Mistakes that are not innocent mistakes: Bell v Tasmania [2021] HCA 42

Criminal responsibility is not usually imposed on people who are in no way morally blameworthy. At least, that is so for serious crimes as opposed to purely regulatory offences. There may be some vagueness over where the boundary is there. And, notoriously, at the opposite extreme, opinions may differ over what can reasonably be required of people who are subject to great levels of stress, such as shipwreck survivors who kill and eat a moribund occupant of their lifeboat.[1]

Mistakes may be made by a person who performs acts that constitute the physical elements of an offence. Some of these mistakes may amount to a denial of the mental elements necessary for criminal responsibility (we call those the mens rea elements). There is nothing controversial about that, as long as the mistake is about fact and not about the law defining the offence. But, sometimes a mistake may be made about some material circumstance (that is, a physical element of the offence, but one that does not engage a mental element), and it may raise a question as to whether it is right to impose criminal liability.

What would the conditions be for an acceptable exculpatory mistake about such a circumstance? Honesty and reasonableness seem to be obvious requirements: the mistake should be one that a reasonable person would have made, and it should have been made honestly, in good faith.[2]

The excuse[3] of mistake was discussed in Bell v Tasmania [2021] HCA 42 (8 December 2021). Here the emphasis was on another aspect of acceptable mistakes: they must be about the “innocence” of the conduct.

The offence here was supplying a controlled drug to a child, and a child was defined as a person under the age of 18 years.

The defendant (appellant) had injected a person with methamphetamine, thinking that the person was over that age.

Supplying the drug to another person was, and is, a separate offence. The mistaken belief, even if it had been honestly and reasonably held, was not an innocent belief: it was belief that another offence was being committed.

The appeal was unanimously dismissed.

Kiefel CJ and Bell J at [9] mentioned historical concerns over the acceptability of a gap between moral culpability and criminal responsibility, and at [12] the common law criterion of the rightness of branding the defendant as a felon and applying punishment. Those concerns were not present where the mistake warranted conviction for an offence: [13], [15]. A provision excusing a defendant from all criminal responsibility for the relevant conduct does not operate in an absence of belief that the conduct was lawful. This contrasts to a corresponding provision in Queensland, whereby the defendant is excused to the extent of the facts that were believed to exist, so that on the facts here the defendant would be excused for supplying the drug to a child but convicted for supplying it to a (non-existent) adult: [18]. The Tasmanian legislature had decided not to follow the Queensland enactment: [23].

Gageler J agreed that under the Tasmanian legislation, the defendant must think the conduct was innocent, that is, not an offence at all. Gordon and Steward JJ came to the same view: [46]-[47], and observed that this conclusion was not harsh, unfair, or contrary to the principles of the criminal justice system: [48]. Edelman and Gleeson JJ were of the same mind: [54], and they embarked upon an extended examination of the development of the common law excuse of honest and reasonable mistake. Again, the mistake must be such as to excuse all the conduct, and here it did not: [105].

It was significant in Bell that no issue was taken over the extent of the mental elements of the offence: the prosecutor only needed to prove the intention to supply the drug. There was no requirement for the prosecutor to prove that the defendant knew or believed the recipient was a child: Kiefel CJ and Keane J at [5], Gordon and Steward JJ at [41], Edelman and Gleeson JJ at [71]. This meant that the issue was not absence of mens rea, but instead was the applicability of the excuse of mistake.

______________________________

[1] You certainly don’t need to be reminded of R v Dudley and Stephens, and the ridiculous prospect of judges making law which they acknowledge they could not themselves obey. But I just mention that case as an aside: it is not a “mistake” case, but rather it is a “boundaries of the criminal law” case.

[2] Self defence can encompass mistakes, but the requirement may be that they are honest mistakes, even if not necessarily reasonable mistakes; the reasonableness requirement attaches to the force used in self defence. See, for example s 48 of the Crimes Act 1961 [NZ]. When I was a law student this provision used the masculine pronoun only, but now we have the recognition that some people are female: “Every one is justified in using, in the defence of himself or herself or another, such force as, in the circumstances as he or she believes them to be, it is reasonable to use.” This plainly needs updating, for now we have more pronouns in recognition of the LGBTQ+ people. Where will this end? To the revisers of statutes, I recommend (consistently with my stubbornly-held view: here) avoidance of all personal pronouns here. Try it.

[3] Is it an “excuse”, or a “justification”? The reasonableness requirement suggests justification (compare self defence in note 2 above). Absence of moral blameworthiness also suggests justification. But the expressions might be used without any intention of drawing a distinction between them.

Conviction appeals and judicial differences - lack of unanimous appellate verdict in a final appeal

In most appeals against conviction the focus is, to use the broadest of non-technical descriptions, on errors and on the soundness of the verdict at trial. In such appeals, differences among judges can occur on each of these matters.

An illustration is Hofer v The Queen [2021] HCA 36 (10 November 2021).

On whether there had been a qualifying error, Kiefel CJ, Keane J, Gleeson J (jointly) and separately, Gordon J, held that there had been such an error. Gageler J, while agreeing with the plurality in dismissing the appeal, held that there had been no such qualifying error.

And on whether the verdict at trial had been sound, the plurality held that it was, but Gordon J held that it wasn’t, and Gageler J didn’t need to address this question.

The case is of wide interest to appellate buffs, partly because of Gageler J’s recognition that he was bound by an earlier decision of the High Court (Weiss v The Queen [2005] HCA 81, discussed critically ten times on this site so far - use the search box) even if he disagreed with it, because its correctness was not in issue in this case: [97], and partly because of his discussion of how Weiss reorientated the function of appellate courts in deciding this type of conviction appeals: [85].

If appellate judges find that an error at trial was significant, in the sense that it created a real chance (etc, [120]) that it affected the verdict, then the appeal judges must themselves reach their own verdict on the available record (and on such new evidence as may be admissible on appeal), unless for reasons usually relating to the issues at trial - such as assessments of witness credibility - it is impossible to do so.[1] The Supreme Court of New Zealand has given itself this function too, by means of what I think is a rather strained application of techniques of statutory interpretation. See my discussion here, on 1 July 2021.

The appellate verdict in Hofer was not unanimous. The majority verdict favoured the upholding of the conviction. As this verdict was not unanimous, should a retrial have been ordered? And, in any case where an appellate verdict unanimously favoured there being a reasonable doubt about the appellant's guilt, would it be appropriate to order a retrial, or should an acquittal be entered?

The plurality judgment in Hofer contains a helpful discussion of the implications of non-observance of the rule in Browne v Dunn in criminal cases: [29]-[37].

___________________________

[1] Update: For illustration of circumstances where the Court considered it was inappropriate for it to reach a verdict, and instead ordered a retrial, see Orreal v The Queen [2021] HCA 44  (16 December 2021).

Rights consistent reading-down of the three strikes sentencing legislation

Our Supreme Court has allowed the appeal against sentence in the Fitzgerald case which I mentioned here on July 17, 2020: Fitzgerald v R [2021] NZSC 131 (7 October 2021).

The case required interpretation of (mainly) s 86D(2) of the Sentencing Act 2002:

(2) Despite any other enactment, if, on any occasion, an offender is convicted of 1 or more stage-3 offences other than murder, the High Court must sentence the offender to the maximum term of imprisonment prescribed for each offence.

It was universally acknowledged that in this case the result of application of a literal interpretation of this subsection was well beyond excessive punishment for the offence and it would shock the conscience of properly informed New Zealanders. This level of punishment was, in this case, a breach of s 9 of the New Zealand Bill of Rights Act 1990:

9 Right not to be subjected to torture or cruel treatment

Everyone has the right not to be subjected to torture or to cruel, degrading, or disproportionately severe treatment or punishment.

Does the phrase in s 86D(2), “Despite any other enactment” include s 9 of the Bill of Rights?

Deciding this can require considering an interpretive provision of the Bill of Rights, s 6:

6 Interpretation consistent with Bill of Rights to be preferred

Wherever an enactment can be given a meaning that is consistent with the rights and freedoms contained in this Bill of Rights, that meaning shall be preferred to any other meaning.

Here, the phrase “can be given” is important. How big a stretch is permitted by this phrase?

The role of s 6 is not necessarily central to reasoning, because the common law has parallel interpretive tools, particularly the principle of legality (discussed here on 11 February 2021 in relation to D (SC 31/2019) v Police [2021] NZSC 2 ). These parallel tools may be used in combination, or separately.

In this appeal the majority judges were Winkelmann CJ, O’Regan and Arnold JJ jointly, and Glazebrook J.

Winkelmann CJ reasoned that s 6 goes beyond the principle of legality [57], describing it as a powerful interpretive obligation [73]. Noting that the common law permits, in appropriate circumstances, the reading-in and reading-down of legislation [59]-[62], and remembering that the result must not be a refusal to apply legislation [66], she concluded that an exception can be read-in to s 86D(2), [112]-[121]. This was consistent with the purpose of the provision [122] and it applied only where s 9 of the Bill of Rights would otherwise be breached [137]. Further, where the exception applies and ordinary sentencing principles come into play, these are supplemented by a principle requiring a stern sentencing response to such recidivism [138].

On the other hand, O’Regan and Arnold JJ relied primarily on the principle of legality as permission for reading-in the exception to s 86D(2). They pointed out [206] that it would have been easy for Parliament to have specified that the provision applied despite the Bill of Rights. Section 6 requires a similar approach to that adopted under the common law principle of legality [207]. After surveying decisions they concluded [215] that apparently unrestricted general words are not sufficient to displace presumptions reflecting core legal values. There are more such values than are included in the Bill of Rights, and in this sense the principle of legality is wider in scope than s 6 [217]. Explicit statutory language is required to override the right protected by s 9 [218]. A rights-consistent meaning of s 86D(2) can be given under s 6 without defeating Parliament’s purpose [219]. This conclusion is supported by another principle of interpretation: that legislation should be read, so far as possible, as being consistent with New Zealand’s relevant international obligations [225].

Glazebrook J, agreeing in the result reached by the other majority judges, was careful to say what she was not making any comment about. See footnote 337, para [243], [244] and footnote 348, [245] and footnote 351, footnote 352. She applied ordinary principles of interpretation, namely the purpose of the legislation [249], the need for reading-down in the light of the principle of consistency with international law and with fundamental human rights, and the constitutional status of the Bill of Rights [250], and the principle of legality [251]. She added in footnote 363: “I do not wish to comment on the relationship between s 6 of the Bill of Rights and the principle of legality, except to say that I agree with Winkelmann CJ that s 6 of the Bill of Rights may go further than the principle of legality.” At footnote 366 she did not agree with the Chief Justice’s comments regarding a stern sentencing principle being added because of s 86D(2).

The majority judges agreed that where the read-in exception resulted in the application of ordinary sentencing principles in a particular case (such as this), a discharge without conviction could be considered, although it might only rarely be appropriate. The question of sentence was remitted to the High Court.

The interpretation of s 6 is now delicately poised: Winkelmann CJ [67] appears to be willing to give it a strong meaning, consistent with Ghaidan v Godin-Mendoza [2004] UKHL 30, [2004] 2 AC 557, while William Young J [294]-[302] (the dissenting judge in this appeal) is more conservative, requiring a rights-consistent interpretation to be reasonably available. The other judges appear to be reserving their positions on this issue.

Personally, and I just say this to get you thinking, my view at present is that the interpretation of s 6 is only a delicately poised issue because Lord Cooke refused to accept he had been wrong in Phillips.

Fair criticism?

Yesterday our media published an article critical of a judge who had repeatedly failed to direct juries correctly on the burden and standard of proof in criminal trials. The article gave me the impression that the judge was stubbornly refusing to follow Court of Appeal corrections of his approach.

I don’t know the judge, but I note that he has now retired.

Close inspection of the four cases shows that the Court of Appeal did not start dealing with these appeals until after the fourth trial. The same pattern applies to other cases alluded to in the article. This means the judge may not have been alerted to the errors as identified by the Court of Appeal.

It may even be that the judge had simply been following an out-dated (or what is now an out-dated) Bench Book model instruction set out for judicial guidance. I don’t know if that is so, because Bench Books are not available for inspection. The directions may have been toughened up after one of the Court of Appeal judgments, delivered on 8 October 2019.

The trials I am referring to were held in June 2016, December 2016, August 2017, and November 2017. The Appeal judgments were given in May 2018, April 2019, July 2019 and October 2019.

The media article linked above refers to the Court of Appeal's decision in R v Wanhalla [2006] NZCA 229, but the judgments in that case, delivered by three judges jointly and two judges each separately, can be read as not being expressed in mandatory terms; for example, from the joint judgment at [48] "something to be said ... at least in broad terms", [49] "inclined to the view that Judges should ...", [52] "there is no single formula which is required", "we are not to be taken as asserting that the formula just stated is mandatory", "It is sufficient to make it clear that the concept involves a high standard of proof which is discharged only if the jury is sure or feels sure of guilt."

And the Court in Wanhalla was not intending to change the way judges directed juries on the burden and standard of proof: [53] "It is most unwise for appellate courts to change course suddenly as to how trial Judges must sum up on the standard of proof."

Problems arising from departure from the Wanhalla direction were only becoming evident to the Court of Appeal by May of 2018, and on 8 September 2019 the Court said that such departure "is a perilous course." This firming-up of the status of the Wanhalla direction only occurred after the cases criticised in the article mentioned above.

The lengthy time it takes to get guidance from appellate courts could be blamed for the numbers of trials containing errors.

It is important to be fair when criticising anyone. It seems to me that the criticisms here may not have been soundly based.

My discussion of our leading case on the burden and standard of proof is here, and the Australian case is discussed here.

Update: You might like to look at a paper by Jason Anthony Aimone et al., "An Experimental Exploration of Reasonable Doubt" (14 September 2021), in which the results appear to suggest that jurors are influenced more by their own pre-conceived ideas of what a reasonable doubt is, than by directions from trial judges, and that differences in assessments of the standard of reasonable doubt correlate strongly with jurors' race.

Convictions arising from "the same facts": availability of the special plea of previous conviction

Whether a plea of previous conviction is available depends, in New Zealand, on the application of s 46 of the Criminal Procedure Act 2011. In particular, the phrase “arising from the same facts”: does the present offence arise from the same facts as an offence for which a conviction has been entered?

I used to think that this has the effect of requiring the prosecutor to allege all the relevant offences in the one proceeding, to avoid dragging things out with multiple separate prosecutions.

However, it has been made clear by the Court of Appeal that it is necessary to examine the essence of each charge and to compare the facts needed for proof to see to what extent there is an overlap. Some overlap will occur, but is it an overlap to the extent of giving rise to “sameness”? See Rangitonga v Parker [2016] NZCA 166, [2018] 2 NZLR 796.

A clear explanation has been given recently (31 August 2021) in Mitchell v Police [2021] NZCA 417. The Court adopted Venn diagrams as suggested by counsel. The issue was whether a conviction for driving with excess breath alcohol was a bar to conviction for driving on the same occasion contrary to the terms of a zero alcohol licence. The common fact was driving with alcohol in the breath (not per se an offence). The offences were held not to arise from the same facts, because in essence the breath alcohol offence arose from the level of alcohol in the driver’s breath, and the licence offence arose from the zero alcohol terms of the licence.

The Court noted at [36] that it did not favour the analysis which seems to have found favour with the Supreme Court of South Australia in two cases advanced by the appellant: Arthur v Police [2008] SASC 213, (2008) 101 SASR 529 at [42] and Jones v Police [2019] SASC 36, (2019) 135 SASR 255 at [51].

Update: on 14 December 2021 the Supreme Court granted leave to appeal.

Another update (7 July 2023): It is now over a year since the Court heard oral argument on this appeal. In the interests of free discussion, I offer the following observations:

1. In this case the charges were filed at the same time and pleas were taken on the one occasion. Usually, charges would be read and pleas entered, a prosecution summary of facts would be read, and the bench would invite submissions before deciding whether convictions should be entered (and if they were entered, they could nevertheless be provisional if an application for a discharge without conviction was to be made). If convictions were entered, the judicial officer would usually enter those by the speech-act of announcing that "on each charge the defendant is convicted". That is, convictions would be entered simultaneously, not sequentially, and no question of there being a previous conviction would arise. But here, as described at [17] of the CA judgment, the judge suggested a plea of previous conviction to the second charge. This unusual intrusion into the decision on plea set this particular ball rolling. However, it did create an opportunity to resolve an important issue: does sequential entry of pleas on one occasion create opportunities for entry of pleas of previous conviction?

2. The application of the rules concerning previous acquittal and previous conviction in ss 46 and 47 of the Criminal Procedure Act 2011 to sequential, not simultaneous convictions, is apparent in the phrases in (for example) s 46: "has been convicted", "currently charged", "was convicted", and "not readily available ... at the time the charging document for the more serious offence was filed". This reflects the policy of promoting finality, which is the main policy underlying the protections against double jeopardy.

3. The appellant correctly submitted on this appeal that the focus of s 46(1)(b) is on the facts rather than the offending. But the reference is to "those facts", and the issue is how broadly to construe that phrase. It seems to me that this points to the date and time of the commission of that offence and of the obtaining of the information about it by enforcement authorities. This is consistent with the exception delineated in s 46(2)(b). Construing "those facts" broadly in this way is also consistent with the breadth of the expression "any other offence" in s 46(1)(b).

4. In this case the question becomes, what information was obtained when the appellant was apprehended? He was driving on a road, he had an alcohol level over the limit, he was in breach of his zero-alcohol licence.

5. The appellant was properly charged with two offences and could have been simultaneously convicted of both. But it would have been a breach of s 47 if he had been charged with and convicted of one of those offences and later, perhaps after he had served his sentence for that, charged with the other. The second offence arose from the same facts as the first.

Update: Mitchell has been decided, as noted here on 11 August 2023.

Registrars and disputed bail conditions

In New Zealand we have no statutory right of appeal against a decision of a court registrar concerning bail. Registrars may determine what, if any, conditions are to be imposed in a grant of bail, pursuant to s 27(2) of the Bail Act 2000. The relevant appeal provision, s 41, does not refer to appeals against decisions of registrars.

I have always thought that a bit strange. True, registrars can only deal with bail if the prosecutor consents, but what sort of safeguard against onerous bail conditions is that?

And in the real world, a defendant may well decide to sign an onerous bond and get out of court as soon as possible, instead of waiting to have the registrar’s decision challenged. I have heard defendants say, “Nah, fuggit, I’m out of here.”

I asked the Minister of Justice in 2012 whether there should be clarification of this in the Act. After seeking the advice of “officials” she replied to the effect that no action was considered necessary.

The inference is that registrars should only impose bail conditions with the agreement of the defendant - although in the aforementioned reality freedom of choice is somewhat limited. And what if a registrar decided to impose a condition over the objection of the defendant?

Matters such as these have received some clarification in a recent case that is primarily concerned with the inherent power of the judges of the District Court to supervise and direct registrars in the conduct of judicial business: District Court at Christchurch v McDonald [2021] NZCA 353 at [38]: 

“… Parliament’s expectation was that while registrars would assist judges by determining some bail applications, that undertaking would remain subject to usual judicial oversight and supervision.”

Clearly, Parliament failed to put its intention into statutory words, presumably having confidence that we would play the game and guess at this.

Rights of appeal are statutory, and it is most unlikely that anyone would have thought to have resort to submitting that a judge of the District Court has inherent power to hear appeals. Indeed, the position alluded to in McDonald, which was really all about the power of a judge of the Family Court (a division of the District Court) to remove bail decisions from registrars, seems to be that the Chief Judge of the District Court would issue a direction that if a defendant objected to a proposed bail condition the registrar would, instead of making any decision, refer the issue to a judge.

This would avoid the suggestion of any appeal or a review. Indeed, review would be inappropriate because it would assume a degree of discretion exercised by the registrar and the judge would only intervene if that discretion had been exercised unreasonably. Far better for the judge to approach the question of bail de novo.

To ensure the point is not buried in the stack of judgments, a practice direction should be published so that, in the event that a registrar may be unaware of the proper practice, the procedure could be pointed out to that person.

Bail conditions that a defendant may think are onerous are usually imposed to help the defendant avoid getting into further trouble. People for whom that is a rather subtle point would be assisted if it were made by a judge rather than a registrar. McDonald speaks of judges, not mentioning the other judicial officers of the District Court: Justices of the Peace and Community Magistrates. Putting disputed bail matters before judges would be the most efficient course, because judges hear appeals from determinations of the other judicial officers, so the issue will end up with a judge anyway.

Update: The Supreme Court refused leave: McDonald v District Court at Christchurch [2021] NZSC 149 (5 November 2021), saying at [11] that it was not persuaded that it was arguable that the Court of Appeal had been wrong in the way it addressed the issues.

Conviction appeals: a virtual proviso?

I am not saying that updating one’s textbooks is a wonderful hobby.

But it can bring to your attention cases that, although currently suppressed, are of great interest to lawyers.

One such opened on my screen this morning. Because of its suppression order, and the possibility that the order may still be in effect (the case is not currently freely available online), I dare not speak its name. Nor will I say anything about its alleged facts, the charge(s), or anything else that could with any seriousness be regarded as a breach of the order.

The legal point is about how appellate courts are to interpret s 232(2)(c) of the Criminal Procedure Act 2011, concerning certain types of appeals against conviction. Specifically, and in historical context, has the removal of the proviso affected the meaning of “miscarriage of justice”? Under the proviso, an appeal could be dismissed if, notwithstanding the existence of a miscarriage of justice, it was not a “substantial” miscarriage of justice.

The new approach, under the current legislation as interpreted by the Court in this case is, if there has been an error relating to the trial that is more than inconsequential or immaterial, ask (1) whether the error has resulted in a real risk that the outcome of the trial was affected adversely to the appellant; and if there is that risk, ask (2) whether on the material available to the appellate court the court can be sure that the appellant is guilty, taking a conservative approach to this assessment in the light of the disadvantages that the appeal court has.

Has the Supreme Court read-in a proviso? On the plain words of the legislation, which are admirable for their simplicity and not in need of modification, there is no proviso. The appellate judges just assess a risk, and do not have to reach a verdict. But the approach set out by the Supreme Court comes down to: if there is a real risk that an error affected the result of the trial adversely to the appellant, allow the appeal “provided that” (my words)  the appeal can be dismissed if the appeal court finds the appellant guilty.

I suggest that it is important to recognise that an appellate court is not permitted to readily accept that guilt is proved on the record, but must be very careful and only dismiss appeals where guilt is obvious.

And to those who say, "But Don, if the answer to question (1) is yes, how can the answer to (2) also be yes?" I can only reply, I feel your pain.

The Court has turned a statutory question which addresses how the fact-finder at trial could have been affected into a question about the appeal court's verdict.

To what extent was there a problem of statutory interpretation here? To what extent should a court rely on the expectations of people who work in the background in the preparation of legislation (and who write explanatory notes to bills), to ascertain the meaning of an enactment? There are statutory directions about how legislation is to be interpreted (we have the Interpretation Act 1999 and the yet to come into force Legislation Act 2019 update: it was fully in  force, except for s 148, on 28 October 2021) and these point to using the text and purpose of the enactment - and here we would look at the purpose provision of the Criminal Procedure Act 2011, s 3, which is simply "to set out the procedure for the conduct of criminal proceedings". On the face of the legislation (s 232 of the Criminal Procedure Act 2011) there is no interpretive difficulty.

On a plain reading, s 232 reproduces the approach to appeals that had been accepted in R v McI [1998] 1 NZLR 696 (CA), as described in Matenga (below) at [12]-[13]. The focus in McI was on what the jury would have done, and the appellate court was not required to come to its own verdict. This raises the question whether the drafters of the legislation were thinking of the law as stated in McI when they said that no change was intended. [1]

It is only when one tries to fit the law as modified in Matenga into the new provision - in accordance with the prediction of the legislation drafters that there would be "no change to the core principles underlying the courts' current approach" to these appeals - that some interpretive stretching is needed. What the core principles underlying the courts' approach were is indeed itself an interpretive question.

I have commented on a case concerning the relevance of the old law to new law on appeals, Baini v The Queen [2012] HCA 59 at [14]-[15]. (In the update to that comment, I have referred to the present case, but with its citation only.)

In the present appeal the interpretive point was obiter, because the Court found that an evidential ruling had rendered the trial unfair, and that therefore the conviction(s) had to be quashed without further consideration of the evidence. Naturally, the Court would want to decide at the earliest opportunity how s 232(2)(c) works, and it did seek submissions on the point, but as it turned out there was no argument because counsel agreed that there had been no change in the law. Normally, failure to take an issue, and the obiter nature of the decision, would make a case weak authority, but obviously here the authoritative decision has been made.

Still, one mustn’t mope. Moving forward with this: the position really isn’t all that bad. The decision is that if an appeal court finds that there was an error at trial that created a real risk of an adverse outcome for the defendant, the appeal against conviction must be allowed unless the appeal court can be sure, on the evidence properly admissible, that the appellant’s conviction was correct. This approach is that which had been clearly set out in R v Matenga [2009] NZSC 18 (where inadmissible evidence had been adduced by the prosecution and the Supreme Court could not be satisfied that the admissible evidence proved guilt). In Matenga the Court noted at [32] that where a case turns on the honesty or reliability of witnesses the appeal court is unlikely to be able to be sure of the defendant’s guilt, and this was the position in Matenga (at [35]). Where inadmissible evidence was relied on at trial, the appellate approach is as stated in Lundy v R [2019] NZSC 152 at [42], and see Ellis v R [2021] NZSC 77 at [29]-[30] (leave to adduce evidence refused).

I think it would be sensible to treat the first question as whether there could have been the requisite risk, and then, if there could have been, ask whether there was such a risk, and to answer that by reference to the appeal court's verdict.

Interestingly, in refusing leave to adduce evidence in the Ellis appeal, the Court rejected a Crown submission that on appeal the Court would essentially be acting as the "final trier of fact" (at [32]), and stated that it would be engaged in "determining whether a miscarriage of justice occurred at the appellant's trial. That is a quintessentially appellate task." This view pushes any thought of the proviso (the old law applies to this appeal) well into the background.

_________________________

[1] Mistakes can occur in the explanatory notes and official commentaries. In the Select Committee Report on the Bill the Commentary on the appeal provisions contains a misunderstanding of the significance of the word "rehearing" (see my note on 2 December 2019). Anyway, the Commentary was added when the original draft of the relevant clause (236) was revised by removing a reference to rehearing, and by defining substantial miscarriage of justice to include reference to trial unfairness, so that after later removal of the word "substantial" by the Committee of the Whole House, the clause  was brought into a form mirrored in what is now s 232. I recently found my submissions to the Select Committee (17 February 2011) in which I raised concerns about the appeal proposals, including an objection to the use of the expression "substantial miscarriage of justice". The then Chief Justice wrote to the Committee a few days later (25 February), also submitting that the word "substantial" should be omitted. The fact that the Bill was called the Criminal Procedure (Reform and Modernisation) Bill does suggest that it shouldn't be assumed that its provisions do not reform the law. And you can see from the linked documents that the Commentary's assertion of no change precedes the deletion of the word "substantive" at the last Committee stage. However, the same assertion is repeated in the Supplementary Order Paper (SOP No 281) which deleted the word "substantial": "The amendment simplifies but does not alter the test for allowing an appeal against conviction." While that may be partly true, you might still wonder why, when the wording was under consideration, no effort was made to expressly give any verdict determining ability to the appellate court by including an equivalent to what the Court has in the present case formulated as the second question.

The police as community caretakers - a "third source" authorisation?

Can the police enter without warrant (or statutory or common law authority) the house of an absent occupier to search for weapons that the absentee might use to commit suicide?

The Supreme Court of the United States has held, no: Caniglia v Strom, 20-157 USSC May 17, 2021.

The Court was unanimous, and was careful to make it clear that this decision did not affect the established law relating to entry without warrant but in exigent circumstances. The exigent circumstances exception has been confined within narrow limits, essentially so that the law conforms with common sense.

Broadly, it is not necessarily unlawful, in exigent circumstances - that is, circumstances of emergency to protect life - for the police to enter private property without first applying for a warrant to authorise that entry.

Independent of exigent circumstances, the police may (in the United States at least) generally take actions that any private citizen might take.

This decision of the Court rejects the proposition that the police may act pursuant to a “community caretaking” duty.

So, caretaking is insufficient to make warrantless entry of private property lawful; there must be exigency.

Recognition that the police can do what any private citizen might lawfully do, is controversial. It calls to mind the “third source” of governmental authority proposed by some jurists. See, for example, BV Harris, “A Call to Maintain and Evolve the Third Source of Authority for Government Action” (2017) 27 New Zealand Universities Law Review 853. See also my comment on Ngan.

The "third source" theory, repeated almost to the point of reifying the concept, is that sources of legal authority - here, the authority of the police as exercisers of the executive power of law enforcement - are legislation, firstly primary, or secondly delegated (regulations or legislative instruments), or "thirdly" facts which call for a response. [1]

Regardless of its source, a power of search must be exercised reasonably. Reasonableness (or, absence of unreasonableness) is not the source of the power, but it is descriptive of how the power must be exercised to continue to deserve recognition as lawful.

Unlawful searches are unreasonable (there are very narrow exceptions to this - for example, where the police in good faith endeavoured to comply with a law that was open to various interpretations and which needed clarification by the legislature or by the courts), and lawful searches are sometimes carried out unreasonably.

We might doubt whether a third source is needed in the context of cases like Caniglia v Strom. The courts can craft rules about police entry into residential property if elaboration of the requirement of reasonableness is required. But the source of lawful authority is primary or secondary legislation. To place the source of legality in the facts of a case is to invite anarchy. [2]

____________________________

[1] I am differing here from the usual counting of sources, in discussions of governmental powers, where the first two are parliamentary and the royal prerogative. In the present context it seems sensible to recognise the two types of legislation, as the prerogative has no application to the law of search. It may be suggested that the third source is really that which is necessarily implied with the grant by legislation of search powers. But implications are not necessarily obvious without a factual context, and where that is so I prefer - for third source purposes - to think of the facts as giving rise to (in the sense of motivating) the interpretation. Implications from the text of legislation are first or second source, whereas implications from the facts might be third source - if indeed the third source is a real thing. See S v Commissioner of Police [2021] NZHC 743 at [64], [80].

[2] Just a bit of exaggeration here, probably. The name "third source" may be misleading: it seems to be not a source of power but a freedom that exists to do what is not prohibited by positive (legislation or judicially-made) law. See Jeff Simpson, "The Third Source of Authority for Government Action Misconceived" (2012) Auckland University Law Review 86. While we are thinking about this, and the hypothesised power of the police to do anything that an ordinary person could lawfully do, let's note the Right Honourable Sir Stephen Sedley's comment in London Review of Books, vol 43, no 13 (1 July 2021), p 19: (in an only slightly different context), that is "a formula that embraces acting out of caprice, greed or spite."

Protecting prosecutors from police prosecution

In Ontario (Attorney-General) v Clark  2021 SCC 18 (30 April 2021) the Supreme Court of Canada explained why Crown prosecutors should have immunity from prosecution by their clients the police for alleged failure to carry out their public duty in the conduct of prosecutions.

The issue of immunity here arose from police complaints that prosecutors had failed to challenge at trial allegations by the defence that officers had assaulted defendants. The result of the prosecutions on charges relating to robbery had been a pre-trial stay of proceedings for one defendant and a reduced sentence for the other who had proceeded to trial.

Prosecutors are not shielded from all claims alleging abuse of powers. There is in Canadian jurisprudence an “accused-centred policy thread woven through the authorities” (majority judgment delivered by Abella J, at [40]). Sometimes, defendants may sue prosecutors, so should police also be able to? No: “allowing police officers to initiate such causes of action would raise profound risks to the rights of the accused and to prosecutorial independence and objectivity, and it would undermine the integrity of the criminal justice system.”

Reasons are: the difference in function of police and prosecutors ([41]), the need for Crown prosecutors to act independently ([43]) being a buffer between the citizen and the police ([44]), the need to avoid injustices and wrongful convictions ([48]), the need to avoid placing prosecutors under pressures which could distort independent decision-making ([51], [53]),the need to protect public confidence in the independent and objective ability of prosecutors to conduct fair trials ([56]), the need to protect defendants’ fair trial rights by freeing prosecutors from accountability to the police whose interests are adverse to those of defendants ([58]).

In summary ([51]):

“Piercing the immunity of Crown prosecutors to make them accountable to police officers puts them in perpetual potential conflict with their transcendent public duties of objectivity, independence and integrity in pursuit of ensuring a fair trial for the accused and maintaining public confidence in the administration of justice.”

Côté J dissented, on the basis that a high threshold for liability for misfeasance in public office would protect prosecutors while also giving remedies for police officers who are subject to severe criticism in criminal proceedings to which they are not parties. The need for a remedy in an individual case should not be obliterated by generalities. For some of the important points made by Côté J, see [117]-[122], [129], [133], [138].

Preventing unfairness of one's own making

My heart missed a beat yesterday at breakfast. A newsreader announced that the Supreme Court of the United Kingdom had held that fairness does not trump public safety concerns.

Not to worry. The fuller story was that a stay of proceedings should be ordered if there was no alternative means of avoiding unfair proceedings.

This was a civil case: R(on the application of Begum) v Special Immigration Appeals Commission [2021] UKSC 7. In civil cases fairness is assessed by taking into account the interests of all parties, and some forensic disadvantage will not necessarily prevent continuation of the proceedings.

A stay of proceedings, when used in the context of fairness, is usually aimed at protecting a party from the unfairness that would occur if proceedings were allowed to continue.

In the unusual circumstances of this case, the stay would prevent Ms Begum from appealing against an order depriving her of her citizenship of the United Kingdom. At least, until the circumstances changed to the extent that she would be able properly to participate in her appeal.

The related issues before the Supreme Court were able to be determined without causing unfairness to her, because no issues of fact were involved.

Some interesting points are illustrated in the judgment of the Court, delivered by Lord Reed P. Examples are the differences in approach to appeals, depending on whether they are against discretions or against evaluative judgements; the need for appellate courts to have an evidential basis for their determinations of facts; the appropriateness of judicial deference to the decisions of ministers who are answerable to Parliament; and the need to recognise when guides to decision-making are not rules and so do not turn a discretion into an evaluative judgement.

(I am spelling judgement with that middle e to draw attention to the point that what is being addressed is the mental process, not the outcome. The outcome is, as we know, spelt in law without that middle e. The courts, in contrast, tend to use the latter spelling most of the time.)

The danger of holding dual-citizenship is illustrated by the circumstances of this case.

The principle of legality, rights limitation by necessary implication

Parliament may make laws that deliberately infringe people’s rights. Including rights contained in a Bill of Rights.

 

But bills of rights may require that legislation is to be interpreted consistently with individual rights, so far as it is possible to do so.

 

The exact wording of this sort of interpretive requirement may vary between bills of rights in different states. [1]

 

However, there is a generally applicable principle which requires that when Parliament intends its legislation to infringe individual rights, it can only do so “by express words or necessary implication.” This is called the principle of legality.

 

Differences over whether rights infringement was a “necessary implication” were central in D (SC 31/2019) v Police [2021] NZSC 2 (9 February 2021). [2]

 

The majority held that the relevant legislation was insufficiently clear to displace the presumption against retrospective penalties. The minority considered that the only available interpretation of the legislation was that it did displace that presumption.

 

It is for Parliament to decide what to do to avoid the consequent incongruities (noted by Glazebrook J, dissenting on this point, at [243]-[248] and referred to in the joint majority judgment of Winkelmann CJ and O’Regan J at [82]).

Given the disagreement here among Judges of the final appeal Court over application of the principle of legality, you might fairly ask whether the majority could have given clear guidance on how such disagreement might be avoided in future. Should the principle of legality have been modified by excluding “necessary implication”, so that clear words are required for legislation to infringe rights? A statutory example of clear words is mentioned at [79].

 

The rejection in New Zealand of the more “far-reaching” interpretive approach in Ghaidan v Godin-Mendoza [2004] UKHL 30, [2004] 2 AC 557 drew mention from Glazebrook J at [253] of academic commentary about when the courts might be prepared to override Parliament’s purpose.

 

The extent to which, on an appeal against sentence, the appellate court should consider evidence of recent (that is, post-sentencing) assessments an offender’s prospects of rehabilitation, also arose for comment in this case.

 

No objection had been taken to the consideration of such material by the appellate courts here, so the point did not need to be decided, but William Young J observed that on sentence appeals the issue is whether there had been an error at sentencing, so he had reservations about the practice ([305]-[307]). Glazebrook J had reservations too, but on the narrower ground that the scheme of the present legislation seemed to be against consideration of such updating material (at [262]).

There are also some interesting remarks on judgment anonymisation in contrast to name suppression: see [136]-[147].

I don’t need to distress you with a more detailed consideration of this appeal, because the Court itself sets out an admirably clear summary of the positions taken by each Judge and the result of the case (at [1]-[11]). Thank goodness for that.

Update: All the judges in this case recognised that Parliament may wish to address the anomalies identified by Glazebrook J (see majority judgment at [82]). A Bill was introduced on 17 March 2021, and it became law on 23 March 2021, to make the retrospective effect explicit in respect of people convicted and sentenced on or after 14 October 2016 (when the Registration Act came into force) for offences committed in New Zealand or overseas before that date. An exception is made for the individual appellant in D (SC 31/2019), who can keep his victory in the appeal. This exception may be for constitutional reasons, illustrating the separation of powers - Parliament, in correcting its legislation after the successful efforts of an appellant, will not interfere with the court's decision in that individual's case.

 

 

[1] For example, s 6 of the New Zealand Bill of Rights Act 1990, s 3 of the Human Rights Act 1998 [UK], and see my comment on the difference between these (8 September 2011). See also my discussion of Momcilovic (9 September 2011).

 

[2] The principle of legality is referred to in this case at [76]. For background, see Bruce Chen “The Principle of Legality: Issues of Rationale and Application” (2015) 41 Monash University Law Review 329. In New Zealand the principle has statutory form in s 6 of the New Zealand Bill of Rights Act 1990. Its rationale as an interpretive aspect of the rule of law is that Parliament understands the way the courts will interpret its legislation, treating it as improbable that there would be a departure from fundamental rights without express and unambiguous statutory wording to avoid the risk that legislation will have unintended consequences.