|
|
Home
| Databases
| WorldLII
| Search
| Feedback
University of Otago Law Theses and Dissertations |
Last Updated: 11 April 2024
The Past, Present and Future
of the Mistake of Fact Ground of Review in New Zealand Administrative Law
Harry Crawford
A dissertation submitted in partial fulfilment of the degree of Bachelor of Laws (Honours) at the University of Otago – Te Whare Wānanga o Otāgo
October 2023
Acknowledgements
First and foremost, to my supervisor, Marcelo Rodriguez Ferrere, for your invaluable expertise, advice, patience and enthusiasm. It has been a privilege to work under your supervision.
To Jack, Merlin and Dad, for giving this thing a read and providing excellent comments.
To the Law Lord Hoffmans – Bella, Jaiden, Nikita, Shani and Merlin – for getting me through law school.
To my parents, for your relentless love and support.
To Brooke, for your support and affection, even from halfway across the country.
To my friends and flatmates, with a special shoutout to the Gallery, the Office and Dusty Duke, for putting up with my terrible jokes and providing me with many lasting memories over these last five years. What a time it has been.
And to Grandad, for teaching me the value of hard work and for choosing to bet on me instead of the horses. You will be missed.
Table of Contents
Including the Availability of Evidence 6
in the Reasoning 8
1 Critique: Unorthodox Approach to Assigning Standards of Review 43
Introduction
“Facts are important. This is a trite observation, but true nonetheless.”
The mistake of fact ground of review has had a tumultuous history in New Zealand administrative law. From its initial championing in 1980 by Cooke J in Daganayasi,2 there has been much debate regarding what is – and what should be – the ground’s scope, parameters and purpose, extending to whether a mistake of fact should constitute a reviewable error at all. The ground’s contested nature has manifested in clashes between legal commentators and inconsistent and uncertain application at all levels of the judicial hierarchy.
Such inconsistency and uncertainty creates an opportunity for the appellate courts to develop a robust mistake of fact doctrine that holds decision-makers to a high standard of factual integrity and avoids the manifest injustices that come with erroneous fact-finding. However, the ground must also not encroach on the decision-maker’s sphere of autonomy, constitute a merits-review or unduly compromise the principle of finality.
The first chapter of this paper sets the scene by defining the mistake of fact ground. First, I employ the well-known case of Daganayasi to demonstrate the manifest injustice that mistakes of facts incur and the necessity of the ground’s existence. I will then detail the United Kingdom’s mistake of fact ground as laid out in E v Secretary of State for the Home Department.3 This decision offers a sensible basis for judicial intervention regarding factual matters and thus forms the foundation of much of my analysis throughout this paper.4 I will then distinguish the ground from similar grounds and principles in New Zealand administrative law.
1 Paul Craig “Judicial Review, Appeal and Factual Error” (2004) PL 788 at 788.
2 Daganayasi v Minster of Immigration [1980] 2 NZLR 130 (CA).
3 E v Secretary of State for the Home Department [2004] EWCA Civ 49, [2004] QB 1044, [2004] 2 WLR 1351. 4 Andrew McCue “Turning the Page: An Analytical Solution to the Law of Jurisdictional Error” (2021) 13 Ital J Public Law 55 at 80.
The second chapter outlines the ground’s historical development. It first sketches its development in the United Kingdom, from Edwards v Bairstow5 to Tameside6 to E.7 I then detail the ground’s introduction to New Zealand before demonstrating that the law has since stagnated and become very unclear, with divided approaches at the High Court level and a lack of authoritative and clear appellate-level guidance.
The third chapter discusses the normative tensions and policy rationales underlying the mistake of fact ground. These inform how conservative or liberal the ground’s elements should be, as well as whether the ground should exist at all.
Lastly, the fourth chapter outlines a novel mistake of fact ground for New Zealand. It presents three potential models: a conservative model, contextual model and Goldilocks model, and discusses the pros and cons of each, informed by analysis conducted in the previous chapter.
5 Edwards (Inspector of Taxes) v Bairstow [1955] UKHL 3; [1956] AC 14, [1955] 3 WLR 410 (HL).
6 Secretary of State for Education and Science v Tameside Metropolitan Borough Counsel [1976] UKHL 6; [1977] AC 1014, [1976] 3 WLR 641 (HL).
7 E, above n 3.
I The Mistake of Fact Ground
This chapter seeks to set the scene by defining the mistake of fact ground. It begins with an exploration of the 1979 Court of Appeal decision of Daganayasi,8 as this case aptly demonstrates the manifest injustice caused by factual errors occurring during the decision- making process and, therefore, the necessity of the ground’s existence in New Zealand administrative law. I will then explore the United Kingdom’s mistake of fact test as laid out in E,9 which provides the basis for much of my analysis including in the development of a mistake of fact ground for New Zealand. Lastly, I will distinguish the mistake of fact ground from similar grounds and principles.
This Court of Appeal decision centres around Joanna Daganayasi, who was being deported from New Zealand after she was declined permanent residency. Joanna appealed the deportation to the Minister of Immigration under s 20A of the Immigration Act 1964, which allowed the Minister to order that she not be deported if he was satisfied that the deportation would be unduly harsh or unjust due to exceptional circumstances of a humanitarian nature.11 The pleaded exceptional circumstance was that her three-year-old son, George, had a rare metabolic disease resulting in a grossly enlarged liver. This was a serious disease that required medication, recurring specialist medical supervision and a precise diet of frequent high-protein meals. Joanna argued that George would only receive proper treatment in New Zealand, with deportation to Fiji potentially entailing drastic health consequences.
This claim had a solid basis. Fiji did not have the protein sources necessary to continue George’s strict diet nor a metabolic clinic capable of providing adequate medical care. All of the doctors cited in the case agreed that deportation would put George’s health at significant risk; all except Dr Blake-Palmer, the medical referee appointed by the Minister to provide advice on the case. Dr Blake-Palmer’s “sanguine” report emphasised favourable progress in
8 Daganayasi, above n 2.
9 E, above n 3.
10 Daganayasi, above n 2.
George’s medical condition and discounted risks to George’s health in returning to Fiji.12 Joanna was at no point given the opportunity to read the report or respond to its prejudicial contents. Following the report, the Parliamentary Under-Secretary to the Minister of Immigration sent a letter of decision to Joanna, declining to grant the s 20A appeal.
The report featured several references to a long and detailed discussion with Professor Elliot, who was in charge of the metabolic clinic where George was a patient. Justice Cooke found that the report gave the impression that Professor Elliot “had been fully consulted about the wisdom of causing the boy to go to Fiji and was in general agreement with the substance of the report.”13 Further, the Parliamentary Under-Secretary’s letter of decision stated that he had “gone to some trouble to obtain the best and most recent medical advice available.”14 Justice Cooke held that such advice would at least include that of Professor Elliot’s clinic, which would have been passed on to the Under-Secretary through Dr Blake-Palmer’s report. As it turned out, Dr Blake-Palmer had not properly consulted Professor Elliot or the clinic. Professor Elliot had spoken with Dr Blake-Palmer via telephone as the report indicated, but was at no point asked whether it would be in the child’s medical interests to return to Fiji. In fact, Professor Elliot’s advice would have been quite contrary to the report – he believed that deportation would see George face “a greater susceptibility to illness and the likelihood of premature death.”15 Justice Cooke therefore found that a mistake of fact had occurred in the decision- making process.
Here, we have a prime example of what the mistake of fact ground would seek to remedy: a decision made on the basis of an incontrovertibly erroneous fact (that George’s clinic had been fully consulted and agreed with Dr Blake-Palmer’s report) and the significant injustice that faced the faultless Joanna and George as a result.
In his judgment, Cooke J advocated for establishing the mistake of fact ground and would have quashed the decision on this basis.16 This was unsuccessful, as the remaining members of the
12 Daganayasi, above n 2, At 139.
13 At 139.
14 At 139.
15 At 139.
Court, Richardson J and Richmond P, demurred from considering this ground of review.17 Fortunately, the case was saved on procedural fairness grounds, with the Court unanimously agreeing that natural justice required the report to be disclosed to Joanna before the decision was made due to its prejudicial content, the Immigration Act’s statutory scheme and the decision’s significant effect on Joanna and George’s rights.18
The 2009 enactment of the Immigration Act is much more restrictive than its 1964 counterpart. It expressly states that the decision-maker need not make inquiries or give reasons when determining whether to cancel a deportation order.19 Arguably, that would mean a court could no longer read in consultation rights as the Court of Appeal did in Daganayasi.20 Hence, if Daganayasi was tried today without an established mistake of fact ground, it is likely that Joanna would have been deported with young George suffering potentially severe health consequences despite the decision being materially affected by an incontrovertible and known mistake of fact.21 This demonstrates the necessity of the mistake of fact ground of review.
B Defining the Mistake of Fact Ground
It is helpful to begin this paper by detailing the British test for mistake of fact as was laid out in E v Secretary of State for the Home Department.22 This test provides an example of what a mistake of fact ground looks like and serves as the foundation for much of this paper’s analysis.
17 At 132 and 149.
18 At 143.
19 Immigration Act 2009, s 177.
20 Likely, but not certain. When granting interim relief in Dean v Chief Executive of the Ministry of Business, Innovation and Employment [2017] NZHC 588, the High Court ruled that it was strongly arguable that consultation rights could be read into the new Act’s s 177 (at [45]). This ruling appears strange given s 177’s restrictive nature, with it expressly not requiring the Minister to make inquiries or give reasons. A number of unique factors were present in this case that could have contributed to this unintuitive result, including an undertaking between the applicant and respondent regarding the deportation (discussed at [27]–[30]), and the fact that it was not a full trial. Regardless, even if consultation rights could have been read-in here, there are many instances where consultation rights are not available despite injustice occurring as a result. See Hanna Wilberg “Re-evaluating Mistake of Fact as a Ground of Judicial Review” (paper presented to Society of Legal Scholars Conference, Preston, September 2019) at 27.
21 It is possible, but unlikely, that the Court could have successfully applied other established grounds of review, such as relevant considerations. Due to length constraints, this is beyond the scope of this paper.
In E, the Court of Appeal held that the ground will be made out if there is a mistake of fact giving rise to unfairness, which requires the following elements:23
The mistake of fact ground concerns a narrow part of the decision-making process: the collection of incontrovertible pure facts.24 These include physical facts, being something that existed or occurred or something that did not exist or occur.25 For example, a physical pure fact would be that Sir Ashley Bloomfield was once the Director-General of Health for New Zealand. In mistake of fact cases, a physical pure fact may go to the availability of evidence. Pure facts may also be mental, taking the form of an opinion.26 For example, a mental fact would be that Dr Bloomfield believes that the COVID-19 vaccine is safe. In a legal context, mental facts are usually expert opinions, such as a scientific consensus on a particular topic. The category of pure facts as defined in this paper largely consists of primary facts, being those that can be determined without inference. However, it also includes incontrovertible factual inferences, for which there is a single objective conclusion or reality that any reasonable person would reach.
Almost every reviewable decision involves the collection of pure facts which form the basis of the decision. There are two main ways in which a decision-maker can get this process wrong. They could have either misunderstood the evidence in front of them and drawn an erroneous conclusion or had incomplete evidence due to ignorance of the availability of additional evidence.
23 At [66].
24 See also Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 3–4.
25 Tameside, above n 6, at 1030.
A court’s analysis of whether a decision-maker has erred in fact is to a correctness or substitutionary standard of review, meaning the courts will not afford deference to the decision- maker and will intervene where they believe that the fact is erroneous.27 This can be contrasted with a rationality standard of review that many similar grounds and principles employ, in which the court gives more deference to the decision-maker by only intervening if the conclusion reached is irrational to a Wednesbury ‘perverseness’ extent.28
For an error of fact to be reviewable, the existing pure fact must be “established” in that it is “uncontentious and objectively verifiable”. This requirement ensures that the courts can only look into fact-finding matters where the facts are incontrovertible. I will refer to this category of fact as ‘incontrovertible facts’.
In contrast, the courts cannot look into any subjective or contested matters, as these are the areas in which Parliament has entrusted the decision-maker to exercise their discretion and judgement.29 This includes cases where the facts are contentious such that determining the correct fact requires evaluating conflicting evidence; for instance, where there is conflicting evidence or conflicting expert opinion.30 This second element also precludes the courts from looking beyond the fact-finding process and into how facts are evaluated for the purpose of applying the law to them, as these matters are not ‘objectively verifiable’.31 I will refer to this category of fact as ‘factual evaluations’.
In limiting the scope of this ground to incontrovertible facts, the second element plays a very critical role in light of the ground’s undeferential correctness standard of review, which allows the court to intervene when it simply disagrees with the factual finding. This can be justified for incontrovertible facts, as ascertaining the correct reality is independent of the entity that
27 Hanna Wilberg "Mistake of Fact as a Ground of Review: Distinct and Defensible" in Joe Tomlinson and Anne Carter (eds) Facts in Public Law Adjudication (Oxford, Hart Publishing, TBC) 201 at 204.
28 Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223, [1947] EWCA Civ 1; and Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 6–7.
29 See Chapter Three, Heading B, Subheading One.
30 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 4.
31 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 4.
does so.32 In contrast, the more-deferential rationality standard of review raises the threshold for intervention such that the conclusion only becomes reviewable if it is irrational to a Wednesbury ‘perverseness’ extent.33 The Edwards v Bairstow principle, which reviews factual evaluations, features this standard of review.34 This is suitable as taking a reasonable view of the facts should not constitute a reviewable error even if the court would have taken another view, as this would usurp the role of the decision-maker.35 Therefore, in limiting the correctness review conducted under this ground to incontrovertible facts, the second element is the key lever preventing the courts from reviewing the merits of a decision.36
Under this element, the applicant or their advisors must not have been responsible for the mistake of fact. This lack of fault is a key element of fairness, which is the ground’s conceptual basis in the United Kingdom.37
The final element is materiality, which states that the mistake of fact must have played a material part in the decision-making process. This ensures the ground does not “invite litigants and the courts to delve into the minutiae of decision-making” such that “every aspect to the decision-maker’s understanding of the factual context becomes a potential target for review”,38 even those that did not affect the decision-making process.
32 Rebecca Williams “When is an Error not an Error? Reform of Jurisdictional Review of Error of Law and
Fact” (2007) PL 793 at 10. This issue is discussed in more depth in Chapter Three, Heading B, Subheading One. 33 Philip A Joseph Joseph on Constitutional and Administrative Law (5th ed, Thomson Reuters, Wellington, 2021) at [23.5.4].
34 See Chapter One, Heading C, Subheading Two. This principle was established in Edwards v Bairstow, above n 5.
35 New Zealand Fishing Industry Association Inc v Minister of Agriculture and Fisheries [1988] 1 NZLR 544 at
552. See Chapter Four, Heading B, Subheading One.
36 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 6–9. This is discussed in much greater detail in Chapter Three, Heading B, Subheading One.
37 E, above n 3, at [63]–[66].
38 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 20.
C Distinguishing the Mistake of Fact Ground
The mistake of fact ground can be easily confused with similar grounds and principles.39 Thus, it is worth highlighting their differences.
The jurisdictional fact ground (also known as precedent fact) serves as the orthodox approach to errors in pure fact-finding.40 Like the mistake of fact ground, the jurisdictional fact ground only concerns incontrovertible facts41 and analysis is conducted to a correctness standard.42 However, its scope is much narrower, being restricted to facts that are a condition precedent for the decision-maker’s power.
For example, in Khera v Secretary of State for the Home Department, the Home Secretary had the legislative power to deport ‘illegal entrants’ and attempted to deport Mr Khera, a migrant.43 Therefore, Mr Khera being an illegal entrant was a jurisdictional fact, as it had to be true in order for the Home Secretary to be able to use their powers to deport him. As Khera was deemed not to be an illegal entrant, the Home Secretary could not deport him.
This conceptualisation of an error of law was established by the House of Lords in Edwards v Bairstow44 and applied by the New Zealand Supreme Court in Bryson.45 It is often confused
39 This section does not discuss the ‘triple hurdle’ test established in Taafi v Minister of Immigration [2013] NZAR 1037 (HC) at [19], which also allows for the review of factual errors. It dictates that those wishing to seek leave to appeal factual findings of the Immigration and Protection Tribunal must overcome a ‘triple hurdle’ test. As the test only applies in such a narrow instance, it does not warrant further discussion. See also Doug Tennent, Katy Armstrong and Peter Moses Immigration and Refugee Law (3rd ed, LexisNexis, Wellington, 2017) at 416.
40 Jurisdictional fact has served as the orthodox approach since it was first established in Terry v Huntington
(1679) Hardres 480, 145 ER 557 (Exch).
41 Hawkins v Minister of Justice [1990] 3 NZLR 486 (HC) at 498–499.
42 Tamil X v Refugee Status Appeals Authority [2009] NZCA 488, [2010] 2 NZLR 73 at [259] and [271].
43 R v Secretary of State for the Home Department, ex parte Khawaja [1982] UKHL 5; [1984] AC 74, [1983] 2 WLR 321 (HL).
44 Edwards v Bairstow, above n 5.
45 Bryson v Three Foot Six Ltd [2005] NZSC 34, [2005] 3 NZLR 721. Note that this point is contentious among commentators. This view is consistent with Hanna Wilberg’s, which I find to be correct: see Hanna Wilberg
“Mistakes about Mistake of Fact: The New Zealand Story” (2017) 28 PLR 248 at 2–9. Phillip Joseph posits that
with the mistake of fact ground, but its differences are important. The Edwards v Bairstow principle is not concerned with pure fact-finding (i.e., incontrovertible facts). Instead, it is concerned with how facts are evaluated for the purpose of applying the law to them (i.e., factual evaluations). This highlights the importance of E’s second element (that the fact must be established), as this requirement distinguishes the mistake of fact ground from the Edwards v Bairstow principle.
The House of Lords in Edwards v Bairstow held that a factual evaluation constitutes an error of law where it is “so insupportable – so clearly untenable – as to amount to an error of law; proper application of the law requires a different answer.”46 The test has otherwise been framed as whether the factual evaluation was ‘open’ to the decision-maker in the circumstances.47 As previously noted, this analysis applies the more-deferential rationality standard of review, only allowing a court to intervene where the factual evaluation is irrational.
The theoretical basis for the Edwards v Bairstow principle is that where a decision-maker has come to a sufficiently irrational determination, the court will assume that the decision-maker has misconceived the law and therefore committed an error of law.48
The New Zealand Court of Appeal established this principle in Re Erebus Royal Commission,49 having derived it from the English Court of Appeal decision of Moore.50 The principle dictates that decisions will be reviewable where the decision-maker fails to base their decision upon evidence that “has some probative value”, being evidence that “tends logically to show the existence or non-existence of facts relevant to the issue to be determined,”51 as opposed to mere
Bryson held that a mistake of fact (of the E type) is not available as a separate ground of review, meaning a factual error is only reviewable where it meets the Edwards v Bairstow principle’s threshold and therefore amounts to an error of law: see Joseph, above n 33, at [23.5.4] and [23.5.5].
46 Bryson, above n 45, at [26].
47 Graham Taylor Judicial Review A New Zealand Perspective (4th ed, LexisNexis, Wellington, 2018) at [15.09].
48 Edwards v Bairstow, above n 5, at 36.
49 Re Erebus Royal Commission; Air New Zealand Ltd v Mahon [1984] AC 808, [1983] NZLR 662 (PC).
50 R v Deputy Industrial Injuries Commissioner, ex parte Moore [1965] 1 QB 456, [1965] 2 WLR 89.
51 Moore, above n 50, at 488.
speculation.52 As Lord Diplock colourfully put it, the decision maker “must not spin a coin or consult an astrologer”.53 The key difference between this principle and the mistake of fact ground is that the former concerns the existence of evidence and its probative nature in justifying factual evaluations, while the mistake of fact ground focuses on the correctness of evidence. However, the Erebus principle’s parameters are quite uncertain.54
D Conclusion
This chapter introduced the mistake of fact ground. I first demonstrated the ground’s necessity with reference to the injustice seen in Daganayasi. I then detailed the United Kingdom’s mistake of fact test as established in E. Lastly, I clarified the scope of the ground with reference to similar grounds and principles. The next chapter charts the mistake of fact ground's development, firstly in the United Kingdom, demonstrating how the E test came to be, and then in New Zealand, outlining its current legal status.
52 Li v Immigration and Protection Tribunal [2018] NZHC 174 at [57].
53 Moore, above n 50, at 488.
54 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 13.
II The Past: Historical Development of the Mistake of Fact Ground
This chapter explores the historical development of the mistake of fact ground, from its development in the United Kingdom culminating in E, before examining its introduction to New Zealand. It will conclude by demonstrating its current murky status in New Zealand administrative law, resulting from inconsistent High Court application and unclear appellate- level guidance.
A United Kingdom
The first landmark case in the development of the mistake of fact ground was the 1956 House of Lords’ decision in Edwards v Bairstow.55 While the principle established in this case is distinct from the mistake of fact ground,56 it set a precedent of the generosity with which courts will interpret the scope of its review powers regarding matters of fact generally, enabling future developments in the mistake of fact ground.57
Following this, Lord Denning MR provided a series of obiter dicta in favour of the ground’s development. In the 1972 decision of ASLEF, he contended that when a decision-maker “plainly misdirects himself in fact [...] it may well be that a court would interfere.”58 He repeated this assertion five years later in Laker Airways, stating that a decision is reviewable where it is “under the influence of a misdirection in fact”.59
The seminal 1977 Tameside decision was perhaps the most significant in the development of the mistake of fact ground.60 At issue was whether the Labour Education Secretary could hold that the local Conservative education authority’s decision to reverse Labour’s comprehensive schooling scheme was unreasonable (which would allow the Secretary to reverse the decision). The Secretary argued that the decision was unworkable, but this claim conflicted with expert
55 Edwards v Bairstow, above n 5.
56 See Chapter One, Heading C, Subheading Two.
57 See, for example, R (Alconbury Developments Ltd) v Secretary of State for the Environment, Transport and the Regions [2001] UKHL 23, [2003] AC 295 at [61]; and E, above n 3, at [49].
58 Secretary of State for Employment v Associated Society of Locomotive Engineers and Firemen (No 2) [1972] 2 QB 455, [1972] 2 WLR 1370 at 493.
59 Laker Airways Ltd v Department of Trade [1976] EWCA Civ 10; [1977] QB 643, [1977] 2 WLR 234 at 705–706.
60 Tameside, above n 6.
advice that the Secretary was not aware of. Scarman LJ held that judicial review is available where the decision-maker labours under a “misunderstanding or ignorance of an established and relevant fact”,61 specifically noting that such facts can include the existence of a body of expert opinion.62 The House of Lords endorsed these views on appeal.63
Subsequent development owes much to obiter comments from Lord Slynn in the House of Lords decisions of Criminal Injuries Compensation Board and Alconbury.64 In both cases, Lord Slynn stated that the Court could quash decisions due to a misunderstanding or ignorance of an established and relevant fact.
The 2001 England and Wales Court of Appeal decision of Haile did not play a significant role in the ground’s development.65 However, it will be referenced throughout this paper as a good example of how even minor factual errors can cause great unfairness.66 Thus, it is worth detailing the case here. The case concerned an asylum claim made on the basis that the claimant was a member of an anti-government coalition of political parties – the Coalition of Ethiopian Democratic Forces (“COEDF”) – and would thus be at risk of political persecution if he were to return to his home country of Ethiopia. Part of the claimant’s evidence was a witness statement from a fellow COEDF member, Captain Santayehu, purporting to confirm the claimant’s claims. The Special Adjudicator deemed the witness statement uncredible as he believed the claimant had stated that Captain Santayehu was a member of EPRDF (the Ethiopian People’s Revolutionary Democratic Front). The EPRDF was the governing party in Ethiopia and a staunch opponent of COEDF, making Captain Santayehu’s membership of COEDF very unlikely. In actuality, the claimant had asserted Captain Santayehu’s party to be EPRP (Ethiopian People’s Revolutionary Party), which was part of the COEDF coalition. This mistake of fact was only one of six reasons that led the Adjudicator to deem the claimant
61 At 1030.
62 At 1030.
63 At 1047. For a more comprehensive commentary on the similarities and differences between Tameside’s ruling in the Court of Appeal and House of Lords, see Taylor, above n 47, at [15.15].
64 R v Criminal Injuries Compensation Board, ex parte A [1999] UKHL 21; [1999] 2 AC 330, [1998] 1 WLR 277 at 345; Alconbury, above n 57, at 344; and Christopher Forsyth and Emma Dring “The Final Frontier: The Emergence of Material Error of Fact as a Ground for Judicial Review” in Christopher Forsyth and others (eds) Effective Judicial Review: A Cornerstone of Good Governance (Oxford University Press, Oxford, 2010) 245 at 250.
65 R (Haile) v Immigration Appeal Tribunal [2001] EWCA Civ 663, [2002] INLR 283.
66 See Chapter Four.
uncredible, however the Court held that it “must inevitably leave a sense of deep injustice in the appellant and it cannot confidently be said to have made no ultimate difference to the result.”67 The decision was quashed on this basis.
It must be noted that the mistake of fact ground did not experience linear progress towards judicial acceptance. Many cases pre-E held that mistakes of fact were only reviewable under more established grounds of review, such as the narrower jurisdictional fact ground, relevant considerations ground or various iterations of Wednesbury unreasonableness.68 Further, many decisions reflected a reluctance to entertain factual review at all, including arguable jurisdictional facts.69
Nevertheless, in 2004 came the authoritative decision of E, where the England and Wales Court of Appeal established mistake of fact giving rise to unfairness as a separate ground of review.70 The case concerned E and K, who were denied asylum by the Home Secretary, adjudicator and Immigration Appeal Tribunal in turn on the basis that neither had a well-founded fear of persecution upon returning to their respective countries of origin. The Court of Appeal granted leave to appeal because new evidence had emerged indicating that E and K’s fears of persecution were indeed well-founded.
In his judgment, Lord Carnwath consulted the existing jurisprudence concerning judicial review for factual error. He referenced Buxton and Wakins LJJ, who took the view that mistake of fact was only available under more established grounds of review.71 He also referenced several judgments that deemed factual review to be available on a broader basis, such as the aforementioned CICB, Alconbury and Tameside decisions, as well as a few planning cases.72 Lord Carnwath argued that these judgments could not be adequately explained by established
67 Haile, above n 65, at [25].
68 See, for example, Pulhofer v Hillingdon LBC [1986] UKHL 1; [1986] AC 484, [1986] 2 WLR 259 (HL) at 518; R v London Residuary Body, ex parte Inner London Education Authority, 24 July 1987, cited in E, above n 3, at [53]; and Wandsworth LBC v A [2000] 1 WLR 1246 (CA) at [59].
69 Forsyth and Dring, above n 64, at 247. See, for example, R v South Hams District Council, ex parte Gibb and Others [1954] EWCA Civ 3; [1955] QB 158 at 170; R v Radio Authority, ex parte Bull [1996] QB 169 at 170; and R v Secretary of State for the Home Department, ex parte Onibiyo [1996] EWCA Civ 1338; [1996] QB 768 at 785.
70 E, above n 3, at [66].
71 At [53].
grounds of review (relevant considerations, a failure to give adequate and intelligible reasons and a failure to base the decision on any evidence) as the factual errors concerned were beyond their scope.73 Instead, Lord Carnwath determined that this group of precedent indicated that factual errors could be dealt with under a separate ground of review based on the principle of fairness.74 He thus outlined the ground in this much-cited dictum:75
In our view, the time has now come to accept that a mistake of fact giving rise to unfairness is a separate head of challenge in an appeal on a point of law [...] First, there must have been a mistake as to an existing fact, including a mistake as to the availability of evidence on a particular matter. Secondly, the fact or evidence must have been “established”, in the sense that it was uncontentious and objectively verifiable. Thirdly, the appellant (or his advisers) must not have been responsible for the mistake. Fourthly, the mistake must have played a material (not necessarily decisive) part in the Tribunal’s reasoning.
B New Zealand
In New Zealand, early development of the doctrine witnessed persistent championing by Lord Cooke in the Court of Appeal, but he was often in the minority on this issue. This initially occurred in the 1980 decision of Daganayasi.76 As outlined earlier, this case saw Joanna Daganayasi and her son George being deported from New Zealand to Fiji after they failed to obtain permanent residency.77 The Minister of Immigration could cancel the deportation if he deemed it unduly harsh and unjust.78 Joanna argued that the deportation should be cancelled as George had a serious and rare medical condition that would not be adequately catered for in Fiji. However, the medical referee advising the Minister wrote a report that discounted risks to George’s health in returning to Fiji. Cooke J argued that the decision was invalid under the mistake of fact ground. He held that the report and letter of decision implied that George’s current clinic had been consulted on the wisdom of the deportation and was in general
73 At [62].
74 At [63]–[64].
75 At [66].
76 Daganayasi, above n 2.
77 See Chapter Two, Heading A.
78 Immigration Act 1964, s 20A(2).
agreement with the substance of the report.79 This was later discovered to be untrue.80 Cooke J thus held:81
The Minister has implied authority to delegate the function of making inquiries, but if as a result the Minister is led into a mistake and a failure to take into account the true facts, it is not right that the appellant should suffer.
The remaining members of the Court – Richardson J and Richmond P – declined to engage with this issue, stating that they preferred to leave the question open given the lack of comprehensive argument from counsel on the matter and the unsettled state of the ground’s law.82
President Cooke again advocated for the ground in New Zealand Fishing Industry Association, although this was in obiter as the ground was not made out.83 Casey J agreed,84 while Richardson J again explicitly left the question of the ground’s existence open with Somers J agreeing, citing the unsettled state of the law on this issue,85 and McMullin J did not comment on the matter. Similarly, Auckland City Council v Minister of Transport saw Cooke P briefly champion the ground in obiter with other members of the Court not commenting.86
However, the 1991 decision of Glaxo saw Cooke P’s advocacy gain the support of the rest of the Bench.87 Writing on behalf of Hardie Boys and Thorpe JJ, Cooke P held that if the decision was “shown to be based on a mistake or misconception of fact, the Court would no doubt be able to grant a remedy.”88 Ultimately, the ground was not made out, as the mistake was not deemed to be sufficiently material, having not been the “basis or probable basis of the Commissioner’s decision”.89 Hanna Wilberg contends that Glaxo “might be considered the
79 At 139.
80 At 139.
81 At 149.
82 At 132 and 149.
83 New Zealand Fishing Industry Association, above n 35 at 552.
84 At 552.
85 At 564.
86 Auckland City Council v Minister of Transport [1989] NZCA 70; [1990] 1 NZLR 264 (CA) at 293.
87 Glaxo Group Ltd v Commissioner of Patents [1991] NZCA 203; [1991] 3 NZLR 179 (CA).
88 At 184.
89 At 184.
strongest contender for binding appellate level authority” for the ground’s existence due to Cooke P’s authoritative statement on behalf of the entire Court.90 However, as Wilberg also points out, Cooke P cited Daganayasi as authority for the ground being available with little further discussion91 even though, as noted above, the majority did not support the ground there.92 Also, Cooke P did not discuss any elements other than materiality.93 Despite Glaxo’s relative strength as an authority for the ground’s existence, it has received very little attention from the courts and commentators.94
Following Lord Cooke’s departure from the Court of Appeal in 1996, the development of the mistake of fact ground has stagnated at all levels of the judiciary.
Firstly, the ground’s application at the High Court has been divided and inconsistent.95 Of the cases that support the ground’s existence, some have directly adopted the English approach, citing Tameside96 and E.97 Others cite New Zealand appellate authority, usually under the assumption that they are binding when, in fact, they are not (as will be discussed shortly).98 Others simply assume the ground’s existence without citing authority.99 The approach of these
90 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 18. Wilberg also notes here that mistake of fact was the only ground discussed by the Court, making the ground’s existence necessary for the decision. This further adds to its value as an authority.
91 Glaxo, above n 87, at 184.
92 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 18.
93 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 18–19.
94 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 18. As of writing, the case has been cited just twice.
95 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 13–14.
96 See, for example, D v M and Board of Trustees of Auckland Grammar School [2003] NZAR 726 (HC) at 737. 97 See, for example, Zafirov v Minister of Immigration [2009] NZHC 419; [2009] NZAR 457 (HC) at [78] and [80]–[81]; and Zhao v Legal Complaints Review Officer [2012] NZHC 3247, [2013] NZAR 193 at [66]–[83]. For more examples,
see Wilberg “Mistakes about Mistake of Fact”, above n 45, at 14–15, fn 69.
98 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 14. See, for example, Ellipse Institute Ltd v New Zealand Qualifications Authority [2012] NZHC 2083, [2012] NZAR 871 at [73]; and Oggi Advertising Ltd v Auckland City Council [2005] NZHC 1259; [2005] NZAR 451 (HC) at [4], [11] and [23]. For more examples, see Wilberg “Mistakes about Mistake of Fact”, above n 45, at 14, fn 68.
99 See, for example, Deliu v Hong and Legal Complaints Review Officer [2012] NZHC 158, [2012] NZAR 209 at [44]–[45]; and Deliu v Hong [2015] NZHC 492 at [22] and [40].
cases can be summarised by the dicta of Harrison J, who held that “[a] finding that [...] the decision is based upon a clear mistake of fact, provides a justiciable basis for setting it aside.”100
Meanwhile, several High Court cases have doubted or advocated against mistake of fact as a separate ground of review.101 For example, in A v Legal Complaints Review Officer, the Court strongly implied that factual review was only available under the no evidence ground and stated, “I am satisfied even to engage in these pleadings will envelop this judgment into examination of the facts, which would be close to a merits examination.”102
A third category of High Court cases deal with incontrovertible factual errors under more- established grounds of review, with some explicitly mentioning the possibility of employing the mistake of fact ground.103 A fourth category of cases makes up many of those that mention the mistake of fact ground.104 These judgments avoid engagement with the issue of the ground’s existence, merely stating that if the ground were available, its requirements were not fulfilled on the facts.105
One could attribute the inconsistent approaches at the High Court to the lack of clear and binding authority from the appellate courts since Lord Cooke’s departure. While several Court of Appeal decisions signal acceptance of mistake of fact constituting a separate ground of review, they are problematically unauthoritative,106 and the Court has denied or doubted the ground’s existence on other occasions.
100 Oggi, above n 98, at [11].
101 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 15. See, for example, Diagnostic Medlab Ltd v Auckland District Health Board HC Auckland, CIV 2006-404-4724, 20 March 2007 at [324]. For more
examples, see Wilberg “Mistakes about Mistake of Fact”, above n 45, at 15, fn 70.
102 A v Legal Complaints Review Officer [2013] NZHC 1100 at [37]; Wilberg “Mistakes about Mistake of Fact”, above n 45, at 15, fn 70.
103 See, for example, Double Eight Ltd v Disputes Tribunal at Manukau [2014] NZHC 1862 at [48]–[58]; Board of Trustees of Salisbury Residential School v Attorney-General [2012] NZHC 3348, [2013] NZAR 228 at [79]– [83]; and Wilberg “Mistakes about Mistake of Fact”, above n 45, at 14.
104 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 14.
105 For examples, see Wilberg “Mistakes about Mistake of Fact”, above n 45, at 14, fn 66.
106 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 13.
On several occasions, the Court of Appeal has declined to engage with the mistake of fact ground, instead characterising factual errors under more established grounds. For example, in the 2003 Squid Fishery decision, the challenge succeeded on alternative illegality grounds,107 so the Court stated that it did “not think it necessary to explore the not entirely easy jurisprudence as to the circumstances in which a fundamental mistake of fact warrants judicial review.”108 Such an apathetic approach is a great shame. The ground is undoubtedly complex and contentious; however, clarifying and settling “not entirely easy jurisprudence” is a crucial responsibility of the appellate courts. Similarly, the 2008 decision of Air Nelson presented an opportunity for the Court to firmly establish or reject the ground when it determined that a mistake of fact invalidated the decision.109 However, the Court preferred to characterise the error under the relevant considerations ground.110
Further, when the Court does engage with the ground, it falls short of laying out its elements in full. At best, the Court discusses one or two elements, and even then, they are rarely explicitly identified as requirements of the ground.111 A lack of a comprehensive test similar to that laid out in E deprives the High Court of critical guidance.
Another common weakness undermining appellate authority is that, like in Glaxo, the Court frequently cites non-binding authority under the view that it is, in fact, binding. For example, in Air Nelson, the Court appeared to adopt a mistaken view of Daganayasi, stating, “The Court held in that case that the Minister was labouring under a mistake of fact”.112 Meanwhile, the 2016 Charter Holdings decision saw the Court state that there existed “clear authority that error of fact may properly constitute a ground of review” but cited Taylor v Corrections and the authorities the Court cited there.113 This cannot be considered “clear authority”, as the scope of
107 Squid Fishery Management Company Ltd v Minister of Fisheries CA39/04, 13 July 2004 at [103].
108 At [104]; and Wilberg “Mistakes about Mistake of Fact”, above n 45, at 17.
109 Air Nelson Ltd v Minister of Transport [2008] NZCA 26; [2008] NZAR 139, [2008] BCL 309 (CA) at [53].
110 At [55]; and Wilberg “Mistakes about Mistake of Fact”, above n 45, at 18.
111 See Charter Holdings Ltd v Commissioner of Inland Revenue [2016] NZCA 499 at [68]–[69] and [75]; Air Nelson, above n 109, at [54]; and Taylor v The Chief Executive of the Department of Corrections [2015] NZCA 477; [2015] NZAR 1648 at [95]–[97].
112 Air Nelson, above n 109, at [52]; and Wilberg “Mistakes about Mistake of Fact”, above n 45, at 17.
113 Charter Holdings, above n 111, at [76] and [76], fn 44.
Taylor’s application is uncertain (for reasons outlined below) and Taylor cites non- authoritative sources: Daganayasi and a list of High Court decisions.114
Then there are cases that see novel unaddressed issues undermine their authoritative value. For example, in the 2015 case of Taylor v Corrections, the main ground argued was a breach of the New Zealand Bill of Rights Act 1990 s 14 right to freedom of expression, with the Court’s discussion devoted to whether the breach was demonstrably justified under s 5.115 As Wilberg notes, it is therefore difficult to tell whether the mistake of fact analysis was relevant as it is a ground of review in its own right or whether it was merely another factor that proved that the limits on the right were unjust per s 5.116
Finally, the Court of Appeal has occasionally doubted or denied the ground’s existence. For example, in the 2000 decision of Lewis, the Court held that factual review was only permitted “where the decision of fact is a condition precedent to the exercise of power or where the error of fact results in a decision which is unreasonable.”117 The Court appeared to take the same approach in the 2013 decision of Vulcan Steel when it held that “it would be an extreme step to set aside a judgment for error of law because of factual findings” aside from under the Edwards v Bairstow principle.118
The 2016 Supreme Court decision of Ririnui was a prime opportunity for the Court to authoritatively and conclusively clarify the law regarding the mistake of fact ground.119 The case concerned Landcorp (a state-owned enterprise) selling land that was subject to a pending Treaty of Waitangi claim by the applicant’s iwi. Normally, this might give the iwi preferential
114 Taylor, above n 111, at [94], fn 67; and Wilberg “Mistakes about Mistake of Fact”, above n 45, at 27–29.
115 At [60]–[91]. At the outset, the Court noted that all of the issues to be discussed fit into a wider balancing
practice between the s 14 right and “the need to maintain order in a prison”: at [2]. See Wilberg “Mistakes about Mistake of Fact”, above n 45, at 28–29.
116 Wilberg “Mistakes about Mistake of Fact”, above n 45, at 28.
117 Lewis v Wilson & Horton Ltd [2000] NZCA 175; [2000] 3 NZLR 546 (CA) at [92]; and Wilberg “Mistakes about Mistake of Fact”, above n 45, at 16.
118 Vulcan Steel Ltd v Wonnocott [2013] NZCA 398 at [12]. As Wilberg notes, the brevity of discussion regarding the mistake of fact ground means a few important questions were left unanswered: see Wilberg “Mistakes about Mistake of Fact”, above n 45, at 9–10.
119 Ririnui v Landcorp Farming Ltd [2016] NZSC 62, [2016] NZLR 1056.
opportunity to acquire the land. However, the Office of Treaty Settlements (OTS) advised Landcorp that the iwi’s historical Treaty claims had been finally settled. This advice influenced Landcorp’s decision to sell the land to a third party and the Minister’s decision not to intervene in the iwi’s favour. The OTS’s advice turned out to be erroneous.
Justice Arnold, on behalf of Elias CJ and with Glazebrook J in agreement, provided some clarity to the law of mistake of fact. The Court held that the factual error was “made in circumstances which would render a decision based on it susceptible to review”, therefore establishing clear authority for the existence of the mistake of fact ground of review.120 Further, the Court cited E and several High Court decisions applying E to support this conclusion.121 The Court also discussed each of E’s elements: the Court noted that the OTS’s advice was wrong, that this was beyond dispute (and therefore ‘established’), that the applicant was not responsible for the error and that the mistake was material.122 It is therefore arguable that the Court was adopting E here.
However, Ririnui did not reach its potential for several reasons. Firstly, engagement with the ground was minimal. While the above points were made, discussion was very brief, occupying a mere two paragraphs. This reduced the majority judgment’s clarity. For example, while Arnold J discussed each of E’s elements, he never explicitly stated that these matters were requirements of the ground, and a test was not otherwise outlined.123
Another critical flaw was that the Court declined to determine whether the mistake was one of fact or law. Justice Arnold noted that there was “scope for argument” on this issue,124 and that the distinction “may, in some instances, be significant in a judicial review context”.125 However, the Court concluded, “in the circumstances of this case, we do not regard the distinction as material” as “[e]ven if it is regarded as a mistake of fact, it is a mistake made in circumstances which would render a decision based on it susceptible to review”.126 While it was correct that the distinction would not have affected the outcome in this case, the distinction
120 At [54]. See also Wilberg “Mistakes about Mistake of Fact”, above n 45, at 21.
121 At [54], fn 58.
122 At [53].
123 Wilberg, “Mistakes about Mistake of Fact”, above n 45, at 22.
124 At [54].
125 At [54].
126 At [54].
is significant for the decision’s authoritative value. For example, as Wilberg notes, while there was an extensive discussion of materiality, this “did not distinguish between the fact and law versions of the mistake, and thus cannot readily be taken as authority on the materiality test for mistake of fact.”127
This leads to the third flaw of the majority judgment: the OTS’s error was not a mistake of fact. As Wilberg notes, the evidence points towards OTS misunderstanding the legal criterion governing the coverage of settlements concerning different claimant groups, which is an error of law.128 This means any engagement with the mistake of fact ground was somewhat flawed from the outset, seriously undermining Ririnui’s value as an authority.
These flaws render Ririnui a missed opportunity. It appeared to establish mistake of fact as an independent ground of review but failed to lay out its elements. Further, it refrained from determining whether the error was one of law or fact. As it turned out, the error was one of law, which undermined the case’s authoritative value. These flaws are particularly egregious given that the Court went out of its way to raise this issue, having introduced the possibility of the error being one of fact from the Bench.129
C Conclusion
While the United Kingdom had its ‘E moment’, New Zealand has yet to experience the same. The United Kingdom saw steady incremental progress, from Edwards v Bairstow to Tameside to E. Meanwhile, the ground was introduced to New Zealand through persistent championing by Lord Cooke, but has stagnated since his departure. Application at the High Court has varied widely, while the Court of Appeal appears to be in general agreement that mistake of fact is available as an independent ground of review (with a few exceptions) but has failed to lay down robust and authoritative precedent. And in the case of Ririnui, the Supreme Court followed suit: the Court voiced support for the ground, but a general lack of engagement rendered the decision a missed opportunity. The mistake of fact ground is very complex, but this makes appellate-level guidance all the more vital. Further progress in establishing the ground therefore
127 Wilberg, “Mistakes about Mistake of Fact”, above n 45, at 23.
128 Wilberg, “Mistakes about Mistake of Fact”, above n 45, at 23–26.
129 Wilberg, “Mistakes about Mistake of Fact”, above n 45, at 21.
requires appellate courts to engage with these complex issues directly when the opportunity next arises. The next chapter outlines the normative tensions and policy rationales underlying the mistake of fact ground for the purpose of arguing that this step should indeed be taken, with this analysis also informing the development of a mistake of fact ground for New Zealand in Chapter Four.
III The Present: Normative Tensions and Policy Arguments Underlying the Mistake of Fact Ground
Commentators have extensively discussed the normative tensions and policy arguments underlying the mistake of fact ground. We must explore these, as they inform how the ground’s parameters should operate, and indeed whether it should exist at all.
A Factors in Favour of the Ground
Factual errors undermine the principles of Parliamentary sovereignty and the rule of law. These are both fundamental constitutional principles generally and make up the constitutional foundations of the courts’ supervisory jurisdiction.130
Parliamentary sovereignty requires decision-making powers to be used according to Parliament’s intent as dictated in legislation. A system of common law rules has developed to assist in this task.131 One of these sees the courts assume that Parliament intends the exercise of discretion to promote the policy and objects of the Act under which it was conferred,132 and as Edward Lui correctly notes, “the policy and objects of each statute can only be promoted if the decision is made on the basis of truth, and not mistakes.”133
Let us return to Daganayasi as an example. We can infer that the policy and object of s 20A of the Immigration Act 1964 was to prevent deportations that caused unduly harsh and unjust consequences to human welfare. Due to the mistake of fact, Joanna and George were (almost) deported, which would have likely imposed unduly harsh and unjust consequences on George’s welfare. Therefore, the mistake of fact (almost) led to the policy and object of this provision, and therefore Parliament’s intent, not being fulfilled. Hence, irremediable mistakes of fact can undermine Parliamentary sovereignty.
130 HWR Wade and CF Forsyth Wade & Forsyth’s Administrative Law (10th ed, Oxford University Press, Oxford, 2009) at 17–18 and 21. See also Senior Courts Act 2016, section 3(2).
131 Wade and Forsyth, above n 130, at 18.
132 Padfield v Minister of Agriculture, Fisheries and Food [1986] A.C. 997 at 1030; and Harry Woolf and others
De Smith’s Judicial Review (8th ed, Sweet and Maxwell, London, 2018) at [5-001].
133 Edward Lui “‘Fairness’” for Mistake of Fact: A Mistake in Fact” (2020) 3 PL 428 at 440.
As a corollary, mistakes of fact also violate the rule of law. This principle is undermined where “powers [are] used in ways which Parliament is not thought to have intended”, as this means the power has not been carried out according to law.134
Daganayasi demonstrates the significant and avoidable injustice that occurs in the absence of an established mistake of fact ground.135 Joanna and George would have faced potentially life- threatening consequences, despite the mistake not being their fault and them having no chance to correct it. Further, as with all qualifying mistake of fact cases that reach the courts, everyone involved knew that the mistake occurred and that it bestowed injustice upon Joanna and George.136 Nevertheless, as aforementioned, if the case were tried today and the mistake of fact ground was not available, the court would not have been able to intervene. Further, even if the Minister did not base his decision on the mistake, it still renders the decision-making process unfair, undermining the legitimacy of the decision overall. This highlights how vital the ground is in ensuring the courts can give effect to the principles of justice and fairness in judicial review.
Judicial review promotes good governance, providing the guardrails of good decision-making and thereby “induc[ing] [decision-makers] to act wisely and well.”137 This is especially the case for public officials, for whom judicial review functions as a feedback mechanism with the goal of preventing decision-making errors from occurring more than once.138 However, carrying this out effectively requires the grounds to be sufficiently comprehensive. A mistake of fact ground is required, as factual errors lead to flawed decision-making. Therefore, an established mistake of fact ground is necessary to ensure good governance.
134 Wade and Forsyth, above n 130, at 17-18; and Wilberg “Mistake of Fact as a Ground of Review”, above n 27, at 203. See also T v Immigration and Protection Tribunal [2012] NZHC 1871 at [22].
135 See Chapter One, Heading A.
136 This is due to the test’s second limb (that the fact was established, being uncontested and objectively verifiable).
137 Waitakere City Council v Lovelock [1997] 2 NZLR 385 (CA) at 417.
138 This is evidenced in by such resources as Crown Law The Judge Over Your Shoulder (2019), a guide that breaks down administrative law principles to “inform and improve the quality of decision-making in
government.” (At 4).
The well-established jurisdictional fact ground displays the courts’ willingness and ability to correct incontrovertible mistakes of fact.139 All that differentiates this ground from the mistake of fact ground is its scope, with the former only reviewing facts that are a condition precedent for the decision-maker’s power.140 This distinction is based on the antiquated doctrine of jurisdiction, which previously restricted the court’s ability to review errors of law on the same basis. The jurisdictional error doctrine was abandoned by the House of Lords some 55 years ago with the decision of Anisminic,141 with many since deeming it to be, at best, as the United Kingdom Supreme Court noted, “artificial and technical”,142 and at worst, vulnerable to manipulation for pragmatic reasons.143 It is therefore strange that courts continue to determine their ability to review factual errors according to an outdated divide that is “difficult if not impossible to draw in any principled manner.”144 Establishing a mistake of fact ground would enable the reviewing of non-jurisdictional facts,145 therefore finally ridding the law of the doctrine of jurisdiction.146
World politics has plunged into a ‘post-truth era’ in which the line between truth and falsehood has become blurred.147 A key element of this post-truth era is the rapid rise of disinformation,
139 See Chapter One, Heading C, Subheading One.
140 See Chapter One, Heading C, Subheading One.
141 Anisminic Ltd v Foreign Compensation Commission [1968] UKHL 6; [1969] 2 AC 147, [1969] 2 WLR 163; Forsyth and
Dring, above n 64, at 247.
142 R (Cart) v Upper Tribunal [2011] UKSC 28; [2012] 1 AC 663 at 111; Michael Fordham Judicial Review Handbook (7th ed, Hart, Oxford, 2020) at 617.
143 Paul Craig Administrative Law (9th ed, Thomson Reuters, London, 2021) at [16–002]; Williams, above n 32, at 3.
144 Craig “Judicial Review”, above n 1, at 797.
145 This is due to the second limb (that the fact be established, being uncontentious and objectively verifiable). Non-jurisdictional factual evaluations are reviewable under the Edwards v Bairstow principle: see Chapter Two, Heading B, Subheading Two.
146 Fordham, above n 142, at 630.
147 Guðmundur Hálfdanarson and Maximilian Conrad "Concluding Reflections on Europe in the Age of Post- truth Politics" in Maximilian Conrad and others (eds) Europe in the Age of Post-Truth Politics (Springer Nature, Cham, 2023) 247 at 248.
exacerbated by the ease of content dissemination in the digital age.148 Decision-makers are not immune from this. They are constantly bombarded with information and misinformation and must conduct the challenging exercise of distinguishing the two. Such is the reality of modern decision-making.
Further, the stakes have never been higher. The issues humanity is currently grappling with are existential in nature, with climate change and pandemics being good examples. It is therefore essential for decisions to have an accurate factual basis.149 A mistake of fact ground would act as a corrective device to counter disinformation and steer decision-makers towards evidenced- based decisions. This is especially important regarding members of the public service, for whom judicial review functions as a key feedback mechanism. A robust mistake of fact ground would send a clear message that it is not acceptable nor lawful for a public servant to act or advise a Minister however they wish if it would mean operating on the basis of a factual error.
B Factors Against the Ground
Judicial review is strictly concerned with the correctness of the decision-making process.150 It is not concerned with assessing the correctness of the decision itself, or as it is commonly called, ‘conducting a merits-review’.151 A potential issue with the mistake of fact ground is that it may allow courts to encroach on a merits-review, as factual findings are arguably part of the substantive decision.152 One could assert that Parliament intended the fact-finding process to come within the decision-maker’s sphere of autonomy due to their constitutional legitimacy
148 Turgay Yerlikaya and Seca Toker Aslan “Social Media and Fake News in the Post-Truth Era: The
Manipulation of Politics in the Election Process” (2020) 22(2) Insight Turk 177; Hálfdanarson and Conrad, above n 147, at 248.
149 Janet McLean “Administration in the constitution: Disaggregating power for accountability purposes” in Carol Harlow (ed) A Research Agenda for Administrative Law (Edward Elgar Publishing, Cheltenham, 2023) 65 at 81.
150 Chief Constable of the North Wales Police v Evans [1982] UKHL 10; [1982] 1 WLR 1155 (HL) at 1174; Joseph, above n 33, at [23.3.3].
151 Evans, above n 150, at 1174; Joseph, above n 33, at [23.3.3].
152 See, for example, A, above n 102, at [37]; Matthew Groves “Judicial Review of Administrative Action in the High Court of Australia” (2008) 33 Queen’s Law J 327 at 360; and Marie Demetriou and Stephen Houseman "Review for Error of Fact – A Brief Guide" (1977) 2(1) Jud Rev 27 at 27–28.
and institutional capacity relative to courts.153 If this is the case, courts should not intervene in this process (other than under more established grounds), as doing so would undermine Parliamentary sovereignty.154 In this way, it would also undermine the rule of law, as it would see the courts making decisions without legal authority.
But should findings of incontrovertible fact be considered part of the merits of a decision? Hanna Wilberg argues that a step in the decision-making process is only part of the merits if it requires the decision-maker to exercise their discretion or judgement in “evaluat[ing] competing options for the purposes of choosing between them.”155 This is a persuasive definition: it is only where there is room for subjectivity and discretion that a decision-maker might reach a different decision from the courts without it being unreasonable. Whereas, where the concerned fact is objective and incontrovertible, the decision-maker cannot reach a different conclusion without it being objectively wrong.156 Further, where there is room for discretion, the intended decision-maker’s greater access to expertise, experience and robust decision- making processes will give them an advantage over the courts and the public may need to hold the decision-maker to account (if they disagree with the conclusion reached). However, where there is only one reasonable conclusion, the courts are as well-placed as the decision-maker, and there is no such need for democratic accountability, as reasonable members of the public could not disagree with the conclusion reached. Institutional characteristics would thus be immaterial.
As already discussed, E’s second element (that the fact be established, being uncontentious and objectively verifiable) constrains the ground to only reviewing factual determinations for which there is a single objective conclusion or reality that any reasonable person would reach.157 It does not concern matters involving evaluating competing options and choosing between them. Therefore, it should not be considered part of the merits of a decision. Nevertheless, the
153 Pulhofer, above n 68, at 518; Joseph, above n 33, at [23.5.2]; Demetriou and Houseman, above n 152, at 27; and Wilberg “Mistake of Fact as a Ground of Review”, above n 27, at 203.
154 Wilberg “Mistake of Fact as a Ground of Review”, above n 27, at 203; and Demetriou and Houseman, above n 152, at 27.
155 Wilberg “Mistake of Fact as a Ground of Review”, above n 27, at 213. For a similar view, see Forsyth and Dring, above n 64, at 258.
156 Williams, above n 32, at 10.
157 See Chapter One, Heading B, Subheading Two; and E, above n 3, at [66].
potential for enabling a merits-review is a valid concern that I will consider further when I discuss the ideal parameters for a mistake of fact ground in New Zealand.158
Courts frequently deal with factual errors by pressing them into more established grounds of review, such as relevant considerations, natural justice and unreasonableness.159 One might therefore wonder whether New Zealand administrative law even requires a separate mistake of fact ground to properly give effect to the interests laid out above, or whether this is already achieved under the status quo.160
Wilberg posits that the mistake of fact ground has substantive novelty, as, unlike other established grounds, it allows the courts to analyse just one step in the decision-maker’s reasoning (a finding or understanding as to a particular fact), even if the fact does not directly go to an applicable legal test or the decision itself.161 Therefore, she concludes, “In order to accommodate all claims that are available under the new ground of mistake of fact, each of the other established grounds would require expansion into an area that is not a well-established part of the ground.”162 This in itself warrants a separate mistake of fact ground, in order to remedy injustice caused by factual errors that does not come within the scope of other grounds.163
However, even if Wilberg is wrong, or if she becomes wrong due to future administrative law developments, substantive overlap with other grounds is not a sufficient reason to reject the mistake of fact ground. This is because, as Taylor notes, substantive overlap between grounds is “a natural incident of judicial review”.164
158 See Chapter Four.
159 See Chapter Two, Heading B, Subheadings One, Two and Three.
160 Taylor frames this issue as determinative of whether it would be “useful and appropriate to recognise [mistake of fact] as a separate ground”: see Taylor, above n 47, at [15.14].
161 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 20–33.
162 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 32.
163 The injustice seen in Daganayasi, above n 2, for example, would probably not have been remediable under established grounds. See Chapter One, Heading A.
164 Taylor, above n 47, at [11.04]
The more important question is whether the mistake of fact ground has conceptual utility. It undoubtedly does, as where an incontrovertible factual error occurs, analysis under the ground gives the most accurate, intuitive and simple description of what went wrong and why this should be rectified, as opposed to the circuitous reasoning employed when explaining mistakes of fact under other grounds. This is very beneficial in a judicial setting. For example, instead of the courts using the language and logic of relevant considerations, which requires finding that a statute implied a duty to consider a given fact and that this duty was not fulfilled, they could remove several steps from their analysis and explain that the decision was materially influenced by a mistake of fact. Simplifying administrative law reasoning brings obvious benefits in making the law more certain and accessible. It also benefits the operation of the public service, whose members look to judicial review to guide them towards proper decision- making processes.
Further, classifying factual errors under the mistake of fact ground makes it easier to discuss the ground according to its novel underlying interests. For example, much of Chapter Four attempts to determine how material a factual error should be in the decision-making process before it constitutes a reviewable error. As we shall see, this involves a detailed analysis of the novel interests the ground seeks to uphold, such as procedural justice and finality. However, if factual errors were reviewed under the relevant considerations ground, construing Parliament’s intent becomes the primary issue. This is an entirely different conversation – one that would not prioritise concerns of procedural justice and finality. Therefore, establishing a mistake of fact ground would allow for more transparent and constructive discussions regarding the review of factual errors.
It should be noted that the Edwards v Bairstow principle could be interpreted as having extensive substantive and conceptual overlap with the mistake of fact ground such that the principle could adequately fulfil the role of the ground. This issue is dealt with in the next chapter, with this argument taking the form of the conservative model.
Those who argue against the mistake of fact ground often cite reasons of finality.165 This is a valid concern, as a lack of finality brings increased uncertainty, prolonged disputes, delays in enacting justice, and wasted resources. Due to this, the finality principle carries some weight.166 For example, Wilberg notes that general appeals see fact-finding generally entrusted to a first instance court or primary decision-maker. Findings are subject to supervision on appeal, but full de novo rehearings of evidence are very rare.167 However, finality is in obvious tension with other valuable principles, namely the interests of justice and fairness. This is why de novo rehearings of evidence, while being uncommon, are not prohibited, as these may be required to prevent an unjust outcome.168 Similarly, where the mistake of fact ground applies, the interests of justice (as outlined earlier in this chapter) again likely outweigh the need for finality. This is especially because the ground is limited to incontrovertible factual errors, reducing the need for expending resources on extensive fact-finding and making injustices more demonstrable and therefore more significantly felt.
Alternatively, one may argue that adopting the ground would ‘open the floodgates’ by bringing in a wave of judicial review that would overwhelm the courts, thereby further legitimising finality concerns. Such an argument was posited in the context of factual appeals as early as the 1956 decision of Edwards v Bairstow.169 However, Chrisopher Forsyth, who has written extensively on the operation of the ground in the United Kingdom since E, argues that “the weakness of the [floodgates] argument [...] needs to be stressed.”170 He notes that the case law has revealed that E’s elements and the discretion infused in the judicial review application process are sufficiently onerous to “provide a certain bulwark against the potential flood”.171 And regardless, even if the establishing of this ground were to ‘open the floodgates’, so be it. The mistake of fact test (especially that recommended in Chapter Four) carefully balances the
165 Shaheen v Secretary of State for the Home Department [2005] EWCA Civ 1294, [2006] Imm AR 57 at [28]; MT (Algeria) v Secretary of State for the Home Department [2007] EWCA Civ 808, [2007] HRLR 41 at [69]; and Forsyth and Dring, above n 64, at 252.
166 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 16–17.
167 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 17. 168 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 17. 169 Edwards v Bairstow, above n 5, at 32.
170 Christopher Forsyth “Error of Fact Revisited: Waiting for the ‘Anisminic Moment’” (2018) 23 JR 1 at [39].
171 Forsyth, above n 170, at [39].
competing interests noted above such that successful cases deserve justice. Also, we should be expecting our public officials to operate from a correct factual basis regardless.
It should be noted here that the 2005 England and Wales Court of Appeal decision of Shaheen raised concerns that enabling review for mistakes as to the availability of evidence may allow for the re-litigating of factual disputes merely because new non-probative evidence emerges.172 This specific concern is addressed in Chapter Four.
C Conclusion
Weighing these arguments reveals that an established mistake of fact ground would play a valuable role in New Zealand’s administrative law landscape. In order to finally unshackle the review of facts from the doctrine of jurisdiction, to give courts a tool to ensure significant decisions are based on correct facts in the post-truth era, and to give effect to the principles of justice, fairness and good governance, as well as the fundamental constitutional principles that underly judicial review, the ground must be established. However, there are valid concerns that must be kept in mind. One must ensure that the ground does not encroach on a merits-review, unduly compromise finality, or ‘open the floodgates’ and inundate the courts. These matters will inform the development of a new mistake of fact doctrine in the next chapter.
172 Shaheen, above n 165, at [27]–[28].
IV The Future: A Mistake of Fact Ground for New Zealand
This chapter outlines a new ground of review that could apply to mistakes of incontrovertible fact in New Zealand. E serves as the starting point due to its success in the United Kingdom in providing “sensible” and “proportionate” supervision of incontrovertible facts.173 E’s first element (that there was a mistake as to an existing fact) and third element (that the applicant must not have been responsible for the mistake) are relatively uncontentious and can be adopted as is. However, there is room for discussion regarding E’s second element (that the fact was established, being uncontentious and objectively verifiable) and fourth element (that the mistake played a material part in the reasoning) and how one might interpret or modify them to strike the right balance in light of the policy reasons and normative tensions outlined in the previous chapter.
Three models are put forward for consideration. The first is a conservative model that would see the ground subsumed into the Edwards v Bairstow conceptualisation of an error of law and therefore employs an onerous materiality threshold. The second is a contextual model that employs a variable approach to both limbs, with the standard of review varying based on the complexity of the facts and the materiality threshold varying based on the type of factual error or the injustice caused by it. The third is a Goldilocks model, which more closely adheres to Lord Carnwath’s formulation in E. I will outline and critique each in turn.
A The Conservative Model
The conservative model abandons the idea of a separate ground of review for mistakes of incontrovertible fact. Instead, mistakes of incontrovertible fact would be reviewed under the Edwards v Bairstow principle. This is arguably conceptually consistent, as the Edwards v Bairstow principle allows for intervention where a finding of fact renders a decision untenable, and mistaken findings of incontrovertible fact certainly render decisions untenable (if it is sufficiently material). Mistakes of fact therefore do not only fail a correctness standard of review as has been discussed throughout this paper,174 but they also fail the Edwards v Bairstow principle’s rationality standard of review, as it would be irrational for a decision-maker to reach an untenable decision due to an error of incontrovertible fact.
173 McCue, above n 4, at 80.
174 See Chapter One, Heading B.
Further, the courts occasionally employ this approach. For example, in Tan v New Zealand Registered Architects Board,175 the High Court dealt with three mistakes of incontrovertible fact under an “untenable findings” header176 citing only Bryson (a case that only discussed Edwards v Bairstow)177 and using Bairstow-like language in querying whether the factual findings were ones the defendant was “entitled to make”178 or whether they were “so untenable that [they] constitute[] an error of law”.179 It is plausible that the courts will increasingly adopt this approach given that the Edwards v Bairstow principle is more established than the mistake of fact ground. This is evidenced by the Supreme Court’s explicit and comprehensive endorsement in Bryson.180 Meanwhile, judicial engagement with the mistake of fact ground has been less comprehensive and clear.181
As alluded to above, the Edwards v Bairstow principle requires factual errors to be sufficiently material to render the resulting decision untenable such that “proper application of the law requires a different answer.”182 This model therefore employs an onerous materiality threshold, as mistakes would need to be so material that they cause a different decision to be reached to a ‘but-for’ extent. There is case precedent for such an approach when dealing with mistakes of incontrovertible fact. For example, the Court of Appeal adopted a similarly stringent materiality threshold in Glaxo, stating that “[o]nly a mistake sufficiently material to be described as the basis or probable basis of the [decision-maker’s] decision” would vitiate the decision.183 Meanwhile, the High Court has applied a stringent materiality threshold in several
175 Tan v New Zealand Registered Architects Board [2016] NZHC 2276.
176 At [137].
177 At [130], fn 67.
178 At [137].
179 At [131].
180 Bryson, above n 45, at [26]. 181 See Chapter Two, Heading B. 182 Bryson, above n 45, at [26].
183 Glaxo, above n 87, at 184. It should be noted that the court deemed the hearing to be a “preliminary stage”, as there were other avenues through which Glaxo could oppose this prima facie case that were yet to be engaged. This likely affected the materiality threshold applied. However, in Attorney-General v Moroney [2000] NZHC 1298; [2001] 2 NZLR 652, the High Court applied this threshold (as noted subsequently), citing Glaxo, despite alternative remedies not being available. This issue is also discussed in Chapter Four, Heading B.
cases.184 For example, in Moroney, the Court cited Glaxo and employed its “basis or probable basis” test;185 and in McGrath, the Court stated that “[t]he factual error must be fundamental or pivotal to the decision.”186
Edwards v BairstowSubsuming the mistake of fact ground within the Edwards v Bairstow principle ignores a significant difference between the two. Specifically, the mistake of fact ground is concerned with procedural fairness. This is evidenced by the original formulation of the ground in E, where Lord Carnwath held that mistakes of fact give rise to unfairness (“an aspect of procedural irregularity”) as they see the claimant deprived of a “fair crack of the whip” through no fault of their own.187 Further, commentators note that procedural fairness grounds are concerned with promoting “the quality, accuracy and rationality of the decision-making process”, thereby enhancing its legitimacy and “instil[ing] a sense of fairness.”188 Ideas of legitimacy and fairness speak to procedural justice, which ensures the decision’s integrity through the applicant subjectively experiencing a fair process.189 The mistake of fact ground clearly gives effect to these goals, as it is concerned with ensuring that information in front of decision-makers is complete and accurate, thereby improving the subjective fairness experienced by the applicant. The ground could also belong to the illegality head of review,190 but this does not weaken the ground’s procedural fairness aspect, as the heads of review are not mutually exclusive.191
184 As well as the following examples, see Deliu v Hong and Legal Complaints Review Officer, above n 99, at [45]; and Northern Inshore Fisheries Company Ltd v Minister of Fisheries HC Wellington CP 235-01, 4 March 2002 at [48]–[49].
185 Moroney, above n 183, at [81].
186 McGrath v Minister of Justice [2014] NZHC 3279, [2015] NZAR 122 at [37].
187 E, above n 3, at [64] – [65]. It should be noted that it is contested whether Lord Carnwath basing the UK mistake of fact test in unfairness is proper, but these concerns go more to the prior case precedent that justified this classification rather than its conceptual appropriateness. See Lui, above n 133. In contrast, see Craig
“Judicial Review”, above n 1, at 798–800.
188 Joseph, above n 33, at [25.1]; and Woolf and others, above n 132, at [6-001].
189 Woolf and others, above n 132, at [6-002].
190 See Lui, above n 133.
191 New Zealand Fishing Industry, above n 35, at 552 (Cooke J: the heads of review “merge[] rather than being discrete”); Joseph, above n 33, at [22.11]; and Taylor, above n 47, at [11.02].
In contrast, the Edwards v Bairstow principle, being a conception of an error of law, is solely an illegality ground. It is only concerned with substantive justice, which, in the context of judicial review, assesses whether the decision was objectively lawful (the focus of the illegality grounds).192
This key conceptual difference is reflected in how the two function in practice, specifically in their materiality threshold. The Edwards v Bairstow principle sets an onerous ‘but-for’ threshold, while the mistake of fact test employs a less demanding threshold which does not require the error to have been decisive. This allows the mistake of fact ground to target a different part of the decision-making process: the preliminary fact-finding procedure the decision-maker undertakes before they derive factual evaluations and apply the law to them. Errors made during this stage may not affect substantive justice, but they certainly affect procedural justice – it makes the decision-making process unfair, therefore harming the decision’s perceived legitimacy. This mirrors other procedural fairness grounds, which can also be successfully invoked where unfairness occurs, even if it had no impact on the ultimate decision. For example, the apparent bias ground does not require finding that bias materially affected the decision. It only requires finding a sufficient risk of bias influencing the decision, as this means justice is not being seen to be done.193 The ground would lose this ability under the more onerous materiality threshold.
Therefore, while it might seem viable for the Edwards v Bairstow principle to cover incontrovertible factual errors, this would result in the court losing its ability to address the other harmful aspect of mistakes of fact: the procedural unfairness they cause. This unfairness is in and of itself worthy of correction.194 It cannot be allowed to exist in the context of executive and administrative decision-making. Haile demonstrates this adequately.195 In that case, the mistake was only one of six factors leading to an adverse credibility finding towards the applicant, resulting in his asylum claim being denied. Under the conservative model, the mistake would not be sufficiently material, as it did not directly go to the ultimate legal test of whether the claimant was at risk of persecution if he returned to Ethiopia, and therefore did not cause the decision to a ‘but-for’ extent, nor result in the decision being untenable. Yet, as the
192 Woolf and others, above n 132, at [6-002].
193 Joseph, above n 33, at [25.5.1].
194 See Chapter One, Heading A and Chapter Three, Heading A, Subheading Two.
195 Haile, above n 65. See Chapter Two, Heading A.
Court noted, this mistake “inevitably [left] a sense of deep injustice”196, with Lui noting, “few would agree the injustice should be left untouched.”197
Polycentric decision-making involves “a large and complicated web of interdependent relationships, such that a change to one factor produces an incalculable series of changes to other factors.”198 In this context, the conservative model’s onerous materiality threshold would be near-impossible to satisfy, as the courts are ill-equipped to assess materiality in such cases.199 I will employ a hypothetical case to demonstrate this.
Suppose the Director-General of the Department of Conservation is told that she will have to cut spending in her department by 6.5% to implement the incoming Government’s campaign promises. One measure she decides to take is decreasing personnel costs by $12 million. The Director-General receives a list of staff members identified as superfluous. At the bottom of this list is the sum of their collective salaries, which equals around $12 million. The public servants are then made redundant. Suppose further that the list the Director-General received miscalculated the redundant employees’ cumulative salaries, which actually only totalled $10 million. If judicial review were sought, how material would a court deem the mistake?
In this case, determining materiality would be near-impossible, as the decision is polycentric. Had the Director-General been aware that the redundancies would have been insufficient, she may have made a different decision (for example, by dismissing staff on higher salaries), in which case the mistake would have been very material. However, she might have instead opted to cut more spending elsewhere (say, the Department’s operating costs), in which case the mistake would have been immaterial. As there are too many moving parts, the Court cannot determine to what extent the mistake influenced the overall decision.
The courts being ill-equipped to determine materiality in the context of polycentric decision- making could justify both a more or less onerous materiality threshold. One could argue that
196 At [25].
197 Lui, above n 133, at 430.
198 Jeff A King “The Pervasiveness of Polycentricity” (2008) PL 101 at 102.
the more onerous ‘but-for’ threshold should be applied, as a lesser threshold would occasionally see the decision quashed when the factual error was not in fact material. These false positives would result in the decision-maker remaking the decision (but this time on a correct factual basis), undermining finality and creating great inefficiency. This is perhaps why the courts have typically invoked deference in cases involving polycentric decision-making.200
However, I would argue that the indeterminacy of materiality in such cases warrants abandonment of the more onerous threshold. Where materiality is indeterminate, a ‘but-for’ threshold would see applicants occasionally fail to invoke the ground despite the factual error indeed being decisive in the decision-making process. And critically, this false negative is more harmful than a false positive. This is due to the significant injustice and unfairness that would ensue, with the applicant left with no means of redress in most cases.201 These critical principles outweigh the need for finality and other potential inefficiencies.
Additionally, a ‘but-for’ materiality threshold would undermine the ground’s ability to function as a critical feedback mechanism in correcting decision-making and ensuring good governance. Decision-makers are not conscious of factual errors when they make decisions. It is therefore plausible that the mistake of fact ground brings the error to the decision-maker’s attention in many instances, thereby correcting the decision-making process. However, if an onerous threshold is established, potential applicants may feel that it is not worth the great time, effort and costs to judicially review polycentric decisions due to the low chances of success. Hence, many errors would go uncorrected, greatly compromising good governance. This is especially important in the modern decision-making context, which features the post-truth era that increases the likelihood of factual errors permeating decision-making and significant polycentric issues such as climate change and pandemics.202
One should also remember that false positives can only occur when decision-makers make a factual error. This should be a rarity – we should expect our decision-makers to always operate from a correct factual basis. While the post-truth era may have made this more difficult,203
200 Smith v Fonterra Co-Operative Group Ltd [2020] NZHC 419 at [26]. See also Smith at [98(f)] and Students for Climate Solutions Inc v Minister of Energy and Resources [2022] NZHC 2116 at [43].
201 See Chapter Three, Heading A, Subheading Two. See also Chapter One, Heading A.
202 See Chapter Three, Heading A, Subheading Four.
203 See Chapter Three, Heading A, Subheading Four.
decision-makers have more access to expertise, experience and robust decision-making processes than ever before. The possibility of having to remake decisions due to factual errors, even when these errors did not affect the ultimate decision, will only incentivise proper decision-making further.
Therefore, as false negatives are more harmful than false positives, the indeterminacy of materiality in polycentric decision-making renders the conservative model and its onerous materiality threshold unsuitable.
In summary, while the courts may prefer the conservative model due to their familiarity with the Edwards v Bairstow principle, it contains two critical flaws. Firstly, subsuming the mistake of fact ground into the Edwards v Bairstow principle would remove the courts’ ability to remedy procedural injustice. Secondly, its onerous materiality threshold will be exceedingly difficult to satisfy in the context of polycentric decision-making, resulting in great injustice. Therefore, another model should be favoured.
B The Contextual Model
This model employs a more liberal approach that sees the ground’s elements vary according to the context. There is support for such an approach, with “a considerable body of academic commentary confirming that common law courts do and should apply variable, variegated or sliding standards of review depending on the context”.204 A variable approach gives effect to the importance of context in administrative law, particularly regarding procedural fairness grounds.205 Unlike the conservative model, this model does not in itself comprise a conducive mistake of fact ground. Rather, the contextual model demonstrates three different ways the court could apply a variable standard targeting different aspects of the ground. These will then be critiqued in turn to determine where a contextual approach might present utility.
204 Deliu v Connell [2016] NZHC 361, [2016] NZAR 475 at [6]. See also Forsyth, above n 170, at [32].
205 Joseph, above n 33, at [25.1].
(a) Variable standard of review based on factual complexity: Craig’s approach
Firstly, the scope of the ground could be extended to include factual evaluations, with the standard of review varying according to the case’s factual complexity. In its E formulation, the ground only assesses mistakes of incontrovertible fact, as only incontrovertible facts are established, uncontentious and objectively verifiable.206 In contrast, this model adopts the approach proposed by Paul Craig, who argues that the ground should be expanded to allow for intervention in “cases involving more complex factual findings which will require a greater degree of evaluative judgement.”207 To minimise possibilities of merits-review and the violation of constitutional principles that this entails,208 Craig proposes a variable standard of review determined by the case’s factual complexity and the degree of evaluative judgement employed by the decision-maker.209 Where the relevant facts are incontrovertible, Craig sees no issue in the court intervening to a correctness standard. At the more complex and evaluative end of the spectrum, Craig anticipates the court usually applying a rationality test.210
This is similar to the conservative model in that it combines the mistake of fact ground and Edwards v Bairstow principle. However, in employing a variable standard of review, incontrovertible factual errors are not required to render the ultimate decision untenable, therefore allowing the ground to continue giving effect to procedural justice.
(b) Variable materiality threshold based on injustice
Secondly, the materiality limb could employ a variable test based on injustice; specifically, the degree to which the applicant’s rights have been affected. Where rights are heavily affected, the mistake causes significant injustice, as seen in cases like Daganayasi and Haile,211 meaning the court should employ a less onerous test. Meanwhile, less injustice has
206 See Chapter One, Heading B, Subheading Two.
207 Craig “Judicial Review”, above n 1, at 794.
208 See Chapter Three, Heading B, Subheading One.
209 Craig “Judicial Review”, above n 1, at 802-803.
210 Craig “Judicial Review”, above n 1, at 802-803. It should be acknowledged that Craig also proposes the American ‘substantial evidence’ test as an option to be applied at this end of the spectrum. However, as Craig acknowledges, there is significant overlap between the two, as a finding unsupported by substantial evidence will be regarded as irrational. Therefore, only the rationality test needs to be discussed for the purpose of this paper.
211 Daganayasi, above n 2; and Haile, above n 65. See Chapter One, Heading A, and Chapter Two, Heading A.
occurred where rights are not heavily affected, so the court could employ a more onerous materiality test. In the interests of clarity, the specific tests to be employed at each end of the spectrum are discussed in the Goldilocks model.
A variable approach based on injustice is consistent with other procedural fairness grounds, which, as Joseph notes, “vary according to the power that is exercised and the circumstances of its use, including the effects of the decision on personal rights or interests.”212 A variable approach would also be consistent with the United Kingdom courts’ approach to mistakes of fact prior to E, where, as Craig stated, “in practice, administrative court judges tended to set aside decisions on this ground when justice so required.”213 One example of this was the 2001 England and Wales Court of Appeal decision of Turgot, where the Court declined to afford any special deference to the decision-maker’s factual conclusions, as the appellant’s rights were significantly affected by the decision.214 This approach was also employed by the New Zealand Court of Appeal in Glaxo, where Cooke P stated:215
Only a mistake sufficiently material to be described as the basis or the probable basis of the commissioner’s decision could, as it seems to us, warrant the Court in putting a stop to proceedings at the preliminary stage [...]
The most likely interpretation of Cooke P’s statement is that the Court was applying a more stringent materiality threshold because an alternative avenue to redress was available to the applicant, making the case a “preliminary stage”. This meant that the applicant’s rights would not be as affected by an adverse ruling. Further, as the case was purely commercial, significant rights were not at issue.
(c) Variable materiality threshold based on the type of factual error
Alternatively, Hanna Wilberg proposes that the materiality threshold could distinguish between the two types of incontrovertible factual error that come within the scope of the ground.216 The
212 Joseph, above n 33, at [25.1].
213 Craig Administrative Law, above n 143, at [17–016].
214 R v Secretary of State for the Home Department ex parte Turgot [2000] EWCA Civ 22; [2001] 1 All E.R. 719; and Craig “Judicial Review”, above n 1, at 804.
215 Glaxo, above n 87, at 184.
216 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 36; and Shaheen, above n 165, at [20].
first type of error is a mistake as to an incontrovertible fact, made due to a misunderstanding of the evidence in front of the decision-maker or gap or error in the evidence.217 A good example is Haile, where a confusion of two parties, EPRDF and ERPR, led the Special Adjudicator to determine that the applicant was uncredible.218 This type of mistake is generally deemed unproblematic.219
The second type involves the decision-maker mistakenly believing that a fact was uncontentious when it was actually contentious.220 This usually involves the applicant bringing forward a new piece of evidence that contradicts the fact previously believed to be uncontentious.221 For example, in Shaheen, an asylum-seeker argued that returning to Pakistan would put her at risk of harm due to her abusive husband, detailing several instances of abuse that had occurred in Pakistan.222 The Tribunal granted asylum on this basis. Subsequently, the Secretary of State brought forward evidence demonstrating that the claimant was in the United Kingdom when the alleged abuse occurred (and not in Pakistan as the claimant claimed). The Secretary of State argued that the claimant had misled the Tribunal and was uncredible.
Therefore, we have a decision based on a factual assumption that the decision-maker believed to be uncontentious (that the claimant was in Pakistan when the alleged abuse occurred), and an applicant bringing evidence to contend this assumption. This second type of factual error is generally deemed more problematic as it “invite[s] litigants and courts to delve into the minutiae of decision-making, and allow[s] every little finding or understanding of fact to be re- opened on review.”223 This is because where the mistake is as to the availability of evidence, the typically-onerous second limb is much easier to satisfy, as the question of whether there was additional unconsidered evidence is unlikely to be contentious.224 In other words, under this type of mistake, “it is not necessary to show that the decision-maker was actually mistaken
217 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 36; and Shaheen, above n 165, at [20].
218 Haile, above n 65. See Chapter Two, Heading A.
219 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 36.
220 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 37; Shaheen, above n 165, at [20] and [28].
221 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 37; Shaheen, above n 165, at [20] and [28].
222 Shaheen, above n 165.
223 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 35 and 37; Shaheen, above n 165, at [27]–[29].
224 Forsyth, above n 170, at [39].
about a relevant fact; it is sufficient to show in relation to some relevant fact that contrary evidence was available but not considered.”225
Thus, Wilberg proposes applying different materiality tests to the two types of factual error. She posits two options. The first would see “missing evidence considered sufficiently material where the decision-maker specifically relied on the absence of evidence to this effect.”226 Whereas, if “the decision-maker relied on a fact that was inconsistent with the missing evidence, but did not specifically turn its mind to the question of whether any contrary evidence existed”, materiality would not be satisfied.227 The second option would see missing evidence deemed material only where it would have determined the answer to a legal test or the decision itself.228
1 Critique: Unorthodox Approach to Assigning Standards of Review
The contextual model departs from the traditional approach to assigning standards of review, which sees the courts rely on “a set of distinct grounds of review with fixed standards of review.”229 As Wilberg notes, the traditional approach would therefore dictate that rationality review for factual evaluations and correctness review for incontrovertible facts should operate as two distinct grounds: the Edwards v Bairstow principle and the mistake of fact ground.230 Similarly, it would see different materiality thresholds, whether that turns on injustice or the type of factual error, constituting separate discrete grounds. The traditional approach is not without exception, as evidenced by unreasonableness’ variable intensity of review, which could indicate a move towards a new approach. However, the courts have thus far adhered to the traditional approach in all other areas.231 This alone does not prevent us from adopting the contextual model, but it would create more inconsistency in the law in this respect. However, adoption could still be justified if aspects of the contextual model present sufficient utility, in
225 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 37. 226 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 39. 227 At 39.
228 At 39.
229 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 9. See also Dean Knight “Modulating the Depth of Scrutiny in Judicial Review: Scope, Grounds, Intensity, Context” (2016) NZ L Rev 63.
230 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 9.
that they further the policy rationales underpinning the ground.232 I will now assess this regarding each aspect of the model.
(a) Variable standard of review based on factual complexity: Craig’s approach
Craig’s proposal of expanding the scope of the mistake of fact ground to encompass factual evaluations makes little change to how the law is applied in practice. Incontrovertible facts would still be assessed to a correctness standard of review, while factual evaluations would usually be assessed to a rationality standard. This may have sufficient utility under Craig’s schema of judicial review generally, however this paper assesses the mistake of fact ground in relation to the landscape of administrative law as it exists in New Zealand today. In this context, the lack of substantive difference in Craig’s approach means it fails to supply sufficient utility to consider adoption.233
Further, this approach would presumably require the courts to abandon the well-established Edwards v Bairstow principle in favour of expanding the mistake of fact ground, which is not even well-established in its narrower form.234 While relative judicial familiarity should not alone determine such matters, more established tools should be favoured where alternatives do not provide additional utility. We should disregard Craig’s approach for now.
(b) Variable materiality threshold based on injustice
The injustice and unfairness caused by errors made during the fact-finding process is magnified where the ultimate decision greatly affects someone’s rights. As principles of justice and fairness are a primary justification for establishing this ground in the first place,235 the ground would see great utility in lowering the bar for intervention where these principles are egregiously contravened through this variable materiality threshold. We can again refer to the United Kingdom asylum case of Haile as an illustrating example.236 The case saw the applicant’s asylum claim rejected due to a factual error, that would have seen him returning to Ethiopia and potentially facing political persecution – a significant effect on his rights. This is
232 See Chapter Three.
233 See also Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 9.
234 See Chapter Two, Heading B.
235 See Chapter Two, Heading A, and Chapter Three, Heading A, Subheading Two.
236 Haile, above n 65. See Chapter Two, Heading A.
likely why the Court noted that the mistake was “most regrettable” and “must inevitably leave a sense of deep injustice in the appellant”, before applying an un-onerous ‘rebuttable presumption’ test, although the court did not explicitly link the two.237 (The rebuttable presumption test is outlined in detail under the Goldilocks model).
In conclusion, a variable materiality threshold based on injustice provides sufficient utility to warrant departure from the traditional approach of assigning standards of review.
(c) Variable materiality threshold based on the type of factual error
To recap, Wilberg proposes two options for a materiality test that would only apply to mistakes as to the availability of evidence. These include materiality only being satisfied where the decision-maker has specifically relied on the absence of the relevant evidence or where the evidence determines the answer to a legal test or the decision itself.238
While I agree with Wilberg that there is a danger that, in some instances, the second type of factual error may allow applicants and the courts to delve too far into the minutiae of decision- making without sufficient justification, the fourth element adequately deals with this concern in its current form. Specifically, if a mistake is made regarding the availability of evidence, the fourth element still requires the mistake to have had a sufficiently material influence on the overall decision. This can be exemplified with reference to Wilberg’s proposed solutions. If the decision-maker specifically relied on the absence of the new evidence, or if the new evidence determines the answer to a legal test (or the decision itself), it is far more likely to be material under the fourth element as it is currently understood and therefore more likely to vitiate the decision. Whereas, if these factors are not present, the new evidence would be far less likely to be deemed material, making success under the ground less likely. Therefore, Wilberg’s approach does not provide sufficient utility to warrant departure from the traditional method of assigning standards of review.
In fact, Wilberg’s suggestions would see the ground worse off. Firstly, her ‘specific reliance’ threshold would constitute a less holistic assessment than that which presently takes place
237 At [25].
238 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 39.
under the fourth element, as it would not consider key factors such as the credibility and probity of the new evidence.
Meanwhile, her second proposed threshold, which sees materiality satisfied where the evidence determines the answer to a legal test or the decision itself, is very onerous. This brings the same issues outlined in critiques of the conservative model, in that it would restrict the ground’s ability to remedy injustice and unfairness. This is critical because, as Wilberg herself notes, this type of factual error can cause significant injustice.239 Wilberg gives the example of R v Criminal Injuries Compensation Board, in which the Board took the view that there was no evidence available to support the applicant’s assault claim as they were unaware of the existence of a police doctor’s report that was consistent with the applicant’s claim.240 The injustice exemplified in this case is not any less than the injustice that may be caused under the first type of factual error. A different materiality threshold is therefore unnecessary and unwise.
C The Goldilocks Model
The Goldilocks model seeks to calibrate the ground in a way that is ‘just right’, informed by the normative tensions and policy arguments discussed in Chapter Three and the lessons learned from analysing the conservative and contextual models.
The preceding examination of the conservative and contextual models reveals that it is inappropriate to deal with errors of incontrovertible fact and erroneous factual evaluations under the same ground, whether that be through an extension of the Edwards v Bairstow principle to encompass mistakes of incontrovertible fact, or through an extension of the mistake of fact ground to encompass factual evaluations. This leads us to conclude that the mistake of fact test should adopt the second element of E (that the fact be established, being uncontested and objectively verifiable) as it is currently understood, such that the ground’s scope is limited to incontrovertible facts.241 The Edwards v Bairstow principle will remain to deal with factual evaluations.242
239 Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 37.
240 Criminal Injuries Compensation Board, above n 64, at 343; and Wilberg “Re-evaluating Mistake of Fact”, above n 20, at 14.
241 See Chapter One, Heading B, Subheading Two.
242 See Chapter One, Heading C, Subheading Two.
A few options are available for the materiality threshold. The first is the moderate materiality threshold outlined in Lord Carnwath’s original formulation in E, which requires the mistake to have played a material, but not necessarily decisive, part in the decision-maker’s reasoning.243 There is plenty of New Zealand case law precedent supporting this approach. For example, the New Zealand Supreme Court applied (what likely equated to) this threshold in Ririnui, finding a mistake to be reviewable as it “materially influenced” the decisions.244 Several New Zealand High Court cases have also adopted this threshold:245 in Zhao, the Court deemed the element met as the decision-maker “was materially influenced by a mistake of fact” (citing E);246 and in Ellipse, the Court noted that the ground must only be available where the “error was influential in [the decision-maker’s] decision.”247
The second option would be to adopt a less onerous materiality threshold that employs a presumption of materiality, under which the mistake is deemed to constitute a reviewable error unless the court is satisfied that it made no difference to the decision. Case law precedent also exists for this approach. For example, in Haile, the English Court of Appeal deemed the mistake material as “it cannot confidently be said to have made no ultimate difference to the result.”248 The Court also applied this approach in Simplex, with Purchas LJ stating:249
It is not necessary for [the applicant] to show that the minister [sic] would, or even probably would, have come to a different conclusion. He has to exclude only the contrary contention, namely that the minister [sic] necessarily would still have made the same decision.
The New Zealand Court of Appeal has applied a similar approach in Taylor, where materiality was satisfied as “it was not inevitable that the same conclusion would have been reached without the error.”250
243 E, above n 3, at [66].
244 Ririnui, above n 119, at [53], [62], and [74].
245 As well as the following examples, see S v Health and Disability Commissioner [2023] NZHC 692 at [195], [196], [199], [204] and [215].
246 Zhao, above n 97, at [77] and [89].
247 Ellipse, above n 98, at [73].
248 Haile, above n 65, at [25]. See Chapter Two, Heading A.
249 Simplex GE (Holdings) Ltd v Secretary of State for the Environment (1988) [2017] PTSR 1041 at 1060.
250 Taylor, above n 111, at [95]–[96].
The third option is the variable materiality threshold set out in the contextual model. This would see the courts apply the less onerous ‘rebuttable presumption’ materiality threshold where rights are heavily affected by the decision and the moderate ‘material influence’ threshold in other cases. The variable threshold gives sufficient utility to warrant departing from the courts’ traditional approach to assigning standards of review, as it allows the test to change as the principles of justice and fairness require.
In prohibiting the review of factual evaluations under the mistake of fact ground, the Goldilocks model would require courts to distinguish between incontrovertible facts and factual evaluations. The line between the two is clear conceptually but can be rather blurry in practice.251 To return to Daganayasi as an example, what if Joanna instead argued that Dr Blake-Palmer’s determination that deporting George to Fiji did not present “any immediate prospect of threat to life” was a mistake of fact?252 All other medical professionals disagreed with Dr Blake-Palmer,253 but would this factual error be deemed incontrovertible and come within the second limb? Would Dr Blake-Palmer’s contrary view be sufficient to render this fact contentious? Would the established lack of protein sources and adequate medical care make this fact objectively verifiable? These questions do not have obvious answers. A judge could reasonably conclude either way.
This blurry distinction between incontrovertible facts and factual evaluations is problematic as they face different standards of review. The distinction could be manipulated such that the courts could intervene for policy reasons, while keeping these hidden under the guise of legal argument.254 To return to the above hypothetical, if Cooke J had a particularly strong anti- immigration stance, he could reasonably define the determination as a factual evaluation, state that the rationality standard had not been satisfied and decline to intervene on that basis, resulting in Joanna and George’s deportation. Alternatively, if Cooke J had more liberal views,
251 Forsyth, above n 170, at [32]–[33].
252 Daganayasi, above n 2, at 137.
253 At 133–134, 135 and 139.
254 Williams, above n 32, at 2–3 notes that the courts already do this when applying other vexed distinctions, such as the law–fact distinction and the jurisdiction distinction.
he could reasonably classify the determination as an incontrovertible fact, state that an error occurred and prevent the deportation.
However, such vexed distinctions are not a rarity in the law. For example, the law–fact distinction has been relevant since the 1960s255 and remains “inherently unstable”,256 often causing “notorious difficulty”.257 Analytical disagreements regarding whether a question is one of law or fact are still common, with specific cases featuring a diversity of opinion from the courts and commentators.258 Nevertheless, the law–fact distinction has remained, and it will likely continue to pervade the law as long as the differing approaches to questions of law and fact make its existence necessary. The incontrovertible fact–factual evaluation distinction should be regarded in the same light. It is undoubtedly imperfect, but if one believes normative considerations demand the two to face differing standards of review, it must also be conceded that the distinction is necessary. As outlined in the analysis of other models, abolishing the distinction is far less palatable.
Further, the incontrovertible fact–factual evaluation distinction may prove to be less problematic than other vexed distinctions, as in a relatively short period, an impressive amount of case law and commentary has developed to assist courts in distinguishing incontrovertible facts and factual evaluations in a consistent and principled manner.259 Where uncertainty inevitably remains, one must trust the courts to operate in this grey area of the law as is a reality of their role.
Therefore, while the distinction between incontrovertible facts and factual evaluations that this ground requires may be somewhat problematic, it by no means warrants an alteration to the
255 Craig Administrative Law, above n 143, at [16–030].
256 Joseph, above n 33, at [13.5.1].
257 E, above n 3, at [44]; and Joseph, above n 33, at [13.5.1].
258 Craig Administrative Law, above n 143, at [16–031].
259 For an example of helpful case law, see New Zealand Fishing Industry, above n 35, at 552 (Cooke J: “it cannot be said to be a mistake to adopt one of two differing points of view of the facts, each of which may be reasonably held.”). At the time of writing, this passage has been cited in 169 New Zealand decisions, having provided determinative guidance in many of these cases. For commentaries that assist in navigating this distinction, see Wilberg “Mistake of Fact as a Ground of Review”, above n 27, at 204–210; and Taylor, above n 47, at [15.17]–[15.19].
ground’s second element. As the ground becomes more established through the development of case law and academic commentary, this distinction will only become clearer.
Confining the ground to incontrovertible facts presents another limitation in the post-truth era.260 As more once-established truths become contested, they fall foul of the second limb (which states that the fact must be uncontested),261 meaning the ground cannot provide redress. More complex matters featuring conflicting expert evidence or competing scientific opinion are particularly vulnerable, as the courts have refused to invoke the mistake of fact ground in such instances,262 and for good reason: executive and administrative decision-makers are better-placed than courts to resolve substantive disagreements involving specialised subject matter.263 This is due to differences in constitutional legitimacy and institutional capacity, with the former being democratically accountable (directly or indirectly) and having greater access to expertise, experience and robust procedures to assist decision-making.264
However, other solutions exist to deal with misinformation in this situation. The Edwards v Bairstow principle will remain available where evaluating conflicting evidence produces an untenable conclusion.265 Solutions beyond the judiciary should also be employed – the post- truth era arose from the political sphere; perhaps the political sphere must take the lead in regulating it. Regardless, this is not a sufficient reason to alter the second element.
If the Goldilocks model were implemented, one would have to determine which of the three proposed materiality thresholds should apply. This decision will likely turn on two matters.
260 See Chapter Three, Heading A, Subheading Four.
261 See Chapter One, Heading B, Subheading Two.
262 Wilberg “Mistake of Fact as a Ground of Review”, above n 27, at 208–209; Brook Valley Community Group Inc v Trustees of the Brook Waimarama Sanctuary Trust [2017] NZHC 1844 at [21]; and New Zealand Climate Science Education Trust v National Institute of Water and Atmospheric Research Ltd [2012] NZHC 2297, [2013] 1 NZLR 75 at [45]–[47], [157], [173] and [182].
263 Demetriou and Houseman, above n 152, at 27–28. See also Marcelo Rodriguez Ferrere “Judicial review of scientific findings” (2012) 11 NZLJ 380.
264 Wilberg “Mistake of Fact as a Ground of Review”, above n 27, at 203.
265 See Chapter One, Heading C, Subheading Two.
The first involves the balancing act between interests of justice and fairness and interests of finality. Those who place more value in upholding justice and fairness are more likely to favour the less onerous rebuttable presumption threshold, while those who prioritise upholding finality and preventing a ‘floodgates’ scenario will probably favour the moderate ‘materially influenced’ threshold. The second is the degree to which one is comfortable in giving the courts more discretion to determine their own jurisdiction. Those who perceive this to be an issue are more likely to favour a fixed materiality threshold, while those who trust the courts not to overstep their constitutional role will be more comfortable with the courts determining the materiality threshold to be applied themselves under a variable threshold.
In my view, none of these options are demonstrably more correct than the others. Any of the three would see the ground adequately give effect to the policy considerations laid out in Chapter Three. Therefore, this question is best left open at this stage.
D Conclusion
The Goldilocks model should be adopted as the test for a mistake of fact ground of review in New Zealand. It dictates that the second element (that the fact was established, being uncontentious and objectively verifiable) should be adopted as is and applied as has been understood throughout this paper – in a manner that restricts the scope of the ground to reviewing incontrovertible facts to the exclusion of factual evaluations266 – eventhough this creates a vexed distinction and weakens the ground’s ability to correct misinformation when the subject matter is complex. The fourth limb could be adopted in its current moderate state, however instating a less onerous or variable materiality threshold should also be considered.
266 See Chapter One, Heading B, Subheading Two.
Conclusion
“Nothing rankles like a mistake of fact.”
– Dame Sian Elias267
As is now clearly evident, the mistake of fact ground of review has had a tumultuous history in New Zealand administrative law. While development of the ground in the United Kingdom has culminated in the authoritative and clarifying decision of E,268 New Zealand has been unable to replicate such efforts, instead seeing divided and inconsistent treatment at the High Court, a lack of clear, comprehensive and authoritative engagement with the ground at the Court of Appeal, and the missed opportunity of Ririnui at the Supreme Court.269
Despite this troubled history, the case for an established mistake of fact ground has never been stronger. Decisions influenced by mistakes of incontrovertible fact continue to impart significant and unnecessary injustice and unfairness on those subject to them. They frequently render decisions unable to give effect to the policy and objects of Parliament’s legislation, thereby undermining Parliamentary sovereignty and the rule of law. Further, courts are already comfortable in correcting incontrovertible factual errors, and widening this scope beyond jurisdictional facts would provide a helpful tool to ensure decision-making occurs from a correct factual basis, which is particularly necessary in the contemporary post-truth era.
There is much debate to be had about what New Zealand’s mistake of fact ground should look like. The conservative model that sees incontrovertible factual errors dealt with under the Edwards v Bairstow principle brings the benefit of judicial familiarity but would see the ground deprived of its ability to give effect to procedural fairness and would cause injustice and poor governance to pervade where decision-making is polycentric. Meanwhile, the contextual model provides a unique prospicient approach, but only the variable materiality threshold based on injustice provides sufficient utility to warrant departure from the courts’ traditional approach of assigning fixed standards of review.
267 Sian Elias “Judgery and the Rule of Law” [2015] OtaLawRw 6; (2015) 14 Otago LR 49 at 55.
268 E, above n 3.
269 Ririnui, above n 119.
Therefore, the more moderate Goldilocks model, which closely mirrors Lord Carnwath’s original formulation in E, provides the best basis for a mistake of fact test for New Zealand. It will see the second element restrict the ground to only reviewing incontrovertible facts, despite this creating a somewhat vexed distinction and reducing the ground’s ability to remedy misinformation in the post-truth era where the subject matter is complex. Three options for the materiality threshold remain: a moderate ‘material influence’ threshold, a less onerous ‘rebuttable presumption’ threshold, or a variable threshold encompassing the two. This paper leaves open the question of which would be most optimal.
Bibliography
A Cases
Air Nelson Ltd v Minister of Transport [2008] NZCA 26; [2008] NZAR 139, [2008] BCL 309 (CA).
ANZ Sky Tours Ltd v New Zealand Tourism Board [2019] NZHC 925, [2019] NZAR 951.
Attorney-General v Moroney [2000] NZHC 1298; [2001] 2 NZLR 652 (HC).
A v Legal Complaints Review Officer [2013] NZHC 1100.
Board of Trustees of Salisbury Residential School v Attorney-General [2012] NZHC 3348, [2013] NZAR 228.
Brook Valley Community Group Inc v Trustees of the Brook Waimarama Sanctuary Trust
[2017] NZHC 1844.
Bryson v Three Foot Six Ltd [2005] NZSC 34, [2005] 3 NZLR 721.
Charter Holdings Ltd v Commissioner of Inland Revenue [2016] NZCA 499.
Daganayasi v Minster of Immigration [1980] 2 NZLR 130 (CA).
Dean v Chief Executive of the Ministry of Business, Innovation and Employment [2017] NZHC 588.
Deliu v Connell [2016] NZHC 361, [2016] NZAR 475.
Deliu v District Court at Auckland [2023] NZHC 658.
Deliu v Hong [2015] NZHC 492.
Deliu v Hong and Legal Complaints Review Officer [2012] NZHC 158, [2012] NZAR 209.
Diagnostic Medlab Ltd v Auckland District Health Board HC Auckland, CIV 2006-404-4724, 20 March 2007.
Double Eight Ltd v Disputes Tribunal at Manukau [2014] NZHC 1862.
D v M and Board of Trustees of Auckland Grammar School [2003] NZAR 726 (HC).
Ellipse Institute Ltd v New Zealand Qualifications Authority [2012] NZHC 2083, [2012] NZAR 871.
Harris v Family Court at Auckland [2023] NZHC 258. Hawkins v Minister of Justice [1990] 3 NZLR 486 (HC). Helu v District Court at Auckland [2013] NZHC 2965.
Kilmarnock Farm Ltd v Canterbury Regional Council [2019] NZHC 2467, [2020] NZRMA 345.
JW v Chief Executive of Ministry of Business, Innovation and Employment [2022] NZCA 286
Lewis v Wilson & Horton Ltd [2000] NZCA 175; [2000] 3 NZLR 546 (CA).
Li v Immigration and Protection Tribunal [2018] NZHC 174. Manawa Energy Ltd v Electricity Authority [2022] NZHC 1444. McGrath v Minister of Justice [2014] NZHC 3279, [2015] NZAR 122.
Minister of Conservation v Māori Land Court [2008] NZCA 564; [2009] 3 NZLR 465, [2009] BCL 185 (CA).
Nabou v Minister of Immigration [2012] NZHC 3365, [2013] NZAR 155.
New Zealand Climate Science Education Trust v National Institute of Water and Atmospheric Research Ltd [2012] NZHC 2297, [2013] 1 NZLR 75.
New Zealand Fishing Industry Association Inc v Minister of Agriculture and Fisheries [1988] 1 NZLR 544 (CA).
Northern Inshore Fisheries Company Ltd v Minister of Fisheries HC Wellington CP 235-01, 4 March 2002.
Oggi Advertising Ltd v Auckland City Council [2005] NZHC 1259; [2005] NZAR 451 (HC).
Problem Gambling Foundation of New Zealand v Attorney-General [2015] NZHC 1701.
Queenstown Lakes District Licensing Agency Inspector v Turnbull Group Ltd [2011] NZHC 598, [2011] NZAR 554.
Re Erebus Royal Commission; Air New Zealand Ltd v Mahon [1984] AC 808, [1983] NZLR 662 (PC).
Rimene v R [2015] NZHC 2619.
Ririnui v Landcorp Farming Ltd [2016] NZSC 62, [2016] NZLR 1056.
Roebeck v Attorney-General [2022] NZHC 3341.
Royal Forest and Bird Protection Society of New Zealand Inc v Minister of Fisheries [2021] NZHC 1427.
Slavich v Wellington District Court [2023] NZHC 251.
Smith v Fonterra Co-Operative Group Ltd [2020] NZHC 419.
Squid Fishery Management Company Ltd v Minister of Fisheries CA39/04, 13 July 2004. Students for Climate Solutions Inc v Minister of Energy and Resources [2022] NZHC 2116. S v Health and Disability Commissioner [2023] NZHC 692.
Taafi v Minister of Immigration [2013] NZAR 1037 (HC).
Tamil X v Refugee Status Appeals Authority [2009] NZCA 488, [2010] 2 NZLR 73.
Tan v New Zealand Registered Architects Board [2016] NZHC 2276.
Taylor v The Chief Executive of the Department of Corrections [2015] NZCA 477; [2015] NZAR 1648.
Te Pou Matakana Ltd v Attorney-General [2022] 2 NZLR 178; [2021] NZHC 3319.
Thorndon Quay Collective Inc v Wellington City Council [2022] NZHC 2356.
T v Immigration and Protection Tribunal [2012] NZHC 1871.
Vulcan Steel Ltd v Wonnocott [2013] NZCA 398.
Waitakere City Council v Lovelock [1997] 2 NZLR 385 (CA).
Winton Property Investments Ltd v Minister of Finance [2022] NZHC 638.
Zafirov v Minister of Immigration [2009] NZHC 419; [2009] NZAR 457 (HC).
Zhao v Legal Complaints Review Officer [2012] NZHC 3247, [2013] NZAR 193.
Anisminic Ltd v Foreign Compensation Commission [1968] UKHL 6; [1969] 2 AC 147, [1969] 2 WLR 163.
Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1947] EWCA Civ 1, [1948] 1 KB 223.
Begum v London Borough of Tower Hamlets [2003] UKHL 5; [2003] 2 AC 430.
Chief Constable of the North Wales Police v Evans [1982] UKHL 10; [1982] 1 WLR 1155 (HL).
Edwards (Inspector of Taxes) v Bairstow [1955] UKHL 3; [1956] AC 14, [1955] 3 WLR 410 (HL).
E v Secretary of State for the Home Department [2004] EWCA Civ 49, [2004] QB 1044, [2004] 2 WLR 1351.
Laker Airways Ltd v Department of Trade [1976] EWCA Civ 10; [1977] QB 643, [1977] 2 WLR 234.
MT (Algeria) v Secretary of State for the Home Department [2007] EWCA Civ 808, [2007] HRLR 41.
Padfield v Minister of Agriculture, Fisheries and Food [1986] A.C. 997.
Pulhofer v Hillingdon LBC [1986] UKHL 1; [1986] AC 484, [1986] 2 WLR 259 (HL).
R (Alconbury Developments Ltd) v Secretary of State for the Environment, Transport and the Regions [2001] UKHL 23, [2003] AC 295.
R v Criminal Injuries Compensation Board, ex parte A [1999] UKHL 21; [1999] 2 AC 330, [1998] 1 WLR 277.
R v Deputy Industrial Injuries Commissioner, ex parte Moore [1965] 1 QB 456, [1965] 2 WLR
89.
R (Haile) v Immigration Appeal Tribunal [2001] EWCA Civ 663, [2002] INLR 283.
R (Cart) v Upper Tribunal [2011] UKSC 28; [2012] 1 AC 663.
R v Radio Authority, ex parte Bull [1996] QB 169.
R v Secretary of State for the Home Department, ex parte Khawaja [1982] UKHL 5; [1984] AC 74, [1983] 2 WLR 321 (HL).
R v Secretary of State for the Home Department, ex parte Onibiyo [1996] EWCA Civ 1338; [1996] QB 768.
R v Secretary of State for the Home Department ex parte Turgot [2000] EWCA Civ 22; [2001] 1 All E.R. 719.
R v South Hams District Council, ex parte Gibb and Others [1954] EWCA Civ 3; [1955] QB 158.
Secretary of State for Education and Science v Tameside Metropolitan Borough Counsel
[1976] UKHL 6; [1977] AC 1014, [1976] 3 WLR 641 (HL).
Secretary of State for Employment v Associated Society of Locomotive Engineers and Firemen (No 2) [1972] 2 QB 455, [1972] 2 WLR 1370 at 493.
Shaheen v Secretary of State for the Home Department [2005] EWCA Civ 1294, [2006] Imm AR 57.
Simplex GE (Holdings) Ltd v Secretary of State for the Environment (1988) [2017] PTSR 1041.
Terry v Huntington (1679) Hardres 480, 145 ER 557 (Exch).
Wandsworth LBC v A [2000] 1 WLR 1246 (CA).
B Legislation
Immigration Act 1964.
Immigration Act 2009. Senior Courts Act 2016.
C Books and Chapters in Books
Paul Craig Administrative Law (9th ed, Thomson Reuters, London, 2021).
Paul Daily “Facticity: Judicial Review of Factual Error in Comparative Perspective” in Peter Cane, Eric Ip, Herwig Hoffman and Peter Lindseth (eds) Oxford Handbook of Comparative Administrative Law (Oxford University Press, Oxford, 2021) 901.
Michael Fordham Judicial Review Handbook (7th ed, Hart, Oxford, 2020).
Christopher Forsyth and Emma Dring “The Final Frontier: The Emergence of Material Error of Fact as a Ground for Judicial Review” in Christopher Forsyth, Mark Elliot, Swati Jhaveri, Anne Scully-Hill and Michael Ramsden (eds) Effective Judicial Review: A Cornerstone of Good Governance (Oxford University Press, Oxford, 2010) 245.
Guðmundur Hálfdanarson and Maximilian Conrad "Concluding Reflections on Europe in the Age of Post-truth Politics" in Maximilian Conrad and others (eds) Europe in the Age of Post- Truth Politics (Springer Nature, Cham, 2023) 247.
Philip A Joseph Joseph on Constitutional and Administrative Law (5th ed, Thomson Reuters, Wellington, 2021).
Janet McLean “Administration in the constitution: Disaggregating power for accountability purposes” in Carol Harlow (ed) A Research Agenda for Administrative Law (Edward Elgar Publishing, Cheltenham, 2023) 65.
MB Rodriguez Ferrere "John Smillie, Judicial Review and New Zealand's Missed Opportunity" in Shelley Griffiths, Mark Henaghan and MB Rodriguez Ferrere (eds) The Search for Certainty: Essays in Honour of John Smillie (Thomson Reuters, Wellington, 2016) 180.
Matthew Smith New Zealand Judicial Review Handbook (2nd ed, Thomson Reuters, Wellington, 2016).
Graham Taylor Judicial Review A New Zealand Perspective (4th ed, LexisNexis, Wellington, 2018).
Doug Tennent, Katy Armstrong and Peter Moses Immigration and Refugee Law (3rd ed, LexisNexis, Wellington, 2017).
Hanna Wilberg "Mistake of Fact as a Ground of Review: Distinct and Defensible" in Joe Tomlinson and Anne Carter (eds) Facts in Public Law Adjudication (Oxford, Hart Publishing, TBC) 201.
HWR Wade and CF Forsyth Wade & Forsyth’s Administrative Law (10th ed, Oxford University Press, Oxford, 2009).
Harry Woolf, Jeffrey Jowell, Catherine Donnelly and Ivan Hare De Smith’s Judicial Review
(8th ed, Sweet and Maxwell, London, 2018).
D Journal Articles
Paul Craig “Judicial Review, Appeal and Factual Error” (2004) PL 788.
Paul Daly “Judicial Review of Factual Error in Ireland” (2008) 30 Dublin U LJ 187.
Marie Demetriou and Stephen Houseman "Review for Error of Fact – A Brief Guide" (1977) 2(1) Jud Rev 27.
Geoffrey A Flick “Error of Law or Error of Fact” (1983) 10 UWAL Rev 193.
Christopher Forsyth “Error of Fact Revisited: Waiting for the ‘Anisminic Moment’” (2018) 23 JR 1.
Lon Fuller “The Forms and Limits of Adjudication” (1978) 92(2) Harv LR 353.
Matthew Groves “Judicial Review of Administrative Action in the High Court of Australia” (2008) 33 Queen’s Law J 327.
Jeff A King “The Pervasiveness of Polycentricity” (2008) PL 101.
Dean Knight “Modulating the Depth of Scrutiny in Judicial Review: Scope, Grounds, Intensity, Context” (2016) NZ L Rev 63.
Nathalie Lieven “Judicial Review: Trends and Forecasts” (2009) 14(1) JR 9.
Edward Lui “‘Fairness’” for Mistake of Fact: A Mistake in Fact” (2020) 3 PL 428.
Andrew McCue “Turning the Page: An Analytical Solution to the Law of Jurisdictional Error” (2021) 13 Ital J Public Law 55.
Tembeka Ngcukaitobi “Error of fact: The recognition of error of fact as a ground for judicial review in South African administrative law” (2013) 26(3) Adv 30.
Esther Pearson “Finding Fairness in Fact Finding: Material Mistake of Fact Review in Asylum Cases” (2019) 26(2) AJ Admin L 100.
MB Rodriguez Ferrere “An Impasse in New Zealand Administrative law: How did we get here?” (2017) 28 PLR 310.
Geoffrey Palmer “Reflections Upon the Drift Away from Democracy in the United States” (2022) 53(3) VUWLR 407.
Marcelo Rodriguez Ferrere “Judicial review of scientific findings” (2012) 11 NZLJ 380. Hanna Wilberg “Mistakes about Mistake of Fact: The New Zealand Story” (2017) 28 PLR 248.
Rebecca Williams “When is an Error not an Error? Reform of Jurisdictional Review of Error of Law and Fact” (2007) PL 793.
Turgay Yerlikaya and Seca Toker Aslan “Social Media and Fake News in the Post-Truth Era: The Manipulation of Politics in the Election Process” (2020) 22(2) Insight Turk 177.
E Conference Papers
Hanna Wilberg “Re-evaluating Mistake of Fact as a Ground of Judicial Review” (paper presented to Society of Legal Scholars Conference, Preston, September 2019).
F Other Materials
Crown Law The Judge Over Your Shoulder (2019).
Yoav Zionov and Hanna Wilberg “Mistake of Fact NZ High Court Cases Since 2004” (2016).
NZLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.nzlii.org/nz/journals/UOtaLawTD/2023/8.html