Opinion
Civil Action No. CCB-20-1082
04-16-2021
NEHA SINGH, et al. v. LENOVO (UNITED STATES) INC.
MEMORANDUM & ORDER
This putative class action raises claims related to a line of "Yoga" computer devices manufactured and sold by Lenovo United States, Inc. ("Lenovo"). The plaintiffs allege Lenovo knowingly sold the Yoga devices with defective hinges in violation of various states' consumer fraud statutes and in breach of express and implied warranties. Before the court is Lenovo's motion to bifurcate class and merits discovery (ECF 44). The court has considered the motion as well as the plaintiffs' opposition (ECF 45) and, for the reasons described herein, will deny the motion.
Whether to order bifurcation, during discovery or at trial, is an issue squarely within the broad discretion of the district court. See Fed. R. Civ. P. 42(b). Generally, "bifurcation of discovery is the exception, rather than the rule, and it is clear that in most instances, regular—that is, unbifurcated—discovery is more efficient." Cardenas v. Resort Sales by Spinnaker, Inc., No. 9:20-cv-00376-RMG, 2021 WL 733393, at *1 (D.S.C. Feb. 24, 2021) (internal quotation marks omitted). At the same time, in the class action context, where a district court must determine whether an action should be certified as a class action at an "early practicable time," Fed. R. Civ. P. 23(c)(1)(A), the Manual for Complex Litigation states that "[d]iscovery relevant only to the merits delays the certification decision and may ultimately be unnecessary" if, for example, individual claims are unlikely to proceed if a class is not certified. Manual for Complex Litigation (Fourth), § 21.14. Thus, where merits bifurcation would cause significant duplication and expense, discovery should proceed concurrently. Id. § 11.213.
Unpublished opinions are cited for the soundness of their reasoning and not for their precedential value.
In the wake of the Supreme Court's decision in Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338 (2011), district courts have been "reluctant to bifurcate class-related discovery from discovery on the merits." Chen-Oster v. Goldman, Sachs & Co., 285 F.R.D. 294, 300 (S.D.N.Y. 2012); see also Cardenas, 2021 WL 733393, at *2 (quoting Ahmed v. HSBC Bank USA, Nat'l Ass'n, No. ED CV 15-2057 FMO (SPX), 2018 WL 501413, at *3 (C.D. Cal. Jan. 5, 2018)); Hunichen v. Atonomi LLC, No. C19-0615-RAJ-MAT, 2020 WL 5759782, at * 1 (W.D. Wash. Sept. 28, 2020). This is because Dukes admonished that district courts were to engage in a "rigorous analysis" to determine class certification requirements, an analysis that often overlaps with the merits of the underlying claims because "the class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiff's cause of action." 564 U.S. at 350-51 (internal quotation marks omitted); see also Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 455, 466 (2013) (clarifying that district courts are not to engage in "free-ranging merits inquiries at the certification stage[,]" but may consider merits questions to the extent they are relevant to determining whether the Rule 23 prerequisites for class certification are met). Given that the distinction between class certification and merits discovery is "murky at best and impossible to determine at worst," separating merits and class discovery "'raise[s] a slew of issues as to what discovery relates to the class, as opposed to the named plaintiffs, thereby causing additional litigation regarding the distinction between the two.'" Cardenas, 2021 WL 733393, at *2 (quoting Ahmed, 2018 WL 501413, at *3).
Lenovo argues that it is well established that bifurcation of discovery is the norm in class actions. But the cases it relies upon either predate Dukes or are otherwise distinguishable. In the unpublished opinion of Witten v. A.H. Smith and Co., for example, the Fourth Circuit Court of Appeals does state in a footnote that prior to class certification, discovery should be limited to issues of certification. No. 84-2269, 785 F.2d 306 (Table), at *4 n.6 (4th Cir. Feb. 13, 1986). Witten was an employment discrimination action in which the plaintiffs sought to certify a class of employees who had worked at any of the defendant's eighteen facilities. Id. at *1. When the defendants produced employment records from just three facilities but balked at producing records from all facilities, the plaintiffs sought to compel complete production. Id. The trial judge refused and the plaintiffs withdrew the class claims; then, the defendant moved for summary judgment and plaintiffs' counsel opposed the motion without relying on any of the records actually produced in discovery. Id. The Fourth Circuit stated that under other circumstances the plaintiffs may well have been entitled to see all the employment records held by the defendant because it would not be very expensive to do so, but in light of plaintiffs' counsel's abdication and refusal to utilize the discovery it had already obtained, there was no abuse of discretion in the trial judge's rulings. Id. at *1, *3. Thus, the lesson of Witten is not so much that bifurcation is the rule, but rather that courts should be mindful of their obligation both to determine class certification at an early stage and to actively manage discovery in proportion to the needs of a case.
The cases which post-date Dukes include Akselrod v. MarketPro Homebuyers LLC, No. CCB-20-2966, 2021 WL 100666 (D. Md. Jan. 12, 2021), and Reid v. Unilever U.S., Inc., 964 F. Supp. 2d 893 (N.D. Ill. 2013). Akselsrod involved a stay in merits discovery pending an imminent Supreme Court decision which might settle a disputed issue of law and dispose of the action. See 2021 WL 100666, at *2. And Reid, which did order bifurcation of discovery in a consumer class action, see 964 F. Supp. 2d at 933, rests on its specific facts and does not reveal bifurcation to be the norm, as other district courts in the same circuit have reached different conclusions. See, e.g., Denney v. Amphenol Corp., No. 19-cv-4757, 2020 WL 5500276, at *4 (S.D. Ind. Sept. 4, 2020); Beezley v. Fenix Parts, Inc., 328 F.R.D. 198, 202-04 (N.D. Ill. 2018).
With that context in mind, the parties agree that in ruling on this motion the court is to consider (1) the overlap between individual and class discovery, (2) whether bifurcation will promote Federal Rule of Civil Procedure 23's requirement that certification be decided "at an early practicable time," (3) judicial economy, and (4) any prejudice reasonably likely to flow from the grant or denial of a stay of class discovery. See 1 McLaughlin on Class Actions § 3:10 (17th ed. 2020). The court considers each factor in turn.
I. Overlap
The parties agree Rule 23's numerosity and adequacy requirements will require very little discovery and may be accomplished expeditiously. As for commonality, typicality, and predominance, the plaintiffs believe these inquiries will be inexorably linked to merits discovery. Lenovo admits there will be some overlap with merits discovery, but contends it will not be substantial and that bifurcation is justified. See Reid , 964 F. Supp. 2d at 933 (some overlap in merits and class certification issues insufficient by itself to overcome efficiency benefits of bifurcated discovery).
Specifically, Lenovo argues the discovery focused on commonality would involve only "a simple inquiry" and the production of "simple data," which it characterizes as:
[T]he named Plaintiffs' individual purchases, warranty claims, and alleged harm, along with simple data on putative class members, including the models purchased (as compared to those purchased by Plaintiffs), any warranty claims made (and the results thereof), similarities/differences of models and alleged defects, Lenovo's knowledge of the alleged defects, and related arguments about harm . . . . [As well as] the alleged omissions . . . relied on by the named plaintiffs and argument as to how this differs from putative class members.(ECF 44, Mot. to Bifurcate at 3). And the typicality inquiry would involve another "simple inquiry" similar to the commonality inquiry but with more focus on the disparity between the plaintiffs' alleged harms as opposed to the harms to putative class members. (Id.). At bottom, Lenovo believes this can be accomplished through written discovery and depositions and by reference to data on the models purchased, the warranty claims, and evidence of harm. (Id.). And with respect to predominance, Lenovo predicts that discovery would focus on the difficulty of applying the laws of up to fifty states to a nationwide class. (Id. at 4).
The plaintiffs counter that this is "more than minor overlap with merits discovery—it is merits discovery." (ECF 45, Opp. at 3). They note that their consumer fraud, express warranty, and implied warranty claims require them to establish a defect in Lenovo's devices and that Lenovo knew or should have known of the defect—and that to certify a class they will need to demonstrate these common issues are susceptible to classwide proof. (Id.). They believe that unique classwide proof is likely in Lenovo's possession and would inform any potential ruling on class certification. (Id.). They would like access not just to the data Lenovo believes is sufficient to complete class discovery, but also to email correspondence, internal reports, and testing results. (Id.). And as for predominance, the plaintiffs argue that inquiry would likely start not with variance in state laws but rather with the underlying causes of action, see City of Cape Coral Mun. Firefighters' Ret. Plan v. Emergent Biosolutions, Inc., HQ, 322 F. Supp. 3d 676, 686 (D. Md. 2018), which would require the court to examine evidence available for each element of a claim to determine whether "one stroke" adjudication is possible, see Dukes, 564 U.S. at 350.
The court agrees with the plaintiffs. The rigorous analysis at the class certification stage would require that the court analyze the causes of action alleged by the plaintiffs to determine whether they are susceptible to classwide proof. While expert testimony may be able to establish a defect, it is likely the plaintiffs will need access to some of Lenovo's records in order to determine whether Lenovo knew of the defect and sold the product anyway. It is difficult to see a way to cleanly and efficiently disentangle these inquiries from the merits of the plaintiffs' claims.
II. Deciding Certification at an Early Practicable Time
Lenovo argues that bifurcation would allow for an earlier determination of class certification and prevent the expense of merits discovery should a class ultimately not be certified, contending that the plaintiffs intend to seek very broad discovery. (ECF 44 at 4). In response, the plaintiffs refer to the parties' deliberations over the proposed schedule, in which Lenovo actually advocated for a later date for class certification than the plaintiffs did. (ECF 45 at 3). At this point, the parties appear to have agreed on March 2022 as the deadline for certification briefing. (Id.).
This is not a case where the parties have vastly different proposals as to the timeline for certification. Lenovo originally had the less optimistic view of how long it would take to complete discovery in advance of class certification, and the plaintiffs accommodated Lenovo's desire to brief class certification in March 2022. The court does not read the plaintiffs' response to suggest that all merits discovery must conclude prior to briefing class certification; rather, they represent that their discovery needs are not the bottleneck. Thus, if the court were to bifurcate discovery, there is no indication that the timeline for certifying a class would advance.
III. Judicial Economy
Lenovo contends that allowing merits discovery to proceed simultaneously may result in an expensive and lengthy process that will prove wasteful should a class never be certified. (ECF 44 at 4). It believes that judicial resources would be saved because the scope of potential discovery disputes would be limited during the first phase. (Id.). In response, the plaintiffs argue bifurcation would actually consume rather than conserve the court's resources. (ECF 45 at 4). They argue they have shown a willingness to cooperate with defense counsel and that if anything, the difficulty of drawing the line between merits and class discovery would likely place greater demands on the court's time. (Id.).
The parties in this action have different ideas about what constitutes merits discovery and what constitutes class discovery. This confirms, as some courts have held, that the line between the two is murky and that the court would be required to expend resources drawing those lines for the parties as disputes arise. And, at least after Dukes, the arbitrary insistence on distinguishing between class and merits discovery may "'thwart[] the informed judicial assessment that current class certification practice emphasizes.'" Cardenas, 2021 WL 733393, at *2-3 (quoting Ahmed, 2018 WL 501413, at *3).
IV. Prejudice
Lenovo argues that it will be prejudiced by having to finance costly discovery efforts unless the court bifurcates discovery because this action is unlikely to proceed if a class is not certified given that each plaintiff is asserting claims worth around $1,000. (ECF 44 at 5). The plaintiffs contend in response that Lenovo would not be prejudiced because of the great extent of overlap between class discovery and merits discovery and because avoiding the expense of litigating frequent disputes over what constitutes the former and the latter may ameliorate any added expense incurred by additional production. (ECF 45 at 5). Rather, the plaintiffs believe it is they who would be prejudiced by an adverse ruling, as they face a high burden to certify a class and it is often easy for a defendant to paint a dismal picture of a plaintiff's prospects at the class certification stage unless plaintiffs have the benefit of some merits discovery. (Id.). They also assert that they intend to proceed even if a class is not certified. (Id.).
Each side's arguments have merit. It may be that the now-five named plaintiffs would not proceed should a class fail to be certified. But it is also true that by artificially limiting discovery at an early stage, the plaintiffs will face a potentially unfair challenge in meeting the predominance and commonality requirements at the class certification stage. See Tait v. BSH Home Appliances Corp., 289 F.R.D. 466, 487 (C.D. Cal. 2012) (noting that defendants may unfairly diminish plaintiffs' prospects at class certification when plaintiffs have not had the benefit of discovery into the merits of a case and defendants have withheld the very information needed to prove plaintiffs' case). Any prejudice to Lenovo that may result from not bifurcating discovery may be minimized or negated by avoiding disputes over what constitutes class discovery and—should a class be certified—by the continuing relevance of merits discovery beyond the class certification stage. See Hunichen, 2020 WL 5759782, at *2.
***
Given the enmeshed nature of class and merits discovery, the unlikelihood that the date of a class certification decision will advance, the expected benefits to judicial economy, and the minimal prejudice it is hoped will result, the court declines to draw bright lines by ordering bifurcation. But neither is the court ordering that all merits discovery be completed prior to the class certification determination; indeed, it may be efficient to defer some merits discovery where it does not plausibly overlap with any of the Rule 23 certification requirements. Counsel are expected, as always, to collaborate and to look to proportionality as their guiding principle.
For the reasons stated herein, it is hereby ORDERED that the defendant's motion to bifurcate class and merits discovery (ECF 44) is DENIED. 4/16/2021
Date
/s/_________
Catherine C. Blake
United States District Judge