|
|
Home
| Databases
| WorldLII
| Search
| Feedback
University of Otago Law Theses and Dissertations |
Last Updated: 21 September 2023
CALIBRATING THE REASONABLENESS TEST OF JUDICIAL REVIEW IN A NEW ZEALAND CONTEXT
CJ SEDDON
A dissertation submitted in partial fulfilment of the degree of Batchelor of Laws (Honours) at the University of Otago – Te Whare Wānanga o Otāgo
Acknowledgements
To Marcelo Rodriguez Ferrere, whose insight, comprehensive constructive criticism and intimidating depth of knowledge as a supervisor made this dissertation possible;
to Elizabeth Yarnall, for her endless support, even at my crankiest, and for her dear and unwavering friendship;
to Nic Wilson, whose research provided the catalyst for this dissertation and who I very much hope will one day realise he is wrong about reasonableness review;
to Joe Garry, for his proofreading and for sending me every article he found even remotely related to reasonableness review;
to Dushanka Govender, for her constant positivity, her style advice, her friendship, and who, I’m sure, definitely thought she was going to be left out of another acknowledgement.
Table of Contents
Chapter I: Introduction pg. 1
Chapter II: Substantive Review in the Status Quo pg. 3
Chapter III: Contextual Review as an Alternative pg. 13
Chapter IV: Developing a Model for Calibrating Unreasonableness pg. 24
Two-Stage Bifurcated Model pg. 31
Chapter V: Justifying the Two-Stage Bifurcated Model pg. 32
Chapter VI: Conclusion pg. 45
Chapter I: Introduction
Reasonableness review in New Zealand is a vexed issue, and one rightly feared by every Administrative Law class for its complexity, nebulousness, and utter lack of consensus across New Zealand courts. The problem is simple to grasp but difficult to solve. There is a substantial body of New Zealand case law indicating that, while Wednesbury may still be applicable in certain discrete situations, it is no longer the monolithic standard of reasonableness that it used to be.1 The demise of Wednesbury as a single unitary standard has ushered in uncertainty to the ground of reasonableness review, with the High Court preferring an approach to review that structures discretion2, and the Supreme Court preferring an approach that leaves discretion unfettered.3 The marked difference in approach has left the High Court without clear guidance, as the Appellate Courts have refused to address the question of how the reasonableness test is to be calibrated. That lack of guidance translates in confusion, not only for the judges of the High Court, but also for litigants and legal commentators.
This paper seeks to propose a solution to the problem of confusion in the status quo of reasonableness review, in the form of a model for calibrating the reasonableness test. That model consists of a two-step approach, first ascertaining whether the decision in question involves a degree of rights-infringement, and second determining the depth of scrutiny based on that content. The paper suggests a “demonstrably justifiable” proportionality test for those decisions involving rights and a “Wednesbury unreasonableness” test for those that do not, similar to the rainbow of review proposed by Michael Taggart.4
In reaching that solution, this paper will be separated into four substantive chapters. The first will be an analysis of substantive review in the status quo; the High Court approach to substantive review; and the lack of guidance by the Appellate Courts. This chapter seeks to identify the problem this paper hopes to solve, namely the confusion in the status quo stemming from the lack of guidance. The second chapter will be a comparison of variable
intensity of review and contextual review using Fuller’s Principles of Efficacy and Virtue. This second chapter looks to resolve the difference in approach between the High Court and
1 Michael Taggart "Proportionality, deference, Wednesbury" (2008) 2008(3) NZ L Rev 423 at 430.
2 Dean R Knight “Mapping the Rainbow of Review” (2010) NZ Law Rev 393 at 408.
3 MB Rodriguez Ferrere “An Impasse in New Zealand Administrative Law: How did we get here?” (2017) 28 PLR 310 at 326.
4 Taggart, above n 1, at 430.
the Supreme Court in favour of variable intensity of review. The third will be an explanation of the model proposed as a solution to the problem in the status quo, working through the two stages. The final substantive chapter will be a justification of the model, both addressing itsinternal coherency and comparing it to the alternative of contextual review using Fuller’s principles.
Finally, it is useful to briefly touch on what this paper is not. This paper does not address the question of whether reasonableness is useful as a test at all. The reasonableness standard of review is firmly established as a ground of review in New Zealand case law.5 Nor will this paper address the question of whether reasonableness should be replaced by grounds such as “the innominate ground” or “cumulative impropriety”. While there has been a suggestion of alternative grounds in recent years,6 New Zealand administrative case law very much acknowledges reasonableness as the primary ground of substantive review.
5 See for example Kim v Minister of Justice of New Zealand [2019] NZCA 209; Deliu v Connell and Ors [2016] NZHC 361; Watson v Chief Executive of Department of Corrections [2015] NZHC 1227, [2015] NZAR 1049.
6 AI (Somalia) v The Immigration and Protection Tribunal & Anor [2016] NZHC 2227.
Chapter II: Substantive Review in the Status Quo
The first stage of discussion is identifying the problem in substantive review that the model seeks to solve. The problem is simple; the status quo of substantive review for New Zealand, in this case reasonableness review, is that there has been a shift away from the monolithic “Wednesbury unreasonableness”7 standard that has been broadly accepted by both the High Court and the Appellate Courts.8 The High Court has opted for an approach known as variable intensity of review9, the Court of Appeal has been consistently vague as to the approach to be applied10, and the Supreme Court has opted for an alternative approach known as contextual review.11 That confusion has posed the primary question of which of those two options of review is preferable, but additionally has left the High Court directionless as to how to conduct reasonableness review in the first instance.This chapter will analyse that problem in three parts. Firstly, the background of substantive review; what it entails in principle, the tension inherent in this form of review, and how the courts have historically resolved that tension from Wednesbury12 onwards. Secondly, the New Zealand High Court’s current approach to reasonableness review through the case studies of Wolf 13, Watson14 and Deliu15 as an articulation of the current confusion in the High Court. Thirdly, the New Zealand Appellate Court’s approach to reasonableness review; the lack of engagement in the Court of Appeal, and the rejection of variable intensity in favour of contextual review in the Supreme Court. Overall, the chapter will set the scene for the two questions this paper must answer; a justification of variable intensity as opposed to contextual review, and a justification of the model as a means of reducing the confusion in the status quo.
7 Associated Provincial Picture Houses Ltd v Wednesbury Corp [1947] EWCA Civ 1; [1948] 1 KB 223 (CA) at 229.
8 Taggart, above n 1, at 430.
9 Knight, above n 2, at 408.
10 Ferrere, above n 3, at 326.
11 Ferrere, above n 3, at 326.
12 Associated Provincial Picture Houses Ltd v Wednesbury Corp, above n 7, at 229.
13 Wolf v Minister of Immigration [2004] NZAR 414.
14 Watson v The Chief Executive of the Department of Corrections [2016] NZHC 1996.
15 Deliu v Connell and Ors [2016] NZHC 361.
The principled basis for the courts to conduct procedural review is clear: the courts have a role in ensuring that administrative bodies abide by the procedures set in place by Parliament to constrain the way in which they make decisions.16 Substantive review is distinct from procedural review in that the courts do not just look to the procedure through which the decision was made, but the substance of the decision itself.17 Judicial review forms the court’s supervisory jurisdiction and the basis of substantive review as a subset of that jurisdiction is that decision-makers are not entitled to unfettered discretion in making decision.18 Underlying any form of judicial review is the principle of the rule of law, that the Executive and related administrative bodies are bound by the law and so the courts have the jurisdiction to enforce the law against them.19In contrast, substantive review creates a tension between the courts’ ability to uphold the rule of law and the principle of the separation of powers: that the roles of the branches of government are distinct and ought to remain so.20 The courts have a role in ensuring that an administrative body is authorised in the decision that it makes and that role will involve the courts examining the substance of a decision to determine whether it has been authorised by Parliament.21 In the case of review on the ground of unreasonableness, there is no clear direction provided by Parliament as to what has been authorised; the review is guided by the concept that there is an acceptable spectrum of decisions and those that fall outside that spectrum have not been authorised.22 When conducting that examination, the courts ought not to look to the merits of a decision because that is outside the competency and jurisdiction of the judiciary.23 Accordingly, a tension emerges in unreasonableness review because in conducting substantive review the court runs the risk of examining a case on the merits of the decision.24
16 Graham Taylor Judicial Review: A New Zealand Perspective (4th ed, LexisNexis, Wellington, 2018) at 3. 17 Dean R Knight “Judicial Review: Practical Lessons, Insights and Forecasts” (2010) VUW-NZCPL002 at 4. 18 DE Paterson An Introduction to Administrative Law in New Zealand (1st ed, Sweet & Maxwell Ltd, Wellington, 1967) at 11.
19 Paterson, above n 18, at 17.
20 Philip A Joseph Constitutional and Administrative Law in New Zealand (4th ed, Brookers Ltd, Wellington, 2014) at 997.
21 Sian Elias “Judicial Review and Constitutional Balance” (Address to the New Zealand Centre for Public Law held at Lecture Theatre 1 Victoria University of Wellington, 28 February 2019).
22 EI Sykes, DJ Lanham, RRS Tracey and KW Esser General Principles of Administrative Law (4th ed, Butterworths, Adelaide, 1997) at 127.
23 Dean R Knight “Simple, Fair and Discretionary Administrative Law” (2008) 39 VUWLR at 109.
24 Joseph, above n 20, at 997.
Historically, courts resolved that tension in unreasonableness review through setting the bar for an unreasonable decision very high. The courts make it clear that they are concerned solely with the legality of the decision in question.25 So, in Associated Provincial Picture Houses Ltd v Wednesbury Corp26 Lord Greene MR set the standard that would become known as “Wednesbury unreasonableness” as a decision that is “so absurd that no sensible person could ever dream that it lay within the powers of the authority”27. In identifying the three primary grounds of judicial review in Council of Civil Service Unions v Minister for the Civil Service (CCSU)28 Lord Diplock articulated that standard as a decision “So outrageous in its defiance of logic or accepted moral standards that no sensible person who had applied his mind to the question to be decided could have arrived at it”.29 English cases applying the standard of Wednesbury unreasonableness following CCSU made it clear that the exercise in substantive review was justified on the basis that the decision was perverse or so absurd as to indicate that the decision-maker had “taken leave of their senses”.30 Setting such a high bar for the standard of review resolved the tension because it restricts the courts from engaging in merit-based review. A high bar leaves no room for a judge to substitute an administrative decision-maker’s view for their own.31Since the late 90s, in the United Kingdom and New Zealand, that bar has been lowered somewhat. That development is the consequence of the Wednesbury unreasonableness standard being practically unattainable in a form that consistently protected rights.32 In a practical context administrative decision-makers may make what would seem to be wrong in the eyes of the court, but very rarely do they act in such a way that throws reason to the wind.33 United Kingdom courts were faced with decisions that seemed to be unjust but were not captured under Wednesbury unreasonableness and so lowered the standard to intervene in decisions that were less obviously absurd.34 That shift in the courts’ approach to unreasonableness review took the form of a closer examination of decisions impacting human rights, while still maintaining that, at its heart, it was still a review of the scope of
25 Robin Cooke "The Struggle for Simplicity in Administrative Law" in Michael Taggart (ed) Judicial Review of Administrative Action in the 1980s – Problems and Prospects (Oxford University Press, Auckland, 1986) 1.
26 Associated Provincial Picture Houses Ltd v Wednesbury Corp, above n 7, at 229.
27 Associated Provincial Picture Houses Ltd v Wednesbury Corp, above n 7, at 229.
28 Council of Civil Service Unions v Minister for the Civil Service [1985] AC 374 (HL).
29 Council of Civil Service Unions v Minister for the Civil Service, above n 28, at 410.
30 Nottinghamshire County Council v Secretary of State for the Environment [1985] UKHL 8; [1986] 1 AC 240 (HL) at 247.
31 Joseph, above n 20, at 997.
32 Taggart, above n 1, at 426.
33 M Taggart Judicial Review of Administrative Action in the 1980s (1st ed, Oxford University Press, Auckland, 1986) at 14.
34 Joseph, above n 20, at 998.
authorisation granted to a decision-maker. In the United Kingdom case of R (Bugdaycay) v Secretary of State for the Home Department35 Lord Bridge set the groundwork for a more variegated test using the language of anxious scrutiny: “when an administrative decision [...] is said to be one which may put the applicants life at risk, the basis of the decision must surely call for the most anxious scrutiny”.36 In R (Smith) v Ministry of Defence37 the court held that “the more substantial the interference with human rights the more the court will require by way of justification”38. Courts in New Zealand also recognised the limitations of “Wednesbury unreasonableness” and followed in the footsteps of the UK courts’ heightened
scrutiny approach.39. In Wolf v Minister of Immigration,40 Wild J explicitly endorsed a sliding
scale of review with reference to Lord Steyn’s misquote in R (Daly) v Secretary of State for the Home Department41 that “in administrative law, context is everything”.42 Overall, both the United Kingdom and New Zealand have lowered the bar on what constitutes an unreasonable decision in certain discrete situations.43
This paper will examine the application of intensity of review in four cases. Firstly, Wild J’s landmark adoption of an intensity of review approach to reasonableness review in Wolf v Minister of Immigration44. Second, the approach to intensity of review taken by Mallon J in Watson v Chief Executive of Department of Corrections45. Finally, the analysis by Palmer J in the High Court in the case of Deliu v Connell and Ors46. Overall, the analysis will indicate that the High Court of New Zealand has accepted intensity of review as the form of substantive review in questions of reasonableness but lacks guidance from the appellate courts in calibrating that approach.47
35 R (Bugdaycay) v Secretary of State for the Home Department [1986] UKHL 3; [1987] AC 514.
36 R (Bugdaycay) v Secretary of State for the Home Department, above n 35, at 531.
37 R (Smith) v Ministry of Defence [1995] EWCA Civ 22; [1996] QB 517.
38 R (Smith) v Ministry of Defence, above n 37, at 554.
39 Dean R Knight Vigilance and Restraint in the Common Law of Judicial Review (1st ed, Cambridge University Press, Cambridge, 2018) at 165.
40 Wolf v Minister of Immigration, above n 13.
41 R (Daly) v Secretary of State for the Home Department [2001] UKHL 26; [2001] 2 AC 532.
42 Wolf v Minister of Immigration, above n 13, at 47.
43 Taggart, above n 1, at 430.
44 Wolf v Minister of Immigration, above n 13.
45 Watson v The Chief Executive of the Department of Corrections, above n 14.
46 Deliu v Connell and Ors, above n 15.
47 Ferrere, above n 3, at 327.
In Wolf v Minister of Immigration48 Wild J articulates the shift in New Zealand case law from the single monolithic standard of Wednesbury unreasonableness to a variable intensity of review in certain cases.49 Wild J states “the tests laid down in GCHQ and Woolworths respectively are not, or should no longer be, the invariable or universal tests of “unreasonableness” applied in New Zealand public law”.50 He indicates a variable intensity of review approach in stating “whether a court considers a decision reasonable [...] depends on the nature of the decision: upon who made it; by what process; what the decision involves [...]; and the importance of the decision to those affected by it”.51 Wild J gives four reasons for the move away from Wednesbury unreasonableness. First, that Wednesbury unreasonableness is a product of administrative law in its infancy and fails to acknowledge the existence of the New Zealand Bill of Rights.52 Second, that variable intensity has been informally recognised in New Zealand administrative law for at least 20 years prior to Wolf.53 Wild J cites the cases of Electoral Commission v Cameron54, Pharmac v Roussel Uclaf Australia Pty Ltd55 and Ports of Auckland Ltd v Auckland City Council56, all of which make reference to a variable intensity of scrutiny depending on the content of the cases. Third, that most legal commentators describe intensity of review as variable in discussing reasonableness review.57 Fourth, that comparable common law jurisdictions such as the United Kingdom and Canada have accepted a variable intensity approach to reasonableness review.58 Wild J articulates the shift towards an intensity of review approach but does not provide the calibration for how that approach should be carried out.
In the case of Watson v Chief Executive of Department of Corrections59 Mallon J reaffirms the approach to reasonableness review adopted in Taylor v Chief Executive of Department of Corrections60, where Randerson, Harrison and Cooper JJ cited the judgement of Dunningham J in Watson v Chief Executive of Department of Corrections61. Mallon J notes that the degree
48 Wolf v Minister of Immigration, above n 13.
49 Dean R Knight “A Murky Methodology: Standards of Review in Administrative Law” (2008) 6 NZJPIL 117 at 130.
50 Wolf v Minister of Immigration, above n 13, at [47]. 51 Wolf v Minister of Immigration, above n 13, at [47]. 52 Wolf v Minister of Immigration, above n 13, at [48]. 53 Wolf v Minister of Immigration, above n 13, at [48].
54 Electoral Commission v Cameron [1997] NZCA 301; [1997] 2 NZLR 421.
55 Pharmaceutical Management Agency Ltd v Roussel Uclaf Australia Pty Ltd [1998] NZAR 58 (CA).
56 Ports of Auckland Ltd v Auckland City Council [1999] 1 NZLR 601.
57 Wolf v Minister of Immigration, above n 13, at [48].
58 Wolf v Minister of Immigration, above n 13, at [48].
59 Watson v The Chief Executive of the Department of Corrections, above n 14.
60 Taylor v Chief Executive of Department of Corrections [2015] NZCA 477.
61 Watson v Chief Executive of Department of Corrections [2015] NZHC 1227, [2015] NZAR 1049.
of deference afforded to Corrections in questions of reasonableness is that adopted in Taylor. She quotes Taylor in saying “where human rights are involved, prison authorities tend to be supervised intensively because they do not have special expertise of authority on rights”.62 She reaffirms the statement by Lord Steyn in Simms “the more substantial the interference with fundamental rights, the more the court will require justification before it can be satisfied the interference is reasonable in a public law sense”.63 The indication given by Mallon J in Watson is that the court ought to engage in a greater depth of scrutiny where there is an interference with fundamental rights, and that it is now settled law that Corrections ought to be subject to heightened scrutiny where human rights are involved.64 This provides useful clarification in the context of Corrections but little guidance in the general application of variable intensity of review.In the case of Deliu v Connell and Ors Palmer J considers substantive review in the context of irrationality, using the same language as that of reasonableness.65 The issue in this case was whether it was reasonable to question the basis of a complaint.66 Palmer J notes that the
issue “invites [Palmer J] to examine the substance of the conclusion rather than whether there was an error in one particular fact”.67 In doing so, Palmer J declines to make a statement of law regarding how to approach issues of substantive review, merely noting the availability of Wednesbury unreasonableness or “some other version of the threshold”.68 What Palmer J does specifically state is that, regardless of the form of review he employs in this case “the threshold for [him] to overturn a conclusion of the Legal Complaints Review Officer (LCRO) on the merits in a judicial review is high”.69 The case acknowledges the existence of thresholds for substantive review outside of Wednesbury unreasonableness and notes that, regardless of the threshold, the standard for overturning decisions is high, but does not go further in articulating how a variegated standard might be applied.
Overall the approach by the High Court is unstructured and confused. The indication from the High Court is that variable intensity of review is accepted practice in substantive review and
62 Watson v The Chief Executive of the Department of Corrections, above n 14, at [35]. 63 Watson v The Chief Executive of the Department of Corrections, above n 14, at [35]. 64 Watson v The Chief Executive of the Department of Corrections, above n 14, at [35]. 65 Deliu v Connell and Ors, above n 15, at [44].
66 Deliu v Connell and Ors, above n 15, at [44]. 67 Deliu v Connell and Ors, above n 15, at [42]. 68 Deliu v Connell and Ors, above n 15, at [45]. 69 Deliu v Connell and Ors, above n 15, at [45].
that Wednesbury unreasonableness is no longer the single monolithic standard.70 Cases like Wolf, Watson and Deliu also indicate that the court will engage in a greater depth of scrutiny where fundamental rights are at issue.71 Beyond that, there is no agreement in the High Court on a specific calibration of variable intensity of review. That means that the High Court has had to rely on the appellate courts to provide guidance on the factors that will determine depth and how to allocate depth against those factors.
The New Zealand appellate courts are currently reluctant to provide any degree of uniform guidance in how to calibrate a spectrum, or intensity, of review.72 The Court of Appeal has been willing to explore a number of versions of intensity of review, without providing one set standard for how substantive review ought to be conducted.73 The Supreme Court, and especially ex-Chief Justice Dame Sian Elias, has been hostile to concepts of intensity or deference and has instead favoured a simpler approach of contextual review that places discretion as to whether a decision is unreasonable in the hands of the judge.74 We will look at Court of Appeal and Supreme Court cases addressing the question, then extra-judicial commentary provided by ex-Chief Justice Elias.The approach in the Court of Appeal of New Zealand has been mixed, with judges either addressing the issue in obiter or refusing to address it at all.75 In the case of Lab Tests Auckland Ltd v Auckland District Health Board,76 Hammond J noted that the area of substantive judicial review is confused and undirected but expresses doubt at the concept of deference.77 In Tamil X v Refugee Status Appeals Authority,78 Baragwanath J conducted a comprehensive examination of intensity of review law with relation to refugees whereas his fellow judges Hammond and Arnold JJ avoided the question of substantive review altogether. In Ye v Minister of Immigration79 Glazebrook J addressed the question of intensity of review
70 John Laws "Wednesbury" in Christopher Forsyth and Ivan Hare (eds) The Golden Metwand and the Crooked Cord: Essays on Public Law in Honour of Sir William Wade QC (Oxford University Press, Oxford, 1998) 185. 71 Wolf v Minister of Immigration [2004] NZAR 414; Watson v The Chief Executive of the Department of Corrections [2016] NZHC 1996; Deliu v Connell and Ors [2016] NZHC 361.
72 Ferrere, above n 3, at 327.
73 Ferrere, above n 3, at 326.
74 Knight, above n 2, at 401.
75 Knight, above n 2, at 406.
76 Lab Tests Auckland Ltd v Auckland District Health Board [2008] NZCA 385, [2009] 1 NZLR 776.
77 Lab Tests Auckland Ltd v Auckland District Health Board, above n 77, at [374].
78 Tamil X v Refugee Status Appeals Authority [2009] NZCA 488, [2010] 2 NZLR 73.
79 Ye v Minister of Immigration [2008] NZCA 291, [2009] 2 NZLR 596.
in obiter but does not resolve the case on it, Hammond and Wilson JJ did not address the question at all and Chambers and Robertson JJ explicitly stated they held no opinion on standards of intensity as it did not arise.80In the case of Kim v Minister of Justice of New Zealand81 in the Court of Appeal, Cooper, Winkelmann and Williams JJ found the standard of review applied by the High Court to be uncontentious.82 They found that, when fundamental human rights are at stake, the appropriate standard of reasonableness review is one of heightened scrutiny.83 The Court of Appeal found that heightened scrutiny in this case required the court to “ensure the decision has been reached on sufficient evidence and has been justified, while recognising that Parliament has entrusted the Minister (and not the courts) to undertake adequate enquiries”.84 Moreover, the court in this case determined that heightened scrutiny required that the court consider whether materially relevant information had been considered by the Minister.85 The case is useful in being higher court authority that identifies human rights as a trigger for greater scrutiny, as well as defining what greater scrutiny entails in a particular context.
In the Supreme Court of New Zealand, the approach to reasonableness review is much more straightforward. The Supreme Court has rejected the notion of intensity of review and explicitly avoided discussing variegated standards of reasonableness. The justification is that a decision to interfere in a decision is inherently based on the context of the decision in question.86 Intensity of review and terminology such as “anxious scrutiny” or “hard look review” is, at best, unnecessary semantics and at worst hampers courts and detracts from the clarity of the law.87 In Ye v Minister of Immigration,88 the case did not involve any substantive discussion on intensity of review save for the statement by Tipping J that the best approach was to “concentrate on the essential points which must be addressed in order to resolve these particular cases”.89 The absence of direct discussion of intensity of review is made notable by the degree to which Glazebrook J in the Court of Appeal addressed the question of intensity in her judgments. Knight notes that a near absolute refusal to address the
80 Ye v Minister of Immigration, above n 79, at 569.
81 Kim v Minister of Justice of New Zealand [2019] NZCA 209. 82 Kim v Minister of Justice of New Zealand, above n 81, at [45]. 83 Kim v Minister of Justice of New Zealand, above n 81, at [45].
84 Kim v Minister of Justice [2017] NZHC 2109, [2017] 3 NZLR 823 at [7].
85 Kim v Minister of Justice of New Zealand, above n 81, at [46].
86 Elias, above n 21, at 4.
87 Knight, above n 2, at 401.
88 Ye v Minister of Immigration [2009] NZSC 76, 2010 NZLR 104.
89 Ye v Minister of Immigration, above n 88, at [11].
partial adoption of intensity of review by the lower courts ought to be considered to be “implicit disapproval”.90 Also notable is the discussion with counsel in Ye v Minister ofImmigration where Elias CJ states that “deference” is a “dreadful word” and holds that “what is reasonable takes its colour from the context”.91
In the earlier case of Discount Brands Ltd v Westfield (New Zealand) Ltd (Discount Brands),92 none of the Supreme Court address the approach to intensity of review taken by Hammond J in the Court of Appeal. Ex-Chief Justice Elias goes so far as to say that the questions were not “helpfully advanced by consideration of the scope and intensity of the High Court’s supervisory jurisdiction to ensure reasonableness in substantive result in the exercise of statutory powers”.93 In the discussion in Astrazeneca Ltd v Commerce Commission94 Elias CJ referred to the concept of a spectrum of unreasonableness as a “New Zealand perversion of recent years”.95 In other cases, despite pressure from the lower courts, the Supreme Court has remained firm in refusing to address the question of calibrating an approach to intensity of review.96 In the case of Lab Tests Auckland Ltd v Auckland District Health Board97 despite Hammond J explicitly notes the lack of uniform guidance provided by the higher courts, and the Supreme Court refuses to hear the appeal of the case, citing that there was no arguable question of public importance.98 While there may have been no question of public importance, this instance forms one in a number of instances of the Supreme Court refusing to provide guidance requested by the Court of Appeal.
Extrajudicially, Dame Sian Elias has spoken out against adopting the concepts of intensity or deference in the New Zealand approach to reasonableness review. She endorsed the Sir David Williams approach to contextualism, quoting from him in a 2008 public lecture to say “the application of the principles of administrative law can sensibly be considered only with the proper regard for the statutory, institutional and broader social or policy context of a particular case”.99 She criticises a doctrinal approach to administrative law and quotes Felix
90 Knight, above n 2, at 401.
91 Knight, above n 2, quoting Ye v Minister of Immigration (NZSC, transcript, 21-23, April 2009, SC 53/2008) at 179-182.
92 Discount Brands Ltd v Westfield (New Zealand) Ltd [2005] NZSC 17, [2005] 2 NZLR 597.
93 Discount Brands Ltd v Westfield (New Zealand) Ltd, above n 92, at [5].
94 Astrazeneca Ltd v Commerce Commission (NZSC, transcript, 8 July 2009, SC 91/2008).
95 Astrazeneca Ltd v Commerce Commission, above n 94, at 52.
96 Attorney General (Minister of Immigration) v Tamil X [2010] NZSC 107 and Huang v Minister of Immigration (Note) [2009] NZSC 77, [2010] 1 NZLR 135.
97 Lab Tests Auckland Ltd v Auckland District Health Board, above n 76.
98 Diagnostic Medlab Ltd v Auckland District Health Board [2009] NZSC 10 (application for leave) at [10]. 99 DGT Williams “Criminal Law and Administrative Law: Problems of Procedure and Reasonableness” in P Smith (ed) Criminal Law: Essays in Honour of JC Smith (London, 1987) 170 at 170.
Frankfurter in saying “[it is] necessary to be on guard against an undue quest for certainty borne of an eager desire to curb the dangers of discretionary power”.100 In her own writing she advocates strongly for a contextual and discretionary based approach to judicial review101 and criticises the exercise of standards of deference by appellate court judges.102Overall, the Court of Appeal has an unstructured and inconsistent approach to substantive review and often sidesteps the question altogether, save for the recent case of Kim. They have indicated to the Supreme Court that the lower courts need some uniform structure and guidance in applying an increasingly common spectrum of review approach to
reasonableness review.103 The Supreme Court has avoided addressing the subject, and when it
has arisen, they have emphasised that substantive review ought to be contextual, rather than structured through doctrine. The model proposed for calibrating intensity of review in this paper is a response to that approach. The approach by the Supreme Court means it is necessary to justify a doctrinal approach before proposing a model for calibrating such an approach.
Overall, reasonableness review in the status quo in New Zealand is without clear direction and lacks guidance in its application.104 It is accepted by the High Court and the Appellate Courts that Wednesbury unreasonableness is no longer the sole standard of reasonableness review. The question then is what the standard of review is. The High Court, and often the Court of Appeal, endorse the application of variable intensity of review,105 and the Supreme Court advocates for a contextual standard of review.106 The outcome of the difference in approach is that the Supreme Court refuses to provide the High Court with structure in applying the reasonableness test, maintaining that it should be assessed on a case by case basis. Currently the High Court applies a variable intensity of review approach in cases of unreasonableness but the way that approach is applied varies from case to case because of the lack of guiding doctrine from the Appellate Courts.
100 Felix Frankfurter “The Task of Administrative Law” (1926) 75 U Pa L Rev 614 at 619.
101 Sian Elias “Administrative Law for ‘Living People’” (2009) 68 CLJ 47 at 66.
102 Sian Elias “’A Painful and Uncongenial Obligation’? Appellate Correction of Error of Fact in the Electronic Age” (address to the Supreme and Federal Court Judges’ Conference, Canberra, 26 January 2010).
103 Knight, above n 49, at 136.
104 Ferrere, above n 3, at 326.
105 Knight, above n 2, at 408.
106 Ferrere, above n 3, at 326.
Chapter III: Contextual Review as an Alternative
Before we propose a model for calibrating intensity of review, we should answer the question of why we ought to favour intensity of review as the preferred form of substantive review, as opposed to contextual review. The core question is whether reasonableness review in a New Zealand context lends itself more towards an approach structured by doctrine, or whether that specific area of judicial review is amorphous to the extent that we would be better to place more discretion in the hands of the judges deciding the cases.107 Importantly, if we are to justify a model for calibrating intensity of review, then it is necessary to justify adopting intensity of review as the preferable form of reasonableness review in New Zealand.This chapter will address the comparison of intensity of review and contextual review in three parts. Firstly, through a descriptive comparison of contextual review and intensity of review as forms of review. Secondly, proposing and justifying Lon Fuller’s Principles of Efficacy and Virtue108 as a means for comparing intensity of review and contextual review with reference to Dean Knight.109 Finally, answering the normative question of why we ought to favour intensity of review and provide a justification of intensity of review in comparison with contextual review. We will compare the two forms of review against Fuller’s principles as articulated in Dean Knight’s Vigilance and Restraint in the Common Law of Judicial Review.110 Then this dissertation will examine analysis as to how each form of judicial review comparatively addresses the tension in substantive review discussed in the previous chapter.
This section will provide a descriptive comparison of intensity of review and contextual review but not a normative analysis. The normative analysis will be provided in Section D. We will look first at intensity of review, then at contextual review. The descriptive analysis of the two will be broadly drawn from Dean Knight’s Vigilance and Restraint in the Common Law of Judicial Review.111
107 Sykes, above n 22.
108 LL Fuller The Morality of Law (1st ed, Yale University Press, 1964).
109 Knight, above n 39.
110 Knight, above n 39.
111 Knight, above n 39.
Intensity of review is a form of substantive review that focuses on the concept of a spectrum of intensity in examining the decisions of administrative bodies and the idea that variegated degrees of deference apply based on the nature of the decision and decision-making body at issue.112 While the language is often complex and peppered with terminology such as “heightened scrutiny” and “light-touch review”, the form of review has fallen into three key aspects.113
112 Philip A Joseph “False Dichotomies is Administrative Law: From There to Here” (Presented to the Centre for Public Law, University of Cambridge, 21 November 2013) at 23.
113 Knight, above n 39, at 184.
114 Knight, above n 39, at 185.
115 Knight, above n 39, at 185.
116 Knight, above n 39, at 186.
117 PP Craig Administrative Law (6th ed, Sweet & Maxwell, 2008) [19-015].
intensity of review across common law jurisdictions. At one end of the spectrum is “correctness review” where there is little to no deference to the decision-maker.118 Then there are the concepts of “structured proportionality review” and “anxious scrutiny” where there is a level of deference, but the burden of justification sits on the decision-making body.119 Then there is “Wednesbury unreasonableness review” and “light touch review” where there is a high degree of deference to the decision-maker, the burden lies on the claimant and the courts would be extremely hesitant to review without something clearly having gone wrong in the decision making.120 This is a brief and simplistic account of varying depths of scrutiny, but it serves to clarify the third aspect of intensity of review.In contrast, contextual review is a form of substantive review that takes its shape from the context of the decisions in question and allocates a great deal more discretionary power to judges in conducting reasonableness review.121 Proponents of contextual review state that such judicial discretion ought to be given primacy for two reasons. First, the courts are the best positioned to be able to identify where a decision has “gone wrong” because it is the job they are trained to do.122 If a decision has been made that is unreasonable to the extent that the court ought to intervene, we should expect that decision to trigger some sort of instinctual response. That response is best recognised by judges because that instinct has been honed sharp by legal practice.123 The test is a “sniff test” and, as Joseph puts it, “seasoned litigators apply the ‘sniff test’ where the decision-making goes palpably awry”.124 Second, regardless of the terminology, contextual review is what the courts undertake at some level whenever they conduct substantive review.125 Doctrinal standards only serve to mask a contextual exercise because courts are either consciously or unconsciously motivated by their instinct when deciding on reviewing a decision.126 Given that, doctrinal standards at best muddy the waters of substantive review, reducing clarity, and at worst allow courts to use the language of a rules-based system to mask discretion. There are two methods of contextual review;
118 Knight, above n 39, at 151.
119 Knight, above n 39, at 151.
120 Knight, above n 39, at 151.
121 Sian Elias “Judicial Review Today” (Maurice Byers Address to the Bar Association Common Room, Thursday 24 April 2008) at 36.
122 TRS Allan The Sovereignty of Law: Freedom Constitution and Common Law (Oxford University Press, Oxford, 2013) at 226.
123 Knight, above n 39, at 231.
124 Philip A Joseph “Exploratory Questions in Administrative Law” (2012) 2 NZULR 75 at 74.
125 Elias, above n 121, at 37.
126 Frankfurter, above n 100, at 619.
strong and weak form. Strong form contextual review rests heavily on the concept of full discretion as outlined above. There are, in essence, infinite varying degrees of depth of scrutiny available to the courts and the judge has full discretion in deciding that depth.127 Weak form contextual review retains the discretion to a lesser extent but focuses that discretion on questions of deference and weight. In its weak form, contextual review borrows the process of weighting factors from intensity of review but removes any doctrinal standards, leaving it to the discretion of the judge.128
We ought to import Lon Fuller’s Principles of Efficacy129 as a metric to compare the forms of review, in the same way that Dean Knight employed them in Vigilance and Restraint in the Common Law of Judicial Review. Fuller identified eight criteria, or principles, with which to judge the efficacy and virtue of legal laws.Those principles are as follows:
127 Knight, above n 39, at 231.
128 Knight, above n 39, at 232.
129 Fuller, above n 108.
130 Knight, above n 39, at 27.
131 Knight, above n 39, at 27.
132 Knight, above n 39, at 28.
133 Knight, above n 39, at 29.
Dean Knight adopts these principles as a metric with which to conduct a normative assessment of competing doctrines of judicial review.138 Firstly, he notes that Fuller is already widely recognised, and his principles of efficacy are “well regarded as a set of standards for examining rule-based systems for their value and virtue”.139 Fuller provides an easy set of metrics for judging doctrinal approaches to administrative law. Secondly, Knight notes that Fuller’s principles are a good judge for the end goal, which is providing the administration of justice. He notes that it is not purely the administration of justice that we care about, given that there are a number of avenues to reach that, but an overall assessment of the efficacy of each of those avenues.140Using Fuller’s principles as a metric for comparison provides us with a means of comparing the ability for forms of review to resolve the current confusion in the application of substantive review. The principles of clarity, stability and public accessibility all promote accountability and certainty in the law in a way that minimises broad judicial discretion and prevents the examination of the merit of decisions.141 The metric focuses on ensuring that the law is clear and known, which limits the use of discretion in judgements where there is the potential to make rulings based on merit.
134 Knight, above n 39, at 29.
135 Knight, above n 39, at 30.
136 Knight, above n 39, at 29.
137 Knight, above n 39, at 30.
138 Knight, above n 39, at 23.
139 Knight, above n 39, at 24.
140 Knight, above n 39, at 25.
141 Fuller, above n 108, at 212.
The metric also has a particular emphasis on the rule of law, being itself a rules-based standard.142 That is open to criticism because a rules-based standard may weight a balancing exercise more in favour of the rule of law with less of a consideration to the violation of the separation of powers. I would argue that using a rules-based standard is a benefit rather than a flaw. Judicial review is principally founded on upholding the rule of law143 and so it ought to be judged in line with its principled basis. Further, it is useful for a rules based standard to place emphasis on the rule of law because it ensures that the courts are properly conducting an inquiry into whether a decision is ultra vires rather than a review on the its merits. Fuller’s principles allow us to police that boundary because the principles of prospectivity and congruence ensure the proposed model upholds the principles of the rule of law. Finally, the principles articulated by Fuller are justified through their relevance to the practical efficacy of the law, not simply their relevance to the principle of the rule of law.144 The basis of the principles lies in practicality and not just in abstract principle.145 That means that doctrine which fulfils Fuller’s principles is doctrine that works practically as law, not just doctrine that meets a principled standard.146 To reject the metric it is not enough to merely value discretion over doctrine, but one also must show that a discretionary standard is as practical in its application as a standard which fulfils Fuller’s principles.147
In this section, we will conduct a normative assessment comparing intensity of review and contextual review on two grounds. Firstly, drawing from Knight’s analysis using Fuller’s principles and looking at the efficacy of the two forms of review as rule-based schema.Second, on a principled basis, assessing how each approach balances the principles in tension in substantive review as described in the previous chapter.
In applying Fuller’s principles as a metric for comparing the two, it is useful to identify where each perform poorly in comparison to the other. In terms of public accessibility and
transparency, contextual review is intended to “channel the judicial impulse”148 and as such does not provide sets of pre-determined justifications for intervention. Knight notes that the
142 Knight, above n 39, at 25.
143 Joseph, above n 20, at 174.
144 Knight, above n 39, at 25.
145 Timothy Endicott “Are There Any Rules?” (2001) 5 Legal Ethics 199 at 207.
146 Knight, above n 39, at 26.
147 Knight, above n 39, at 26.
148 Knight, above n 39 at 235.
justification for intervention is provide ex post facto, after the instinctual decision has already been made.149 This provides little in the way of transparency of judicial decision-making and discretion is inherently inaccessible to the public. In applying contextual review in R (Abdi & Nadarajah) v Secretary of State for the Home Department150 Laws LJ notes that the approach is “not merely simple, but simplistic”151. In comparison, intensity of review performs well under this metric because it explicitly provides the exercise in calibration performed by the judge.152 The three aspects force judges to engage in what is essentially a three-stage process of reasoning. That engagement and justification that separates out the factors and weight being addressed provides a degree of transparency that contextual review lacks. That transparency allows intensity of review decisions to be more easily scrutinised by the public.153Against the principle of clarity, contextual review scores poorly. By its nature, the lack of a doctrinal standard decreases certainty in outcomes and makes judicial decision-making difficult to predict.154 Courts use broad statements such as “common sense” and “abuse of power” to justify triggers for judicial intervention, but the meaning of those statements and the weight given to the factors is discretionary. That discretion means judges set their preferred meanings of those statements and so triggers for intervention are unpredictable and vague.155 Intensity of review also encounters issues with this metric. In particular, the depth of scrutiny is dependent on the weight allocated to factors; that weight tends to be viewed as discretionary and so encounters the same uncertainty as does contextual review.156 It should be noted that in each case the lack of clarity is a product of discretion: in contextual review the entire analysis is discretionary whereas in intensity of review, only the weighting of factors is discretionary. Accordingly, intensity of review always performs comparatively better in terms of clarity but is not necessarily clear itself.
In terms of stability, both contextual review and intensity of review perform relatively well. Contextual review is based on contextual discretion and so is theoretically able to move easily with evolutions in judicial thought. Intensity of review can address such changes
149 Thomas Poole “Between the Devil and the Deep Blue Sea” in Linda Pearson, Carol Harlow and Michael Taggart (eds.) Administrative Law in a Changing State (Hart Publishing, 2008) at 39.
150 R (Abdi & Nadarajah) v Secretary of State for the Home Department [2005] EWCA Civ 1363.
151 R (Abdi & Nadarajah) v Secretary of State for the Home Department, above n 150, at [67].
152 Knight, above n 39, at 192.
153 Knight, above n 39, at 192.
154 Knight, above n 39, at 238.
155 Knight, above n 39, at 238.
156 Knight, above n 39, at 238.
through shifts in the factors and their weight. Knight does note that the lack of transparency in contextual review means that people are left unable to predict judicial outcomes when there are changes in judicial thought.157The non-doctrinal nature of contextual review means that there is no attempt to promote coherence and non-contradiction in the regime.158 The mechanism for achieving coherence and avoiding contradiction is through individual judicial reasoning contributing towards a holistic understanding of the triggers for intervention. In principle, coherence is produced as a by-product of clear reasoning, but in practice Knight notes that coherence tends to break down in practical application when judges apply their individual instincts.159 Intensity of review relies on a similar mechanism, but only in the aspect of developing the weight allocated to factors. So again, there is a lack of coherence to a lesser degree, with the more doctrinal standard promoting non-contradiction.160
Contextual review performs well against the principle of non-impossibility because it is a practical judicial tool. All that is required is an articulation of the judicial instinct and the common-sense approach is easy and practical and can be applied across an infinitely broad set of circumstances.161 However, cases that rely on triggering the judicial instinct place an incentive on the litigants to advance as much material as possible on the off-chance that something triggers a “gut reaction” in the judge. That incentive generates a feedback loop in which the plaintiff and defence ratchet up the level of evidence provided in each case, costing judicial time and resources and increasing the costs of litigation for the litigants.162 Litigants under an intensity of review framework face similar incentives, but not to the same extent as contextual review. Contextual review gives primacy to the personality of the judge, unconstrained by the limiting effect of doctrine, exacerbating the incentive to flood the court with corpus.163
Against the principles of congruence and candour, Knight holds that contextual review presents the appearance of congruence, but it manifests poorly in the judicial instinct. Ex post facto reasoning, when it occurs, means that the judicial reasoning hides the operation of
157 Knight, above n 39, at 194.
158 Knight, above n 39, at 239.
159 Knight, above n 39, at 239.
160 Knight, above n 39, at 194.
161 Knight, above n 39, at 239.
162 Knight, above n 39, at 240.
163 Knight, above n 39, at 195.
individual judges’ instinct and biases, resulting in a lack of judicial candour.164 In comparison, intensity of review provides a rules-based schema that sets out the stages of judicial reasoning and necessitates justification for these stages. The schema is “opentextured and malleable” to suit a variety of circumstances, but it provides a framework within that incentivises judicial candour, and structures decision-making.165
Finally, the two forms of review ought to be compared on the basis of how well they resolve the tension identified in the previous chapter. Contextual review is uniquely poor at resolving the tension between the rule of law and the separation of powers for four key reasons.
First, given that contextual review performs poorly against Fuller’s principles, it also follows that this indicates it poorly resolves the tension because the principles are a measure of a schema’s ability to uphold the rule of law.166 Fuller’s principles highlight a lack of transparency and indicates where the form of review lacks clarity, undermining the ability to recognise and address merit-based review.167 Contextual review scores poorly because the metric is rules-based and opposes judicial discretion, so it weighs more heavily on the side of the rule of law. But, as outlined in the previous section, emphasis on the rule of law holds the judiciary to account in ensuring that review of a decision is on the basis of the power the decision-maker was authorised, not the merits of the decision.
Second, the lack of a framework and rules-based guidelines leaves contextual review open to the discretion of the judge in question, even in the case of weak contextual review.168 That discretion is more open to merit-based review because what naturally subconsciously turns the mind of the judge in a “gut reaction” is whether it is a “good” decision, not whether the decision was objectively reasonable in the circumstances. Further, the lack of a framework means that there is no metric to assess those decisions and hold them to account. This means the judiciary is more likely to overstep their role in terms of the separation of powers and less likely to properly enforce the rule of law.169 Extensive discretion was criticised in Duport
164 Knight, above n 39, at 240.
165 Knight, above n 39, at 196.
166 Knight, above n 39, at 25.
167 Paul Daly “Substantive Review: Categories, Context, Controversy” (22 February 2019) Paul Daly Administrative Law Matters <https://www.administrativelawmatters.com/blog/2019/02/22/substantive-review- categories-context-controversy/>.
168 Knight, above n 49, at 140.
169 Knight, above n 49, at 140.
Steels Ltd v Sirs170 by Lord Scarman: “Justice [...] is not left to the unguided, even if experienced, sage sitting under the spreading oak tree”.Third, as outlined above, under contextual review, litigants have an incentive to ratchet up the amount of corpus presented to the court in the hope that something will trigger an instinct in the judge.171 The accompanying cost to court resources of this incentive corresponds with less clarity in the trial and in the reasoning because the judge must filter through all of the evidence provided.172 Less clarity stymies accountability and a cost to court resources undermines the quality of the court reasoning.173 That detracts from the ability of the judge to conduct the balancing exercise necessary in substantive review.
Finally, as touched on above, contextual review does not promote transparency and accountability because there is no framework through which to assess judicial decisions.174 That means that, even if the schema successfully works to resolve the tension, it is more difficult to recognise the cases where judges overstep their role and engage in merit-based review. While in both contextual review and intensity of review the appellate courts act as a check and balance, it is comparatively easier to recognise merit-based review with the framework provided through intensity of review.
At the heart of the matter the question is a simple one. Even if it is true that, regardless of the terminology, courts are wholly motivated by instinct in reasonableness review, should we have a framework of rules that structures the way in which those decisions are made? I would argue that we should, because even if judges do operate on instinct, a framework reduces that instinctual response, while contextual review grants that instinct primacy. The framework serves to identify the separate elements of the instinctual trigger, forcing a judge to justify each of the elements. That promotes accountability and sets a metric that judges can be assessed based on, reducing instances of merit-based review.
170 Duport Steels Ltd v Sirs [1980] 1 WLR 142 at 168.
171 Knight, above n 39, at 240.
172 Knight, above n 39, at 240.
173 Knight, above n 2, at 408.
174 Jeffrey Jowell and Anthony Lester "Beyond Wednesbury: Substantive Principles of Administrative Law" (1988) PL 368.
A comparison of contextual review and intensity of review is necessary because the New Zealand appellate courts either do not address intensity of review or explicitly advocate for a contextual standard of review. Intensity of review provides a framework based on factors and weight allocated to those factors that sets the level of depth of scrutiny.175 Contextual review puts primacy on judicial discretion and calibrates the triggers for intervention through common-sense and the process of judicial reasoning.176 It is necessary to justify intensity of review as the preferable form of reasonableness review before providing a model for calibration. According to the metrics of efficacy provided by Fuller’s principles, contextual review overall scores comparatively worse than intensity of review. Through the balancing exercise designed to resolve constitutional tensions, contextual review provides no checks and balances against merit-based review and enables such review through the discretion of the judge.177 Overall, although the Supreme Court has favoured contextual review, intensity of review is both more effective as a schema for reasonableness review and better balances constitutional principles.
175 Joseph, above n 112, at 23.
176 Elias, above n 121, at 36.
177 Knight, above n 49, at 140.
Chapter IV: Developing a Model for Calibrating Unreasonableness
Any proposed model for calibrating intensity of review in reasonableness review needs to provide clarity and direction for the lower courts that engage in that calibration exercise. The model need not solve every issue with reasonableness review, it need only provide a better alternative than the status quo: confusion about how to engage in that calibration exercise resulting from the lack of guidance from the Court of Appeal and the Supreme Court. In this chapter we will first lay out a brief overview of my proposed model: a Two-Stage Bifurcated Model. This chapter is purely descriptive: chapter five engages in the normative assessment of the proposed model.
As mentioned in the previous chapter, Dean Knight identifies a basic intensity of review exercise as comprising of three core elements.178 Those elements are:
(1) identifying the factors relevant to calibrating intensity of scrutiny;179
(2) identifying the weight allocated to each relevant factors;180 and
(3) the intensity indicated by the specific combination of weights and factors.181
Those three elements are the building blocks that underly any model calibrating intensity of review.182 The Two-Stage Bifurcated Model tries to undertake those three elements in a two- stage format.
The two stages of this model are, first, an entry point stage that evaluates the existence of the relevant factors, and, second, a review stage that provides two different depths of scrutiny depending on the factors identified in the first stage. The first stage is a rights-based enquiry: the only relevant factor to identify at the entry point is the existence or absence of rights- based considerations in a decision. The second stage provides the depth of scrutiny. If there are rights-based considerations in the administrative decision, then the courts should take an “anxious scrutiny” approach. That approach is proportionality-based and is articulated below
178 Knight, above n 39, at 184.
179 Knight, above n 39, at 184.
180 Knight, above n 39, at 187.
181 Knight, above n 39, at 188.
182 Knight, above n 39, at 188.
as a “demonstrably justifiable” test. If the decision is absent of rights-based considerations, then the courts are to apply the “Wednesbury unreasonableness” test as articulated below.This model addresses Dean Knight’s three elements of intensity review in a way that maximises clarity in the law. The first stage of the model both identifies the relevant factors and attributes weight to them. Barring some exceptions below, the only relevant factor to identify for the courts is the presence or absence of rights considerations in an administrative decision. The weight given to that factor is binary. The court need not conduct a complex balancing exercise: the existence of the factor of rights necessitates one depth of scrutiny, the absence necessitates another. The depth of scrutiny is simplified as well. Rather than a continuum of varying levels of depth, there are binary options depending on the content of
the administrative decision. The model fulfils Dean Knight’s articulation of the elements of intensity review but does so in a way that provides a clear structure that removes elements of judicial discretion that unhelpfully blur the lines between intensity review and contextual review.
Stage one of the model acts as the entry point to unreasonableness review and provides the identification and weighing of factors necessary for a structured approach to intensity review. The core is of this first stage is an investigation as to whether an administrative decision involves rights-based considerations. Rights-based considerations act as the entry point because the presence or absence of rights in an administrative decision is the sole indicator of the depth of scrutiny the court ought to apply to the decision. Stage one of the model is, therefore, merely an inquiry by the court as to whether the administrative decision concerns rights. ‘Rights’ in this context are those enumerated in an exhaustive set of rights instruments detailed below. For example, an administrative decision concerning company mergers by the Commerce Commission, such as in the recent Court of Appeal judgement on NZME Limited v Commerce Commission183, is likely not to involve rights-based considerations, while a decision by the Minister of Immigration to deport someone, such as in Wolf,184 is likely to involve rights-based considerations.
183 NZME Limited v Commerce Commission [2018] NZCA 389.
184 Wolf v Minister of Immigration, above n 13.
In order to provide structure and certainty to stage one of the model, the list of rights the court can draw from must be defined and exhaustive, and the sources from which the court can identify rights will not be open to judicial discretion. The sole sources that the court can use to identify rights from are rights instruments, both national and international, to which the Executive is a signatory (or is otherwise bound by). Those instruments applicable are only those that purport to articulate rights. The applicable instruments would be:
185 New Zealand Bill of Rights Act 1990.
186 Universal Declaration of Human Rights GA Res 217A (1948).
187 International Covenant on Economic, Social and Cultural Rights GA Res 2200A (XXI) (1976).
188 International Covenant on Civil and Political Rights GA Res 2200A (XXI) (1976).
189 United Nations Convention on the Rights of the Child GA Res 44/25 (1989).
190 Convention on the Rights of Persons with Disabilities [A/RES/61/106] (opened for signature 13 December 2006, entered into force 3 May 2008).
191 Convention on the Political Rights of Women UNTS 193 (opened for signature 20 December 1952, entered into force 7 July 1954).
192 Convention against Torture and Other Cruel, Inhumane or Degrading Treatment or Punishment UNTS 1465 (opened for signature 10 December 1984, entered into force 26 June 1987).
193 Convention against Discrimination in Education UNESCO (entered into force 22 May 1962).
194 Right to Organise and Collective Bargaining Convention ILO Co98 (opened for signature 1 July 1949, entered into force 18 July 1951).
195 Geneva Convention Relative to the Treatment of Prisoners of War (Third Geneva Convention) 75 UNTS 135 (entered into force 21 October 1950).
196 Geneva Convention relative to the Protection of Civilian Persons in Time of War (Fourth Geneva Convention) 75 UNTS 187 (entered into force 21 October 1950).
Those instruments comprise an authoritative source of rights and any rights-based considerations identified in the administrative decisions must be drawn from those instruments. In conducting the inquiry, the court does not have judicial discretion to identify rights that are not present in those instruments.The inquiry, like any inquiry identifying rights, will involve a process wherein the court identifies abstract rights in practical contexts. The court naturally has a degree of discretion to read rights broadly to be able to capture sets of facts that may not involve rights-based considerations on a narrow reading of the instruments. Structuring discretion in this case is difficult, because firstly there is always a degree of natural discretion in the way in which courts read documents, and secondly in an overly restrictive model it becomes more likely that arbitrary distinctions will be made in cases that exist on the margins of engaging rights- based considerations. The check and balance in this case will just be that of judicial reporting. The fact that there are objective metrics at this stage – whether rights are engaged or not – makes it comparatively easier for an appellate court to review this classification stage. Given that, the judiciary will moderate itself through internal criticism of judicial reporting. Judges must justify broad interpretations and following judgements can choose whether they consider the justification legitimate. Internal criticism corrects against outliers and, over time, establishes a body of authoritative case law.
Where stage one requires the Court to identify rights-based considerations in an administrative decision, the bifurcated approach at stage two classifies the depth of scrutiny depending on the presence or absence of those considerations. If rights-based considerations are present in the decision, the court will engage in an “anxious scrutiny” approach to substantive review. That approach would be in the form of “structured proportionality review” as identified in De Smith,197 or the “demonstrably justifiable” test employed in s 5 of the New Zealand Bill of Rights Act 1990198 and in the Canadian approach to substantive review articulated in Dunsmuir v New Brunswick.199 That approach would require a court to engage in an inquiry as to whether the decision in question, with particular regards to the rights-based content of the decision, was demonstrably justifiable in a free and democratic
197 De Smith S Judicial Review of Administrative Action (6th ed, Sweet & Maxwell Ltd., 2007) at 592.
198 New Zealand Bill of Rights Act 1990, s 5.
199 Dunsmuir v New Brunswick [2008] 43 SCLR (2d) 1.
society. The test is an inquiry into the result intended by the administrative decision-maker and whether that result is proportional to the means used by the decision-maker in achieving it.200 Importantly, the test is logically consistent with the purpose of substantive review. If there is an administrative decision that interferes with the rights of an individual in a way that is both procedurally fair and non-illegal, then it would be odd for us to say that it was a ‘reasonable’ decision if the court finds that the rights interference was disproportionate, or not demonstrably justifiable, considering the ends that the decision-maker was intending to achieve.201The first exception to the application of the “demonstrably justifiable” test is where there is a clear and unarguable indication that Parliament either intended for the power to be used in the particular way. In the first instance, if there is legislation that indicates that a certain means may be used to achieve a certain end, then that would always indicate that the use of such a means to achieve an end is demonstrably justifiable and proportionate. That applies even so far as the means interferes with the rights of an individual or group, so long as Parliament has indicated that such interference is acceptable in this instance.
The second exception to the “demonstrably justifiable” test is the existence of a privative clause in the legislation in question, indicating that the court ought not to review a decision made under that legislation. In the instance of the existence of a privative clause the approach by the court in judicial review is guided by the Ramsay202 principle; “if the perceived error or invalidity cannot be fitted within [the statutory] procedure, then the exclusion of other
remedies will not apply”.203 That test is still for the court to determine in cases of
unreasonableness, what the existence of a privative clause indicates is a Parliamentary rejection of a greater depth of scrutiny. If it is the case that the statutory procedure is not available, the existence of a privative clause would necessitate the application of the test in stage two (b) below, that test being the “Wednesbury unreasonableness” standard as opposed to the “demonstrably justifiable” test. The justification for privative clauses acting as an exception is that they provide the clearest indication of Parliamentary intent with regards to depth of scrutiny by the courts. Where that indication exists, Parliamentary intent takes precedence on the basis of comity.
200 Knight, above n 39, at 153.
201 Paul Craig “Proportionality, Rationality and Review” [2010] NZ L Rev 265 at 270.
202 Ramsay v Wellington District Council [2006] NZAR 135 (CA).
203 Ramsay v Wellington District Council, above n 202, at [33].
In the case where the court has conducted the stage one inquiry and found that the administrative decision is without rights-based content – as will often be the case in purely regulatory decisions – the bifurcated approach in stage two would mandate a lower scrutiny of review. The degree of scrutiny applicable to these decisions would be at the same depth as standard “Wednesbury204 unreasonableness” review. The specific test applicable would be that articulated in cases that have adopted the language of Wednesbury, specifically Wellington City Council v Woolworths205 and Richardson P’s articulation of the test as “they must be so “perverse”, “absurd” or “outrageous” in [their] defiance of “logic” that Parliament could not have contemplated such decisions being made by an elected council.206” The decision would only be found to be unreasonable if it were to trigger the judicial language of “outrageous” or “arbitrary” as engaged in the cases of CCSU207 and Pro-Life Alliance.208 The intention is to give deference to decision-makers where a decision’s content does not engage rights-based concerns. “Wednesbury unreasonableness” would only be engaged as a test in outlier cases where decision-makers have acted in a way that is clearly outside of the bounds of the powers implicitly allocated to them by Parliament.209 It would only be necessary to engage reasonableness review in these instances because they lack the rights-based content that would indicate more anxious scrutiny.
Overall, the model does away with the previous complex and discretion-based articulations of intensity of review as a “sliding scale” of review. It does so by removing the array of different factors at varying levels of weight and replacing that array with two stages of inquiry with binary results at each stage. The first stage asks, “are there rights-based considerations present in this administrative decision?” The inquiry then seeks to identify the presence of rights-based considerations using an exhaustive set of authoritative sources of rights as a means of structuring the approach and limiting judicial discretion. The outcome of
204 Associated Provincial Picture Houses Ltd v Wednesbury Corp, above n 7.
205 Wellington City Council v Woolworths New Zealand Ltd (No 2) [1996] 2 NZLR 537 (CA).
206 Wellington City Council v Woolworths New Zealand Ltd (No 2), above n 205, at [13].
207 Council of Civil Service Unions v Minister for the Civil Service, above n 28.
208 R (Pro-Life Alliance) v British Broadcasting Corporation [2003] UKHL 23; [2004] 1 AC 185.
209 See for example the application of Wednesbury in Unison Networks Ltd v Commerce Commission [2007] NZSC 74 at [55]; and the discussion on “public wrongs” in Taggart, above n 1, at 469.
the first stage is either (a) there are rights-based considerations present, or (b) the decision is absent of rights-based content. If (a); the second stage mandates that the depth of inquiry is “structured proportionality driven by a “demonstrably justifiable” test. The outcome of the test is either the decision is or is not demonstrably justifiable and so, is or is not unreasonable: if (b); the second stage mandates greater deference with the backstop of the Wednesbury unreasonableness standard. The outcome of that inquiry is either that the decision is outrageous or arbitrary in which case it is unreasonable.
Two-Stage Bifurcated Model
Rights-Based inquiry guided by an exhaustive set of rights-articulating instruments
Stage Two
Bifurcated standard of intensity of review depending on the presence or absence of human rights concerns
Human Rights
Test: “Demonstrably justifiable in a free and democratic society”
No Human Rights Test: Wednesbury unreasonableness
Exceptions
Chapter V: Justifying the Two-Stage Bifurcated Model
This chapter will examine how the two-stage bifurcated model will work practically, as well as seek to justify the decisions made in constructing the model. This normative stage of analysis will seek to prove that the model is fit for purpose, internally coherent and workable in a wider scheme of judicial review. In particular, this analysis will show that the model solves the core problem with the status quo: a lack of clarity and guidance in the lower court’s exercise of intensity review in reasonableness review.This chapter will thus do two somewhat different things. First, it will justify the Two-Stage Bifurcated Model internally: even before we ask questions about whether the model is better than the alternatives, we must ensure that the model is internally coherent and allows the court to perform reasonableness review in a way that is neither constrained arbitrarily or so open to discretion as to become an exercise in semantics. Likewise, the model must allow for reasonableness review to exist as an independent and useful exercise in judicial review and must not be open to the court straying into the realms of merit-based review. Second, the chapter will compare the model against the alternative of a directionless status quo and a shift towards contextual review by the Supreme Court. The chapter will do both those things in five sections, the first four being a breakdown of the choices made when constructing the model: (1) the choice of rights as the entry point, (2) the sources of rights in the judicial
inquiry, (3) the use of the “demonstrably justifiable test”, and (4) the use of Wednesbury unreasonableness as a backstop in stage two. The fifth section is an evaluation against Fuller’s Principles of a rule of law system as a metric to measure the degree of coherence and guidance a doctrinal approach – such as the model – provides.
Core to the model is the idea that the existence of rights considerations in an administrative decision is the only relevant factor for determining the depth of review. The model rejects the concept of a continuum of depth based on a series of different factors and weights and seeks to reduce judicial discretion creating more structure in the intensity of review inquiry. The concept of intensity of review still exists in form, but the way courts evaluate cases is no longer based on a spectrum of intensity: now it is a binary choice. That choice, and so theentry point to the model, is based on rights and rejects all the other factors the courts have used in the past in evaluating depth. This section will look at how rights-based considerations work in practice as an entry point, elaborating on the analysis in the previous chapter, before then outlining the normative justification for that entry point.
The practical evaluation of rights-based content is straightforward: the Court examines the content of the decision in question and decides, based on the sources considered in the next section, whether the decision contains rights considerations. There are two initial things to note about the practical evaluation. First, it is the existence of rights considerations in the decision in question that matters at this stage, not the scale of the rights infringed, because if rights-based considerations do exist it is the role of the demonstrably justifiable test to assess their scale against the ends the decision-maker is seeking to achieve.210 Second, the rights- based content ought to be relevant to the aspect of the decision that the court is considering for reasonableness review. For example, if there is a decision in multiple stages, and only one of those stages is under consideration for being unreasonable, it would not be correct for the court to decide that the decision had rights-based content on the basis that the content existed in another, unrelated, stage of the decision. Those two clarifications serve to ringfence the way that the courts can identify rights as existing in a decision. That ringfencing of the inquiry adds structure to the intensity of review exercise and limits judicial discretion in a way that having a spectrum of factors and weights does not. If a concern with intensity of review is that it becomes an exercise in semantics, limiting the inquiry to rights and creating a binary approach to depth addresses that concern by limiting the scope of power for the court.
In terms of the justification for solely using rights as the factor that determines depth of review, the ringfencing clarification is particularly important. Rights-based content is the sole measure of depth of scrutiny because rights act as the most powerful judicial indicator of the seriousness of a decision and its consequences.211 It indicates clearly the content of the decision more so than the spectrum of review articulated in Wolf212 or the multifactorial approach in Dunsmuir,213 involving factors such as the position of the decision-maker, the context in which the decision was made and the importance of the decision to those affected. It also indicates the way in which the content of the decision will impact the parties involved.
210 Geeti Faramarzi “The Bill of Reasonable Rights: Solving a Conundrum and Strengthening an Enactment (2009) Canterbury Law Review Vol. 5 at 47.
211 Jeff King “Proportionality: A Halfway House” [2010] NZ L Rev 327 at 340.
212 Wolf v Minister of Immigration, above n 13 at [47].
213 Dunsmuir v New Brunswick, above n 199, at [57-61].
If the courts have recognised that the standard for reasonableness has been relaxed in recent years, that judicial shift came as a result of an appreciation that decisions that infringe on people’s lives in a meaningful way should be held to a higher degree of scrutiny than those that do not.214 The most readily accessible trigger in evaluating that infringement is whether the decision concerns rights, because if it does concern rights then, necessarily, it must constitute a significant impact on someone’s life.215 That use of rights as a trigger for greater scrutiny was evident in the English cases which shifted the standard away from Wednesburyunreasonableness. As covered previously, in R (Bugdaycay) v Secretary of State for the Home Department216 Lord Bridge held that a move away from Wednesbury unreasonableness was justified on the basis that some decisions had a greater impact on individual’s lives than others. He stated: “when an administrative decision [...] is said to be one which may put the applicants life at risk, the basis of the decision must surely call for the most anxious scrutiny”.217 Likewise, in R (Smith) v Ministry of Defence,218 the court held the trigger for greater scrutiny of a decision to be the extent to which the decision impacts individual rights, stating “the more substantial the interference with human rights the more the court will require by way of justification”.219 In the last year in New Zealand the Court of Appeal in the case of Kim held that the appropriate standard of reasonableness review would be one of heightened scrutiny when it was found that fundamental rights were at issue in the decision.220
Standards for the rights-based inquiry refers to the scope of rights-articulating instruments available to the court. The model, at each opportunity, ought to reduce ambiguity in the application of reasonableness and promote clarity. The first stage of inquiry in the model is open to judicial discretion in how the court identifies what aspects of a decision involve rights considerations. The model structures that discretion by pointing to an exhaustive set of
214 See for example R (Bugdaycay) v Secretary of State for the Home Department [1986] UKHL 3; [1987] AC 514 at 531; R (Smith) v Ministry of Defence [1995] EWCA Civ 22; [1996] QB 517 at 554; Wolf v Minister of Immigration [2004] NZAR 414 at [48]; Tamil X v Refugee Status Appeals Authority [2009] NZCA 488, [2010] 2 NZLR 73 at [259-276]; Kim v Minister of Justice of New Zealand (CA) at [45].
215 Taggart, above n 1, at 467.
216 R (Bugdaycay) v Secretary of State for the Home Department, above n 35.
217 R (Bugdaycay) v Secretary of State for the Home Department, above n 35, at 531.
218 R (Smith) v Ministry of Defence, above n 37.
219 R (Smith) v Ministry of Defence, above n 37, at 554.
220 Kim v Minister of Justice of New Zealand, above n 81, at [45].
authoritative sources for rights. The problem that emerges in setting any scope of inquiry is finding the balance between an inquiry so broad that it is effectively an exercise in discretion, and one so narrow that the distinctions it makes are arbitrary and not fit for purpose.221 The model chooses to set that balance by constraining the inquiry to just instruments that articulate rights so that there is an exhaustive set of rights to be drawn from.The wider scope is justified on two grounds. First, even unincorporated international instruments hold some relevance to the considerations of the court. A 1996 Law Commission Report stated that unincorporated treaties were “relevant to the determination of the common law”.222 The case of Hoskings v Runting found that international instruments were “a vital source of relevant guidance”223 and the case of C v Holland held that there was a presumption that domestic law ought to be applied consistently with ratified international conventions.224 Second, a larger range of rights means that the courts are not arbitrarily confined in determining what is a right. If the basis of the model is that closer scrutiny is required in decisions which impact rights then it would be inconsistent with that basis to exclude impacts on individuals just because the range of sources for rights has been drawn too narrowly.225 Courts are granted a degree of discretion in interpreting rights but the
model’s purpose is to provide doctrinal guidance and that guidance is undermined where courts are effectively given unfettered discretion.
The other, more principled, consideration is that the use of international instruments grants too much weight to Executive decisions where the signing of international instruments is a unilateral decision by one element of the Executive.226 Decisions by the Executive to sign up to instruments unconstrained by Parliament would become considerations by the court that would be relevant in declaring decisions to be unreasonable. The concern is that this means international instruments not enacted as domestic law now bind the Executive without Parliament’s consent. The concern can be addressed in two ways. Firstly, the process of ratification means that the Executive is not acting unconstrained by Parliament in signing international instruments. The constitutional convention in New Zealand is that the Executive is not to bind to a treaty without a period of Parliamentary scrutiny, and that convention is set
221 Knight, above n 23, at 119.
222 Law Commission A New Zealand Guide to International Law and its Sources (NZLC R34, 1996) at [68].
223 Hosking v Runting [2004] NZCA 34; [2005] 1 NZLR 1 (CA) at [6].
224 C v Holland [2012] NZHC 2155 at [69].
225 Geoffrey Palmer “What the New Zealand Bill of Rights Act Aimed to Do, Why it Did Not Succeed and How it Can Be Repaired” (2016) 14 NZJPIL at 179.
226 Petra Butler “Human Rights and Parliamentary Sovereignty in New Zealand” (2004) 35 VUWLR at 349.
out in the Cabinet Manual.227 After scrutiny by a Select Committee the treaty Parliament ratifies the treaty.228 The Executive is not acting unilaterally, it is acting with the consent of the House. Secondly, it is already established case law that rights beyond those in the New Zealand Bill of Rights Act will determine a court’s decision. In Wolf the rights identified were those articulated in the Convention on the Rights of the Child, not the New Zealand Bill of Rights Act.229 In Baigent’s Case the judicial activism was informed by the Optional Protocol to the International Covenant on Civil and Political Rights.230 The New Zealand courts have established themselves as willing to bind the Executive based on non- incorporated international instruments, and the above cases of Hoskings and Holland support that conclusion.231 Those instruments do not determine an outcome, but merely a depth of review.Finally, in practical terms the stage one inquiry looks like the court identifying rights in the practical outcomes of a decision.232 Like any rights-based inquiry, the inquiry is an imperfect matchmaking exercise between abstract rights in instruments and practical impacts. As covered earlier, the check on judicial discretion in this case would be simply judicial reporting and criticism by courts and practitioners. The previous chapter covers how this check and balance would work in practice; the ideal is that it would develop a body of case law that would guard against the extremes of judicial discretion. That check is the same as the theoretical check and balance on contextual review, but the difference in this case is that the check is a part of a larger structured whole. If the problem with contextual review is that reliance on judicial discretion, then it is comparatively better to have a model that provides doctrinal structure, for which a part is regulated by judicial reporting, than to have contextual review where there is no structure and entirely relies on judicial reporting. Where judicial reporting is the only check and balance the problems with it are magnified.233
227 Cabinet Office Cabinet Manual 2017 at [5.79].
228 Cabinet Office Cabinet Manual 2017 at [5.78].
229 Wolf v Minister of Immigration, above n 13, at [11].
230 Simpson v Attorney-General (Baigent's case) [1994] NZCA 287; [1994] 3 NZLR 667; (1994) 1 HRNZ 42 (CA) ["Baigent's case"].
231 Hosking v Runting [2004] NZCA 34; [2005] 1 NZLR 1 (CA); C v Holland [2012] NZHC 2155.
232 Taggart, above n 1, at 467.
233 Knight, above n 39, at 231.
Where there are rights-based considerations in an administrative decision, the second stage mandates an “anxious scrutiny” approach. In the model, that approach takes the form of a “demonstrably justifiable” test of the decision. The wording and form of the test is drawn from s 5 of the NZBORA 1990: “whether the decision in question is demonstrably justifiable in a free and democratic society”.234 The court would conduct the test in the same form as it is conducted in section 5 inquiries, except against the content of administrative decisions asopposed to primary legislation. The following case law on section 5 inquiries would guide the exercise of the court’s discretion in performing that test. The New Zealand High Court in Ministry of Transport v Noort235 adopted the section 5 inquiry from the Canadian case R v Oakes236, which was later affirmed in the Supreme Court in R v Hansen237. In Hansen the Supreme Court proposed proportionality analysis consisting of three stages;
(1) Establishing a rational connection between the limit on rights and the goal pursued;
(2) Establishing whether the right is infringed as little as possible; and
(3) Establishing whether the limit is proportionate to the goal pursued.
That three-stage analysis can be imported for use against administrative decisions in the same way that it is used to assess primary legislation. Each of those stages is equally applicable against the goals of an administrative decision-maker as against the goals of Parliament.
Further, the proportionality test is an evaluation of the means used to achieve the end desired by the decision maker.238 That form of test necessarily allows the court to calibrate the test against the seriousness of the rights infringed, in the same way that Taggart proposes a bifurcated approach incorporating proportionality and rights-review.239 While the test itself is of heightened scrutiny, within the test the court is able to calibrate what is reasonable against the scale of the right infringed.240 Whether means are proportionate to ends, or whether a
means is “demonstrably justifiable”, is an exercise by the courts in evaluating the importance of rights, evaluating the importance of intended ends, and weighing those factors against each
234 New Zealand Bill of Rights Act 1990, s 5.
235 Ministry of Transport v Noort [1992] NZCA 51; [1992] 3 NZLR 260 at 296.
236 R v Oakes [1986] 1 SCR 103 (SCC) at [69-70].
237 R v Hansen [2007] 3 NZLR 1, [42] per Elias CJ; [64] per Blanchard J; [103-104], per Tipping J; [203-204] per McGrath J; [269] and [272] per Anderson J.
238 Woolf, Jowell & Le Sueur De Smith's Judicial Review (6th ed, Sweet & Maxwell, 2007) at paras 11-101-11- 102.
239 Taggart, above n 1, at 472.
240 See for example R (Daly) v Secretary of State for the Home Department [2001] UKHL 26; [2001] 2 AC 532 at [26]; Huang v Secretary of State for the Home Department [2007] UKHL 11; [2007] 2 AC 167 (HL) at [19].
other.241 That exercise requires a degree of discretion, but that discretion can be guided by the pre-existing framework New Zealand has for conducting section 5 inquiries.The reason for adopting the “demonstrably justifiable” test in stage two is because the pre- existing body of case law around the test provides two benefits. Firstly, the courts have pre- existing New Zealand-based guidelines in how to conduct the inquiry that would not exist if the inquiry was new or adopted from another jurisdiction. The inquiry involves an exercise in assigning importance and comparatively weighing that importance.242 It is useful for the courts to have a pre-existing structure for how to perform that inquiry, in the interests of uniformity in judicial decisions, but also predictability by public litigants. Secondly, judges are familiar with a “demonstrably justifiable” test in a way they would not be if the test was created as entirely new, in that it provides a pre-established framework through Hansen.243 Notably in New Zealand there has been comparatively little engagement in section 5 of the NZBORA 1990 outside of using it in an evaluative sense,244 which may indicate that judges are not familiar with the test. However, that criticism by Butler is 17 years old and existed before the analysis in Hansen245 and the application in Taylor.246 It is useful to note that Sian Elias refers to proportionality as providing well-established and clear structure and that proportionality is preferable in issues concerning rights infringements.247
A means/end proportionality test can also be seen in Canadian case law on substantive review. The two stages of the model are similar to the two stages of the approach set forward in Dunsmuir v New Brunswick,248 albeit the Two-Stage Bifurcated Model is unifactorial where the Dunsmuir approach is multifactorial.249 Both involve a two-stage test where the courts first assess the nature of the decision at hand and then use that assessment to apply a depth of review. Where the model improves on the approach taken in Dunsmuir is that the model focuses solely on the presence of rights-based considerations. So, the model does away with much of the complexity that contributed to the lack of clarity in the Dunsmuir
241 See also the section 5 inquiry conducted in Taylor v Attorney-General of New Zealand [2015] NZHC 1706.
242 Andrew S Butler Limiting Rights in (2002) 33 VUWLR at 538.
243 R v Hansen, above n 237. 244 Butler, above n 242, at 539. 245 R v Hansen, above n 237.
246 Taylor v Attorney-General of New Zealand [2015] NZHC 1706.
247 Sian Elias “The Unity of Public Law?” (paper presented to Public Law Conference, University of Cambridge, September 2016) at 17.
248 Dunsmuir v New Brunswick, above n 199.
249 Dunsmuir v New Brunswick, above n 199, at [62].
approach250 but imports the means/end proportionality test that existed as one of the depths of review available in Dunsmuir.251 While Canada’s supreme Charter provides the constitutional justification to use rights as a trigger, the justification in New Zealand comes from the body of case law that acknowledges rights as a trigger for a heightened depth of scrutiny.252
If no rights-based considerations are found in the administrative decision, then Wednesbury unreasonableness would act as the test. That test is engaged only when the decision is what Michael Taggart refers to as a “public wrong” and is not rights-infringing.253 If rights considerations act as an indicator of the seriousness of the decision, then the lack of rights indicates that the decision is comparatively less serious or has less serious consequences.254 In the event where individual rights do not need to be protected, the court is able to grant a greater level of deference to the decision-making power of the decision-maker.The second stage of the test would be simple to implement and would draw from the pre- existing case law established around the use of the Wednesbury unreasonableness test. Courts would have the guidance of the case law established both in New Zealand and the United Kingdom. In particular, New Zealand courts would be able to draw guidance from the New Zealand case of Woolworths,255 which established the use of the Wednesbury test in certain administrative decisions, and later New Zealand cases which identified when Wednesbury review may be applicable.256 The articulation of the test in Woolworths is particularly useful, because the test in that case is applied concerning high-level policy decisions made by local government.257 The context of the Woolworths decision mirrors the contexts in which the Wednesbury unreasonableness test is likely to be used under the model. This provides a
250 See for example Paul Daly “Dunsmuir and the hows and whys of judicial review (Eddie Clark)” (16 February 2018) Paul Daly Administrative Law Matters
<https://www.administrativelawmatters.com/blog/2018/02/16/dunsmuir-and-the-hows-and-whys-of-judicial- review-eddie-clark/> ; David Mullan “Dunsmuir v. New Brunswick, Standard of Review and Procedural Fairness for Public Servants: Let’s Try Again!” (2008) 21 Can J Admin L & Prac 117 at 150.
251 Knight, above n 39, at 156.
252 See for example Wolf v Minister of Immigration [2004] NZAR 414; Watson v Chief Executive of Department of Corrections [2015] NZHC 1227, [2015] NZAR 1049; Taylor v Chief Executive of Department of Corrections [2015] NZCA 477; Kim v Minister of Justice of New Zealand [2019] NZCA 209.
253 Taggart, above n 1, at 448.
254 Taggart, above n 1, at 427.
255 Wellington City Council v Woolworths New Zealand Ltd (No 2), above n 205.
256 See for example Ports of Auckland Ltd v Auckland City Council [1999] 1 NZLR 601 at 681; Unison Networks Ltd v Commerce Commission [2007] NZSC 74 at [77].
257 Wellington City Council v Woolworths New Zealand Ltd (No 2), above n 205.
justification for the use of the test in these contexts, as court in Wolf258 expanded the scope of reasonableness but never challenged that the Wednesbury test was used appropriately in Woolworths.259 It also provides good and clear law to act as a guide in future cases, and specifically law that would be relevant to the contexts in which the test would be employed. The use of the Wednesbury test itself is justified because, if the reason the scope of substantive review has been widened is because Wednesbury was too onerous a test in rights- based cases,260 that is no reason to remove the test as a backstop in cases not involving rights. In fact, Wednesbury continues to be useful conceptually because the purpose it serves is to indicate that a decision is so unreasonable as to clearly not be in the scope of power that Parliament has impliedly granted to the decision-maker.To conclude in justifying the model internally it is useful to illustrate the proposed model through the use of two worked examples, drawing on the facts of the cases of Deliu v Connell and Ors,261 and Kim v Minister of Justice of New Zealand.262 In Deliu, the claim of irrationality against the LCRO was based on the contention that a finding of vexatious action was made “without a scintilla of evidence” to support it.263 If the court were employing the proposed model then the first stage would be whether the decision by the LCRO contained rights-based content, with reference to the rights-articulating instruments listed above. It would be likely that the court would find that the content of a disciplinary review proceeding does not engage the rights articulated in those instruments. The second stage would be the application of the Wednesbury standard of review, as dictated by the absence of rights-based considerations. As noted in the decision in Deliu, the threshold to overturn a decision by the LCRO is high264 and, while the court may not have come to the same decision as the LCRO,265 the decision is neither perverse nor absurd. It is likely that, using the model, the court would not find the decision unreasonable.
In Kim, the challenge to the decision by the Minister involved the claim of unreasonableness on the basis that the result of the decision would be the torture and execution of Mr Kim.266 In this case, using the model, the court would first examine the decision for rights-based
258 Wolf v Minister of Immigration, above n 13.
259 Wolf v Minister of Immigration, above n 13 at [47].
260 Taggart, above n 1, at 426.
261 Deliu v Connell and Ors, above n 15.
262 Kim v Minister of Justice of New Zealand, above n 81.
263 Deliu v Connell and Ors, above n 15, at [42]. 264 Deliu v Connell and Ors, above n 15, at [45]. 265 Deliu v Connell and Ors, above n 15, at [44].
266 Kim v Minister of Justice of New Zealand, above n 81, at [49].
content and would be very likely to find that content present in the form of the right to freedom from torture267 and the right to a fair trial.268 Given the existence of those considerations, the second stage would be the application of the “demonstrably justifiable” proportionality test. The end that the Minister is trying to achieve is the fair execution of international justice. The means is an extradition process with the potential of resulting in torture and execution. It would be likely that the court would find that the extent and risk of the potential harms meant the means were not demonstrably justifiable in a fair and democratic society.
As articulated by Dean Knight,269 as well as earlier in this paper, Lon Fuller posits a set of eight criteria as Principles of Efficacy.270 Given their use by Dean Knight,271 this section will use those criteria as a metric for judging the efficacy of the model as a structured, doctrinal approach. The section will go over each principle, evaluate the model against it, and provide a comparative assessment against the alternative of contextual review.In terms of the first principle of generality, the two-stage model seeks to draw a balance between flexibility on one hand and predictability on the other. That balance is achieved through the structured discretionary stages. Overall structure is provided by the stages of the model and the tests outlined in each of the stages, as well as doctrinal approaches that guide the court in performing those tests. Discretion is engaged in aspects of the tests that allows for a degree of flexibility in how they are applied. Equally, structured proportionality allows for a degree of flexibility in how rights are weighed and assessed against outcomes. The model is generally applicable and contains a degree of both flexibility and predictability that means it can apply equally across a range of situations. In comparison, contextual review necessarily must tend more towards flexibility if it prioritises discretion without the guidance of doctrinal structure. Dean Knight assesses contextual review as performing particularly poorly against the principle of generality.272
267 Convention against Torture and Other Cruel, Inhumane or Degrading Treatment or Punishment, above n 193.
268 New Zealand Bill of Rights Act 1990, s 27.
269 Knight, above n 39.
270 Fuller, above n 108.
271 Knight, above n 39, at 23-32.
272 Knight, above n 39, at 230.
Public accessibility relies on a regime that encourages “reason giving” in a court’s adjudication of decisions.273 The model encourages judges to justify their decisions and provides a degree of structure through which that happens. The structure aids in public understanding as all related decisions are structured in the same way and so can be compared against each other. Justification is also encouraged through the model as parts of the model rely on the check and balance of that justification being provided adequately. Contextual review relies solely on justification being the only check and balance in the exercise of review. The model is similar to contextual review in that aspect but has the added benefit of a structure that simplifies reason-giving and aids public understanding.The model has very little to do with the principle of prospectivity, that the law ought not be retrospective. Dean Knight acknowledges that in judicially-adjudicated disputes there are some instances where that adjudication may have retrospective effect.274 The model provides a structure that can be best understood as being prospective, even though there may be retrospective effect in some cases of the adjudication of decisions. Likewise, contextual review may have retrospective effect but is broadly prospective in nature.275
Clarity is an important element of Fuller’s rule of law schema.276 The law must be clear so that judges have consistency in their adjudication and litigants have a degree of predictability in the outcome of cases. The proposed model provides that clarity in substantive review. The model provides a clearly defined process and calibrates what would otherwise be a vague exercise in determining the intensity of review. The specific binary aspect of the model, the fact that at each stage the court is provided with one of two options to determine, provides the clarity in the law that a vaguer spectrum of review could not. Dean Knight notes that contextual review performs poorly against this principle.277 Discretion in substantive review may be well justified post-fact, but a schema that relies entirely on discretion provides no direction for judges or litigants in conducting cases.
Non-contradiction and coherence focus on the seamlessness of the particular doctrine.278 The model does not run into any major problems in terms of this principle. The model is built to be internally coherent and to logically move from stage one to stage two. The model does not
273 Knight, above n 39, at 27.
274 Knight, above n 39, at 28.
275 Knight, above n 39, at 237.
276 Knight, above n 39, at 29.
277 Knight, above n 39, at 238.
contradict itself because it is formed in such a way as to lead from a broader to a narrower inquiry. Contextual review is, at least at its face, more incoherent as a doctrine because it abandons the use of structure and legal devices to encourage coherency.279Both the model and contextual review fulfil the principle of non-impossibility. The model is practical for the reasons given above, the inquiries that the court would perform through the model are broadly derived from existing law that we can rely on to be fit for purpose. The rights-based inquiry in stage one is guided by structure and an authoritative set of rights- based documents. Contextual review is broadly practical as a form of review. As noted earlier, it tends to create an incentive to overload the court, which undermines practicality on the comparative.280
Stability is just the principle that the laws ought not to change too quickly.281 There is no explicit provision within the model to encourage any degree of rapid change. Predictability is also encouraged through the structure of the doctrine. Comparatively, contextual review does not really encounter issues with this principle. The discretion given to judicial decision- making, and the lack of structure in reason-giving undermines predictability to some extent.282
The model encourages congruence and candour in that it provides structure through which the judge must justify their decision with reference to the outline of the model. The constraints placed on the sources of rights in the first stage of inquiry especially force the court to be faithful to declared rules and to justify decisions in terms of those rules. Fidelity is encouraged between judges and the model because checks and balances exist to ensure that decisions are made only according to the model. In comparison, contextual review can sometimes mask judicial instincts and biases through the lack of a structure for the courts to reference when performing the exercise.283
Overall, the choices in the model attempt to address the concerns that the complexity of intensity of review makes the exercise of such review merely a matter of semantics. That
279 Knight, above n 39, at 239.
280 Knight, above n 39, at 239.
281 Knight, above n 39, at 29.
282 Knight, above n 39, at 238.
283 Knight, above n 39, at 240.
concern is addressed in the first stage by structuring the inquiry into factors and weighting as simply a matter of the existence or absence of rights. That inquiry is guided by an exhaustive set of rights-based instruments. The use of those instruments is justified through the Executive having made the choice to bind themselves to those instruments.284 That allowsclarity in what the rights are and sets justifiable limits on the court’s discretionary power. The two separate tests in the bifurcated second stage are justified in the depth of review they seek to achieve. The “demonstrably justifiable” test assesses the proportionality of any rights- based interference against the desired ends in a way already accepted and used by New Zealand courts.285 The Wednesbury unreasonableness test acts as a backstop where rights are not being infringed. Finally, the model scores well against the eight principles of Fuller’s Principles of Efficacy and, in comparison, performs better than the alternative of contextual review.
284 See for example Hosking v Runting [2004] NZCA 34; [2005] 1 NZLR 1 (CA); C v Holland [2012] NZHC 2155.
285 See for example R v Hansen [2007] 3 NZLR 1; Taylor v Attorney-General of New Zealand [2015] NZHC 1706.
Chapter VI: Conclusion
Reasonableness review in New Zealand lacks clarity because the High Court and the Appellate Courts have been unable to agree on a simple core premise; whether court discretion in conducting substantive review ought to be structured or unfettered. The High Court has consistently applied a variable intensity of review standard, structuring the discretion of the court. The Supreme Court has rejected variable intensity in favour of contextual review, leaving it to the judicial instinct to intervene in cases where “something has gone wrong”. The divergence in approach has left us with two questions; should we favour contextual review or variable intensity of review; and if we opt for variable intensity, how ought we calibrate that test?
Variable intensity of review outperforms contextual review, both in terms of Fuller’s principles and its comparative ability to uphold the rule of law. The reliance “judicial
instinct” naturally leaves contextual review vulnerable to the judge engaging in merit-based review because what naturally turns the mind of the judge is whether a decision is good, rather than whether a decision is substantively reasonable. Even if engagement in merit-based review is symmetrical in both variable intensity and contextual review, variable intensity provides a framework for structuring a judge’s reasoning that promotes transparency and accessibility where that structure is absent in contextual review.
The Two-Stage Bifurcated Model provides a means for calibrating variable intensity of review that draws direction from the High Court while also structuring judicial discretion. The model simplifies the multifactorial spectrum of review proposed in Wolf to a set of two stages with binary results that indicate the subsequent tests applicable. The first stage draws on New Zealand case law indicating that the existence of rights-based considerations in a decision should be the trigger for more anxious scrutiny. That rights-based inquiry is structured through reference to an exhaustive list of rights-articulating instruments. The second stage bifurcates depending on the existence of rights-based considerations. If rights are present, then the court engages in a “demonstrably justifiable” proportionality test. If rights are absent, the test is simple “Wednesbury unreasonableness”.
The use of rights as a binary trigger of depth of review is supported by the direction reasonableness case law has taken in New Zealand, with the Court of Appeal in Kim this year acknowledging as uncontentious that heightened scrutiny was appropriate where rights were
at stake. The sole use of rights as a trigger promotes internal coherency in the model through addressing the semantic complexity of a more multifactorial approach. The provision of an exhaustive list of rights-articulating instruments structures judicial discretion and the use of Wednesbury as a non-rights-based test is appropriate given the decision in Unison.
To put it simply, the courts have been unable to agree on an appropriate balance of doctrinal structure and discretion and that lack of agreement has been destructive to the court’s ability to carry out its supervisory jurisdiction. This papers states that, in this case, we ought to consider structure to be more important and, given that, provides a means of calibrating reasonableness review where such calibration has until now been painfully absent.
Bibliography
Cases: New Zealand
Cases: United Kingdom
Cases: Other Jurisdictions
- R v Oakes [1986] 1 SCR 103 (SCC) at [69-70]
New Zealand Legislation
International Materials
(XXI) (1976)
Books
University Press, 2009)
ed, Cambridge University Press, Cambridge, 2018)
Brookers Ltd, Wellington, 2014)
Articles/Chapters
Fairness for Public Servants: Let’s Try Again!” (2008) 21 Can J Admin L & Prac 117
Parliamentary & Government Material
Reports
Dissertations
Internet Resources
<https://www.administrativelawmatters.com/blog/2018/02/16/dunsmuir-and-the- hows-and-whys-of-judicial-review-eddie-clark/>
<https://www.administrativelawmatters.com/blog/2019/02/22/substantive-review- categories-context-controversy/>
NZLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.nzlii.org/nz/journals/UOtaLawTD/2019/23.html
