Home | Databases | WorldLII | Search | Feedback Otago Law Review

Rodriguez Ferrere, M B --- "The unnecessary confusion in New Zealand's appellate jurisdictions" [2012] OtaLawRw 8; (2012) 12 Otago LR 829

Last Updated: 16 April 2014

The Unnecessary Confusion in New Zealand’s Appellate

^{Jurisdictions}

M B Rodriguez Ferrere*

December 2012 was the fifth anniversary of the Supreme Court of New Zealand’s delivery of its judgment in Austin, Nichols & Co Inc v Stichting Lodestar.¹ Few will have marked the occasion. The Supreme Court itself believed it was delivering an innocuous judgment that simply restated “well-established” principles surrounding appellate review.²

However, when December 2012 also saw the delivery of a judgment in a case where counsel “expressed a level of frustration that conflicting decisions of the High Court have created uncertainty over what [appellate] test is to be applied by the High Court”³ – a situation directly caused by Austin, Nichols – this comment argues that far from being innocuous, the judgment instead has the dubious honour of forming the centrepiece of New Zealand’s complicated and confusing approach to appellate jurisdiction.

This paper examines that approach, attempts to identify the reasons why it it is confusing and lastly, suggests a potential path for reform and improvement.

The many types of appeal

Chief Justice Elias’s judgment of the Court in Austin, Nichols was “arguably the most important statement ever made by a New Zealand court on how to go about deciding an appeal”.⁴ It held that on a “general appeal, the appeal court has the responsibility of arriving at its own assessment of the merits of the case.”⁵ Whilst an uncontroversial statement in itself, confusion arises from what it does not say: if that is the responsibility of an appellate court in a “general appeal”, what is its responsibility when dealing with other types of appeal?

A “general appeal” – ie, an appeal on a question of fact and law – is perhaps the most common type of appeal, including most appeals against judgments of the District Court⁶ and the High Court.⁷ However, this is

^* Lecturer, Faculty of Law, University of Otago. This comment first took the form of a Faculty of Law seminar at the University of Otago and I appreciate my colleagues’ helpful comments and suggestions at that forum.

^{1 Austin, Nichols & Co Inc v Stichting Lodestar [2007] NZSC 103, [2008] 2}

^{NZLR 141.}

^{2 At [6] per Elias CJ.}

^{3 Roberts v A Professional Conduct Committee of the Nursing Council of New}

^{Zealand [2012] NZHC 3354 at [22].}

^{4 Andrew Beck “Litigation” [2008] NZLJ 19 at 19.}

^{5 Austin, Nichols, above n 1, at [5].}

^{6 District Courts Act 1947, s 72.}

^{7 Judicature Act 1908, s 66.}

only one type of appeal, and others include appeals that are limited to questions of law, appeals from exercises of discretion or appeals by way of case stated. For each of these appeals, the role of an appellate court and how deeply it can examine the decision of the court or decision-maker below is different.

Austin, Nichols provided guidance to appellate courts about their role when they hear general appeals, and other decisions have provided guidance on the other main types of appeal, as listed below:

• General appeals: the Court considers a decision or judgment under appeal in its entirety, and must arrive at its own conclusion on both the facts and law. The Court uses the evidential record that was available to the court or decision-maker below, unless the appeal provision provides for an appeal de novo, in which case the Court hears evidence afresh.⁸

• Appeals against exercises of discretion: when the decision of a court or decision-maker below is “discretionary”, the Court does not come to its own conclusion, and must not disturb the decision unless the decision-maker used the wrong principle, failed to take into account some relevant matter, took account of some irrelevant matter, or was “plainly wrong”.⁹

• Appeals against questions of law: the Court must accept factual findings but can intervene when a decision-maker or court below has applied a wrong legal test, come to an unreasonable conclusion based on the evidence before it, failed to take into account some relevant matter, or taken account of some irrelevant matter.¹⁰

• Appeals against questions of law by way of case stated: the Court’s role is further restricted to answering the questions of law agreed to and submitted by a court below, coming to its own conclusion.¹¹

One might think that this guidance makes the responsibilities of an appellate court easy to discharge: decide what type of appeal lies before it, then apply the appropriate guidance as detailed above. However, this guidance is for nought if a court has difficulty in distinguishing one type of appeal from another or difficulty in actually applying the guidance. These two difficulties are discussed below.

Problems of classification

The first concern is the potential difficulty in determining what type of appeal lies before an appellate court. Sometimes, this potential is not realised. For example, a case-stated appeal and a general appeal are easily distinguishable from one another because there is a different procedure

⁸ Austin, Nichols, above n 1, at [4].

^{9 May v May (1982) 1 NZFLR 165 (CA) at 170.}

^{10 Countdown Properties (Northland) Ltd v Dunedin City Council [1994] NZRMA}

^{145 (HC) at 153.}

^{11 See, eg, the provisions in Part 4 of the Summary Proceedings Act 1957.}

for bringing each to Court and the notices of proceeding will indicate the type of appeal.

However, there is no such procedural distinction between a general appeal and an appeal against the exercise of discretion. Similarly, the distinction between a question of law and a question of fact is not clear. It is not necessarily discernible from simply looking at the notice of proceeding, because an appeal may involve decisions of fact, law and discretion all at once. Given the Court’s role will significantly differ depending upon what type of decision it is examining, this classification question becomes fundamentally important.

Discretionary decisions compared to non-discretionary decisions

The distinction between an exercise of discretion and a determination of fact and law is significant, because the Court must intervene and come to its own conclusion on issues of law and fact, but it must not intervene on issues of discretion, save in egregious circumstances. However, it is a distinction beset by confusion, acknowledged by the Supreme Court as “not altogether easy to describe in the abstract.”¹² Statutory provisions provide little assistance, and although some will indicate that a decision is discretionary¹³ or instruct the Court to treat any appeal as if it were an appeal against the exercise of discretion, these are rare instances of clear legislative direction.¹⁴ Instead, it is usually the Court that must determine whether the decision before it is discretionary.

This is no easy task, a fact brought into sharp relief by the recent judgment of Collins J in Roberts v A Professional Conduct Committee of the Nursing Council of New Zealand,¹⁵ delivered almost exactly five years after the Supreme Court’s delivery of Austin, Nichols. Roberts involved appeal against a decision of the Nursing Council to suspend Mr Roberts’ registration as a nurse for three years. Mr Roberts appealed on the basis that this was an excessive and unreasonable penalty. The preliminary question before the Court prior to its consideration of this argument was whether the decision by the Nursing Council on penalty was an exercise of discretion. As discussed above, this preliminary question was important because it would determine whether the Court had to come to its own conclusion on the appropriate penalty or instead defer to the Council’s decision.

This preliminary question also took on particular importance because both counsel and the Court were confronted by conflicting jurisprudence on the issue. Justice Collins cited six decisions that had applied Austin, Nichols, and three that applied May v May in the same situation; ie the

¹² Kacem v Bashir [2010] NZSC 112, [2011] 2 NZLR 1 at [32].

^{13 See, eg, s 14 of the Bail Act 2000, which deems the grant of bail pending}

^{sentencing an exercise of discretion.}

^{14 See, eg, s 18(4) of the Broadcasting Act 1989: “The Court shall hear and}

^{determine the appeal as if the decision or order appealed against had}

^{been made in the exercise of a discretion.”}

^{15 Roberts, above n 3.}

jurisprudence appeared split as to whether a disciplinary tribunal’s decision on penalty is an exercise of discretion.¹⁶ Such is the difficulty of the distinction between discretionary and non-discretionary decisions that two completely opposing approaches had persisted in the High Court since the Supreme Court’s decision in Austin, Nichols.

In Roberts, Collins J attempted to break this deadlock. He provided classical instances of discretionary decisions – decision on whether to grant bail or name suppression – before deciding that the determination of a penalty is a discretionary decision. His rationale for doing so was as follows:¹⁷

The process of evaluating penalty options and deciding what penalty to impose involved an exercise of discretion by the Tribunal in the same way that a decision about bail or name suppression also involves the exercise of discretion by judicial officers. All involve the careful evaluation of options and the choosing of the most suitable option that is available. In this respect, the Tribunal’s penalty decision can be distinguished from its role when interpreting the law, deciding facts and/or applying the law to established facts when determining if a practitioner has committed a disciplinary offence.

^{Justice Collins was faced with conflicting decisions because there is no}

^{one accepted method of determining whether a decision is discretionary.18}

^{One definition holds that “[d]iscretion arises when an official is}

^{empowered to exercise public authority and afforded scope to decide}

^{how that authority should be exercised in particular circumstances.”19}

^{However, this general definition does not assist, for it encompasses far}

^{too many decisions hitherto acknowledged by the Court as not being}

^{discretionary.}

To wit: the Commissioner of Trade Marks, when considering a proposed trade mark, “shall, after hearing the parties, if so required, and considering the evidence, decide whether, and subject to what conditions or limitations, if any, registration is to be permitted.”²⁰ One could expect the broad powers given to the Commissioner by this provision affords him scope as to how to decide whether he registers a trade mark, thus making his decision “discretionary” under the definition above. Instead, the decision-making power of the Commissioner is almost the paragon of non-discretionary making power, given it was the basis of Austin, Nichols.

^{Like most definitions of discretion, both Collins J’s approach in Roberts}

^{and the general definition focus on “choice”.21 Instead of one right}

¹⁶ At [24]–[35].

^{17 At [43].}

^{18 David Feldman “Discretions, choices and values” [1994] PL 279 at 288.}

^{19 Anna Pratt and Lome Sossin “The Dilemmas of Discretion: A Brief}

^{Introduction of the Puzzle of Discretion” (2009) 24 Can JL & Soc’y 301 at}

^301.

^{20 Trade Marks Act 2002, s 27(5).}

^{21 John Bell “Discretionary Decision-Making: A Jurisprudential View” in}

^{Keith Hawkins (ed) The Uses of Discretion (Clarendon Press, Oxford, 1992)}

answer, when making a discretionary decision, the decision-maker is faced with several, equally valid answers from which she can choose. Instead of following a mandated process to arrive at this answer, the decision-maker exercises her discretion to choose the answer she thinks fits best. The key element here is that this exercise of discretion is an irreplicable decision-making process. There may be factors that guide the discretion, but there is no way for an appellate court to replicate the process and thus assess its validity. The inability of the Court to do so provides a practical reason why it will only intervene in situations where the answer arrived at by the decision-maker – her choice – is clearly beyond the limits of her discretion.

However, the problem that the Court encounters when applying

^{general definitions – and thus why there is a problem with classification}

^{– is that the “discretionary” versus “non-discretionary” distinction is}

^{not a binary one. Instead, there is a spectrum of discretionary decision-}

^{making: some is close to unfettered, while other decision-making is}

^{severely constrained by statute and factual circumstances. Just as the}

^{size of discretion waxes and wanes, so too will the Court’s ability to}

^{replicate the decision-making process, reducing the rationale for its}

^{non-intervention.}

Such variability in the degree of discretionary decision-making was recently demonstrated by the Supreme Court in Kacem v Bashir when it held that “the fact that the case involves factual evaluation and a value judgment does not of itself mean the decision is discretionary.”²² This was despite House of Lords precedent describing ‘value judgments’ as the very manifestation of discretionary power and “an acceptable price to pay for the flexibility of ... discretion”.²³ If an element of discretion does not have the ability to make a decision a discretionary one, then a firm dividing line between discretionary decisions and non-discretionary decisions is almost impossible to draw. The problem is that the guidance for appellate courts nevertheless requires them to do so. A binary theory of discretion is undermined by the reality that discretionary power lies on a spectrum.

Questions of fact and questions of law

The distinction between questions of law and questions of fact is similarly difficult to draw. As William Wade observed: “No two terms of legal science have rendered better service than ‘law’ and ‘fact.’ They are the creations of centuries. What judge has not found refuge in them? The man who could succeed in defining them would be a public enemy.”²⁴

The Supreme Court of Canada has possibly had the most success in

93.

^{22 Kacem v Bashir, above n 12, at [32].}

^{23 Piglowska v Piglowski [1999] UKHL 27; [1999] 1 WLR 1360 (HL) at 1373 per Lord Hoffmann.}

^{24 HWR Wade Administrative Law (5th ed, Clarendon Press, Oxford, 1982),}

^{cited in Harry Woolf “A hotchpotch of appeals: the need for a blender”}

^{(1988) 7 CJQ 44 at 52.}

^{defining questions of fact, law and mixed fact and law, at least in broad}

^terms:25

Briefly stated, questions of law are questions about what the correct legal test is; questions of fact are questions about what actually took place between the parties; and questions of mixed law and fact are questions about whether the facts satisfy the legal tests.

However, despite this definition being accepted in New Zealand, it is not one that is reflective of reality.²⁶ For example, in The Wanaka Gym Ltd v Queenstown Lakes District Council,²⁷ the Court was asked to consider the question of “whether, on all the evidence, the proper classification for [a] dwelling was as a single household or residential unit”²⁸ was a question of law. Justice Lang held that it was not a question of law, and that a “factual issue is not converted into an issue of mixed fact and law merely because it produces legal consequences.”²⁹ He held further that since the applicant was seeking to challenge the conclusion reached by the judge below as to whether the dwelling was a single household unit, rather than the test itself, the proposed question was simply one of fact and degree, and not a question of law. However, application of a test to particular facts is the classical issue that amounts to a question of mixed fact and law under the Canadian definition above.

This case shows that broad distinctions do not work in practice when an issue has the potential to be both factual or legal depending upon whether one looks at it through a lens of existence or application. It is this problem of classification that leads to the same problem that occurs with discretionary decisions: when so much depends upon how an issue is classified, is it satisfactory to have such amorphous definitions?

Problems of application

Assuming an appellate court navigates its way through the difficult process of classifying a proceeding, it is then confronted by a further problem: how to approach the decision on appeal. As indicated earlier, there is no shortage of guidance on what the Court ought to do. The difficulty is how to follow that guidance.

This problem of application becomes clear once the guidance is examined, and the centrepiece of that guidance – the Supreme Court’s decision in Austin, Nichols – is a good place to start. Austin, Nichols stands for the proposition that no presumptive deference to the original decision-maker is required in general appeals. This guidance is not

²⁵ Canada (Director of Investigations and Research) v Southam Inc [1997] 1 SCR

^{748, 144 DLR (4th) 1 at [35].}

^{26 Accepted in Nixon v Walker HC Auckland CIV-2007-404-1372, 13 July 2007}

^{at [25].}

^{27 The Wanaka Gym Ltd v Queenstown Lakes District Council [2012] NZHC}

^2662.

^{28 At [13].}

^{29 At [22].}

equivalent, however, to requiring an appellate court to disregard the original decision, for the Court will still receive and consider the original decision before coming to its own conclusion.

What the Supreme Court requires is, in other words, a second opinion: an independent assessment of and conclusion on the same facts that were before the original decision-maker. This requirement, however, ignores the basic psychological constraints on a second decision-maker ’s ability to arrive at an independent conclusion having already considered the first decision-maker’s conclusion. Phenomena such as “information cascades” and “reputational cascades” impair a second decision-maker ’s ability to arrive at an independent conclusion, with that second decision-maker free-riding on or being influenced by the information and expertise of the decision-maker below.³⁰ Put simply, it is difficult to imagine noting someone else’s conclusion on a matter but then not in some way being influenced by it.

The Supreme Court acknowledged the inevitability of this influence in Kacem v Bashir, its most thorough consideration of Austin, Nichols to date. In confirming the Court of Appeal’s chastisement of the High Court’s approach to Austin, Nichols, Tipping J for the majority observed the following:³¹

The Court of Appeal was right to say that Courtney J had rather overstated the effect of Austin, Nichols when she indicated she should approach the appeal to the High Court “uninfluenced” by the reasoning of the Family Court. The High Court was required to reach its own conclusion, but this did not imply that it should disregard the Family Court’s decision. What, if any, influence the Family Court’s reasoning should have was for the High Court’s assessment.

However, although the Supreme Court acknowledged the near psychological impossibility of the Court remaining uninfluenced by the decision on appeal, it gave no guidance as to how appellate courts should determine the appropriate amount of influence. Appellate courts must undertake the near impossible task of striking the balance between deference and disregard unassisted.

Similarly, the guidance given to appellate courts on how to approach discretionary decisions ignores practical realities. May v May instructs the Court not to intervene in such decisions unless they are, for example, “plainly wrong”. Notwithstanding the questionable distinction between “wrong” and “plainly wrong”,³² the latter term has come to mean a decision that no reasonable decision-maker would make³³ or one that is outside the permissible boundaries of that decision-maker ’s discretion.³⁴

30. Adrian Vermeule “Second Opinions and Institutional Design” 97 Va L Rev 1435 at 1455.

^{31 Kacem v Bashir, above n 12, at [31].}

^{32 Roberts, above n 3, at [42].}

^{33 Shiloh Spinners Ltd v Harding (1973) 25 P & CR 48 (HL) at 63 per Lord}

^{Simon of Glaisdale.}

^{34 Te Wini v R [2011] NZCA 617, citing G v G [1985] 1 WLR 647 (HL) at 652}

This is, of course, simply a redux of the test under the Wednesbury unreasonableness ground in judicial review, derided by the House of Lords as a “tautologous formula”³⁵ and for which criticism is de rigeur in legal circles.³⁶ The fundamental problem with such a standard is that an appellate court defines the permissible boundaries of a discretion, thereby allowing itself the ability to justify intervention if and when it sees fit, rather than working to an objective standard.

Any theoretical purity in the guidance given to a court on how to exercise its appellate jurisdiction is tainted when put into practice, either by the limits of psychology when it must come to its own conclusion or the dangers of subjectivity when it must assess the validity of others’ conclusions.

The underlying issue: calibrating deference

Problems of classification and application belie a more fundamental problem: the misapplication of the concept of deference. One theme in the guidance given to appellate courts is that it sometimes requires deference to decision-makers’ conclusions; other times not. Occasionally, this theme is made explicit, as per the Supreme Court in Austin, Nichols:³⁷

The appeal court must be persuaded that the decision is wrong, but in reaching that view no “deference” is required beyond the “customary” caution appropriate when seeing the witnesses provides an advantage because credibility is important.

In essence, the Supreme Court has said: “no deference”. In other cases, say, involving discretionary decisions, guidance has amounted to: “full deference”. The guidance on New Zealand’s appellate jurisdiction, when viewed in totality, is essentially a binary of either full or no deference depending upon the type of appeal.

The problem is that this is not how deference works. Aileen Kavanagh provides a good outline of what deference means in public law: “judicial deference occurs when judges assign varying degrees of weight to the judgments of the elected branches, out of respect for their superior expertise, competence or democratic legitimacy.”³⁸ As Kavanagh notes, the notable features about this definition is that deference is a variable concept and is partial rather than absolute.³⁹ As with the concept of discretion, discussed above, rather than deference being a binary, all- or-nothing affair, the degree of deference given by the Court will vary with the circumstances. It is less a traditional light switch than it is a

per Lord Fraser.

^{35 R v Chief Constable of Sussex ex parte International Trader’s Ferry Ltd [1999]}

^{2 AC 418 (HL) at 452 per Lord Cooke of Thorndon.}

^{36 Paul Daly “Wednesbury’s reason and structure” [2011] PL 238 at 238.}

^{37 Austin, Nichols, above n 1, at [13] (citations omitted).}

^{38 Aileen Kavanagh “Defending deference in public law and constitutional}

^{theory” (2010) 126 LQR 222 at 223.}

^{39 At 224.}

modern dimmer dial. Moreover, a corollary to deference being a concept that will vary with the circumstances is that “deference does not apply automatically with reference to certain subject-matters.”⁴⁰ The inquiry into how much deference an appellate court must give to a decision- maker is determined on a case-by-case basis, using a series of indicia that measure the relative expertise of the Court and the decision-maker.⁴¹

This means that despite deference underpinning the structure of New Zealand’s appellate jurisdictions, that structure is at odds with the nature of deference. New Zealand’s appellate jurisdictions constrain deference to an absolutist, binary concept that is dependent upon the nature of the proceeding, ie, everything that deference is not. It is this dissonance between structure and concept that exacerbates the problems of classification and application outlined above.

Pre-determined levels of deference that accompany each type of appeal makes the classification question far more significant than it needs to be. If an appellate court is confronted by a decision in a general appeal upon which the Court feels the decision-maker had a great degree of expertise, competence and mandate to adjudicate, then its options are either to follow Austin, Nichols and ignore its instincts for deference or reframe the appeal as one against an exercise of discretion. Such was the issue E v Director of Proceedings,⁴² which like Roberts, dealt with an appeal under the Health Practitioners Competence Assurance Act 2003. Justice Ronald Young held that the degree of deference hitherto given to the Health Practitioners’ Disciplinary Tribunal was no longer appropriate in the wake of Austin, Nichols given that appeals against its decisions were general appeals. However, he still identified some situations where deference was appropriate given some discretionary aspects of the Tribunal’s decision-making.⁴³ In all, both the judgments of Roberts and E v Director of Proceedings had one quarter of their reasons dedicated to resolving this thorny classification question, indicating just how much this preliminary issue unjustifiably dominates proceedings.⁴⁴

Similarly, it is the nature of deference as a variable concept that causes problems of application. When the Court is confronted by circumstances that justify deference but is restricted by guidance that prevents any being afforded, it makes it even more difficult for the Court to ignore the persuasiveness of the decision-maker ’s reasoning. Perhaps this is why Ronald Young J ignored the Supreme Court’s dictum in Austin, Nichols by holding that:⁴⁵

Deference to the Tribunal’s decision may be appropriate where the Tribunal has a particular advantage such as medical expertise or an assessment of credibility of witnesses.

⁴⁰ At 226.

^{41 At 226.}

^{42 E v Director of Proceedings [2008] NZHC 879; (2008) 18 PRNZ 1003 (HC).}

^{43 At [22].}

^{44 21 out of 78 and 15 of 58 paragraphs respectively.}

^{45 E v Director of Proceedings, above n 42, at [21].}

Justice Ronald Young’s approach reflected the normal and variable nature of deference. However, in doing so, he did not necessarily follow the guidance of the Supreme Court. This is the effect of trying to place a flexible concept into hard categories.

A potential path for reform

The problems with New Zealand’s appellate jurisdiction above are not simply theoretical or academic. The confusion surrounding the classification of proceedings and application of guidance has led to frustrated counsel,⁴⁶ contrary approaches, and importantly, appealable errors by the Court, including arguments that it:

• gave an insufficient amount of deference given the circumstances;⁴⁷

• incorrectly determined that a decision was discretionary;⁴⁸ and

• unsuccessfully selected or applied the appropriate guidance for the particular appeal.⁴⁹

As long as the underlying problem that afflicts the structure of appellate jurisdictions persists, time and resources will unnecessarily be wasted in the classification and application process and appeals against errors in that process. Thus, that problem – the dissonance between the nature of deference and how it is accommodated in the appellate structure – is a prompt for reform.

Successful reform to assuage this problem does not require drastic legislative change, for its locus is in the common law. Statutory provisions providing the rights of appeal are certainly not blameless, given the guidance they do not provide. However, as is hopefully evident from the discussion above, the critical issues involving distinctions between decisions of fact, law and discretion are less creatures of statute than they are of the courts.

Austin, Nichols swept away a relatively settled – although not altogether justified – approach to appellate procedure, which essentially saw appellate courts stand back from questions of fact and discretion and only examine questions of law.⁵⁰ Austin, Nichols “meant that the thinking regarding appellate intervention changed fundamentally. As a result, all courts were compelled to grapple with the whole of the dispute between the parties in order to come to a proper decision.”⁵¹ The problem was

⁴⁶ Acknowledged by Collins J in Roberts, above n 3, at [22].

^{47 Fassler v Parry [2012] NZCA 327 at [7]; Bashir v Kacem [2010] NZCA 96; [2010] NZFLR 865}

^{(CA) at [39].}

^{48 Glynbrook 2001 Ltd v Official Assignee [2012] NZCA 289 at [87]; Blackstone}

^{v Blackstone [2008] NZCA 312; (2008) 19 PRNZ 40 (CA) at [8]; and Wildbore v Accident}

^{Compensation Corporation [2009] NZCA 34; [2009] 3 NZLR 21 (CA).}

^{49 L v K [2010] NZCA 618; (2010) 28 FRNZ 692 (CA) at [18]–[22]; and Rowley v Commissioner of}

^{Inland Revenue [2011] NZCA 160 at [12]–[20].}

^{50 Andrew Beck “Farewell to the Forum Otiosum?” (2011) NZLJ 269 at}

^269–270.

^{51 At 271.}

that such compulsion was unaccompanied by any guidance on how to grapple with the decision, leading to the confusion outlined above, and the Supreme Court has appeared reluctant to provide that further guidance.

Accordingly, this paper proposes two aspects of reform:

• Eliminating illusory distinctions: questions of fact and law shade into one another. There is always arguably some modicum of discretion in all decisions. By eliminating the distinction between questions of fact, law and discretion, and accepting that there are degrees of each in every decision, the problems of classification would all but disappear.

• Accepting a singular approach to all appeals: the corollary of eliminating the distinctions between different types of appeals is to apply the same approach to all appeals. This would eliminate the confusion suffered by both judge and practitioner as to which approach is appropriate and any potential to exploit that confusion.

Of course, the immediate objection to homogenising the approach to all appeals is that no singular approach is appropriate. Removing the distinction between the different types of appeals does not mean that questions of fact, law and discretion would disappear, and no one approach can successfully accommodate those different questions that can arise on appeal and varying levels of expertise that the Court possesses to deal with such questions. The response to this objection is that so long as the approach is context-dependent and variable, differentiations in the expertise between Court and decision-maker will not present an issue.

There is insufficient room in this comment to give a comprehensive account of what such an approach would look like, but there are existing models that New Zealand could follow. For example, the Supreme Court of Canada – acknowledging that deference is a variable concept

– accepted the notion of “deference as respect” first mooted by David Dyzenhaus.⁵² The key aspect of that model is that it views deference as a process, requiring respectful attention by the Court to a decision-maker ’s reasons, and for the decision-maker to adequately justify her decision.⁵³

Such a process-based model is a reflection of Kavanagh’s definition of deference described above. Accordingly, it is more effective at adjusting to the context and nature of a decision and critically, does not depend on pre-determined categories and levels of deference.

The proposed reforms are not a reversion to the pre–Austin, Nichols jurisprudence, termed by Beck as “a clear abdication of the

⁵² David Dyzenhaus “The Politics of Deference: Judicial Review and Democracy” in Michael Taggart (ed) The Province of Administrative Law (Hart, Oxford, 1997) 279, accepted in Baker v Canada (Minister of Citizenship

^{& Immigration) [1999] 2 SCR 817 at [65].}

^{53 Dyzenhaus, above n 52, at 286.}

responsibilities resting on an appellate court”.⁵⁴ Indeed, there would be even more engagement by appellate courts, for they would scrutinise each decision for adequate justification. In any case, no longer would the Court be able to inappropriately step back from or intervene in a decision simply because it is contained in a particular type of appeal as per the status quo.

Conclusion

A plethora of different appeals and incomplete guidance from the Supreme Court has led to confusion in lower courts on how to classify appeals and apply that guidance. These problems of classification and application are problems ultimately stemming from a paradigm that requires appellate courts to calibrate the level of deference it gives to decision-makers depending on the type of appeal.

If deference is best understood as a process, presumptive levels of deference indicated by fixed categories of appeal is the wrong way to approach it. To truly improve the guidance given to appellate courts on how to approach appeals, the Supreme Court must remove unnecessary distinctions between different types of appeals and should prescribe one singular approach. That would allow appellate courts and their inferior decision-makers to truly engage with one another, rather than spending their time classifying appeals, applying guidance and getting it wrong.

⁵⁴ Beck, above n 50, at 270.