them in R v H [2003] UKHL 1; [2003] 1 WLR 411 at p 413, “irresponsible in the eyes of the law”. For reasons of public protection however there are statutory powers which are triggered by a special verdict. These include the power to detain such a person in a secure hospital (a hospital order) with a restriction that he or she is not to be released until permission is given by the Secretary of State (a restriction order (Footnote: ¹⁶)): see sections 37 and 41 of the Mental Health Act 1983. The medical evidence before the court meant the Claimant could be made the subject of a hospital order; and the return of the special verdict in respect of the murder charges (a charge where the sentence on conviction is fixed by law) meant that this had to be coupled with a restriction order: see section 5(1)(a) and (2) of the Criminal Procedure (Insanity) Act 1964 as amended by the Domestic Violence, Crime and Victims Act 2004.

The Claimant was made the subject of a hospital order and a restriction order. He will remain in detention until the Secretary of State considers it is safe to release himor he is discharged by a Mental Health Tribunal: see section 41(3)(c)(iii) of the Mental Health Act 1983. In thanking the jury for their note, the trial judge said this: “It has been a disturbing case on so many levels, three dead, two badly injured whose cases lie on the file, at the hands of somebody whom you have found on strong psychiatric evidence was grossly, floridly psychotic at the time and therefore not criminally responsible for his actions on 10 February last year. The result, as you have heard, is the same. He will be cared for in hospital with a Restriction Order, which means he will not be allowed into the community, even in a step by step way, until the agencies are absolutely content that it is safe for him to be released.”

In February 2021, the Claimant brought these civil proceedings in negligence and under section 7 of the Human Rights Act 1998 (for breaches of articles 3 and/or 8 of the European Convention on Human Rights) against four parties: G45 Services (UK) Limited, the Chief Constable of Devon and Cornwall Police, Devon Partnership NHS Trust and Devon County Council. The claim under the Human Rights Act is proceeding, as it is conceded by the defendants that the doctrine of ex turpi causa non oritur actio (no claim arises from a dishonourable cause of action, or the common law defence of criminality as Sedley LJ describes it in Al Hassan-Daniel v Revenue and Customs Commissioners [2010] EWCA Civ 1443;[2011] QB 866 at para 6) does not operate as a defence to such a claim.

The facts pleaded by the Claimant in support of his claim and which we are required to accept as correct for the purposes of this appeal, relate to events which occurred in the days prior to and on the day of the killings.

The Claimant had a known history of mental illness and had been compulsorily detained in 2016 and 2017 under the Mental Health Act 1983. He was diagnosed with schizophrenia and psychosis and received treatment in a psychiatric intensive care unit. On 7 February 2019, the Claimant interfered with electrical and security equipment at a farm in Ilfracombe. He opened animal enclosures and left the farm with a bicycle and a drill. No arrests were made but the police were called following reports of his strange behaviour. On 8 February 2019, the Claimant stole a pony from its enclosure at a farm. He was arrested at 8.41 am that day on suspicion of burglary and detained at 10.04 at Barnstable Police Station. At 2.49 am on 9 February 2019 the Claimant was released on bail and taken to the Freedom Centre. On 9 February 2019, the Claimant let animals out of their enclosure at a small holding in Barnstaple. He went on to assault the owner, who was 84 years old with a long double handed saw. The Claimant released more animals from their enclosure on a neighbouring farm. The Claimant was arrested on suspicion of grievous bodily harm and re-detained at Barnstaple Police Station at 10.06. At 11.05 am, whilst at that police station, the Claimant tried to grab an officer’s taser gun. He was restrained and taken to his cell where his clothing was removed and he was given a self-harm suit. At 9.38 am on 10 February 2019 the Claimant was released on bail for a second time. The killings and assaults occurred within hours of his release. The Claimant was arrested on 11 February 2019, following an assault on a night manager of a hotel in Exeter and taken to Exeter police station. During the medical examination in his cell, he punched the examining nurse in the face. He was detained under the Mental Health Act 1983 later that afternoon.

The Particulars of Claim set out in detail what was done or not done by various police officers, medical and other health care professionals during this period. They describe the increasingly bizarre, delusional, psychotic and violent behaviour of the Claimant. They also describe his mother’s grave concerns, communicated to the police, about what might happen were he to be released on bail. These pleaded events are recorded in paragraphs 9 to 31 of the judgment of Garnham J, which is the subject of this appeal.

The Claimant alleges that all four of the defendants were negligent in their treatment of him during the period 8 to 10 February 2019 and acted in breach of his rights under articles 3 and 8 of the European Convention on Human Rights. He seeks damages for personal injury, loss of liberty, loss of reputation and loss of dignity; and an indemnity in respect of any claim brought against him as a consequence of his violence towards others in the period 9 to 11 February 2019.

In relation to negligence, the Claimant’s claim in summary is this. There was inadequate provision of mental health services when he was in police custody between 8 to 10 February 2019; as a consequence of systemic and individual failings, he was not assessed under the Mental Health Act; despite warnings that he posed a risk to the safety of others he was released from custody (twice) and within hours of his second release, whilst in a psychotic state and acting under delusions, he went on to kill the three men. It is alleged that had he been assessed, he would have been detained under the Mental Health Act and would not have gone on to kill the three men. On his pleaded case, the Claimant was the victim of a high order of negligent conduct on the part of the defendants.

The criminal law provides a defence for people who as a result of their mental condition should not be held responsible for what would otherwise be criminal conduct. The fundamental rationale for this is that it is unjust to hold people criminally responsible who could not have avoided committing the alleged crime, through no fault of their own. The defence of insanity is difficult to establish and as the Law Commission noted in 2013, it is little used. There is no reason to think things are different now. One of the psychiatrists at the Claimant’s criminal trial described the defence as difficult to establish and very rare.

The Claimant might have avoided a trial by entering a partial defence of manslaughter on the grounds of diminished responsibility to the charges of murder. But he did not do so. Instead, he proved on the balance of probabilities that he was not guilty of the charges of murder. Though he had committed the actus reus of the crime charged, his mental state meant he was not responsible in the eyes of the law, for what he had done. His acts were not quasi criminal. They were not criminal at all.

It is against this background that the application of the doctrine of ex turpi causa to this civil claim needs to be considered.

In Clunis, after a review of the authorities, Lord Justice Beldam, giving the judgment of the court, concluded that the illegality defence was restricted to cases in which the person seeking redress must be presumed to have known that what they were doing was unlawful. He drew a clear distinction between cases of manslaughter on the grounds of diminished responsibility and insanity cases (at p. 989):

“In the present case the plaintiff has been convicted of a serious criminal offence. In such a case public policy would in our judgment preclude the court from entertaining the plaintiff’s claim unless it could be said that he did not know the nature and quality of his act or that what he was doing was wrong. The offence of murder was reduced to one of manslaughter by reason of the plaintiff’s mental disorder but his mental state did not justify a verdict of not guilty by reason of insanity. Consequently, though his responsibility for killing Mr Zito is diminished, he must be taken to have known what he was doing and that it was wrong. A plea of diminished responsibility accepts that the accused’s mental responsibility is substantially impaired but it does not remove liability for his criminal act. …The court ought not to allow itself to be made an instrument to enforce obligations alleged to arise out of the plaintiff’s own criminal conduct.”

The subsequent case of Gray decided in 2009, was considered by the Supreme Court in Henderson in 2020. In Henderson, at [105] Lord Hamblen (with whom Lord Reed, Lord Hodge, Lady Black, Lord Lloyd Jones, Lady Arden and Lord Kitchin agreed) said:

“The key consideration as far as the majority in Gray were concerned was that the claimant had been found to be criminally responsible for his acts. That he had been convicted of manslaughter on the grounds of diminished responsibility meant that responsibility for his criminal conduct was diminished but it was not removed. It was not an insanity case and so, as Beldam LJ pointed out in Clunis at (at p 989): “he must be taken to have known what he was doing and that it was wrong”.”

He went on to say at [142]:

“As to whether there was a marked disparity in the parties’ respective wrongdoing the appellant [Ms Henderson] was convicted of culpable homicide. Whilst she may not bear a significant degree of responsibility for what she did, she knew what she was doing and that it was morally and legally wrong.”

Each of these cases draws a coherent and bright line distinction for the purposes of the ex turpi causa doctrine, between those who are criminally responsible for their acts whether fully or partially, and those who are not responsible for their acts because they do not know what they are doing is morally and legally wrong. In my judgment, this common thread running through the criminal and civil law, is consistent with principle, a proper understanding of the true implications of acute mental illness and is one that would not offend the sensibilities of ordinary right-thinking members of the public or undermine public confidence in the law.