Morabito, Vince --- "Class Actions against Multiple Respondents" [2002] FedLawRw 11; (2002) 30(2) Federal Law Review 295

On one reading of the approach by some Judges of the Federal Court to complex cases involving more than one respondent, particularly where the conduct in question involves a complicated or lengthy series of transactions, those cases should simply not be brought as class actions. If that reading is correct, then to put it bluntly, defendants are more likely to escape liability if by their conduct they cause harm or loss to more people over a greater period of time, and if they do so in concert with others.^[1]
Class actions are unlikely to flourish in Australia without a change in attitudes, both in the profession and the judiciary ... Many Australian judges may still view class actions as predatory litigation instituted for the advantage of entrepreneurial lawyers.^[2]

In March 2002 the class action regime introduced by the Commonwealth Parliament, through the addition of a new Part IVA to the Federal Court of Australia Act 1976 (Cth),^[3] celebrated its tenth 'birthday'. During this time, a number of important issues, concerning the interpretation and application of Part IVA, have emerged.^[4] One such issue has been the circumstances in which a proceeding may be brought, in pursuance of the provisions of Part IVA, against more than one respondent ('multiple respondents suits').^[5] It is the aim of this article to assess the operation of the Part IVA regime with respect to such suits.

The picture that will emerge from this analysis is a largely unsatisfactory one as it will reveal (a) conflicting decisions of the Federal Court and, consequently, uncertainty as to the precise ambit and impact of applicable principles; and (b) the existence of significant and unnecessary barriers to the availability of the Part IVA regime in relation to multiple respondents suits.

In Section II below, the relevant provisions of Part IVA will be summarised. Sections III, IV and V will explore the impact on multiple respondents suits of the judicial interpretation of the three requirements which, according to s 33C(1), must be complied with before aspiring representative plaintiffs may avail themselves of the Part IVA regime.^[6] A summary of the author's conclusions is provided in Section VI.

II THE PART IVA REGIME

The central provision of Part IVA is s 33C(1) as it sets out the conditions which need to be satisfied in order to activate the Part IVA regime.^[7] It provides that where

(a) 7 or more persons have claims against the same person;^[8] and
(b) the claims of all those persons are in respect of, or arise out of, the same, similar or related circumstances; and
(c) the claims of all those persons give rise to a substantial common issue of law or fact;
a proceeding may be commenced by one or more of those persons as representing some or all of them.

In light of the importance of the three requirements set out above, each will be considered in Sections III, IV and V below respectively.

Section 33D(1) addresses the standing of the applicant to commence a class action on behalf of the represented group:

(1) A person referred to in paragraph 33C(1)(a) who has a sufficient interest to commence a proceeding on his or her own behalf against another person has a sufficient interest to commence a representative proceeding against that other person on behalf of other persons referred to in that paragraph.

Section 33C(1) reflects a deliberate decision by the Commonwealth Parliament not to impose on potential class representatives the requirement of seeking leave from the Federal Court before they may proceed with a class suit under Part IVA. This rejection of a mandatory certification requirement^[9] was based upon the recommendations of the Australian Law Reform Commission ('ALRC').^[10] However, a properly commenced class proceeding, that is, a proceeding which complies with all three requirements of s 33C(1), may be discontinued, as a Part IVA suit, by the Federal Court if one of the scenarios specified in ss 33L,^[11] 33M^[12] or s 33N is found to exist. Section 33N constitutes the most significant source of the Federal Court's power to terminate a Part IVA suit. It allows the Court, on application by the respondent or of its own motion, to order that the action no longer continue as a class action where

it is satisfied that it is in the interests of justice to do so because:
(a) the costs that would be incurred if the proceeding were to continue as a representative proceeding are likely to exceed the costs that would be incurred if each group member conducted a separate proceeding;^[13] or
(b) all the relief sought can be obtained by means of a proceeding other than a representative proceeding under this Part; or
(c) the representative proceeding will not provide an efficient and effective means of dealing with the claims of group members; or
(d) it is otherwise inappropriate that the claims be pursued by means of a representative proceeding.^[14]

Section 33P provides that where the Court orders the discontinuance of a class suit, under ss 33L, 33M or 33N, the proceeding may be continued as a proceeding by the representative party on his/her own behalf and 'on the application of a person who was a group member for the purposes of the proceeding, the Court may order that the person be joined as an applicant in the proceeding'. O'Loughlin and Drummond JJ of the Federal Court have explained that:

where the Court determines that a proceeding commenced under Part IVA is not authorised by s 33C to be commenced as a representative proceeding or where it orders, under any of ss 33L, 33M or 33N, that a representative proceeding no longer continue under Part IVA, the effect of such a determination is to extinguish the representative aspect of the proceeding but to leave unaffected its character as a proceeding brought for the benefit of the representative party itself. The implication in each of ss 33L(b), 33M(c) and 33N(1) is that once an order terminating the representative character of a proceeding brought under Part IVA is made, that same proceeding continues, but as an action brought by and for the sole benefit of the representative party. What is implied in these provisions is expressly stated in s 33P(a).^[15]

Section 33H(1) outlines the requirements with respect to the originating process in Part IVA proceedings. The pleadings must describe or otherwise identify the group members to whom the proceeding relates, specify the nature of the claims made on behalf of the group members and the relief claimed as well as the questions of law or fact common to the claims of the group members. Section 33H(2) makes it clear, however, that in describing, or otherwise identifying group members, it is not necessary to name, or specify the number of, group members.^[16] Section 33H(2) reflects the fact that an opt out mechanism was selected by the Commonwealth Parliament.^[17]

III SEVEN OR MORE PERSONS MUST HAVE A CLAIM AGAINST THE SAME PERSON

In Philip Morris (Australia) Ltd v Nixon,^[18] Sackville J of the Federal Court (with whose reasons Spender and Hill JJ agreed) enunciated the following principle ('the Philip Morris principle') concerning the impact of s 33C(1)(a) upon multiple respondents suits:

s 33C(1)(a) requires every applicant and represented party to have a claim against the one respondent or, if there is more than one, against all respondents. This conclusion follows from the language of s 33C(1)(a) itself and is consistent with the approach taken by the [ALRC] in Grouped Proceedings. It is also consistent with the structure of the legislation. For example, s 33D(1)(a) (which provides that a person who has a sufficient interest to commence a proceeding on his or her own behalf against another person has a sufficient interest to commence a representative proceeding against that person on behalf of other persons referred to in s 33C(1)(a)) is clearly drafted on the assumption that all applicants and represented persons will have claims against the same person.
It follows that s 33C(1)(a) is not satisfied if some applicants and group members have claims against one respondent (or group of respondents) while other applicants and group members have claims against another respondent (or group of respondents). The requirement in s 33C(1)(b) that the claims of all group members are in respect of or arise out of the same, similar or related circumstances, is a necessary but not sufficient condition for the commencement of representative proceedings. Of course, if there are two sets of claims against two sets of respondents, it may well be that each can be the subject of representative proceedings. It may even be that directions can be made for them to be heard together: Ryan v Great Lakes Council (1997) 149 ALR 45, at 48, per Wilcox J. But they cannot both be the subject of the same representative proceedings.^[19]

In the remainder of this section an analysis of the cogency and the impact of the Philip Morris principle will be provided. In particular, the appropriateness of Sackville J's reliance on the ALRC's report,^[20] which led to the enactment of Part IVA, will be considered (see Section IIIA below). This will be followed by a study of whether the Philip Morris principle is, in fact, in accordance with the intention of the Commonwealth Parliament (see Section IIIB below). The approach to the application of s 33C(1)(a) in multiple respondents suits, dictated by the Philip Morris principle, is then contrasted with the approach adopted by the Federal Court before Philip Morris and by courts in British Columbia and Ontario (see Section IIIC and Section IIID below, respectively). Section IIIE will canvass the availability of other procedures, such as joinder and consolidation, in relation to Part IVA proceedings affected by the Philip Morris principle. Finally, in Section IIIF it will be submitted that the Philip Morris principle may, and should, be applied in a less strict fashion than what occurred in Philip Morris itself.

A The grouped proceedings proposal of the ALRC

Part IVA was based upon the grouped proceedings model recommended by the ALRC in 1988.^[21] As the passage from Sackville J's judgment set out above indicates, his Honour placed significant reliance on the views of the ALRC in reaching the conclusion that s 33C(1)(a) requires every representative plaintiff and every group member to have a claim against every respondent. Sackville J was of the view that the ALRC's

recommendations were specifically designed to provide an effective procedure to enable people suffering loss or damage in common with others as a result of a wrongful act or omission by the same respondent (par 69, 95, 133). It therefore plainly did not envisage that the grouped procedure could be employed to bring a proceeding against more than one respondent, in circumstances where some members of the group make a claim against one respondent only and others make a claim against another respondent.^[22]

Before considering the report of the ALRC in some detail, some preliminary comments, concerning Sackville J's views quoted above, are called for. His Honour was correct when he noted that the ALRC made numerous references to 'respondent' rather than 'respondents'. However, it is equally clear that the ALRC did envisage the commencement of grouped proceedings against more than one respondent. The ALRC furnished a number of examples of legal grievances which could be appropriately dealt with by the model proposed by the ALRC. Three of these examples, provided by the ALRC, involved causes of action against more than one respondent.^[23]

A central component of the model advocated by the ALRC was that the class representatives and the group members would all be parties to the proceeding. In the words of the ALRC:

all persons with relevant and related claims [would] be made parties to their own separate proceedings. This implies that a distinct proceeding be commenced on behalf of each member of the group even if they have different causes of action, claim different relief or rely on different bases of jurisdiction. The proceeding of each member of the group could be 'bundled together' and conducted by one of the group on behalf of all of them together with his or her own proceedings. The result would be that all such persons would be bound by the decision in a matter to which that person was a party.^[24]

The model proposed by the ALRC should be contrasted with the class action model, pursuant to which

one party commences an action on behalf of other persons who have a claim to a remedy for the same or similar perceived wrong. That party conducts the action as 'representative plaintiff'. Only the 'representative plaintiff' is a party. Other persons having claims that share questions of law and fact in common with those of the representative plaintiff are members of the 'class'. Once the class has been determined, the class members are bound by the outcome of the litigation even though, for the most part, they do not participate in the proceedings.^[25]

What the ALRC had in mind bears some similarity to the group litigation regime introduced in England in May 2000.^[26] In fact, under this new English system, 'each group litigant is a member of a procedural class as a party, rather than as a represented non-party'.^[27] However, one crucial difference between the English model and the ALRC's grouped proceedings model is that in the former model the requirement of making group members parties was the result of an evaluation of the relevant policy considerations.^[28] The decision of the ALRC, on the other hand, was based entirely on its concern that adherence to a class action model for the Federal Court—pursuant to which only the applicant would be a party on the record—would violate the Commonwealth Constitution.^[29] Obviously, the Commonwealth Parliament did not share the views of the ALRC, concerning the compatibility of a class action model with the Commonwealth Constitution, as it selected, for the Federal Court, a class action model.^[30]

Under the regime advocated by the ALRC, it was essential that, if a group action was initiated against more than one respondent, each member of the group was to have a claim against each respondent. Otherwise, the constitutional problem identified by the ALRC,^[31] and which dictated (in the ALRC's view) a grouped proceedings model, would re-emerge, rendering the relevant legislation unconstitutional in its application to multiple respondents suits. In light of the Commonwealth Parliament's rejection of the model proposed by the ALRC, the views of the ALRC on the issue of multiple respondents suits—which views were, as indicated above, totally dictated by the essential features of the model the ALRC was putting forward—can shed no light upon the crucial issue of what s 33C(1)(a) requires in multiple respondents suits.^[32] Consequently, Sackville J's reliance on the views of the ALRC, in respect of multiple respondents suits, was, with respect, inappropriate.^[33]

B That elusive intention of Parliament

The reasoning in favour of the Philip Morris principle that was embraced by the Federal Court essentially revolved around the terms of s 33C(1)(a) and s 33D(1).^[34] As was indicated by Wilcox J in Ryan v Great Lakes Council:

Subsection (1) of 33D may have been inserted out of an abundance of legislative caution. The standing it confers may be implicit in the closing words of s 33C(1) anyway; nonetheless, in confirming that standing, s 33D(1) also limits it. The subsection takes one of the seven or more claimants referred to in s 33C(1)(a) (the first person) whose individual interest is sufficient to support a proceeding brought by the first person against a particular person, and gives the first person the further entitlement to make claims on behalf of others against 'that other person'. The 'other person' is the person referred to earlier in s 33D(1) as 'another person', that is, the person against whom seven or more members, including the applicant, have claims.^[35]

It is submitted, however, that the terms of these provisions would not necessarily preclude the commencement of one Part IVA action against more than one respondent where in relation to each respondent there are at least seven persons, including a representative party, who are all claiming against that respondent.^[36] As is explained in Section IIIC below, this was essentially the approach adopted by the Federal Court before Philip Morris.

The objectives which the Part IVA device, and indeed any class action regime,^[37] seeks to attain were explained as follows in the Second Reading Speech of the Federal Court of Australia (Amendment) Bill 1991 (Cth), which introduced Part IVA:

The Bill gives the Federal Court an efficient and effective procedure to deal with multiple claims. Such a procedure is needed for two purposes. The first is to provide a real remedy where, although many people are affected and the total amount at issue is significant, each person's loss is small and not economically viable to recover in individual actions. It will thus give access to the courts to those in the community who have been effectively denied justice because of the high cost of taking action.
The second purpose of the Bill is to deal efficiently with the situation where the damages sought by each claimant are large enough to justify individual actions and a large number of persons wish to sue the respondent. The new procedure will mean that groups of persons, whether they be shareholders or investors, or people pursuing consumer claims, will be able to obtain redress and do so more cheaply and efficiently than would be the case with individual actions.^[38]

It is submitted that the existence of these objectives of fundamental importance dictates that a generous approach be taken to the construction and application of Part IVA. This point was made most effectively by the Supreme Court of Canada in Hollick v Toronto (City)^[39] in relation to Ontario's counterpart to Part IVA:

the legislative history of the Class Proceedings Act 1992 makes clear that the Act should be construed generously ... it is essential ... that courts not take an overly restrictive approach to the legislation, but rather interpret the Act in a way that gives full effect to the benefits foreseen by the drafters.^[40]

The Philip Morris principle is not in accordance with the desirable philosophy outlined above. This can be seen by exploring the impact which this principle has had upon (a) the class members in the Philip Morris litigation itself and (b) other multiple respondents suits that sought to rely on the Part IVA regime. The Philip Morris litigation involved claims by those suffering the ill effects of smoking and were made against Australian manufacturers and distributors of cigarettes. Sackville, Spender and Hill JJ delivered separate judgments but all agreed that the proceeding was not properly brought as a Part IVA suit because it failed to adhere to s 33C(1)(a) (as well as s 33C(1)(b) and s 33C(1)(c)). Section 33C(1)(a) had not been satisfied because, in the words of Sackville J, 'the statement of claim pleads that some applicants and group members have claims against one respondent, while others have claims against the other individual respondents'.^[41] Spender and Hill JJ (Sackville J dissenting) held that the applicants should not be permitted once again to replead the application as a Part IVA action. As is explained in Section IIIF below, a predominant reason for this ruling by Spender and Hill JJ was their conclusion that it would be impossible for the applicants to plead a case that would demonstrate that every applicant and group member had a claim against each respondent. Hill J also explained that

to allow a case to continue which involves investigation of the conduct of all the respondents over such a period of time [40 years] might well be productive of considerable injustice to them or some of them who will incur enormous costs in giving discovery in respect of matters which may not be relevant at all to any actual applicant. This difficulty would disappear if those applicants who have a genuine case bring individual proceedings where discovery and other interlocutory processes can be limited to what is actually alleged, rather than to what may hypothetically be alleged.^[42]

Hill J's comments, and his (and Spender J's) decision not to allow the applicants to replead their case, are, with respect, disappointing on both a conceptual and a practical level. Failing to provide the applicants with another opportunity to plead their case as a Part IVA proceeding by, instead, categorically holding that under no circumstances could the applicants in Philip Morris comply with s 33C(1)(a)^[43] as well as concluding that a more satisfactory scenario would ensue, if individual proceedings were commenced by the group members, demonstrate an inability, and/or an unwillingness, on the part of Hill and Spender JJ, to embrace the philosophy underlying class action regimes.^[44] The comments made by Goldberg J of the Federal Court, in a different context, are apposite here:

counsel for the applicants correctly pointed out ... [that the] group members ... will be free either to bring, or join in the bringing of, further proceedings against the respondent. This may be true as a matter of theory and principle but ... it fails to recognise the practical advantage of the current representative proceeding. The whole purpose of a representative proceeding is to enable persons with relatively small claims to have their claims pursued, where the cost of pursuing individual claims is impractical having regard either to the quantum or nature of the claim.^[45]

On a more practical level, the approach of Hill and Spender JJ displays an insufficient understanding of, and/or regard for, the circumstances of the class members. Individual proceedings were simply not an option for most, if not all, of the representative plaintiffs and group members.^[46] As was revealed to the High Court by Tobin QC, acting for the applicants, in Philip Morris:

there is no way that the tobacco litigation will continue as individual proceedings. It is financially impossible ... there is another reason ... and it is this ... for an individual plaintiff to take on the tobacco companies while suffering from a terminal disease is, in the lingo, a big ask, and it is not a realistic ask ... because individual plaintiffs, in these circumstances, are probably unlikely to see the outcome of their own litigation.^[47]

Kirby J of the High Court was entitled to lament that the group members in Philip Morris were 'put out of court under the provisions that are reformatory provisions which, at least on my understanding, were enacted for cases such as this'.^[48] If, on the other hand, individual proceedings were feasible, a state of affairs inconsistent with the 'judicial economy' goal of class actions would result. This was precisely the scenario witnessed in Ontario after the decision of the Ontario Court (General Division) not to certify a class proceeding in Sutherland v Canadian Red Cross Society^[49] concerning claims by persons who received HIV-infected blood as well as those cross-infected by the primary infected persons:^[50]

as a consequence, there are at least 84 individual tainted blood cases before the Ontario courts and another possible 300 cases forthcoming;^[51] and
the filing of individual claims resulted in three very long trials without significant plaintiff-wide settlements.^[52]

The adverse practical impact of the Philip Morris principle extends beyond preventing access to justice to smokers. To the knowledge of the author, there have been five Part IVA proceedings—that were either on foot at the time of, or were started after, the judgment in Philip Morris that encompassed multiple respondents suits—that were found by the Court not to have been properly instituted, mainly as a result of an inability to adhere to the Philip Morris principle.^[53] The application of the Philip Morris principle by the Supreme Court of Victoria to the Victorian equivalent of Part IVA has also resulted in the termination of the proceedings against multiple defendants in Cook v Pasminco Ltd.^[54] Furthermore, certain comments by Hill^[55] and Sackville JJ^[56] in Philip Morris raise legitimate concerns as to the future availability of the Part IVA regime in certain types of multiple respondents suits brought pursuant to the Trade Practices Act 1974 (Cth).

It is also vital to note that comments made by Senator Tate, the then Minister for Justice and Consumer Affairs, indicate that the government responsible for the introduction of Part IVA envisaged the use of the class action procedure in circumstances now precluded by the Philip Morris principle. In a media release issued on 17 April 1991 Senator Tate drew attention to the fact that

the legislation will confer no new legal rights—but it will allow people to assert existing rights collectively, effectively and more cheaply. If such a measure had already been in operation, the present actions by haemophiliacs and others who have medically acquired AIDS, for example, could have proceeded with one action representing the general group of AIDS sufferers.^[57]

The litigation referred to by Senator Tate was E v Australian Red Cross.^[58] This action was brought in the Federal Court against three respondents: the Australian Red Cross Society, the New South Wales Division of that Society and the Central Sydney Area Health Service. Wilcox J indicated that there were also thirty other individual actions before the Federal Court 'each seeking damages from the Australian Red Cross Society ..., the New South Wales Division of that Society ... and one or more respondents [emphasis added]'.^[59] The information above tends to suggest, quite clearly, that the only respondents against whom all the applicants were claiming were the Australian Red Cross Society and the NSW Division of that Society. And yet, as was indicated above, the Minister who introduced the Part IVA Bill in the Senate expressly singled out these individual proceedings as an excellent example of the type of multi-party legal dispute that could appropriately be dealt with by the new regime his Government was introducing. Similarly, Wilcox J noted that the grouped proceedings regime recommended by the ALRC would have provided an excellent vehicle for the resolution of these disputes.^[60]

In light of the adverse impact which the Philip Morris principle has had, and will continue to have, upon the ability of Part IVA to attain the benefits it was created to secure, an obvious question is whether it is in fact possible for a class action regime to operate in a satisfactory manner, in relation to multiple respondents suits, in the absence of a requirement that every applicant and group member have a claim against every respondent. The approach followed by the Federal Court itself, in a number of multiple respondents suits which preceded Philip Morris, provides an affirmative answer to the question posed above.

C The pre-Philip Morris era

The first Part IVA proceeding where it was apparent that some of the applicants and group members were not claiming against each respondent was McMullin v ICI Australia Operations Pty Ltd.^[61] The proceedings were initiated by Mr and Mrs McMullin against seven respondents. The proceedings concerned losses allegedly suffered as a result of the accumulation of chlorfluazuron in tissues of cattle. The contamination was said to have occurred when the cattle were fed cotton waste that had been sprayed with an insecticide called Helix. The applicants carried on business as farmers and graziers in New South Wales. They represented other persons, such as graziers and abattoirs, who suffered the losses described above. The class members resided and carried on their businesses in either New South Wales or Queensland. The respondents included, in addition to the manufacturers and distributors of Helix, the States of Queensland and New South Wales. The negligence claims against these States essentially concerned the decisions of the relevant functionaries of each State to register Helix under State law.

There was no discussion by the Federal Court, in the various judgments it delivered in this class action, as to whether the proceeding complied with s 33C(1)(a). But the fact that the proceeding was permitted to continue under Part IVA clearly shows that the Court was of the view that all of the threshold requirements of s 33C(1) had been satisfied; otherwise the Court would have been compelled to terminate the proceeding as Part IVA proceeding.^[62] What was equally clear, however, was that the decision of the State of Queensland to allow the registration of the impugned product in that State could hardly affect (and thus provide them with a cause of action against) those members of the class, including the applicants, who carried on business in New South Wales. Accordingly, not all class members, including the applicants, had claims against every respondent. Indeed, given that New South Wales class members had no claims against Queensland and that Queensland class members had no claims against New South Wales, it would seem highly likely that not a single member of the group could be said to have claims against each respondent.^[63] The validity of this analysis appears to be confirmed by the following comment of Wilcox J, the trial judge in McMullin:

during the hearing to which these reasons pertain, a question arose as to whether the applicants are entitled to represent group members who carried on business in Queensland, especially in relation to those members' claims against the State of Queensland. In order to resolve any difficulty ... , I constituted a sub-group consisting of those group members who conducted a grazing or other relevant business in Queensland. With his consent, I appointed Christopher John Blomfield as a sub-group representative party, representing the members of that sub-group in the proceeding ... Since 1975 Mr Blomfield has conducted a grazing business [in] ... Queensland. He also claims to have been affected by chlorfluazuron contamination.^[64]

The fact that not all members of the group were claiming against all respondents did not attract any adverse comments from the Court. Nor did it prevent a ruling by the Court as to the crucial issue of whether the respondents had in fact been negligent, as alleged by the applicants.^[65] Had the Court formed the view that the proceedings, as Part IVA proceedings, would not provide an efficient and effective means of dealing with the claims of group members or that it was simply inappropriate for the claims to be pursued under Part IVA, an order to discontinue the representative nature of the proceedings would have been made by the Court under s 33N.^[66] No such order was made.

The Federal Court's first discussion of what compliance with s 33C(1)(a) entails for multiple respondents suits can be found in the judgment handed down, several months after the McMullin proceedings, by Wilcox J in Symington v Hoechst Schering Agrevo Pty Ltd.^[67] In Symington, the named plaintiffs were graziers who claimed that their cattle had ingested a chemical, called endosulfan, marketed as a pesticide by each of the respondents under various brand names. The plaintiffs sued on behalf of themselves and other persons who allegedly suffered losses when meat was rejected and cattle impounded following the detection of endosulfan residues in beef and beef cattle. There was, however, a problem. The representative parties claimed that the product that contaminated their cattle was that of the first respondent only. Wilcox J explained as follows the impact of s 33C(1)(a) on this representative proceeding:

I think it is clear that the applicant—that is to say, the representative party—must himself or herself have standing to sue the particular respondent and, where there is more than one respondent, each of them. It is not enough that the applicant has standing to sue one respondent and other people have claims against some other respondent which arise out of similar or related circumstances and give rise to a substantial common issue of law or fact.
The present applicants claim the product supplied to their neighbours, which contaminated their cattle, was that of the first respondent, Hoechst Schering Agrevo Pty Ltd. They concede they personally have no claim against the second, third, fourth, fifth or sixth respondents. The result is they have no standing to bring an action against those respondents. Accordingly, the action must be dismissed as against those respondents.^[68]

Less than two weeks later, in Ryan v Great Lakes Council,^[69] Wilcox J was provided with another opportunity to canvass the ambit of s 33C(1)(a) in the context of multiple respondents suits. In Ryan the class representative sued on behalf of himself and all other persons who suffered injury as a result of eating oysters from the Wallis Lakes in New South Wales, which were contaminated with the Hepatitis A virus.

The action was brought against the Great Lakes Council ('the Council'), which was said to have certain legal responsibilities in regard to the quality of the water at the lake, a number of oyster farmers and oyster distributors. The representative plaintiff had a personal claim against the first and second respondents, namely, the Council and an oyster farmer, Graham Barclay Oysters Pty Ltd. He had no claims against the remaining ten respondents. Furthermore, the evidence showed that there were a number of respondents against whom it did not yet appear that at least seven persons had claims. The only respondent against whom all class members had a claim was the Council. After reviewing the terms of s 33C(1)(a) and s 33D(1), and the submissions put forward on behalf of the applicant and of the respondents, his Honour arrived at the following general conclusion regarding the scope of s 33C(1)(a):

it follows that, in order to utilise the Pt IVA procedure against a given respondent, the applicant must have a personal claim against that respondent that is shared by at least six other persons. The legislation does not prevent several respondents being joined to a single Pt IVA proceeding, so long as the commencement and standing requirements are met by the applicant in respect of each of them. Nor is there any reason why an applicant could not, within the one action, use the Pt IVA procedure against one or more respondents but not against others; the action is in the applicant's name alone, and it might be convenient to determine connected non-representative claims at the same time as the claims the applicant brings for the benefit of the group. Nor is it forbidden to consolidate the hearing of two or more representative proceedings, brought by different representatives but having, as between them, such similarity as to warrant their being heard together.^[70]

The application of these principles to the proceedings before him led Wilcox J to make the following orders:

as regards those respondents against whom the applicant personally makes no claim, the present proceeding will, in any event, need to be dismissed. If it is possible to organise a separate Pt IVA proceeding against a particular oyster grower or distributor using a different representative party, it may be appropriate to consolidate that proceeding with this one; the claim against the council, at least, will be common to both proceedings, and the expert and other evidence seems likely to be substantially similar in nature and effect. Similarly, group members affected by the dismissal of the current proceeding against particular respondents, who could not satisfy the s 33C requirements, might wish to bring individual claims against those respondents which, again, might appropriately be consolidated with the current proceeding.^[71]

If the exercise of ascertaining the views of the Court, as to what compliance with s 33C(1)(a) entailed in multiple respondents suits, was restricted to the comments in Symington and Ryan set out above, then one would be entitled to conclude that, subject to one exception, the Court adhered to the view that compliance with s 33C(1)(a) required each group member to be claiming against each respondent. The exception referred to above concerns Ryan. It will be recalled that whilst the applicant in Ryan had a personal claim against the first two respondents, the only respondent against whom all class members had a claim was the first respondent. Consequently, complete adherence to the Philip Morris principle would have also required an order that s 33C(1)(a) had not been complied with in relation to the action against the second respondent.

As the trial judge in the Philip Morris proceedings, Wilcox J said:

in Symington ... I held this paragraph [s 33C(1)(a)] requires that the applicant, or each one of several applicants, and each group member must have a claim against each respondent; it is not sufficient for one applicant to make a claim against one respondent and another applicant or a group member to make a claim against some other respondent.^[72]

A substantially different picture, concerning the pre-Philip Morris practice of the Federal Court in relation to multiple respondents suits, emerges, however, if one concentrates (a) on the subsequent class action in Schneider v Hoechst Schering Agrevo Pty Ltd;^[73] and (b) on the orders that were subsequently made in Ryan. The action in Schneider was described as follows by the Federal Court:

this case is about damaged wheat crops ... A common feature in relation to many of these [damaged] crops was that they had been sprayed with a chemical herbicide called Puma S. One of the farmers, Mr Schneider, commenced a representative proceeding against the manufacturer of Puma S ... Hoechst ... on behalf of all affected farmers. He also commenced proceedings against IAMA Limited ('IAMA') from whom he had purchased the Puma S ... The proceedings against IAMA were said to be brought on behalf of a sub-group of the represented parties, namely those who had purchased Puma S from IAMA.^[74]

It is apparent from the description above that while the applicant had a claim against each of the two respondents, the same cannot be said about every group member. The use of the term 'subgroup' by the Court suggests quite clearly that not all of the represented persons had a claim against the second respondent, the distributor.

An even clearer illustration of the Federal Court permitting multiple respondents suits to commence and continue, despite the inability of the plaintiffs to demonstrate that every member of the group, including the representative parties, had a claim against every respondent, is provided by the developments in the Ryan proceeding itself. Additional representative plaintiffs were added. Wilcox J explained, as follows, how, in his view, the addition of further applicants now ensured compliance with the requirements of s 33C(1)(a):

on 18 September 1997 I ruled that Mr Ryan was not competent to maintain a representative action against a person in relation to whom he had no personal claim ... However, I subsequently gave leave Mr Ryan to amend the proceeding in such a manner as to join additional applicants; each being a person who made a personal claim against a particular grower or distributor and was therefore competent to represent other group members who had claims against that grower or distributor.^[75]

After Wilcox J delivered a judgment on the 'merits' of the action, there was an appeal to the Full Federal Court. In the judgment of one of the members of the Court, Lindgren J, appear the following comments, which are directly relevant to the issue currently being considered:

Originally Mr Ryan alone was the applicant but as a result of Ryan v Great Lakes Council (1997) 78 FCR 309 the proceeding was reconstituted to add the other six applicants. As a result, each applicant who made a personal claim against a particular oyster grower or distributor was competent to represent other group members who claimed against the same grower or distributor. Following the reconstitution of the proceeding, the group members to whom the proceeding related were defined as 185 people named in eleven annexures ... Each annexure listed the group members who claimed against a particular grower or distributor, the Council or the State. All 185 people claimed against the Council and the State but not all of those people claimed against a particular grower or distributor of oysters.^[76]

It is clear from the summary above that, with the exception of the Council and the State, not all applicants and class members in Ryan were making a claim against every respondent.

In light of the discussion above, it can be safely concluded that despite the comments in Symington and Ryan—which appear to accord largely with what the Full Federal Court subsequently held in Philip Morris—the litigation in McMullin, Ryan and Schneider was allowed to progress despite the fact that not all representative parties and group members were claiming against all respondents. Equally important is the fact that in McMullin, Ryan and Schneider, the proceedings reached the final stage of a judgment on the common issues.^[77] This indicates quite clearly that non-adherence to the Philip Morris principle will not necessarily lead to unmanageable proceedings^[78] or generate unfair scenarios for the class members and the respondents,^[79] as long as there is sufficient similarity between the claims advanced against the various respondents. This scenario is consistent with the experience in the Canadian provinces of Ontario and British Columbia. Attention will now be turned to how courts in those jurisdictions have dealt with multiple respondents suits.

D Ontario and British Columbia

The Class Proceedings Act, SO 1992, c 6 ('Ontario Act') introduced, in January 1993, a modern class action regime in Ontario. Pursuant to s 5 of the Ontario Act, the Court is required to certify^[80] a class proceeding if:

• the pleadings disclose a cause of action;
• there is an identifiable class^[81] of two or more persons that would be represented by the representative plaintiff or defendant;
• the claims or defences of the class members raise common issues;
• a class proceeding would be the preferable procedure for the resolution of the common issues; and
• there is a representative plaintiff or defendant who would fairly and adequately represent the interests of the class; has produced a plan for the proceeding that sets out a workable method of advancing the proceeding on behalf of the class and of notifying class members of the proceeding; and does not have, on the common issues for the class, an interest in conflict with the interests of other class members.

As early as August 1993, courts in Ontario were faced with the question of whether multiple respondents suits should be certified where some of the plaintiffs and some of the class members do not have claims against all defendants. The case in question was Bendall v McGhan Medical Corp.^[82] The class was defined as follows:

all persons who have had silicone gel breast implants placed in their bodies, whose implants were manufactured, developed, designed, fabricated, sold, distributed or otherwise placed into the stream of commerce by the named Defendants.^[83]

The judge presiding over this class proceeding, Montgomery J of the Ontario Court (General Division), explained that

the plaintiff Bendall allegedly received implants manufactured by the defendant McGhan Medical Corporation ('McGhan') while the plaintiff Wise allegedly received implants manufactured by the defendants Dow Corning Canada Inc, or Dow Corning Corporation ('Dow Corning').^[84]

This meant, of course, that not all representative parties were making claims against all defendants. The same conclusion applied to the class members, as those women who received McGhan implants had no cause of action against Dow Corning and those class members who received Dow Corning implants had no cause of action against McGhan. However, the Court rejected the defendants' proposition that this state of affairs warranted the denial of certification under s 5. Montgomery J's conclusion was consistent with, and probably the product of, his views concerning the appropriate approach to the interpretation of the Ontario Act:

in my opinion, the court should err on the side of protecting people who have a right of access to the courts. This is remedial legislation and it should be addressed with a purposive approach. This is not inconsistent with the duty to look carefully at the facts to see if they meet the requirements of s 5.^[85]

Bendall was recently endorsed by Cumming J of the Ontario Superior Court of Justice in Gariepy v Shell Oil Company.^[86] Cumming J confirmed that 'it is unnecessary in a class proceeding for the representative plaintiff to personally have a cause of action against each named defendant. It is sufficient for a representative plaintiff to represent a proposed class of persons who may be able to assert claims against the various defendants'.^[87] Consequently, the position in Ontario regarding multiple respondents suits is quite clear. As long as it is possible to demonstrate that for each of the various defendants there are some class members who have claims against that particular defendant, a class proceeding may proceed, if the prerequisites for certification contained in s 5 are satisfied. It is not necessary to show that the plaintiffs and the class members are all claiming against each defendant.

A modern class action regime was introduced in British Columbia in August 1995 as a result of the enactment of the Class Proceedings Act, RSBC 1996, c 50 ('BC Act'). This Act was based on, and is substantially similar to, the Ontario Act.^[88] One of the first judicial pronouncements on the BC Act was that of Mackenzie J of the Supreme Court of British Columbia in Harrington v Dow Corning Corp.^[89] His Honour certified a proceeding in which the representative party was acknowledged to have a cause of action against only five of the sixteen defendants. Mackenzie J also indicated that 'negligence is a cause of action which involves the manufacturers severally and it may be appropriate to divide the class into sub-classes by manufacturer, with separate representatives for each sub-class'.^[90] However, the Court went on to certify the class action without requiring a representative plaintiff for each manufacturer.

Harrington was applied by the Court of Appeal for British Columbia in Campbell v Flexwatt Corporation:

there is no requirement that there be a representative plaintiff with a cause of action against every defendant; the legislation simply requires that there be a cause of action. If a class includes a subclass whose members have claims that raise common issues not shared by all members of a class then the court must appoint a representative plaintiff for the subclass if the court determines that the representative plaintiff for the class could not fairly and adequately represent the interests of the subclass.^[91]

As was the case with the practice of the Federal Court in the pre-Philip Morris era, permitting the use of the class action device in multiple respondents suits, without requiring compliance with a principle such as the Philip Morris principle, has not generated problems for Canadian courts—in managing class proceedings—nor has it resulted in unfairness for either class members or defendants in Ontario and British Columbia. At the same time, this approach has expanded the availability of the class action device in multiple respondents suits thereby facilitating the attainment of the important social goals of class actions.^[92] A striking illustration of this state of affairs is provided by the litigation in Bendall.^[93] In that case, the Court indicated that 'certification is the only way a large number of women can access a legal system that would otherwise be denied to them'.^[94] The outcome of this class proceeding and similar breast implants class actions in Canada was that a global settlement for more than $50 million was attained on behalf of all class members.^[95]

E Associated issues

In Philip Morris, the Full Federal Court did not consider what options or procedures are available to the Court once a multiple respondents suit is found not to have satisfied the condition which, according to the Court, is contained in s 33C(1)(a), namely, that all members of the group have a claim against each respondent. A leading Victorian QC, Jonathan Beach, argued, at a seminar held in October 2000, that

in summary, on the authority of Ryan, it may be permissible directly, or through separate proceedings and then consolidation, for:
(i) applicant A to bring representative claims for a group against respondent A only and an individual claim against respondent B;
(ii) applicant A to bring representative claims for one group against respondent A and applicant B to bring other representative claims for a separate group against respondent B (providing that applicant A and six others at least have claims against respondent A and applicant B and six others at least have claims against respondent B);
(iii) applicant A to bring representative claims for one group against respondent A, and for one of those group members as a separate applicant B to bring an individual claim against respondent B.
Nixon would appear in its terms not to foreclose any of those possibilities. It only appears to stand as authority for the proposition that if an applicant(s) brings claims on behalf of group members (one group) against multiple respondents, then every group member must have a claim against every respondent.^[96]

It is submitted that scenario (ii) may not be attained directly. This is clearly shown by the passage from Sackville J's judgment set out above^[97] and by the recent judgments in Milfull v Terranora Lakes Country Club Limited,^[98] Cook v Pasminco Ltd,^[99] Batten v CTMS Ltd^[100] and Hunter Valley Community Investments Pty Ltd v Bell.^[101] Whether each of those three scenarios proposed by Beach are available through consolidation of proceedings is not clear, especially in light of the recent judgments of Kiefel J in Batten and Milfull. Order 29 r 5 of the Federal Court Rules provides that

where several proceedings are pending in the same Division, then, if it appears to the Court:
(a) that some common question of law or fact arises in both or all of them;
(b) that the rights to relief claimed therein are in respect of, or arise out of, the same transaction or series of transactions; or
(c) that for some other reason it is desirable to make an order under this rule;
the Court may order those proceedings to be consolidated or may order them to be tried at the same time or one immediately after the other or may order them to be stayed until after the determination of any of them.^[102]

In Batten Kiefel J indicated that in the proceedings before her:

the 1st applicant (and the group members he represents) does not have a claim against the PCS companies or their directors, and the 2nd applicant (and his group) has none against the 1st to 5th respondents. The proceedings cannot therefore be continued as representative proceedings.^[103]

Her Honour then considered whether it was appropriate to grant leave to join the parties under O 6 r 2(b) of the Federal Court Rules. This rule provides that two or more persons may be joined as applicants in any proceeding where the Court gives leave to do so.^[104] Kiefel J held that it was appropriate to exercise this power based on considerations such as cost, delay and fairness. What happened after Kiefel J delivered her judgment in Batten was explained as follows by her Honour in Milfull:

in Batten v CTMS Ltd ... I ordered two separate representative proceedings be treated as joined proceedings, on the basis that they shared a common question of fact. That order was set aside by a Full Court, with the consent of the parties. The proceedings were later ordered to be heard together.^[105]

Kiefel J explained why her earlier ruling in Batten was incorrect, as follows:

upon further reflection, it would not seem appropriate to apply the rule for joinder, given that Part IVA ... makes express provision for the circumstances in which claims, other than those constituting the representative proceeding itself, may be added to the proceeding and for the procedures which are to be followed. I refer in particular to ss 33Q, 33R and 33S.^[106]

In order to consider the cogency of Kiefel J's analysis it is necessary to consider the provisions referred to by her Honour. Section 33Q(1) empowers the Court to give directions in relation to the determination of issues which have been left unresolved by the determination of the common issues.^[107] Section 33Q(2) allows the Court to establish sub-groups so as to deal with issues common to the claims of some only of the group members.^[108] Section 33R(1) allows individual group members to appear in the class proceeding for the purpose of determining an issue that relates only to the claims of that member.^[109] Section 33S comes into play where an issue cannot properly or conveniently be dealt with under s 33Q or s 33R. Section 33S(a) provides that if the issue concerns only the claim of a particular member, the Court may give directions relating to the commencement and conduct of a separate proceeding by that member; whilst s 33S(b) provides that if the issue in question is common to the claims of all members of a sub-group, the Court may give directions relating to the commencement and conduct of a representative proceeding in relation to the claims of those members.^[110]

It is submitted that Kiefel J's reliance on ss 33Q, 33R and 33S to justify the inappropriateness of her reliance in Batten on O 6 r 2(b) is, with respect, unpersuasive. None of those provisions of Part IVA deal with the issue of joinder of parties which, of course, is the topic with which that rule deals. Furthermore, ss 33Q,^[111] 33R and 33S^[112] empower the Court to dispose of non-common issues only in respect of the parties bound by a properly constituted Part IVA proceeding, that is, a proceeding which complied from the outset with the requirements of s 33C(1), which include, of course, the Philip Morris principle. In a situation where there is, for instance, a Part IVA proceeding ('proceeding A') by a group of persons ('group A') against respondent A and the question is whether, to these proceedings, can be joined another group of persons ('group B') who have claims against another respondent ('respondent B') ss 33Q, 33R and 33S can have no role to play. This is because those provisions may only be employed by the Court presiding over proceeding A in relation to those who are bound by that proceeding and, as a result of the Philip Morris principle, proceeding A cannot encompass group B and respondent B. Once the limited operation of these provisions is appreciated, it becomes clear why one of the provisions of Part IVA, s 33ZG(c)(iii), expressly and unambiguously provides that 'except as otherwise provided by this Part, nothing in this Part affects ... the operation of any law relating to joinder of parties'.

The reasoning in Milfull, if applied in future cases, is likely to prevent the employment of the consolidation procedure in relation to each of the three scenarios depicted by Beach. If ss 33Q, 33R and 33S are sufficient, as Kiefel J held, to displace the Court's general powers with respect to joinder of parties (despite the fact that Parliament sought to expressly save these powers through s 33ZG(c)(iii)), then a similar fate should logically be suffered by the Court's general powers with respect to consolidation of proceedings. This should logically follow in spite of the fact that, as with the joinder powers, Parliament sought to expressly save these powers, through the enactment of s 33ZG(c)(iv).^[113]

F Alternative approaches to the application of the Philip Morris principle

Sections IIIA to IIIE above have sought to demonstrate that the requirement that every applicant and group member have a claim against every respondent has highly undesirable consequences and is not, in any event, necessary. In this section it will be submitted that, if this interpretation of s 33C(1)(a) is to remain,^[114] it should not be applied in a restrictive manner, as was done by the Court in Philip Morris; instead, the flexible approach evident in the judgment of a differently constituted Full Federal Court in King v GIO Australia Holdings Ltd^[115] should be followed.

In Philip Morris, when appearing before the Full Federal Court, counsel for the applicants submitted that the primary judge's comments, set out below, provided an accurate statement of the case the applicants intended to plead:

the applicants do not seek to make a product liability case, in which it would be essential to relate the injury suffered by a particular claimant to a deficiency in the product of a particular manufacturer or distributor ... The applicants' case is painted on a larger canvas. They claim the three sets of respondents—who, they say, at all relevant times dominated the Australian retail cigarette market—embarked individually and collectively on a course of conduct designed to create a false community perception about the risks associated with cigarette smoking. If that claim can be made good, it would seem not to matter that a particular claimant smoked cigarettes manufactured by only one of the respondents; indeed, logically, by none of them, although the present claim is not so wide.^[116]

Sackville J was of the view that the statement of claim lodged on behalf of the applicants did not plead a case of collective conduct on the part of all three respondents; it was, instead, the conduct of one or more of the respondents which was said to have influenced the group members to smoke or continue to smoke the cigarettes.^[117] His Honour then considered what result would ensue if the statement of claim had clearly pleaded that each class member was influenced by the conduct of each respondent. He concluded that

further pleading and endless management issues would be raised. Would it possible to particularise such a case in a manner that makes it clear how class members are said to have been influenced by advertisements or public statements they may never have seen? Is it feasible to contemplate continuing representative proceedings when the smoking history of and factors influencing members of the represented class are likely to vary so substantially?^[118]

The first question posed by his Honour may be criticised for placing greater importance on the procedural requirements contained in s 33H than on the crucial issue of whether a proceeding satisfies the requirements of s 33C(1). As was clearly indicated by the High Court in Wong v Silkfield Pty Ltd,^[119] the requirements for the originating process in a class action, set out in s 33H, are simply designed to show that the threshold requirements in s 33C(1) have been met.^[120] It is also difficult to see how the latter consideration canvassed by Sackville J could be relevant to the question of compliance with s 33C(1)(a). It would only be relevant to the issue of whether the Court should exercise its discretion, under provisions such as s 33N, to order the discontinuance of a properly constituted class suit.^[121]

Equally unpersuasive was, with respect, the approach of Hill and Spender JJ in relation to the issue of whether the proceedings satisfied s 33C(1)(a). Their approach is accurately captured by the following passage from Hill J's judgment:

it is impossible to conceive how a case could be pleaded, consistent with instructions, which would allege that each and every applicant class member suffered loss or damage as a result of misleading and/or deceptive conduct of every respondent as principal, or as a result of negligence on the part of every respondent as principal ... I share the view of Spender J as to the equal impossibility of its being suggested that each and every class member was caused to smoke, discouraged from giving up or encouraged to continue smoking as a result of each lobbying effort being part of a single campaign to which each company was a party.^[122]

These comments evince, with respect, an excessively narrow application of the requirements of s 33C(1)(a) which is at odds with both (a) the philosophy underpinning Part IVA and with (b) the fact that s 33C(1)'s purpose is not to authorise the Court to assess the likelihood that the claims brought on behalf of the class will succeed.^[123] This fallacy in the reasoning of their Honours becomes apparent when the judgment of a differently constituted Full Federal Court in King is considered.

The proceeding in King concerned advice provided to the shareholders of GIO as to whether a takeover offer by AMP should have been accepted. The class was described as 'GIO shareholders who did not accept the AMP takeover offer "by reason of the conduct ... of all (or alternatively, any) of the Respondents and who suffered loss as a consequence"'.^[124] The use of the alternative 'any' was relied upon by the respondents for the proposition that the proceeding failed to comply with s 33C(1)(a) in that it did not disclose a claim by each group member against each respondent. Wilcox, Lehane and Merkel JJ, in their joint judgment, upheld the judgment of the trial judge and rejected the argument of the respondents^[125] on the following grounds:

the alternative 'any' means that a person is a group member if he or she, as a matter of fact, suffered loss as a result of the conduct of any respondent. This is necessary to cover the situation of a group member who, although claiming against all respondents, only suffered loss by reason of the conduct of one of the respondents. That person is still to be regarded as a group member and, accordingly, is bound by the result.
The fact that a person is ultimately adjudged to be entitled to succeed against only one respondent, does not mean that a person makes a claim against only that respondent. There is a world of difference between a claim and success on the claim.^[126]

The reasoning above is directly relevant to Philip Morris.^[127] The fact that it was highly likely that, once the claims of the class were adjudicated upon, it would transpire that very few of the class members in Philip Morris had succeeded against each respondent was not relevant to the question of whether every member of the class was claiming against all respondents.^[128] Furthermore, it is submitted that Wilcox J's formulation of what the applicants in Philip Morris needed to demonstrate, in order to succeed on the cause of action that relied upon s 52 of the Trade Practices Act 1974 (Cth), is to be preferred to that provided by the Full Court. His Honour concluded that

The claimant would establish a s 52 claim if he or she established:
(a) the publication by a particular respondent or respondents of misleading information;
(b) that this information caused the claimant to commence, or continue, to smoke cigarettes; and
(c) that such smoking—that is, the smoking caused by the misleading conduct of a particular respondent or respondents—caused the claimant to contract a particular disease.
For the purposes of the s 52 claim ... it would seem unnecessary for a claimant to establish an identicality between the respondent (if only one) whose promotional material caused him or her to commence or continue smoking and the respondent whose product caused the relevant disease.^[129]

The Federal Court will soon be provided with the opportunity, in Bray v F Hoffman-La Roche Ltd,^[130] to decide upon the most appropriate approach to the application of the Philip Morris principle. In Bray the applicant commenced a Part IVA proceeding against 13 respondents claiming damages in respect of an international price fixing and market sharing arrangement in respect of a number of vitamin products. The group members were described in the pleadings as follows:

persons who between 5 March 1992 and 5 July 1999 purchased in Australia all or some of vitamins ... either directly or indirectly by way of purchase of foods, beverages, vitamin pills or capsules or other products which contained one or more class vitamins supplied by one or more of the respondents.^[131]

The highlighted part of the passage above bears some similarity to the pleadings in Philip Morris.^[132] In a judgment handed down in March 2002, Merkel J, the trial judge in Bray, indicated that he has not yet had an opportunity 'to consider any questions concerning the pleadings [and] ... questions arising under Part IVA'.^[133]

IV THE CLAIMS MUST BE IN RESPECT OF, OR ARISE OUT OF, THE SAME, SIMILAR OR RELATED CIRCUMSTANCES

The inclusion of the term 'related' in s 33C(1)(c)^[134] tends to suggest that this provision was not intended to place, in the path of potential representative parties, a significant barrier. This impression is confirmed by s 33C(2)(b) which makes it clear that a representative proceeding may be commenced whether or not the proceeding:

(i) is concerned with separate contracts or transactions between the respondent in the proceeding and individual group members; or
(ii) involves separate acts or omissions of the respondent done or omitted to be done in relation to individual group members.^[135]

As was indicated by the High Court, 'the broadening provisions in subs (2) of s 33C emphasise the width of the entitlement conferred by s 33C(1) to commence a representative proceeding'.^[136]

In Zhang v Minister for Immigration, Local Government and Ethnic Affairs,^[137] French J outlined the parameters of s 33C(1)(b). His analysis, set out below, has been referred to with approval by the High Court^[138] and by the Full Federal Court:^[139]

[s 33C(1)(b)] contemplates a relationship between the circumstances of each claimant and specifies three sufficient relationships of widening ambit. Each claim is based on a set of facts which may include acts, omissions, contracts, transactions and other events. As appears from s 33C(2), the circumstances giving rise to claims by potential group members do not fall outside the scope of the legislation simply because they involve separate contracts or transactions between individual group members and the respondent or involved separate acts or omissions of the respondent done or omitted to be done in relation to individual group members.
The outer limits of eligibility for participation in representative proceedings are defined by reference to claims in respect of or arising out of related circumstances. The word 'related' suggests a connection wider than identity or similarity. In each case there is a threshold judgment on whether the similarities or relationships between circumstances giving rise to each claim are sufficient to merit their grouping as a representative proceeding. At the margins, these will be practical judgments informed by the policy and purpose of the legislation. At some point along the spectrum of possible classes of claim, the relationship between the circumstances of each claim will be incapable of definition at a sufficient level of particularity, or too tenuous or remote to attract the application of the legislation.^[140]

French J held that the proceedings before him adhered to s 33C(1)(b) despite the fact that 'the relationship between the circumstances of each group member is defined by a few common integers which leaves room for considerable diversity in circumstances which might support individual claims to set aside the review decisions'.^[141] The Court's conclusion was based on 'the need for a purposive approach to the construction of s 33C(1)(b)'.^[142]

Unfortunately, a far more restrictive construction of s 33C(1)(b) was embraced by Sackville and Spender JJ in Philip Morris.^[143] The former justice explained that:

The representative procedure provided for by Part IVA ... is plainly designed to accommodate a case where the applicants and group members rely on a series of related but not identical transactions, such as similar representations being made separately to different individuals: Grouped Proceedings, para 134. But this case involves vastly different forms of advertising, promotions and other public statements by the three respondents over four decades. It is true that the applicants allege that the various public statements—ranging from a single brand name on a billboard at a sporting match to a submission to a Senate Committee—all make substantially the same representations. Yet to test that allegation it would be necessary to examine each of the public statements made over four decades in its own context, having regard to the characteristics of the likely audience. This is a far cry from the kind of case envisaged by the [ALRC] as falling within the purview of the representative procedure.^[144]

With respect, Sackville J appears erroneously to have injected into the task of ascertaining whether s 33C(1)(b) has been complied with, considerations pertaining to whether the class action device would provide the most effective and efficient vehicle for the resolution of the legal dispute in question. These considerations are relevant to the exercise of the powers conferred upon the Federal Court by s 33N but ought not to be entertained in the context of s 33C(1). As was clearly indicated by the High Court in Silkfield,

[s 33C(1)] is concerned with the commencement, not subsequent conduct, of litigation using the procedures provided in Pt IVA.^[145] ...
The circumstance that proceedings which pass the threshold requirement of s 33C may later be terminated as representative proceedings, by order made under s 33N, confirms rather than denies ... a [broad] construction of s 33C(1).^[146]

Accordingly, once considerations such as the number and different types of representations made by the tobacco companies and the period over which those representations were made are put to one side, it becomes clear that the conduct of the various respondents challenged by the applicants essentially entailed the making of, as was conceded by Sackville J, 'substantially the same representations'.^[147] It is therefore difficult to understand why the facts in Philip Morris did not constitute an illustration of one of the scenarios which the ALRC envisaged would satisfy this prerequisite, namely, similar representations being made separately to different individuals.^[148]

Sackville J's narrow construction of s 33C(1)(b) is also not in accordance with, and should not be preferred to, the approach of Einfeld J in Marks v GIO Australia Holdings Ltd:

I am not convinced that mere volume of evidence disqualifies a proceeding from being undertaken as a class action. Nor in my experience has complexity ever been a reason for failure to determine an issue. I am also not persuaded that substantial differences in individual circumstances disqualify a case from being a class action. Part IVA anticipates that individuals in the group will have differing circumstances ... As far as group actions provided for by Pt IVA are concerned, what is relevant is similarity not difference.^[149]

Spender J also concluded that s 33C(1)(b) had not been satisfied by the class in Philip Morris. He explained the reasons for this conclusion by furnishing the following hypothetical scenarios:

suppose one company, Widget Retailing Ltd ('Widget') had distributed its goods by road throughout Australia over four decades. Suppose further that various individual members of the public were injured by the negligent driving of different employees of Widget, as follows: A in Melbourne in 1960; B in Townsville in 1974; C in Sydney in 1984; and ... Z in Perth in 1996. Claims for damages by each injured individual against Widget, in my opinion, would not properly be described as claims arising out of the same, similar or related circumstances. This is so, even if it be the position that on each occasion the negligence consisted of, say, driving at an excessive speed. In my opinion, a representative proceeding against Widget could not properly be brought under Pt IVA ... on those assumed facts. The ALRC reports (which led to the introduction of Pt IVA) do not suggest that it was intended that such disparate claims could properly be brought in representative proceedings.
Closer to the present case, if each of A, B, C ... and Z had suffered loss or damage in reliance on a particular deceptive advertisement by Widget at the times and places posed, with the advertisements in each case being different (including, for instance, advertising different products), but each one being part of Widget's campaign over the years to persuade people to buy its products, again the claims, in my opinion, would not arise out of the 'same, similar or related circumstances', within the proper meaning of those words in s 33C(1)(b).^[150]

The author is not as confident, as Spender J was, that the two scenarios above would not have been seen by the ALRC as satisfying the prerequisites for the commencement of group litigation. It should not be forgotten that s 33C(2) was based upon a specific recommendation of the ALRC. The ALRC explained that without a provision such as s 33C(2) there was a risk that the same, similar or related circumstances requirement might be interpreted in a way that 'would prevent the grouping of claims if the persons affected were involved in different transactions or series of transactions even if those transactions involved similar or related acts of the respondent'.^[151]

Furthermore, one of the examples provided by the ALRC, of situations which could be dealt with through the group litigation model it was proposing, could possibly be regarded as involving circumstances as disparate as those covered by Spender J's examples. The illustration provided by the ALRC was that of litigation encompassing 'approximately 1500 Australian women ... involved in a claim against AH Robins' Australian subsidiary for injuries arising from the use of the Dalkon Shield birth control device'.^[152] Similarly, the use made by Senator Tate, of the litigation brought by sufferers of medically-acquired HIV virus, as an excellent example of the intended impact of Part IVA, tends to indicate that s 33C(1)(b) was not intended to constitute a significant hurdle for representative plaintiffs. In fact, the claims of the various applicants in the proceedings referred to by Senator Tate involved substantially different circumstances.^[153]

The interpretation and application of s 33C(1)(b) implemented by the Full Federal Court in Philip Morris have the potential to drastically curtail the use of the Part IVA regime by those wishing to bring multiple respondents suits.^[154] As was perceptively noted by Beach,

the Court appeared [to assume] that, to satisfy the requirements of sub-paragraph (b), an applicant had to plead a case which related to the injury suffered by a particular claimant to either:
(a) the use by that claimant of a product produced by each and every respondent; or
(b) conduct by each and every respondent which was directed towards that particular claimant and to which that particular claimant was exposed.
On one view, it is arguable that in making these assumptions the Court gave s 33C(1)(b) a more restrictive operation than its plain words warrant and placed a gloss on the paragraph.^[155]

V THE CLAIMS MUST GIVE RISE TO A SUBSTANTIAL COMMON ISSUE OF LAW OR FACT

In Silkfield a majority of the Federal Court held that in order to comply with s 33C(1)(c) the representative parties were required to satisfy the Court that determination of the issue common to the claims of all group members was likely to have a major impact on the conduct and outcome of the litigation by resolving wholly or to any significant degree the claims of all group members.^[156] On appeal, however, a different view was adopted by Gleeson CJ, McHugh, Gummow, Kirby and Callinan JJ of the High Court:

Clearly, the purpose of the enactment of Pt IVA was not to narrow access to the new form of representative proceedings beyond that which applied under [the traditional representative action procedures] ... This suggests that, when used to identify the threshold requirements of s 33C(1), "substantial" does not indicate that which is "large" or "of special significance" or would "have a major impact on the ... litigation" but, rather, is directed to issues which are "real or of substance".^[157]

The ruling of the High Court in Silkfield has rendered compliance with s 33C(1)(c) a reasonably easy task.^[158] However, as was the case with sub-ss (a) and (b) of s 33C(1), the application of sub-s (c) to multiple respondents suits has generated not insignificant problems for representative plaintiffs. In fact, s 33C(1)(c) has been held to require that the claims of all applicants and group members against all respondents must give rise to at least one substantial common issue of fact or law.^[159] The problems which this requirement can generate were evident in Hunter Valley.^[160]

This proceeding related to the Hunter Valley Community Investments Limited Partnership ('the Partnership') which was established as part of a scheme which was promoted to potential investors as providing a tax-effective form of investment suitable for wage and salary earners. When the expected tax benefits did not eventuate, a Part IVA proceeding was commenced on behalf of the partners of the Partnership. There were four respondents to the class suit: Mr Bell ('the first respondent'), the promoter of the scheme; and Lawler Davidson ('the second respondent'), Lawler Davidson Partners ('the third respondent') and Walker & Co ('the fourth respondent'), which are all accounting firms. According to the applicants, there was one common question of law or fact which related to the claims of all applicants against all respondents.

In order to understand this common issue, a brief description of the claims advanced by the applicants is necessary. It was alleged that Bell, Lawler Davidson and Lawler Davidson Partners made representations with respect to the scheme that breached s 52 of the Trade Practices Act 1974 (Cth). Lawler Davidson and Lawler Davidson Partners provided accounting services to the applicants for the years 1992 to 1995 and it was alleged that in providing such services to the applicants, the second and third respondents breached their retainer with, and/or duty of care to, each of the applicants. Walker & Co was retained to audit the financial statements of Bell, as General Partner of the Partnership ('the Retainer'). It was alleged that Walker & Co breached the Retainer and/or breached its duty of care to each of the applicants as partners in the Partnership by failing to advise the applicants that such financial statements were not a true and fair view of the financial performance of the Partnership and that the applicants should not rely on the financial statements in their current form.

The common issue of law or fact identified by the applicants was whether the failure to advise on the part of Walker & Co, as outlined above, 'resulted in each of the Applicants losing the opportunity to have discovered in May 1995 the true nature of the misleading and deceptive conduct, breaches of retainer and/or negligence on the part of [Mr Bell], [Lawler Davidson] and [Lawler Davidson Partners]'.^[161]

The Court, however, held that the paragraph above did not identify

an issue of fact or law common to the claims of all applicants against all respondents. In the first place, the proposed statement of claim does not allege that a consequence of Walker & Co's breach of retainer or duty was to deprive the applicants of the opportunity to have discovered in May 1995 'the true nature of the misleading and deceptive conduct ...' of Mr Bell, Lawler Davidson and Lawler Davidson Partners. In any event, [the alleged common issue] identifies the issue as whether the applicants, by reason of Walker & Co's breach of retainer or duty, lost that opportunity. That issue is not common to all claims of the applicants against the other respondents, let alone the unpleaded claims of group members.^[162]

The requirement of an issue of fact or law common to the claims of all applicants and group members against all respondents also proved to be too difficult to satisfy in other class proceedings.^[163]

VI CONCLUSION

As was recently noted by Sackville J, 'it is becoming increasingly frequent for representative proceedings to be brought by more than one representative applicant against more than one respondent'.^[164] Consequently, the question of how accessible the Part IVA regime has been, to those wishing to act on behalf of a group of persons with similar legal grievances against multiple respondents, is of considerable importance in assessing the general operation and impact of Part IVA over the last ten years.

The analysis developed in this article has revealed the existence of significant barriers to the commencement of Part IVA proceedings against more than one person.^[165] It has been shown that this scenario is the result of a number of factors. The most significant of these factors has been the fact that s 33C(1)(a), as construed by the Full Federal Court in Philip Morris, requires each applicant and group member to be claiming against each respondent. This has resulted in an inability to bring within the one Part IVA proceeding claims brought on behalf of various groups of persons against various respondents, even where the causes of action against the various respondents were linked by similar or common issues of fact or law.

This scenario is difficult to reconcile with the clear intendment of the Part IVA regime, of allowing groups of related claims to be determined, so far as possible, in one representative proceeding. This state of affairs becomes apparent when one compares the Philip Morris principle with O 6 r 13 of the Federal Court Rules.^[166] This rule governs the so-called traditional representative action procedure^[167] and provides that

where numerous persons have the same interest in any proceeding the proceeding may be commenced, and, unless the Court otherwise orders, continued, by or against any one or more of them as representing all or as representing all except one or more of them.^[168]

Part IVA was intended to allow the bringing of group litigation in circumstances which would fall outside the purview of the traditional representative action procedure.^[169] It is, therefore, disappointing and ironic that, at least in relation to multiple respondents suits, Part IVA operates in a more restrictive manner than Order 6 rule 13. In fact, in SZ v Minister for Immigration & Multicultural Affairs Lehane J of the Federal Court recently noted that

although it is by no means clear that claims are asserted by all the represented persons against all respondents, there is no explicit requirement under ... FCR O6 r 13 ... that all applicants claim against all respondents (compare s 33C(1)(a) of the Federal Court of Australia Act ...).^[170]

The Philip Morris principle is also difficult to reconcile with the class proceedings that were allowed to be brought by the Federal Court under Part IVA in McMullin, Ryan and Schneider.^[171] Those cases, and the approach of courts in Ontario and British Columbia, substantiate the view that permitting one class proceeding to encompass applicants and/or group members who do not have claims against all respondents need not equate to proceedings which are either unmanageable or provide unfair advantages to the plaintiffs.

Moreover, if abandoning the Philip Morris principle were to lead, in some circumstances, to the unsatisfactory scenarios noted in the preceding sentence, the provisions which were mentioned in Section IIIE above—ss 33Q,^[172] 33R and 33S—provide the Court with adequate powers to deal with those problems.^[173] Furthermore, s 33ZG(b) provides that nothing in Part IVA affects 'the Court's powers under provisions other than this Part, for example, its powers in relation to a proceeding in which no reasonable cause of action is disclosed or that is oppressive, vexatious, frivolous or an abuse of the process of the Court'.^[174] But perhaps the most important provision is s 33ZF(1) which empowers the Federal Court to make 'any order ... [it] thinks appropriate or necessary to ensure that justice is done in the proceedings'.^[175] If those powers are not sufficient in a given class action, the Court may terminate the representative nature of the proceedings pursuant to the power conferred upon it by s 33N.^[176]

The Philip Morris principle has been accompanied by a recent refusal, on the part of the Federal Court, to allow representative plaintiffs to employ procedures, such as the joinder procedure (and presumably the consolidation procedure), to achieve, to some extent and indirectly, the result which the Philip Morris principle does not authorise directly. The difficulties faced by those wishing to bring multiple respondents suits have been exacerbated by a narrow interpretation of the concept of identical, similar or related circumstances by the Federal Court in recent cases and by an interpretation of s 33C(1)(c) which entails the need to demonstrate the existence of a substantial common issue of fact or law in relation to the claims of all members of the group against all respondents.

In light of the picture depicted above, it does not appear unreasonable to conclude that, unless the conduct of more than one person that is being challenged by a group of aggrieved persons entails a single transaction, act or event;^[177] a single document^[178] or, perhaps, a limited number of very similar transactions, acts or events,^[179] persuading the Court that the claims in question comply with the terms of s 33C(1) will be a difficult task indeed.^[180] Allowing this state of affairs to continue would be tantamount to an admission, on the part of the legislature^[181] and of the judiciary, that opponents of the class action procedure have been correct in asserting that the benefits which are associated with this procedure are more theoretical than real.^[182] Such an admission would constitute a tragic event for those who regard access to justice as a principle of cardinal importance.

---

^[*] Associate Professor, Department of Business Law and Taxation, Monash University; Co-Editor of the Journal of Australian Taxation. The author wishes to thank Associate Professor Peta Spender of the Australian National University for her comments and suggestions on an earlier draft of this article.

^[1] Peter Gordon and Lisa Nichols, 'The Class Struggle' (2001) 48 Plaintiff 6, 10.

^[2] Justice Stephen Charles, 'Class Actions in Australia' (Paper presented at the Australian Bar Association Conference, San Francisco, 18–21 August 1996) 32. An example of the judicial approach to class actions highlighted by Justice Charles is provided by the recent comments of Callinan J of the High Court in Mobil Oil Australia Pty Ltd v Victoria [2002] HCA 27, para 183: 'the problems to which I have just referred are likely to be aggravated by the increasingly competitive entrepreneurial activities of lawyers undertaking the conduct of class or group actions, in which, in a practical sense, the lawyers are often as much the litigants as the plaintiffs themselves, and with the same or even a greater stake in the outcome than any member of a group'.

^[3] See the Federal Court of Australia (Amendment) Act 1991 (Cth).

^[4] For a discussion of some of these issues, see Australian Law Reform Commission, Managing Justice—A Review of the Federal Civil Justice System, Report No 89 (2000) paras 7.87–7.128; S Stuart Clark and Christina Harris, 'Multi-Plaintiff Litigation in Australia—A Comparative Perspective' (2001) 11 Duke Journal of Comparative and International Law 289; Vince Morabito, 'Security for Costs and Class Actions in the Federal Court of Australia' (2001) 20 Civil Justice Quarterly 225; Jocelyn Kellam and S Stuart Clark, 'Multi-Party Actions in Australia' in Christopher Hodges (ed), Multi-Party Actions (2001) 269; Rod Freeman, 'Class Actions the Australian Way' (1999) 10 Australian Product Liability Reporter 109; Justice Murray Wilcox, 'Representative Proceedings in the Federal Court of Australia: A Progress Report' (1996-97) 15 Australian Bar Review 91; Peter Cashman, 'Consumers and Class Actions' [2001] UWSLawRw 2; (2001) 5 University of Western Sydney Law Review 9; Peta Spender, 'Securities Class Actions: A View from the Land of the Great White Shareholder' (2002) 31 Common Law World Review 123; Maggie Doyle, 'The Nature of Representative or Class Actions in the Context of Compensation Claims Against Resources and Utilities Companies' [1999] Australian Mining and Petroleum Law Association Yearbook 277.

^[5] Surprisingly, this significant issue has not been the subject of any extensive analysis in the legal literature. See, however, Jonathan Beach, 'Representative Proceedings—Some Current Issues' (Paper presented at a Seminar on Recent Developments in Class Actions, Melbourne, 12 October 2000) 18–29.

^[6] In November 2000 the Victorian Parliament introduced a legislative framework for class actions in the Supreme Court of Victoria which is almost identical to the pt IVA regime: Supreme Court Act 1986 (Vic) pt 4A. Consequently, the analysis developed in this article is equally relevant to the Victorian regime. The constitutional validity of pt 4A was upheld in June 2002 by the High Court in Mobil Oil Australia Pty Ltd v Victoria [2002] HCA 27. On the Victorian regime, see Beach, above n 5, 1-17; Schutt Flying Academy (Australia) Pty Ltd v Mobil Oil Australia Pty Ltd [2000] VSCA 103; (2000) 1 VR 545; Vince Morabito, 'Ideological Plaintiffs and Class Actions—An Australian Perspective' (2001) 34 University of British Columbia Law Review 459, 462; Cook v Pasminco Ltd [2000] VSC 534.

^[7] See Philip Morris (Australia) Ltd v Nixon [2000] FCA 229; (2000) 170 ALR 487, 514 (Sackville J) (‘Philip Morris’):

A proceeding is not properly commenced unless it satisfies each of the three threshold requirements specified in s 33C(1). If the proceeding does not comply with these requirements ... the proceeding is liable to be dismissed or the applicants' pleading struck out. (An alternative procedure was adopted in Silkfield v Wong, where the Full Federal Court made a declaration that the proceedings continue as proceedings brought by the respondents on their own behalf, to give effect to the majority holding that s 33C(1)(c) had not been complied with ...).

See also Femcare Ltd v Bright [2000] FCA 512; (2000) 172 ALR 713, 734 (Black CJ, Sackville and Emmett JJ); Silkfield Pty Ltd v Wong (1998) 159 ALR 329, 333–4 (Foster J); Wong v Silkfield Pty Ltd [1999] HCA 48; (1999) 199 CLR 255, 265–6 (Gleeson CJ, McHugh, Gummow, Kirby and Callinan JJ) (‘Silkfield’). See, however, Bright v Femcare Ltd [2000] FCA 1179, para 14 (Lehane J): 'I am not convinced that it is clear that a failure to comply with s 33C(1)(a), at the time a proceeding is commenced, is necessarily fatal to its continuance as a proceeding under Pt IVA'.

^[8] In Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 511 Sackville J explained that 'the expression "the same person" in s 33C(1)(a) is to be read as including more than one person (see Acts Interpretation Act 1901 (Cth), s 23(b))'.

^[9] The existing class action regimes in the US federal courts and in the Canadian provinces of Quebec, Ontario, British Columbia, Saskatchewan and Newfounland and Labrador all require certification of a class action: see, respectively, United States Federal Rules of Civil Procedure, Rule 23(c); Code of Civil Procedure of Quebec, arts 1002–1005 ('Quebec Act'); Class Proceedings Act, SO 1992, c 6, ss 2-8 ('Ontario Act'); Class Proceedings Act, RSBC 1996, c 50, ss 2-10 ('BC Act') ; The Class Actions Act, RSS 2001, c C-12.01, ss 4-12 ('Saskatchewan Act'); Class Actions Act, RSN 2001, c C–18.1, ss 3–11 ('NL Act'). In May 2002 the Attorney-General of the Canadian province of Manitoba introduced in that province's legislature the Class Proceedings Bill 2001 (Bill 16). A central component of this proposed modern class action regime is the certification model.

^[10] Australian Law Reform Commission, Grouped Proceedings in the Federal Court, Report No 46 (1988) ('Grouped Proceedings'). For more details, see Vince Morabito, 'Dinning v Federal Commissioner of Taxation—The Dawn of A New Era in Tax Litigation in Australia?' [2000] CanterLawRw 4; (2000) 7 Canterbury Law Review 487, 501–2.

^[11] It provides that where, at any stage of the class action, it appears likely that there are fewer than seven class members, the Court is empowered to order (a) that the proceeding continue as a class action or (b) that the proceeding no longer continue as a class action under pt IVA.

^[12] It empowers the Court to order the termination of a class action where the cost to the respondent of identifying the class members and distributing to them the damages won by the representative plaintiff would be excessive.

^[13] See Gold Coast City Council v Pioneer Concrete (Qld) Pty Ltd (Unreported, Federal Court, Drummond J, 9 July 1997).

^[14] As Kellam and Clark (above n 4, para 15.51) have indicated:

where the respondent believes it is not appropriate to have the matter dealt with as a representative proceeding ... the respondent will generally launch a multi-faceted attack asserting that (1) the proceedings do not satisfy the requirements of commonality prescribed by section 33C; and (2) that, for one or more of the reasons set out in subsection 33N(1), the proceedings should be terminated.

See also Vince Morabito and Judd Epstein, Class Actions in Victoria—Time for a New Approach (1997) (Report commissioned by the Victorian Attorney-General's Law Reform Advisory Council), para 6.16.

^[15] Silkfield Pty Ltd v Wong (Unreported, Federal Court, O'Loughlin and Drummond JJ, 17 December 1998), 3.

^[16] See Bright v Femcare Ltd [2000] FCA 1179, para 19 (Lehane J): 'it is an inevitable aspect of proceedings under Pt IVA, I should think, that in many cases a substantial number of members of the represented group will be unknown'; Australian Competition and Consumer Commission v Golden Sphere International Inc [1998] FCA 598; (1998) 83 FCR 424, 428 (O'Loughlin J).

^[17] Section 33E permits a class suit to be commenced by the representative plaintiff without the express consent of the class members. However, an opportunity is offered to the class members to exclude themselves from the class action; that is, to opt out (see s 33J). Section 33ZB provides that a judgment given in a representative proceeding binds all persons other than those who have opted out and requires that the class members affected by it be described or identified. Section 33ZB was recently described by the Full Federal Court as the 'pivotal provision of Part IVA': Femcare Ltd v Bright [2000] FCA 512; (2000) 172 ALR 713, 719 (Black CJ, Sackville and Emmett JJ).

^[18] [2000] FCA 229; (2000) 170 ALR 487, 514.

^[19] Ibid 514.

^[20] Grouped Proceedings, above n 10.

^[21] Ibid. For a discussion of the background to the enactment of pt IVA, see Vince Morabito, 'Class Actions—the Right to Opt Out Under Part IVA of the Federal Court of Australia Act 1976 (Cth)' [1994] MelbULawRw 5; (1994) 19 Melbourne University Law Review 615, 617.

^[22] Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 511–12.

^[23] Grouped Proceedings, above n 10, paras 64–5.

^[24] Ibid para 94.

^[25] Alberta Law Reform Institute, Class Actions, Report No 85 (2000) para 57.

^[26] UK, Civil Procedure Rules, Rules 19.10–19.15; and UK, Civil Procedure Rules, Practice Direction 19B—Group Litigation. For more details see Neil Andrews, 'Multi-Party Proceedings in England: Representative and Group Actions' (2001) 11 Duke Journal of Comparative and International Law 249; Hodges, above n 4, chs 1–8; Alberta Law Reform Institute, above n 25, paras 86–87; Robyn Trigge, 'Representative Actions under the Uniform Civil Procedure Rules' (2001) 21 Queensland Lawyer 110, 111.

^[27] Andrews, above n 26, 249.

^[28] See Lord Woolf, Access to Justice: Final Report (1996), ch 17; Lord Chancellor's Department, Multi-Party Situations: Draft Rules and Practice Direction (June 1999).

^[29] See Grouped Proceedings, above n 10, para 93:

the Constitutional requirement that the federal judiciary power only be exercised in respect of 'matters' [see Part III of the Commonwealth Constitution] would lead to difficulties if this course was followed for the Federal Court, or in respect of the exercise of federal jurisdiction. Those represented would be bound only if there was a 'matter' between them and the respondent ... In the Commission's view, there is considerable doubt whether there would be a 'matter' in the constitutionally required sense if the persons to be bound were not parties to the proceeding.

^[30] No reasons were furnished by the Government of the day for its rejection of the ALRC's grouped proceedings model. A leading commentator welcomed this move: 'this recommendation was, rightly, not adopted in Pt IVA ... since the effect of its adoption would have been to set at nought one of the essential characteristics and advantages of representative proceedings' (Michael Tilbury, 'The Possibilities for Class Actions in Australian Law' (Paper presented at the Australian Legal Convention, Hobart, 1993), 2 n 7). It is interesting to note that Senator Durack, the then Deputy Leader of the Opposition, lamented that 'it is very bland and misleading to say that the Government has not adopted the Commission's group proceedings. It may be that that is literally true, given the special definition in the report. The fact of the matter is that the Government has adopted a proceeding which is basically the same proposal': Commonwealth, Parliamentary Debates, Senate, 13 November 1991, 3020. See also Courtney v Medtel Pty Limited [2002] FCA 957, para 36 (Sackville J).

^[31] A constitutional challenge to the validity of pt IVA was rejected by the Full Federal Court in Femcare Ltd v Bright [2000] FCA 512; (2000) 172 ALR 713. An application for special leave to appeal to the High Court from the judgment of the Full Federal Court has been lodged by Femcare Ltd. However, the chances of this application being granted have been diminished somewhat by the fact that the Federal Court has recently made an order under s 33N that the proceedings against Femcare Ltd be terminated as pt IVA proceedings: see Bright v Femcare Ltd [2001] FCA 1477; Transcript of Proceedings, Femcare Ltd v Bright (High Court of Australia (30 October 2001).

^[32] See King v GIO Australia Holdings Ltd [2000] FCA 617; (2000) 100 FCR 209, 221 (Moore J): 'I acknowledge, however, that the proposal of the [ALRC] for grouped proceedings was not adopted by Parliament ... and comments ... made by the [ALRC] ... must be approached with that in mind'. See also Nixon v Philip Morris (Australia) Ltd [1999] FCA 1107; (1999) 165 ALR 515, 543–4 (Wilcox J):

bearing in mind that, in drafting the Bill eventually presented to Parliament, the Government departed from the Commission's model in several important respects, it is a large assumption to ascribe to the Government all the Commission's views. Although the Commission's report undoubtedly provides relevant background to the Government's proposal, on matters of detail the Government must be allowed to speak for itself.

^[33] Lawyers for applicants in pt IVA suits have attacked Sackville J's reliance on the ALRC's report upon the following grounds: 'with respect, the text of Part IVA itself, which provides that once the threshold requirements are met, an action may be brought, does not require and is in fact inconsistent with the use Sackville J sought to make of the [ALRC] report' (Gordon and Nichols, above n 1, 10).

^[34] See the quote accompanying note 19 above.

^[35] (1997) 149 ALR 45, 48. It is fascinating to note that there is a difference of opinion among commentators as to how liberally s 33C(1) has been construed and applied by the Federal Court. See, for instance, Gordon and Nichols, above n 1, 8: '[there has been] an overly-rigorous reading of the threshold requirements allowing representative proceedings to be commenced, resulting in an outcome inconsistent with the legislative intent'; and Freeman, above n 4, 109. Clark and Harris, on the other hand, have indicated that 'the threshold requirements [of s 33C(1)] have been liberally interpreted by the courts such that they are easily satisfied in practice': above n 4, 307.

^[36] This is essentially the interpretation advocated by Beach. However, he argues that this interpretation can co-exist with the Philip Morris principle: Beach, above n 5, 23. But, as is demonstrated in Section IIIE below, this latter proposition is incorrect. The Philip Morris principle and the interpretation of s 33C(1)(a) advanced by Beach are mutually exclusive.

^[37] See, for instance, Hollick v Toronto (City) [2001] SCC 68, para 14 (McLachlin CJ); Western Canadian Shopping Centres Inc v Bennett Jones Verchere (2001) 201 DLR (4th) 385, 397 (McLachlin CJ); Mobil Oil Australia Pty Ltd v Victoria [2002] HCA 27, para 12 (Gleeson CJ); Note, 'Developments in the Law—The Paths of Civil Litigation: IV. Class Action Reform: An Assessment of Recent Judicial Decisions and Legislative Initiatives' (2000) 113 Harvard Law Review 1806, 1809–10; Manitoba Law Reform Commission, Class Proceedings, Report (1999), 23–30; Ontario Law Reform Commission, Report on Class Actions, Report No 48 (1982), 117–46; Vince Morabito, 'Federal Class Actions, Contingency Fees, and the Rules Governing Litigation Costs' [1995] MonashULawRw 10; (1995) 21 Monash University Law Review 231, 232; Scottish Law Commission, Multi-Party Actions, Report No 154 (1996), para 2.23.

^[38] Commonwealth, Parliamentary Debates, House of Representatives, 14 November 1991, 3174 (Michael Duffy, Attorney-General). Similar reasoning was embraced by the Victorian Attorney-General, Mr Hulls, when he introduced in the Victorian Parliament Victoria's counterpart to pt IVA: Victoria, Parliamentary Debates, Legislative Assembly, 31 October 2000, 1252. It is interesting to note that Peter Gordon from Slater & Gordon, the firm which has acted for applicants in many pt IVA actions, expressed the view that the class action procedure should not 'be considered an appropriate legislative vehicle for ... cases ... where the plaintiffs have individually suffered significant damages': Peter Gordon, 'Class Actions: The Victorian Direction—the Plaintiff's Perspective' (Paper presented at a Seminar on Class Actions, Melbourne, June 1996), 3.

^[39] [2001] SCC 68.

^[40] Ibid paras 13-14 (McLachlin CJ). See also Femcare Ltd v Bright [2000] FCA 512; (2000) 172 ALR 713, 728, 730 (Black CJ, Sackville and Emmett JJ); Silkfield (1999) 100 CLR 255, 266–7 (Gleeson CJ, McHugh, Gummow, Kirby and Callinan JJ); Johnson Tiles Pty Ltd v Esso Australia Pty Ltd [1999] FCA 56, para 49 (Merkel J); Johnson Tiles Pty Ltd v Esso Australia Pty Ltd (No 3) [2001] VSC 372, paras 32, 33, 50 (Gillard J).

^[41] Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 520–1.

^[42] Ibid 493.

^[43] It is true that various opportunities to produce appropriate pleadings were provided to the applicants. But these opportunities were granted to the applicants by the trial judge, before the applicants were 'exposed' to the significantly more restrictive approach adopted by the members of the Full Court. If, after having had the benefit of the guidance provided in the judgments of the Full Court as to what was required, the applicants continued, in the view of their Honours, to provide pleadings which displayed a failure to comply with s 33C(1) then, and only then, would the Court have been entitled, in the author's view, to put an end to the pt IVA suit.

^[44] Gordon and Nichols have formulated another criticism of Hill J's reasoning by arguing, essentially, that his Honour placed more emphasis on the rights of the respondents than on the rights of the class members: above n 1, 10.

^[45] Williams v FAI Home Security Pty Ltd (No 4) [2000] FCA 1925, para 39. See also Graham Barclay Oysters Pty Ltd v Ryan (No 2) [2000] FCA 1220, para 6 (Lee, Lindgren and Kiefel JJ); Commonwealth, Parliamentary Debates, House of Representatives, 26 November 1991, 3290 (P Duncan):

whilst people can say, rather facetiously, that any individual who has a right can exercise that right, we know that that is not the reality; the reality is that ordinary people, as consumers or injured workers, for example, are not able to exercise their rights in the courts as more well-to-do people are able to. This legislation seeks to enable those people who have a common cause of action to club together so that they are able to have a matter litigated and, hopefully, result in a satisfactory outcome.

^[46] Before the Full Federal Court handed down its judgment in Philip Morris, another attempt, which also failed, was made to bring a pt IVA proceeding against the tobacco companies: Tobacco Control Coalition Inc v Philip Morris (Australia) Ltd [2000] FCA 1004. As Francey, who appeared on behalf of the applicant in the latter action, explained:

[the Tobacco Control Coalition Inc] proceeding was designed to 'dovetail' with the Nixon proceeding in so far as it did not purport to represent smokers with disease but rather was brought on behalf of health and medical organisations and smokers not yet diagnosed with smoking-related disease in order to establish a fund under Section 87 of the Trade Practices Act 1974 (Cth).

Neil Francey, 'Tobacco Litigation: The Australian Experience in a Global Context' (Paper presented at the World Health Organisation Consultation on Litigation & Public Inquiries as Public Health Tools for Tobacco Control, Amman, Kingdom of Jordan, 5–7 February 2001), 14.

^[47] Transcript of Proceedings, Nixon v Philip Morris (Australia) Ltd (High Court of Australia, Tobin QC, 21 June 2000) 5, 6. The High Court (Gleeson CJ and Callinan J; Kirby J dissenting) refused an application for special leave to appeal from the judgment of the Full Federal Court.

^[48] Ibid 10. Professor Pengilley has also drawn attention to the impact of this judgment on the behaviour modification goal of class actions: Warren Pengilley, 'Representative Actions Under the Trade Practices Act: The Lessons for Smokers and Tobacco Companies' (2000) 8 Competition & Consumer Law Journal 176, 179.

^[49] (1994) 112 DLR (4th) 504.

^[50] See S John Page, 'Class Actions in Canada: How They Work and Their Impact on Health Organizations and Businesses' (2000) 21 Health Law in Canada 1, 11: 'this case was decided in 1994. In view of the more flexible and liberal approach taken for certification by the Ontario courts, it is possible that a different result would be reached if the matter were considered for certification today'. See also Garry D Watson, 'Class Actions: The Canadian Experience' (2001) 11 Duke Journal of Comparative and International Law 269, 270 n 6: 'subsequently, two judges who presided at lengthy HIV trials have questioned the wisdom of the decision in Sutherland'.

^[51] Manitoba Law Reform Commission, above n 37, 27.

^[52] Watson, above n 50, 270.

^[53] See Bright v Femcare Ltd [2000] FCA 742, para 81 (Lehane J); Batten v CTMS Ltd [2000] FCA 915; Hunter Valley Community Investments Pty Ltd v Bell [2001] FCA 201 and [2001] FCA 1148; Sereika v Cardinal Financial Services Ltd [2001] FCA 1715; Milfull v Terranora Lakes Country Club Limited [2002] FCA 178.

^[54] [2000] VSC 534.

^[55] [2000] FCA 229; (2000) 170 ALR 487, 492–3:

there is a difficulty in representative proceedings being brought where some of the respondents are what may be called principal offenders by virtue of breaches of s 52 of the Trade Practices Act 1974 committed by those respondents on the one hand and others have an accessory liability on the other. That problem is compounded where the possibility arises that the so-called common issue involves in one case one group of respondents as principals and another group of respondents with accessory liability and in another a quite different group of respondents as principals and another differently constituted group of respondents with accessory liability ... The difficulties are both legal and practical and would suggest that it would be preferable that each case be separately pleaded and tried, rather than an attempt be made to combine each case into a representative application.

^[56] Ibid 522:

I should add that there may be a further question raised by the applicants' contentions. It is by no means clear that s 33C(1)(a) can be satisfied where the statement of claim pleads that some group members have claims under the TP Act against one respondent (A) and others against another respondent (B), even if A and B are each alleged to have aided and abetted the other's contravention of the TP Act.

^[57] As quoted in Commonwealth, Parliamentary Debates, House of Representatives, 26 November 1991, 3293.

^[58] [1991] FCA 20; (1990-91) 99 ALR 601. See also Commonwealth, Parliamentary Debates, Senate, 13 November 1991, 3027.

^[59] (1990–91) [1991] FCA 20; 99 ALR 601, 605.

^[60] Ibid 606:

if there were in force in this court provisions relating to grouped proceedings, such as those recommended by the [ALRC] ... it would have been possible for the court to determine all common questions of fact or law at a single hearing in such a manner as to make the result binding on all applicants and all respondents ... The result would have been to avoid the repetition in each of the later cases of most of the evidence in this case, with consequential savings in costs and the earlier finalisation of the whole litigation. But that recommendation has not become law. So it will be necessary to deal with each of the cases separately.

^[61] [1997] FCA 541; (1997) 72 FCR 1 (‘McMullin’).

^[62] See Vasram v AMP Life Limited [2000] FCA 1676, para 9 (Stone J): '[if s 33C(1) has not been complied with] then the proceeding is not properly commenced and is liable to be dismissed or the applicant's pleading struck out. The matter then is not one of the court's discretion under s 33N but of a failure to meet the threshold requirements of s 33C'.

^[63] The only class members who had claims against both States were, of course, those who were running relevant businesses in both States.

^[64] [1997] FCA 541; (1997) 72 FCR 1, 7-8.

^[65] Ibid, McMullin v ICI Australia Operations Pty Ltd [1999] FCA 1814.

^[66] See Clark and Harris, above n 4, 313:

even where threshold tests are satisfied at the commencement of a class action, Australian courts will be willing to entertain a Section 33N strike-out application later in the proceedings. In this sense, after the issues in dispute have been properly clarified, Section 33N provides courts with an 'escape hatch' through which the issues may pass in order to terminate inappropriate class actions.

^[67] (1997) 149 ALR 261 (‘Symington’).

^[68] Ibid 264.

^[69] (1997) 149 ALR 45 (‘Ryan’).

^[70] Ibid 48.

^[71] Ibid 48–9.

^[72] [1999] FCA 1107; (1999) 165 ALR 515, 528.

^[73] [2000] FCA 154 (‘Schneider’).

^[74] Ibid paras 1 and 2 (Mathews J) (emphasis added).

^[75] Ryan v Great Lakes Council [1999] FCA 177, para 5. Beach QC, who appeared on behalf of one of the applicants in Ryan, revealed (above n 5, 22) that:

in fact, in Ryan a subsequent application for leave to add additional applicants representing different groups making claims against different respondents was granted on the logic that the same result could in any event be achieved by having two separate proceedings consolidated—thus there was no good reason to prevent such a result through the more direct route of an application to amend by adding new applicants and new groups and modifying the claims.

^[76] Graham Barclay Oysters Pty Ltd v Ryan [2000] FCA 1099, para 84 (emphasis added).

^[77] As was noted by Gordon and Nichols, above n 1, 12: 'the fact that some matters have proceeded to judgment (even those including numerous defendants) demonstrates that Part IVA cases are capable of delivering what they were intended to deliver: determination of common issues in an economic manner'.

^[78] One commentator has colourfully indicated that 'it is clear that representative proceedings are a complex and often unwieldy animal': Doyle, above n 4, 288. See also Tiemstra v Insurance Corporation of British Columbia (1996) 22 BCLR (3d) 49, 61 (Esson CJ): 'class actions have the potential for becoming monsters of complexity and cost'.

^[79] This is, of course, an important consideration: 'fairness is a "two-way" concept which must include fairness for defendants as well as plaintiffs' (Kellam and Clark, above n 4, para 15.83).

^[80] The Supreme Court of Canada has recently indicated that:

in British Columbia, Ontario, and Quebec, a class action may proceed only after the court certifies that the class and representative meet certain requirements. In Alberta, by contrast, courts effectively certify ex post, only after the opposing party files a motion to strike. It would be preferable if the appropriateness of the class action could be determined at the outset by certification.

Western Canadian Shopping Centres Inc v Bennett Jones Verchere (2001) 201 DLR (4th) 385, 399 (McLachlin CJ). For a summary of the perceived benefits of the certification model, see South African Law Commission, The Recognition of Class Actions and Public Interest Actions in South African Law, Report—Project 88 (1998), para 5.5.5. Cf Grouped Proceedings, above n 10, para 147.

^[81] See Western Canadian Shopping Centres Inc v Bennett Jones Verchere (2001) 201 DLR (4th) 385, 401 (McLachlin CJ):

the class must be capable of clear definition. Class definition is critical because it identifies the individuals entitled to notice, entitled to relief (if relief is awarded), and bound by the judgment. It is essential, therefore, that the class be defined clearly at the outset of the litigation. The definition should state objective criteria by which members of the class can be identified.

See also JA Prestage and S Gordon McKee, 'Class Actions in the Common Law Provinces of Canada' in Hodges, above n 4, 223, 226.

^[82] (1993) 14 OR (3d) 734.

^[83] Ibid 735 (Montgomery J).

^[84] Ibid.

^[85] Ibid 744. See also Bywater v Toronto Transit Commission 1998 Ont Sup CJ Lexis 254, 14-15 (Winkler J); Endean v Canadian Red Cross Society (1997) 148 DLR (4th) 158, para 58 (Smith J):

however, the object of the [BC] Act is not to provide perfect justice, but to provide a 'fair and efficient resolution' of the common issues. It is a remedial, procedural statute and should be interpreted liberally to give effect to its purpose. It sets out very flexible procedures and clothes the court with broad discretion to ensure that justice is done to all parties.

^[86] 2000 Ont Sup CJ Lexis 2279.

^[87] Ibid 5. See also Millard v North George Capital Management Ltd 2000 Ont Sup CJ Lexis 874, 54 (Farley J). See, however, Ragoonanan v Imperial Tobacco Canada Limited [2000] OJ No 4597.

^[88] The certification requirements are set out in s 4 and are similar to the certification requirements set out in s 5 of the Ontario Act. The only major difference is that, unlike the Ontario Act, s 4(2) of the BC Act lists five factors which the Court must consider in determining whether the class action device would be the preferable procedure for the resolution of the common issues.

^[89] (1996) 22 BCLR (3d) 97 (‘Harrington’).

^[90] Ibid para 51.

^[91] (1997) 44 BCLR (3d) 343, para 42 (Cumming, Newbury and Huddart JJ); leave to appeal to the Supreme Court of Canada denied—[1998] SCCA No 13. In this case the plaintiffs sued four distinct classes of defendants: three designers/manufacturers; three distributors; one certification organisation; and nine public regulatory authorities. See also Collette v Great Pacific Management Co [2001] BCD Civ J 1100, para 52 (Macaulay J); Pausche v British Columbia Hydro & Power Authority [2000] BCD Civ J 130, paras 24-25 (Bauman J); Scott v TD Waterhouse Investor Services (Canada) Inc [2001] BCD Civ J 2644, para 32 (Martinson J); Pearson v Boliden Ltd [2001] BCD Civ J 2453, para 71 (Burnyeat J): 'a representative plaintiff must not necessarily have a cause of action against each defendant in order to certify a proceeding as a class proceeding ... Accordingly, it is also the case that members of a subclass need not necessarily have a cause of action against each of these Defendants'.

^[92] Consequently, Canadian courts have been able to 'achieve a balance between increasing access to justice and avoiding inappropriate and burdensome litigation': Lord Chancellor's Department, Representative Claims: Proposed New Procedures (2001), para 18.

^[93] This phenomenon in Canada also applies beyond multiple respondents suits. See, for instance, Watson, above n 50, 285:

on the whole, the Canadian judiciary seems to have recognised that class actions have an important and valuable role to play by affording access to justice to many who could never in their wildest dreams hope to litigate individually ... On the issue of whether class actions have been a good thing for the citizenry of Canada, if the 'proof of the pudding is in the eating' one could ask the class members who shared in the proceeds of the class actions settled or litigated to date (eg, the ... $140 million recovery in the vanishing premium cases and the $1.5 billion recovery in the Hepatitis C litigation), keeping in mind that without a class action regime, it is unlikely that any of this money would ever have been recovered in Canada. On the other hand, there is little if any evidence to date of the 'successful' use of class actions as a form of litigation blackmail.

See also Michael J Peerless and Michael A Eizenga, 'Class Actions in Breast Implant Litigation' (1996) 16 Health Law in Canada 78, 84; Lynn Pierce, 'Raising the Roof on Community Housing for People with Disabilities: Class Actions in Canada' (2000) 6 Appeal: Review of Current Law and Law Reform 22, 22.

^[94] (1993) 14 OR (3d) 734, 747 (Montgomery J). See also Webb v K-Mart Canada Ltd (1999) 41 OR (3d) 389, 402 (Brockenshire J); Harrington v Dow Corning Corp 2000 BCCA 605, paras 67-68 (Huddart JA):

from an individual plaintiff's perspective, a class proceeding is probably the only way she might have a chance to press her claim effectively. The cost of a risk assessment in resources of time and money would burden even the plaintiff with extremely serious injuries. For those with more modest claims the cost would be prohibitive ... As with pacemakers in Nantais v Telectronics Proprietary (Canada) Ltd (1995) 25 OR (3d) 331 ... toilet tanks in Chace v Crane Canada Inc (1997) 44 BCLR (3d) 264 (CA), and heating panels in Campbell ... this case about breast implants seems ideally suited for resolution by a class action, in a multi-staged proceeding, with trials of both common and individual issues.

^[95] Watson, above n 50, 285.

^[96] Beach, above n 5, 22–3 (emphasis in original).

^[97] '[I]f there are two sets of claims against two sets of respondents, it may well be that each can be the subject of representative proceedings. It may even be that directions can be made for them to be heard together: Ryan v Great Lakes Council (1997) 149 ALR 45, 48, (Wilcox J). But they cannot both be the subject of the same representative proceedings': Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 514.

^[98] [2002] FCA 178 (‘Milfull’). Each of these four cases essentially entailed scenario (ii) above and the Court held that such a scenario was inconsistent with the Philip Morris principle.

^[99] [2000] VSC 534. For more details on this case see Morabito, above n 6, 482–4.

^[100] [2000] FCA 915 (‘Batten’).

^[101] [2001] FCA 201.

^[102] See, generally, Kellam and Clark, above n 4, paras 15.16–15.20; Grouped Proceedings, above n 10, paras 49–50.

^[103] [2000] FCA 915, para 12.

^[104] See, generally, Charles, above n 2, 11-15; Morabito and Epstein, above n 14, ch 5; Tilbury, above n 30, 6.

^[105] [2002] FCA 178, para 20.

^[106] Ibid para 22.

^[107] See Johnson Tiles Pty Ltd v Esso Australia Pty Ltd (No 3) [2001] VSC 372, para 37 (Gillard J):

[section 33Q] gives a wide power to the Court to give directions in relation to questions. In my view, the wide power given in sub-s 1 is not made subject to sub-s 2. Sub-section (2) merely indicates that the Court may, if it so desires, include directions establishing a sub-group. But that does not seem to me to qualify or read down the wide power given to the Court to decide other questions, which will not finally determine the claims of all group members.

^[108] See Report of the Attorney-General's Advisory Committee on Class Action Reform (Toronto, 1990), 33:

sub-classing is a process by which the larger class is divided into more distinct and representative groups. A sub-class will have an issue of law or fact common to itself and therefore requires separate representation in order to protect interests that it has separate from the larger class. Inherent in sub-classing is the need to ensure that the sub-class is not prejudiced by being in conflict with the larger classes' interests.

^[109] Section 14 of the Ontario Act provides that:

(1) in order to ensure the fair and adequate representation of the interests of the class or any subclass or for any other appropriate reason, the court may, at any time in a class proceeding, permit one or more class members to participate in the proceeding;

(2) participation under subsection (1) shall be in whatever manner and on whatever terms, including terms as to costs, the court considers appropriate.

See also s 15 of the BC Act, s 16 of the NL Act and s 17 of the Saskatchewan Act.

^[110] To the author's knowledge, there has been no judicial discussion and application of this provision. It is also interesting to note that this section was not based on any recommendations of the ALRC.

^[111] See Explanatory Memorandum, Federal Court of Australia (Amendment) Bill 1991 (Cth) para 26:

in some cases determination of the common issues in a representative proceeding will still leave some issues relating to the particular claims of group members to be determined. This section [s 33Q] enables the Court to provide for the most convenient method of resolving such issues by giving directions.

The ALRC provided the following example of the circumstances in which subgroups may need to be created:

in a principal proceeding claiming damages for a breach of the Trade Practices Act 1974 (Cth) s 52, some group members may have relied on a representation made by the respondent in relation to a product while others may claim only that the product was not of merchantable quality. A further principal applicant may need to be appointed to represent those group members who do not have claims under the Trade Practices Act 1974.

Grouped Proceedings, above n 10, 178.

^[112] See Explanatory Memorandum, Federal Court of Australia (Amendment) Bill 1991 (Cth) para 28:

after determining the common issues in a representative proceeding it may not be appropriate, in some cases, to deal with remaining issues in the same proceeding. An example would be a case where, after a determination of liability is made, there are remaining issues in relation to one or more group members which are complex and diverse. In such cases it may be more efficient for separate proceedings, limited to those remaining issues, to be brought either by individual group members or as a separate representative proceeding. This section [s 33S] enables the Court to give directions relating to the commencement and conduct of such an individual proceeding or representative proceeding.

^[113] Section 33ZG(c)(iv) provides that 'except as otherwise provided by this Part, nothing in this Part affects ... the operation of any law relating to consolidation of proceedings'.

^[114] In light of the High Court's refusal to grant the applicants in Philip Morris leave to appeal (see above note 47), this is a likely scenario.

^[115] [2000] FCA 1543 (‘King’).

^[116] Nixon v Philip Morris (Australia) Ltd [1999] FCA 1107; (1999) 165 ALR 515, 529–30 (Wilcox J).

^[117] Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 521.

^[118] Ibid.

^[119] (1999) 199 CLR 225, 259–60 (Gleeson CJ, McHugh, Gummow, Kirby and Callinan JJ).

^[120] The ALRC has recently noted that 'a lawyer for applicants in representative proceedings commented that "pleadings have the potential to undermine Part IVA—they are the new battleground"': ALRC, above n 10, para 7.101. See also Gordon and Nichols, above n 1, 12:

the more important point that ought to be stressed is that pleadings are a vehicle to allow the issues in the case to be understood and ventilated. They are not an end in themselves and should not become a battleground in which pedantry and endless etymological exposition become a priority over the efficient and fair adjudication of the real issues in a case.

The opposite view has been advanced by Pengilley, above n 48, 179. See, generally, Clark and Harris, above n 4, 313-17.

^[121] '[Pursuant to s 33N] the court will look beyond the factual similarities between the claims of applicants gathered together in a representative action and balance the advantages and disadvantages to both sides, and the court, of allowing it to proceed in that form': Justice Donnell Ryan, 'The Development of Representative Proceedings in the Federal Court' (1993) 9 Australian Bar Review 131, 137.

^[122] [2000] FCA 229; (2000) 170 ALR 487, 492.

^[123] The ability of courts presiding over class actions in Canada and the US to consider the merits of the litigation, as part of the process of determining whether the suit should be certified, as a class action, has been described as follows by the Manitoba Law Reform Commission (above n 37, 48):

the Quebec regime requires the certification judge to consider the merits of the claim in deciding a certification application ... For the most part, the American class action regimes provide that the applicant need not demonstrate the merits of the case in order to obtain certification, although some American jurisprudence does require an inquiry into the merits of the plaintiff's claim. Ontario [and] British Columbia ... merely require the court to find that the claim discloses a cause of action.

^[124] [2000] FCA 1543, para 6 (Wilcox, Lehane and Merkel JJ).

^[125] 'The result is that 33,000 of the 68,000 shareholders who declined the AMP offer and expressed interest in the representative proceedings can now have these proceedings continue as representative proceedings. No one can deny that such litigation is a formidable weapon against companies, their directors and their advisers': Warren Pengilley, '33,000 Shareholders Can Take Class Action Against GIO, its Directors and Advisers' (2001) 12 Australian Product Liability Reporter 14, 16. See also, Michael Duffy, 'Shareholder Representative Proceedings: Remedies for the Mums and Dads' (2001) 39(7) Law Society Journal 53.

^[126] [2000] FCA 1543, paras 6–7.

^[127] The other significant aspect of King is the ruling of Moore J, the primary judge, that the claims of the class members need not be identical: [2000] FCA 617; (2000) 100 FCR 209, 220–1. In refusing leave to appeal the Full Federal Court made no comment regarding Moore J's conclusion. In Cook v Pasminco Ltd [2000] VSC 534, para 47, Hedigan J of the Supreme Court of Victoria indicated that in King, Moore J 'appeared to place some limits on the decision' in Philip Morris. But in Hunter Valley Community Investments Pty Ltd v Bell [2001] FCA 201, para 58, Sackville J endorsed Moore J's conclusion. See also Johnson Tiles Pty Ltd v Esso Australia Pty Ltd (No 3) [2001] VSC 372, paras 42, 45, 48 (Gillard J).

^[128] See also Femcare Ltd v Bright [2000] FCA 512; (2000) 172 ALR 713, 734 (Black CJ, Sackville and Emmett JJ).

^[129] [1999] FCA 1107; (1999) 165 ALR 515, 529–30.

^[130] [2002] FCA 243.

^[131] Ibid para 2 (Merkel J) (emphasis added).

^[132] In Philip Morris the pleadings alleged that the illnesses suffered by the class members were caused by 'the conduct of any one or more of the respondents': [2000] FCA 229; (2000) 170 ALR 487, 499.

^[133] [2002] FCA 243, para 199.

^[134] See Wilcox, above n 4, 92–3:

Where the court is not satisfied that the claimants' circumstances are sufficiently related, it has no power to allow the action to continue as a representative proceeding ... The advantage of representative proceedings is that issues common to all the claimants can be determined at one hearing ... But an absence of 'relatedness' undermines the rationale of the proceeding; any hearing would necessarily degenerate into a jumbled trial of disparate actions.

^[135] Similarly, the Full Federal Court has relied on the terms of s 33C(2) to reach the conclusion that 'plainly, Parliament envisaged cases involving non-common material; that was not to be a disqualification from using Part IVA': Finance Sector Union of Australia v Commonwealth Bank of Australia [1999] FCA 1250; (1999) 166 ALR 141, 145 (Wilcox, Ryan and Madgwick JJ).

^[136] Silkfield [1999] HCA 48; (1999) 199 CLR 255, 267 (Gleeson CJ, McHugh, Gummow, Kirby and Callinan JJ).

^[137] [1993] FCA 489; (1993) 118 ALR 165.

^[138] Silkfield [1999] HCA 48; (1999) 199 CLR 255, 267 (Gleeson CJ, McHugh, Gummow, Kirby and Callinan JJ).

^[139] See, for instance, Chen Zhen Zi v Minister for Immigration and Ethnic Affairs [1994] FCA 985; (1994) 121 ALR 83, 88 (Black CJ, Lee and Heerey JJ); Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 522 (Sackville J).

^[140] [1993] FCA 489; (1993) 118 ALR 165, 184–5.

^[141] Ibid 185. This action was brought on behalf of all persons who had been refused refugee status between March 1992 and June 1993. It was submitted by the representative plaintiff that the principles of natural justice conferred upon all persons applying for refugee status a legal right to an oral hearing by the relevant decision-maker.

^[142] Ibid: 'If the application were to succeed, all group members would be entitled to the offer of an oral hearing by the decision-maker. ... If the application fails, then a principle applicable to each group member would be established, namely that there is no entitlement in any member of the group to an oral hearing by reason only of the fact that the member is an applicant for administrative review of the refusal to grant refugee status'. See also Tang Jia Xin v Minister for Immigration and Ethnic Affairs (Unreported, Federal Court, Wilcox J, 11 April 1996).

^[143] Conversely, 'the criteria in Quebec, especially "identical, similar, or related ..." were initially interpreted conservatively by the Courts, but since a judgment of the Court of Appeal in 1990, they have been interpreted more liberally': Rules Committee of the Federal Court of Canada, Class Proceedings in the Federal Court of Canada, Discussion Paper (2000), 42.

^[144] Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 523.

^[145] Silkfield [1999] HCA 48; (1999) 199 CLR 255, 266 (Gleeson CJ, McHugh, Gummow, Kirby and Callinan JJ).

^[146] Ibid 267.

^[147] Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 523.

^[148] Grouped Proceedings, above n 10, para 134.

^[149] (1996) 63 FCR 304, 311. Einfeld J held that s 33C(1)(b) had been complied with 'notwithstanding the differences in individual circumstances in the present case and the necessity at the hearing to consider the evidence on reliance and damages on an individual basis': 315. See also Tropical Shine Holdings Pty Ltd v Lake Gesture Pty Ltd (1994) 118 ALR 510, 516 (Wilcox J); Lek v Minister for Immigration, Local Government and Ethnic Affairs [1993] FCA 297; (1993) 43 FCR 100; Metcalfe v NZI Securities Australia Ltd [1995] ATPR 40,645. See also Johnson Tiles Pty Ltd v Esso Australia Pty Ltd [1999] FCA 56, para 61 (Merkel J):

the magnitude of the claims is a factor that arises from the magnitude of the 'mass wrong' alleged by the applicants. Obviously, the greater the extent of the alleged wrong, the greater will be the number of claimants and the magnitude of the claims ... it would be anomalous for such an action to be said to be inappropriate as a representative proceeding because of its magnitude and the practical difficulties which arise in achieving resolution of the claims.

^[150] Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 490.

^[151] Grouped Proceedings, above n 10, para 135.

^[152] Ibid para 63. Ironically, a representative proceeding concerning faulty surgical sterilisation procedures was discontinued as a pt IVA proceeding by the Federal Court in Bright v Femcare Ltd [2001] FCA 1477.

^[153] See E v Australian Red Cross (1990–91) [1991] FCA 20; 99 ALR 601, 606 (Wilcox J):

despite the common elements in the claims, there are differences in the circumstances under which the various applicants ... became HIV infected. Some applicants claim to have been infected as a result of a single post-operative or post-parturition transfusion. Some claim to be haemophiliacs who have received blood products on numerous occasions. The applicant whose case was discontinued ... was an infant infected as a result of breast feeding by her mother, who had received the virus in a post-parturition transfusion. Moreover, the date at which the particular applicants allegedly received the blood or blood product is an important matter. Most, if not all, claims relate to donations received during years 1983, 1984 and 1985.

^[154] See, for instance, Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 522–4 (Sackville J); Hunter Valley Community Investments Pty Ltd v Bell [2001] FCA 201, paras 68-72 (Sackville J); Cook v Pasminco Ltd [2000] VSC 534, paras 50-53 (Hedigan J).

^[155] Beach, above n 5, 26.

^[156] (1998) 159 ALR 329, 343-344 (O'Loughlin and Drummond JJ).

^[157] Silkfield [1999] HCA 48; (1999) 199 CLR 255, 267. For more details, see Warren Pengilley, 'What is a Class Action? The High Court Overturns the Full Federal Court in the Silkfield case' (1999) 15 Australian & New Zealand Trade Practices Law Bulletin 69; Morabito, above n 10, 495–8.

^[158] See Vasram v AMP Life Limited [2000] FCA 1676, para 13 (Stone J):

the 'substantial' element of the s 33C(1)(c) requirement is easier to meet since the decision of the High Court in Wong v Silkfield Pty Ltd ... established that to be substantial the claim need only be real or of substance. It need not be the major or core issue and it is not necessary to show that litigation of the common issue would be likely to resolve the claims of all group members wholly or to any significant degree.

It has been noted by Emmett J of the Federal Court that 'in one sense, the High Court has simply substituted different words for the words of the statute': Murphy v Overton Investments Pty Ltd [1999] FCA 1673, para 16.

^[159] See Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 524 (Sackville J); Hunter Valley Community Investments Pty Ltd v Bell [2001] FCA 201, paras 58-60 (Sackville J); King v GIO Australia Holdings Ltd [2000] FCA 1543, para 13 (Wilcox, Lehane and Merkel JJ); Cook v Pasminco Ltd [2000] VSC 534, para 54 (Hedigan J). See also Grouped Proceedings, above n 10, para 138; Harrington v Dow Corning Corp 2000 BCCA 605, para 24 (Huddart JA):

'common' means that the resolution of the point in question must be applicable to all who are to be bound by it. I agree with the appellants that to be applicable to all parties, the answer to the question must, at least, be capable of extrapolation to each member of the class or subclass on whose behalf the trial of the common issues is certified for trial by a class proceeding. As the appellants note, this requirement will, of necessity, require that the answer be capable of extrapolation to all defendants who will be bound by it.

^[160] [2001] FCA 201.

^[161] Ibid para 54.

^[162] Ibid para 62. In Campbell v Flexwatt Corporation (1997) 44 BCLR (3d) 343, para 53 (Cumming, Newbury and Huddart JJ), the Court of Appeal for British Columbia highlighted the fact that:

when examining the existence of common issues it is important to understand that the common issues do not have to be issues which are determinative of liability; they need only be issues of fact or law that move the litigation forward. The resolution of a common issue does not have to be, in and of itself, sufficient to support relief. To require every common issue to be determinative of liability for every plaintiff and every defendant would make class proceedings with more than one defendant virtually impossible.

^[163] See, for instance, Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 524 (Sackville J); Cook v Pasminco Ltd [2000] VSC 534, paras 54–9 (Hedigan J). See also Beach, above n 5, 27: 'one may be able to identify an issue common to all claims against a particular respondent, but it may be difficult (absent a conspiracy, joint tortfeasor or other analogous type situation) to identify such an issue common to all claims against all respondents'.

^[164] Hunter Valley Community Investments Pty Ltd v Bell [2001] FCA 201, para 57.

^[165] See Jocelyn Kellam and Madeleine Kearney, 'Product Liability—A Decade of Change' (2001) 12 Australian Product Liability Reporter 49, 56. This scenario is in stark contrast to the following comment of Foster J of the Federal Court concerning the impact of s 33C on potential applicants: 'in my opinion, s 33C provides a very wide gateway for the commencement of representative proceedings' (Silkfield Pty Ltd v Wong (1998) 159 ALR 329, 333–4).

^[166] There are similar rules in every Australian jurisdiction: see Vince Morabito, 'Taxpayers and Class Actions' [1997] UNSWLawJl 31; (1997) 20 University of New South Wales Law Journal 372, 374.

^[167] The differences between the class action procedure and the traditional representative action procedure have been described as follows:

modern class action statutes elaborate and improve upon the 'representative action' procedure. Like the historic action, in a modern class action a representative plaintiff conducts the proceeding on behalf of other persons. However, a class member's claim need only be similar to the representative plaintiff's claim; it need not be exactly the same. Here again, all members of the class are bound by the outcome on the common issues. However, provision is made for the formation of subclasses and the separate resolution of issues relating to individual class members in addition to the resolution of issues common to the main class or a subclass. A number of statutory safeguards and an expanded role for the court help to ensure that the interests of the class members are protected.

Alberta Law Reform Institute, above n 25, xix-xx.

^[168] See, generally, Morabito and Epstein, above n 14, ch 4; Jillaine Seymour, 'Representative Procedures and the Future of Multi-Party Actions' (1999) 62 Modern Law Review 564. According to Ormiston JA of the Court of Appeal of Victoria, the first representative suit took place 'almost precisely 800 years' ago: Schutt Flying Academy (Australia) Pty Ltd v Mobil Oil Australia Pty Ltd [2000] VSCA 103; (2000) 1 VR 545, para 39 n 41.

^[169] Silkfield [1999] HCA 48; (1999) 199 CLR 255, 266 (Gleeson CJ, McHugh, Gummow, Kirby and Callinan JJ); Mobil Oil Australia Pty Ltd v Victoria [2002] HCA 27, para 34 (Gaudron, Gummow and Hayne JJ).

^[170] [2000] FCA 458, para 7. A similar approach appears to have been taken by the Supreme Court of Canada in relation to Alberta's counterpart to O 6 r 13: Western Canadian Shopping Centres Inc v Bennett Jones Verchere (2001) 201 DLR (4th) 385.

^[171] The following comments of Kiefel J in Milfull v Terranora Lakes Country Club Limited [2002] FCA 178, paras 19 and 25 tend to suggest a possible return to the practice adopted by the Federal Court in those three cases:

it was first submitted for the applicant that the problem [created by the failure of the proceedings to adhere to the Philip Morris principle] could be got around by joining four other persons as additional applicants to these proceedings, to represent the sub-groups. ... The proceedings are at present not properly constituted as a representative proceeding. ... During submissions it was indicated that the applicant may bring an application for directions concerning the sub-groups already identified and as to their representation. If such an application is not made, those respondents [against whom the applicant has no personal claim] may apply to have the claim against them struck out.

^[172] Nixon v Philip Morris (Australia) Ltd [1999] FCA 1107; (1999) 165 ALR 515, 546 (Wilcox J):

if any stage a conflict of interest emerges, between particular classes of group members or particular individuals, that will not necessarily make it impossible or inappropriate to maintain the proceeding as a representative action. It might prove possible to meet any difficulty by the constitution of sub-groups, and the appointment of sub-group representatives.

See also Alberta Law Reform Institute, above n 25, para 165:

dividing plaintiffs into subclasses could lead to a more efficient resolution of claims than might be possible otherwise ... It could ... be useful to divide plaintiffs into subclasses for the purpose of assessing liability or damages where plaintiffs have obtained a defective product from different distributors who have made different representations about the product.

In Rumley v British Columbia [2001] SCC 69, para 42, the Supreme Court of Canada has explained that '[the BC Act] contemplates the possibility of subclasses ... [and] provides the court with ample flexibility to deal with limited differentiation amongst the class members as and if such differentiation becomes evident'.

^[173] See Wilcox, above n 4, 97–8:

the procedure has the potential to handle cases more efficiently than otherwise and to resolve cases that might otherwise remain unresolved. Its use will often require innovative answers to practical problems. Imaginative case management, and sensible attitudes by both bar and bench, will ultimately demonstrate that the representative proceeding provides a valuable addition to traditional procedures.

See also Johnson Tiles Pty Ltd v Esso Australia Pty Ltd [1999] FCA 56, para 62 (Merkel J):

the Federal Court's ample discretionary and procedural powers under the Rules of Court, the individual docket system of case management and Pt IVA (see for example ss 33Q, 33R and 33S) will ensure that the practical difficulties that may arise can be overcome.

^[174] '[The Court] will retain all its current powers including the power to stay or dismiss any application it considers to be vexatious, oppressive or trivial': Commonwealth, Parliamentary Debates, House of Representatives, 14 November 1991, 3175 (Michael Duffy, Attorney-General). As was recently explained by the Full Federal Court,

s 33ZG gives effect to the specific recommendation made by the [ALRC] that the Court's general powers to protect against abuse of process should apply to representative proceedings. The recommendation, in turn, was based on the policy that the Court must leave adequate power to ensure that the representative procedure is not abused or used inappropriately or inefficiently.

Femcare Ltd v Bright [2000] FCA 512; (2000) 172 ALR 713, 733–4 (Black CJ, Sackville and Emmett JJ).

^[175] For more details on the purpose of this provision, see McMullin v ICI Australia Operations Pty Ltd [1998] FCA 658; (1998) 156 ALR 257, 260 (Wilcox J). Similarly, s 12 of the BC Act provides that

the Court may at any time make an order it considers appropriate respecting the conduct of a class proceeding to ensure its fair and expeditious determination and, for that purpose, may impose on one or more of the parties the terms it considers appropriate.

See also s 12 of the Ontario Act, s 14 of the Saskatchewan Act and s 13 of the NL Act.

^[176] In fact, even with the Philip Morris principle in force, s 33N is more likely to be employed by the Court in multiple respondents suits than in single respondent suits: see Beach, above n 5, 28-29.

^[177] See, for instance, Johnson Tiles Pty Ltd v Esso Australia Pty Ltd [1999] FCA 56 (class proceedings with respect to the gas explosion in Victoria). See also Kirsty Sutherland, Ian Dallen and Matthew Flood, 'Class Actions: the Increasing Risk for Manufacturers and Suppliers' (2000) 3(10) Inhouse Counsel 113, 114; Clark and Harris, above n 4, 317.

^[178] See, for instance, King v GIO Australia Holdings Ltd [2000] FCA 617; (2000) 100 FCR 209 (concerning statements contained in a document issued by, among others, directors of a company); Patrick v Capital Finance Corporation (Australasia) Pty Ltd [2001] FCA 1073, para 13 (Heerey J):

I am satisfied that this is an appropriate case for a representative proceeding under Pt IVA. Indeed it seems typical of the kinds of claim which the legislature probably had in mind ... The complaint of the investors revolves around the one prospectus and what is said to be misuse of funds invested in a way contrary to what the prospectus stated.

^[179] See, for instance, Batten v CTMS Ltd [2001] FCA 1493; Philip Morris [2000] FCA 229; (2000) 170 ALR 487, 525 (Sackville J):

but a pleading of this kind, if it is to survive scrutiny, requires a considerable degree of commonality in the claims made by or on behalf of group members. That degree of commonality may well be present where, for example, the group members all claim to have brought a defective product in reliance on substantially the same misleading representation or to have suffered loss or damage by reason of a particular event caused by the respondent's negligence.

^[180] An effective way of highlighting the unsatisfactory nature of this scenario is by drawing attention to how this scenario bears a striking similarity to the views expressed by O'Loughlin and Drummond JJ of the Federal Court in Silkfield Pty Ltd v Wong (1998) 159 ALR 329, 347 as to what type of case should be allowed as a pt IVA proceeding:

the kind of case that can best be run as a representative proceeding is one arising out of a 'mass wrong', ie, out of a single act, omission or course of conduct or the same act, omission or course of conduct repeatedly made or engaged in.

This concept of a pt IVA case was essentially based on their Honours' extremely narrow construction of s 33C(1)(c); a construction which, as was shown in Section V above, was unequivocally rejected by the High Court.

^[181] Two recent developments demonstrate that the current government is unlikely to take any action to address the problems identified in this article. In the first place, it failed to implement the ALRC's recommendation that a review be commissioned by the Commonwealth Government on the operation of pt IVA: ALRC, above n 4, recommendation 81. In the second place, it recently legislated to prohibit the use of the pt IVA procedure in migration litigation: see Susan Harris, 'Another Salvo Across the Bow: Migration Legislation Amendment Bill (No 2) 2000 (Cth)' [2000] UNSWLawJl 60; (2000) 23 University of New South Wales Law Journal 208.

^[182] See, for instance, FG Hawke, 'Class Actions: the Negative View' (1998) 6 Torts Law Journal 70; Mobil Oil Australia Pty Ltd v Victoria [2002] HCA 27, paras 172 and 183 (Callinan J).