Plaintiff v Defendants

1 Each of these proceedings is a group proceeding (a class action) issued under Part 4A of the Supreme Court Act 1986 (the Act) and each concerns "flex commission" arrangements in retail lending to consumers purchasing motor vehicles.

2 Alannah Fox and Bridget Nastasi bring claims against Westpac Banking Corporation and its subsidiary St George Finance Ltd, (the Fox proceeding); Steele Crawford brings claims against Australia and New Zealand Banking Group Limited (ANZ), for loans taken with the Esanda car finance business, and against Macquarie Bank Limited and Macquarie Leasing Pty Ltd (the Crawford proceeding). The plaintiffs allege that under the flex commission arrangements car dealers were authorised by the relevant banks to set their own interest rates for loans that the banks provided to consumers introduced to them by the dealers, by setting the rates charged to consumers higher than the base rate set by the banks. Where a higher interest rate was set, the dealer was paid a commission calculated as a proportion of the difference. The plaintiffs say that these arrangements, which were not disclosed or required to be disclosed to customers, incentivised the dealers to set higher interest rates than they would otherwise have set. The dealers are alleged to have been acting on behalf of the lenders, and engaging in conduct that was, among things, unfair within the meaning of s 180A(1)(b) National Consumer Credit Protection Act 2009 (Cth).[1] The claims are disputed.

3 The plaintiff in each case is seeking a group costs order pursuant to s 33ZDA of the Act, to the effect that the legal costs payable to the solicitors for the plaintiff and group members (Maurice Blackburn) be calculated as a percentage of the amount of any award or settlement that may be recovered in the proceedings, that the percentage be 25% (subject to further order) and that liability for payment of the legal costs be shared among the plaintiff and all group members.

4 Group costs orders (or GCOs) have been foreshadowed in a number of group proceedings commenced in this Court since the introduction of s 33ZDA into Part 4A of the Act, in July 2020,[2] but these are the first applications of their kind to be determined. Victoria is the first Australian jurisdiction to introduce a provision of this kind.

5 With the consent of the parties, the applications in these proceedings were heard jointly. The existing contractual arrangements between Maurice Blackburn and the plaintiffs in each case governing legal costs and funding are materially identical. The plaintiffs relied on common evidence and submissions advanced by jointly retained counsel. The reasons that follow apply to both proceedings, save where I have distinguished between them.

6 In keeping with the direction in the Court's practice note,[3] having determined to seek group cost orders, the plaintiffs each sought them soon after commencing their respective proceedings. Each proceeding is still at an early stage. Pleadings have closed and the ambit of discovery is presently being determined. Orders fixing a date by which group members may opt out of the proceedings have not yet been made.

7 I appointed a contradictor who appeared in both matters,[4] and directed that the parties make submissions concerning the principles that should govern the determination of an application made under s 33ZDA. The plaintiffs relied upon affidavits of Andrew Watson, a principal of Maurice Blackburn, and reports setting out the opinion of Greg Houston, an economist. By orders made on 8 June 2021, some of the material on which the plaintiffs relied was redacted on the basis that it was properly characterised as confidential to the plaintiffs. That material was provided to the contradictor but not to the defendants. The parts of my reasoning that describe the plaintiffs' confidential material are set out in a confidential schedule for each proceeding provided to the plaintiffs and the contradictor but not to the defendants, and not published. The necessity of proceeding this way is apparent from the nature of the material relied upon and the way the applications were put, which appears from the reasons that follow.

8 In summary, I have decided that the plaintiffs have not established a sufficient basis for the exercise of the discretion conferred by s 33ZDA to make a group costs order in each case, at the proposed rate. In particular:

(a) Whether the making of a group costs order (at all or at a particular percentage rate) is appropriate or necessary to ensure that justice is done in the proceeding (the statutory criterion for the exercise of the discretion), will depend upon a broad, evaluative assessment of the relevant facts and the evidence before the court. In making that assessment the interests of group members must be given primacy. In that assessment price, or the costs that group members are likely to pay, is a relevant consideration, but not the only consideration.

(b) The central thrust of each of the plaintiffs' cases was that fixing a group costs order at 25% of the recovered amount would cause the group to be "better off" than under alternative arrangements. Calculating legal fees in that way would deliver a better price and therefore a better financial return to group members. The proposed rate was also appropriate because it would not result in remuneration that would be disproportionate to the risks to be assumed by Maurice Blackburn in funding the proceeding. Further, a group costs order would ensure transparency and certainty of funding arrangements.

(c) The plaintiffs said that the costs to group members and thus the return to them under the proposed group costs order should be assessed against the costs and likely returns that would be achieved should third party funding be obtained for the proceedings. Historically, third party funding has delivered returns to group members in the range of about 45-64%. The proposed GCO would, by comparison, guarantee to group members a 75% return of recovered funds.

(d) In each of these cases the plaintiffs are the beneficiaries of existing funding arrangements in which Maurice Blackburn is acting on a "no win no fee" basis, and has indemnified the plaintiffs against the risk of adverse costs. I have rejected the submission that those arrangements are interim arrangements. Maurice Blackburn subjectively intended them to be so, but on proper construction they are not. Making group costs orders in these cases would effect a fundamental change in the arrangements from one funding model to another, and to make an order effecting such a change, I must be positively satisfied that doing so would be appropriate or necessary to ensure that justice is done in the proceeding. That requires, in this case, an assessment against the existing "no win no fee" funding arrangements.

(e) On the question of outcome, the proposition that group members will be "better off" under a group costs order is founded on predictive modelling that is riven with significant uncertainty. In the Fox proceeding, that modelling does not, on its face, indicate that group members will be better off under the proposed group costs order. In the Crawford proceeding, the modelling does support that contention, but it, too, is founded on significantly uncertain assumptions, and the evidence is otherwise presently unsatisfactory. Ultimately, the present evidence is insufficient to support the exercise of the discretion.

(f) The answer to the statutory question in this case turns on whether the proposed group costs order is more advantageous to group members than the present funding arrangements. That is not a general proxy for the statutory test. Section 33ZDA does not, as a matter of construction, require in every case that a proposed group costs order be demonstrated to likely yield a better outcome than a counterfactual funding arrangement. The facts of these cases may well be anomalous.

(g) In the circumstances I consider it appropriate to adjourn the applications to permit the plaintiffs to consider their respective positions and if so advised, to re-apply at a later time, for group costs orders.

9 The parties were largely in agreement as to the nature of the inquiry required for the purposes of s 33ZDA, but in dispute about its application. Before turning to the parties' positions and the relevant facts, it is necessary to consider the text and context of the provision so as to expose the considerations that should inform its application.

10 In engaging in a purposive construction of the section[5] the statutory text must be considered in its broad context, including its legislative history and extrinsic material. Understanding context has utility if and insofar as it assists in fixing the meaning of the statutory text. Legislative history and extrinsic materials cannot displace the meaning of the statutory text.[6]

(1) On application by the plaintiff in any group proceeding, the Court, if satisfied that it is appropriate or necessary to ensure that justice is done in the proceeding, may make an order -

(a) that the legal costs payable to the law practice representing the plaintiff and group members be calculated as a percentage of the amount of any award or settlement that may be recovered in the proceeding, being the percentage set out in the order; and

(b) that liability for payment of the legal costs must be shared among the plaintiff and all group members.

(2) If a group cost order is made -

(a) the law practice representing the plaintiff and group members is liable to pay any costs payable to the defendant in the proceeding; and

(b) the law practice representing the plaintiff and group members must give any security for the costs of the defendant in the proceeding that the Court may order the plaintiff to give.

(3) The Court, by order during the course of the proceeding, may amend a group costs order, including, but not limited to, amendment of any percentage ordered under subsection (1)(a).

(4) This section has effect despite anything to the contrary in the Legal Profession Uniform Law (Victoria).

(5) In this section -

group costs order means an order made under subsection (1);

legal costs has the same meaning as in the Legal Profession Uniform Law (Victoria).

12 As the statutory text makes clear, s 33ZDA facilitates the funding of group proceedings by introducing what might be described as a statutory common fund[7] of three parts: when a group costs order is made the plaintiff's liability to pay its own legal costs is contingent on recovery of an award or settlement, and the quantum of the costs payable to the legal practice representing the plaintiff and group members is calculated as a percentage of that award or settlement (sub-s 1(a)). An order permitting the calculation of fees in this way must also require that liability for payment of legal costs be shared among all group members (sub-s 1(b)), and where such an order is made the statute shifts the plaintiff's risk of paying adverse costs and any requirement to give security for the defendant's costs to the law practice (ss (2)).

13 In that way, the provision addresses and links these things: first, when a proceeding is funded this way, how legal costs may be calculated (as a percentage of the award or settlement recovered in the proceeding, as specified in the Court's order); second, where a proceeding succeeds, who shares in the liability for the costs of having brought the proceeding (the plaintiff and all group members); third, who bears the financial risks of bringing a group proceeding (the law practice representing the plaintiff and group members).

14 Section 33ZDA is thus a law regulating the calculation of, and liability to pay legal costs. Relevantly, the provision is concerned with regulating the liability of the plaintiff and group members for legal costs vis-à-vis the law practice representing them. The section adopts the definition of legal costs set out in the Legal Profession Uniform Law (Victoria), which (by s 6) defines legal costs to mean:

(a) amounts that a person has been or may be charged by, or is or may become liable to pay to, a law practice for the provision of legal services; or

(b) without limitation, amounts that a person has been or may be charged, or is or may become liable to pay, as a third party payer in respect of the provision of legal services by a law practice to another person

• including disbursements but not including interest.

15 It is clear then that the provision is concerned only with the liability of the plaintiff and group members to pay the law practice representing them. This has two implications. First, the law does not directly or on its face concern the defendants; it is prima facie a law directed to matters on the plaintiff's side of the record as it were.

16 Secondly, although it was not necessary to decide this point on these applications, it seems clear that the provision does not displace the usual position that from the perspective of an unsuccessful defendant costs would follow the event, and a successful litigant will receive her costs in the absence of special circumstances justifying some other order.[8] Nor does s 33ZDA appear to constrain the Court's discretion under s 24 of the Supreme Court Act in relation to the making of adverse costs orders against an unsuccessful defendant. Neither the text nor the legislative context suggests that the provision is intended to displace existing laws or principles relevant to the assessment of or taxation of costs to be paid by an unsuccessful defendant. The plaintiffs' position was that any adverse costs order made against the defendants in these proceedings would not be set by reference to the measure of legal costs fixed by any group costs order, but would be assessed in the ordinary way. There may be aspects of the relationship between the indemnity principle and the effect of group costs orders that will need to be worked out in other cases.

17 Although s 33ZDA permits, in group proceedings, a method of calculating legal fees that is not permitted in other contexts,[9] it does not do so by generally authorising law practices who represent the group in a group proceeding to charge contingency fees. Fees may be calculated in that way if the Court so orders, at the percentage set out in the order, and with the consequence that the other elements of the funding regime contemplated by s 33ZDA (the sharing of liability for costs among the group and the assumption of risk by the law practice) will apply. Although that part of the provision permitting the calculation of fees as a percentage of moneys recovered is novel, the regulation by courts of charges made by lawyers for legal work done, is not.[10]

18 The regulation by courts of fees charged to group members in class actions commonly occurs at the conclusion of a proceeding when the plaintiff seeks court approval of a settlement including the payment of legal costs, under s 33V of the Supreme Court Act. Unlike a "funding equalisation order" which at the end of a proceeding distributes among all group members the burden of a litigation funders' commission agreed privately between the funder and some group members,[11] a group costs order requires group members to share liability for costs calculated at a rate approved by the court on application by the plaintiff.

19 A group cost order permits the plaintiff's legal costs to be calculated as a percentage of the amount of any award or settlement that may be recovered in the proceeding. The words, "award or settlement" are not defined in the Supreme Court Act or in the Uniform Law.[12] The parties submitted, and I agree, that the phrase should be construed as a reference to any principal monetary sum awarded or recovered, not inclusive of any costs award against the defendant in favour of the plaintiff. Parliament cannot be taken to have intended that a law practice in respect of which a group costs order is made would be entitled to be paid both a percentage of the monetary award, and a percentage of any costs recoverable from the defendant.

20 By incorporating the elements it does, s 33ZDA implicitly permits the linking of risk and reward in the calculation of a group costs order. I say that because the section provides that a legal practice the subject of a group costs order will be made liable to pay the defendant's adverse costs and to give any security for the defendant's costs. It follows from the text that the calculation of legal costs in the manner permitted by s 33ZDA may properly take into account not only the value of legal services performed, but the assumption of financial risk by the law practice. The policy reflected in the risk-reward model was discussed in the Victorian Law Reform Commission's Access to Justice - Litigation Funding and Group Proceedings Report[13] in response to which s 33ZDA was introduced,[14] in these terms -

Class actions are an appropriate forum for lawyers to absorb the risks of litigation and be rewarded for this, because the representative plaintiff has a disproportionate exposure to the financial risk of an unfavourable outcome, compared to both the value of their own claim and the exposure of other class members. The risk is a significant disincentive to taking on the role and is only partly mitigated when lawyers act on a 'no win, no fee' basis.[15]

21 The reference in the VLRC Report to disincentives to a person becoming a plaintiff is a particular manifestation of the broader purpose of s 33ZDA, which was described in the second reading of the Bill introducing the provision, as enhancing access to justice in Victoria "by reducing potential barriers to commencing class actions in the Supreme Court".[16] Section 33ZDA sits within Part 4A of the Supreme Court Act which permits and governs the conduct of group proceedings in this Court. The principal object of that part of the Act is enhancing group members' access to justice.[17] Section 33ZDA then, builds on the existing provisions of Part 4A of the Act by conferring on the Court the power, in an appropriate case, to facilitate access to justice for group members by making a GCO, subject to the statutory pre-conditions to the exercise of the discretion being met.

22 The provision does not, explicitly or implicitly, require that a group costs order be made only at a particular juncture after the commencement of a group proceeding. There are clear contextual indicators however, that the legislature intended that group costs orders may (but not must) be put in place early in the life of a proceeding. Where such an order is made, s 33ZDA regulates aspects of funding that, as a matter of practical reality, ordinarily require consideration early in a proceeding - who will give security for a defendant's costs, on what basis the law practice which represents the group is to be paid, and who can be made liable for any adverse costs. Sub-section 33ZDA(3) provides that a group costs order may be amended during the course of the proceeding, including by amending the percentage fixed under sub-s (1). The section thus contemplates that circumstances that may inform the terms of and appropriate rate for a group costs order, may change during the life of a proceeding. Whether or not a GCO ought be made at an early stage in any particular case will depend on the circumstances of the case and the evidence supporting the application.

23 The power to amend a group costs order in sub-s (3) is generally expressed. Although the time at which a Court might amend an order and the basis for doing so are not constrained by the statute, an obvious use of the provision would be the adjustment of the percentage specified in an order, at the time of the settlement of a proceeding, having regard to the recovery achieved by the plaintiff (among other relevant considerations).

24 On application by the plaintiff in a group proceeding the Court may make a group costs order if satisfied that it is appropriate or necessary to ensure that justice is done in the proceeding. The provision does not set out any other criteria for the exercise of the discretion. Section 33ZDA is an example of an open-textured legislative provision that "leave[s] courts with a large measure of significantly unguided discretion in making orders considered to be appropriate to do justice in all the circumstances of a given case",[18] as Nettle J said, in BMW v Brewster, of the similarly worded s 33ZF of the cognate Part IVA of the Federal Court of Australia Act 1976 (Cth) (Federal Court Act).[19] A group costs order is novel in this jurisdiction as a means of regulating litigation funding, but the conferral on a Court of a discretion of this kind, where the court is to determine what is "appropriate or necessary", not by an idiosyncratic assessment but in order to achieve what is "just" in a proceeding, is not.[20]

25 As Nettle J went on to say in Brewster, generally speaking, provisions of that kind "may be seen to reflect a legislative intention to confer on courts the widest possible power to do what is appropriate to achieve justice in the circumstances".[21] Section 33ZDA is a provision, the interpretation of which attracts the long-established principle stated in Owners of the Ship 'Shin Kobe Maru', that it is "quite inappropriate to read provisions conferring jurisdiction or granting powers to a court by making implications or imposing limitations which are not found in the express words".[22]

26 Because of the way that some of the parties' submissions were framed, it is necessary to say something about the different functions of s 33ZDA and s 33ZF of the Act.

27 Unlike s 33ZF of the Act, whose cognates in Federal and New South Wales legislation were recently considered by the High Court in BMW v Brewster,[23] s 33ZDA is not a supplementary or gap-filling power. Section 33ZF of the Act provides that, "in any proceeding ... conducted under this Part the Court may, of its own motion or on an application by a party, make any order the Court thinks appropriate or necessary to ensure that justice is done in the proceeding". Although the criterion for the exercise of power conferred by s 33ZDA also employs the words "appropriate or necessary to ensure that justice is done in the proceeding", those words do not condition a general power to make "any order" in the proceeding; they condition the exercise of the power to make a very particular kind of order (a GCO). Contrary to Westpac's submission, s 33ZDA does not "reflect the test" employed in s 33ZF. It employs some of the same language, but in a different context. The criterion for the exercise of the power must be read together with the grant of power itself. Some guidance might be gleaned from the reasoning in cases in which s 33ZF and its cognates have been considered by courts, but that reasoning cannot be applied as though s 33ZDA and s 33ZF were provisions of the same or a substantially similar kind. They are plainly distinct and serve very different purposes.

28 Given its nature, the statutory criterion for making a group costs order must be construed in the context of the particular facts on which the applications are founded. However, some general observations may be made.

29 The parties in both the Fox and Crawford proceedings directed their submissions to the question whether the making of a group costs order would be appropriate to ensure that justice is done in the proceeding, implicitly proceeding on the basis that the words necessary and appropriate have separate work to do. That reading sits naturally with the statutory language. As will become clear, the parties' positions did not, and the result in each case does not turn on any differentiation between those expressions.[24]

30 Each part of the statutory criterion permits of a range of meanings and is capable of satisfaction in myriad ways. The dictionary meaning of "appropriate" is suitable, fitting or proper in the circumstances. To "ensure" is to make certain of something.[25] In the course of considering s 33ZF of the Federal Court Act the Full Court of the Federal Court said in Money Max[26] that the word "necessary" and the phrase, "to ensure" do not bear fixed meaning but depend upon and must be construed in their contexts.[27] The Court there cited the United States Supreme Court which, in McCulloch v State of Maryland,[28] said of the word necessary, "we find that it frequently imports no more than that one thing is convenient, or useful or essential to another".[29] The Full Court in Money Max went on to conclude that in s 33ZF there was less of a difference between the expressions, "necessary to ensure justice" and "appropriate to ensure justice" than might initially appear. It further considered that in s 33ZF, the word "necessary" identifies a connection between the proposed order an identified purpose as to which the Court must be satisfied before making an order, and that the expression, necessary to ensure that justice is done required that the proposed order be "reasonably adapted to the purpose of seeking or obtaining justice in the proceeding".[30]

31 In the present context, it may be said that before making a group costs order the Court must be satisfied that doing so would be a suitable, fitting or proper way to ensure that justice is done in the proceeding. As the contradictor submitted in this case, such an order would be reasonably adapted to the purpose of seeking or obtaining justice in the proceeding.

32 As to ensuring that the justice is done in the proceeding, it must be recalled that the criterion governs the making of an order concerning the funding of the proceeding in question, specifically the calculation of legal costs payable by the group to the law practice representing it on the conditions set by the statute. A court making a group costs order will, then, have regard to what is appropriate or necessary to ensure justice in the proceeding in respect of the fees payable to the law practice representing the group.

33 The framing of the criterion for the exercise of the power in s 33ZDA by reference to whether a group costs order will be "appropriate or necessary to ensure that justice is done in the proceeding" indicates that what is required in determining whether to make a GCO is a broad, evaluative assessment. In that assessment, the question whether to make an order, and the question what is the rate that ought be set by the order, will be intertwined.

34 Because of its subject matter and its place within Part 4A of the Act, s 33ZDA requires that in exercising the power to grant a group costs order the Court must be astute to protect the interests of group members.[31] Having regard to the Court's role in ensuring that group members are not prejudiced by the conduct of litigation on their behalf,[32] the effect on group members of a proposed order must be a primary consideration informing that evaluation.

35 In Brewster, Kiefel CJ, Keane and Bell JJ said that s 33ZF "assumes that an issue has arisen in a pending proceeding between the parties to it, and that the proceeding will be advanced towards a just and effective resolution by the order sought from the court".[33] That analysis reflects the fact that s 33ZF, unlike s 33ZDA, permits the court to make "any order the court thinks appropriate or necessary to ensure that justice is done in the proceeding". Section 33ZDA, by contrast, regulates the funding and legal costs of the plaintiff and group members. It is not directly concerned with resolving issues necessary to advance the resolution of the inter partes dispute the subject of the plaintiff's claim.

36 Kiefel CJ, Keane and Bell JJ went on to cite Wigney J in Blairgowrie[34] with approval, to the effect that the requirement in s 33ZF that justice be done in the proceeding suggests that the proposed order must be fair and equitable, which will ordinarily involve a consideration of the position of all parties.[35] An order that is appropriate to ensure that justice is done in a proceeding will require fairness and equity, and an order that unjustly affects a party to a proceeding could not be described in those terms.

37 However, the requirement that the position of all parties be considered will not apply as generally or as directly in the case of s 33ZDA as it does for the purposes of s 33ZF, because the subject matter of s 33ZDA concerns the plaintiff's liability in respect of legal costs. I accept the contradictor's submissions that evidence of the nature and extent of the risk that the legal practice would be prepared to accept might inform the fixing of an appropriate percentage, but the interests of the law practice are not directly in issue. Whether or not a defendant has a legitimate interest in an application for a group costs order (or in aspects of that application) is a question to be decided on a case by case basis. Questions as to the capacity of the law practice to give security may well legitimately concern a defendant. Because these applications were the first of their kind, I indicated to the parties at an early stage that I would receive submissions from all parties on matters relevant to their interests, including as to the principles that should govern the interpretation of s 33ZDA.

38 Westpac submitted that even if the statutory pre-condition to the exercise of the power is enlivened here, the power should be exercised "cautiously" because a group costs order is an exception to an otherwise general prohibition on the charging of contingency fees. I reject that submission. Statutory construction must begin and end with the text.[36] There is no warrant in the statutory text for requiring presumptive caution in exercising the discretion to make a group costs order, whatever a "cautious" approach might be taken to mean. The criterion for making an order is specified and is to be applied according to its terms. As discussed earlier, s 33ZDA is a provision of the kind that reflects a legislative intention to confer on courts the widest possible power to do what is appropriate to achieve justice in the circumstances, and it is inappropriate to read provisions conferring jurisdiction or granting powers to a court by making implications or imposing limitations which are not found in the express words.[37] Reading down an empowering provision because what it expressly empowers was not formerly permitted is not a proper approach to the construction of the statute.

Part C - Summary of the plaintiffs' case and parties' positions

39 The plaintiffs in each case submitted that the making of a group costs order fixing 25% as the percentage at which the costs payable to Maurice Blackburn would be calculated if their actions were successful (subject to further order), satisfied the statutory criteria and would "better secure the plaintiff and group members to achieve justice in the proceeding" for these reasons:

(a) A GCO at that rate will provide a better return to the plaintiff and group members than alternative funding arrangements for the proceedings;

(b) A GCO would fairly distribute the burden of legal costs incurred in pursuit of common questions among all group members;

(c) A GCO would make funding arrangements for the proceeding certain and transparent;

(d) A GCO would not expose the plaintiff to significant and disproportionate exposure to financial risks as a result of assuming that role; and

(e) A GCO at that rate is proportional to the risks to be incurred by Maurice Blackburn in funding the proceeding. This would serve as a cross-check of the reasonableness of the proposed percentage and therefore of its appropriateness.

40 If circumstances later materialised the demonstrated the rate to have been disproportionate or unreasonable, the rate could be adjusted under sub-s 33ZDA(3). The remaining factors said to support the application are not particular to the circumstances of these proceedings and describe characteristics inherent in a group costs order.

41 Put that way, the argument required the plaintiffs to identify the alternative funding arrangements against which the proposed group costs order might be compared. Their case was that the appropriate comparator was third party litigation funding in which a funder would charge a commission in addition to reimbursement of legal costs. Third party funders have historically charged commission in the range of about 21-29%, with an average of 24% (on top of reimbursement for legal costs). The combined legal and litigation funding fees in historical proceedings was in the range 36% to 55%, with a resulting recovery to group members, of 45-64%.[38] The proposed GCO would, by comparison, guarantee to group members a 75% return of recovered funds.

42 The plaintiffs and Maurice Blackburn presently have in place funding arrangements for the proceedings. Maurice Blackburn is retained by each plaintiff under a "no-win no-fee" (NWNF) or conditional costs agreement in which the plaintiffs will pay costs only if the action is successful, in which case an uplift fee calculated at 25% of the costs will be charged. Those agreements are considered below. They anticipate that the plaintiff may apply for a group costs order and that if an order were made, it would supersede the relevant terms of the costs agreements. Mr Watson's evidence was that the present arrangements were intended to operate on an interim basis, until GCOs were made, and if the applications were unsuccessful, Maurice Blackburn would seek to obtain third party funding, failing which it would continue to fund the proceedings on a NWNF basis. The plaintiffs' case was that NWNF funding was not the proper comparator against which the proposed group costs orders were to be assessed, however ultimately it did not matter which comparator was chosen because in either case the group would be better off.

43 The contradictor expressed the issue this way: the Court's central inquiry in this case ought be a comparison between the features of the proposed group costs order and the features of any realistic alternative funding arrangement for the advancement of the plaintiffs' and group members' claims, including in particular the comparative likely returns to the plaintiffs and group members. The question is whether group members' interests will be better advanced by the proposed GCO or the funding arrangement that will prevail without the GCO.

44 The contradictor submitted that properly construed, the existing funding arrangements, including the indemnities, are not in fact interim arrangements. The plaintiffs have the benefit of secure funding agreements under which they are not exposed to any risk of paying the defendants' costs in the event that the proceedings fail, and will not themselves have to fund any obligation to give security for the defendants' costs. They will not have to pay Maurice Blackburn's costs unless the proceedings result in a recovery against the defendants. Given those circumstances, the answer to the question whether a group costs order is "appropriate or necessary to ensure that justice is done in the proceeding" turns on whether the proposed GCO is more advantageous to group members than the present funding arrangements. The evidence shows that for the purposes of that assessment, the existing arrangements are the real alternative to a GCO, as opposed to other possible or hypothetical alternatives.

45 The contradictor submitted that on that basis, on the plaintiff's own evidence, the existing arrangements clearly favoured group members. On the plaintiffs' modelling, on some projected recovery scenarios group members would be better off under a group costs order funding scheme than under the existing arrangements. However, on other 'realistic' scenarios the group would be worse off. Projections that posit that group members will be better off, or even not worse off under a GCO, compared with their existing funding arrangements, are subject to very considerable uncertainty. On the evidence I could not be satisfied that the statutory criterion for the exercise of the discretion to make a group costs order has been met. Other considerations such as transparency in funding arrangements did not warrant a different conclusion.

46 The defendants largely echoed the submissions of the contradictor (although the defendants and contradictor were not in agreement on every point made in submissions).

47 As ANZ put the central proposition, the statutory test directs attention to justice being done in the proceedings, which is a broad and evaluative test. However, in this case factors that might otherwise establish that a group costs order is necessary or appropriate to ensure that justice is done in the proceeding - in particular, the provision of an indemnity against the risk of adverse costs and the requirement to give security for costs - are not present. That risk has already been neutralised by the existing arrangements, as has the requirement that the group fund the costs of their proceeding. A GCO must, if it is to be made in accordance with the statutory test, in some way better secure the position of the plaintiff and group members in the proceeding. On these facts, there is no need for a GCO in order to ameliorate the risks assumed by the plaintiff, or to otherwise secure the funding of the proceeding. There is no risk of justice not being done in the sense of the group not being able to readily pursue their claims in the proceeding.

48 As a result, ANZ submitted, the inquiry whether a group costs order is "appropriate or necessary" directs attention to the "price" of the relevant funding mechanisms, meaning whether a GCO is likely to deliver a better financial result to group members than other funding arrangements. The alternative is to be identified by asking what would occur if a GCO were not made. In this case, the existing NWNF funding is the relevant alternative and is the model against which the proposed order should be assessed, because that funding would remain in place pursuant to a binding contract, as would the indemnities already given by Maurice Blackburn. ANZ said that the particular arrangements in this case do not present a good vehicle for the development of principles generally applicable to applications under s 33ZDA. ANZ's submission was that on evidence available to it (which did not include the plaintiff's modelling specific to this case) the existing arrangements plainly favoured group members. Westpac made a comparable submission.

49 The plaintiffs, contradictor and defendants were in agreement then, about how the statutory test should be applied in this case. They were also in agreement concerning the analytical process by which the existing NWNF funding may be compared with the proposed GCO (discussed below).

50 I agree that in these particular cases it is appropriate, in answering the question whether the making of a group costs order is appropriate or necessary to ensure that justice is done in the proceeding, to consider whether there is sufficient evidence on which to conclude that the proposed GCO would likely deliver a better financial outcome to group members than other funding arrangements would deliver. I also consider, for reasons discussed in Part D, that "better" in this instance means better than the existing NWNF arrangements.

51 Whether the terms and effect of the proposed order are generally advantageous to group members is not the statutory question[39] and, as discussed below, whether group members are likely better off under the proposed GCO than under another arrangement is not a general proxy for the statutory test. In this case, the plaintiffs' central contention is addressed to a comparison between funding models and to outcome, because that is what presents on facts as the substantive and most significant basis on which the two models may be distinguished.

52 I accept that the relevant inquiry is an evaluative one, and that in this case there are relevant factors other than price. However, the plaintiffs and indeed all parties rightly emphasised price (the cost of funding assessed in relation to outcome) as the most significant indicator of whether making a group costs order was justified.

53 The significance of the relevant considerations is further considered in Part D, in the context of the evidence.

Part D - The plaintiffs' contractual position and alternative funding arrangements

54 The plaintiffs said that third party litigation funding was the alternative arrangement against which the proposed GCO should be assessed, for two reasons, expressed this way: first, Maurice Blackburn entered into the NWNF agreements with the plaintiffs "in contemplation that the present applications for the GCO would be made". Second, if a GCO were not made, "the resulting scenario would entail Maurice Blackburn approaching [the litigation funder] Vannin Capital Operations Ltd (Vannin) to negotiate whether they may fund the proceeding on the basis of a traditional third party litigation funder".

55 Whether making the proposed group costs orders would be a proper exercise of discretion under s 33ZDA is to be informed in this case by the contractual position of the plaintiffs and the evidence as funding arrangements for the proceedings. It is convenient to set out the contextual matters first.

How the application was framed - Maurice Blackburn's intention

56 Mr Watson's evidence was that Vannin had brought to Maurice Blackburn the idea of class actions arising from the payment of flex commissions in the car finance industry, which Vannin had investigated and determined to be meritorious. Vannin provided information about the proposed claims confidentially to Maurice Blackburn to enable it to conduct its own investigations. Maurice Blackburn and Vannin agreed that if Maurice Blackburn determined the proceedings had merit and if it was willing to act as the solicitor in respect of proposed proceedings, Vannin would fund the proceedings, with Vannin and Maurice Blackburn to negotiate commercial terms upon which the proceedings would be commenced.

57 At that time, Mr Watson was aware that the Justice Legislation Miscellaneous Amendments Bill was likely to become operative in July 2020. Mr Watson formed the view that as compared with ordinary third party litigation funding (where the funder would charge a commission in addition to being reimbursed for legal costs), group members would be better off if Maurice Blackburn commenced the proceedings on a NWNF basis and subsequently sought a group costs order. Maurice Blackburn then entered into two types of agreements with a view to facilitating the funding of these proceedings in anticipation of seeking a group costs order - a "costs-sharing agreement" between Maurice Blackburn and Vannin, and NWNF retainer and conditional costs agreements with each of the plaintiffs. Mr Watson's evidence was not that he considered that group members would be better off under a GCO than under the existing NWNF arrangements - that point was developed later in submissions, through the "tipping point" analysis (discussed in Part E, below). Rather, his evidence was that the NWNF arrangements were intended to apply until a GCO were made, or if it were not made, until third party funding was arranged.

NWNF costs agreements - Plaintiffs' contractual position

58 The terms of the retainer and costs agreements (the NWNF Agreements) are materially identical in the Fox and Crawford proceedings. Save for a reference to a proposed group costs order they are in a standard form. Although nothing turns on this, the plaintiffs are in fact "representatives" in their respective cases, but the agreements they have each executed appear to be in a form that Maurice Blackburn would provide to group members, who are described as "claimants". The Agreements relevantly provide as follows:

(a) The NWNF Agreements set out the terms on which Maurice Blackburn will act for the claimant and supersedes any prior agreement. The NWNF Agreements constitute an offer by Maurice Blackburn to enter into a legally binding costs agreement with the claimant.

(b) The claimant instructs MB to provide such legal services to the claimant or for the claimant's benefit as MB considers reasonably necessary prosecute the claims in the proceeding, perform investigative, common benefit and individual work, and negotiate a settlement.

(c) The claimant instructs Maurice Blackburn to take and act upon instructions from the representative (unless separate instructions are required from the claimant).

(d) The claimant instructs Maurice Blackburn to seek approval from a court for the firm to be paid a sum for legal costs or a group costs order from money received in a settlement in a class action.

(e) The claimant agrees that the representative will give binding instructions to Maurice Blackburn and make binding decisions on behalf of the claimant in relation to the claims.

(f) Maurice Blackburn's professional fees, and disbursements, will be payable on a conditional basis, meaning that they will be payable only if there is a successful outcome in the proceeding. A successful outcome includes an award of money in favour of the claimant in connection with the claims of the claimant and/or the representative, the recovery of money as a result of any settlement or settlement approval by a court in a class action.

(g) 100% of the professional fees are conditional.

(h) An uplift percentage of 25% applies. The uplift fee is charged as a premium for conducting the proceeding on a conditional fee basis and is warranted because of the complexity, nature and circumstances of the proceedings and the risk to Maurice Blackburn in entering into the conditional costs agreement.

(i) Conditional fees are calculated at the hourly rates set out (which may be varied) and disbursements are charged at cost. Estimates for fees payable for common benefit work (including uplift) are given, as is advice that the estimates are not quotes and are subject to change. Major variables affecting the total legal costs are set out.

(j) Maurice Blackburn's costs will never exceed the "resolution sum" which means the amounts received on account of settlement or judgment in respect of the claims, as defined.

(k) The amount that Maurice Blackburn will be paid for its legal costs will be subject to court approval.

(l) The representative may apply to the court for a GCO. If the court makes a GCO, the amount that Maurice Blackburn will be paid will be subject to the court's orders, which to the extent of any inconsistency will supersede the agreements.

(m) These are not funded claims.

(n) Maurice Blackburn has a reasonable belief that a successful outcome is reasonably likely in the proceeding.

(o) Maurice Blackburn may reasonably change the terms of the NWNF Agreements and will provide 30 days' written notice to the claimant of any such change. The claimant is presumed to agree to the change unless the claimant gives written objection to Maurice Blackburn prior to the date the change takes effect (clause 13.5).

(p) If, under clause 13.5, the claimant objects to Maurice Blackburn changing the terms of the NWNF Agreements , MB may terminate the NWNF Agreements (clause 12.1(f)).[40]

(q) The NWNF Agreements otherwise set out obligations of the parties including in respect of giving instructions, confidentiality and privacy, billing and statements of account, the role of the representative, and the definition of work to be performed.

59 It is apparent from these terms that the NWNF agreements contemplate that the plaintiffs may apply to the Court for a group costs order and if a group costs order is made the Court's order would, to the extent of any inconsistency in the agreement, "supersede" that part of the agreement. It appears to have been intended that the effect of a group costs order would be that that part of the agreement providing for costs to be calculated and charged conditionally on an hourly rate basis with an uplift would no longer be operative, but the agreement would otherwise remain in place.

60 It is plain that the agreements are not in any sense interim or conditional meaning that they would come to an end or cease to be binding on Maurice Blackburn in the event that a group costs order were not made. There is no indication in the agreements that Maurice Blackburn will cease to act for the plaintiffs in the event that a group costs order application is not successful; rather, the NWNF Agreements contemplate that Maurice Blackburn will continue to act.

61 Further, the NWNF Agreements do not provide that, failing the grant of a group costs order, Maurice Blackburn may obtain third party funding that will "supersede" the NWNF funding arrangements, or operate as a variation to which the plaintiff's consent is taken to have been given.

62 Ultimately, Maurice Blackburn did not submit that NWNF Agreements were not binding or that they could necessarily be properly varied by substituting the current funding terms with third party funding. The highest the submission was put was that pursuant to clause 13.5 Maurice Blackburn is entitled to vary the agreements, subject only to the variation being "reasonable".

63 The presence of the variation provisions (clauses 13.5 and 12.1(f)) does not have the effect that the agreements can be properly characterised as effecting an interim arrangement such that if a group costs order were not made, Maurice Blackburn would be at liberty to substitute a fundamentally different set of terms.

64 Maurice Blackburn submitted that any future variation under clause 13.5 would have to be assessed on its terms and that it is not self-evident that a variation switching to third party funding would necessarily constitute an unreasonable change to the agreement. The first part of that submission is evidently correct. The second part of the submission is correct insofar as it follows from the first, but it fails to engage with the evidence on the plaintiff's application that third party funding generally requires the payment of legal fees in addition to the payment of a commission, and that funding commissions have historically been charged in the general range 21-29% of recoveries. A future variation of the existing arrangements to switch to third party funding could conceivably amount to a reasonable change, depending on the circumstances, which may include considerations other than the cost of funding. But there was no evidence, on this application at this time, sufficient to support a conclusion that such a change (amounting to a reasonable change within the meaning of clause 13.5) was likely to occur. The general evidence of Maurice Blackburn's intentions was not a sufficient basis on which to so conclude.

65 To be clear, Maurice Blackburn submitted that it would not likely take the step of seeking to change the terms of the NWNF Agreements and would only do so after careful consideration of the plaintiffs' and group members' interests and on the instructions of the plaintiffs. I do not doubt the sincerity of that submission, and I apprehend that that reserve underpinned the ultimate approach to the application which was to say that the arrangements had to be assessed by reference to which would provide the "better deal" to group members, focussing centrally but not exclusively on projected outcomes.

66 There was no evidence as to whether in the absence of a GCO the plaintiffs wished or would be prepared to renegotiate a fresh retainer and costs agreement accommodating third party funding and associated funding commission payments.[41]

67 It follows that, notwithstanding a stated intention on the part of Maurice Blackburn to seek third party funding in the event that group costs orders are not made, the default contractual arrangement is the present NWNF agreement, and not third party funding.

68 I accept Mr Watson's evidence that Maurice Blackburn entered the NWNF Agreements in anticipation of a group costs order being made. However, ultimately, the evidence established no more than a subjective intention on the part of Maurice Blackburn that the NWNF Agreements act only as an interim arrangement, in effect as a bridge between the commencement of the proceedings and the Court awarding a GCO or the plaintiffs and Maurice Blackburn making alternative arrangements with a third party funder, and only failing that, reverting to a NWNF arrangement. That subjective intention, however, was not reflected in the contractual agreements between Maurice Blackburn and the plaintiffs. It was not Maurice Blackburn's contractual intention, as objectively discerned from the contractual documents.

69 Maurice Blackburn did not put its submission on the basis that the terms of the written NWNF Agreements as set out above should be read as having been varied in light the "surrounding circumstances" relevant to its intention, or subject to a condition subsequent.

70 The evidence was that Mr Watson was advised by each of the plaintiffs that without an indemnity or undertaking being offered from Maurice Blackburn or a third party funder to pay any adverse costs that may be made against the plaintiffs in the proceedings:

(a) the plaintiffs would not have either the means to satisfy or the willingness to be exposed to the risk of adverse costs orders of the magnitude that would be likely in the event the proceedings were unsuccessful;

(b) in the event that the proceedings were unsuccessful, the plaintiffs would not have either the means to satisfy or the willingness to incur legal costs that are likely to be incurred in prosecuting the proceedings and the plaintiffs would not have agreed to act as the representative plaintiffs in any class action were they so exposed.

71 On the question of an indemnity against the plaintiffs' risks of conducting the proceeding, Mr Watson's evidence was that "in contemplation of a group costs order being made", Maurice Blackburn informed the plaintiffs that it would protect them from any costs exposure in the proceedings and would provide them with an undertaking to pay any adverse costs order. The provision of those indemnities was in keeping with the costs sharing arrangement with Vannin (discussed below), in which Maurice Blackburn and Vannin assumed obligations to each other about the provision of indemnities to the plaintiffs.

72 The evidence was that it is not Maurice Blackburn's practice to indemnify plaintiffs for whom it acts on a conditional basis in class actions against the risk of paying adverse costs - Mr Watson said the provision of indemnities was the exception rather than the rule. However, in these proceedings Maurice Blackburn has provided to each plaintiff an indemnity in respect of any liability to pay the defendants' costs and has agreed to pay any security for the defendants' costs that the plaintiffs might be ordered to pay, for which purpose it will obtain "after the event" (ATE) insurance.

73 I accept that the indemnities were provided "in contemplation" of group costs orders being made, however the terms of the indemnities were not in fact confined to the circumstances in which a group costs order is made, and will remain in place even in the context of a NWNF funding arrangement. The terms of the indemnities are set out in the confidential schedules.

74 Mr Watson said that if the costs sharing agreement with Vannin had not been negotiated in contemplation of a group costs order, Maurice Blackburn would have proceeded under a third party funding model. His evidence as to what would happen in this case if the GCO is refused is that he presently anticipates that Maurice Blackburn would approach Vannin to see whether they may fund the proceeding on the basis of a traditional third party litigation funding arrangement and, if so, negotiate terms. If Vannin and Maurice Blackburn were not able to reach terms amenable to group members' interests, then Maurice Blackburn would consider negotiating with other third party litigation funders or proceeding on a no win, no fee basis with appropriate adverse costs protection. There was no specific proposed third party funding model in place or described in the evidence, because it was not contemplated to proceed that way unless a GCO were not made. Mr Watson's evidence was therefore necessarily addressed to a hypothetical circumstance.

75 Mr Watson's evidence was that for the purposes of investigating and prosecuting the proceedings Maurice Blackburn entered into a costs sharing agreement with Vannin on 21 August 2020. The agreement is to the following effect:

(a) Vannin's obligations under the agreement will commence only on the "effective date", which is the date on which a GCO is made in each of the proceedings or a date which Vannin, in its sole discretion, may designate.

(b) Maurice Blackburn has commenced the proceedings and agrees (with Vannin) that it will conduct the proceedings on a 'no win, no fee' arrangement until such time as a GCO is made.

(c) Maurice Blackburn will cause an application to be filed in the proceedings seeking orders including a GCO.

(d) The costs of litigating the claims will be incurred in accordance with an agreed project costs budget as updated from time to time and varied in accordance with the agreement.

(e) Maurice Blackburn is and will be instructed by the plaintiffs.

(f) Subject to the effective date occurring, Vannin will pay half of the project costs (expenses incurred in relation to the litigation as defined in the agreement) as at the effective date and subsequently, on monthly invoices submitted by Maurice Blackburn.

(g) Maurice Blackburn will pay half of the investigation expenses incurred by Vannin.

(h) Maurice Blackburn will pay to Vannin half of the amount of any contingency payment, without set-off (any payment received by Maurice Blackburn or a related entity pursuant to a GCO or similar order).

(i) The parties acknowledge that once a GCO is obtained, Maurice Blackburn will be required to indemnify the plaintiffs or otherwise provide an undertaking to pay in respect of any amount of adverse costs ordered against the plaintiffs. Vannin agrees to pay to Maurice Blackburn half of the amount of any adverse costs Maurice Blackburn has to pay on behalf of the plaintiffs pursuant to Maurice Blackburn's indemnity or undertaking.

(j) To the extent that security for costs is ordered, Vannin agrees to provide half of the total amount of security ordered.

(k) To the extent that the plaintiffs are the beneficiary of any adverse costs recovered during the proceeding or following settlement, those funds will be distributed between Vannin and Maurice Blackburn in accordance with the applicable rate of proceeds sharing between the parties at the time of receipt of the funds.

(l) If the effective date does not occur within the time contemplated by the agreement (the sunset date), Vannin will have no further interest in the claims the subject of the proceeding.

(m) The agreement annexed a project budget for each proceeding for professional fees and disbursements on the assumptions set out.

76 Three things may be said about the costs sharing agreement.

77 First, the plaintiffs are not party to it. It is necessary, then, to look to the contractual relationship between Maurice Blackburn and the plaintiffs to determine the existing funding terms. Those terms, unlike the costs sharing agreement, did not stipulate that Maurice Blackburn would conduct the proceedings on a NWNF basis until such time as a group costs order is made.

78 Secondly, the costs sharing agreement puts in place a regime between Maurice Blackburn and Vannin that comes into effect only if a group costs order is made. It does not contemplate any ongoing relationship between Maurice Blackburn and Vannin in respect of these proceedings in the event that a GCO is not made. It is not evidence of the existence of third party funding for the proceeding (and nor was it put as such).

79 Thirdly, the plaintiffs distinguished the costs sharing agreement from a tripartite third party litigation funding agreement, emphasising that the plaintiff and group members have no contractual relationship or any other relationship with Vannin as they would a third party litigation funder. Vannin is not to provide "legal services" and would not be receiving "legal costs" as defined in s 3 of the Uniform Law. From the perspective of the plaintiff and group members, Maurice Blackburn assumes all liability to pay any costs payable to the defendants and to give any security for costs to the defendants that the court may order, as per s 33ZDA(2). As Maurice Blackburn put it, the effect of the arrangement between Vannin and Maurice Blackburn is that Vannin is a financier to Maurice Blackburn, and not a third party litigation funder of the proceedings.

80 Westpac submitted that s 33ZDA does not authorise a group costs order where a funder is involved, in reliance on the words, "legal costs payable to the law practice" in s 33ZDA(1)(a), in that legal costs may not be made payable to a funder. ANZ said, on the same basis, that it questionable whether s 33ZDA authorises any payment to a financier to a law practice.

81 The contradictor submitted (substantially agreeing in this respect with the plaintiffs' position but not expressing a definitive position), that any group costs order would necessarily be expressed as permitting the payment of a percentage of settlement or award by way of "legal costs" to the "law practice" rather than to any third party litigation funder, and any award or settlement would be paid by the plaintiff and group members to the law practice directly, and not to a third party litigation funder. Any subsequent sharing of those moneys between a law practice and a third party litigation funder would occur pursuant to a "commercial side arrangement" and not by order of the Court. The third party funder would not strictly be receiving any payment for "legal costs" pursuant to any group costs order. Once that is accepted, the separate financing arrangement is not relevant to the power to make a GCO. There is force in these submissions.

82 In a similar vein, it may be said that by operation of s 33ZDA, if a group costs order is made, the law practice is liable to pay any costs payable to the defendant in the proceeding and must give any security for costs of the defendant that the Court may order the plaintiff to give. The fact that the law practice might defray that risk by obtaining its own financing or insurance does not alter the liabilities and entitlements that flow from the making of a group costs order, namely that the legal costs payable to the law practice be calculated as permitted by sub-s 33ZDA(1) (at the percentage set out in the order), and that the law practice becomes liable for costs as provided for in sub-s 33ZDA(2).

83 The fact that the law practice is not bearing all of the risk might inform the exercise of the discretion in respect of whether to make a group costs order, and what percentage to fix. As the contradictor put it, there may be circumstances in which the court considers that the proposed arrangement to provide funding to the law practice is an acceptable mechanism by which the law practice is justified in reducing in part, even in substantial part, the risk assumed, particularly in the context of a proceeding of substantial scale and risk. Such an arrangement might also be viewed as analogous to a loan facilitating the law practice funding the action. On the other hand, there might be circumstances in which the Court considers that the nature of an arrangement with a third party is such that a group costs order ought not be made, for example where the law practice is acting as a "mere front" for a third party funder.

84 As the plaintiffs rightly submitted, any such consideration must take into account the whole of any such financing arrangement, including the costs to the law practice.

85 The defendants' submissions were not directed to any detailed consideration of the provisions of the costs sharing agreement between Maurice Blackburn and Vannin.

86 I am not persuaded by the general submissions that the existence of the cost sharing agreement has the effect that there is no power to make a group costs order, or that by itself, the existence of the arrangement would be reason to refuse an order. However, because of way I have decided this application it is not strictly necessary to consider this point further.

87 Maurice Blackburn submitted that the existence of what was described in submissions as a "contractual fall-back position" in the NWNF Agreements facilitated the opportunity to negotiate other funding arrangements. Had the NWNF Agreements provided that they were to terminate on the failure of a group costs order application, that would not be protective of the plaintiff and group members' interests. Maurice Blackburn submitted that its approach was both responsible and reasonable.

88 It was submitted that if the contradictor's analysis is accepted the practical effect would be commercially unworkable for plaintiffs, law practices and litigation funders, and contrary to the policy behind the introduction of s 33ZDA. Taking the contradictor's submission to its logical conclusion, applications for group costs orders would only be successful in the event that retainers between law practices and plaintiffs were conditional and terminated in the event that a GCO was unsuccessful or plaintiffs, law practices and litigation funders spent considerable time and resources negotiating short-lived funding agreements that would soon become redundant upon the making of group costs orders. Moreover, it would not be practicable to inform a potential representative plaintiff that a law practice is prepared to act on his or her behalf but if a GCO not obtained, the law practice will cease acting. Very few, if any, persons would assume the considerable burden and responsibility of becoming a representative plaintiff in such a scenario. It is plainly not in the interests of group members and inimical to the pursuit of access to justice for such outcomes to become "standard practice".

89 Although those factors explain why Maurice Blackburn put in place the arrangements it did, they do not permit a different legal characterisation of the NWNF Agreements (and were not said to have that effect). They do not provide a basis to overlook the legal effect of the NWNF Agreements, particularly in the circumstances in which these applications are framed - focusing on what is the better deal for the plaintiffs and group members. Furthermore, a general proposition about access to justice in matters of this kind is not a substitute for the particular analysis of the interests of the plaintiffs (and through them, the group members) in these proceedings.

90 Finally, whilst the broad policy considerations articulated have some attraction, the difficulties of accommodating a genuinely interim arrangement in contractual terms may be overstated. It ought also be recalled that the significance of the comparative analysis of returns, and hence the focus on the contractual terms, has arisen because of the particular circumstances of this case. The present facts may well be anomalous (see further, Part E).

Part E - Should a discretion be exercised in these cases?

91 As noted, the plaintiffs' case was that even if the existing NWNF funding was the proper comparator against which the proposed group costs order should be tested, in the Fox proceeding the group would be better off, and in the Crawford proceeding the group would be at least no worse off, or better off, on revised modelling.

92 For the most part the considerations relevant to the Fox and Crawford proceedings are the same, and I differentiate between them only where necessary.

93 For the purposes of the comparison between funding models, "return" and "recovery" in this context refer to the amount remaining in the plaintiffs' hands after the deduction of legal costs from any monetary amount awarded or recovered on settlement (whether calculated in the traditional way under the NWNF agreements (with an uplift), or as a percentage of the recovered amount, under a group costs order). The analysis did not take into account the fact that costs might be recovered from the defendant. For the purposes of comparing the effects of the different funding arrangements the analysis assesses both models in the same way, permitting a valid comparison. While no submissions were directed to this question and it is not necessary to consider it further on this application, it should be noted for clarity that in reality, on the premise that the indemnity principle is not displaced by s 33ZDA, recovery of costs against a defendant would favourably impact the flow of funds to the plaintiffs' account.

94 The precise interaction of s 33ZDA, the indemnity principle (and in practical terms, the structuring of settlements) are questions for another occasion. There are also questions, for another day, about how a group costs order would take effect in the context of a judgment or judgments in favour of a plaintiff which, in the usual course, would result in the determination of common questions and a monetary award in the first instance on the plaintiff's personal claim, with recovery by group members determined or resolved later.

95 The existing NWNF and proposed group costs order arrangements can be compared this way:

(a) Both provide an indemnity to the plaintiffs against adverse costs.

(b) Both provide an indemnity to the plaintiffs against any obligation to give security for the defendants' costs.

(c) In both models, only if a recovery is made in the proceedings (whether by settlement or judgment) will the plaintiffs pay costs.

(d) Under a GCO, costs will be shared among the plaintiffs and all group members. The statute has that effect without the need for the plaintiffs to seek an order at a later stage. Under the existing arrangements the plaintiffs can be expected apply for an order under s 33V of the Act effecting an equitable distribution of costs upon settlement, or an order under s 33ZJ for reimbursement of their costs if an award is obtained consequent upon a judgment.

(e) Under the proposed GCO group members will be guaranteed a 75% percent return of any recovery. That is a consequence of fixing costs at 25% of recovered amount subject to further order, by which the percentage may be adjusted. I construe the present applications to intend that the fee not exceed 25% of the recovery sum, regardless of any future adjustment. The existing NWNF Agreements (which I infer will apply to both the plaintiffs and group members) provide that costs will not exceed the recovery sum. They therefore permit recovery by the plaintiff and group members at less than 25% of the total recovery, and permit a zero recovery result, because they allow for the possibility that the recovery sum might may be eroded by legal costs, up to but not exceeding the total recovery amount. Fees are calculated and will be charged on an hourly rate basis, with an uplift. A preliminary fee estimate is given but, as group members are made aware, it is not binding - it is the best estimate the solicitors are able to give and may be revised.

(f) Mr Watson's evidence, which I accept, was that class actions are complex, often difficult and hard fought in a dynamic legal environment and may be delayed by lengthy periods beyond the control of the most judicious plaintiff's lawyer. Costs in class actions can and do blow out under time-based billing arrangements, which impacts on the proportionality of returns to group members. In Mr Watson's experience it is more common than not for legal costs estimates and budgets to be revised upward over the life of a class action. That is subject to the significant caveat that the amount of costs that Maurice Blackburn will be paid will be subject to Court approval, which is both a contractual term and a reflection of the principles informing the application of ss 33V and 33ZJ of the Act.

96 The contention that group members would be better off under a group costs order was not limited to the question of what the group might eventually recover. The plaintiffs submitted that a group costs order would engender certainty, transparency and the alignment of the economic interests of the group and the law practice (those factors are considered below). But they accepted (as the thrust of their submission demonstrated) that the most significant consideration was the financial outcome to group members. The plaintiffs emphasised that what a group costs order affords is not just a certain costs measure, but a guarantee against the erosion of returns to the group by legal costs, which is particularly relevant where recoveries are poor. The "insurance" against poor outcomes is a benefit that has to be considered in conjunction with the analysis of possible returns.

97 Turning to the question of the financial benefit to group members of making a group costs order (the "outcome" measure) the plaintiffs, contradictor and defendants were ad idem concerning the means by which the potential returns group members should be evaluated, namely by identifying for each proceeding the "indifference point" (or "tipping point") which is the settlement or judgment amount at which the outcome to the group would be the same under a GCO or NWNF regime, and concomitantly the point above which group members would receive better returns on a NWNF basis, and below which they would be better off under a group costs order.

98 As Mr Houston explained (and is obvious in any event) a comparison of the outcomes that NWNF and GCO funding regimes will deliver to group members starts with the proposition that in a NWNF model the proportionate relationship between legal costs and the recovery amount will vary with the recovery amount, whereas in a GCO model the return to the class is a constant proportion of the recovered sum. Under both models, where a proceeding is unsuccessful the result is the same to group members, if (and only if) both provide indemnities. Regardless of the rate at which at GCO is set, where a positive recovery results but is less than or equal to the total costs incurred, a GCO alone will deliver a return to the group. Where a settlement or award is greater than the total legal costs incurred but less than the "indifference point", the GCO represents a higher return to the group. In those situations the GCO model transfers the risk of a "poor" outcome to the law practice. For all outcomes in which the recovered sum is beyond the indifference point, the NWNF model results in a higher return to the group.

99 As a matter of basic arithmetic an order which permits the allocation of 25% of recoveries in a proceeding to costs, will produce the same result to the group as a NWNF-funded proceeding in which the award or settlement happens to equal four times the amount of the costs charged. Similarly, a 20% contingent costs order will produce the same result as a NWNF funded proceeding in which the award or settlement is equal to five times the costs charged, and so on. If one starts with a notionally fixed quantum of costs the recovery amount which will produce this coincidence marks the "point of indifference" between the GCO and NWNF models. For a proposed GCO at 25% the amount of settlement or award at the point of indifference is calculated this way:

(total estimated legal fees x 1.25 [to account for the 25% uplift] + disbursements + cost of ATE insurance premium) ÷ 0.25[42]

100 One must then compare the indifference point to the range of estimated outcomes for the proceedings. That analysis requires plotting the tipping point against projected outcomes, to which the inputs are, in turn, projections as to group size, group members' prospective damages entitlements, and prospects. It will be immediately apparent that an assessment of that kind will commonly (and does in this case) entail significant uncertainty if made at an early stage in the proceedings.

101 The plaintiffs on the one hand, and the contradictor and defendants on the other, were in agreement as to how the "tipping point" should be identified, but were at odds as to how that uncertainty should be evaluated in the context of these applications.

102 Each of the inputs to the outcome projections was quantified in the evidence of Mr Watson, and the reasoning supporting the inputs and Mr Watson's analysis of each component was described in some detail in confidential parts of Mr Watson's affidavits provided to the Court and the contradictor. In my reasoning set out here I have omitted content that would reveal Mr Watson's assessment and opinions as to costs and prospects of the proceedings that are properly confidential to the plaintiffs. I elaborate on my reasoning in the confidential schedules provided to the plaintiffs and contradictor in each case. The inputs to the outcome projections were derived in the following way:

(a) The estimated costs amounts were taken from Maurice Blackburn's litigation budgets for solicitors' fees and disbursements (which estimates are likely to change as the proceedings progress) and a premium payable for ATE insurance.

(b) The damages pool was estimated by assessing material made public in connection with the Financial Services Royal Commission, in which features of the loan books of the defendants were described, including the extent and quantum of lending involving flex-commissions. Mr Watson estimated the rate at which affected consumers could be expected to participate in the proceedings, expressed as a percentage range, which was applied to the headline total pool figure to reach a range of damages for each proceeding (at this point not discounted for risk).

(c) For the purposes of projecting prospects of success in the proceedings, Mr Watson initially gave a very generally expressed opinion, acknowledging the uncertainties affecting the inputs to any assessment made at this point of the proceedings. As discussed below, analysis initially prepared for the purposes of estimating Maurice Blackburn's potential return on investment under a GCO (as a cross-check of reasonableness) was developed to address prospects more generally. Mr Watson developed five scenarios reflecting his assessment of prospects (projected outcomes) in five bands. Those projections were described by assigning to each scenario a total recovery amount in round numbers, and the likelihood of that outcome occurring, expressed as a percentage. Those assessments reflected the analysis described above, Mr Watson's and Maurice Blackburn's own considerable experience in class actions, on which the firm keeps detailed data, and publicly available data quantifying the class action outcomes in Australia.

(d) Mr Watson performed the modelling himself. He personally undertakes modelling of that kind regularly, in his role as head of class actions at Maurice Blackburn. He draws on his deep and broad experience in class actions, of which he is a leading practitioner in Australia, and his skills include a degree in mathematics.

(e) From the projected recovery amounts and percentages for the five scenarios a probability weighted average was calculated. Mr Houston (employing simple mathematics) multiplied the quantum for each scenario by its probability (inputs to his calculation that were provided by Mr Watson), and added the outputs together. The result of that calculation was a single figure for each proceeding.

The position in Fox v Westpac - what the modelling shows

103 For the Fox proceeding there are two factors that undermine the plaintiffs' principal "better off" contention, namely what the plaintiffs' predictive modelling of outcomes shows on its face, and the uncertainties inherent in the modelling at this stage of the proceeding. Those considerations may at first glance appear to be in tension, but both point in the same direction.

104 First, on its face, the modelling does not sufficiently support the conclusion that the outcome (meaning funds in group members' hands) will likely be better for group members under a group costs order. For the reasons discussed in the confidential schedule, Mr Watson's opinion as to what scenarios are most likely (subject to the uncertainty inherent in the modelling, which Mr Watson acknowledges and explains), does not support the conclusion that at GCO at 25% will deliver a better result to group members than the existing NWNF arrangements provide. Even if the modelling could be placed on a more certain factual foundation, for example, if it were performed again at a later stage in the proceeding when some of the variables affecting it were less doubtful but yielded the same conclusions, it would not favour the grant of a GCO at 25%, subject to the additional considerations set out below.

105 The plaintiffs relied on the fact that the weighted average of Mr Watson's assessments (the mathematical calculation made by Mr Houston) is below the tipping point, meaning that it points to the GCO providing a better outcome to group members. The weighted average calculation is a mathematically derived synthesis of Mr Watson's probability assessments, although Mr Watson did not express his evidence in these terms. A weighted average calculation might have utility in other contexts but it is not sufficiently helpful in this context because assessments of the kind in question are matters of judgment. Mr Watson (as he explained) expressed his opinion by using numerical values to connote likelihood in respect of each scenario, but the assessment is ultimately a qualitative evaluation based on experience which I find persuasive, subject to the uncertainty inherent in the inputs to the evaluation. The weighted average calculation, used in this context, distorts the effect of the evaluative assessments and does not displace Mr Watson's the opinion as to what is most likely to occur, given the reasoning set out by Mr Watson.

106 Secondly, as Mr Watson candidly said, while predicting litigation outcomes is an essential part of litigation practice, it requires assessments based on several inputs that are uncertain. Both he and counsel properly emphasised that the possible outcomes for the proceedings remained, at this early stage, highly uncertain.

107 Mr Watson prefaced his assessments of prospects with that caveat, but explained that his analysis was his professional judgment, of the kind he regularly undertakes in connection with the firm's class action practice. The intellectual process reflected in the modelling is evidently sound and was not challenged by the contradictor, who had unrestricted access to the confidential modelling. Rather, the inputs to the process are uncertain.

108 In this case the uncertainties are embedded in the foundation of the analysis, because at this stage the size of the represented group can only be estimated within a range that has significant breadth. The estimate for the class size was that it was likely to be in the high tens of thousands or hundreds of thousands, based on an analysis of the bank's loan books made on public material.

109 On the plaintiffs' submission, it is essential to appreciate that the benefit conferred by a group costs order is an insurance against poor outcomes. The plaintiffs submitted that apart from comparing "pinpoint" hypothetical outcomes which may be suggested from the tipping point analysis discussed above, they and the group members would be "better off" under a group costs order because it will insure them against lower positive outcomes, which may otherwise be consumed by costs. The quid pro quo for that insurance is a lower return than would otherwise be the case, where there is a bigger recovery. The plaintiffs submitted that when a reward is higher, group members are less concerned about their "loss", in the form of extra funds going to the lawyers under a GCO than a NWNF agreement, because they already have a positive outcome. Group members are to be taken not to be concerned that their great outcome is essentially expensive, as they have been insured against worse outcomes when recoverability would be much more in issue, so the submission went.

110 The submission was not supported by any particular evidence from the plaintiffs, or developed by reference to any particular feature of this case, other than the general risk assessment. The turning point about which this transfer occurs is the tipping point. Below the tipping point, funds that would have gone to lawyers costs under a traditional arrangement instead supplement the (smaller) pool available for distribution to the group. Above the tipping point, some funds (which would have gone entirely to group members, costs having been, by definition, covered), go to the lawyers, though always only 25 cents in each dollar. In essence, therefore, with a group costs order fixed at 25%, group members are overall better off by having ready access to the "insurance" aspect of a GCO across the likely range of outcomes in each proceeding, exchanging better shares in the lower range for lesser shares in higher ranges.

111 The contradictor submitted that in all or most litigation low or poor outcomes would be possible to some extent, and that that fact did not of itself sufficiently demonstrate that the statutory criterion had been met, so as to justify the exercise of the discretion in this case.

112 The proposition that a group costs order would insure the group members against low or poor outcomes does not provide a sufficient basis on which to exercise the discretion to make an order in this case. It is true that were I to made a group costs order the group would be insured against the possibility that if the proceeding delivered a result below the tipping point their damages would be eroded by costs - they would instead be guaranteed a 75% return. For that, group members would trade away the prospect they might do better above the tipping point. That might be an advantageous trade, but the evidence did not permit me to so conclude. Otherwise, the plaintiffs' submissions described an inherent feature of the group costs order regime. Generally speaking, the fact that a group costs order provides a guaranteed proportion of whatever sum is recovered in litigation to group members is a feature of the legislation that, in a policy context, it might be fitting to describe as beneficial. But in a particular case, a general description to that effect is not sufficient. As I have said, on other facts, the trade-off between higher and lower outcomes might not present as the basis for the exercise of the discretion. In these cases, the plaintiffs are inviting the court to assess that trade-off and conclude that insurance against low outcomes is a more significant benefit to group members than the prospect of a greater return, without a sufficient evidentiary basis on which to do so.

113 Some observations must be made about the uncertainty inherent in modelling outcomes for the proceeding and the significance of uncertainty in this application more generally.

114 The uncertainty inherent in the modelling in this case at this stage matters for this application, because the plaintiffs are in fact the beneficiaries of an existing, stable arrangement that covers the risk assumed in conducting the proceedings on behalf of group members. Making a group costs order would effect a fundamental change in the arrangements from one funding model to another. In exercising a discretion under s 33ZDA I must be positively satisfied that making a GCO would be appropriate or necessary to ensure that justice is done in the proceeding, being mindful of the protective role in relation to group members that the Court is required to assume. So much was accepted in the framing of the plaintiffs' application, the crux of which was a comparison between different funding models. As I have said, in this case, the plaintiffs' central contention is addressed to outcome, because that is what presents on the facts as the substantive and most significant basis on which the two models may be distinguished.

115 Predictive modelling that is riven with uncertainty significantly undermines the essential proposition that group members will be "better off" under a group costs order. The contradictor, who was given full access to the plaintiff's material, put the point this way: there are realistic scenarios where the group will be worse off under a GCO and you cannot say with any certainty that they will be better off. The evidence was simply too uncertain to discharge the legislative onus of proving that the order sought satisfied the statutory condition for the exercise of the discretion. I accept that submission.

116 Furthermore, as I have said, on Mr Watson's analysis a "worse off" result is in fact the most likely scenario. Allowing for the uncertainty, other outcomes apart from that scenario are of course possible, as the predictive analysis itself shows. As the plaintiffs' counsel put it, there are also "realistic scenarios" below the tipping point in which the group will be advantaged by a GCO. That may be so but considered as a whole, the modelling is insufficient to support a change of the funding basis in these circumstances.

117 The "better off" analysis in the plaintiff's case has not consistently invoked language that suggests a particular standard of proof ("likely", "possible", "would", "might", and so on). That is because the predictive modelling is concerned with projections and hypothetical scenarios based on significant factual assumptions, notwithstanding that it is rationally based and draws on considered professional judgment. It would be unhelpful to seek to formulate a general proposition about how one should approach the question of outcomes-focused assessments in this context. It suffices to say that in making the evaluative assessment required by s 33ZDA in this case, there is a demonstrable absence of positive proof supporting the "better off" contention. Ultimately, the plaintiffs asked the Court to exercise the discretion by concluding that the group costs order would advantage group members in the eventual outcome, and their evidence was not capable of establishing that proposition.

118 A different case would require a different analysis. The relevance of predictive modelling (if any) in a given case will depend on the subject matter to which the "appropriate or necessary" examination is directed. A different case might present a different and sufficient reason why a group costs order would be appropriate or necessary, for example because it would ameliorate existing risks to the plaintiff of continuing with the proceeding.

119 In a different context, the proper administration of justice sometimes requires the Court to manage the existence of competing class actions. In that context, the Court is required to solve the multiplicity question posed but not answered by Part 4A, which does not prevent the filing of a second representative proceeding against a defendant in relation to a given controversy.[43] There, comparative forward-looking projections of litigation outcomes, made on the available evidence despite its limitations, can provide a sufficient foundation on which to choose between competing proceedings and make related or similar orders.[44] However in this context, where the Court is asked to exercise a discretion under s 33ZDA, substantially on the basis that a group costs order would cause group members to be better off than under their existing arrangements or alternative arrangements, the question confronting the court is a different one.

120 The difficulties inherent in the forward looking projections in this case illuminate the fact that an outcome-focused analysis is not a particularly apposite touchstone for the question of whether a group costs order is appropriate or necessary to ensure that justice is done in a proceeding, when the question is asked at an early stage in that proceeding. As discussed at the outset, s 33ZDA is intended to be capable of operation at an early stage in the proceeding, and should sit comfortably with the existence of risk in a proceeding, given the risk-reward relationship at the heart of the funding model it embodies. I accept the plaintiffs' submission that s 33ZDA should not be construed as embodying threshold requirements not present in the legislative text, especially when those requirements would make it harder, not easier, for plaintiffs and group members to conduct representative proceedings. But that is not really to the point. In this case, the removing of the plaintiffs' risk from the equation by the particular "NWNF with indemnity" arrangement available here, has resulted in a focus on outcomes that might not have otherwise arisen.

121 That is not to say that, as this case shows, an outcomes based analysis is not an available way of approaching the question whether a group costs order is "appropriate or necessary"; it is only to say that it can be a difficult test to satisfy, and that whether or not it may be satisfied in that way will depend on the particular evidence in question.

122 To put the plaintiffs' predictive analysis in context, it should be observed that the five scenarios reflecting Mr Watson's evaluation of projected settlement outcomes in proceedings were initially formulated in order to instruct Mr Houston to undertake financial modelling to estimate Maurice Blackburn's return on investment should a GCO be made, as a cross-check of the reasonableness of the rate proposed.

123 Mr Houston had advanced the contentions in his first report that a GCO fixed at 25% would produce a materially higher return to group members than the 45-64% retained by the class in historically approved settlement outcomes; that the historical rates of return to group members in third party funded proceedings was a sound comparator against which the proposed GCO may be assessed, and that in matters where the complexity of the case and the size of the disbursements would be expected to be material, a NWNF model is "not likely to be appropriate in remunerating a law practice (or any party) for bearing these financial risks".

124 Shortly before these applications were heard the plaintiffs submitted a further report in which Mr Houston elaborated upon the ways in which the GCO and NWNF arrangements might be compared, proffering the probability weighted sum of the projected outcomes in Mr Watson's scenarios as "one way of synthesising the range of different possible settlement outcomes" and comparing the tipping points derived earlier, with the probability weighted single outcome figure for each proceeding. On that basis it was said that under a GCO group members would be better off in the Fox proceeding and in the Crawford proceeding the expected outcome value coincided almost exactly with the tipping point for that case, so that group members would not be worse off under a GCO. Mr Watson's detailed reasoning in relation to his five scenarios was provided in a further affidavit in response to a question from the Court as to the basis for those scenarios as set out in Mr Watson's first affidavit. This context underscores the fact that the plaintiffs' outcomes-based comparative analysis of the GCO and NWNF regimes was very much a fall-back position; they did not frame their application setting out to prove, on the basis of predictive modelling, that the proposed GCO was better for group members than the NWNF arrangement.

125 Separately, the plaintiffs said that the likely outcomes in these cases compared favourably to outcomes indicated by quantitative data on the historical rates of the return to group members in class actions including consumer class actions. The data came from the reports of published inquiries into class actions in which rates of return to group members was considered, and from Maurice Blackburn's own data pertaining to the cases it has conducted over a number years. Drawing on that data Mr Watson set out a hypothetical comparative analysis of the kind described above (turning on a tipping point at which the outcomes were the same under both funding models). The historical data modelled in that way, assuming the actual estimated costs for these proceedings, supported the GCO model as delivering a better outcome to group members. In these cases, where the principal benefit of the proposed GCO is tied to outcome and the primary contention is that the order will cause group members to be better off, it is unsatisfactory to rely on historical data which averages returns in other cases. As the plaintiffs accepted (and as the contradictor rightly emphasised) these cases are comparable to other cases only at a high level of generality.

126 Brief mention of the defendants' position should be made. The defendants did not have access to the confidential parts of the plaintiffs' evidence. They each submitted that Maurice Blackburn had said in submissions to various inquiries, and by a statement in their standard fee agreement, that their costs in class actions had historically been on average approximately in the range of 12-15% of the recovered sum, and if their costs were in keeping with that range, then group members would do better under the existing NWNF arrangements than under the proposed group costs order. So much was uncontroversial, but it did not address the particular circumstances of these cases, as the defendants accepted. Their submissions focused on the plaintiffs' requirement to discharge the burden of persuading the court to exercise the discretion on a sufficient basis. Westpac made a separate submission about the likely rate of return that the proposed GCO would deliver to Maurice Blackburn, which is considered in the confidential schedule.

127 I do not consider that the other factors to which the plaintiffs point, warrant a different conclusion.

128 The plaintiffs' applications were each founded on a set of factors, namely that a group costs order would deliver a better outcome, would engender simplicity, transparency and certainty in respect of funding, would facilitate a fair distribution of costs among group members, and was pitched at a percentage that was proportional to the risks to be incurred by Maurice Blackburn. Those features were relied on cumulatively, however it was not put that if the outcome measure was neutral, unfavourable to group members or not established, that the other measures would of themselves sufficiently support the exercise of discretion in the circumstances.

129 I agree with the plaintiffs that group costs orders make simple, certain and transparent the basis on which legal costs will be calculated. However, I do not consider that those factors in this case, justify the exercise of the discretion, given the primary submission on which the plaintiffs' "better off" case was put, in which the plaintiffs invited the Court to exercise a discretion to effect a substantive change in the funding of the proceeding. I accept that a group costs order funding regime would entail a simpler (indeed very simple) method of calculation of legal costs. Transparency and certainty are related virtues. Once again, as a matter of policy, those are features inherent in the group costs order regime that may be characterised as beneficial. In other cases they might assume more significance. But given the way the applications were put, and the other substantive considerations on which they turn it is relevant that the evidence does not suggest that the means by which costs are calculated a NWNF model, or the need for the legal practice to explain their costs to group members, has adversely affected group members' ability to participate in this proceeding or that that is a feature of class actions more generally. Indeed, experience demonstrates the contrary.

130 The sharing of costs is facilitated when a group costs order is made, by operation of s 33ZDA(2). That feature of the regime engenders both certainty and convenience. But in this case, their existence does not warrant a different outcome to that otherwise indicated. The plaintiffs in this case can be expected to apply for, and can reasonably expect to obtain, an order that the costs of their proceedings incurred on a NWNF basis will be distributed equitably between group members. Because third party funding is not, on the facts before me, in issue, a "funding equalisation order" would not be required, unless the circumstances change.

What the "better off" analysis does not stand for

131 The conclusion as to how the statutory test is to be applied in this case is not intended to suggest that s 33ZDA requires that an applicant for a group costs order in every case positively prove that group members' recovery will likely be greater under a GCO than under any other potential funding regime that could possibly or reasonably be obtained.

132 On the present evidence, predictive modelling was employed to compare the outcomes under different funding arrangements because of the particular arrangements in this case.

133 The defendants pointed to the fact the plaintiffs did not proffer evidence of the terms on which third party funding would be made available to the plaintiffs. That fact was relevant in assessing the plaintiffs' contention that the costs to group members under the proposed GCO should be compared against likely costs under third party funding.

134 However, some of the defendants' submissions went further. The Macquarie parties submitted that on its proper construction s 33ZDA would require that in the event that the plaintiffs' lawyers were not prepared to act on a no-win no-fee basis, the plaintiffs would have to elicit evidence as to why that was so. Were that circumstance to arise in a future case it would have to be decided on the facts there presented. But to the extent that any of the defendants meant to submit that a Court must approach s 33ZDA by first identifying the relevant counter-factual funding scenario and then by determining whether the proposed GCO was proved to be more advantageous to group members than the counter-factual funding, and that an application for a GCO must necessarily proceed in that way, I reject that submission. I do so for several reasons.

135 To read the statutory test in that way is to read it down by implying into it conditions not present in the text. The statute requires that a group costs order be appropriate or necessary to ensure that justice is done in the proceeding, but the statutory text, read in its context, does not require or suggest that a GCO is intended to be available only as a funding model of last resort. It does not require or suggest that a group costs order may be awarded only if it can be positively proved that it would deliver a better financial outcome to group members than some other funding model. Whether or not that comparative assessment is a relevant consideration will vary from case to case. A comparative analysis may be relevant on the facts of a particular case (as it is here).

136 As the plaintiff submitted, construing the statute in that way is not a purposive reading.[45] Reading into s 33ZDA a general threshold requirement to construct hypothetical counterfactual funding scenarios that the plaintiffs do not intend to take up, against which plaintiffs must positively prove that their proposed group costs order compares favourably, would read into the statute conditions that do not sit comfortably with its overarching purpose of facilitating access to justice by securing the funding of group proceedings.

137 Some of the defendants' submissions emphasised that the extrinsic material[46] identified the primary purpose of the statute as removing the risk to prospective plaintiffs in group actions, of having to pay the defendants' costs, which was described as a particular disincentive to plaintiffs in taking up the Part 4A procedure. If the risk has been ameliorated by the provision of an indemnity under another funding model, then the primary purpose for which s 33ZDA was been introduced could not be satisfied, so the submission went.

138 The submission did not appear to go so far as to say that where the plaintiff's risk of becoming liable for adverse costs was not in issue a group costs order could not be made. If that was the intended effect of the submission I would reject it. Regard to the purpose of a statute is not a warrant for re-writing the plain language of the statute so that it conforms with presumed statutory intention.[47] The statutory text does not admit of such a limited reading, and as Gleeson CJ said in Carr v Western Australia, "[l]egislation rarely pursues a single purpose at all costs. Where the problem is one of doubt about the extent to which the legislation pursues a purpose, stating the purpose is unlikely to solve the problem".[48]

139 In any given case there might be a range of reasons why the making of a group costs order is appropriate or necessary to ensure that justice is done in the proceeding. The comparative exercise serves a particular purpose in this case. In a different case, for example where a GCO would ameliorate the financial risks that a plaintiff would otherwise be required to assume, or where the plaintiff had obtained only genuinely interim arrangements for funding, an analysis of the likely outcome to the group might assume less importance. That was not this case.

Setting a rate - proportionality and reasonableness

140 The plaintiffs submitted that in addition to the "better off" measure, the court ought also to have regard, when assessing the proposed GCO rate, to these factors:

(a) the proportionality of costs sought by way of a GCO percentage;

(b) the rate of return that a GCO should provide to Maurice Blackburn and whether it is reasonable having regard to the anticipated costs and risks to be incurred by it;

(c) historical outcomes and returns to group members in representative proceedings of similar size and risk,

to which the contradictor added that in general, when assessing the reasonableness of a proposed commission rate under s 33ZDA, it might be relevant to consider,

(d) the commission rates presently charged by third party litigation funders with respect to representative proceedings of similar size and risk, and if the commission rates are different to the proposed GCO, whether the GCO percentage sought is nevertheless justifiable.

141 I have no difficulty in accepting that those considerations can meaningfully inform the setting of an appropriate percentage under s 33ZDA.

142 On the evidence, as discussed below, there is nothing to suggest that the proposed rate of 25% is inherently unreasonable, considered generally against third party funding commissions that have historically been offered in Australia, or that projected rates of return to Maurice Blackburn suggest that such a rate would produce disproportionate profits.

143 However, for the reasons discussed earlier, this case was pitched at a particular standard - whether a group costs order would provide a better outcome than relevant alternative funding arrangements. The plaintiff's primary measure of appropriateness has answered the statutory question (but not favourably to the plaintiffs) and in this particular case general considerations of reasonableness and proportionality, that might assume greater significance in other applications of this kind, are not sufficient to displace the conclusions following from the plaintiff's primary point.

144 It is appropriate however, to make some brief observations about those factors and the evidence addressed to them.

Proportionality and reasonableness - legal context

145 The plaintiffs and contradictor submitted that proportionality is a measure that should inform the fixing of the percentage rate for a group costs order. The defendants submitted that the prospect of "windfall" returns to the law practice must be avoided. The profit return to a law practice under a group costs order does not appear on its face to be properly a matter affecting the interests of the defendants. That said, proportionality and reasonableness are more legally significant touchstones for assessing the appropriateness of a proposed percentage rate, and likely better express the concerns underlying objections to potential "windfalls". They are not, of course, substitutes for the statutory test, but will assist in answering the statutory question.

146 Proportionality is a measure that is embedded in the legal context within which s 33ZDA is to be construed, including because of the overarching obligations that apply to parties and legal practitioners to use reasonable endeavours to ensure that legal costs incurred in civil proceedings are reasonable and proportionate to the complexity or importance of the issues in dispute and the amount in dispute.[49] The overarching obligations are not directly concerned with inter partes relations but are obligations owed to the Court.

147 Proportionality is a measure of the relationship between things. Section 24 of the Civil Procedure Act 2010 (Vic) directs attention to the relationship between costs and the issues and amount in dispute. Section 33ZDA, as noted earlier, engages with risk and reward, in that legal costs calculated as permitted under the section may reward the legal practice not only for the effort they contribute in legal work, but for the risk they accept in funding the proceedings and assuming obligations in respect of adverse costs. It therefore invites the question whether the reward proposed is (among other things) proportional to the risk to be undertaken.

148 The question whether the return to the law practice under a group costs order is or is likely to be reasonable and whether it bears a proportionate relationship to the assumption of risk or to any other relevant measure, may be considered prospectively, but with real limitations on the Court's ability to make an informed assessment. That is where sub-s 33ZDA(3) assumes significance. A review under that sub-section, of a percentage fixed at an earlier time, once information informing questions of proportionality is available, will facilitate the Court ensuring that the percentage to which the law practice is ultimately entitled, remains appropriate. Such a review might be informed by the Court having regard to the practitioners' obligations under s 24 of the Civil Procedure Act.

Proportionality and reasonableness - learnings from other jurisdictions

149 In the course of making an order under s 33ZF for the distribution of moneys paid under a settlement and making a "common fund order" fixing a funders' commission rate to be shared between group members at the conclusion of proceedings, Beach J said in Blairgowrie (No 3) that:

Whether a court should set a commission rate and the rate to be used is largely a forensic question depending upon the material available to the judge at the time the order is sought. ... [J]udges set legal costs by scales, rates, individual amounts and total or capped amounts, whether ex ante or ex post; a commission or funding rate may be seen as a relevant analogue. Further, some judges in fixing remuneration of external insolvency practitioners also readily engage in such exercises, including ex ante or ex post rates of remuneration. In other contexts, judges have set rates of remuneration for trustees administering trust assets. In yet other contexts, judges set discount rates on some aspects of common law damages. They may also set less than statutory interest rate entitlements. In general, the question is not whether the rate setting for a common fund order is a suitable subject matter for the exercise of judicial power. Rather, the question is whether, in a particular case, a judge is in a position to or should set a rate by the application of the appropriate judicial method tailored to the circumstances of the individual case.[50]

150 Not all of the reasoning applicable to the approach that a court should or might take when seeking to distribute the burden of a privately agreed third party funding commission will apply in this context. Section 33ZDA presents a different task to the Court, and the standard of evidence required to assess any particular application will depend upon its facts. Nevertheless, the work of courts in that context is instructive, and as this passage reminds, there are numerous contexts in which courts are called on to fix remuneration, including where the criteria for doing so is not specified in legislation, and where an evaluative assessment is required.[51]

151 In the context of common fund orders, the Federal Court has considered and identified those considerations relevant to fixing a commission rate for a litigation funder at the end of a class action, which include the litigation risks of funding provided in the proceeding, which must be assessed avoiding the risk of hindsight bias and recognising that the funder took on those risks at the commencement of the proceeding; the quantum of adverse costs exposure that the funder assumed at the commencement of the proceeding; the legal costs expended and to be expended, and the security for costs provided by the funder; a comparison of the funding commission with commissions in other group proceedings and what is available or common in the market by reference to broad parameters; pre-existing funding arrangements including class members' likely recovery in hand; and the amount of any settlement or judgment.[52]

152 On the latter consideration, the Full Court said in Money Max that it will be important to ensure that the aggregate commission received is proportionate to the amount sought and recovered in the proceeding and the risks assumed by the funder.[53] In relation to the commission rates offered in the "market", the Federal Court has tended to regard as relevant whether the funding commission rates sought in common fund orders were within the broad parameters of rates available and "towards the middle of the range of rates offered or accepted by funders for class actions in Australia".[54]

153 Some of the factors discussed in that context can only be assessed with any measure of certainty at the conclusion of the litigation. In the context of a GCO, facts that can only be ascertained with any certainty at the end of a proceeding may be considered in adjusting the rate fixed earlier, under the explicit power to do so in sub-s 33ZDA(3). The capacity to adjust a rate under that sub-section does not mean, of course, that the Court must not be sufficiently satisfied at the time at which a GCO is sought, that it is appropriate or necessary to ensure that justice is done in the proceeding.

154 Brief mention may be made of other jurisdictions. Lawyers' fees assessed as a proportion of total recovery have long been in place in Canada and the United States. In each jurisdiction class counsel are permitted to make an application to the Court to approve payment of fees assessed as a percentage of recovery as previously agreed with their clients. In approaching these applications, the courts have developed a generally consistent set of inquiries for assessing the appropriateness of the order sought, expressed as adequately remunerating counsel for their time, skill and expertise, and avoiding "windfalls" which might bring the legal profession into disrepute. Chief amongst these inquiries is the amount of the recovery and the risk or uncertainty associated with the litigation, though the courts also consider other factors such as the time expended, the legal complexity of the matter, the importance of the matter to the client or class, the monetary value in issue, the degree of skill and competence demonstrated by class counsel and their professional standing, and the opportunity cost to class counsel in taking on this matter rather than others.[55]

155 Judges deciding those cases have acknowledged that the inquiry is somewhat unscientific; the fact of the longstanding use of percentage based fees in these jurisdictions has however allowed the emergence of something of a "market", in which the reasonableness of the rate sought may be assessed laterally by reference to the percentages approved in similar proceedings. Empirical reviews of these orders have found that historically courts determined an award of 20% to 30% to be an appropriate benchmark, but courts have maintained this benchmark does not displace a close analysis of the particular case in all its circumstances.[56]

156 In the United Kingdom, "damages based fees" are rarely used,[57] and the only opt-out group procedure in that jurisdiction expressly prohibits their use.[58]

157 The plaintiffs assembled and analysed historical rates in litigation funding and legal fees from public material. That analysis was directed to the plaintiffs' primary case, which was that the proposed group costs orders would deliver a better outcome to group members than they would obtain via third party litigation funding. Otherwise, the rates available in the "market" were proffered as an indicator of the reasonableness of the proposed rate of 25%.

158 The analysis was undertaken by the plaintiffs' expert, Mr Houston. The analysis of data in that way was convenient and helpful, and Mr Houston employed technical skill in his mathematical or basic statistical analysis of the data. Although nothing turns on it, it should be remarked that that work was not the expression of an opinion. The opinion that Mr Houston did proffer, namely that "an appropriate methodology for arriving at an appropriate rate for a costs inclusive GCO is one that has regard to historical rates of funding commissions ... and historical legal fees in previous representative proceedings involving litigation funders", was not admissible. The appropriate methodology is a matter for the court, to be informed by statutory construction, legal principle, and the evidence. Although Mr Houston, an economist, has been briefed in numerous class actions to address questions concerned with remuneration, his qualifications do not permit him to advance an opinion on such a question.

159 Turning to the question of historical data, there are a number of publicly available data sets, including data compiled by the Australian Law Reform Commission and the Law Council.

160 The ALRC's data was compiled from judgments in cases decided by the Federal Court under Part IVA of the Federal Court Act in relation to proceedings finalised by judgment or settlement approval. The data set contains 104 cases finalised during the period 1997 to 2016. Separately, the Law Council compiled a set of data for the period 2001 to 2020 which it described as a summary of most of the Australian class action proceedings resolved in the period 2001 to 2020 and the distribution of settlement amounts among class members, legal representatives and third party funders.[59] The Law Council noted that it had not been possible to capture all class action settlements in Australia with all of the required information during the relevant timeframe. However, the Council considered the information to be of value because it identified most settlement sums, legal costs and funding commissions where actions have proceeded on a third party funding basis. The results did not include cases settled for no compensation or those that had failed. The Council said that the work was not intended to be scientifically rigorous.[60]

161 Mr Houston analysed and combined those data sets. He removed cases that appeared in both sets and those that did not disclose a funding fee or because a litigation funder was not involved, and identified a further ten cases from his own searches. His final data set compiled in that way comprised 61 cases of which 37 were shareholder class actions, and six were product liability or consumer class actions. Mr Houston noted that the available data reflects rates of commissions approved at the end of the process, with the benefit of after the fact assessments, and that the commission rates might have been adjusted by courts on settlement approval, but there was no evidence about that. Mr Houston said that there were intrinsic challenges associated with the identification, collection and verification of those data. Many of the significant items were kept confidential. There were some discrepancies between the ALRC and Law Council data. Mr Houston attempted to verify the reported numbers by reference to published court orders and judgments, but in many instances he was unable to verify one or more components of the settlement. Where a discrepancy was identified, Mr Houston adopted the minimum rate of funding commissions identified in conflicting sources.

162 Subject to those caveats, Mr Houston reported that:

(a) 50% of the commissions in that data set fell within the range 21% to 29% of the gross settlement amount;

(b) the median rate was 25%, while the mean was 24%. The mean was lower than the median because of a small number of very low commission rates;

(c) 75% of commission rates in the sample were at or above 21%, and 75% of commission rates were at or below 29%;

(d) the data set was not sufficiently large to detect any systemic differences by reference to the subject matter of the actions.

163 On the basis of that analysis, Mr Houston concluded that the range of 21% to 29% was broadly reflective of the market rates for the remuneration of funding services. As Mr Houston put it, a commission of 21% to 29% is a rate at which funders are generally prepared to put forward their capital and which has then been accepted by plaintiffs and endorsed by the courts.

164 From the data set Mr Houston further calculated the combined proportion of any award or settlement that allowed for both legal costs and litigation funding commissions by adding together those two data points. The data showed that the median proportion of an award or settlement deducted by funders in respect of legal and funding fees was 46%, while the mean was 45%. The interquartile range of combined legal and litigation funding fees in historical proceedings on that data set was 36% to 55%.

165 I accept that that analysis, acknowledging its limitations, provides a meaningful indicator of the reasonableness of the proposed rate, by comparison with historical results achieved in third party funded cases. By that measure, a return to group members of 75% is evidently significantly better than a return of somewhere around 45%. However, as the applications were put, the question was not simply whether the proposed rates were "reasonable". The comparative analysis was directed to the plaintiffs' proposition that that third party funding was the appropriate comparator against which to assess the existing arrangements, a contention that I have rejected.

166 As the defendants would have it, reference to "market rates" for third party funding is an insufficient basis against which to assess the proposed GCO rates. The relevant touchstones for the evaluative assessment required under s 33ZDA are best considered in the context of particular facts, and the significance of the evidence going to the historical results in funded litigation need not be considered further in these applications.

167 The plaintiffs submitted that although estimating a prospective return on investment was problematic, including because the inputs to the assessment were uncertain, such assessment could properly serve as a check on the reasonableness of the proposed percentage range.

168 The plaintiffs' analysis is considered in the confidential schedule. I accept that, generally speaking, it is a relevant consideration in this context whether the rate at which a group costs order is sought appears proportional to the risks being assumed by the law practice. The utility of that consideration will depend upon the available evidence. Other models might be developed in other cases, for example an insurance-based actuarial calculation might assist in assessing why a proposed return is likely to be reasonable for an investor with the particular funder's characteristics, in the circumstances.

169 The plaintiffs' analysis, based on the assumptions and limitations identified, suggests that there is nothing inherently unreasonable in the proposed GCO rate, considered from that perspective. For clarity, the modelling appeared to assume that the return on investment was solely comprised of the amount fixed under group cost order. Any costs recovery against the defendant would, I infer, be offset against the liability of the plaintiff and group members to pay Maurice Blackburn's costs, but that point was not discussed. This is not the occasion for the further scrutiny of that evidence.

170 As with the other factors discussed here, the fact that the rate at which a GCO is sought might appear to be proportional to the risks to be assumed or otherwise to be reasonable does not answer the question posed by the particular circumstances of this case and the framing of the applications that resulted from those circumstances.

171 The plaintiffs elicited from Mr Houston what might be described as an opinion in relation to efficiency considerations embedded in the group costs order funding model, as follows.

172 Mr Houston was of the opinion that in a representative proceeding the law practice acting for the plaintiff and group members has a significant degree of autonomy to make both strategic and day-to-day decisions on behalf of group members in relation to the conduct of a proceeding. In economics, those characteristics describe a principal-agent relationship. Economic principles show that optimal outcomes are more likely to be achieved when the principal-agent arrangements are structured so as to align the agent's economic interests with that of its principal so that the actions taken by the agent are more likely to be in the mutual best interests of both parties. The best interests of group members are likely to be served when the proceedings are managed in a timely, cost-effective manner.

173 Both the group costs order and NWNF models offer strong financial incentives for the law practice to seek a successful outcome on behalf of the plaintiff and group members. Under both funding models, the law practice does not receive any compensation for its legal services in the event of an unsuccessful outcome. From an economic perspective, this brings the incentives of the law practice (agent) in line with that of the group (principal).

174 However, the group costs order model aligns the economic interests of the parties more precisely in that remuneration of the law practice moves together with that of the group in respect of all potential settlements or awards.

175 This was an opinion framed at a high level of generality and might reflect economic aspects of the policy that led to the introduction of s 33ZDA. In this case though, given the evidentiary basis of the application, it does not materially assist the case for the exercise of the discretion.

176 The Fox plaintiffs' comparative analysis has encountered two difficulties: one concerning what the modelling appears to show, and one concerning the uncertainties inherent in that modelling. For the reasons set out, the evidence does not support the application for the proposed group costs order.

177 I have considered whether making a group costs order at a different rate than the rate proposed would meaningfully change the analysis, resulting in the conclusion that a GCO at a different rate would be appropriate or necessary to ensure that justice is done in the proceeding.

178 I am inclined to the view that section 33ZDA would permit me to make a group costs order at a different rate than the one sought, but as a matter of fairness, not without permitting the plaintiff in particular (and the other parties to the extent appropriate) to make submissions on any alternative rate. The plaintiffs did not propose a GCO at an alternative rate.

179 Employing the "tipping point" methodology an adjusted rate might be set that, by the plaintiffs' modelling, would show an advantage to group members. I have set out a brief illustration of that point in the confidential schedule.

180 However, that would deal with the first difficulty but not the second. The modelling itself is limited by its inputs.

181 It may be that some of the uncertainties that impact on the modelling can be resolved as the proceeding progresses. The most significant sources of uncertainty are identified in Mr Watson's evidence, as set out in the confidential schedule. As some uncertainties are resolved, the inputs to the modelling will likely change, producing different results.

182 The plaintiffs submitted that a group costs order transfers from group members to the law practice the risk of a small positive settlement or award and the risk that costs will erode the recovered sum, and those risks are incapable of being transferred at the conclusion of the proceeding because at that point the risks no longer exist - they have crystallised - either materialising or evaporating. I accept that the transfer of risk that the group costs order model contemplates may become less relevant or meaningless as time progresses. Whether or not making a GCO would be necessary or appropriate at a later stage (including at the end of the proceeding) would have to be assessed on the evidence presented at that time.

183 Both the plaintiffs and the contradictor submitted that in the event that I was not persuaded to make a group costs order at the proposed rate at this time, I should adjourn the application to permit the plaintiffs to further consider their position, and specifically whether a reformulated application should be pressed at a later time. I consider that to be an appropriate course. I take that view because, although for the reasons set out the present NWNF funding arrangements are not, contractually speaking, interim arrangements, they do expressly contemplate that the plaintiffs may apply for a GCO (a fact of which group members were informed), but also because these are the first applications of their kind. In framing their applications the plaintiffs did not have (but now have) the benefit of these reasons. The Fox and Crawford proceedings became the first cases in which the application of s 33ZDA was to be considered because they happened to be the first cases ready to proceed, and not because the plaintiffs formulated them as test cases. I will hear submissions in due course as to the time-frame for the adjournment of the plaintiffs' summons.

184 Much of my reasoning set out above (which I will not repeat here) applies to the Crawford proceeding, with some modification.

185 There is one important difference. In contrast to the position in the Fox proceeding, for reasons set out in the confidential schedule, the predictive modelling on its face, could be said to support a group costs order at 25%.

186 That conclusion is subject to the sufficiency of the evidence supporting the conclusions in the modelling which, as explained in the confidential schedule, was given only in a short-hand way in this case, without the more detailed and specific reasoning provided in the Fox proceeding in relation to the designation of particular probability assessments.

187 However, the same uncertainties are embedded in the modelling applicable to this case, as set out above. The uncertainties are significant in this case for the same reasons explained above in the context of the Fox proceeding, noting that the applications were put on the same basis, and the same contractual position applies to both proceedings.

188 The same analysis set out in respect of the Fox proceeding concerning the significance of the other factors, also applies to this case.

189 There is a better basis then, in this case, for the making of the proposed GCO subject to the limitations of the evidence of the reasoning supporting the outcome modelling. The fundamental uncertainty that affects the inputs to the modelling remain, however.

190 Both the plaintiffs and the contradictor submitted that in the event that I was not persuaded to make a GCO at the proposed rate at this time, I should adjourn the application to permit the plaintiffs to further consider their position, and specifically whether a reformulated application should be pressed at a later time. I consider that to be an appropriate course. I take the view for the same reasons applicable to the Fox proceeding and will hear submissions in due course as to the time-frame for the adjournment of the plaintiffs' summons.

[1] It is also alleged that the conduct for which the defendants are responsible was misleading or deceptive conduct under s 1041H of the Corporations Act 2001 (Cth) or s 12DA of the ASIC Act 2001 (Cth) and that, for similar reasons, the payments by the plaintiffs and group members were vitiated by an actionable mistake. The plaintiffs seek damages, and restitution.

[2] By commencement of s 2 of the Justice Legislation Miscellaneous Amendments Act 2020.

[3] Practice Note SC Gen 10 Conduct of Group Proceedings (Class Actions), s 14.1.

[4] Mr Bernard Quinn QC, who appeared with Ms Eugenia Levine.

[5] Project Blue Sky Inc v Australian Broadcasting Authority (1998) 195 CLR 355, [69]; Interpretation of Legislation Act 1984 (Vic), ss 35(b)(iii) and (iv).

[6] Commissioner of Taxation v Consolidated Media Holdings Ltd [2012] HCA 55; (2012) 250 CLR 503 at [39] (French CJ, Hayne, Crennan, Bell and Gageler JJ).

[7] The descriptor is used for convenience; it does not appear in the text.

[8] See, e.g., Oshlack v Richmond River Council [1998] HCA 11; (1998) 193 CLR 72, [67] (McHugh J), [134] (Kirby J).

[9] Section 33ZDA takes effect despite s 183 of the Legal Profession Uniform Law (Vic), which provides that "a law practice must not enter into a costs agreement under which the amount payable to the law practice or any part of that amount, is calculated by reference to the amount of any award or settlement or the value of any property that may be recovered in any proceedings to which the agreement relates". Under that Law in Victoria and in similar legislation in other states conditional billing (where the lawyer's fee only becomes payable to the lawyer if a successful outcome is achieved), and uplift fees (being an additional fee payable on success) are permitted: Legal Profession Uniform Law (Victoria and NSW) s 181 and 182; Legal Profession Act 2006 (ACT) ss 283 and 284; Legal Profession Act (NT) ss 318 and 319; Legal Profession Act 2007 (Qld) ss 323 and 324; Legal Practitioners Act 1981 (SA) ss 25 and 26 of Schedule 3, as applied by s 41; Legal Profession Act 2007 (Tas) ss 307 and 308; Legal Profession Act 2008 (WA) ss 283 and 284. See also Clyne v Bar Association (NSW) [1960] HCA 40; (1960) 104 CLR 186, 203 in which the High Court held that it was perfectly proper for a lawyer to provide representation on a conditional basis.

[10] See BMW Australia Ltd v Brewster; Westpac Banking Corporation v Lenthall [2019] HCA 45; (2019) 374 ALR 627 (Brewster HCA), [124] (Nettle J).

[11] Brewster HCA, [88] (Kiefel CJ, Bell And Keane JJ).

[12] The same expression is found in s 183(1) of the Uniform Law.

[13] Victorian Law Reform Commission, Access to Justice - Litigation Funding and Group Proceedings (Report, March 2018) (VLRC Report).

[14] Explanatory Memorandum, Justice Legislation Miscellaneous Amendments Bill 2019, clause note 5.

[16] Victoria, Parliamentary Debates, Legislative Assembly, 28 November 2019 at 4586, 4590.

[17] See Brewster HCA at [110] per Gageler J, in the context of the federal counterpart to Part 4A, Part IVA of the Federal Court of Australia Act 1976 (Cth).

[18] Brewster HCA, [123] (Nettle J), and the examples there cited.

[19] Section 33ZF of the Federal Court Act is in the same terms as s 33ZF of the Supreme Court Act 1986 (Vic) and shares the criterion "appropriate or necessary to ensure justice is done in the proceeding" with s 33ZDA.

[20] See, eg, the discussion in Thomas v Mowbray [2007] HCA 33; (2007) 233 CLR 307 at [79] and [91] (Gummow and Crennan JJ), and the statement there approved of Professor Leslie Zines, now in Stellios, Zines's The High Court and the Constitution (6ed, Federation Press, 2008) 258; see also Mitchell v The Queen [1996] HCA 45; (1996) 184 CLR 333, 346-347.

[21] Brewster HCA, 657, [123] (Nettle J).

[22] Owners of the Ship 'Shin Kobe Maru' v Empire Shipping Company Inc (1994) 181 CLR 404, 421.

[23] Being s 33ZF of the Federal Court Act and s 183 of the Civil Procedure Act 2005 (NSW); see Brewster HCA, [69], [70], [145], [170].

[24] As Beach J observed of s 33ZF of the Federal Court Act in Earglow Pty Ltd v Newcrest Mining Ltd [2015] FCA 328; (2015) 230 FCR 469, [33], the requirement that an order be "appropriate" may introduce a lower threshold than a requirement that it be "necessary", but in the expression, "appropriate or necessary to ensure that justice is done", both are coupled with the words, "to ensure".

[25] See Kiefel CJ, Bell and Keane JJ in Brewster HCA at [49] to the effect that orders contemplated by s 33ZF of the Federal Court Act are orders which may be thought to make certain justice is done in the proceeding.

[26] Money Max Int Pty Ltd v QBE Insurance Group Ltd [2016] FCAFC 148; (2016) 245 FCR 191 (Money Max).

[27] Money Max, [161] citing the New South Wales Court of Appeal in Fairfax Digital Australia and New Zealand Pty Ltd v Ibrahim [2012] NSWCCA 125; (2012) 83 NSWLR 52 to the effect that the word "necessary" can have shades of meaning and "admits of all degrees of comparison". See also Brewster v BMW Australia Ltd (2019) 343 FLR 176, where at [10] the Full Court of the Federal Court held that s 33ZF was capable of supporting the making of a "common fund order" at an early stage in a group proceeding. The High Court rejected that view on appeal, however, the High Court's reasons did not turn on any different view about the construction of the expression, "appropriate or necessary to ensure the justice is done in the proceeding".

[28] McCulloch v State of Maryland [1819] USSC 5; 17 US 316, 414 (McCulloch); cited in Thomas v Mowbray [2007] HCA 33; (2007) 233 CLR 307, 353 and Money Max, [162].

[29] McCulloch, 414; Money Max [161]-[162] and the passages there cited.

[31] Wigmans v AMP Limited [2021] HCA 7 [116] (Gageler, Gordon and Edelman JJ).

[32] Mobil Oil Australia Pty Ltd v Victoria (2012) 211 CLR, [21] (Gleeson CJ).

[34] Blairgowrie Trading Ltd v Allco Finance Group Ltd (recs and mgrs apptd) (in liq) [2015] FCA 811; (2015) 325 ALR 539, [112]-[114].

[36] Commissioner of Taxation v Consolidated Media Holdings Ltd [2012] HCA 55; (2012) 250 CLR 503 at [39] (French CJ, Hayne, Crennan, Bell and Gageler JJ).

[37] Owners of the Ship 'Shin Kobe Maru' v Empire Shipping Company Inc (1994) 181 CLR 404, 421.

[38] It was implicit but not certain, that those figures took into account recovery of costs against the defendants. These figures were calculated by Mr Houston in research set out in his reports submitted as evidence for the plaintiffs (discussed below). They are broadly in line with the results of a similar survey carried out by the Australian Law Reform Commission of Federal Court group proceedings in its report Integrity, Fairness and Efficiency - An Inquiry into Class Action Proceedings and Third-Party Litigation Funders (Report 134, December 2018), which found at a median return to group members in funded class actions in the Federal Court between 2013 and 2018 of 51%.

[39] See, in the context of s 33ZF, Gordon J's observation in Brewster HCA, at [151].

[40] Clause 12.3 provides that if there is a termination under clause 12.1 and there is a successful outcome after the date of termination, legal costs are payable for legal work performed prior to the date of termination. There are other termination provisions that are not presently relevant.

[41] The position is further set out in the confidential schedules for each proceeding.

[42] The tipping point for a 20% GCO may be calculated in the same way, but dividing the result by 0.2, and so on.

[43] See Wigmans v AMP Limited (2021) 388 ALR 272; [2021] HCA 7, [77] (Gordon, Gageler and Edelman JJ).

[44] See Wigmans v AMP Limited (2021) 388 ALR 272; [2021] HCA 7.

[45] Project Blue Sky Inc v Australian Broadcasting Authority (1998) 195 CLR 355, [69].

[46] Victoria, Parliamentary Debates, Legislative Assembly, 28 November 2019, 4590; Explanatory Memorandum, Justice Legislation Miscellaneous Amendments Bill 2019, 2-3.

[47] Greater Shepparton City Council v Clarke [2017] VSCA 107; (2017) 56 VR 229, [64] (the Court).

[48] Carr v Western Australia (2007) 232 CLR 128, [5].

[49] Civil Procedure Act 2010 (Vic) s 24.

[50] Blairgowrie Trading v Allco Finance (No 3) [2017] FCA 330 (Beach J). The decision in Blairgowrie (No 3) was delivered before the High Court's decision in Brewster HCA. On the subject of post settlement common fund orders, see Brewster v BMW [2020] NSWCA 272 (Bathurst CJ, Bell P and Payne JA); Davaria Pty Ltd v 7-Eleven Stores Pty Ltd [2020] FCAFC 183; (2020) 384 ALR 650 (Middleton, Moshinsky and Lee JJ).

[51] See also Wigmans HCA, [115]; Asirifi-Otchere v Swann Insurance (Aust) Pty Ltd (No 3) (2020) 385 ALR 625; [2020] FCA 1885 (Asirifi-Otchere), [23] and [24] (Lee J).

[52] Money Max at [80]; see also Asirifi-Otchere at [21] (Lee J).

[54] See, for example, Blairgowrie (No 3) and Asirifi-Otchere at [25].

[55] See, eg, Brown v Canada 2018 ONSC 3429, [39]-[41] (Belobaba J); Goldberger v Integrated Resources Inc, [2000] USCA2 112; 209 F3d 43 (2d Cir 2000), 47.

[56] See M Legg, 'Contingency fees - Antidote or poison for Australian civil justice?' in (2015) 39 Australian Bar Review 244, 268; see also Cannon v Funds for Canada Foundation 2013 ONSC 7686 (Belobaba J).

[57] United Kingdom Ministry of Justice, Post-Implementation Review of Part 2 of the Legal Aid, Sentencing and Punishment of Offenders Act 2012 (LASPO) - Civil litigation funding and costs, February 2019 [11].

[58] Competition Act 1998 (UK), s 47C(8).

[59] Law Council of Australia, Submission to the Parliamentary Joint Committee on Corporations and Financial Services, Litigation Funding and the Regulation of the Class Action Industry (16 June 2020), 7.