NEW ZEALAND GREYHOUND RACING ASSOCIATION v NEW ZEALAND RACING BOARD [2017] NZHC 1771

NEW ZEALAND GREYHOUND RACING ASSOCIATION v NEW ZEALAND RACING BOARD [2017] NZHC 1771 [28 July 2017]IN THE HIGH COURT OF NEW ZEALANDWELLINGTON REGISTRYCIV-2015-485-1041[2017] NZHC 1771BETWEEN NEW ZEALAND GREYHOUNDRACING ASSOCIATIONApplicantAND NEW ZEALAND RACING BOARDRespondentHearing: 18 July 2016Counsel: R Fowler QC and M Smith for ApplicantJ E Hodder QC and B Davies for RespondentJudgment: 28 July 2017JUDGMENT OF WILLIAMS JTable of ContentsIntroduction [1]The legislation [6]The Racing Board [6]The Racing Board's 'levers' [12]History [23]Structural challenges [23]Distribution agreement 2011–2015 [31]The 2016 agreement [45]Terms [45]Negotiations [58]Submissions overview [67]Analysis overview [76]Who allocates the surplus? [78]Personal reflections on the legislative process [78]History and meaning [83]Is the Racing Board exercising a statutory discretion? [88]Are multi-year allocations lawful? [91]Can the codes delegate the power to the Racing Board to extend the life of any methodology and reduce distributions? [96]Was the substantive proposal otherwise inconsistent with the purposeand scheme of the Act or the Racing Board's functions? [101]The purpose of s 16 [103]The purpose and scheme of the Act [105]Did the Racing Board (and the codes) create a legitimate expectation that the 2016 agreement would not be a simple rollover of the 2011 agreement? [112]Did the Racing Board make mistakes of fact, fail to consider relevant factors, or otherwise act in a manner unfair towards GRNZ? [120]Mistakes of fact [120]Relevant considerations [135]Fairness and even-handedness [138]The question of relief [141]Introduction[1] The New Zealand Greyhound Racing Association (GRNZ)1 represents the greyhound racing code. There are two other code organisations: New Zealand Thoroughbred Racing (NZTR), representing the gallops; and Harness Racing New Zealand (HRNZ), representing the trots. These codes are funded by distributions from the New Zealand Racing Board (the Racing Board). The Racing Board controls all racing and sports betting in New Zealand. After making any necessary deductions from its betting revenue, it must divide the surplus between the three codes.[2] This litigation is about the surplus funds distribution model for 2016 (and potentially for 2017 and 2018). The model was agreed by the Racing Board, NZTR and HRNZ in the face of opposition by GRNZ. It is common ground that the model requires GRNZ to subsidise NZTR. To that extent, it is a continuation of a cross- subsidising model put in place in 2011 for the period 2011 to 2015. The 2011 to 2015 model was unanimously supported (that is with GRNZ's agreement).1 This is the abbreviation most commonly used by the parties. However, in some of the documents "NZGRA" is used.[3] GRNZ says this cross-subsidisation was only ever intended to be temporary but as a result of the new agreement, it has become the standard methodology with potentially indefinite application. It says the distribution model is unfair and unlawful and the Racing Board exceeded its statutory mandate by joining NZTR and HRNZ in signing it.[4] The Racing Board replies that it complied with all statutory provisions relevant to its role and, crucially, that the decision to adopt the 2016 distribution model did not belong to the Racing Board at all. It says under the applicable legislation, distribution is a matter for the codes to decide, and this may be done by majority if unanimity cannot be achieved. The Racing Board says that, in addition to having a poor case on the merits, GRNZ has sued the wrong party. It says GRNZ should have sued NZTR and HRNZ.[5] Before I turn to discuss the history of this case and the 2015 agreement containing the distribution model, it is necessary to explain the relevant components of the applicable legislation, the Racing Act 2003. The Act describes the codes and establishes the Racing Board, and it regulates the relationship between these bodies.The legislationThe Racing Board[6] The Racing Act 2003 establishes the New Zealand Racing Board and its"governing body" of seven members comprising code representatives, and ministerial appointees including an independent chair. The Racing Board administers all racing and sports betting in New Zealand and it is the only entity authorised by law to do so. It operates the TAB brand. Its objectives are broadly stated: to promote the racing industry and racing and sports betting, and to maximise the profits of the industry for the long-term benefit of New Zealand racing.2[7] Relevant functions of the Racing Board are in s 9. They include to develop policies for economic development in the industry and the economic wellbeing of2 Racing Act 2003, s 8.people in organisations that depend on it;3 to use its resources (financial and otherwise) in a way the Racing Board thinks will directly or indirectly benefit New Zealand racing4 and (importantly for the purposes of this case) to distribute funds obtained from betting to the racing codes in accordance with ss 16 and 17.5[8] Section 9 also provides that, when it is carrying out its functions, the Racing Board must comply with the principles of natural justice and be socially responsible.6[9] The Racing Board has a "governing body" and all decisions of the Racing Board must be made by or under its authority. Membership of the governing body is carefully controlled by statutory prescriptions. Each of the codes has one appointment to the governing body but the chairperson of each code is ineligible for such appointment unless that chairperson first resigns the chairpersonship before taking appointment.7 In addition, three further members are appointed by the Minister on the advice of a nomination advisory panel.8 The final member is an independent chairperson appointed by the Minister following consultation with the industry.[10] The nomination advisory panel comprises the Minister as chair, the independent chair of the Racing Board, and the three racing code chairpersons. But the panel is directed that it may not recommend any appointment unless:9(a) the Minister has first identified the skills and qualifications required of candidates and sought nominations;(b) industry organisations have been given a reasonable period to make them;103 Section 9(1)(a).4 Section 9(1)(h).5 Section 9(1)(d).6 Section 9(2).7 Section 11(1)(b)-(d) and (2).8 Sections 11(e) and 12, but no code chairperson may be so appointed while in that role.9 Section 12(3).10 Not less than four weeks.(c) the Minister has distributed to industry organisations, the list of those who have accepted nomination;(d) the codes have had an opportunity to consult affiliated organisations about the list and those organisations have had an opportunity to comment;11 and(e) the advisory panel has met to consider the result of such consultations and the skills and qualifications desired. Desired skills and qualifications include knowledge and experience in the industry or expertise in business, marketing or economics.12[11] Thus, the scheme of the Act is that the Racing Board has an overarching mandate to steward the New Zealand racing industry generally. It does this through the development of policies and the application of resources in addition to its other specific statutory functions. This overarching role is reflected in the makeup of its governing body. There is a balance between those who are at the table to represent their respective codes and those, who, following a careful vetting process including industry consultation, are at the table because of the relevant skills or experience they possess. All six of these appointees are led by a chair who is required by law to be independent.13The Racing Board's 'levers'[12] The Racing Board has three key levers in relation to the codes (and through the codes, their constituent clubs). The first is found in s 23 of the Act. Section 23 provides that each year the codes must deliver their statements of intent and business plans to the Racing Board. These documents must relate to the forthcoming three racing years.14 The Racing Board can either approve the statement and plan, or11 Not less than seven days.12 Section 12(4).13 See discussions in this respect in NZ Greyhound Racing Association (Inc) v Minister for Racing[2013] NZHC 245.14 Section 23(1).refuse to do so15 leaving it to the code to resubmit an amended document or documents for approval.16[13] Here, the Racing Board's role is powerful but not strictly directive. TheRacing Board may persuade, encourage and even cajole codes in this process, but it has no statutory power to impose its will directly. Central to this dynamic is the Racing Board's ability to withhold code funding until it has approved that code'sstatement and plan.17 According to John Allen, CEO of the Racing Board, funding has never been withheld for that reason, but negotiations between the Racing Board and various codes has resulted in statements and plans being revised before they were approved.[14] The second lever is that the Racing Board convenes the "Dates Committee" –a special statutory committee of the Racing Board that sets the racing calendar for each year.18 The Dates Committee consults closely with the codes over it, but the Committee itself is the final decision-maker.[15] The domestic racing calendar is very important to the codes. The number of race meetings each code is able to hold and the timing of those meetings is the crucial component in achieving its annual domestic betting turnover. To some extentat least, the codes will compete with one another for the punters' discretionarybetting dollar. The number of meetings each code may hold and when in the week the meeting is set vitally affects that code's prospects. There is, one would expect, a certain amount of jostle in discussions between the codes and the Dates Committee.[16] Again, this form of Board influence over the codes is indirect, but nonetheless significant.[17] The third lever of the Racing Board is that which is the subject of this proceeding and it may be explained as follows. Sections 53 and 57 of the Act direct how the Racing Board must apply its revenue from racing (s 53) and sports betting15 Section 23(3).16 Section 23(5).17 Section 17(3).18 Section 42(1).(s 57). Both sections provide that payments must be applied to: refunds and winnings to punters,19 GST,20 totalisator duty,21 then finally Racing Board expenses.22 In the case of sports betting, payments to sports organisations must also be deducted.23 The surplus available after those deductions must be applied in accordance with ss 15 to 17 of the Act.24[18] Section 15 then allows the Racing Board to set aside reserves in its discretion. The sum remaining after all statutory deductions have been made and reserves set aside is dealt with under s 16. This is the key provision in this case. It provides the process for distribution to codes of the surplus. Its terms are as follows:(1) The Board must, as soon as practicable following the end of a racing year, determine the amount to be distributed among the racing codes for that year from any surpluses referred to in sections 53(2) and 57(2), or any other source whether capital or income.(2) Unless a majority of the racing codes otherwise agrees in writing, the amount referred to in subsection (1) must be not less than the total of the surpluses referred to in sections 53(2) and 57(2) for that racing year less the total amount credited to reserves for that year from those surpluses.(3) Unless a majority of the racing codes otherwise agrees in writing, the amount referred to in subsection (1) must be distributed among the racing codes in the same proportions that the Board considers are the proportions to which the codes contributed to the New Zealand turnover of the Board for that racing year.(4) In subsection (3), New Zealand turnover of the Board means the total gross amount received by the Board from racing betting placed in New Zealand on races run in New Zealand.[19] Subsections (1) and (2) effectively reiterate the rules already discussed for calculating the surplus for distribution and then direct when the calculation is to be made. The amount must be calculated soon after the end of each racing year. It must equal surpluses pursuant to the ss 53 and 57 deduction methodology less the Racing Board's provision for reserves – unless a majority of the racing codes otherwise agrees in writing. I was not advised in evidence whether it has ever been19 Sections 53(1)(a), 57(1)(a).20 Sections 53(1)(b), 57(1)(b).21 As prescribed by the Gaming Duties Act 1971 – ss 53(1)(c), 57(1)(c).22 Sections 53(1)(d), 57(1)(e).23 Sections 57(1)(d), 55.24 Sections 53(2), 57(2).the case since 2003 that the Racing Board's calculation of surplus for distribution has been overridden by a majority of the codes.[20] Subsection (3) is the provision in particular focus in this case. It provides a default distribution formula. That is that the surplus must be distributed according toeach code's contribution to overall domestic betting turnover. It does not matter that some turnover comes from sports betting, from New Zealand punters betting at the TAB on overseas races (as we shall see an increasing proportion of Board revenue), or from other sources such as foreign punters betting into New Zealand races through TAB contracts with overseas bookmakers. In the absence of code agreement, code domestic betting turnover is how a surplus must be distributed. This is interpreted to mean that the Racing Board takes a subset of overall turnover –local race betting – and calculates distribution of the entire surplus by reference to the inter-code proportions in that subset.[21] The codes may override that default by agreeing, two to one if necessary, an alternative distribution method. That is what happened in this case. Under the 2016 agreement, GRNZ receives significantly less than its proportion of domestic turnover per the s 16(3) default.[22] Section 17 provides the mechanical rules by which the surplus is to be handed over. It need not detain us further.HistoryStructural challenges[23] The history that follows is drawn from the affidavits of the key deponents in this case, many of whom have long been involved in the racing industry. Mr Leach provided the primary evidence for Greyhounds. He was the General Manager of GRNZ from 2010 to March 2016. Prior to that, he worked for the Totalisator Agency Board and the Racing Industry Board (the precursor to the Racing Board) for 14 years in senior roles. Mr Allen is the CEO of the Racing Board. He has been in that position since March 2015. Mr Rennell, CEO of HRNZ, and Mr Purcell,CEO of NZTR, also provided affidavits largely in support of the Board's position inthis case. They have held their positions for 19 years and five years respectively. Further affidavits were filed by Mr McArthur, past chairman of GRNZ; Mr Deed, past director of GRNZ; and Graham Cooney, a current member of the RacingBoard's governing body.[24] The domestic racing industry has been under stress for some considerable time. This is due to a downturn in the fortunes primarily of equine racing and betting. Partly this reflects greater competition for the discretionary consumer dollar. More direct effects are from domestic racing having to compete with overseas racing, particularly in Australia. New Zealand punters can now bet on overseas races through the TAB as a result of agreements between the TAB and relevant overseas organisations. This has meant, perhaps ironically, that the Racing Board's bettingrevenue has significantly increased but at some cost to the local product. For example, on an average Saturday in New Zealand, around 180 to 200 thoroughbred races are imported from Australia and more are imported from elsewhere. These are broadcast on the Sky racing channels. At least some local race meetings have had to move to accommodate the timing of imported races. For example, they have moved to weekday slots which attract fewer on-course punters and lower domestic revenue than, say, Saturday or Sunday racing. The domestic industry has put up with this because the hit on the domestic market has been more than made up for by an overall growth in revenue for the industry.[25] The Racing Board has also introduced fixed odds betting into New Zealand as an offering alongside the traditional totalisator betting. This has been necessary to compete with imported fixed odds races. Fixed odds betting is popular with higher end punters but the return is generally lower. It is possible for the TAB to lose money on a fixed odds race, an outcome impossible with totalisator betting. The whole industry is affected by this.[26] The cost structure in the equine codes is also a factor in the stress on domestic racing and betting. New Zealand has a high number of old horse racing tracks with aging infrastructure. There are 64 tracks around the country, most forthoroughbred racing. By contrast, the United Kingdom's far larger market only has 60 tracks. The cost of maintaining this aging capital infrastructure is high. There isalso the fact that horses, being large animals, are inherently expensive to maintain, train and move around the country.[27] GRNZ has bucked this negative trend in domestic turnover. Domestic income for GRNZ has risen sharply since the Act was enacted – from $89 million to $147 million. Export income has also increased.[28] Part of that story has been because greyhound racing is seen by the punters asless of an "on-course experience". It is made for television. There are lots of races in a day, run in quick succession. GRNZ relies on the harness code venues, broadcasting and betting infrastructure to hold its race meetings. The upfront cost for greyhound racing is therefore much lower than that for the equine codes. And the cost in breeding, maintaining, training and racing greyhounds is far lower than that for the equine codes.[29] An additional factor is that there has (by agreement of the Dates Committee) been a significant increase in the number of greyhound meetings held in New Zealand. In 2003, GRNZ held between 70 and 100 greyhound race meetings. There are now over 450 per year. Part of the reason for this is GRNZ has sought and obtained additional race meetings through the Dates Committee for animal welfare reasons. In short, there were too many dogs being bred for the number of races being held. In 2011, the equine codes agreed to an increase to meet this need but on certain conditions to which I will return.[30] The long and short of this overall difference in fortunes between the equine and canine codes is that, as Mr Allen frankly conceded in his evidence, domestic horse racing has not made sufficient revenue to cover its capital costs for some years and cannot now be sustained without cross-subsidy. If the sole object of the Racing Board were to maximise profit, it would have jettisoned its higher cost/lower return horse race meetings (and venues) in order to focus on low cost imported product. The Racing Board has not done that. Instead, the Board and the codes have relied on the (relatively) booming greyhound racing code as the subsidiser. This fact has been reflected in code distribution at least since 2011.Distribution agreement 2011–2015[31] In 2010, a sharp downturn in racing revenue, reflective of wider global economic fortunes at the time, encouraged the codes to come to a multi-year agreement for the distribution of the Racing Board's surplus under s 16(3).[32] Following discussions, the codes agreed unanimously on a four year distribution formula. It is necessary to discuss this agreement in some detail because the GRNZ case is that the 2016 agreement was, effectively, a rollover of this agreement.[33] The purpose section of the agreement explained its underlying thinking as follows:The arrangements outlined in this document provides (sic) the terms of the calendar and Code funding model as a necessary compromise between the respective positions of each Code, and the direction of the Board of the NZ Racing Board, as expressed in meetings and correspondence over recent weeks.The funding model provides an annual agreement, with provision for rollover into subsequent racing seasons subject to annual CRIG appraisal of the review criteria, as outlined in this paper, being met.The intention of each Code and the NZ Racing Board is to support the proposed NZTR revitalisation initiatives, and work within the framework of this funding and calendar model, provided that trading conditions permit and NZTR successfully adheres to its package of reform.[34] The agreement then set out the funding model. Clause 1.1 provided that the"base" funding of $126.6 million would be split on the following basis:$69.5 million to NZTR, $37.8 million to HRNZ and $19.3 million to GRNZ.25Clause 1.2 then provided:If the NZRB Board determines that distributable earnings in any of the four years of the funding model will be in excess of $121.6 m26 (before cost recovery charges), then the excess will be distributed on the following basis.a. Where the increment is between $0-$2m, the additional funds will be distributed in the following proportions: 20% to GRNZ & 80%25 Numbers are rounded.26 This figure must be in error. It seems to mistakenly exclude the $5 million "racing share ofgaming proceeds" that is included in the $126.6 million figure.divided between the 3 Codes in accordance with domestic market share for the relevant season;plusb. For any portion of an increment that exceeds $2m the additional funds will be divided between the 3 codes in accordance with domestic market share for the relevant season.The proviso to this sharing of distribution uplift, is that if any code materially underperforms expectations, then NZ Racing Board may diminishthat Code's share of the upside benefit.c. Should racing's share of NZ Racing Board gaming proceeds exceed $5m, then the additional proceeds will be distributed in accordance with domestic market share for the relevant season.[35] Finally, if funds available for distribution were less than $123.6 million, cl 1.5 provided that the Racing Board could determine an amount of reduction in distribution, which would be applied in proportion to the base funding allocations.[36] NZTR's revitalisation plans were referred to in cl 2 of the agreement. These plans were NZTR's response to its challenges as described above and it is clear that they were an important component of the agreement. Indeed GRNZ said the revitalisation plans were the reason it eventually agreed to the proposal. Clause 2 provided as follows:The funding model has been negotiated to accommodate NZTR revitalisation strategy. This strategy is supported by NZ Racing Board, GRNZ and HRNZ in the expectation that increased revenues will be generated for the entire racing industry.With regard to each Code's provisions of racing product, and theimplementation of NZTR's revitalisation plan, a series of reviews will beundertaken as outlined below. Should CRIG determine that criteria have been met, then the funding model will remain in place for a further period. If criteria are not met, then CRIG reviews the position.The determination of whether the review criteria are met will require agreement of the majority of Codes, with consideration moderated by the NZ Racing Board. Decisions are to be based solely on the review criteria set out below.2727 The reference to CRIG in this clause is the Combined Racing Industry Group. I explain its make-up and role below.[37] Some review criteria applied to all codes. These related to racing product –scheduling and field sizes – but in relation to NZTR's revitalisation plan, cl 2.1.4 provided as follows:The NZTR revitalisation plan incorporates a range of initiatives that set outto revitalise NZ's thoroughbred racing industry.A criteria for periodic review will be NZTR implementation of the race calendar (including the number of race meetings and races scheduled), race programming, handicapping and club benchmark reforms, as outlined on pages 7 and 8 of the NZTR presentation to CRIG (21 June). Adherence to initiatives such as rollout of infrastructure maintenance programme and decisions regarding synthetic track construction will not be within the scope of review, since funding for such initiatives is not yet able to be committed.[38] Attached to the clause was an implementation plan for NZTR's revitalisationinitiatives. It is unnecessary to set out the detail of that but it included the development and rollout of a strategic infrastructure and maintenance programme, successful negotiation of race date changes, handicapping, stakes, race funding reforms, and the possible construction of a synthetic track in Waikato.[39] Review dates were every six months beginning December 2011 and ending May 2014.[40] An important, perhaps the most important, innovation in the revitalisation plan was the introduction of Monday thoroughbred racing. This was to be how NZTR would achieve greater access to the Australian export market. By rearranging its calendar and introducing Monday racing, so that New Zealand races could be broadcast on Sky 1 in Australia – that country's premier racing channel – NZTR sought to tap into a new and potentially very large pool of Australian punters. In the agreement GRNZ accepted that NZTR could take an extra $3 million per annum above what its domestic turnover would justify in order to fund the Monday racing/Sky 1 initiative. In the result, the initiative did not work out. Sky 1 did not broadcast enough of NZTR's Monday races and earnings were low. This combined with the cost of running race meetings in New Zealand seven days a week meant Monday racing could not pay its way and was eventually dropped.[41] Greyhounds also got concessions that it needed. In addition to having a priority call on 20 per cent of the income above expected base revenue, GRNZ also obtained agreement from the Racing Board and the equine codes to the introduction of Friday racing, and then in 2012-13, Tuesday racing, involving in all an extra 85 meetings per year. The need, as I have said, was based not just on enhanced profitability, but also animal welfare concerns. There were nonetheless also concerns in the equine codes that the introduction of new greyhound races would cannibalise thoroughbred and harness racing revenue, rather than tap into new revenue. To meet this concern, GRNZ agreed that the income from these additional meetings would not be factored into the calculation of gross domestic revenue for Greyhounds. In other words, the additional meetings could be held, but revenue would be treated as overall industry revenue rather than specifically tagged to greyhound domestic product for the purposes of s 16(3).[42] GRNZ said the NZTR revitalisation plan was never properly executed and itexpected an independent annual review of the plan's implementation undertakenthrough the Racing Board. It says such review was referred to in a letter of 5 July 2011 from the Racing Board (10 days before the agreement itself). The letterreferred to "an annual review of that implementation [plan's] progress". For GRNZ, Mr Leach said he considered that review would be undertaken by the Racing Board, and it was never done. For its part, NZTR and HRNZ (according to the affidavit of Mr Rennell) rejected that suggestion, and said that any review of the revitalisation plan would be done within the auspices of "CRIG" in accordance with cl 2.1.4.[43] CRIG is the Combined Racing Industry Group. It consists of the chairs and chief executives of the Racing Board, and the three codes. It is a forum for co-ordinating the industry's governing bodies and maintaining good and productive working relationships between them. Its chair is Mr Cooney, an independentmember of the Racing Board's governing body. Mr Cooney deposed that CRIGnever discussed NZTR's revitalisation plans as a stand-alone agenda item althoughNZTR's "initiatives and strategies" were regularly discussed and probably amountedto the same thing. He said neither Mr Leach, nor his predecessor Mr McArthur raised this issue by name in CRIG meetings until 2015 when the codes began to discuss a new agreement. The thrust of his evidence was that revitalisation wasnever a burning issue in CRIG meetings. Mr Leach, for his part, said that was because reviewing NZTR's performance in this respect was the task of the Racing Board not CRIG.[44] Mr Leach said in his evidence that he was lead to believe that a core component of that revitalisation plan would be the closure of uneconomic infrastructure. Without that, he considered, the plan could have no efficacy. Mr Rennell rejected that suggestion. He said the focus in the revitalisation plan was on the shifting of race dates and in particular, Monday racing. He said the closure of venues was never discussed or contemplated. In any event, no venues were closed during that four year period.The 2016 agreementTerms[45] There was agreement during the negotiations that all codes should receive baseline funding in a specific sum below which no code should fall. NZTR and HRNZ (and the Racing Board) say this figure was the same as that contained in the 2011-15 agreement. The effect was to maintain the cross-subsidy. GRNZ says it agreed there was to be a baseline, but there was never an agreement that it should be the same as that under the 2011-15 agreement. GRNZ said the actual baseline figure was still to be negotiated and settled.[46] Principle 2 of the 2015 agreement reflected this issue:Ideally no Code should earn less in Base Funding than a mutually agreed amount in dollar terms by reference to current funding levels. While it is desirable that no Code should receive less funding from NZRB than the Code currently receives, this cannot be guaranteed.[47] The objectives of the agreement are contained in cl 2:The objectives of this agreement are to:(a) grow the total income of the industry by increasing NZRB'sprofitable turnover;(b) provide a framework that incentivises the Codes and the industry to provide quality product at the right time;(c) provide an alternative and improved framework for calculating the distribution of NZRB funding to the three Codes as permitted by section 16(3) of the Act; and(d) provide each Code with some degree of certainty as to the level of funding that the Code can expect to receive from NZRB for a particular Racing Year, and to enable NZRB to provide updates to the Codes throughout the Racing Year.[48] The agreement would remain in place for up to three years: an initial term of one year and then renewal at the unilateral election of the Racing Board for up to two further years:28NZRB can unilaterally elect to extend this Agreement for up to two further periods of 12 months (each further period of 12 months being a Renewal Term) by giving notice to the Codes no later than three months prior to the end of the Initial Term or the current Renewal Term (as the case may be).[49] The possibility of a further multi-year agreement after that was signalled in these terms:The Parties intend that on the expiry of this Agreement, the parties will seek to negotiate in good faith to determine whether or not they wish to enter into a new funding agreement for a period of at least three years.[50] The distribution model was as follows:297.1 Allocation of Base Funding amount: The calculation and allocation of Base Funding is as follows:2014/15 respective Final Code Distribution as indicated in 2014/15 Statement of Intent (and subsequently confirmed by finalised NZRB financial results)Minus: 2014/15 net Export Earnings (calculated by each respective Codes contribution)Equals: Initial year and subsequent years Base Code Funding amount (allocated by respective Code).Section 11 outlines the payment process to respective Codes of these amounts.7.2 Allocation of Additional Funding: Additional Funding will be distributed by NZRB between the three Codes as follows:28 Clause 4.2.29 GBR is gross betting revenue.(a) 50% in proportion to the Code's contribution towards NZRBDomestic GBR; and(b) 50% in proportion to the Code's contribution towards NZRBDomestic and Import GBR.The Additional Funding will be paid to the Codes in accordance with clause 11.2.7.3 GRNZ one-off payment: For the Initial Term only, NZRB will make a one-off payment to GRNZ of $500,000 by 31 July 2016, to support the needs of the greyhound industry.[51] The effect of cl 7.1 was to maintain 2014/15 levels of funding to each code for the initial and subsequent years. According to Appendix 1 the 2014/15 total for all codes was $134.2 million. But as cl 7.2 shows, beyond that base figure, code domestic turnover became a more prominent factor.[52] A working example of the agreed model was also attached in as Appendix 1. It related to the 2016 financial year. It is difficult to read the copy I have, but it appears GRNZ would receive funding in the order of $21 million out of a projected pool of around $134 million for that year.[53] The effect of this agreement was thus that the 2011-15 cross-subsidy from GRNZ to NZTR would remain for another year, and more likely, three years, and income from Tuesday and Friday GRNZ meetings would not be credited to GRNZ. GRNZ says that although greyhound racing produced an average 19 per cent of domestic turnover in 2015, it would receive only 13.5 per cent of the surplus under the agreement – even less by proportion than GRNZ received in 2011-15.[54] This would be slightly ameliorated from GRNZ's perspective by a one-off $500,000 animal welfare payment, though GRNZ asked for $1.75 million in 2015/16 and $1.95 million for each of the two following years.[55] A further aspect of the agreement requires mention. Clauses 3.8 and 9.2 covered the contingency of a profit under-performance. They provided that in theevent of such under-performance, the Racing Board would be empowered in its discretion to reduce code allocations. The terms were as follows:30Principle 7: If NZRB, in its absolute discretion, believes there is a risk that NZRB will not achieve its target profit within a particular Racing Year and therefore the projected Code distributions will not be achieved, NZRB will undertake a clear set of actions to communicate with the Codes.[56] The machinery provision was cl 9.2. It provided:(a) The Codes recognise that the SOI is best guidance and represents a target which NZRB aims to achieve, however there is risk associated with the SOI numbers.(b) NZRB may make a reduction in the distribution to be made to the Codes if NZRB considers that NZRB will not achieve its target net profit for a particular Racing Year. In determining any such reduction, NZRB will take into account (having regard to committedrace meetings and prize money) the Codes' ability to absorb a lesser Distribution within the remaining months of the Racing Year.(c) The Codes will be given three months' notice of any reduction in theamount of Distribution, before such reduction is applied.(d) The parties recognised that open and ongoing dialogue between NZRB and the Codes will be essential to ensure the Codes are aware of potential funding shortfall or performance challenges the NZRB is facing.[57] The target profit figure was to be set by the Racing Board and the only criterion would be an assessment of the codes' ability to absorb the reduction overthe course of the year. These provisions are to be contrasted with cl 1.5 of the 2011 agreement in which a trigger distribution figure is identified and the reduction is agreed to be in equal proportion to the base funding allocation.31Negotiations[58] The first meetings over surplus funds distribution occurred in January 2014 and were initiated by NZTR. But the Racing Board quickly came to take a lead role and by March 2015, was advancing its own draft proposal. From that point on, the Racing Board led all discussions. The 30 July 2015 text, which was the draft finally signed by NZTR and HRNZ, and countersigned by the Racing Board, was proposed30 Clause 3.8.31 See above at [35].by the Racing Board. It followed 18 months of meetings, proposals, takings of position and counter-proposals.[59] The process of engagement was complex. During its course, the Racing Board advanced and sought feedback on three successive drafts. It is unnecessary to traverse their detail. But the evidence is clear that GRNZ engaged fully in the process. It set its position out clearly and repeatedly in letters to the Racing Board, copied to the other codes, and around the table at Board and CRIG meetings.[60] Correspondence between the Racing Board and GRNZ over the months leading up to July 2015 traversed all of the issues the subject of argument before me. GRNZ advised the Racing Board that it had never agreed to a principle that minimum funding levels in the new agreement should equal those in the old agreement. All GRNZ had agreed was that a minimum level of funding should be settled upon. GRNZ complained that the proposal took no account of NZTR'suntouched reserves of more than $7 million. GRNZ suggested that NZTR should be required to access that first before looking to subsidy from GRNZ.[61] GRNZ complained that the cross-subsidy arrangement was only agreed to assist a sibling code make necessary adjustments to its business model but in 2015 had become a permanent fixture without proper justification. It complained thatNZTR's revitalisation programme had failed and there had been no effective reviewof its progress as promised in the 2011 agreement and in correspondence prior to that agreement. It also complained that the proposed model sent the wrong economic signals and discouraged structurally inefficient codes from making necessary adjustments to their business model – especially the closure and consolidation of race tracks. GRNZ complained that the animal welfare payment needed to be of the order of $1.75 million in the first year and even more in the two subsequent years. And it complained finally that the allocation was simply unfair and lacked even- handedness. It was for that reason alone, GRNZ said, unsustainable. GRNZ suggested that the default model in s 16(3) should be applied as it was the fairest and most efficient model.[62] On 30 July following further correspondence, the Racing Board issued a fresh draft agreement.[63] According to Mr Allen:From the Board's perspective, this proposed 2015 agreement was a fully considered new agreement for the racing industry. It followed months of negotiations and consideration by the NZRB and the codes of a wide range of industry concerns. It was in no way a rollover of the 2011 agreement. The fact that the base funding percentage allocations are derived from percentage funding allocations in the final year of the 2011 agreement, is a deliberate, considered term of the new agreement. The reason for importing these percentages is because the Board (and I) were supportive of the concept that no code should get materially less funding than the previous financial year, subject to overall NZRB performance targets being met. It was to provide some stability in funding over time. We saw this as part of an overall suite of terms decided upon as being in the long term best interests of the New Zealand racing industry as a whole.[64] By letter dated 10 August 2015, GRNZ also rejected that proposal. Greyhounds argued:The only positive part of the agreement is that NZRB will make a one-off payment to NZGRA of $500,000. However, on a s 16 distribution, proportionate to the Code's contribution to NZRB domestic turnover, this is a sixth of the amount which NZGRA contributes under the Racing Act'sprovisions. NZGRA has identified that it should receive $3.2 million, over and above the proposed amount for distribution by NZRB.[65] Greyhounds continued:In effect, NZRB is asking NZGRA to agree for up to three years to a funding model that not only is inconsistent with s 16 of the Act, but which substantially disadvantages NZGRA relative to the equine Codes, in that those Codes are entitled to distributions from NZRB significantly greater than the proportions to which the code contribute (sic) to the New Zealand turnover of the NZRB for the racing year.[66] On 20 August, the Racing Board held a meeting and discussed Greyhounds'position. The Racing Board resolved to reject GRNZ's arguments and counter-sign the agreement which by that stage had been signed by NZTR and HRNZ. In his evidence, Mr Allen identified the reasons as follows:(a) NZRB had engaged with the codes equally over an extensive period;(b) This was not a code funding agreement designed by two codes to disadvantage the third. This was an agreement which the NZRBsupports (and to some extent put forward) as being in the best interests of the industry for the 2016 financial year;(c) The (1+1+1) term of the agreement meant that the agreement terms would be reviewed by the NZRB annually having regard to the overarching statutory obligation of the NZRB to promote the best interests of the racing industry; and(d) The agreement was not, as alleged by GRNZ, a simple "roll-over"but was in the best interests of racing for the 2016 financial year. The NZRB is firmly of the view that it is in the best interests of the industry that there is some stability in funding (i.e. similar to the previous year), despite changes to pure domestic turnover, reiterating that domestic turnover is not a helpful matrix for profit distribution. Under a default distribution method, the codes would be incentivised to expand domestic racing (to support the turnover model) at a time when reducing the number of facilities and incentivising their use is the ambition of the NZRB.Submissions overview[67] Although Mr Fowler QC clothed GRNZ's case in orthodox judicial review grounds, the essence of his argument is that the Racing Board acted outside the objects of the Act and its own purposes and functions when it proposed to the codes an effective rollover of the 2011-15 agreement and then countersigned it. That is so, GRNZ says, because the Racing Board and the codes all knew GRNZ's concessions were temporary. They were made only to allow and encourage NZTR to get its house in order. That context meant GRNZ was entitled to expect the Racing Board to propose, and the codes to accept, that the subsidy would end in 2015.[68] There are two related arguments that (effectively) attack the agreement rather than the Racing Board even though the other two codes are not joined as parties to this proceeding. The first argument boils down to the proposition that it is unlawful for the agreement to delegate to the Racing Board the power to extend its terms for two more years or to vary the allocation by reference to Board generated criteria because these are decisions for the codes alone under the Act. Put simply either the codes dictate the outcome or, in the absence of that, the default in s 16(3) applies. The codes have no right to delegate that function.[69] The second argument, closely related to the first, is that allocations must be done in annual tranches. There is no ability to impose multi-year agreementsbecause s 16 requires the Racing Board to determine the amount to be allocated toeach code "for that year" and to do so "as soon as practicable following the end of aracing year".[70] GRNZ mounts further arguments with respect to mistake of fact and failure to consider relevant matters. And GRNZ argues that the failure to have regard to these relevant considerations can also be characterised as procedural unfairness, because of the resulting failures to:(a) treat GRNZ in an even-handed manner; and(b) correctly ascertain the relevant facts and law.[71] I will address the detail of these arguments in my analysis.[72] In reply, the Racing Board argues that the 2016 agreement is between a majority of the codes. The Racing Board says it has no authority or discretion under s 16 to do anything other than comply with the direction of a majority (at least) of the codes, or apply the default in s 16(3) if there is no agreement. That said, the Racing Board submits that it is entitled, in pursuit of its broader objects, purposes and functions, to form a view as to what allocation methodology is in the overall best interests of the industry, and to propound that view to the codes; but it is still not the decision-maker.[73] The Racing Board argues that the 2011-15 agreement has expired and no longer has any effect. The new agreement was carefully negotiated on its own terms and should not be viewed merely as a rollover of the prior arrangement. Nothing was said or done in the context of its negotiation that might have created the substantive expectation that a cross-subsidy would not continue.[74] Furthermore, the Racing Board argues, it was open to the codes, by a majority, to contemplate an extension of the new agreement beyond 2016 at the election of the Racing Board given its overarching industry mandate. And while s 16 contemplates annual allocations, that relates to the actual amount allocated, ratherthan the methodology which was what the equine codes agreed. There is nothing in the Act that prevents the adoption of a multi-year methodology. On the contrary, theAct's focus on long-term viability and profitability suggests multi-year formulae are to be preferred if they can be agreed, for the sake of certainty in the industry.[75] The Racing Board says that the 2011 agreement provides no basis for expectations about distribution methodology after its expiry.Analysis overview[76] GRNZ's case is pitched both at the Racing Board's role in the development and execution of the agreement, and at the agreement itself. It is unfortunate therefore that the other codes were not joined as formal parties in this proceeding since they are its signatories and GRNZ's case is inherently also a challenge to their decision. This gap is ameliorated to some extent by the fact that NZTR and HRNZ provided their own evidence in support of the Racing Board's case, and there can beno doubt that they are very much aligned with the Racing Board in this case and in support of the agreement. But that is not the same as being separately representedthrough counsel and engaging with GRNZ's case from a code perspective via legal submissions.[77] For present purposes, I have come to the view that this litigation raises six core questions. Those questions are raised, though sometimes implicitly, in the multiple and overlapping grounds of judicial review advanced by GRNZ. They are:(a) Who allocates the Racing Board's surplus?(b) Are multi-year allocations lawful?(c) Can the codes delegate the power to the Board to extend the life of any methodology and reduce distribution?(d) Was the substantive proposal otherwise inconsistent with the purpose and scheme of the Act or the Racing Board's functions?(e) Did the Racing Board (and the codes) create a legitimate expectation that the 2016 agreement would not be a simple rollover?(f) In advancing the agreement, did the Racing Board make mistakes of fact, fail to consider relevant factors, or otherwise act in a manner unfair towards GRNZ?Who allocates the surplus?Personal reflections on the legislative process[78] For GRNZ, Mr Leach provided extensive evidence of his engagement with the drafting of the 2003 Act, where, as then interim CEO of the Racing Industry Board (the predecessor of the Racing Board), he was responsible with Sir Geoffrey Palmer for co-ordinating negotiations between the codes on the structure of the Bill. This evidence included extensive personal recollections of the attitudes various players brought to the reform process at the time. On the basis of this material, Mr Leach sought to draw conclusions as to the meaning of the Act, and in particular s 16. He argued this material showed that the Racing Board was given only a limited role in surplus allocations, and it had exceeded this role in its promotion of the 2015 agreement.[79] Although I accept there is no definitive appellate ruling in New Zealand on the subject,32 I am sceptical about the admissibility of evidence of Mr Leach'spersonal and subjective recollections, recorded many years after the relevant event, to assist in the construction of s 16. Reliance on such material brings real dangers to the task of statutory interpretation. In the absence of contemporaneous documentation, there is a risk that any recall will contain the same self-serving flaws that have led the courts to avoid such material in contractual interpretation. And even if admissible, the reflections of a side player in the process will have less value than those of the actual deciders in the Parliamentary process.[80] Mr Leach referred to no material contemporaneous to the reform to confirm those recollections with the exception of official Parliamentary explanatory notes to32 See Vector v Commerce Commission [2012] NZSC 99; [2013] 2 NZLR 445 at [66].the draft provisions. These clearly are admissible and speak for themselves. I will come back to those documents in due course. But even if Mr Leach's evidence is reliable, it speaks only to the subjective perspectives of someone who contributed to, but was nonetheless outside, the Parliamentary process. Reliance on such material may well distract the Court away from the primary task of divining what it was that Parliament intended the words it employed to mean.[81] Clifford J made a similar point with respect to the extensive historical material provided to him in Wellington Airport v Commerce Commission.33 That material related to early departmental reports and Cabinet papers as well as submissions made by private sector parties to officials – material which was, in some ways, a good deal less problematic than Mr Leach's evidence. Clifford J preferred totake the "traditional approach". He said:34A particular explanation of, or opinion on, relevant issues expressed by officials in background materials – for example in advice in preliminary papers or provided to Ministers and the Select Committee from time to time– or even by a Minister (albeit no doubt prepared by officials) in a cabinet paper is, in our view, of no great relevance. What matters is what Parliament considered and decided. Moreover, the complexity and ambiguity of that wider record counts against its utility as an appropriate guide to statutory interpretation. Naturally, the further back one goes in the policy development process the more there will be seen to have been competing ideas, views and proposals.[82] That view is applicable to the material put before me. Furthermore, in reality Mr Leach's recollection take me no further than a plain reading of the words of s 16 when combined with a brief perusal of earlier iterations of the draft and explanatory notes. I set Mr Leach's personal reflections to one side accordingly.History and meaning[83] Section 16(3) currently provides two allocation options: either an allocation based on code domestic turnover or one imposed by a majority (at least) of the racing codes. The original draft of that provision (cl 15 of the Bill at first reading) contained the same default allocation methodology, but provided that it could be overridden only by unanimous agreement of the codes. This draft thus gave each33 Wellington Airport v Commerce Commission [2013] NZHC 3289.34 At [126]. Emphasis added.code a veto and, I assume, made it more likely that in any given year, the default would be resorted to. Perhaps that is why, despite the code veto, the explanatory note to the Bill provided:The Board will be required to distribute betting profits to the 3 racing codes after determining appropriate reserves. Funds will be allocated to the 3 racing codes in proportions that reflect each code's contribution to domesticbetting turnover.[84] Supplementary Order Paper (SOP) 282 replaced cl 15(3) with a proposed clause that shifted all allocation power to the Racing Board. It proposed that codedistribution be "in such proportions and on such basis as the Racing Board from timeto time determines". The explanatory note to the SOP explained that it was"unnecessary, inappropriate and dictatorial to tie down the discretion of the Board",in the manner proposed under cl 15(3), and suggested the answer was to hand the decision to the Racing Board. That proposal was dropped when SOP 16 was introduced four months later.[85] SOP 16 contained the majority override provision currently in place. Itsexplanatory note suggested that the new draft would "provide the Board with moreflexibility in relation to the formula for distributing betting profits to the racingcodes."[86] The evolution to the final form of s 16(3) provides important insights intoParliament's intention. Parliament did not want the Racing Board to decide the allocation, but it fully expected the Racing Board to engage with the codes over the methodology. It was not to be merely a passive player acting on instructions from the codes. The requirement of code unanimity was seen by Parliament as an unhelpful fetter on the ability of the Racing Board to develop, with the codes, an allocation mechanism that suited the exigencies of the particular time. The middle ground adopted in what became s 16(3) was designed to make that easier.[87] The Racing Board was thus always seen to be a central player under s 16(3), but its role is a facilitator or broker. It is not the decider. The decider is either the codes themselves or Parliament by way of the default methodology drafted into s 16(3). I thus agree with the Racing Board that its signature on the 2016 agreementwas superfluous. It was no more than a token of the Board's role as promoter of the draft a role that Parliament specifically envisaged.Is the Racing Board exercising a statutory discretion?[88] Although the Racing Board's role is limited, it is nonetheless a role foundedin the statute. By that I mean it is mandated by the Racing Board's s 8 objectives of promoting the racing industry, facilitating and promoting racing betting, and maximising profits for the long-term benefit of New Zealand racing. Further, Parliament must also have considered that by promoting a particular s 16 allocation methodology, the Board was performing a s 9 function, whether that be seen as making policies for the economic development of racing,35 using its resources to benefit New Zealand racing,36 or just keeping all aspects of racing "under review".37[89] Broadly stated though these discretions are, they are still statutory discretions and anything the Racing Board does pursuant to one or other of them, is reviewable on the usual public law grounds, at least in theory. I see no difficulty in the fact thatthe Racing Board's draft agreement could not have legal effect until signed by the codes. Its role as a promoter to the codes is akin to that of advisors in the public sector. Such advice, if flawed, can be reviewable, on a realistic assessment of its practical impact on a subsequent decision.38[90] In answer to the question I conclude that the codes allocate the surplus but the Racing Board is entitled to promote to the codes its own view of an appropriate methodology. The problem for GRNZ in this proceeding is that once any RacingBoard proposal is adopted by a majority of the codes, reviewing the Racing Board'sadvocacy is likely to be too late. I will come back to this point below.35 Section 9(1)(a).36 Section 9(1)(h).37 Section 9(1)(i).38 Ririnui v Landcorp Farming Ltd [2016] NZSC 62, [2016] 1 NZLR 1056.Are multi-year allocations lawful?[91] GRNZ says that the clause in the 2016 agreement enabling extension of the agreement for two more years is contrary to s 16(3).39 The extension provision is unlawful because the statute requires distributions to be determined year by year.[92] The allocation is by definition and business necessity an annual exercise. The Racing Board must determine the amount to be distributed to the codes at the end of each racing year (subs (1)). That amount cannot be less than the sum of the surpluses for racing and sports betting minus the Racing Board's reserves for that year (subs (2)). And in the absence of a code agreement, the Racing Board'scalculation of proportional contributions to domestic code turnover for that racing year will be the primary element in the calculation of the amount (subs (3)).[93] But the Racing Board is right that the debate in this litigation is not about the amount allocated annually. It is about the methodology. It is not possible to identify any actual amount until all the receipts are in and the final accounting done. That can only be an annual exercise because business accounting is undertaken on that basis. But that statutory prescription does not prevent the codes, with the facilitation of the Racing Board, from developing longer term horizons for allocation methodology.[94] Longer term planning arguably provides better certainty for the codes and the Racing Board. Provided they are not misused to oppress a minority code for reasons unrelated to the long-term viability of the industry, multi-year methodologies can plausibly enhance the long-term benefit of New Zealand racing in accordance with the purposes of the Act and the objectives of the Racing Board. And longer term planning can be said to provide sustainable economic development in the industry, and allow both the Racing Board and the codes to plan the use of their respective resources in a manner that directly or indirectly benefits New Zealand racing in terms of s 9(1)(a) and (h). Multi-year allocation methodologies can therefore be both permitted and rationally connected to the Act's purposes and the Racing39 I deal separately with the issue of the Racing Board's role in extending the agreement in the next question.Board's objectives without falling foul of the requirement to calculate annualamounts in s 16(3).[95] I do acknowledge that the 2016 agreement does purport to set a minimum amount, but that still does not breach the annual approach in s 16(3). It is aminimum. It is not the actual amount to be distributed unless certain contingencies occur. It is still therefore a component of a larger calculation, albeit a precise one. I conclude that multi-year allocation methodologies are lawful. It follows that the 2016 agreement applied a methodology to allocation, but it did not determine the actual amounts referred to in s 16.Can the codes delegate the power to the Racing Board to extend the life of any methodology and reduce distributions?[96] As I have already said, the Racing Board has a broad mandate to pursue courses that in its view will benefit New Zealand racing. Parliament specifically ensured that a majority of members on the Racing Board's governing body would be independent of the codes. This was to help ensure that its focus would be industry- wide rather than code specific, and its planning horizon would be the long-term benefit of the industry. For that purpose, as I have also said, the Racing Board has important levers in relation to the codes. It must approve code statements of intent and business plans and it can refuse to allocate them money in the absence of such approval. Through the Dates Committee it indirectly controls the all-important racing calendar. And as Parliament made clear in the process that lead to the final form of s 16(3) with its industry-wide mandate, the Racing Board is expected to be the facilitator or broker of code agreement on annual allocation methodology.[97] But there are limits. To briefly recap, by rejecting SOP 282, Parliament rejected in clear terms any suggestion that the Racing Board would be the allocator. Parliament preferred to embed in allocation a natural tension between the self- interest of the codes and the wider purview of the Racing Board. The compromise settled upon in the Parliamentary process was that no single code could hold the Racing Board to ransom if a majority of codes supported the Board's alternative to the default methodology. But, subject to that, the codes' substantive agreement toeach allocation was still required. This was, I consider, a delicate balance intentionally struck.[98] It must follow that it is inconsistent with the scheme of the Act for the codes to hand over to the Racing Board the power unilaterally to extend the term of the 2015 agreement for up to a further two years because that would upset this balance. That is an option Parliament rejected. In the absence of an upfront code agreement to a multi-year methodology, the codes' agreement to that course is required at the time of each extension. At that point, the codes and the Racing Board must turn their minds to the purposes of the Act, and reconsider whether they are still served by the methodology adopted earlier. The requirement to reconsider rather than delegate thatresponsibility is a way of ensuring that GRNZ's issues (among others) are not lost. To reiterate, multi-year methodologies are lawful, but the codes must decide whether to adopt them, and they cannot abdicate that decision to the Racing Board.[99] Further, it is also inconsistent with the allocation of responsibilities under s 16(3) to empower the Racing Board in its absolute discretion to reduce distributions if the Racing Board's "target net profit" is not met, in accordance with cls 3.8 and 9.2. The agreement ought at least to contain code agreed criteria for when and how reductions are to be made. This would demonstrate that in exercise of their s 16(3) discretion, the codes have turned their minds to that contingency in agreeing to the methodology. There is always a spectrum in these cases, but in the absence of such criteria, it is my view that the codes have delegated too much of their power under s 16(3) to the Racing Board.40[100] I conclude that the codes may not delegate to the Racing Board either the power to extend a particular methodology or to reduce distribution by reference to its own criteria. To the extent that the Racing Board promoted an agreement that would allocate these roles to the Racing Board, they are inconsistent with the text of the relevant provision, and the wider scheme of the Act seen in light of its purpose and history.4140 For completeness, I note that the 2011 agreement did not contain that flaw. Clause 1.5 identified the line below which reduction would be triggered and detailed how the reduction would be allocated between the codes.41 Commerce Commission v Fonterra Co-operative Group Ltd [2007] NZSC 36, [2007] 3 NZLRWas the substantive proposal otherwise inconsistent with the purpose and scheme of the Act or the Racing Board's functions?[101] GRNZ says that the principle enshrined in the agreement that no code should suffer a reduction in its distributions is inconsistent with the purposes of the Act. This is because such a policy, in the context of the equine codes becoming lessprofitable, and NZTR's failure to implement reforms, cannot be said to promote the long-term viability of racing;42 promote the racing industry;43 maximise profits;44develop the industry;45 comply with natural justice;46 or have regard to the interests of the community.47 Nor, GRNZ argued, does it comply with the function to distribute funds in accordance with ss 16 and 1748 (because it is merely a blanket policy); or the requirement for the Racing Board to operate in a financially responsible manner.49[102] GRNZ placed a great deal of weight on the argument that allocationaccording to each code's domestic turnover is a purpose of the Act, or at least of s 16 and (in effect) had to be applied in the 2016 agreement. GRNZ also took the matter further. It argued the Racing Board's view that due to changes in the racing business, the default allocation formula was no longer "fit for purpose", amounted,effectively, to conspiring to defeat the Act's purpose. It also argued the departure from s 16 was a fettering of its discretion, and a pre-determination of its approach.The purpose of s 16[103] I am unable to agree with the argument in relation to the purpose s 16(3) specifically. It tends to add a certain mana to the default methodology that the plain words of s 16(3) did not intend it to carry. Despite the terms of the explanatory note to the Bill,50 it is merely the default methodology. It applies as a last resort. The codes can allocate on an entirely different basis provided it is consistent with s 3 and767 at [22].42 Section 3(c).43 Section 8(a).44 Section 8(c).45 Section 9(1)(a).46 Section 9(2)(a).47 Section 9(2)(b).48 Section 9(1)(d).49 Section 18.50 Above at [80].the scheme of the Act. It is open to the Racing Board (and the codes) to conclude that they do not want the default to apply. That is the plain structure of s 16(3). It entitles the Racing Board and the codes to reject the default because they do not think it suits their purposes – subject of course, to the alternative being able to attract the agreement of at least two of the three codes. In fact as Mr Rennell deposed and Mr Leach accepted, the default has never been applied in its pure form.[104] GRNZ is really arguing that the Racing Board and the codes cannot stray too far from domestic turnover as a core component of any allocation unless the departure is temporary and/or unanimous. That is to read a great deal into s 16(3) that is simply not there.The purpose and scheme of the Act[105] Nor am I able to agree that the agreement was inconsistent with the Act more broadly.[106] On the facts in this case, the Racing Board (and the codes) had to come to a view of what form of allocation would best facilitate the long-term viability of racing. The Racing Board had to turn its mind to how it could use its resources to benefit New Zealand racing directly or indirectly, and to develop policies that facilitated economic development in the industry. The mandate could hardly be more broadly stated. GRNZ was of the view that adopting an allocation methodology other than code domestic turnover simply perpetuated the current inequities between the canine and equine codes – particularly the thoroughbreds. Any other allocation allowed NZTR to avoid making hard decisions about its uneconomic infrastructure. On the other hand, the Racing Board and the equine codes took the view that stability over time was in the long-term best interests of the industry as a whole. They adopted an approach that kept funding for each code at an historically consistent level for 2015-2016 (and potentially beyond).[107] Whatever one thinks of the merits of that proposition as against those ofGRNZ's take, it is not for this Court to impose its view. As McGrath J found inUnison Networks Ltd v Commerce Commission, while even broadly stated powers have limits and must be exercised to promote rather than thwart the policy andobjects of the Act, it is important to keep clearly in mind, the distinction between the merits of a decision and its legality:51In this area the Courts are concerned with identifying the legal limits of the power rather than assessing the merits of its exercise in any case. They must be careful to avoid crossing the line between those concepts.[108] McGrath J concluded that when an expert body is given expansive powers to achieve statutory objectives that are themselves expansively expressed, Parliament usually intends wide policy considerations to be taken into account and the courts are unlikely to interfere (or even, I venture, be equipped to interfere) unless there is bad faith, error of law or irrationality.52[109] While reasonable minds are very likely to differ on the best allocation methodology for the industry, it simply cannot be said that the 2016 agreement is an irrational departure from the purposes of the Act. Nor is there any evidence that there was bad faith in the sense for example, that the equine codes ignored the Act'spurposes and simply ganged up on the canine code in order to obtain a permanent source of unearned and undeserved subsidy. Such a take on the lengthy negotiation process is, in my view, too simplistic at this stage. Reasonable minds will differ on whether margin or turnover provide the best incentives for the long-term development of the industry. They will also differ on what best promotes long-term viability and profitability and what might be good for the racing community, for example. Equally, it was open to the Racing Board to take account of NZTR'spredicament. As Mr Allen said, NZTR cannot meet its costs without cross-subsidy. That remains stubbornly the position following the failure of the revitalisation plan that GRNZ said never went far enough. I do not see any of these issues as questions for a Judge unless it could be established that no rational Racing Board acting in good faith could possibly have developed such a proposal.[110] In the end, I conclude that the allocation methodology was adopted in good faith, was open to the codes to adopt, and equally open to the Racing Board to promote, subject only to what I have said about delegation of authority to extend the agreement and reduce allocations.51 Unison Networks Ltd v Commerce Commission [2007] NZSC 74; [2008] 1 NZLR 42 at [54].52 At [55].[111] I make these findings with real sympathy for the position in which GRNZ finds itself. It perhaps justifiably fears that what it thought was a temporary problem it agreed to help fix, has or will become a permanent state of affairs to its long-term detriment. GRNZ no doubt feels that it is being taken advantage of, by what will become a permanent coalition of the Racing Board and the equine codes. I agreethat it would be very unfair for GRNZ's subsidy to NZTR to become indefinite, andeffectively structural. NZTR will be required to address these difficult issues as part of any ongoing allocation negotiation. To fail to do so must eventually invite the conclusion that the good of the whole industry has not been a factor in any succeeding agreement. That would expose the methodology to challenge on orthodox public law grounds. But the evidence with respect to the 2016 agreement does not yet establish wilful avoidance of the Act's objects, or the Racing Board'sfunctions.Did the Racing Board (and the codes) create a legitimate expectation that the 2016 agreement would not be a simple rollover of the 2011 agreement?[112] GRNZ submits that a legitimate expectation in relation to the 2016 agreementarises from the background to the 2011 agreement. In return for GRNZ's willingnessin 2011 to allow a departure from the default methodology to its own considerable detriment, there were two concessions to GRNZ: that during the four years of theagreement, NZTR would take steps to "revitalise" its business, and that the retreatfrom the default model was a temporary compromise to give NZTR a chance to put its house in order. It was therefore legitimate for GRNZ to expect there would be no rollover at the expiry of that agreement.[113] Whether substantive legitimate expectation is a doctrine applicable in New Zealand public law remains controversial although the courts have often said the ground is available in theory.53 Mr Fowler argues that the expectation was both substantive and procedural, but the essence of the case is that the arrangement in 2011-15 agreement would be a temporary expedient to allow for NZTR revitalisation.53 See for example B v Waitemata District Health Board [2016] NZCA 184, [2016] 3 NZLR 569 at [55]; Hampton v Canterbury Regional Council [2015] NZCA 509, (2015) 18 ELRNZ 825 at [87]; Green v Racing Integrity Unit Ltd [2014] NZCA 133, [2014] NZAR 623.[114] In Comptroller of Customs v Terminals (NZ) Ltd, the Court of Appeal confirmed that a legitimate expectation can be created by a promise that an authority will act in a particular way or if there is a settled practice or policy in place.54 That,the Court said, should be distinguished from "a mere hope that a course of action will be pursued or a particular outcome gained."55 Any inquiry, the Court found, would involve three steps:(a) establishing the nature of the commitment made by promise or practice. This is a question of fact to be determined by reference to allthe surrounding circumstances. The Court cautioned "a promise orpractice that is ambiguous in nature is unlikely to be treated as givingrise to a legitimate expectation ";56(b) considering whether any reliance on the promise, policy or practice was reasonable in the circumstances;57 and(c) inquiring into the appropriate remedy.58[115] I have no difficulty in agreeing with GRNZ that the 2011 agreement was intended to be a temporary expedient. Further, I accept that the NZTR revitalisation programme was an important ingredient in that agreement and in that respect at least, all parties hoped the revitalisation would succeed and cross-subsidisation would not need to continue beyond the life of that agreement. But it is wrong in my view to interpret the 2016 agreement as if it was merely an unthinking extension of the earlier one. The evidence would have to have established a narrative whereby the Racing Board and equine codes agreed to put off making any substantive decisions or addressing relevant statutory criteria for three more years, and to roll the 2011 agreement over in the meantime. GRNZ would need to have established that the Racing Board in the end made no decision at all and so resorted to maintaining the old decision in the meantime. In my view, the evidence does not support that narrative.54 Comptroller of Customs v Terminals (NZ) Ltd [2012] NZCA 598; [2014] 2 NZLR 137 at [123].55 At [124].56 At [125].57 At [126].58 At [127].[116] The new agreement was, as the evidence shows, arrived at after intensive and lengthy negotiations between the parties over 18 months in which GRNZ set out its position in clear and repeated terms. The fact that the 2015 levels of funding were maintained and the cross-subsidy continued does not mean that the 2011 agreement lasted longer than the Racing Board promised it would. It was independently justified on the premise of certainty and stability.[117] GRNZ could legitimately expect, given the background, that the 2015 negotiation would begin afresh without any particular expectation that the 2011 agreement would simply be rolled over. But the history of the 2014-15 negotiations is that the Racing Board did not seek to rollover the 2011 agreement. Rather, I accept Mr Allen's evidence that each aspect of the 2015 agreement was calibrated toaddress the issues confronting the whole industry as they obtained at that time. It was a fresh agreement capable of justification on its own terms and in its own context. The product is not proof that the process that produced it was unlawful.[118] In reality, GRNZ's argument was (without actually saying so) that it had aright to return to the statutory default methodology and an end to the cross-subsidy. An expectation that an agreement would not be extended was never going to be enough without a follow-up expectation in relation to what would replace it. That iswhere GRNZ's case inevitably fails. The record demonstrates only that all parties accepted that after 2015 all bets were off. GRNZ wanted the cross-subsidy to end. That hope was honest and entirely understandable given the generosity it exhibited in 2011. But it was no more than a hope. It was certainly not based on any unambiguous commitment by the Racing Board in terms of the Comptroller of Customs decision.[119] I conclude that GRNZ was entitled to expect there would be no rollover of the 2011 agreement, but that does not assist because there was no rollover. More to the point, GRNZ had no legitimate expectation about the content of the new model for 2016.Did the Racing Board make mistakes of fact, fail to consider relevant factors, or otherwise act in a manner unfair towards GRNZ?Mistakes of fact[120] GRNZ identifies four mistakes of fact it says the Racing Board made andargues they were each material to the Racing Board's draft of the agreement. They are that GRNZ supported:(a) the continuation of the 2011 cross-subsidy;(b) the nil reduction starting point;(c) the exclusion of Tuesday and Friday racing from funding calculations; and(d) the proposition that $500,000 was sufficient for animal welfare.[121] As to the first alleged mistake of fact, GRNZ says the Racing Board assumed that GRNZ were comfortable with an indefinite continuation of the cross-subsidy. Inmy view, that is not borne out by the evidence. GRNZ refers to Mr Allen's letter of2 June 2015 as evidence of the Racing Board's mistake. But it is not. The relevantextract is as follows:You know that the analysis of history provided in your letter is contested by your colleagues. There is a strong feeling that NZGRA is walking away from commitments it made to not include additional races (which were required for animal welfare purposes and actively supported by colleagues) in your market share calculations for distribution purposes. I understand your view is that those commitments were time-bound, but it is clear that this is not accepted by all those involved.[122] Two points become obvious. The first is the letter does not address the question of indefinite continuation of the cross-subsidy at all. The second is, GRNZ made it clear that in its view, its concessions in relation to extra race meetings were time-bound and could not be extended or reintroduced. But, as Mr Allen says, that view was simply not accepted.[123] The next alleged mistake relates to the so-called nil reduction starting point. GRNZ said it never agreed to funding for the 2015/16 year being no less than that received in the previous year. There is something in what GRNZ contends. The same 2 June 2015 letter by Mr Allen opens in these terms:We are disappointed that NZ Greyhound Racing Association Incorporated (NZGRA) has again rejected a Code Funding Distribution proposal from NZRB. As you know, the proposal reflected principles developed by Codes CEOs and their Chairs through more than 14 months of consultation. These principles were specifically endorsed by all Codes and the CRIG meeting of 22 January 2015. Importantly the principles agreed include one, that no Code should be worse off (in dollar terms) (principle 2); a proposition that NZGRA is now objecting to.[124] That rendition is not quite right. Principle 2 to which Mr Allen refers is as follows:Ideally no Code should earn less than a mutually agreed amount in dollar terms subject to current funding levels. While it is desirable that no Code should earn less than currently, this cannot be guaranteed.[125] This same wording was adopted as principle 2 in the final agreement. The minutes of the CRIG meeting on 22 January 2015 record the meetings' response tothe draft principle 2 in these terms:There were some queries relating to principle 2 with GC [Mr Cooney] expressing concern whether this was providing a guaranteed amount irrespective of NZRB performance. It was clarified that the principle was about fairness in the apportionment. GP [Mr Purnell] noted that this is consistent with the current agreement, and that if the amount distributed is lower, then the principle will ensure the lower amount is apportioned the same way.[126] The first point is that an actual figure was never the subject of firm agreementby any of the codes although the term "ideally" suggests that all parties were of the view that this would be a proper outcome if the Racing Board's income continued to justify it. GRNZ is not recorded in the minutes as dissenting from that proposition. This is not quite the same as a firm agreement, but, realistically, it is not that far from it.[127] The next point is that the 2 June letter from Mr Allen to GRNZ was sent during the course of negotiations. It was open to GRNZ to point out to Mr Allen thathe had overstated the extent of code agreement. In fact GRNZ responded in writing on 15 June. The nil reduction question is not raised.[128] As to the third alleged mistake with respect to the exclusion of Tuesday and Friday racing from funding calculations, this issue is also covered in the 2 June 2015letter from Mr Allen. GRNZ's reply on this issue (in the same 15 June letter) was inthe following terms:[Y]ou have failed to acknowledge that the agreement entered into in 2011 was for a four year period and that NZGRA accepted a lower distribution rate for those four years, and for you to provide NZGRA with copies of the Revitalisation Plan performance reviews as previously requested. Your reference to the Tuesday racing calendar meetings has been confused with this issue. NZGRA agreed during the four year period, since 2012 that Tuesdays would not be included. That was not an agreement in perpetuity.[129] If the Racing Board was unclear about GRNZ's view of this matter on 2 June 2015, the 15 June letter can only have disabused the Board of that. Mr Allen responded to that letter on 30 June in the following terms:It is clear that a variety of arrangements designed to accommodate the specific challenges of individual Codes have been agreed and incorporated in Code funding distribution over the years. It is also clear that there are differences of view about those arrangements and, in my view, as a newcomer to the industry, there is little insight that I can bring to these issues and less benefit in my trying to do so. What the NZRB is committed to doing is continuing to work to grow the industry, put it on a sustainable footing, and ensure the Codes have more money and greater funding certainty as they look to the future.[130] In other words, what is in the past is in the past, and the new agreement is not bound to its assumptions.59[131] As to the fourth alleged mistake of fact, the matter of the animal welfare payment, GRNZ argues that Mr Allen mistakenly believed that the amount given was sufficient for GRNZ's purposes. In his affidavit Mr Allen said:59 I note GRNZ also refers in this context to a Board paper on the Code Funding Distribution Agreement of 15 April 2015 where there is an acknowledgment that the issue of accounting for additional race meetings was acknowledged to be time-bound. It related only to the 2010/11 racing year and could be re-negotiated thereafter. I do not consider this to be as important as GRNZ does. What is more important is that GRNZ repeatedly put its perspective to the Racing Board in this regard during the course of negotiations and particularly in the crucial period May to July immediately prior to execution of this agreement in August 2015.My recollection is that the sum of $500k was the amount that Jim Leach identified as being required to address the shortfall in funding for welfare issues in the 2016 year.[132] GRNZ wrote to Mr Allen on 24 April 2015 outlining its animal welfare needs so its expectations are on record. An overall sum of $1.75 million for the 2015/16 year was identified followed by $1.95 million for each of the two following years. Mr Allen is thus wrong in his recollection. But the GRNZ letter did identify$500,000 as an aspect of its "re-homing" project. Mr Leach wrote:The current gap is approximately 360 dogs per year which would require re- homing to obtain a zero euthanasia objective. We estimate this cost to be between $500k and $600k which is based on the current cost under the GAP [Greyhounds as Pets] programme for approximately the same number of dogs.[133] So Mr Allen's recollection is accurate up to a point but covers only a subset of all that GRNZ felt was needed. The thrust of this passage is that euthanising dogs had to end and unless that sum was forthcoming 360 dogs per year would be killed. The need was obviously compelling.[134] In reality, the $500,000 one-off figure can only have been arrived at by reference to a number of subjective and qualitative factors. These will have included actual canine need but other factors such as affordability and acceptability to the other codes will no doubt have been in the Racing Board's mind when it promotedthe figure. On the other side of the equation, it was plainly felt that a substantive gesture to GRNZ was required given its consistent and now reluctant cross-subsidy of thoroughbreds. This was a payment with a political dimension too. The fact that the payment was one-off only underscores that point.Relevant considerations[135] GRNZ says the Racing Board failed to consider:(a) the text of the 2011 agreement; and(b) NZTR's lack of success in complying with its revitalisation plan over 2011-2015.[136] It also says the Racing Board failed to have proper regard to:(a) the policy of the Act; and(b) the array of matters raised in correspondence by GRNZ to the Racing Board and already traversed in previous grounds.[137] These relevant considerations arguments similarly have no merit. The Racing Board was well aware that NZTR's revitalisation plan had not panned out,primarily because Monday racing had been a failure. As Mr Rennell and Mr Cooney noted in their affidavits the record suggests that NZTR's progress on revitalisationwas the subject of extensive discussion at Board meetings, CRIG meetings, and other combined code meetings, even when it was not a formal agenda item. The additional factors referred to in argument, and in extensive correspondence between GRNZ and the Racing Board were also clearly before the Racing Board andconsidered. GRNZ's argument is really that if all these factors had been taken into account, a different agreement would inevitably have been proposed by the Racing Board – the results, GRNZ argued, speak for themselves. For the reasons I have set out in my analysis of whether the Racing Board acted consistently with the purpose of the Act, I simply do not agree that that is so.Fairness and even-handedness[138] Finally, I do not agree that the Racing Board conducted itself in a way that led to GRNZ being treated unfairly or in a manner that lacked even-handedness. The argument is that GRNZ is being discriminated against but I do not see that in the evidence so far. I can see a rational reason for the limited continuation of the cross- subsidy and the nil reduction principle. Mr Rennell also pointed out why it made sense in terms of code agreement for the extra dog racing meetings to be excluded from the code turnover calculation. And the animal welfare payment was a figure that no doubt took account of a number of relevant factors in addition to actual need including affordability and political acceptability. That said, continuation of some of the allocation elements may over time come to look suspiciously like they have become structural and as though the Racing Board has ignored the needs of GRNZin a manner inconsistent with the scheme of the Act, but that point has not yet been reached.[139] In the meantime, and during the tenure of the new agreement, the industry generally and NZTR in particular, will need to make some hard decisions in relation to its own structural problems. GRNZ is right that these things cannot be put off indefinitely by way of a structural cross-subsidy that must eventually be seen, by definition, to ignore rather than apply the scheme of the Act.[140] I conclude that the Racing Board made no mistakes of fact, considered relevant factors and cannot be said at this point to have behaved unfairly towards GRNZ.The question of relief[141] I have concluded that the Racing Board exceeded its statutory mandate in one respect. That is by promoting an allocation methodology that purported to delegate to itself, and in their entirety, key decisions under s 16(3). These were decisions to extend the life of the methodology beyond 2016, and to vary the methodology by reference to its own criteria.[142] I have concluded that promoting an unlawful agreement is an exercise of theRacing Board's statutory powers set out in s 9 in pursuit of its objectives in s 8. Such promotion is therefore reviewable and was itself unlawful.[143] As I have presaged, the difficulty is that the codes then adopted the RacingBoard's draft and made it their own. The decision to adopt it was theirs, not theBoard's. They are, however, not named in this proceeding. It is therefore not open to me, in their formal absence from this proceeding, to set aside what, in my view, is now their agreement.[144] I acknowledge that chief executives and former chief executives of the equinecodes have given evidence in support of the Racing Board's case, but, on reflection,that cannot cure what in the end must be seen as a fundamental structural flaw in the proceeding. HRNZ and NZTR should have been joined and the separate exercise oftheir respective discretions to agree an alternative allocation under s 16(3) should have been directly challenged. The codes would then have had an opportunity to advance their own legal submissions.[145] In the result, I am prepared to grant GRNZ a declaration that by promoting an agreement that delegated to itself the power to extend the life of the methodology and the power to vary it by reference to its own unstated criteria, the Racing Board acted unlawfully. I am not prepared to go further as invited by GRNZ to set aside the agreement itself or to consequentially direct the Racing Board to reconsider the agreement on any particular terms.[146] I accept that this leaves matters in an unsatisfactory state but I do not consider it is within my power to correct the structural flaw in the proceeding that leads me to this position.[147] There will be a declaration accordingly. In light of the result, GRNZ will be entitled to its costs. If the parties cannot agree on a figure, brief memoranda may be filed.Williams JSolicitors:Treadwells Solicitors, Wellington for ApplicantMinter Ellison Rudd Watts, Wellington for Respondent