Abstract
In 2013, the applicant, then a student at Coventry University, was convicted of murder. He now sought an extension of time (five years) and leave to appeal his conviction, relying on the Court of Appeal decision in Brennan [2014] EWCA Crim 2387 and the Supreme Court decision in Golds [2016] UKSC 61. These had, subsequent to his conviction, clarified the courts’ approach to cases involving mental health considerations outside the jury’s experience where the sole issue at trial was the partial defence of diminished responsibility and the psychiatric experts were substantially agreed as to the mental health of the killer at the time of the homicide.
On 16 January 2013, the applicant drove to Manchester where he attacked and fatally stabbed a young man. The attack was unprovoked and lacked an apparent motive. Once in custody, the applicant’s mental health was assessed by three different medical professionals. Having initially denied any symptoms of mental illness, in subsequent police interviews, he claimed to have been hearing voices taunting and abusing him. On his account, he had travelled to Manchester, armed with a knife, in order to seek out his abusers and to make them stop swearing and racially abusing him. The voices directed him to a particular place where he came across the victim whom he hit three times in the stomach, believing him to be one of those abusing him. He claimed that he had only wanted to scare the victim and did not think he had killed him.
At a pre-trial hearing, the judge directed the applicant’s treating doctor, a forensic psychiatrist, to provide a report. The defence also instructed a forensic psychiatrist and so did the prosecution. The prosecution’s expert was not called at trial but his opinion was elicited via the defence experts. All three were agreed that the applicant was suffering from paranoid schizophrenia at the time of the killing and that this had substantially impaired his responsibility. While they could not rule out the possibility that the applicant was feigning his symptoms, they believed that he was genuinely experiencing the voices, and the other behavioural evidence was consistent with this diagnosis.
At trial, the prosecution sought to disprove the experts’ diagnosis and its impact, relying inter alia on evidence demonstrating extensive planning, Closed Circuit Television (CCTV) footage showing the applicant driving past the victim on five separate occasions (which was said to disprove the random nature of the attack), lies told in police interviews, inconsistencies in the accounts given to various psychiatrists, the lack of previous reporting of mental health issues, the absence of relevant symptoms when first medically assessed while in custody and the applicant’s denials that he had heard voices or had intrusive thoughts until he knew the extent of the evidence against him. It was the prosecution case that the applicant was not suffering from paranoid schizophrenia, that he had feigned his symptoms and that the doctors had been duped. In the alternative, the prosecution argued that any mental illness present had not substantially impaired his responsibility for the killing.
The defence case was that the applicant was not guilty of murder but guilty of manslaughter by reason of diminished responsibility. He was suffering from paranoid schizophrenia which had substantially diminished his mental responsibility in that it had affected his ability to form a rational judgment and exercise self-control. It was a common misperception that individuals with this condition would not be able to lie, plan or ‘think straight’, while it was not uncommon for them to be too ashamed, confused, distressed or embarrassed to admit their condition and to delay seeking medical help for months or even years.
In his application for leave and an extension of time to appeal, the applicant (having changed his legal representation) advanced two grounds for appeal: First, he contended that there was no proper evidential basis upon which the jury could have rejected the unanimous expert psychiatric evidence, wherefore the trial judge should have withdrawn the case from the jury at the close of the evidence. Secondly, the judge’s directions, although fair and measured at the time, were inadequate in the light of the guidance advanced subsequently to the applicant’s conviction by the Supreme Court (UKSC) decision in Golds. This would have required a direction to the jury to the effect that they must not act as ‘amateur psychiatrists’ and that where there was undisputed expert evidence they would probably wish to accept it, unless there was some identified evidential reason for not doing so.
The single judge referred the case to the full court for consideration whether exceptional leave should be granted pursuant to the test for change of law cases set out in Jogee [2016] UKSC 8 and on the merits.
The court concluded that this was not one of those rare cases where the trial judge should have withdrawn the murder charge from the jury. The prosecution were entitled to pursue the murder charge and to challenge the medical evidence. Even if the applicant had not been faking his symptoms, the prosecution had been entitled to argue that the defence had not established, to the relevant standard, that all four questions in s 2 of the Homicide Act 1957 [as set out by Lord Hughes in Golds at [8] these are as follows: ‘(1) Did the accused suffer from an abnormality of mental functioning? (2) If so, did it arise from a recognised medical condition? (3) If yes to (1) and (2), did it substantially impair one or more of the abilities listed in section 1A? (4) If yes to (1), (2) and (3), did it cause or significantly contribute to the killing of the deceased?’] had been answered in favour of the applicant. The question remained: was the recognised medical condition sufficient, at the time of the killing, to provide an explanation for the killing and substantially impair his responsibility? The court had no doubt that had the judge had the benefit of Lord Hughes’ observations in Golds [that judges should direct the jury that if there is a proper basis for rejecting the expert evidence the decision is theirs and that it was wise to direct them not to turn themselves into amateur psychiatrists] he would have added those words, but the fact that he did not did not necessarily undermine the safety of the conviction. The summing up in its entirety ‘could not have been fairer to the defence’.
Commentary
Hussain is the latest in a series of cases that required the appellate courts to reflect on the role of the jury where the sole issue to be determined at trial was the partial defence of diminished responsibility, in circumstances where the psychiatric evidence as to the defendant’s mental health at the time of the homicide was unanimous or uncontradicted. The decision in the present case confirms that the assessment whether medical evidence supports a plea of diminished responsibility under s 2 of the Homicide Act 1957, as amended by s 52 of the Coroners and Justice Act 2009, is ultimately a matter for the jury, not psychiatrists. Under its amended terms, which have been said to have ‘medicalised’ the defence, juries remain free to reject expert opinion, albeit that the freedom to do so is qualified by the requirement, emphasised in the present case, that the jury’s attention ought to have been drawn to some rational evidential basis for challenging the agreed expert evidence. This is so even where the psychiatric evidence points unanimously one way, namely towards the defendant having suffered from a relevant condition at the time of the killing and where the experts are agreed that it must have impacted upon the accused’s abilities to understand the nature of his conduct, rationally form a judgment or exercise self-control. The issue is fact-sensitive and the Court of Appeal in the present case was ‘confident all parties would have been aware…that there had to be evidence and a rational basis to challenge the expert evidence’ and that ‘the case as presented to the jury and as left to them by the judge provided sufficient safeguards of the kind Lord Hughes had in mind in Golds’. The outcome in Hussain is thus unremarkable.
What is remarkable is the way in which the instant Court supports its decision. It cites with approval paragraphs from the (seven member strong) UKSC decision in Golds which considered, but only partially endorsed, the Court of Appeal’s observations in Brennan on the role of the jury and the circumstances in which a judge ought to withdraw a murder charge from the jury, in trials where the psychiatric evidence supporting diminished responsibility is uncontradicted or unchallenged.
It is worthwhile to recall the salient facts of that case and the Court’s considerations in some detail. The appellant in Brennan appealed his murder conviction (returned by a majority) on the basis that (although no application to this effect had been made by the defence at the close of the evidence) the judge should have withdrawn the murder charge from the jury, since the unchallenged psychiatric evidence had concluded that he met all the criteria for the diminished responsibility defence (as amended). Allowing the appeal and substituting a verdict of manslaughter, the Court held that juries may not depart from uncontradicted and unchallenged expert evidence, where there is simply no rational or proper basis for doing so (Brennan at [44]). If unchallenged expert evidence were to be rejected by a jury, then it needed to be rejected for reason (Brennan at [45]). Reviewing the relevant authorities, Davis LJ concluded that while juries were not bound uncritically to accept unchallenged expert evidence, they were not free, on the other hand, uncritically to reject evidence on a matter calling for scientific expertise (Brennan at [47]). He determined that trial judges have power to withdraw a charge of murder from the jury (Brennan at [64]). Drawing on the principles in Galbraith [1981] 1 WLR 1039 [that on a submission of no case to answer a judge ought to stop the case only when there is either no evidence or the evidence is so tenuous that taken at its highest a jury could not properly convict], Davis LJ went on to suggest that judges should withdraw a murder charge where the medical evidence supporting the defence was uncontradicted and there was no other evidence rebutting diminished responsibility. In cases where there was some other evidence at least capable of rebutting the defence, the judge should withdraw the murder charge from the jury if the other evidence was ‘too tenuous or, taken at its highest, insufficient’ to permit a rational rejection of the defence of diminished responsibility (Brennan at [65]).
Subsequently, in Golds, the UKSC considered the pronouncements made in Brennan. It agreed that juries were not bound by expert evidence (Golds at [49]). It further agreed that if juries were to be invited to reject expert opinion, some rational basis for doing so must at least be suggested. It was not open to the Crown simply to invite the jury to convict the defendant of murder without suggesting why the expert evidence ought not to be accepted (Golds at [49]). However, the UKSC went on to state that while the principles in Galbraith might be capable of being applied in a trial where the sole issue was diminished responsibility, a court ought to be cautious about doing so. In the UKSC’s view, the Galbraith process ‘was generally a conclusion that no jury, properly directed, could be satisfied that the Crown has proved the relevant offence so that it is sure’ (Golds at [50]). In the context of diminished responsibility, murder could only be withdrawn from the jury if the judge was satisfied that no jury could fail to find that the defendant had made out the defence. That the judge might entertain little doubt about what he thought the right verdict ought to be was not sufficient reason for withdrawing from the jury issues which were properly theirs to decide (Golds at [50]).
In Hussain, a differently constituted Court of Appeal has now concluded that the Court’s earlier observations in Brennan have survived Golds only to the extent that they emphasised that there ought to be some rational evidential basis for challenging agreed expert evidence and that the decision as to whether a defendant falls within the bounds of the diminished responsibility defence is for the jury, not the doctors, to decide. The clear implication of this is that the passages in Brennan considering the circumstances in which a judge ought to withdraw a murder charge from the jury (where uncontradicted psychiatric evidence supports the defence) did not survive Golds and, what’s more, should not now be followed: in asserting that ‘in future, reliance should not be placed on any judgment predating Golds on this issue’, Hussain treats Golds as authoritative on this issue. This is remarkable, for, as noted by Richard Percival in Archbold Review (2019) 5, 2–3, the relevant passages in the UKSC’s judgment were actually obiter: the issues in Golds were the correct interpretation of the phrase ‘substantially impaired’ in s 2 of the Homicide Act 1957 (as amended) and whether trial judges are required to direct juries on the meaning of that phrase in every case. The paragraphs in Golds analysing Brennan followed determination of these issues and were not necessary to answer the certified questions and resolve the appeal (but note that commentators had called upon the UKSC ‘to take the opportunity to approve Brennan’ when deciding Golds, see, for example, Jonathan Rogers ‘The amended diminished responsibility plea’ (2015) CLJ 201, 205).
It is true that the Court Martial Appeal Court in Blackman [2017] EWCA Crim 190 anticipated this development when it noted, and acknowledged, the advice in Golds for trial judges to exercise caution before accepting the partial defence and removing the case from the jury. But whereas the appellate court in Blackman sought to anchor this in ‘the broad approach of this court in Khan (Dawood) [2009] EWCA Crim 1569’ (Blackman at [43]), Hussain goes further by suggesting that some of the general observations put forward in Brennan are no longer authoritative post-Golds.
While it is, of course, helpful to know whether a particular approach has the support of the highest appellate court, by treating the UKSC’s obiter dicta in Golds as authoritative, Hussain exemplifies an inclination by the present Court of Appeal to ignore the rules of precedent or, at least, to ‘not apply the doctrine of stare decisis with the same rigidity [as applied in its civil jurisdiction]’ (Sir Brian Leveson P in DPP v Patterson [2017] EWHC 2820 at [16], citing Diplock LJ in Gould [1968] 2 QB 65, 68G) so as to adopt the UKSC’s position, even where this contradicts technically still binding Court of Appeal authority. In doing so, the Court facilitates tendencies in the current UKSC to seize opportunities to clarify and develop the common law, in passing and whenever it sees fit. Commentators will be divided as to whether this development is deeply concerning (from a procedural propriety, rule of law and separation of powers perspective) or to be welcomed (as a means of keeping the common law tidy and updated, when legal aid cuts might prevent cases that raise the issue directly from making it to the appellate courts).