Dianne Castano v. The American Tobacco Company

Rule 23(b)(3) requires a rigorous determination that common questions predominate and that a class action is superior to other methods of adjudication. Conditional certification does not relax those requirements. A court may later revise a certification order, but it cannot use conditional certification to postpone deciding whether Rule 23 has been substantially satisfied at the time of certification.

In a multistate class action, differences in governing state law can overwhelm common questions. Before deciding predominance, the court must identify the substantive legal issues that control the claims and determine which jurisdictions’ laws apply. The class proponents, not defendants, bear the burden to make a credible and extensive showing that legal variations can be managed, such as by grouping materially similar state-law rules.

The proposed class asserted numerous causes of action on behalf of millions of people across every state. State laws differed in material ways on fraud and reliance, duties to disclose, products liability, comparative fault and assumption of risk, negligent infliction of emotional distress, consumer-protection statutes, punitive damages, and other issues. Those differences could alter both the elements plaintiffs must prove and the defenses available to defendants.

The district court acknowledged that the parties had only briefly addressed conflicts of law, yet it treated the inquiry as premature. Its cursory review of surveys and other cases did not explain how the many legal differences could be handled in this case. Nor did it undertake a genuine manageability assessment. A generalized concern about repetitive litigation, drawn from a much smaller one-state asbestos case, could not substitute for case-specific analysis of this sprawling nationwide action.

The district court believed that Eisen v. Carlisle & Jacquelin and Miller v. Mackey International barred it from looking beyond the pleadings at certification. The Fifth Circuit held that this misread those cases. They prohibit deciding certification based on whether plaintiffs are likely to win on the merits; they do not prohibit probing issues intertwined with the merits when that inquiry is needed to decide whether Rule 23’s requirements are met.

Predominance is not established merely because a common issue appears in every individual case. The court must determine whether the common issues are a significant part of each case when compared with individual issues. Without understanding how an addiction-as-injury claim would be presented and defended at trial, the district court could not know whether defendants’ allegedly common conduct would be significant enough to predominate.

The district court itself recognized that injury in fact, causation, reliance, affirmative defenses, and compensatory damages were overwhelmingly individualized. Each smoker’s history, knowledge, reasons for smoking, exposure to alleged misrepresentations, addiction, injury, and comparative fault could differ. Those individual questions had to be weighed against the proposed common questions rather than set aside simply because the court labeled the latter ‘core liability.’

The court also wrongly postponed consideration of reliance in the fraud claims. Under Fifth Circuit precedent, a fraud class cannot be certified where reliance will require individual proof. Although some state laws may permit an inference of reliance for omissions, the district court had to determine, rather than assume, whether reliance would actually be individualized. Otherwise, it risked holding a class trial and later discovering that the class did not satisfy predominance.

Rule 23(c)(4) permits a court to sever particular issues for class treatment, but it is a case-management device rather than a means of manufacturing predominance. The court must assess the common issues chosen for class treatment against the individual issues that remain in the overall cause of action.

If a court could isolate any shared question and certify it without considering the rest of the claim, predominance would disappear as a meaningful limitation. Virtually every case containing one common issue could be certified. That interpretation would conflict with Rule 23(b)(3), particularly in mass-tort litigation where liability, causation, defenses, and damages often vary substantially by claimant.

Here, severing defendants’ conduct and punitive damages did not cure the individual nature of reliance, causation, comparative negligence, addiction, injury, and damages. The district court therefore could not rely on issue certification to avoid deciding whether the plaintiffs’ claims, considered as a whole, were predominated by common questions.

Superiority asks whether a class action is the fair and efficient way to resolve the controversy. The proposed class posed exceptional management difficulties: individualized choice-of-law decisions, variations across eight claims and many state-law regimes, potential subclasses, notice to millions of people, determination of nicotine dependence, and separate proceedings for causation, defenses, and damages. The district court could not treat those problems as outweighed by a speculative prediction that millions of individual cases would inevitably be filed.

The court rejected the premise that this case was a negative-value suit. Individual claims could produce substantial damages, including punitive damages in many jurisdictions, and some statutes permitted fee shifting. Plaintiffs’ own counsel represented that individual suits would be brought if certification failed. Where individual recoveries are sufficiently large to support litigation, the benefits of aggregating claims are weaker.

The asserted tort was novel and untested: no court had tried an addiction-as-injury claim of the sort alleged. Without a record of individual cases, the district court had no reliable basis to determine the significance of common issues, typical injuries, levels of damages, workable trial procedures, or whether a litigation crisis would arise. Individual and smaller-group litigation would allow claims, defenses, causation issues, and governing state laws to develop before a court considered nationwide aggregation.

The proposed bifurcation also created a substantial Seventh Amendment concern. A class jury would decide defendants’ conduct, while later juries would decide comparative negligence and other individual matters. Comparative negligence requires comparison of plaintiff and defendant conduct, creating a serious risk that later juries would reexamine factual issues already decided by the first jury. That risk further undermined the claim that the proposed class structure was superior.

The court concluded that individual adjudication would allow plaintiffs to narrow claims, permit state courts to develop their own law on the novel addiction theory, and generate the experience needed for any future aggregation decision. Because the district court’s procedural errors could not be meaningfully corrected while the tort remained immature, the Fifth Circuit reversed and remanded with instructions to dismiss the class complaint.