Issues have concerned the power of appellate courts to disallow appeals against conviction if, notwithstanding that there had been, for example (s 385(1)(c)) a miscarriage of justice at trial, that miscarriage was not "substantial": the proviso to s 385(1). When is a miscarriage of justice "substantial"?
In Sungsuwan the appeal was based on alleged misconduct of counsel at the trial, in failing to adduce evidence that, it was claimed, would have supported the defence, from two prosecution witnesses. The Supreme Court unanimously held that the verdict was safe and that there was no miscarriage of justice in the conviction. Three judgments were delivered: individually by Elias CJ and Tipping J, and jointly by Gault, Keith and Blanchard JJ (delivered by Gault J).
Elias CJ held that it would be unlikely to be appropriate to apply the proviso where a miscarriage of justice in terms of s 385(1)(c) is made out (para 6).
Tipping J was firmer on this point, holding that in cases coming within s 385(1)(a) and (c) the proviso was fused with the error, in the sense that there was no need to apply the proviso once the error had been established (para 113). At this point, Tipping J footnotes R v McI  1 NZLR 696 (CA), illustrating that there was a degree of "awkwardness" in the relation between para (c) and the proviso. In that case, delivering the majority judgment of himself and Keith J, Tipping J had not gone so far as to separate para (c) from the application of the proviso, and had held that before the proviso could be applied, the Court "must be sure that the jury would without doubt have convicted had the matters giving rise to the initial miscarriage of justice not been present" (p 712, line 22). There are two points here worth noting.
First, the attempt in McI to imagine what might have happened had the error not occurred, is not used as the test in Sangsuwan. Admittedly, it makes sense in some contexts, but it can be a distraction from the real point which is, given that the error had occurred, what was its effect? In Sungsuwan the question is put variously as "whether the verdict is unsafe" (Elias CJ at para 7), whether the error prejudiced the accused’s chance of an acquittal (Gault J at para 67), whether the error was likely to have had an effect on the trial outcome (Gault J at 69), whether there was a real risk that the error affected the trial outcome, whether there was a real concern for the safety of the verdict (Gault J at 70), whether the error prejudiced the accused’s prospects of acquittal or a lesser verdict (Tipping J at 101), whether the error led to a real risk of an unsafe verdict (Tipping J at 107, 108, 110, 111, 116), whether the accused was deprived of a reasonable possibility of a more favourable verdict (Tipping J at 115).
Second, the quoted passage from McI shows that there is a high standard on the prosecution to show that the miscarriage of justice was not "substantial", implying that tolerance of miscarriage of justice is low. This appears consistent with Sungsuwan, as the formulae referred to in the previous paragraph indicate. Tipping J, however, and perhaps unintentionally, highlighted an inconsistency by referring in Sungsuwan to "the high threshold" for showing that the trial had been unfair (para 115, footnote 45). Is unfairness tolerated to a greater extent than other forms of miscarriage of justice?
One thing made clear in Sungsuwan is the relevance of the effect of the error on the verdict in cases where the error has led to the trial being unfair. Trials may, of course, contain errors, yet not amount to unfair trials. Trial unfairness is a narrow concept, focused on the questions of bias and application of the law to the facts. Some erroneous directions on the law may not cause a jury, overall, to misapply the law to the facts; they may, in the particular context, be insignificant slips. There has been some difference in the cases (see especially, Howse, below)about whether an error has to affect the verdict before the trial can be said to have been unfair.
Elias CJ held that it was difficult to envisage that a verdict reached without fair trial could not amount to a miscarriage of justice and it would not be likely that the proviso could apply (para 6). Gault J referred to errors that deny the accused a fairly presented defence, saying that these may readily permit the Court to find prejudice and in extreme cases may not require the Court to ask whether they affected the verdict (para 65). Tipping J addressed the point more directly, saying that, rarely, things may have gone so wrong at trial that a miscarriage of justice occurred without there being any need to refer to whether there was a real risk of an unsafe verdict (para 111, 112, citing, inter alia, the judgment of himself, and Richardson P and Blanchard J in R v Griffin  3 NZLR 577, 587).
When judges say that things will happen only "rarely", they happen all the time. To say "rarely" is to indulge in wishful thinking. The point is that whenever the trial has been unfair, there is no need to enquire whether the unfairness caused the guilty verdict. No matter how correct the guilty verdict may be, it must not be arrived at by an unfair trial.
It makes no sense to put the necessary risk of trial unfairness, before a substantial miscarriage of justice is found on that ground, at any different from the necessary risk of an unsafe verdict. Difficulty can arise in deciding between the two grounds, as is illustrated by Howse v R  UKPC 31 (19 July 2005), noted in this blog on 23 July 2005. There, the majority, in a case where it was not in issue that the admissible evidence supported the verdict, rather the issue was whether the trial had been unfair, applied the criterion of what course the trial would have taken if the errors had not been made. This was precisely the wrong test, as Sunsuwan shows the appropriate question is, given that the errors occurred, what part did they play. The correct approach would have been to ask, as did the minority in Howse, whether the errors could have played a part in the task of applying the law to the facts.
Readers of these blogs will be deeply satisfied to note that in our entry for 1 April 2005, commenting on Teeluck v State of Trinidad and Tobago, we asserted that loss of a fair trial was itself a substantial miscarriage of justice and did not require that the verdict be unsafe.