Air New Zealand Limited v Commerce Commission
[2025] NZHC 3230
•29 October 2025
IN THE HIGH COURT OF NEW ZEALAND WELLINGTON REGISTRY
I TE KŌTI MATUA O AOTEAROA TE WHANGANUI-A-TARA ROHE
CIV-2025-485-94
[2025] NZHC 3230
UNDER the Judicial Review Procedure Act 2016, Part 30 of the High Court Rules 2016 and the
common law
IN THE MATTER of Airport Services Input Methodologies (IM
Review) Amendment Determination [2023] NZCC 34
AND an application for judicial review BETWEEN
AIR NEW ZEALAND LIMITED
First Applicant
QANTAS AIRWAYS LIMITED
Second ApplicantBOARD OF AIRLINE
REPRESENTATIVES OF NEW ZEALAND INC
Third Applicant
… continued
Hearing: 16 and 17 September 2025 Counsel:
A M Boberg and S C Trevella for Air New Zealand Ltd
A S Butler KC, S C Keene and V M A Fowler for Qantas Airways Ltd
T J A Lindsay and S J McNae for Board of Airline Representatives of New Zealand Inc
J D Every-Palmer KC, R S May, S T Hartley and A A Cox for Commerce Commission
K M Massey for New Zealand Airports Association and Auckland International Airport Ltd
J B M Smith KC and T D Smith for Wellington International Airport and Christchurch International AirportJudgment:
29 October 2025
JUDGMENT OF RADICH J
AIR NEW ZEALAND LTD v COMMERCE COMMISSION [2025] NZHC 3230 [29 October 2025]
continued …
AND COMMERCE COMMISSION
First Respondent
NEW ZEALAND AIRPORTS ASSOCIATION
Second Respondent
AUCKLAND INTERNATIONAL AIRPORT LTD
Third Respondent
WELLINGTON INTERNATIONAL AIRPORT LTD
Fourth Respondent
CHRISTCHURCH INTERNATIONAL AIRPORT LTD
Fifth Respondent
Table of Contents
The legislative scheme [8]
The purpose and nature of Part 4 regulation [18]
Input methodologies [21]
Information disclosure regulation [25]
The determination of input methodologies by the Commission [28]
The building block method [28]
Weighted average of capital component of input methodologies [31]
The 2023 Input Methodology Review and the Determination [35]
The 2023 Input Methodology Review [37]
The appeals [50]
The legal framework for input methodology appeals [52]
The hearing of the appeals and the Closed Record Judgment [59]
The amendment commitment [63]
The process that led to the judicial review proceeding [67]
Causes of action in the judicial review proceeding [75]
The nature and relevance of the R code errors [79]
Expert evidence on the interplay between the R code errors and the WACC [83]
The impact of the appeals on relief that is available through
judicial review [94]
Authorities [95]
The Airlines’ position [106]
Discussion [108]
The appeal pathway is adequate and more appropriate [108]
The place of judicial review [114]
Other discretionary factors in considering relief – collateral
attack and delay [122]
The judicial review causes of action [133]
First cause of action – Incontrovertible mistake of fact [134] The parties' positions [135]
A distinction between an error of law and a mistake of fact? [139] Application to this case [160]
The second and third causes of action – taking into account irrelevant considerations, failing to take into account relevant considerations and unreasonableness [164]
Outcome [165]
Costs [166]
[1] Over eight days in late July this year, the Court1 heard appeals by the Airports Association Inc and Auckland, Wellington and Christchurch International Airports2 against a decision of the Commerce Commission.3 The applicants in this judicial review proceeding4 were joined as respondents to the appeals.
[2] The decision that is the subject of the appeals was an input methodology determination relating to the ongoing regulation of airports under Part 4 of the Commerce Act 1986.5 I will refer to it in this decision as the Determination.
[3] The bespoke appeal regime for input methodology determinations under the Act6 includes a merits-based test and a prescribed set of things the Court may do if an appeal is allowed.
[4] Central to both the appeals and this proceeding is a set of errors in the computer code (the R code) written by the Commission to carry out various statistical operations that fed into weighted average cost of capital and leverage figures that are a prominent part of the Determination (the R code errors).
[5] The R code errors are one of three grounds of appeal advanced by the Airports in the appeal proceedings. The Commission has admitted the errors in the appeal proceedings and argument has been heard on ways in which the Court may or may not deal with the errors in the context of orders that are available on the appeal.
1 Sitting with two lay members under s 52ZA of the Commerce Act 1986.
2 In this decision, I refer to New Zealand Airports Association Inc, Auckland International Airport Ltd, Wellington International Airport Ltd and Christchurch International Airport Ltd collectively as the Airports.
3 New Zealand Airports Association Inc and Others v Commerce Commission and Others – CIV- 2024-485-61, CIV-2024-485-66 and CIV-2024-485-68. In this decision, I refer to the Commerce Commission as the Commission.
4 In this decision, I refer to Air New Zealand Ltd, Qantas Airways Ltd and Board of Airline Representatives of New Zealand Inc collectively as the Airlines.
5 Input methodologies are the rules and processes the Commission sets upfront to help provide certainty on how it will regulate services that are subject to Part 4 of the Act; which include airline services.
6 Commerce Act 1986, ss 52Z–53.
[6] However, the Airlines – as applicants in this proceeding – say that the errors are of such significance that the Determination should be quashed in its entirety through this judicial review proceeding; rather than being addressed through the appeals. The appeal regime, they say, is not equipped to deal with them.
[7]The issues, then, in this proceeding are:
(a)whether, in the exercise of its discretion in granting relief, relief could be contemplated on judicial review in light of the remedies available through the appeal process;7 and
(b)if so, whether the grounds of review – centred primarily on mistake of fact as an independent ground of review – can be made out such that the Determination should be quashed.
The legislative scheme
[8] The background to this proceeding is a little complex. It is essentially the same as that for the appeals, which took eight days to be heard. It is necessary to set the background out methodically in order to have an understanding of the nature of this judicial review proceeding. The starting point is the unique legislative scheme that sits behind it all.
[9] But first, let me endeavour to capture the scheme and the decision-making that forms part of it in a nutshell.
[10] Part 4 of the Commerce Act provides for the regulation of the price and quality of goods or services in markets where there is little or no competition and little or no likelihood of a substantial increase in competition.8 Part of the regulation involves the Commission making decisions about the methods it will use to determine (described in the Act as input methodologies) the asset base and cost of capital for firms that are
7 Consideration of relief would, ordinarily, follow consideration of the grounds of review. But it is a determinative issue here so it is addressed first.
8 Commerce Act, s 52.
regulated under Part 4.9 A material part of the Determination involves the Commission calculating the weighted average cost of capital (WACC) for regulated firms. The weighted average cost of capital, together with other costs that form input methodologies, will play a material part in establishing prices that may be charged for goods or services by regulated firms.10
[11] One part of the assessment that leads to the determination of the WACC for regulated firms in a particular market involves fixing a value for the measure of their market risk; their equity beta.11 Another involves the fixing of a value for the ratio of debt to equity; their leverage.12 In order to fix those values for a future period, the Commission looks at data over recent years from a number of airport companies, internationally, that it sees as being comparable to New Zealand airports.
[12] The Commission reviewed its input methodologies for airport services in 2023. In the review process, the Commission applied different criteria – to identify the airport companies it saw as being comparable – to the criteria it had used when it first determined input methodologies in 2010 and when it next reviewed them in 2016. The different criteria saw it using data from nine airports, rather than from the 25 to 26 airports it had used previously.
[13] In addition, in setting the equity beta in the 2023 amendment process, it faced a conundrum in terms of what it should do about the data from the comparable airports during the time of the Covid-19 pandemic; the asset beta data in particular.13 Because airport companies were, largely, shut down altogether during the pandemic, their asset beta figures were considerably higher than they would otherwise have been. The
9 The High Court said in Wellington International Airport Ltd v Commerce Commission [2013] NZHC 3289 [Wellington International Airport] at [20] that input methodologies “set the rules” for making the calculations and determinations that are “central to price regulation under Part 4”.
10 However, the form of regulation under the Act for airports is known as information disclosure regulation which, as discussed in below, does not compel the charging of certain prices.
11 The equity beta is expressed as a number in the Determination. For example, a beta of 1.0 shows that a firm’s stock is as volatile as the broader market, while a beta that is larger than 1.0 indicates greater volatility. The equity beta need not necessarily be expressed as a number in an input methodology determination. It could instead be expressed as a method or a formula.
12 Leverage is expressed as a percentage in the Determination. Equally, it could be expressed as a method or a formula.
13 In past reviews, the Commission had looked at average asset beta data from comparable airports using multiple rolling five-year periods. The asset beta removes the effect of leverage from an equity beta so as to reflect the underlying risk of the business.
Commission decided to exclude data over a time period during which it considered the equity markets were affected by the pandemic and then applied an uplift to the beta calculation for the pre-Covid period, in order to account for the potentially increased forward-looking systematic risk of another pandemic (the pandemic adjustment).
[14] In the appeal proceedings, the Airports challenged the Commission’s change in its selection of comparator airport companies and the pandemic adjustment it made during the estimation period. They challenged also the sample period used for the Commission’s leverage calculations. They challenged, as a consequence the numeric values given for the equity beta and leverage inputs in the Determination.14
[15] The story becomes more complex than that. In its decision-making process, the Commission carried out various statistical calculations using R code authored by the Commission.15 The outputs from those calculations informed the Commission’s ultimate exercise of judgment in determining the equity beta which, in turn, forms a material part in the calculation of the WACC.
[16] In the event that one of the first two grounds of appeal succeeds (relating to the set of comparator firms and the Covid adjustment) then the R code errors are no longer an issue. However, in the appeals, in the event that neither of the first two grounds of appeal succeed, such that the R code errors, admitted by the Commission, are the only remaining issue, a live issue arises as to whether, on the appeal test that applies (described further below), the Court is able to allow the appeal and to refer the Determination back to the Commission with appropriate directions. That, the Airlines say, is where the judicial review proceeding comes in because, in their view, the Court is not able to refer the Determination back. In all of the circumstances, they say that the errors are such that the proceeding has miscarried altogether and that judicial review is the appropriate remedy because the Determination needs to be quashed in its entirety and the process needs to start again.
[17]I turn now to outline the statutory scheme in more detail.
14 Commerce Commission Airport Services input Methodologies (IM Review) 2023 Amendment Determination 2023 [2023] NZCC 34.
15 The R code is a series of instructions to carry out statistical calculations in a programming language known as R.
The purpose and nature of Part 4 regulation
[18] In determining an appeal against an input methodology determination, a Court may only exercise its powers if it is satisfied that an amended or substituted input methodology will be “materially better” in meeting the purpose of Part 4 of the Act, or the purpose of input methodologies themselves or both.16 Part 4 applies because airports provide aeronautical services in markets in which there is little or no competition and little or no likelihood of a substantial increase in competition.
[19] The purpose of Part 4 of the Act is to promote the long-term benefit of consumers in markets referred to in markets of that sort by promoting outcomes that are consistent with outcomes produced in competitive markets such that, by way of summary, suppliers of regulated goods or services have incentives to innovate, invest and improve efficiency while sharing with consumers the benefits of efficiency gains and such that they are limited in their ability to extract excessive profits.17
[20] Section 52A does not seek to mimic features of, or behaviours in, workably competitive markets because markets that are regulated under Part 4 are not workably competitive at all. Rather, Part 4 introduces machinery of a type that does not feature in workably competitive markets in order to produce outcomes of a kind that are produced in workably competitive markets.18
Input methodologies
[21] A material part of the machinery introduced through Part 4 is known as “input methodologies”. They are rules, requirements and processes that must be applied to determine prices or quality standards applying to the goods or services in question.
[22] The purpose of input methodologies is to promote certainty for suppliers and consumers in relation to the rules, requirements and processes that apply to regulation under Part 4.19
16 Commerce Act, s 52Z(4).
17 Section 52A.
18 Major Gas Users Group Inc v Commerce Commission [2024] NZHC 959 [Major Gas] at [14].
19 Commerce Act, s 52R.
[23] Under s 52T, input methodologies are to be prepared for evaluating or determining the cost of capital, the valuation of assets, the allocation of common costs, and the treatment of taxation.20
[24] In determining input methodologies, the Commission is to follow a prescribed process including through publishing a draft methodology and giving interested people a reasonable opportunity to give their views.21 The Commission must review each input methodology at, at least, seven-year intervals.22
Information disclosure regulation
[25] There are three different types of regulation under Part 4: “information disclosure regulation”, “negotiate/arbitrate regulation”, and “defaults/customised price-quality regulation”.23
[26] The Airports are subject to information disclosure regulation.24 Accordingly, they are not restricted in the level at which they may set charges (unlike the case under default or customised price-quality regulation). Rather, regulation operates alongside the price-setting regime in the Airports Authorities Act 1966 under which, approximately every five years, airports are to consult with substantial customers on charges and any major capital expenditure plans.25
[27] The Commission will review these five-yearly price-setting events to determine whether prices set are consistent with the purpose of Part 4 of the Act. It will compare the airports targeted returns against the Commission’s estimated WACC, calculated in accordance with its input methodologies. Moreover, a mechanism is provided for more onerous forms of regulation to be applied for airports should the Commission conclude that information disclosure is not achieving the Part 4 purpose of restraining excessive profits.26
20 Section 52T(1)(a). Other matters to be included in input methodologies are described in s 56T(1)(b)–(d).
21 Section 52V.
22 Section 52Y.
23 Under sub-parts 4, 5 and 6, respectively of Part 4.
24 Sections 56A and 56C.
25 Airports Authorities Act 1966, ss 4B and 4C.
26 See Commerce Act, ss 56F–56L.
The determination of input methodologies by the Commission
The building block method
[28] The design of the input methodologies incorporates what is known as the “building block” method of economic regulation. The method is based on the notion that workably competitive markets are those that produce prices based on costs.27
[29]The building block method calculates regulated revenue on the following basis: regulatory asset base × cost of capital
+ depreciation + operating expenditure + tax
- revaluation gains (or + revaluation losses)
- other income
[30] As can be seen, one of the input methodologies determined by the Commission under s 52T relates to the cost of capital component of the building block model. It is determined on a weighted basis.
Weighted average of capital component of input methodologies
[31] The methodology used by the Commission to determine the WACC involves adding the cost of debt (multiplied by leverage) to the cost of equity (multiplied by leverage). A material part of the calculation of the cost of equity involves setting an equity beta. And a material part of determining the equity beta involves determining the asset beta.28
[32] In order to determine the asset beta, the Commission takes a set of comparable firms and, through the use of Bloomberg data, estimates the average asset beta for those firms. A six-step process is involved:29
(a)Step 1 – identify a set of relevant comparator firms;
27 Wellington International Airport, above n 9, at [43]; and Major Gas, above n 18, at [27].
28 See n 11 and n 13 for descriptions of these terms.
29 The six-step process has been used by the Commission in its 2016 input methodology review, in the 2020 fibre input methodology review and in the 2023 input methodology review.
(b)Step 2 – estimate the equity beta for each firm in the set;
(c)Step 3 – de-lever each equity beta estimate to get an estimated asset beta for each firm in the set;
(d)Step 4 – calculate an average asset beta for the set;
(e)Step 5 – apply any adjustments for regulatory differences or differences in systematic risk across services to get the average asset beta for the set; and
(f)Step 6 – re-lever the average asset beta for the set to an equity beta estimate using the Commission’s assumed notional leverage.
[33]In the appeals, the Airports say that:
(a)The Commission changed its approach to the selection of the set of comparator firms, preferring a materially narrower set than used in previous Determinations. They say that the Commission’s previous approach is materially better.
(b)The Commission changed its approach to the estimation window used to calculate the average asset beta for the set. Rather than taking the average of the two previous five-year periods of observations, it excluded observations in a time-period in which the Commission considered equity markets were affected by the Covid-19 pandemic and then applied an adjustment to the beta calculated in the pre-Covid period to reflect the risk of future pandemics. It is said that, in applying that pandemic adjustment, the Commission used a method modelled on that developed by the economic consultancy, Flint Global. It is said that, not only was it inappropriate to apply the adjustment but, when applied, it was applied incorrectly. It is said that the Commission’s previous approach is materially better.
[34] In addition, the Airports say that the Commission’s implementation of the pandemic adjustment methodology contained mathematical errors – described further below.
The 2023 Input Methodology Review and the Determination
[35] The Determination amends the Commerce Act (Specified Airport Services Input Methodologies) Determination 2010, amended previously in 2016. It applies to specified airport services relating to aircraft and freight activities, airfield activities and specified passenger terminal activities.30 The Determination is divided into parts providing methodologies for cost allocation, asset valuation, treatment of taxation and, in Part 5, the cost of capital.
[36] Part 5 provides the methodology for estimating the weighted average cost of capital. It provides numeric values or methods for determining each aspect of the WACC calculation. In some cases, a single figure is provided. In others, a methodology is described through words and formulae. The two entries in question in the appeal proceedings are under the heading, in 5.2, “Fixed WACC parameters”. They are as follows:
(1) ‘Leverage’ is 23 per cent
…
(5) ‘Equity beta’ is 0.87.
The 2023 Input Methodology Review
[37] The 2023 review process began in July 2022 when the Commission issued an “open letter” seeking views on emerging issues for electricity networks, gas networks and airports; including on the impacts of Covid-19.
[38] Having received views on the letter, the Commission issued a “Process and Issues Paper” and then, subsequently, a “Draft Framework Paper” and then, subsequently, a “Final Framework Paper” which set out its proposed approach to the 2023 review.
30 Commerce Act, s 56A.
[39] It received submissions on each of those papers and then prepared a draft “Cost of Capital Topic Paper” and draft reasons. It sought and received submissions and cross-submissions on the drafts and then issued the Determination.
[40] Sitting behind the Determination are three supporting papers. Each is dated 13 December 2023, as is the Determination itself.
[41] The first is entitled “Context and Summary of Final Decisions”. It discusses the review process, the feedback received, provides summaries of the decisions and describes factors that have led to change.
[42] The second is entitled “Report on the Input Methodology Review 2023”. It identifies the decisions made in the 2023 review31 and compares them with the terms of comparable methodologies in the 2010 determination and in the 2016 amendment.
[43]The third is the “Cost of Capital Topic Paper” (the Reasons Paper).32
[44] Among the matters that are the subject of discussion in the Reasons Paper are the two issues that are central to the appeals.
[45] The first relates to the selection of the set of comparator firms – Step 1 of the six-step process referred to in [32] above. The Commission explained that it had reduced the number of firms in the comparator set through the use of several new factors – or filters – to exclude firms that were not seen as being comparable to a major airport trading in New Zealand.33 The new filters reduced the number of firms in the comparator set from 25 in the 2010 input methodology assessment and 26 in the 2016 amendment, to nine.
[46] The second relates to the sample period used and the pandemic adjustment made to the asset beta calculations. For step 2, in the six-step process referred to in
31 It does so for the separate determinations issued for electricity distribution services, for Transpower, for gas distribution and transmission services and for airport services.
32 Commerce Commission Cost of Capital Topic Paper: Part 4 Input Methodologies Review 2023 – Final Decision (Commerce Commission, ISBN 978-1-991085-69-6, 13 December 2023) [Reasons Paper].
33 Reasons Paper, above n 32, at [4.73].
[32] above, it used a sample period from October 2007 to February 2020 so as to exclude the Covid-19 period.34 It divided the sample period into two five-year windows plus a two-year window and used the weighted average of those windows.35
[47] For steps 3 and 4, the smaller comparator set and the new sample period was used to calculate a pre-Covid asset beta of 0.63.36 The Commission then made the pandemic adjustment: it uplifted the pre-Covid asset beta to account for the risk- adjusted impact of a Covid-like pandemic on the asset beta over the next regulatory period. The uplift was modelled on a method developed by Flint Global. It produced a pandemic uplift range of 0.02 to 0.08.37 The Commission chose an uplift of 0.04 which gave a final asset beta of 0.67.
[48] For step 5, the Commission did not apply, as it had in the past, a 0.05 downward asset for any difference in the asset betas for the aeronautical and non-aeronautical parts of an airport’s business.
[49] For step 6, the Commission considered the average leverage of the comparator set over different periods and then exercised judgment to determine a notional leverage of 23 per cent, which was then used to re-lever the pandemic-adjusted asset beta to give a final equity beta of 0.87.
The appeals
[50] The Airports have, in the appeal proceedings, each appealed from the amendments made in Part 5 of the Determination that are set out in [36] above on the basis of the workings summarised in [31]–[32]. They say that the changes made in 2023, through the comparator firms used, the sample period, and the pandemic adjustment applied should not have been made and that, in terms of s 52Z of the Act,
34 In 2010, it had used a sample period of the most recent 10 years. It divided the sample period into six five-year overlapping windows and used the average of those windows. In 2016, it used a sample period of 10 years but refined its approach by dividing the sample period into two five- year windows and using the average of the windows.
35 Reasons Paper, above n 32, at n 289.
36 The Commission made R code errors in this calculation, which are addressed further below.
37 Reasons Paper, above n 32, at [4.189] and [4.209]. The R code errors affected this calculation also.
a materially better input methodology will be one that, in calculating the leverage and equity beta figures, does not involve an assessment that includes those changes.
[51] In addition, in their notices of appeal, the Airports each alleged that the Commission made R code errors:
(a)In its notice of appeal, New Zealand Airports Association said that there were “computational errors in the non-Covid-19 asset beta calculation” and a “calculation error” in “the Commission’s implementation of the Flint Global Method to estimate pandemic risk”.
(b)In its notice of appeal, Auckland International Airport said that the Commission erred by “making errors when calculating the asset beta” and that it “incorrectly calculated asset beta estimates and incorrectly implemented a weighted regression to estimate a pandemic uplift”.
(c)In their notices of appeal, Wellington International Airport and Christchurch International Airport said that the Determination does not accurately or appropriately implement the decisions and reasons set out in the Reasons Paper because it contains “data errors and mathematical errors” and contains “programming that is inconsistent with the decisions and reasons set out in the [Reasons Paper]”.
The legal framework for input methodology appeals
[52] The Commerce Act prescribes a unique “materially better” test in appeals from input methodology determinations. Section 52Z(3) and (4) of the Act are in the following terms:
52Z Appeals against input methodology determinations
…
(3) In determining an appeal against an input methodology determination, the court may do any of the following:
(a)decline the appeal and confirm the input methodology set out in the determination:
(b)allow the appeal by—
(i) amending the input methodology; or
(ii) revoking the input methodology and substituting a new one; or
(iii) referring the input methodology determination back to the Commission with directions as to the particular matters that require amendment.
(4)The court may only exercise its powers under subsection (3)(b) if it is satisfied that the amended or substituted input methodology is (or will be, in the case of subsection (3)(b)(iii)) materially better in meeting the purpose of this Part, the purpose in section 52R, or both.
[53] As the Court said in Major Gas Users Group Inc v Commerce Commission, it is not necessary to consider, in the first instance, whether the Commission erred and to then go on to address the materially better question. Rather, the single question is whether an amended or substituted input methodology will be “materially better” than the Commission’s.38
[54] The onus is on the Airports to satisfy the Court that a different input methodology would be materially better.39 In undertaking that assessment, a level of deference to the expertise of the members of the Commission is not strictly needed. While the Commission is a specialist tribunal, so too is the Court when it sits on appeal with two lay members.40
[55] Another feature of the appeal regime for input methodology decisions is the closed record requirement. Section 52ZA(2) provides that the appeal “must be by way of rehearing and must be conducted solely on the basis of the documentary information and views that were before the Commission when it made its determination, and no party may introduce any new material during the appeal”.
[56] Accordingly, in the appeals, the Court needs to consider whether the changes in approach proposed by the Airports would be such as to result in an input methodology, with a focus on the particular numeric values in question, that would be materially better. If that is the conclusion it reaches, then, in terms of s 52Z(3)(b), it could allow the appeal if it is in a position to amend the input methodology (by replacing the numeric values that are in question), it could revoke the input
38 Wellington International Airport, above n 9, at [38], [164] and [157].
39 At [143].
40 Commerce Act, s 52ZA(3); and Major Gas, above n 18, at [40].
methodology and substitute a new one, or it could refer the Determination back to the Commission with directions as to the particular matters that require amendment.
[57] As the Court in Wellington International Airport said, the “reference back” power is drafted narrowly and is considerably more constrained than the broad reference-back position in s 94 of the Act, which relates to appeals from determinations other than input methodology determinations.41
[58] That is because a reference-back order under s 52Z(3)(b)(iii) must achieve a final known outcome: the Court must be able to specify the amendments it wants the Commission to undertake with sufficient detail to enable it to be satisfied that the materially better outcome will in fact be produced.42
The hearing of the appeals and the Closed Record Judgment
[59] The appeals were heard by the Court (the members of which were Radich J, Dr Jill Walker and Professor Stephen King) from 21–30 July 2025. The Court’s decision is reserved.
[60] During the course of the hearing, an evidential point arose. The issue was whether, in circumstances in which the Commission had accepted the R code errors, the closed record rule (referred to in [55] above) was such as to prevent the Commission from going on to explain how the equity beta and WACC figures might be affected if the R code errors were corrected. The Airlines’ position was that, as the corrected figures were not before the Commission when it made the Determination, the closed record rule would prevent their introduction on appeal. In a judgment given during the course of the appeal hearing,43 it was held that the amended figures – for the adjusted pre-Covid asset beta and the calculation of the pandemic uplift – could be put before the Court in argument.44 I refer to this judgment as the Closed Record Judgment.
41 Wellington International Airport, above n 9, at [182].
42 At [191].
43 Following an interim decision of 16 May 2025 where I (sitting alone) held that the references to the corrected figures should be withheld, New Zealand Airports Association Inc v Commerce Commission [2025] NZHC 1205 at [34] and [37].
44 New Zealand Airports Association Inc v Commerce Commission [2025] NZHC 2128 [the Closed Record Judgment]. The Closed Record Judgment followed a results decision given orally
[61] Importantly, for the purposes of this judicial review proceeding, a central plank in the Court’s decision in admitting the corrected figures into evidence in the appeal was to enable it to be in a position to understand and make a decision based upon the spectrum of potential relief available on the R code errors.45 The Court observed that, in the first place, in the event that one or both of the first and second grounds of appeal are made out, and if relief is granted on them, then the specific relief for the R code errors is unlikely to be necessary. However, if neither the first nor the second ground of appeal is made out, the Court observed that the positions of the parties reveal a spectrum of potential relief on the R code errors:
(a)On the Airports’ case, knowing what the corrected figures are would enable the Court to refer the Determination back to the Commission and direct it to recalculate the equity beta in light of the admitted R code errors.46
(b)The Commission’s position is that determination of the equity beta is not purely mechanistic and that the asset beta and pandemic adjustment numbers are but two aspects in a multifaceted process in which the Commission draws upon multiple threads to consider asset beta, equity beta and resulting WACC figures. Adjustments in the numbers in question may lead it to reassess the way in which it applies the broad range of considerations and data points in reaching a final outcome. Accordingly, the Court would not, in the Commission’s view, be in a position to refer the R code errors back to the Commission for reconsideration.47
(c)The Airlines agree with the Commission but they go further and say that the whole process has miscarried to such an extent that Part 5 of the Determination should be set aside altogether through this proceeding.48
following the hearing of argument: New Zealand Airports Association Inc v Commerce Commission HC Wellington CIV-2024-485-61, 23 July 2025.
45 At [30].
46 At [30(a)].
47 At [30(b)].
48 At [30(c)].
[62] The Court found that correcting the admitted R code errors is a mathematical exercise that can be undertaken from the information that was available on the closed record and that an understanding of the numbers was material as the Court considered the grounds of appeal and relief in the appeal proceeding.49
The amendment commitment
[63] Coming back for a moment to the spectrum of potential relief on the R code errors described in [58], the Commission’s position is that, while the R code errors are admitted:
(a)the mechanical application of the corrected numbers alone would not produce an input methodology that was “materially better”50; and
(b)relief on the appeal under s 52Z(3)(b) is not available because there is too much judgment involved on the Commission’s part for the Court to be able to amend, revoke or refer the R code errors back to the Commission for reconsideration.
[64] But, so that an outcome can be achieved in the event that the R code errors are the only remaining issue in the appeals, the Commission in those circumstances has committed to undertake a process under s 52X of the Commerce Act to amend the input methodologies to correct the errors, with the benefit of the Court’s guidance on issues of principle that arise (the Amendment Commitment).
[65] Further details of the Amendment Commitment were given in the Commission’s memorandum of 15 May 2025, in which it was said:
[2] [ …] in the event these appeals are dismissed without relief being granted under s 52Z(3)(b) of the Commerce Act 1986, it would promptly undertake an amendment process under the Act to address the R code errors (if still relevant) and any other comments made by the Court regarding the other appeal grounds (Amendment Process).
…
[4] The Commission confirms that it:
49 At [31] and [32].
50 In terms of s 52Z(4) of the Commerce Act.
4.1intends to conduct consultation of the type referred in s 52V of the Act as part of the Amendment Process (regardless of whether the threshold of a proposal “to amend an input methodology by making a material change” in s 52X(1) is met) and
4.2accepts that, where a s 52V process is conducted, s 52Z affords participants in that process with a significant interest in the matter (which it is accepted would include Air New Zealand) a right of appeal.
[66] In an exchange of correspondence between counsel for the Commission and for the Airlines and the Airports, the Commission provided further details of the Amendment Commitment to take effect if the Airports’ appeals are unsuccessful. Paragraph 3 of the letter from counsel for the Commission to counsel for the Airports and Airlines of 8 September 2025 is in the following terms:
3. In terms of the scope and detail of the exercise contemplated under the Amendment Commitment, the Commission can indicate the following, which represent a staff expectation that has not yet been finally approved by the Division of the Commission responsible for determining these matters:
3.1The review would use the 2023 IM Determination as its starting point.
3.2The primary subject matter of the review would be to remedy the R Code errors (including the “third error” if established) and to address any matters raised by the Court in the merits review appeal.
3.3The Commission would not, however, rule out revisiting any aspect of Part 5 of the 2023 IM Determination as part of that exercise. It has acknowledged in submissions that remedying the R Code errors may involve revisiting other aspects of its reasoning, including the downwards adjustment (which the Airlines have asked about specifically).
3.4The Commission would expect to take into account new evidence and data that is reasonably available at the time of the review.
3.5The Commission anticipates that this process would take approximately nine months following receipt of the Court’s judgment, with the precise timing of commencement to be determined (potentially – but not necessarily – subject to any appeals).
3.6The Commission may include other steps, such as a consultation on the scope of the review or an issues paper, but it would be premature to set out any staff expectations at this stage.
…
The process that led to the judicial review proceeding
[67] The Airports identified the R code errors in broad terms before the expiry of the appeal period and included them as a ground of appeal. The Airports obtained an expert report on the subject51 and on 2 April 2024 wrote to the Commission, attaching the report and setting out the basis of their claim that R code errors had arisen. The letter and report were published on the Commission’s website on 9 September 2024, following the Commission’s verification of the errors.
[68] The Commission then filed a memorandum in the appeal proceedings on 25 October 202452 in which it recorded its concession that the R code errors had arisen. Having made the concession, the Commission said that it did not make any admissions as to how its analysis or assessment of the asset beta for airports might have differed had the errors not occurred or whether any changes to its methodology might have resulted.
[69] In a further memorandum on 4 December 2024, the Commission admitted that the R code errors affected the pandemic uplift range used in estimating the equity beta and went on to explain how the equity beta and WACC might be affected if the figures resulting from the correction of the R code errors were applied mechanically – without the Commission exercising further judgment or undertaking reasonableness checks.
[70] On 14 November 2024, the Airlines had, through a memorandum of counsel, indicated their view that it would be preferable for the appeals to be stayed pending the Commission amending the input methodologies to address the R code errors. Their position is the same as that advanced in this proceeding – that addressing the R code errors through the appeals is not possible given the scheme of the Act such that they ought to be addressed before the appeals are determined.
[71] Ultimately, the Airlines sought an adjournment of the hearing of the appeals and a stay of the appeal proceedings pending the Commission amending the Determination in the first instance under s 52X of the Commerce Act.
51 Competition Economics Group, Incenta and Housten Kemp Responding to Coding Errors made by NZCC (1 April 2024).
52 Having circulated a draft to counsel for the parties to the appeals in advance.
[72] In a decision of 19 December 2024, I declined the application.53 My reasons for doing so were, by way of summary, as follows:
(a)I was not satisfied that it was appropriate to determine the R code errors conclusively in advance of considering all issues on appeal, together. The issues raised by the Airlines were (and continue to be) substantive questions on the availability and form of relief. The determination of the two other alleged errors in the appeals (relating to the comparator set and the pandemic adjustment) could overtake the R code errors and render consideration of the R code errors unnecessary. The Airports raised three primary sets of issues in the appeals and are entitled to have answers from the Court on them.54 In the event that the only remaining issue on appeal was the Court’s jurisdiction to consider R code errors, then the Amendment Commitment came into play. To reach conclusions on the availability of relief before hearing the appeals would be to get ahead of ourselves in circumstances in which the Airports are entitled to their appeal on all issues.55
(b)It was not clear – in advance of the determination of the appeal proceedings – that there is no jurisdiction on the Court’s part to refer the R code errors back to the Commission under s 52Z(3)(b)(iii) as the Airlines submit. The parties’ respective positions on that is the subject of considerable contest and a finding from the Court, one way or the other, is needed in the first instance.56
(c)The Commission’s position was (and remains) that if these appeals had not been brought and if the Commission had found R code errors in the Determination, then it would most likely have used a s 52X process to correct them. But where, as here, appeals are on foot, the orders sought by the Airlines would interrupt vested appeal rights and those rights
53 New Zealand Airports Association Inc v Commerce Commission [2024] NZHC 3948 [Decision declining stay of appeals].
54 At [32].
55 At [33].
56 At [34]–[38].
should not be stifled by the potential judicial review proceeding the Airlines had at that stage proposed.57
(d)At a practical level, if the appeal proceedings were adjourned, then the Commission would be unlikely to undertake the amendment process in the face of the extant appeals and, accordingly, whether the judicial review application was successful or unsuccessful, there would be significant delays in the hearing of the appeals, or any related appeals.58
[73] However, the Airlines maintained their position. On 11 February 2025, they filed this proceeding and then applied for it to be heard and determined on an expedited basis, before the fixture for the appeals. In a decision of 28 February 2025, I declined the application and directed that the judicial review proceeding would be heard immediately after the hearing of the appeals.59
[74]By way of summary, I found that:
(a)Any of the grounds for bringing appeals from input methodology decisions could equally be pleaded in judicial review language. But it could not properly be said that, because the outcome of a judicial review proceeding can be a finding of invalidity, a judicial review cause of action should be pursued in the first instance. That would undermine the appeal process altogether.60
(b)The R code errors will be addressed one way or another in the appeals, either through an order allowing the appeal and revoking or amending the input methodology, or referring it back, or through the Amendment Commitment.61
57 At [44]–[48].
58 At [51]–[53].
59 Wellington International Airport Ltd v Commerce Commission [2025] NZHC 359 at [36].
60 At [20].
61 At [22].
(c)In the event that the R code errors were determined in advance, the other issues on appeal are likely to remain such that it is most efficient to determine them collectively.62
(d)There is, in any event, the prospect of the Court declining to grant any relief on review given the existence of the bespoke appeal process. Accordingly, dealing with the appeals first is the most efficient course.63
(e)A range of practical considerations were such that it was in the interests of justice for the judicial review proceeding to be heard alongside the statutory appeals.64
Causes of action in the judicial review proceeding
[75] With that rather lengthy background recounted, I turn to consider the causes of action that are advanced.
[76] The first cause of action alleges a material mistake of fact. It is said that the R code errors are an “obvious mistake as to existing fact”. It is said that calculation or other mathematical errors have been held to ground causes of action for mistake of fact, or a related ground of irrationality. It is said that the error is as to a verifiable and established fact, that the Commission has “appropriately” taken responsibility for the mistake, and that the errors played a material part in the Commission’s reasoning.
[77] The second cause of action is pleaded as an error of law on the grounds that the Commission took into account incorrect, and therefore irrelevant, considerations and that it failed to take into account relevant considerations, namely the correct results of the calculations the Commission intended to carry out.
62 At [24] and [25].
63 At [26].
64 At [30]–[35].
[78] The third, and alternative, cause of action is in unreasonableness on the basis that, as a result of the R code errors, the Determination is one that no reasonable person in the Commission’s shoes could have made.
The nature and relevance of the R code errors
[79] As explained previously in general terms, there were two primary impacts from the R code errors. In particular:
(a)When calculating the pre-Covid asset beta, the Commission intended to average weekly and four-weekly asset betas, but it only averaged weekly asset betas. If the R code had implemented correctly what the Commission had intended, then the pre-Covid asset beta would have been 0.65, instead of 0.63.
(b)When calculating the pandemic adjustment range, the Commission intended to implement a weighted least squares regression,65 but it failed to do so correctly. If the R code had implemented correctly what the Commission had intended, the pandemic adjustment range would have been 0.07–0.15 instead of 0.02–0.08.
[80] A curious feature of this proceeding is that the Airlines support the Determination they seek to set aside. In the appeal proceeding, the Airlines were respondents along with the Commission and said (in joint written submissions) that, while they did not agree with every judgement call made by the Commission, they considered that the 2023 Determination as a whole “struck a reasonable balance between the positions advanced by the various stakeholders and fell within the range of reasonable outcomes that could be expected in workably competitive markets”. They said that they elected not to appeal because, amongst other factors, they did not consider it likely that there was a materially better input methodology to that determined by the Commission. Therefore, they said, they “generally agree with the submissions filed by the Commission in support of its determination”. They went on to say that the Commission’s judgement should not be disturbed in the appeals.
65 A linear regression that assigns different weights to each relevant data point.
However, that position was expressed on the basis that it was “putting aside the Airlines’ position in the judicial review proceedings”.
[81] By way of contrast, in the introductory section of their submissions in this proceeding, the Airlines say:
In short, had the Commission not made these errors, it may well have reached a different conclusion by different reasoning, regarding asset beta, equity beta, and leverage – and, thus, WACC.
[82] They say that errors are material, that they “cannot sensibly be addressed within the context of the appeals” and that, therefore, the appropriate remedy for the Court is to quash the decisions set out in Part 5 of the Determination. In this sense, it is a proceeding that is advanced essentially on a principled basis.
Expert evidence on the interplay between the R code errors and the WACC
[83] Considerable expert evidence has been filed in this proceeding on the materiality of the R code errors. Evidence of this type was introduced seemingly in order to show that, given the closed record rule, there is an information vacuum in the appeals such that it could not be known what submissions might have been made, or evidence produced for the Commission, if the errors had been known and before the Determination was made. The point advanced is that information of this type needs to be interrogated by the parties before a decision could be made on the R code errors; an option that is not available in the appeal process.
[84] An overriding point is that the Closed Record Judgment allowed the corrected figures to be introduced into evidence on the appeal. So, the appeal court has a clear understanding of what the raw asset beta, equity beta and leverage figures would be with the R code errors corrected. Really nothing more than that is needed for the appeal court to be able to consider the range of options that are available to it in order to determine the ground of appeal relating to the R code errors. An understanding of submissions that would have been made, or of what the Commission would have done, had the R code errors been known at the outset, are not directly relevant to the issues on judicial review. The errors have been admitted. Corrected figures are known and options exist to address them.
[85] Dr Taylor, for the Airlines, and Mr Balchin, for the Airports, have advanced a range of possible WACC figures in order to illustrate what could follow if the Commission was to engage (or had engaged) in reasoning unaffected by the R code errors. The short point is that there is no one inevitable outcome. A level of judgement is needed to arrive at an estimate of an appropriate asset beta and, then, to arrive at an appropriate estimate of the WACC. The difference between the experts is as to what the reasonableness boundaries for those estimates might be. For example, Dr Taylor explored the plausible range for the asset beta; with and without the R code errors and on the basis of different scenarios, with or without a pandemic adjustment, and with different comparative periods. A purely mechanical assessment with the errors corrected would, Dr Taylor suggested, produce a counter-intuitive outcome that would have been unlikely to have been adopted by the Commission.
[86] Mr Quach’s evidence, for the Airports – which was supported by Mr Balchin for the Airports – was that, if the R code errors had not been made, a range of 0.72–
0.80 for the pandemic-adjusted asset beta would have resulted and it would have been entirely open to the Commission to have selected a figure within that range (recalling that the figure chosen by the Commission in its Determination for the asset beta was 0.67). In Mr Quach’s view, 0.72 was the “lowest plausible asset beta”.
[87] A little time was spent by the experts on what might be taken, in terms of corrections to the R code errors, from the outcome of Auckland International Airport’s fourth price-setting event (PSE4).66
[88] On 7 June 2023, Auckland International Airport set its prices for the July 2022 to June 2027 period through PSE4. The Commission then conducted a review of Auckland International Airport’s PSE4 pursuant to the information disclosure provisions in Part 4 of the Commerce Act.67 Although Auckland International Airport made its PSE4 decisions when the 2016 input methodology was still in force, the Commission decided to use the 2023 input methodologies, but making provision for
66 PSE4 is a price-setting event under the Airports Authorities Act 1966, as described in [26] and
[27] above.
67 On 3 September 2024, New Zealand Airports Association and Auckland International Airport made submissions in the review of PSE4. The submissions addressed, among other things, the R code errors.
the R code errors. That exercise resulted in the Commission using a mid-point for the asset beta in the range of 0.715.68
[89] The Commission says that, in those circumstances, Mr Quach’s “lowest plausible asset beta” of 0.72 could not be correct while the Airports, in response, say that there is no basis for that submission.
[90] The Airlines say, with reference to Mr Taylor’s evidence, that a merely mechanical correction of the R code errors would result in a material increase in asset beta and push the beta to 0.75 (assuming all else is equal), significantly above the asset beta in the Determination of 0.67 and outside the Commission’s “plausibility range” in the Determination of 0.65–0.74, with consequential effects on the equity beta. This, in turn, the Airlines say, would result in a significant increase in the Commission’s estimation of a vanilla WACC (mid-point) for regulated airport services from
8.39 per cent to 8.92 per cent.
[91] However, the Airlines say, with reference in particular to the Commission’s amended statement of defence, that it is not possible to mechanistically ‘draw a line’ between the corrected R code outputs and the asset beta the Commission would have determined. They say that, as a result, a change to the outputs of the R code may have led to a change in the judgement exercised in relation to the pandemic adjustment or to where, within a reasonableness range, the ultimate beta would be set.
[92] I mean no discourtesy to the parties or to the expert witnesses by not undertaking a comprehensive assessment of the evidence. I provide these few examples to underline the point that there is no real difference between them as to the significance of the R code errors in terms of what the R code sought to produce. The dispute between them is on the extent of the errors and of their materiality in terms of the difference they might have made to the ultimate estimate of beta and, in turn, to the WACC. However, what the Commission would in fact have done, had it been aware of the errors, is not relevant to the issues in this proceeding. It would be wrong for the Court to pre-empt the correction process that will follow. It is enough to say
68 It is to be recalled that the asset beta used in the Determination – with the R code errors – was 0.67.
that the corrected R code is a material consideration that can be taken into account by the Court in the appeal proceedings – should it allow the appeal and grant relief under s 52Z(3)(b).
[93] But, whether it would result in a substantial change to the beta and the WACC itself is not something on which a firm conclusion can be reached in this proceeding. The most that could be said in this proceeding, viewed in its most favourable light, is that the errors are in the nature of a reviewable flaw such that, regardless of the appeals, the Determination should be set aside altogether. The expert evidence does not advance the answer to that question. Rather, the focus is on the appropriateness of the proceeding in the wake of the outcomes and remedies that are available on appeal.
The impact of the appeals on relief that is available through judicial review
[94]I address the question of whether in the circumstances of this case relief could
be available on review first, because it is a determinative question.
Authorities
[95] The Airlines emphasise the constitutional significance of judicial review. As Mr Butler KC has put it, the merits appeals for input methodologies were “intended to be additive, not subtractive”.
[96] The starting point for this analysis is a series of fundamental planks in our constitutional arrangements:69
(a)Senior courts have a constitutional responsibility to uphold values that constitute the rule of law. Judicial review is the common law means by which the courts hold public officials to account.
69 Tannadyce Investments Ltd v Commissioner of Inland Revenue [2011] NZSC 158, [2012] 2 NZLR 153 [Tannadyce] at [3]–[5] per Elias CJ and McGrath J (who were in the minority) and at [56]–
[61] per Blanchard, Tipping and Gault JJ. While the minority differed on the extent to which the privative clause in s 109 of the Tax Administration Act 1994 precluded the applicant from pursuing judicial review, their descriptions of these fundamental principles coincided.
(b)Accordingly, statutes limiting recourse to judicial review to challenge statutory decisions raise issues of constitutional concern.
(c)The concern is such that the courts will presume when interpreting legislation that it was not Parliament’s purpose to allow decision makers power conclusively to determine any question of law.
(d)Therefore, a Judge should be slow to conclude that statutory provisions were intended to preclude applications to the High Court for judicial review alleging unlawfulness of any kind. That can interfere with the full supervision by the court that is necessary to ensure that government activity conforms with the rule of law.
[97] Within these fundamental propositions sits the notion that a right of appeal does not exclude the courts’ jurisdiction in judicial review proceedings to grant relief in relation to the same subject matter.70
[98]However, as Elias CJ and McGrath J in Tannadyce went on to say:71
The courts nevertheless recognise that statutory challenge and appellate processes can provide a better means of judicial supervision of government decision making than judicial review.
[99] Elias CJ and McGrath J then approved an extract from a judicial review text which referred to the prospect of statutory appeal scheme providing a superior remedy for (in the example given there) a breach of natural justice and as being able to provide a strong justification for allowing comprehensive appeal rights to cure defects rather than through seeking judicial review.72
[100] As Kós J put it in DFS New Zealand Ltd v New Zealand Customs Service, it is clear that the existence of an alternative remedy, through a statutory appeal, is a consideration affecting the exercise of the Court’s discretion to grant relief where a
70 Judicial Review Procedure Act 2016, s 16(3).
71 Tannadyce, above n 69, at [6].
72 At [6], citing Mark Aronson, Bruce Dyer and Matthew Groves Judicial Review of Administrative Action (4th ed, Lawbook Company, Pyrmont, 2009) at 496.
reviewable error is found.73 He said, with reference to Wislang v Medical Council of New Zealand:74
Review may be refused where the appeal rights are available, adequate and more appropriate.
[101] However, he cautioned that this discretionary consideration must not be taken too far – in the case of fair trial rights in particular.
[102] In DFS, it was found that judicial review was available to quash the decision because the appeal period had expired before the applicant had become aware of the issue.75 The circumstances here differ. The R code errors were capable of being discovered within the appeal period. They were so discovered by the Airlines. And in any event, the Airlines were joined to the appeals as respondents and could raise points as they wished.
[103] There are many examples of cases in which it has been found that active appeal rights are such as to adequately protect an applicant’s interests so that judicial review relief was not necessary nor appropriate,76 or so that it was best to hear the appeal before the judicial review proceeding.77 As the Court of Appeal put it in Rees v Firth
– in the context of the adjudication scheme under the Construction Contracts Act 2002
– courts must be vigilant to ensure that judicial review does not cut across a statutory scheme that has a clear and specialised focus.78 While, in principle, any ground of review may be raised, it will be difficult to satisfy a court that intervention by way of relief is necessary in the case of a merits-based statutory scheme.
[104] In Rees, the adjudication scheme was part and parcel of giving effect to the Act’s purpose of providing a mechanism to enable money flows to be maintained on the basis of preliminary and non-binding assessments of the merits. In such
73 DFS New Zealand Ltd v New Zealand Customs Service [2012] NZHC 3279, [2013] NZAR 175 at [56].
74 At [56], citing Wislang v Medical Council of New Zealand [2002] NZAR 573 (CA) at [31].
75 At [60] and [61].
76 See, for example, Auckland Acclimatisation Society Inc v Commissioner of Crown Lands (South Auckland) [1985] 2 NZLR 94 (CA) at 103,
77 New Zealand Kiwifruit Growers Inc v Gisborne District Council [2021] NZHC 2198 at [15] and
Blair v Upper Hutt City Council HC Wellington, CIV-2005-485-1961, 21 June 2006 at [15].
78 Rees v Firth [2011] NZCA 668, [2012] 1 NZLR 408 at [27].
circumstances, it was said that it would be unlikely that errors of fact by adjudicators could give rise to a successful application for judicial review.79
[105] Another factor in considering the exercise of the Court’s discretion to grant relief that is relevant here is the principle that relief is unlikely to be granted where there is no utility in a review application. As the Court of Appeal said in Te Whakakitenga o Waikato Inc v Martin, courts will not make orders that have no utility – the court’s time is precious and it is not the function of courts to provide abstract opinions.80
The Airlines’ position
[106] The Airlines come at it from a different angle altogether. They start with the proposition that judicial review should always be available and that, if a ground of review is made out, relief would only be declined for good reason. They say that the court’s powers on appeal are not as good as those of a review court: a review court has broad powers to set aside a decision and to order reconsideration while the power on appeal will only enable amendment, revocation or referral back in the context of the “materially better” test. The successful application of the materially better test is such that – expressed in broad terms – the court can essentially craft or direct a replacement methodology. It is said that the appeal court could not apply that test or grant relief of that kind here because of the closed record and in circumstances where the type of evidence filed in this proceeding is not available on appeal.
[107] The court, it is said, has a discretion as to how it wishes to sequence the proceedings. The Airlines say that it is conceivable, for example, that the court would issue judgment on appeal followed by a judicial review decision quashing the Determination. Judicial review, it is said, is the superior regime.
79 At [27].
80 Te Whakakitenga o Waikato Inc v Marton [2016] NZCA 548, [2017] NZAR 173 at [39]. That case involved a challenge to a decision removing the respondent as chairperson of the appellant. Ultimately the case had no utility because the respondent was, subsequently, re-elected to her position. A declaration of invalidity would have made no difference to the present situation – see [46]–[50].
Discussion
The appeal pathway is adequate and more appropriate
[108] It is not as I see it a matter of asking, first, whether judicial review can provide a remedy and then saying that, if so, a particular right of appeal could not be ‘subtractive’ from it. As Kós J said in DFS, the focus should be on whether the appeal is available, adequate and more appropriate. If it is, then the constitutional responsibility of the court to uphold values constituting the rule of law will be met.
[109] Here, the right of appeal under s 52Z is available. It is being exercised actively by the Airports on the primary grounds of appeal, and on the R code errors. The Airlines are active participants in the appeal. They support the Commission’s Determination. In the appeals they describe the Determination as being reasonable and sustainable. But, in this proceeding, say that it was flawed.
[110] The appeal pathway is adequate and more appropriate. That is because the R code errors will be resolved comprehensively through the appeal process, whatever the outcome:
(a)If the outcome of the appeal is that one of the first two grounds of appeal (the comparator set of firms and the pandemic adjustment) is made out, then the R code errors are overtaken. The Court might, for example, direct the Commission to adjust the set of comparator firms or to reassess the equity beta with no pandemic adjustment, each of which provides the opportunity for the Commission to correct the R code errors at the same time. Directions of that type could be given if the appeal court thought that a materially better input methodology would result.
(b)Alternatively, if one of the first two grounds is not made out but the Court is of the view that a materially better input methodology would result with the R code errors corrected, then it might, for example:
(i)find that the information it has on appeal is sufficient to enable it to, itself, amend the input methodology to correct the R code errors or to revoke it and substitute a new one;81 or
(ii)direct the Commission to reassess the input methodology with reference to a Court-identified mid-point pandemic adjustment after correcting for R code errors;82 or
(iii)give a broader set of directions for amendment by the Commission. It might, for example, provide framework directions for the Commission to establish the asset beta and to apply an uplift derived from a correctly calculated R code output.83
(c)Alternatively, if the appeal Court decides either or both that:
(i)it cannot conclude that options of the type described in (b) above would produce an input methodology that would be materially better; or
(ii)a clear direction cannot be made as to an amendment on remission back because the influence of the correctly calculated uplift on the equity beta is not discernible;
then the Amendment Commitment comes into play.
[111] It is important that this spectrum of possible outcomes be maintained. To proceed otherwise would see the Court as pre-empting or pre-judging the appeals. Parliament has indicated clearly its intent through the bespoke statutory scheme it has created. If relief is available through the appeals, that is the most appropriate form of relief. As Mr Every-Palmer KC put it, it would not be appropriate for this Court to
81 Under s 52Z(3)(b)(i) or (ii) of the Commerce Act.
82 Under s 52Z(3)(b)(iii).
83 Again, under s 52Z(3)(b)(iii).
pre-empt (and gazump) any relief available on the appeals by, essentially, determining the appeals as a preliminary issue through this judicial review proceeding.
[112] The Airlines have levelled a range of criticisms at the terms of the Amendment Commitment. For example, they are concerned about it being conditional on the appeals being dismissed in their entirety; about the 2023 input methodology being the proposed starting point; about whether other aspects of Part 5 of the Determination would be revisited also; and about the extent to which new evidence would be taken into account.84
[113] However, I am satisfied that the Amendment Commitment is sufficiently detailed to enable an appropriate suite of possible outcomes to be assessed in the appeals. It is considerably more comprehensive than could be the case through an order in a judicial review proceeding directing reconsideration. Its final terms would depend on the substance of the Court’s decision determining the appeals. And it is hinged upon the Commission’s commitment to the Court that – if it was to come into play – it would ensure that the R code errors will be addressed (so as to ensure that the input methodology determination in force for the time being represents the best forward-looking estimate of equity beta and leverage, so as to produce a reasonable estimate of the WACC), unless superseded by the result on the appeals.
The place of judicial review
[114] The explanatory note to the Commerce Amendment Bill 2008 (which made substantive amendments to Part 4 of the Commerce Act) explained the new “merits review” provisions that are now found in ss 52Z and 52ZA. It said:85
This is in the form of an appeal to a High Court Judge assisted by 2 expert lay members (in most circumstances). The appeal provides accountability for the Commission, helps ensure that input methodologies deliver on the purpose statement, and promotes business confidence.
[115] The explanatory note goes on to say that the nature of the proposed appeal right (now embedded in s 52Z) “may be appropriate to help reduce gaming risks”.
84 Other concerns are raised about what precisely “remedying the R code errors” entails; about the length of time the review will take; and about the nature and scope of consultation.
85 Commerce Amendment Bill 2008 (201-1), explanatory note.
[116] After then referring to the more general right of appeal provided in s 91 of the Act, the explanatory note on the appeal provisions ended by saying “[f]irms will also have judicial review available to them”.
[117] That will always be the case. It will be an appropriate vehicle in some circumstances. The question here is only whether, in the circumstances of this case, the Court could exercise its discretion to grant relief in the event that a ground of review was available. For the reasons given, my conclusion is that it would not.
[118] An approach that would entertain a judicial review application that is concerned with the substance of an input methodology determination, and which would give a proceeding of that sort priority over an appeal proceeding, would risk undermining the statutory appeals scheme. The scheme could be circumvented altogether. A party would not need to meet the materially better test. It would not be constrained by the closed record. It would not have the benefit of lay members considering the issue from a substantive perspective. It would enable fresh factual and expert evidence to be introduced, rather than being limited to the closed record. It could enable proceedings to be brought outside the 20-day appeal period.86
[119] That would not, in terms of the language in the explanatory note, help to ensure that input methodologies deliver on the purpose statement or promote business confidence. It would not help to reduce gaming risks.
[120] Judicial review will be a more natural fit for issues that are not aimed at the substance of an input methodology determination; that are not capable of being the subject of a merits appeal. There is no bright line here but an example might be a proceeding founded upon a fundamental breach of natural justice or apparent bias. But a judicial review proceeding that challenges a topic that is the subject of the appeal and capable of being addressed appropriately through the appeal is at the other end of the spectrum.
[121] Accordingly, the Court could only properly exercise its discretion against granting any relief in this judicial review proceeding.
86 There is no limitation period for judicial review proceedings.
Other discretionary factors in considering relief – collateral attack and delay
[122] The Commission says that this proceeding should be seen as an inappropriate collateral attack that, essentially, is designed to bolster the prospect of the Court, one way or another, not being in a position to direct, through the appeals, a mechanical correction to the equity beta and leverage in the Airports’ favour.
[123] As mentioned above, the Airlines’ evidence does not express concerns with the equity beta and leverage figures in Part 5 of the Determination. Dr Taylor’s evidence relates to concerns about mechanical adjustments to the beta and leverage if the R code errors are corrected. But, ultimately, and as mentioned earlier, the Airlines support the Commission’s decision in the appeal proceeding. The challenge is brought by the Airlines primarily on a principled basis.
[124] In her evidence, Ms Roden for Air New Zealand,87 has said that, had Air New Zealand been made aware of the R code errors and their significance at the time the Determination was published, she believes that Air New Zealand would have sought to challenge the Determination either by way of judicial review or appeal. And Mr McKinnon, for Qantas Airways88 has said in his evidence that, had Qantas been aware of the errors at the time of the draft Determination, then it would have argued that a mechanical adjustment, resulting in an upward adjustment of the asset beta, would be perverse.
[125] The reality is that these arguments have been advanced comprehensively on behalf of the Airlines in the appeals; a factor that distinguishes this case from the likes of Tannadyce89 and DFS90 where appeal rights were not exercised.
[126] But, whatever the case, I do not see this proceeding as having been an inappropriate collateral attack. It is certainly forthright, in circumstances in which the Court had concluded in December that there was no basis for the appellants’ appeal rights to be stifled by a judicial review proceeding,91 and in which the Airlines support
87 Air New Zealand’s senior legal counsel.
88 Head of Commercial for Airports at Qantas Airways Ltd.
89 Tannadyce, above n 69.
90 DFS New Zealand Ltd v New Zealand Customs Service, above n 73.
91 Decision declining stay of appeals, above n 53, at [48].
the input methodology in question. But the Airlines have acted appropriately in promoting an issue they have seen as being of constitutional significance.
[127] The Airports say that the Airlines delayed unnecessarily in bringing the judicial review proceeding and that the delay should count against the exercise by the Court of any discretion in granting a remedy.92
[128] They say that the Airlines had sufficient notice of the R code errors from the beginning. They could, the Airports say, have been identified by a person with the relevant expertise. They refer to the Airports having been able to do that.
[129] They refer to the Airports’ notices of appeal having identified R code errors and, for example, New Zealand Airports Association and Auckland International Airport having referred to the R code errors in their submissions on the review of Auckland International Airports PSC4; submissions on which Air New Zealand cross- submitted and in which, in those cross-submissions, it referred also to the R code errors.
[130] The Airports say that, if the Airlines had been reasonably diligent, they could have applied for judicial review after the final Determination on 13 December 2023 or shortly after the Airports’ appeals were filed on 1 February 2024. Mr Smith KC put it on the basis that the Airlines have sat on their hands, believing that judicial review grounds existed but choosing neither to explore them further let alone issue proceedings on them, knowing also that arrangements were being made and pursued in good faith by the Commission and the Airports to address the R code errors as part and parcel of the appeals.
[131] The Airlines say that the errors were expressed only in the most general of terms until, having been pressed for information, the Commission filed its 4 December 2024 memorandum in which the errors were particularised. They say that, at that time, the Airlines applied promptly for an adjournment of the hearing of the appeal pending
92 Drawing upon authorities including New Era Energy v Electricity Commission [2010] NZRMA 63 at [64]; Macpherson v Napier City Council [2013] NZHC 2518, [2015] NZAR 342 at [94]; and Wellington City Council v Minotaur Custodians Ltd [2017] NZCA 302, [2017] 3 NZLR 464 at [74]–[76].
an amendment process. When that application was declined, they filed the judicial review proceeding early in the New Year and sought its urgent hearing, leading to the Court’s decision of 28 February 2025, described already.93 Moreover, as Ms Boberg said, correspondence between the Airports and the Commission during 2024 about the nature and extent of the issues were not copied to the Airlines.
[132] There is as I see it nothing deliberate in the Airlines’ approach; no conduct on their part that would affect any entitlement to relief that would otherwise be available. The Airports were seeking to establish through exchanges of correspondence with the Commission the points they were making about errors. When the Airlines did become aware of the issues, they acted appropriately. I do not see that delay as entering into the equation.
The judicial review causes of action
[133] I have found that the Court could only properly exercise its discretion against granting relief. Therefore, I express my findings on the grounds of review in relatively short form.
First cause of action – Incontrovertible mistake of fact
[134] In the first cause of action it is alleged that the R code errors are an incontrovertible mistake about a material fact. The primary issue that arises is whether mistake of fact can be a standalone grand of judicial review or whether it is not a ground in its own right and, rather, is a part of an error of law ground – where the facts found are such that a decision maker could not tenably have made the decision in decision. The threshold in the latter case is particularly high and, the respondents say, could not be met here.
The parties’ positions
[135] The Airlines say that there is sound authority for the proposition that an incontrovertible mistake of fact will, in and of itself, constitute a ground of judicial review. They say that, as is often the case in judicial review, the grounds tend to
93 At [73] above.
overlap so that it is not unusual to see the label used to describe an incontrovertible and material mistake of fact varying from one case to another, or being regarded as overlapping with other grounds (such as taking into account irrelevant considerations, failing to take into account relevant considerations, unreasonableness, or unfairness). This is why, the Airlines say, they plead three related grounds of review – error of fact, relevant and irrelevant considerations, and unreasonableness. Each ground is based on the R code errors, that impact the accuracy of the equity beta and leverage figures, which then inform the estimate of the WACC.
[136] They say that those R code errors are obvious mistakes of fact. They are computational errors that then form the basis for the exercise of judgement by the Commission in making the decision that is subject to review. They say that the error has been identified objectively and verified as a matter of pure statistical calculation; to such an extent that the Commission has taken responsibility for the mistake.
[137] The respondents say94 that there is no clear appellate statement to the effect that mistake of fact exists as a standalone ground of review in New Zealand. Rather, they say, the cases to date are explicable as going no further than providing examples of the second manifestation of error of law in Edwards v Bairstow (discussed below).95
[138] They say that the high threshold that results cannot be met here; particularly where the R code errors are not mistakes of “fact”. The Commission’s decision, it is said, while including the R code calculations, was the product of the exercise of judgement by the Commission and the Edwards v Bairstow formulation (as expressed by courts subsequently) excludes matters of evaluation where reasonable minds can differ.96
94 Mr May, for the Commission, made the primary arguments for the respondents on the substantive grounds of review.
95 Edwards v Bairstow [1956] AC 14, [1955] 3 All ER 48 (HL), at 57.
96 Referring to CREEDNZ Inc v Governor-General [1981] 1 NZLR 172 (CA) at 200; and Winton Property Investments Ltd v Minister of Finance [2023] NZCA 368 at [103].
A distinction between an error of law and a mistake of fact?
[139] The speech of Lord Radcliffe in Edwards v Bairstow isolated the two manifestations of errors of law in the following terms:97
If the Case contains anything ex facie which is bad law and which bears on the determination, it is, obviously, erroneous in point of law. But without any such misconception appearing ex facie, it may be that the facts found are such that no person acting judicially and properly instructed as to the relevant law could have come to the determination under appeal. In those circumstances, too, the court must intervene. It has no option but to assume that there has been some misconception of the law, and that this has been responsible for the determination. So there too, there has been error in point of law.
[140] In this way, first manifestation described by Lord Radcliffe captures pure errors of law while the second captures outcomes that are simply untenable.
[141] The Edwards manifestations were adopted by the Supreme Court in Bryson v Three Foot Six Ltd. The Court described the second Edwards manifestation as a state of affairs in which the ultimate conclusion of a fact-finding body is so unsupportable
– so clearly untenable – as to amount to an error of law.98 That would be the position, the Court said, only in rare cases.99
[142] When applying the second Edwards manifestation, an important distinction is to be drawn between a finding on the facts that is entirely unsupportable and a finding on which there would be a tenable basis for reasonable minds to differ. An error of law will exist in the former case, but not in the latter.
[143] There is, it seems to me, a relatively clear distinction between a factual evaluative process that miscarries to such an extent as to be captured by the second Edwards manifestation, and the identification of an error about a fact that is incontrovertible in the sense discussed by United Kingdom and Wales Court of Appeal in E v Secretary of State for the Home Department.100 There is a case to say that an error of fact of that type can stand as a separate ground of review; a ground that is not simply subsumed within the second Edwards manifestation.
97 Edwards v Bairstow, above n 95, at 57.
98 Bryson v Three Foot Six Ltd [2005] NZSC 34, [2005] 3 NZLR 721 at [25]–[26].
99 At [27].
100 E v Secretary of State for the Home Department [2004] EWCA Civ 49, [2004] QB 1044.
[144]The Court of Appeal in E said:101
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, at least in those statutory contexts where the parties share an interest in co- operating to achieve the correct result. Asylum law is undoubtedly such an area. Without seeking to lay down a precise code, the ordinary requirements for a finding of unfairness are apparent from the above analysis of CICB. 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.
[145] The mistaken fact that was under consideration in that case was an error in relation to which “if attention had been drawn to the point, the correct position could have been shown by objective and uncontentious evidence”.102
[146] As the Airlines say, calculation or other mathematical errors have been held to enable judicial review on the ground of mistake of fact, or on the related ground of irrationality.103
[147] A number of relatively recent, and prominent, cases have considered factual issues through the lens of the second Edwards manifestation, in my view primarily because the proceedings were advanced on that basis. In Charter Holdings Ltd v Commissioner of Inland Revenue, the Court said that “[t]here is now clear authority that an error of fact may properly constitute a ground of review”.104 However, in that case the error in question was in fact an error of law in the second Edwards manifestation sense because the error was in an evaluative process that lacked factual foundation.105
101 At [66].
102 At [63].
103 See, for example, Sutton v Canterbury Regional Council [2015] NZHC 313, [2015] NZRMA 93; Minister for Primary Industries v Austral Fisheries Pty Ltd (1993) 40 FCR 381, (1993) 112 ALR 211; R (Law Society) v Lord Chancellor [2018] EWHC 2094 (Admin), [2019] 1 All ER 638 at [142]; and R Torbay Quality Care Forum Ltd) v Torbay Council [2014] EWHC 4321 (Admin) – a finding not challenged on appeal in [2017] EWCA Civ 1605 at [3].
104 Charter Holdings Ltd v Commissioner of Inland Revenue [2016] NZCA 499 at [76].
105 See [66]–[75].
[148] The same is so with Vodafone New Zealand New Zealand Ltd v Telecom New Zealand Ltd.106 It was, after all, an appeal on a question of law. Accordingly, it was accordingly concerned only with the second Edwards manifestation.107
[149] Similarly, in Ririnui v Landcorp Farming Ltd, the alleged mistake of fact as to the status of the iwi represented by the applicant was only ever advanced in argument as an error of law in the second Edwards manifestation sense.108 While the Court did acknowledge that whether mistake of fact is a standalone ground for judicial review is an open question that may be significant in some instances, it did not see the need to determine that question in that case, saying that “even if [the error] 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”.109 Accordingly, the case was argued and considered through the lens of the second Edwards manifestation.110
[150] The same is so of other authorities advanced by counsel for the Commission: in Queenstown Lakes District Licensing Agency Inspector v Turnbull Group, Whata J described the issue as being an alleged “mistake in the evaluation of primary facts” and applied Bryson in considering the issue as an error of law through the lens of the second Edwards manifestation.111 In Winton Property Investments Ltd v Minister of Finance, the Court of Appeal, in considering a ministerial grant of consent under the Overseas Investment Act 2005, applied the Bryson test and considered the issue as one of error of law through the second Edwards manifestation lens.112 It did so, acknowledging that the scope of mistake of fact as an independent ground of review is a “head of review [that] is not yet fully settled”.113
106 Vodafone New Zealand Ltd v Telecom New Zealand Ltd [2011] NZSC 138, [2012] 3 NZLR 153.
107 See at [52]–[53] and [65].
108 Ririnui v Landcorp Farming Ltd [2016] NZSC 62, [2016] 1 NZLR 1056 at [3(b)] and [54].
109 At [54].
110 At [96] and [98].
111 Queenstown Lakes District Licensing Agency Inspector v Turnbull Group [2011] NZHC 598, [2011] NZAR 554 at [1], [33] and [41].
112 Winton Property Investments Ltd v Minister of Finance [2023] NZCA 368 at [103].
113 At [103]. The nature of the alleged error there in question was such that the second Edwards
manifestation of error of law approach applied.
[151] As Philip Joseph KC has put it, “Bryson checked the over-zealous application of the mistake of fact ground”.114 It is argued that the court should be reluctant to review determinations of fact because review focuses on legality; because factual challenges are generally more appropriate for statutory appealed processes; and because factual inferences and findings should be reserved to the deciding body. However, as Professor Joseph went on to say, “[s]ome decisions, it is conceded, ought to be set aside where they are made in ‘ignorance of or defiance of an incontrovertible fact’”115 and the error materially affects the outcome of the decision.
[152] That, as I see it, is where the distinction lies. The approach captured by the second Edwards manifestation, which was adopted in Bryson, is there to deal with errors that are made by a decision maker in the evaluative process. An error of law will arise if there is simply no basis in fact upon which a particular decision could have been reached. The cases discussed in [147]–[150] above all involved alleged errors of that kind.
[153] But that is something different from an incontrovertible factual error: an incontrovertible mistake of fact. Addressing an error of that kind is concerned with procedural fairness.
[154] As Hanna Wilberg has said, a mistake of fact ground of review – only applicable when a mistake is both incontrovertible and material – does not encroach on the merits. It does not interfere with fact-finding – because there is no conflicting evidence.116 As Ms Wilberg has put it, a mistake of fact ground clearly does apply to invalidate decisions that are affected by an incontrovertible mistake concerning one single relevant fact – subject to a materiality test.117 For example, in R (Haile) v Immigration Appeal Tribunal, a mistaken understanding represented one of six reasons for an adverse creditability finding against an asylum claimant.118 Even although, as the Court said in that case, the claimant may have few grounds for
114 Philip A Joseph Joseph on Constitutional and Administrative Law (5th ed, Thomson Reuters, Wellington, 2021) at [23.5.5].
115 Citing Isaac v Minister of Consumer Affairs [1990] 2 NZLR 606 (HC) at 637.
116 Hanna Wilberg “Mistake of Fact as a Ground of Review: Distinct and Defensible” in J Tomlinson and A Carter (eds) Facts in Public Law Adjudication (Hart Publishing, Oxford, 2023) 201 at 202.
117 At 218.
118 At 218, citing R (Haile) v Immigration Appeal Tribunal [2021] EWCA Civ 663, [2002] INLR 283.
optimism on reconsideration, the mistake was found to be material and incontrovertible and so the decision was set aside.
[155] As the Court of Appeal put it in Taylor v Chief Executive of the Department of Corrections:119
Although the scope for judicial review for error of fact may not be definitively determined, as the Judge noted, there is support for the proposition that error of fact may constitute a ground of review. This ground of review may alternatively be put on the basis that the decision-maker must take reasonable steps to ascertain the facts and circumstances relevant to the decision.
[156] The point turns on the conceptual differences between the second Edwards manifestation (as adopted in Bryson) ground of review and the mistake of fact ground of review;120 between an untenable evaluative process, on the one hand, and an incontrovertible mistake about a material fact, on the other. The standalone mistake of fact ground can be seen as being concerned with procedural fairness. As Lord Carnwath J said in E, incontrovertible mistakes about material facts give rise to unfairness in the sense that, because of the decision-maker’s mistake and through no fault of the claimant’s own, the claimant had not received “a fair crack of the whip”.121
[157] There are those who say that seeing mistake of fact as a standalone ground takes judicial review beyond its traditional bounds and that courts are ill-equipped to determine materiality in the context of polycentric decision-making. However, an incontrovertible mistake of fact does not involve fact-finding procedure. There simply is no conflicting evidence.
[158] It might equally be said that the ground belongs under the illegality head of review,122 but that does not detract from the ground’s procedural fairness aspect. And in either case, the ground stands in contrast with the second Edwards manifestation
119 Taylor v Chief Executive of the Department of Corrections [2015] NZCA 477, [2015] NZAR 1648 at [94].
120 See, generally, Harry Crawford “The Past, Present and Future of the Mistake of Fact Ground of Review in New Zealand Administrative Law” (LLB (Hons) Dissertation, University of Otago, 2023) at 35–36.
121 E v Secretary of State for the Home Department, above n 100, at [65], a comment made with reference to Fairmount Investments v Secretary of State [1976] 1 WLR 1255 at 1266A, per Lord Russell.
122 See, for example, Edward Lui “‘Fairness’ for mistake of fact: A mistake in fact” (2020) 3 PL 428 at 440.
ground of review, which is a manifestation of an error of law and, therefore, solely an illegality ground.
[159] Moreover, the incontrovertible mistake of fact ground targets a different part of the decision-making process. The second Edwards manifestation targets the evaluation carried out by a decision maker while the mistake of fact ground targets a particular fact that is incontrovertibly wrong and was material. It targets a pillar that is cemented in place in the decision maker’s mind before their factual evaluation begins. It cannot be shoehorned smoothly into the Edwards and Bryson principles.
Application to this case
[160] The numbers produced by the R code, which was, in turn, used by the Commission as one of the pillars in its decision-making process were, incontrovertibly, in error. The Commission says that this does not, however, amount to a mistake of fact. It says that Dr Taylor and the other experts all agree that, in estimating the beta, the Commission is engaged in a forward-looking predictive exercise, which is a matter of judgement, rather than calculation, as the equity beta of the Airports during the regulatory period is not directly observable or ascertainable.
[161] However, there can be little doubt that the erroneous numbers that were produced because of the code formed a central pillar in the reasons for the Commission having arrived at set beta figures and, consequently, the equity beta, WACC and leverage figures. They were material. The Commission did exercise judgement – it did not apply the numbers alone. But, the numbers were part and parcel of the Commission’s determinative process. The R code errors underlying the numbers were an incontrovertible mistake and the mistaken numbers were clearly material to the Commission’s decision. That gives rise to procedural unfairness that should be corrected.
[162]For these reasons, the first cause of action succeeds.
[163] However, as I have discussed earlier in this decision, the procedural unfairness will in this case be corrected through the appeals, whatever their outcome.
The second and third causes of action – taking into account irrelevant considerations, failing to take into account relevant considerations and unreasonableness
[164] While grounds of review will commonly overlap, such that labels are often unhelpful,123 I do see a true mistake of fact – an incontrovertible mistake as to a material fact – as sitting uncomfortably with notions of relevant and irrelevant considerations or of unreasonableness. To put them in the blender would detract from the narrow bounds of the ground of incontrovertible mistake of fact and would see the court being open to fact-finding in pursuit of relevance and reasonableness. That is not what judicial review is about and, accordingly, the second and third causes of action could not in my view succeed as a consequence.
Outcome
[165]For these reasons:
(a)The first cause of action succeeds.
(b)The second and third causes of action do not.
(c)The Court declines to exercise its discretion to grant relief as a result of the bespoke appeal regime in Part 4 of the Commerce Act, which is in play.
(d)The judicial review proceeding is dismissed.
Costs
[166] In the event that costs cannot be resolved by discussion between the parties, they should be the subject of memoranda to be filed as follows:
(a)By the Commission and the Airports by 5 pm on Friday, 21 November 2025.
123 Matthew Smith New Zealand Judicial Review Handbook (2nd ed, Thomson Reuters, Wellington, 2016) at [4.2].
(b)By the Airlines by 5 pm on Friday 12 December 2025.
[167] Memoranda (including any schedules) should not exceed seven pages in length.
[168] The Commission has indicated that it wishes to be heard on costs. If that remains the case for the Commission, or is the case for any other party, they should liaise with the registry to make provision for a two-hour hearing.
Radich J
Solicitors:
Bell Gully, Auckland for Air New Zealand Webb Henderson, Auckland for Qantas
Lindsay Francis & Mangan, Wellington for Board of Airline Representatives of NZ Inc Luke Cunningham Clere, Wellington for Commerce Commission
Russell McVeagh, Auckland for New Zealand Airports Association and Auckland International Airport Ltd
Chapman Tripp, Wellington for Wellington International Airport Ltd and Christchurch International Airport Ltd
3
18
0