United States: Class Action Defence

Class actions are ‘an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only’. ¹ In the United States, the class action device allows hundreds, if not thousands, of plaintiffs to combine their claims into a single lawsuit, where one or more named plaintiffs purport to represent the interests of and litigate the case on behalf of absent class members. In large-scale antitrust litigation, where damages are aggregated, trebled and therefore often exceed hundreds of millions of dollars, class certification becomes ‘the defining moment’ in the case – the denial of class certification ‘may sound the “death knell” of the litigation on the part of plaintiffs’, whereas the grant of class certification may ‘create unwarranted pressure to settle nonmeritorious claims on the part of defendants’. ²

To obtain class certification, plaintiffs must satisfy the four requirements in Rule 23(a) of the Federal Rules of Civil Procedure (numerosity, commonality, typicality and adequacy) and at least one of the requirements in Rule 23(b). ³ Plaintiffs bringing antitrust class actions usually proceed under Rule 23(b)(2) (injunctive relief) and 23(b)(3) (damages).

In addition to the express requirements of Rule 23, many courts require as ‘an essential prerequisite of a class action . . . that the class . . . be currently and readily ascertainable based on objective criteria’. ⁴ The threshold requirement that the class be ‘ascertainable’ is not codified in the Federal Rules of Civil Procedure; rather, whether the class members are readily ascertainable is a logical antecedent question that the court must address before reaching the formal requirements of Rule 23. Other courts, however, do not require that the class be ascertainable based on objective criteria, and some permit unnamed class members to self-identify as belonging to the class. ⁵

The US Supreme Court has emphasised that it ‘may be necessary for the court to probe behind the pleadings before coming to rest on the certification question’ and that certification is proper only if the trial court is satisfied, after a ‘rigorous analysis’, that the prerequisites of Rule 23 have been satisfied. ⁶

Numerosity – Rule 23(a)(1)

Rule 23(a)(1) requires the putative class to be ‘so numerous that joinder of all members is impracticable’. ⁷ The Supreme Court has noted that a class of 15 is ‘too small to meet the numerosity requirement’. ⁸ Lower courts have found that classes of 40 are sufficient to meet the numerosity requirement. ⁹ Where the number falls in-between, some courts have adopted a non-exhaustive list of factors to determine whether joinder is feasible, including: ‘judicial economy, the claimants’ ability and motivation to litigate as joined plaintiffs, the financial resources of class members, the geographic dispersion of class members, the ability to identify future claimants, and whether the claims are for injunctive relief or for damages’. ¹⁰ ‘While all factors are relevant . . .  judicial economy and the ability to litigate as joined parties are of primary importance.’ ¹¹

In In re Androgel Antitrust Litigation, ¹² the court denied certification to a proposed class of 33 direct purchasers of the drug Androgel, because the ‘large’ and ‘sophisticated’ companies that made up most of the proposed class could proceed through joinder. In the Eleventh Circuit, ‘generally less than twenty-one [class members] is inadequate, more than forty adequate, with numbers between varying according to other factors’. ¹³ The court found that a putative class of 33 members ‘places it squarely in the middle of the Eleventh Circuit’s gray area’, which required the court to look to a list of factors, including the ‘ease of identifying [the proposed class’s] numbers and determining their addresses, facility of making service on them if joined and their geographic dispersion’. ¹⁴ The court found that it would not be impracticable for the proposed plaintiffs to proceed through joinder. While geographically dispersed, the ‘vast majority’ of class members had ‘revenue of at least tens of millions of dollars per year’, and at least 10 of which had ‘revenue in the billions’. ¹⁵ Additionally, two-thirds of the class members had treble damages claims over US$1 million, and only six had treble damages claims under US$100,000. ¹⁶ Accordingly, these ‘very large, sophisticated companies with very large claims’ were unlike typical class members. ¹⁷ Moreover, most class members had ‘litigated similar actions before’, showing they had ‘the means and the motivation’ to proceed through joinder. ¹⁸ Therefore, class certification was denied, because the class was not ‘so numerous that joinder of all members [was] impracticable’. ¹⁹

Commonality – Rule 23(a)(2)

Rule 23(a)(2) requires that each class member’s claim involves ‘questions of law or fact common to the class’. ²⁰ ‘What matters to class certification . . . is not the raising of common “questions” . . . but, rather the capacity of a class-wide proceeding to generate common answers. . . . Dissimilarities within the proposed class are what have the potential to impede the generation of common answers.’ ²¹ ‘That common contention, moreover, must be of such a nature that . . . determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke.’ ²² Unlike the predominance requirement in Rule 23(b)(3), which asks whether individual or common issues will predominate over the case (discussed below), Rule 23(a)(2) only analyses whether common questions exist at all – ‘a single question of law or fact common to the members of the class will satisfy the commonality requirement’. ²³

In In re Capacitors Antitrust Litigation, ²⁴ the court certified a class of direct purchasers of capacitors – which are ‘used to temporarily store and even out the flow of electrical energy’ in electronic devices ²⁵ – that likely consisted of uninjured class members. The defendants argued that the plaintiffs’ expert failed to ‘calculate separate overcharges for class members on an individual basis’, and instead merely compared ‘each customer’s actual prices with the “but-for” predicted prices for each customer, where the “but-for” price was based purely on . . . aggregate overcharge percentages’, indicating that some putative class members might not have been injured at all. ²⁶ The court disagreed, because the defendants’ argument would require the plaintiffs to ‘prove that each and every putative class member was harmed before certification can be granted’. ²⁷ ‘Rule 23’, the court noted, ‘does not require proof of impact on each purchaser before a class can be certified’. ²⁸ The court also found that the expert’s method was supported by another expert’s testimony, documents indicating an agreement was afoot and prior criminal guilty pleas, all allowing for class-wide proof of impact. ²⁹ The court thus denied defendants’ commonality argument and certified the class.

Typicality – Rule 23(a)(3)

Rule 23(a)(3) requires that ‘the claims or defenses of the representative parties are typical of the claims or defenses of the class’. ³⁰ ‘The representative party’s interest in prosecuting his own case must simultaneously tend to advance the interests of the absent class members’. ³¹ Typicality should fail if ‘it cannot be said that a class member who proves his own claim would necessarily prove the claims of other class members’. ³² ‘To determine whether a plaintiff’s position is markedly different from the class as a whole’, the court will look to whether: the representative’s legal theory advanced, and its underlying facts are, the same as those of the class; the class representative is subject to a unique defence; and the representative’s ‘interests and incentives’ are ‘sufficiently aligned with those of the class’. ³³

In In re Automotive Parts Antitrust Litigation, ³⁴ the court denied certification to a class of direct purchaser plaintiffs (DPPs) of automotive bearings, in part, on typicality grounds. The plaintiffs argued that manufacturers of steel ball and roller bearings coordinated their price increases in response to price increases imposed by steel suppliers, and rigged bids and allocated markets in response to requests for quotations issued by automotive original equipment manufacturers (OEMs) and component manufacturers, and annual price reduction (APR) requests made by OEMs.

The court found that ‘the disparities in the circumstances’ under which named and unnamed plaintiffs purchased the products and ‘the divergent means through which Defendants allegedly conspired’ rendered named plaintiffs’ claims ‘not typical of the claims asserted by absent class members’. ³⁵ Moreover, the named plaintiffs lacked the incentive to introduce at trial ‘evidence . . . of bid rigging or APR collusion for customized bearings purchased by the dominant members of the putative class’. ³⁶ In particular:

the named [plaintiffs] only have the incentive to focus their trial efforts on the [steel price increase] issues that allegedly did affect them, and to pursue expert damages methodologies that result in a significant portion of any purported overcharges being attributed to the [steel price increase] agreements, as opposed to either bid rigging or APR collusion with respect to particular OEMs and tier suppliers. ³⁷

‘[T]he evidence of Defendants’ alleged bid-rigging and collusion in response to APR requests will be of no assistance in the named DPPs’ effort “to prove that list prices paid by ordinary purchasers such as themselves were fixed”’. ³⁸ Accordingly, the court denied certification, in part, because the named plaintiffs’ claims were atypical of the unnamed class members’ claims.

Adequacy – Rule 23(a)(4)

Rule 23(a)(4) requires named plaintiffs to ‘fairly and adequately protect the interests of the class’. ³⁹ The purpose of the adequacy requirement is to avoid the potential for conflict between named plaintiffs and unnamed class members. ⁴⁰ To defeat class certification, the conflict of interest must be fundamental and go to a specific issue in controversy. ⁴¹ Adequacy can fail if the representative’s economic incentives and objectives differ from those of the class members. ⁴²

In the antitrust context, courts have identified several circumstances where intra-class conflicts arise that render named plaintiffs inadequate representatives of the class, including: where a defendant’s challenged conduct benefitted certain absent class members; ⁴³ where the named plaintiff and absent class members are competitors; ⁴⁴ where named plaintiffs seek to represent indirect purchasers at different parts of the supply chain, because who actually bore the overcharge becomes a conflict within the class; ⁴⁵ and where the named plaintiff pursues a different legal strategy than could be pursued by absent class members. ⁴⁶

In In re Automotive Parts Antitrust Litigation, ⁴⁷ plaintiffs advanced three theories of liability, one theory for named plaintiffs’ claims and two theories for unnamed class members’ claims. The court found that the named plaintiffs were three ‘relatively small distributors . . . primarily engaged in the business of purchasing and reselling aftermarket bearings’, ⁴⁸ and they lacked the incentive to pursue theories of recovery that would support unnamed plaintiffs’ claims. The named plaintiffs ‘lack[ed] the incentive to pursue claims based on Defendants’ alleged bid-rigging or APR collusion’ and therefore could not ‘show that they will adequately represent the interests of class members that complain of injury as a result of these allegedly unlawful activities’. ⁴⁹ The court continued that:

To the contrary, [the named plaintiffs] have an incentive to see that the damages proven at trial are attributed as much as possible to unlawful activities that impacted bearings purchasers like them. . . . Given this likelihood that the interests of the named [plaintiffs] might diverge from the interests of other class members, the named [plaintiffs] cannot be deemed adequate representatives of the putative class. ⁵⁰

The court thus denied certification.

Injunctive Relief – Rule 23(b)(2)

Rule 23(b)(2) allows for class certification for injunctive or declaratory claims if ‘the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole’. ⁵¹ Courts generally deny certification of injunctive classes where the primary form of relief sought is damages. ⁵²

Damages Relief – Rule 23(b)(3)

Putative classes seeking damages must establish that ‘questions of law or fact common to class members predominate over questions affecting only individual members, and . . . a class action is superior to other available methods for the fair and efficient adjudication of the controversy’. ⁵³

Predominance

The predominance requirement is similar to, but more demanding than, the commonality requirement – it requires not only that common issues exist, but that those issues be more predominant than those affecting only individual class members. ⁵⁴ Predominance focuses on whether common questions ‘can be proved through evidence common to the class’, and whether the issues capable of resolution through generalised proof are more substantial than the issues subject only to individualised proof on a given question. ⁵⁵ Some courts also consider whether plaintiffs can prove with common evidence: the existence of an antitrust violation; that class members suffered antitrust injury as a result of the conduct; and measurable damages. ⁵⁶

In In re Asacol Antitrust Litigation, ⁵⁷ the First Circuit reversed the district court’s certification of a class of indirect purchasers because plaintiffs could not prove injury-in-fact on a class-wide basis. ⁵⁸ A few months before Warner Chilcott’s patent protection for its drug Asacol expired, which treats mild to moderate ulcerative colitis, and while its drug Asacol HD was still under patent protection, Warner pulled Asacol from the market and ‘simultaneously introduced a similar but not exactly identical substitute drug called Delzicol’, which had years left on its patent. ⁵⁹ Plaintiffs alleged that Warner’s practices (frequently called product-hopping or evergreening) precluded the entry of ‘lower-cost alternative[s] to . . . Delzicol and Asacol HD’. ⁶⁰

Based on both parties’ experts, the district court presumed that approximately 10 per cent of the class members would have opted for Asacol HD or Delzicol even if a generic version of Asacol was introduced, ⁶¹ meaning that ‘ten percent of the class had not suffered any injury attributable to defendants’ allegedly anticompetitive behavior’. ⁶² Yet, the district court concluded that a claims administrator could remove the de minimis number of uninjured class members and certified a class of all Asacol purchasers who ultimately switched to Delzicol or Asacol HD. ⁶³ The First Circuit found this certification approach ‘at odds’ with Supreme Court and Circuit precedent and reversed. ⁶⁴

First, the court of appeals rejected the plaintiffs’ argument that a claims administrator could feasibly remove the uninjured class members. There was no showing of an administrable method for defendants to challenge the affidavits of unnamed (potentially uninjured) class members, and acceding the authority to adjudicate such a crucial element of the plaintiffs’ claims to a claims administrator would run the risk of ‘jettisoning’ the rules of evidence, the rules of civil procedure, the Seventh Amendment, the Due Process Clause and the Rules Enabling Act. ⁶⁵ ‘[T]his is not a case’ where a small number of class members can be ‘picked off in a manageable, individualized process at or before trial’, it ‘is a case in which any class member may be uninjured, and there are apparently thousands who in fact suffered no injury’. ⁶⁶ The need to identify the uninjured individuals will predominate and render adjudication unmanageable. ⁶⁷

Next, the court rejected the plaintiffs’ argument that the Supreme Court’s decision in Tyson Foods, Inc v Bouaphakeo ⁶⁸ permits plaintiffs to use representative evidence like affidavits to establish class members’ injuries. ⁶⁹ There, an expert calculated the average time employees spent ‘donning and doffing’ protective equipment for the purpose of establishing all class members’ Fair Labor Standards Act claims. ⁷⁰ However:

using the average time it takes a person to don and doff clothes to estimate how long it takes a given individual to do so . . . is quite different than saying . . . that a given person wore certain clothes merely because most but not all others did so. . . . [The plaintiffs’ expert’s] estimate that a generic drug would achieve roughly ninety percent market penetration, if used to prove that each individual would have likely purchased the generic drug and was thus injured . . . leads to the demonstrably wrong conclusion that one hundred percent of individuals were injured. ⁷¹

Accordingly, the First Circuit reversed.

Likewise, in In re Thalomid & Revlimid Antitrust Litigation, ⁷² the court denied certification to a consumer class that alleged that Celgene delayed the entry of generic versions of its blockbuster drugs Thalomid and Revlimid. Plaintiffs alleged that Celgene violated antitrust laws ‘by listing and suing to enforce invalid patents, refusing to sell samples necessary to develop generics, and encouraging the FDA to reject generic applications based on sham safety concerns’. ⁷³ The court denied certification because the plaintiff did not offer class-wide proof of antitrust impact, and class membership and variations in state law required individualised attention that would predominate at trial. ⁷⁴

The court agreed with the defendants’ expert, who concluded that up to 10 per cent of the proposed consumer class was brand loyal and would have purchased the branded drug even if a generic was available. ⁷⁵ The court found ‘that identifying these [uninjured] members would require extensive individualized inquiry’. ⁷⁶ The court disagreed with the plaintiffs’ reading of Nexium, Solodyn and Lidoderm ⁷⁷ as permitting uninjured class members to be sorted out at the damages stage. Solodyn and Lidoderm both relied on Nexium, but ‘the First Circuit [recently] limited the reach of Nexium, bringing it into closer consistency with the Third Circuit’s standards for certification’. ⁷⁸ The court also agreed that there exists ‘significant differences between the states’ consumer protection statutes and plaintiffs’ causes of action. These distinctions ‘lessen the predominance of common legal issues’ and ‘would demand significant attention from this Court, not the least of which would be instructing the jury or juries consistent with the law of each relevant states’. ⁷⁹ Thus, the court denied certification on predominance grounds while allowing the plaintiffs to renew their motion, which they did on 14 December 2018.

Finally, in In re Automotive Parts Antitrust Litigation, ⁸⁰ the plaintiffs advanced a single theory of conspiracy ‘to raise bearings prices’ and argued that this ‘was accomplished through several mechanisms, including rigging bids, fixing prices, and allocating markets for bearings’. ⁸¹ The defendants, however, argued that the plaintiffs ‘allege[d] three separate types of antitrust liability and impact in this case: (i) bid rigging of specific RFQs . . . (ii) collusion with respect to APR requests . . . and (iii) five separate [steel price increase] agreements’. ⁸² The defendants argued that ‘these three separate theories of liability give rise to the very same problem that defeated class certification in Comcast Corp: a disjunction between a party’s theory of liability and antitrust impact and its theory of damages’. ⁸³

The court agreed:

If this suit goes forward, nothing would prevent the trier of fact from concluding that Defendants engaged in one or two of these price-fixing mechanisms . . . but not the others. Under this scenario, the only class members that could claim antitrust impact and resulting damages would be those that purchased bearings through the means that were subject to conspiratorial activity. . . . [I]n other words, each of the three price-fixing mechanisms identified by the [plaintiffs] is independent from the others, in terms of both Defendants’ capacity to invoke it and its effect upon the relevant subset of Defendants’ customers. DPPs and their experts claim only that there is a correlation in the prices paid for bearings across various segments of the putative class. ⁸⁴

However, ‘evidence of correlation does not establish causation’. ⁸⁵ The court therefore denied certification.

Superiority

To certify a damages class under Rule 23(b)(3), plaintiffs must also show that ‘a class action is superior to other available methods for the fairly and efficiently adjudicating of the controversy’. ⁸⁶ Courts examine four factors: class members’ interests in individually controlling the prosecution or defence of separate actions; the extent and nature of any related litigation already begun by or against class members: the desirability or undesirability of concentrating the litigation in the particular forum; and likely difficulties in managing a class action. ⁸⁷

Certification of classes consisting of uninjured class members

In 2018, several courts were confronted with the question of whether to certify a class that consists, or at least potentially consists, of uninjured class members. As in all litigation in federal court, antitrust plaintiffs must show that they suffered an ‘injury-in-fact’. ⁸⁸ However, certification of classes that consist of uninjured class members ‘has been the source of much debate among the circuits’. ⁸⁹ The Seventh Circuit has permitted certification so long as the number of uninjured class members is not ‘a great many’. ⁹⁰ How many uninjured class members, or what percentage of the class, constitutes ‘a great many’ has not been defined. Nonetheless, from July 2018 to June 2019, two courts denied certification where there was evidence that at least some class members were uninjured, while three courts granted certification faced with similar evidence.

In In re Lamictal Indirect Purchaser & Antitrust Consumer Litigation, ⁹¹ the court found ‘no significant presence of uninjured’ class members. ⁹² In finding that plaintiffs met the predominance inquiry, the court relied on In re Lidoderm, In re Solodyn and In re Nexium for the proposition that the presence of uninjured class members does not defeat certification, because the uninjured members can be sorted out at the damages stage by the court or by a claims administrator. However that line of cases has been called into question by other district courts within the Third Circuit.

In In re Thalomid & Revlimid Antitrust Litigation, ⁹³ the court denied certification of a class consisting of uninjured members, and in doing so, disagreed with the plaintiffs’ reading of In re Nexium, In re Solodyn and In re Lidoderm, ⁹⁴ as permitting uninjured class members to be sorted out at the damages stage. ⁹⁵ The court pointed out that the courts in Solodyn and Lidoderm both relied on Nexium in reaching the conclusion that a class consisting of uninjured members can be certified. ⁹⁶ However, ‘the First Circuit subsequently limited the reach of Nexium, bringing it into closer consistency with the Third Circuit’s standards for certification’. ⁹⁷

In In re Asacol Antitrust Litigation, ⁹⁸ the First Circuit reversed the district court’s decision to grant certification, in part, because it was ‘not a case where a small number of class members can be picked off in a manageable, individualized process at or before trial’; rather, it was ‘a case in which any class member may be uninjured’, and identifying those uninjured class members would predominate and render adjudication unmanageable. ⁹⁹ ‘[A]fter all, one does not ordinarily set out to find a needle in a haystack by examining only ten percent of the straw’. ¹⁰⁰

In contrast, in In re Capacitors Antitrust Litigation, ¹⁰¹ the court rejected the defendants’ argument that the plaintiffs’ expert ‘did not calculate separate overcharges for class members on an individual basis’, but instead merely compared ‘each customer’s actual prices with the “but-for” predicted prices’ based on ‘aggregate overcharge percentages’ for three different capacitor products, indicating that some class members might not have been injured. ¹⁰² The court reasoned that ‘Rule 23 does not require proof of impact on each purchaser before a class can be certified’. ¹⁰³ Moreover, the expert’s method was supported by another expert’s testimony, documents indicating an agreement was afoot and prior criminal guilty pleas, all of which demonstrated that class-wide proof of impact was available. ¹⁰⁴

Similarly, in Wortman v Air New Zealand, ¹⁰⁵ the court granted certification to two damages classes of transpacific airline passengers and rejected the defendants’ argument that ‘large segments’ of the classes ‘paid lower prices as a result of the price-fixing conspiracy, and therefore did not suffer an antitrust injury’. ¹⁰⁶ The court found that the presence of ‘a small number’ of uninjured class members ‘is not an obstacle to class certification’. ¹⁰⁷ The court was satisfied that, ‘for now, Plaintiffs have adequately demonstrated that “all or virtually all” of their proposed classes suffered an injury’. ¹⁰⁸

Courts appear to be drawing a distinction ‘between class members who were not harmed and those who could not have been harmed’. ¹⁰⁹ Where some class members might not have been harmed, courts may still grant certification if defendants offer no concrete evidence that class members were in fact uninjured. ¹¹⁰ And even faced with evidence that some class members could not have been harmed, courts may still grant certification if the number of uninjured class members is not ‘a great many’. ¹¹¹ However, where there is evidence that ‘any class member may be uninjured’, some courts will deny certification if identifying those uninjured class members is likely to predominate and render adjudication unmanageable. ¹¹²

Conclusion

Developments in antitrust class actions over the past year highlight the critical role that the certification decision plays. While the decision to grant or deny class certification continues to rest on the facts of a particular case, three lessons can be drawn from the major class certification decisions in 2018. First, since the Third Circuit’s landmark decision in In re Hydrogen Peroxide Antitrust Litigation, ¹¹³ courts have increasingly vetted the testimony of economists hired to prove class-wide impact and calculate multimillion-dollar damages claims. The Asacol and Automotive Parts cases continue this trend. Second, lower courts continue to apply and define the contours of the Supreme Court’s decisions in Comcast and Tyson Foods in determining whether class plaintiffs have satisfied the predominance inquiry, particularly as to proof of class-wide impact. Third, the question of whether courts should certify classes that contain uninjured class members (and, if so, how many uninjured class members may defeat class certification) continues to be an emerging issue in this area.

Notes