Abstract
Legislation:
Criminal Procedure Ordinance (Cap.221) s.83E(1)
This is an unusual case with a tragic background. The appellant brutally killed his cohabitee and was convicted of murder three times following three separate jury trials. That exceptionally brought the appellant the second time before the apex court in Hong Kong, the Hong Kong Court of Final Appeal (“CFA”).
In 2009, the appellant killed his cohabitee by chopping her with a knife 213 times after discovering her alleged infidelity. He was first convicted of murder following a jury trial in 2010, but the Court of Appeal (“CA”) allowed his appeal on the basis of misuse of hearsay evidence and ordered a retrial in 2013. In 2014, the appellant was again convicted of murder in a jury trial. The CA, by a majority, upheld his conviction and the case went through to the CFA. The issue in the first and second trials was that of provocation. The appellant had admitted to the killing and offered to plead guilty to manslaughter every time, but the prosecution refused.
In 2017, the CFA quashed the conviction due to a misdirection given to jury at the trial and ordered a second retrial since “the issue of credibility of the facts on which the defence was based ought to have been left to the jury to determine” (HKSAR v Liang Yaoqiang (2017) 20 HKCFAR 1 at [127].).
In 2018, the appellant for the third time was convicted of murder. The issue of diminished responsibility was raised for the first time in this third trial. In 2020, his conviction was quashed by the CA for the reason of “a failure to direct the jury on the relevance of the psychiatric evidence in relation to the defence of provocation” (HKSAR v Liang Yaoqiang [2021] 1 HKLRD 26 at [145]). However, the majority of the CA (Macrae VP and Zervos JA with McWalters JA dissenting) ordered a third retrial upon its exercise of discretion. All of the judges gave their reasons as follows.
The brief reasons given by Macrae VP were that, first, the crime at issue was murder, one of the most serious crimes; second, the killing was undisputed; third, the fading of memories could be remedied; fourth, the defence of diminished responsibility was run for the first time; and fifth the misdirection to the jury was contributed to by the appellant's counsel (at [3]-[7]).
Zervos JA gave detailed reasons, which were summarized as eleven considerations directing his exercise of discretion (at [257]): (1) a fourth trial would not be oppressive to the appellant; (2) the lapse of time would not post material prejudice to the appellant; (3) there was a need for finality determined by a jury; (4) the case for murder was strong; (5) a retrial would be straightforward; (6) the issues were of the nature suitable to be determined by a jury; (7) the interests of the public required a determination of the most serious crime; (8) the sentence of murder would be life imprisonment; (9) a fair trial was possible; (10) the appellant and his counsel respectively caused the delay and misdirection to jury; and (11) the appellant's mental condition did not preclude a retrial.
On the other hand, McWalters JA concluded in his powerful dissenting judgment that (at [143]): the appellant had already served a sentence of approximately 23 years (taking a one-third discount upon his early plea to manslaughter and another one-third discount to allow for remission by the Commissioner for Correctional Services); a third retrial would be oppressive to the appellant; there were “considerable difficulties” in ensuring a fair trial; a substituted verdict would not leave the appellant unpunished and would bring finality to the case; and, significantly, the evidence of murder did not “remain very powerful” as opposed to manslaughter (citing R v Bell (Phillip James) [2010] EWCA Crim 3 at [46]).
Cheung CJ first acknowledged that the discretion to order a retrial is a statutory power under section 83E(1) of the Criminal Procedure Ordinance (Cap.221) (at [30]. For cases in the United Kingdom, see sections 7 and 8 of the Criminal Appeal Act 1968 and section 43 of the Criminal Justice Act 1988).
His Lordship then referred to the leading judgment delivered by Ma CJ in HKSAR v Zhou Limei (No 2) (2020) 23 HKCFAR 169 for the relevant principles of the exercise of that discretion (at [31]). Ma CJ therein revisited the authorities in commonwealth jurisdictions including Hong Kong and the United Kingdom and, for the purposes of that case, concluded that: first, whether the interests of justice are best served is to be assessed in the widest sense (per Lord Bingham in B (a Child) v R [2001] UKPC 19 at [39]) and this is the “critical question” upon which the exercise of discretion entirely depends; secondly, that requires considering the interests of all parties, including the public, the accused, and also the complainant (per Lord Diplock in Au Pui Kuen v Attorney General of Hong Kong [1980] AC 351 (HL) at 359); thirdly, all relevant factors for and against a retrial should be considered and they may carry different weights in each case; fourthly, the court must put significant weight on the factor that a trial has been gone through so that “cogent and compelling reasons” would be needed to warrant a retrial; and lastly the time an accused has spent in custody should be weighed against the likely sentence upon a retrial. Notably, Cheung CJ also added that precedents are “bound to be of limited assistance” because “each case turns on its own facts” (at [32]) and that the CFA would only intervene the CA's exercise of discretion in “limited circumstances” (at [33]).
Cheung CJ then examined the CA judgment and found that the major reason for disagreement was whether the appellant's counsel contributed to the trial judge's misdirection to jury (at [34]). His Lordship held that the CA was wrong to hold that the appellant's counsel made “a deliberate tactical decision” for withholding the relevance of that evidence also to the question of loss of self-control in order to achieve some “tactical benefit” (at [36]). It was because, first, neither the judge nor the prosecution could realise the relevance that evidence so that all of them “were on equal footing” (at [37]), and, secondly, no tactical benefit could possibly be achieved (at [38]). Furthermore, the appellant should also not be held responsible for not running the defence of diminished responsibility in the first two trials, particularly due to the technicality of this defence (at [40]). However, for the general question of whether an error made by a counsel is attributable to an accused, his Lordship held that it is fact-sensitive (at [39]). Since the CA took into account irrelevant considerations, the discretion miscarried which justified the CFA's intervention.
After dispelling the CA's erroneous exercise of discretion, the CFA exercised the discretion afresh. Together with the factors considered by the CA, the CFA was of the view that five matters were of significant importance (at [44]-[49]): (1) a fourth trial would be oppressive to the appellant; (2) the appellant had already been remanded in custody for a long period of time beyond the higher end of a sentence for manslaughter; (3) given the long lapse of time, the quality of the evidence from both the appellant and the witnesses would be irremediably affected; (4) a murder verdict would not be inevitable; and (5) although the appellant's counsel contributed to the mistakes upon which the appellant's convictions were quashed, he should not be held primarily responsible and he did not make any deliberate mistake.
Accordingly, the CFA substituted the verdict with a conviction for manslaughter and immediately released the appellant.
Commentary
As the CFA put it, this is a case “testing the limits of the discretion to order a retrial” (at [47]). It illustrates how discretion can be exercised differently by different judges in different levels of court. In that regard, the CFA's reasoning triggers some reflections on how discretion ought to be exercised in future cases particularly in the context of retrial.
The Relevance of the Irrelevant Factors
First, with the utmost respect, the CFA’s decision that the CA’s exercise of discretion miscarried must be right, but it is not entirely clear about the threshold of when an exercise of discretion would miscarry. The miscarriage seems to be simply due to taking into account irrelevant considerations (at [42]). But there could be two lines of explanations, either that taking into account any irrelevant considerations per se would render an exercise of discretion miscarry, or that the irrelevant considerations had led to the miscarriage.
The former does not sit easily with the CFA's ruling that it “would only intervene in limited circumstances” (at [33]) which is a high threshold. As Bokhary PJ (as his Lordship then was) pithily stated: “This Court is here to do justice. To allow an appeal just because something has gone wrong at the trial even though it has not resulted in a miscarriage of justice would not be doing justice” (Tang Siu Man v HKSAR (1997–98) 1 HKCFAR 107 at [129]).
However, if the latter line of explanation is adopted, the court would have to be satisfied that the irrelevant factors considered by the lower court were indeed the effective (or at least significant) causes to the erroneous determination. It would turn on the question of weight which causes other problems.
In this case, Macrae VP gave five factors whereas Zervos JA gave eleven factors for their conclusions. It is not clear if the irrelevant factors did lead them to order a third retrial or, hypothetically, they were not effective causes in the sense that taking those factors away would have made no difference. As such, the higher court must assess the weight of any irrelevant factors and that logically follows that the weights misgiven to the irrelevant factors would justify its intervention. It would become a de facto re-exercise of discretion.
Therefore, it is respectfully submitted that it is not entirely clear on which basis the CFA substituted the discretion exercised below, and it may need further judicial clarification.
When Would Irrelevant Factors Become Relevant?
Secondly, the CFA seems to leave open that those factors it considered irrelevant could potentially be relevant in certain circumstances. On the one hand, the appellant's counsel alleged mistake was irrelevant because the CA mistakenly thought that it was “a deliberate tactical decision” to achieve “tactical advantage”, and overlooked that the trial judge and the prosecution were “on equal footing” (at [37]-[38]). On the other hand, raising diminished responsibility for the first time at the third trial was irrelevant partly because it was of “highly technical nature” (at [41]). They seem to be irrelevant merely on the facts of this case but not in principle. There remains interesting (but unanswered) questions as to whether they could be relevant and, if so, when and how.
Would that make a difference if the appellant's counsel did make a “deliberate tactical decision” to gain strategic advantage? Or would it be the same if the defence was non-technical and could have been raised earlier but was not? These point to a more general question of law as to under what condition(s) can any errors or negligence at trial be attributed to an accused. The CFA indeed accepted that a mistake made by counsel is attributable to an accused dependent on the facts of each case (at [39]), but the technicality of a defence was not discussed. Regrettably, the CFA did not provide an answer to the more general question of law which is what circumstances would turn the irrelevant factors in this case into relevant factors in other cases. It might be because the exact question did not arise from the facts and was not necessary for the disposition of this case so that the CFA did not address it.
One possible answer could be that it is incorrect to ask whether any errors or negligence is attributable to an accused because, ultimately, the “critical question” is how the interests of justice are best served (per Ma CJ in Zhou Limei (No 2)). Therefore, the interests of all parties, not only the accused, should be considered in the round. It takes the issue to a higher level of abstraction to preserve flexibility to avoid fettering discretion, but it comes with the cost that it is of little assistance to future cases.
How Should the Discretion be Exercised in Future Cases?
Lastly, this case sheds light on a wider, more fundamental question of law on the exercise of discretion. I can do no better than respectfully cite Lord Millett on the inherent problems of discretion: “Reported cases are merely illustrations of circumstances in which particular judges have exercised their discretion […]. Since they are all cases on the exercise of a discretion, none of them is a binding authority on how the discretion should be exercised. The most that any of them can demonstrate is that in similar circumstances it would not be wrong to exercise the discretion in the same way. But it does not follow that it would be wrong to exercise it differently” (per Millett LJ (as Lord Millett then was) in Jaggard v Sawyer [1995] 1 WLR 269 (CA), 288 approved in Coventry (t/a RDC Promotions) v Lawrence [2014] UKSC 13 at [120]).
Ma CJ held that a second retrial would be “a ‘departure from accepted norms’ sufficient to constitute a substantial and grave injustice” if there is no “special or compelling reason” (Zhou Limei (No 2) at [9(5)] cited at [31]). Similarly, Lord Judge CJ held that a second retrial should be ordered if “the evidence” of the occurrence of an extremely serious crime “on any fair minded objective judgment remains very powerful” (R v Bell at [46] applied by McWalters JA in the CA at [143] cited at [17]). One way to interpret the aforesaid is that there is a presumption against a second retrial
The second way to give appropriate guidelines is to set down an inexhaustive list of potentially relevant factors in certain circumstances so that it would not be fettering the courts’ discretion. Along the development of case law, the list can be expanded to provide for certainty and predictability. Nonetheless, it is indubitable that too much of it would fetter the courts’ discretion so that the list “[cannot] be made too precise, but should remain relatively abstract and open-textured” to avoid “[leading] to odd or inequitable results” (D. J. Galligan, Discretionary Powers: A Legal Study of Official Discretion (OUP: Oxford, 1990) at p.46). It is telling that in the CA the three judges did not consider the same set of factors and even the numbers of factors considered varied amongst them. Professor Galligan wrote that: “[T]here may be guiding and constraining factors which must be taken into consideration (‘factors S1, S2, S3…’); the sergeant may be left to pick whichever men he thinks suitable, but be instructed to take into account the hours which each has already done on patrol. The statute […]
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.