IN THE COURT OF APPEAL OF CA284/05

BETWEEN LIMITED Appellant

AND THE COMMERCE COMMISSION Respondent

Hearing: 19 and 20 July 2006

Court: Hammond, O'Regan and Ellen France JJ

Counsel: D J Goddard QC and L Theron for Appellant R A Dobson QC and J S McHerron for Respondent

Judgment: 19 December 2006 at 10 am

JUDGMENT OF THE COURT

A The appeal is dismissed.

B No order for costs.

REASONS OF THE COURT

Hammond and Ellen France JJ [1] O’Regan J (dissenting) [91]

UNISON NETWORKS LIMITED V THE COMMERCE COMMISSION CA CA284/05 [19 December 2006] HAMMOND AND ELLEN FRANCE JJ

(Given by Ellen France J)

Table of Contents Para No Introduction [1] Issues [7] Factual background [8] The pleadings [16] Legislative context [19] The judgment of the High Court [26] What is the purpose of the thresholds? [32] Did the Commission’s thresholds meet the statutory purpose? [48] (i) Initial threshold [48] (ii) Revised threshold [63] Impact of finding initial threshold not lawful [79] Result and costs [89]

Introduction

[1] Under the Commerce Act 1986, the Commerce Commission administers a regulatory regime applicable to large electricity lines businesses such as Unison Networks Limited. Unison challenges the two price-related thresholds which have been set by the Commission under this regime. Breach of these thresholds triggers further inquiry by the Commission and the possibility of control of the services provided by lines businesses. A business which does not breach the thresholds is immune from control.

[2] The first challenge is to the initial price path threshold applicable from 6 June 2003. That threshold required electricity lines businesses, like Unison, to maintain, in nominal terms, their weighted average prices at their 8 August 2001 level.

[3] The second challenge is to the reset threshold applicable from 31 March 2004. That threshold took the form of CPI-X. A lines business will breach the threshold if its average price changes at an annual rate exceeding the change in the consumer price index (CPI), less the annual rate of X% that is set by the Commission for that business.

[4] By the time the matter came before the High Court, Unison had breached both revised thresholds on each of the three assessment dates set by the Commission. On 9 September 2005 the Commission published in the Gazette its notice of intention to declare control over Unison.

[5] Unison says that in setting these two thresholds, the Commission did not turn its mind to the statutory purpose of the thresholds as set out in s 57E of the Commerce Act 1986 and that neither threshold is capable of meeting the statutory purpose. The Commission says it did consider the statutory purpose and maintains that both thresholds fulfil the purpose.

[6] Wild J, in a decision delivered on 28 November 2005, dismissed Unison’s application for judicial review challenging these decisions: HC WN CIV-2004-485-960. Unison appeals.

Issues

[7] The appeal raises the following issues:

(a) What is the purpose of the thresholds?

(b) Did the thresholds meet this purpose?

(c) What are the consequences if either or both of the thresholds do not meet the statutory purpose?

Factual background

[8] Part 4A of the Commerce Act, under which the Commission acted in this case, is a part of a number of reforms of the electricity industry. Those reforms have seen the industry change from public to private ownership and other changes designed to make the retailing of electricity competitive. This background is helpfully discussed in [10] to [16] of Wild J’s judgment.

[9] Under Part 4A, the Commission is required to take various steps to promote the efficient operation of the markets related to electricity distribution and transmission services. The steps include setting thresholds for the declaration of control. The decisions to set the initial and revised thresholds in issue followed an extensive consultation process. Wild J describes that process and the associated development of the Commission’s thinking at [30] to [56] of his judgment. The parties agree that description is accurate and helpful and so we do not repeat it.

[10] For ease of reference, we do need to repeat the Judge’s description of the initial and revised thresholds. The initial thresholds were set by the Commerce Act (Electricity Lines Thresholds) Notice 2003 published in the Gazette on 6 June 2003. There was a price path threshold and a quality threshold. The quality threshold is not in issue.

[11] The Commission published a “Thresholds Decisions” paper of the same date which summarised the thresholds as follows: Threshold Assessment Dates Criteria for Compliance Average price (i.e. base-weighted notional Price Path (First assessment date) annual revenue) at the first assessment date does not exceed the lowest average price at any 6 September 2003 time between 8 August 2001 and the publication date of the Gazette notice (all lines businesses including and Transpower) No increase in average price since the publication date of the Gazette notice. (Second assessment date) No increase in average price (i.e. base- 31 March 2004 weighted notional annual revenue) since the (distribution businesses) first assessment date.

30 June 2004 (Transpower) No material deterioration in recent levels of Quality 31 March 2004 reliability (distribution businesses) and 30 June 2004 (Transpower) meaningful consumer engagement on the “price-quality trade-off” [12] The revised thresholds were published in the Gazette on 31 March 2004 (Commerce Act (Electricity Distribution Thresholds) Notice 2004). Again there were two thresholds – a price path threshold of the form CPI-X and a quality threshold. In its paper of 1 April 2004, the Commission described the X factor as comprising the sum of two factors, a B and a C factor, as follows:

• a B factor, reflecting expected industry-wide improvements in efficiency, determined through total factor productivity (TFP) analysis; and

• a C factor, reflecting the relative performance of groups of distribution businesses, and found from the sum of two component factors:

- a relative productivity component (C1 factor), determined through multilateral total factor productivity (MTFP) analysis; and

- a relative profitability component (C2 factor), determined by comparing ‘residual’ rates of return.

[13] The Commission explained that distribution businesses had been ranked as follows:

on the basis of their average MTFP [average multilateral total factor productivity] index values from 1999 to 2003, and grouped into above- average performers, average performers, and below-average performers, taking account of any clear step points that occur in the rankings. Below- average performers have been assigned a C1 factor of +1%. Businesses performing near the industry average in terms of their relative cost efficiencies have received a C1 factor of zero. Businesses with relatively higher productivity have been assigned a C1 factor of –1%.

… A ten-year time frame has been used to derive the magnitude of the C1 factor, because the Commission acknowledges that rapid improvements in efficiency for infrastructure industries may be difficult to achieve.

[14] The X factor was found by totalling the common B factor of 1%, with a composite C factor that combines the productivity (C1) and profitability (C2) component factors described above. Wild J at [55] set out the X factors as follows:

Table 1 – X Factors for Lines Businesses (Regulatory Period Beginning in 2004)

Lines Business X(=B+C) B C(=C1+C2) C1 C2 Centralines 2% 1% 1% 0% 1% Counties Power 2% 1% 1% 0% 1% Eastland Network 2% 1% 1% 1% 0% Electra 2% 1% 1% 0% 1% MainPower 2% 1% 1% 1% 0% Marlborough Lines 2% 1% 1% 1% 0% 2% 1% 1% 0% 1% The Lines Company 2% 1% 1% 0% 1% WEL Networks 2% 1% 1% 0% 1% 1% 1% 0% 0% 0% Aurora Energy 1% 1% 0% 1% -1% Buller Electricity 1% 1% 0% 1% -1% Electricity Ashburton 1% 1% 0% 1% -1% Horizon Energy 1% 1% 0% -1% 1% Nelson Electricity 1% 1% 0% -1% 1% Network Tasman 1% 1% 0% -1% 1% Orion 1% 1% 0% 0% 0% Transpower 1% 1% N/A N/A N/A 1% 1% 0% 1% -1% Electricity Invercargill 0% 1% -1% -1% 0% Network Waitaki 0% 1% -1% -1% 0% Scanpower 0% 1% -1% -1% 0% The Power Company 0% 1% -1% 0% -1% 0% 1% -1% 0 -1% Unison 0% 1% -1% 0% -1% Vector 1% 1% 0% 1% 0% -1% 1% -2% -1% -1% Otago Net -1% 1% -2% -1% -1% Waipa Networks -1% 1% -2% -1% -1%

[15] Hence, Wild J said:

[56] Assuming inflation (a CPI) of 3%, the effect of this revised price path threshold was:

(a) [Large electricity lines businesses] such as Electra (which might be viewed as the worst performers) with an X factor of 2% could increase prices by 1% i.e. were required to reduce prices in real terms by 2%.

(b) [Large electricity lines businesses] such as Otago Networks (the “star performers”) could increase prices by 4% i.e. an increase in real terms of 1%.

The pleadings

[16] Unison’s claim before the High Court advanced four causes of action. The fourth, relating to the decision to exercise control, is not in issue. [17] The first cause of action alleges that both thresholds are invalid because they are inconsistent with the statutory purpose as set out in Subpart 1. The second cause of action avers that the Commission did not take into account relevant considerations and had regard to irrelevant matters in setting both thresholds. The considerations derive from Unison’s view of the statutory purpose. The third cause of action is one of unreasonableness and is again linked to Unison’s view of the statutory purpose.

[18] The relief sought for all three causes of action is the same, namely:

(a) an order setting aside the initial decision and the revised decision; and

(b) an order requiring the Commission to reconsider the initial decision and the revised decision in accordance with such directions as the Court may give.

Legislative context

[19] We will come back to the detail of the statutory provisions but first we outline briefly the legislative context. Under the Commerce Act, there are two regimes under which goods or services may become the subject of control. The first of these regimes is found in Part 4A. This part was inserted in 2001 by the Commerce Amendment Act (No 2) 2001 with effect from 8 August 2001, and applies to goods and services supplied by electricity lines businesses. The Commission decides whether a business subject to Part 4A is to be controlled. Part 4 applies to other goods and services and it is the Minister of Commerce who makes the decision that these goods and services will be controlled.

[20] Under the Part 4 regime, s 52 provides that goods or services may be controlled if the market is anti-competitive and it is necessary for the goods or services to be controlled in the interests of consumers or suppliers. Pursuant to s 53 the Minister can make a recommendation for the control of goods or services by Order in Council on his or her own initiative or after a report from the Commission. Under s 56, the Commission may report to the Minister on whether or not an Order in Council under s 53 should be made, amended or revoked. [21] The critical section in Part 4A is s 57E. That sets out the purpose of Subpart 1 and how that purpose is to be achieved in the following terms:

The purpose of this subpart is to promote the efficient operation of markets directly related to electricity distribution and transmission services through targeted control for the long-term benefit of consumers by ensuring that suppliers-

(a) are limited in their ability to extract excessive profits; and

(b) face strong incentives to improve efficiency and provide services at a quality that reflects consumer demands; and

(c) share the benefits of efficiency gains with consumers, including through lower prices.

[22] The Part 4A regime has four components. First, there is provision in s 57G for the Commission to “set” thresholds “for the declaration of control” and publish the thresholds. The thresholds are to be set after consultation with participants in the electricity distribution and transmission markets.

[23] Second, the Commission may declare by Gazette notice that goods or services are to be controlled (s 57F). The control decision is made once a business has “breached” a threshold. The Commission has to “assess” businesses against the thresholds and, if there is a breach, the Commission decides whether or not to impose control. That decision can only be made after the process set out in s 57H has been undertaken. That process involves giving an opportunity for interested parties to give their views. The Part 4A regime is unique in that equivalent regimes overseas simply provide for price control of businesses.

[24] The third component of Part 4A regime is the requirement in Subpart 3 for businesses to disclose to the Commission information about their businesses including financial statements, asset values and prices and pricing methodologies (s 57T).

[25] Finally, Subpart 4 requires the Commission to carry out a review of valuation methodologies for line business system fixed assets as soon as practicable. The judgment of the High Court

[26] After setting out the areas of common ground, Wild J discussed the competing arguments and concluded that the Commission had asked itself the right question and the thresholds met the statutory purpose.

[27] Wild J based his conclusion, first, on the interpretation of s 57E. The Judge said s 57E was open, giving the Commission a broad discretion. Wild J relied in that respect on the direction in s 57E to “promote”. The Judge saw that term as consistent with encouraging and persuading over time. Wild J said that Unison’s approach was wrong because it treated the elimination of the ability to extract profits as an end in itself. In this context, Wild J took the view the legislature recognised there was a general link between price and incentives to improve efficiency and quality.

[28] Second, Wild J treated the thresholds as one part of the overall Part 4A regime. In other words, the thresholds alone did not have to achieve all of the desired outcomes.

[29] The third factor influencing the Judge was that there was no evidence that any of the businesses were gouging in an excessive manner. That is, there was no basis for Unison’s concern that the thresholds would miss the very worst of the candidates for control. Associated with this factor, Wild J said it was not unlawful for the Commission to adopt thresholds which would take some time to capture candidates for control.

[30] Fourth, Wild J saw it as relevant that there was some urgency in setting thresholds so the Commission had to do its best within the time constraints.

[31] Finally, Wild J considered there were difficulties in utilising the Po approach at that time favoured by Unison. (The Po approach would involve an adjustment in the price path formula to take account of the initial level of profitability of the business.) What is the purpose of the thresholds?

[32] Unison’s argument is that the thresholds do not achieve the statutory purpose and are therefore unlawful: Padfield v Minister of Agriculture, Fisheries and Food [1968] AC 997. It is necessary then to consider, as a matter of statutory interpretation, what is the purpose of the thresholds. The differences between the parties lie over the extent to which the thresholds have to be a proxy for the attributes ultimately relevant to control.

[33] Unison’s argument is that while the thresholds may be a rough proxy they do nonetheless have to act in that way. In other words, using an analogy with applications for a job, the thresholds must operate to shortlist the prospective applicants for that job.

[34] The Commission’s approach is that nothing in the Act suggests that the thresholds themselves are required to answer the question as to whether a business ought to be controlled. Rather, the thresholds must act as a trigger for the Commission to identify businesses whose performance may warrant further examination and, if necessary, control. The thresholds also have a role to play in providing incentives for businesses to modify their behaviour.

[35] In support of this submission, the Commission relies first on the width of its discretion. Second, the Commission emphasises that the Act envisages two stages, the setting and possible breach of thresholds, and then a decision about control. There is no necessary correlation between breach and control.

[36] The Commission’s discretion in setting the thresholds is undoubtedly broad. It follows that there may well be numerous ways in which the Commission can achieve the statutory purpose. It is true also, as Wild J said, that the Commission through the thresholds does not have to achieve all of the statutory purposes at once. The thresholds may also act as an incentive for businesses to act in a way that, in turn, will encourage the attainment of the objectives in s 57E(a) to (c). That approach is consistent with the overall purpose of the Act which is to “promote” the long-term benefits of consumers. [37] We have, however, concluded that the Commission’s approach and that of the High Court on the purpose of the threshold is wrong because it does not give adequate emphasis to a number of relevant factors.

[38] The first factor we emphasise is that these thresholds have an unusual characteristic in that they are also a “safe harbour”. That is because compliance with them immunises a business from control. It would be odd then if a “threshold for the declaration of control” did not have to attempt, at all, to throw up those at the extremes who are plainly candidates for control. That does suggest some filtering should occur at the threshold stage, albeit it may take time before the filter works to throw up all of the potential candidates for control.

[39] We do not see any great difficulty in a threshold which is overly inclusive. That is because the Commission is right that the second assessment stage is there to enable some weeding out. Having said that, there is some merit in Unison’s argument that the thresholds should not just catch everyone all the time with all of the work being done at the assessment stage. That, too, suggests a filtering is envisaged.

[40] There is also support for the view that some isolation of potential candidates for control is envisaged at the threshold stage by the reference in s 57E to the achievement of the purposes through “targeted control”. That phraseology carries with it an implication that the regime is designed to ensure that control is focused on those who identify themselves as a target for it. It is also an indication of some structure being imposed on the Commission’s discretion which is not apparent in relation to the Minister’s decision under Part 4.

[41] That latter point is supported by an excerpt, relied on by Mr Goddard QC for Unison, from the Parliamentary debates on the Electricity Industry Bill from which Part 4A emerged. Hon Pete Hodgson observed that, unlike proposals from the previous government, the amendments to the Commerce Act then before the House were not based on the notion that all of the lines companies needed to be regulated without evidence: (25 July 2001) 593 NZPD 10411. Instead, Mr Hodgson at 10412, explained that the Government’s approach was to regulate the lines companies that are “out of control”, if there are any, and “not those that are not out of control”. The idea was to “pay attention to the lines companies according to a set of criteria … . There will be a range of criteria by which lines companies are chosen for investigation by the Commerce Commission”.

[42] The second factor which merits emphasis is the use of the term “threshold” itself, particularly in combination with “breach” in s 57H. Section 57H provides as follows:

The Commission must –

(a) assess large electricity lines businesses against the thresholds set under this subpart; and

(b) identify any large electricity business that breaches the thresholds; and

(c) determine whether or not to declare all or any of the goods or services supplied by all or any of the identified large electricity lines businesses to be controlled, taking into account the purpose of this subpart; and

(d) in respect of each identified large electricity lines business, -

(i) make a control declaration; or

(ii) publish the reasons for not making a control declaration in the Gazette, on the Internet and in any other manner (if any) that the Commission considers appropriate.

[43] The dictionary definition of “threshold” does not assist particularly. The usual meaning is the piece of timber below a door which has to be “crossed” in entering a house: The Shorter Oxford English Dictionary on Historical Principles (3 ed). That does suggest, at least, something to be surmounted.

[44] The dictionary definition of “breach” includes “The breaking of any legal or moral bond or obligation; violation, infraction”: The Shorter Oxford English Dictionary on Historical Principles. The use of the term “breach” rather than “cross”, the verb commonly used in association with “threshold”, implies that the thresholds describe an expected standard of behaviour which has not been met. In other words, the business who has breached is one that has identified itself as a candidate for control subject of course to having the opportunity for a dialogue with the Commission about whether control should be imposed.

[45] Finally, there is some assistance in a comparison with the role of thresholds under Part 4. Section 54(1) in Part 4 states that the Minister may require the Commission to advise him or her on the thresholds that would “assist” the Minister “in assessing whether goods or services should be controlled under section 52”. Those thresholds can be expressed in quantitative or qualitative terms (s 54(2)). The thresholds in the Part 4 regime do not have to fulfil the additional role of setting incentives. However, the notion that the thresholds operate to “assist” the decision maker suggests that Part 4A thresholds do need to be precise enough to satisfy the regulator that, at least at a general level, the relevant business will not need to be subject to price control.

[46] For these reasons, we accept Unison’s argument that the statutory purpose of the threshold is to perform a screening or filtering function, which, over time, should capture those who are potential candidates for control. The statute does not require anything more than a rough approximation and, in assessing compliance with the statutory purpose, the incentive effects of the threshold will be relevant.

[47] We turn then to consider whether the thresholds set by the Commission met the statutory purpose.

Did the Commission’s thresholds meet the statutory purpose?

(i) Initial threshold

[48] The initial threshold was essentially a price freeze that operated as at 8 August 2001, the date on which Part 4A became effective. The associated quality threshold similarly froze the status quo telling businesses they complied if quality was no worse than in the previous five years.

[49] In challenging both of the thresholds, Unison says the threshold must fulfil the screening function already discussed as well as constraining the Commission’s discretion whilst creating the specified incentives. In terms of the initial threshold the submission is that it achieves none of these objectives. The submission made by Unison is that there is no correlation at all between a business holding its nominal prices constant and a business acting consistently with s 57E(a) to (c). There is, Unison says, no logical link between price changes above a certain magnitude and the long-term benefit of consumers.

[50] The Commission maintains that the initial threshold was a lawful response given the constraints under which the Commission was operating, particularly the statutory direction in s 57G to set thresholds as soon as practicable and the paucity of information about the comparative state of the businesses in the industry.

[51] There is not a great deal of assistance in the Commission’s papers as to how it saw the initial thresholds fitting the statutory criteria. The Commission did, plainly, appreciate that its role differed from that of the Minister under Part 4. That is apparent from the Commission’s initial discussion paper of 21 March 2002. Similarly, the Commission started from the premise that the choice to use “targeted”, and not universal, control was a deliberate one.

[52] Calum Gunn, a Chief Advisor in the Network Performance Group at the Commission, deposes that the Commission was aware that the level of complexity of the thresholds would affect how much work had to be done at the post-breach stage and potentially the number of businesses that would breach the thresholds. This consideration is apparent in the Commission’s March 2002 paper. There the Commission observed that simple thresholds may catch businesses that were not of concern, “while others who ought to be caught” may not. Further, the Commission noted that more complex thresholds could better target “bad” businesses.

[53] The Commission did turn its mind to the possibility of false negatives i.e. not capturing those who warrant control. The consideration arose in the context of the March 2002 paper where the Commission was reviewing options for thresholds. In particular, in the course of a discussion of an outlier approach in which businesses are ranked on relevant dimensions and a specific proportion are deemed to breach the threshold, the Commission observed that one possible disadvantage of the outlier approach was that “[more] businesses than those caught by the specified fraction might be behaving contrary to the Purpose Statement.”

[54] The options assessed by the Commission included a profit threshold and using a Po adjustment. In rejecting those and other options or variations on them, the Commission did turn its mind to the relevant incentive effects in s 57E(a) to (c). However, what is missing is an explanation of how the Commission sees the threshold it initially adopted as meeting the statutory purpose.

[55] In the executive summary of the Commission’s decisions paper of 6 June 2003, there is a statement that the purpose of the price path threshold is to “provide incentives for lines businesses to reduce their prices in real terms, and therefore to improve efficiency, to be limited in their ability to extract excessive profits, and to share the benefits of efficiency gains with consumers.” There is, however, no substantive discussion of these incentives in the body of the paper. Similarly, the executive summary refers to thresholds as a screening mechanism but there is nothing that reflects this in the body of the decision.

[56] Ultimately, it appears that what the Commission decided to do was to follow a two stage approach. The first stage was a fairly crude initial threshold that was set on the basis that it would be followed by a more sophisticated revised threshold as soon as that could be properly prepared. It has to be said that this approach was favoured by the majority of submitters to the Commission.

[57] It is also clear that the idea behind the initial threshold was that it would operate as an anti-gaming device and avoid the problem that it would be otherwise almost two years before a threshold would be in place with no incentives in the meantime for future action (see, for example, the Commission’s draft decisions paper of 23 December 2002). Businesses could have increased prices in the meantime creating a “buffer”, as the chairperson of the Commission, Paula Rebstock, put it, against the way in which control might later apply.

[58] It is appropriate in assessing the lawfulness of the initial threshold to give the Commission a wide margin of appreciation given the newness of the task and the width of the discretion. The other aspect relevant to that margin of appreciation is that the Commission has indicated that no business will suffer solely because of a breach of the initial threshold if it would not have breached the revised threshold.

[59] There is, as Wild J said, some focus on prices in the legislation. Further, Ms Rebstock makes the point in her affidavit that price is “one of the simplest objective measurements of performance”. In support also of the initial threshold, there is some long-term benefit to the consumer in terms of the matters in s 57E(a) to (c) in implementing an anti-gaming threshold.

[60] Despite the width of the discretion and the valid reasons for operating in this way, we have concluded that the initial threshold did not meet the statutory purpose. An inefficient and high charging business could put itself out of reach of the potential for control simply by maintaining its monopoly pricing and low quality service. There is no element of screening at all in that and indeed a disincentive to act in a manner that will achieve the objectives in s 57E(a) to (c). On the other hand, a business that was charging too little to be able to afford to maintain its network properly and provide a proper quality of service would breach the thresholds by putting its prices up to a level necessary to allow it to reinvest in its network. That business could establish there was no need for control at the assessment stage but these two sides of the coin demonstrate the bluntness of the initial threshold.

[61] The Commission says the alternative to the initial threshold was to do nothing and that doing nothing would not have achieved any of the objectives. In our view, the problems facing the Commission did not make lawful a threshold that did not meet the statutory purpose. Further, as Unison points out, it did not take very long for the revised thresholds to be set so something which did meet the purpose could have been put in place quite quickly.

[62] The Commission also argues that given the links between the initial and the revised thresholds, it is artificial to consider the initial threshold on its own. That linkage may be relevant in considering the impact of a finding of unlawfulness in respect of one of the thresholds, a matter to which we return, but it does not mean the Commission did not have to turn its mind to the statutory purpose and meet that purpose.

(ii) Revised threshold

[63] The revised threshold is challenged by Unison on two principal bases. First, because it is linked to the initial threshold. As the initial threshold is flawed, so too is the revised threshold. Second, Unison says that the revised threshold is flawed because it does not capture or potentially does not capture the worst achievers while capturing too many businesses which should not be caught. A particular concern is the failure to capture those businesses which are under-recovering. For them, Unison says, the threshold provides a strong incentive not to improve efficiency and not to seek to earn normal profits. It follows, Unison says, that the revised threshold does not screen for control; does not provide incentives in the terms envisaged by s 57E; and provides no constraint on the Commission’s discretion.

[64] In developing the first basis of its criticism, Unison points out that the revised threshold incorporates the initial threshold. The price path threshold in the revised threshold involves a formula incorporating R2004. R2004 is defined as “the maximum notional revenue at the reference date which would not have caused the business to breach the [initial price path threshold]” subject to other variables. The Commission accepts that if the initial threshold was set aside, R2004 would be meaningless for the purposes of the 2004 notice. Setting aside the initial threshold would also impact on another part of the formula (R1) dealing with allowable notional revenue for subsequent assessment dates. There is another linkage between the two thresholds in the calculation of allowable price changes during the year in issue.

[65] We do not view this linkage as fatal to the lawfulness of the revised threshold. If the revised threshold meets the statutory purpose, the use of a figure from the initial threshold does not alter the position.

[66] On the second aspect, that is, the failure to capture the worst performers and the under-recovering businesses, Unison emphasises that the objectives in s 57E(a) to (c) are to be achieved through targeted control. The target should be those who warrant control or at least are strong candidates for control. Unison submits that the goals in s 57E(a) to (c) cannot be met if the thresholds are likely to be breached by conduct consistent with those paragraphs or if conduct inconsistent with those paragraphs is unlikely to breach. In this context, Unison relies on the meaning of the various terms used in s 57E. For example, Mr Goddard submits that there is no direct correlation between being limited in the ability to extract “excessive profits” and not increasing prices. Unison also relies on the evidence of Alexander Sundakov, an economist, who makes this point.

[67] Unison complains that the effect of the price path threshold set in the revised threshold is that the average of the worst group will not be “brought to heel” for a number of years and some not at all. Mr Sundakov says that the slow adjustment for companies earning above the normal rate of return may also result in false negatives. Hence, Mr Goddard described the revised threshold as a “slow glide to an uncertain destination in groups”.

[68] A related concern is the failure of the thresholds to take proper account of the situation of the business before the thresholds come into play. Unison argues that the thresholds pay no attention to how efficient a business was at the start and say nothing about the s 57E concerns at the start. That means the threshold cannot achieve the statutory purpose.

[69] Other forms of threshold may better serve the statutory purpose. However, if the complaint is simply a matter of preference for a particular type of threshold, there has been no unlawfulness. We have concluded that Unison’s complaint about the revised threshold boils down to a view there were other, preferable, ways of achieving the statutory purpose. The revised threshold does screen, and does provide incentives in terms of the matters in s 57E(a) to (c) albeit criticism could be levelled at just how successfully the threshold achieves these things.

[70] Unison accepts that the revised threshold is an advance on the initial threshold. The revised threshold places businesses into four groups based on assessments of relative productivity and relative profitability. Each group is assigned a different price path which is designed so that the prices of the businesses over time move (“glide”) to more efficient levels. The revised threshold will identify those businesses that elect to increase prices beyond the constraint which the glide path imposes. In this way, the threshold operates as a filter albeit a fairly rough one.

[71] Unison is critical of the fact that a high proportion (22 of 28 businesses) have breached one of the thresholds at one or more of the assessment dates. It does not follow that no screening is occurring and this factor may simply reflect the fact that the new regime is in its early days.

[72] As we have accepted in the context of determining the statutory purpose, there is merit in Unison’s concern that the worst performers not be missed because those businesses otherwise can evade control. No reliance can be placed on the assessment process to resolve that problem. However, although there are still difficulties in terms of the information base on which the Commission is working, there is enough information before the Commission to support the view that there are no companies in the worst performing category which are being missed by the thresholds. Unison points to the figure in the report of the Commission’s advisers, Meyrick and Associates, which shows Nelson Electricity as having a 15.2% rate of return. However, that is explained by asset revaluations which affected the residual rate of return. The figures from Meyrick’s report show that there are no businesses in the category of both low productivity and high profitability.

[73] We also accept the Commission’s submission that the revised threshold is a means of promoting the statutory objectives because, over time, businesses will be constrained in their ability to extract excessive profits and there will be associated incentives to improve efficiency in order to make the same revenue go further. A business will have to look at ways other than increasing prices in order to maximise profits. The quality threshold means that a business cannot look to reductions in quality to maximise profits. In other words, the threshold does provide incentives to maintain the quality of services while reducing prices in real terms. In this way, consumers share in the efficiency gains, for example, through the lower prices that will result. [74] There is support for the Commission’s “glide” path approach in the Meyrick report. Meyrick and Associates point out that the glide path is a more reasonable approach to addressing the profitability problem. That report also refers to the difficulties of seeking to remove large productivity gaps in a short space of time given the capital intensive nature of the businesses and the long lived nature of the assets. Meyrick and Associates suggest that a decade or two five-year regulatory periods, “is likely to be necessary for businesses performing near the bottom of the range to lift themselves into the middle of the pack”.

[75] In assessing the matter overall, it is also relevant that the thresholds are not set in stone and can be reset where circumstances change. The Commission’s power is to set thresholds, “from time to time” (s 57G(1)). If information comes to light of “out of control” companies which are not being caught, the Commission can reset the thresholds to address that. The Commission flagged the possibility of changing the thresholds at its conference on the thresholds in March 2003. Ms Rebstock pointed to the Commission’s suggestion in its March 2002 paper that a profit threshold may be adopted if the Commission found evidence of “significant excessive returns”. The rejoinder from counsel for PowerCo was that the Commission had not found evidence of excessive returns, or, if it had, had not disclosed that to the industry.

[76] It is also pertinent that, as Ms Rebstock says, the measures in Subpart 1 are complemented by the information disclosure regime under Subpart 3.

[77] Neither of these two matters provide a complete answer but again, are part of the context in which the lawfulness of the Commission’s approach is to be assessed. In a similar vein, it has to be acknowledged that the thresholds reflect the Commission’s analysis of where the appropriate balance lies between light-handed and more heavy-handed forms of regulation. There are trade-offs to be made between the costs imposed on businesses and the regulatory imperatives and these are matters for the Commission.

[78] For these reasons, we consider that Wild J was right to conclude that the revised threshold was lawful. Impact of finding initial threshold not lawful

[79] Further memoranda were filed by the parties after the hearing on the question of relief. In essence, the Commission says that the initial and revised thresholds are linked. If the initial threshold were set aside the Commission says that, nonetheless, the principle of severance would operate to preserve the quality threshold. As a variation on this, the Commission submits the Court should exercise its discretion not to grant relief given the impact of setting aside on the industry and consumers.

[80] Unison’s submission is that if the initial threshold is set aside, the other thresholds (the quality and the revised thresholds) could not stand although that is not how the matter was pleaded. As to the suggestion relief could be declined as a matter of discretion, Unison says that subsequent lawful action cannot save the earlier illegality. Nor should it, given the ongoing effect of the initial thresholds. Finally, Unison argues that the Commission over-states the disruption to the industry arising from the need to start afresh.

[81] There are strong cautions against exercising the discretion not to set aside an unlawful decision: see, for example, the authorities referred to in Fordham Judicial Review Handbook (4 ed 2004) at 24.3.4.

[82] However, it is plain that relief is discretionary. The authors of English Public Law (2004) edited by Professor Feldman observe at 18.52:

Although it is unusual to do so, the court may decide to refuse remedies … possibly allowing invalid public action to stand, because countervailing public considerations justify withholding relief.

[83] Reasons identified as warranting refusal include the implications of granting the remedies for third parties and public administration (Feldman at 18.52). Lord Bingham of Cornhill “Should Public Law Remedies be Discretionary?” [1991] Public Law 64 at 73 notes that the Court has exercised the discretion to refuse declarations where relief would serve no purpose; the applicant has achieved the substantial result sought; where a public body has shown that it is doing all it can genuinely do to meet the statutory duty; and where an error has been substantially cured. [84] The last factor is a reference to R v Secretary of State for Social Services, Ex parte Association of Metropolitan Authorities [1986] 1 WLR 1. The error there was an absence of consultation prior to the making of regulations. The regulations had been consolidated into new regulations and there was no objection to the substance of the new regulations. In that case, relief in the form of quashing the regulations was not seen as appropriate. The fact the regulations had been in force for a period and applications for housing benefits dealt with under the regulations was also relevant.

[85] In R v Monopolies and Mergers Commission Ex parte Argyll Group plc [1986] 1 WLR 763 the need for the financial public to be able to rely on the finality of a decision about a takeover was a reason for declining relief.

[86] We accept that relief here would have some utility because of the ongoing effect of the initial threshold on the revised threshold. Unfortunately, despite the Court having signalled some concern over the remedial issues raised by the case, we have not been provided with a great deal of information about the effect of a decision to set aside the initial threshold whilst upholding the legality of the revised threshold. We have little information, for example, about the impact of relief in that form on third parties. We consider, however, that it is preferable to resolve the matter without a further hearing on the question of relief.

[87] Doing the best we can on the information we have, we have decided that this is an appropriate case to decline to grant relief. Essentially, we accept the Commission’s submissions that to set aside the regime would cause considerable disruption to the industry and to the consumers. The thresholds regime has been operating now for several years. Four rounds of assessments have occurred and several post breach inquiries commenced. All this in a situation where the majority of the industry supported the initial threshold and where the interests of consumers are at issue. The reality is that the revised thresholds are lawful and there is a public interest in their continued operation.

[88] For these reasons, we decline to set aside the initial threshold. Results and costs

[89] The appeal is dismissed.

[90] We make no order for costs given that both parties have had some measure of success. Costs accordingly lie where they fall.

O’REGAN J

[91] I agree with my colleagues that the initial threshold was unlawful but the revised threshold was lawful. However, I have come to a different view on the question of remedy. On the information currently before the Court, I am not satisfied that this is an appropriate case to decline to grant relief.

[92] I accept that the granting of relief to a successful applicant in judicial review proceedings is discretionary. But the effect of our finding in relation to the initial threshold is that the Commission has exercised its power unlawfully. Its statutory power under s 57G of the Commerce Act is to set thresholds. In fact, what it set were not thresholds at all, yet a large electricity lines business (LELB) which fails to comply is exposed to the risk of the imposition of price control, which would obviously have very significant and potentially adverse consequences for that business. Unison currently faces that risk (although we understand that there are now proposals for an administrative settlement to avoid the imposition of price control on Unison). Having succeeded in persuading the Court that the Commission’s action in relation to the initial thresholds was unlawful, in my view Unison should be entitled to have the initial thresholds quashed, unless the Commission can convince the Court that this case is truly exceptional.

[93] The Commission records in its decision on its intention to declare price control on Unison (9 September 2005) at [47] that Unison’s breach of the revised price path reflects the price increases implemented by Unison on 1 March 2004, rather than any subsequent actions by Unison. That means the essential reason that Unison faces exposure to the possible imposition of price control is its breach of the initial thresholds which we have found to be unlawful. Because it faces that exposure, Unison has had to agree to an administrative settlement to avoid the imposition of price control on it.

[94] The starting point for my analysis is Berkeley v Secretary of State for the Environment [2001] 2 AC 603. In that case, Lord Hoffmann described the exercise of discretion not to quash a decision which had been found to be ultra vires as “exceptional” (at 616). Lord Bingham of Cornhill said that the discretion of a Court to do other than quash an order or action where an excessive exercise of power is shown “is very narrow” (at 608). The issue in this case is whether the circumstances are so exceptional as to justify the declining of a remedy.

[95] We asked counsel to address this question in supplementary submissions. Counsel for the Commission, Mr Dobson QC submitted that the quashing of the initial thresholds would bring down the revised thresholds as well. The reason for this is that the revised thresholds build on the initial thresholds and the starting point for the price path in the revised thresholds is specifically linked to the initial thresholds.

[96] The essence of the Commission’s argument is that a decision to quash the initial thresholds would cause the revised thresholds, which are not inherently unlawful, to fail as well. The Commission would then be required to undertake the process for setting thresholds again, which would be costly and inconvenient and would create uncertainty in the relevant electricity markets.

[97] I do not underestimate the significance of those matters. But nor can I ignore the reality that many LELBs have been exposed to regulatory action in circumstances where the event which triggered those processes was the breach of a threshold which we have found to be unlawful. For many, including Unison, this has meant they have had to undertake expensive and intrusive processes as a result of investigations by the Commission.

[98] Mr Dobson accepted that there would need to be “extremely strong reasons to decline any relief”, citing: R v Attorney-General Ex parte Imperial Chemical Industries [1987] 1 CMLR 72 at 109. He argued that the extremely strong reasons in this case were:

(a) There was no useful purpose in setting aside the initial thresholds because they were now spent;

(b) Any unlawfulness in the initial thresholds did not substantively taint the revised thresholds;

(c) Setting aside the whole thresholds regime would cause major disruption to the electricity distribution industry and consumers, given the lengthy process involving considerable consultation with the industry and consumer groups leading up to the promulgation of the initial thresholds and the revised thresholds. In that regard the Commission pointed out that a decision invalidating the thresholds for LELBs would also affect the thresholds for Transpower.

[99] Unison’s counsel, Mr Goddard QC, disputed all of these points. He said the Commission’s submission that the initial thresholds were spent, and did not taint the revised thresholds, could not be reconciled with the Commission’s concession that the revised thresholds were inextricably linked with the initial thresholds. He said if the Commission had made the revised threshold with effect from 1 April 2004 but taking an earlier date as the starting price, it could have been expected that the price path would not have featured a price freeze element for the period up to 1 April 2004, as had been the case under the combined initial thresholds and revised thresholds. He disputed the extent of the disruption that would be caused by setting aside the whole regime, given that the Commission has now undertaken the necessary research and analysis, and the market is now well informed as to the way in which the initial thresholds and revised thresholds operate. He said that if the initial thresholds were unlawful, then Unison was entitled to an order setting aside the action taken by the Commission consequent upon Unison’s breach of the initial thresholds, in particular the Commission’s publication of its intention to declare price control over Unison. [100] Mr Goddard suggested that, if the Court was in doubt as to the appropriate approach, then a further hearing should be held on the question of remedies, following the issue of a judgment ruling on the legality of the thresholds.

[101] In my view, the Court should grant an effective remedy in the present case. I would not classify this case as exceptional in the sense described in Berkeley. I accept that the consequences of setting aside the initial thresholds would be that the revised thresholds would also fail, and that this may cause some disruption in the relevant electricity markets. I do not accept that the Commission has established that this case is so exceptional that relief should be withheld. I would therefore set aside the initial thresholds.

[102] I have considered whether the Court should accept Mr Goddard’s suggestion that we should convene a further hearing on remedies. I agree with the majority that submissions on this aspect of the case were not as fully developed as they might have been. That occurred because the possibility of the initial thresholds being unlawful but the revised thresholds being lawful was addressed only at the hearing and in the submissions we received after the hearing. On my view of the case, the principal point of a further hearing would be to allow the parties to address us on the practical impact of the setting aside of the initial thresholds decision. For the Commission, that would be a chance to bolster the case for the withholding of a remedy. In the light of the majority decision, it is unnecessary for the Commission to do so. I agree therefore that no purpose would be served in re-opening the case for further argument.

Solicitors: Simpson Grierson, Wellington for Appellant Crown Law Office, Wellington for Respondent