Arguing Class Actions is a monthly column by Adam J. Levitt for the National Law Journal.
Reprinted with permission from the July 6, 2026, edition of the National Law Journal. © 2026 ALM Media Properties, LLC. Further duplication without permission is prohibited. All rights reserved.
Predominance has become one of the most misunderstood concepts in modern class-action jurisprudence. What Rule 23(b)(3) frames as a comparative inquiry—whether common questions predominate over individual ones—has increasingly been recast as something far more rigid: a demand that every element of every claim be proven through identical evidence for every class member before certification may be granted. Variations in damages? Fatal. Differences in reliance? Fatal. Individualized injury inquiries? Fatal. While that shift sounds technical, it isn’t. It actually rewrites the rule.
Predominance is neither a test of uniformity, nor a demand for perfection. Rather, it’s a simple question of primacy: Will the common or the individual issues actually decide the case? Rule 23(b)(3)’s text makes this understanding unmistakably clear. Courts must only determine whether “questions of law or fact common to class members predominate over any questions affecting only individual members.” The rule assumes that individual issues will exist. It anticipates variation and directs courts to weigh its relative importance—not to eliminate all differences or conclude that differences among class members on certain issues doom certification.
Yet predominance has increasingly been treated as an element-by-element requirement of perfect identity. If reliance varies, certification fails. If damages differ, certification fails. If some class members may ultimately require individualized injury showings, certification fails. But that’s not Rule 23(b)(3) predominance; that’s perfection masquerading as rigor.
The proper inquiry is practical. If this case were tried tomorrow, what would the trial actually be about? Would it focus on the defendant’s uniform contract, its standardized disclosure, its centralized policy, its pricing algorithm, its product design? Would liability rise or fall for each class member on common corporate documents, common data, and common testimony? Or would the proceeding fracture into thousands of disconnected mini-trials untethered from any shared core?
In most modern consumer, antitrust, employment, and privacy cases, the answer is straightforward: At the heart of the case, the conduct is standardized and the evidence is centralized. The legality of that conduct is a common and predominant question in every meaningful sense. Individual damages may vary and certain defenses may require tailored treatment. But those issues are downstream, and they certainly don’t displace the central question: Was the defendant’s uniform course of conduct lawful?
The Supreme Court’s modern Rule 23 decisions, read together, rather than in isolation, reinforce this understanding.
Wal-Mart Stores v. Dukes, 564 U.S. 338 (2011), while often invoked as a cautionary tale, is better understood as a lesson in discipline. The plaintiffs in Wal-Mart challenged millions of discretionary decisions by thousands of managers. The court held that they failed to identify a common answer to the critical “why” question—what corporate policy caused the alleged discrimination, finding that there was no unifying practice generating common answers.
Wal-Mart, however, doesn’t stand for the proposition that variation defeats certification. Rather, it stands for the proposition that plaintiffs must identify the “glue” holding the case together. Where the challenged conduct is standardized—a uniform fee, a single misrepresentation, a centralized algorithm, a common product defect—that glue exists. The liability question becomes common not by abstraction, but by structure.
Significantly, Dukes doesn’t require that every class member experience identical effects, only that the legality of the defendant’s conduct be susceptible to common proof. That’s why courts continue to certify Rule 23(b)(3) classes post-Dukes where a defendant’s standardized conduct is the central issue, even if damages and certain defenses require individualized follow-through.
Comcast v. Behrend, 569 U.S. 27 (2013), has likewise been stretched far beyond its holding. Comcast didn’t hold that individualized damages defeat predominance, nor did it require plaintiffs to prove damages at certification. Rather, it required alignment: A damages model must measure damages attributable to the theory of liability certified for class treatment. The defect in Comcast wasn’t variability; it was mismatch.
That principle is nothing more than a straightforward instance of the need for a causal link between injury and damages. A price-inflation theory requires a model that isolates inflation. An overcharge theory requires a method that measures the overcharge. A uniform unlawful fee can be calculated directly from transactional data. When liability theory and damages model align, Comcast is satisfied—even if class members’ recoveries differ.
To convert Comcast into a merits trial at certification is to confuse methodological coherence with ultimate proof. The predominance inquiry asks whether a workable, classwide method exists—not whether plaintiffs have already won. Indeed, multiple courts of appeal have warned against turning Rule 23 into a front-loaded substitute for summary judgment. The point of class certification isn’t to decide who wins; it’s to decide how the case should be tried.
Tyson Foods v. Bouaphakeo, 577 U.S. 442 (2016), underscores a related point: Variation doesn’t defeat predominance where representative proof is appropriate. In Tyson Foods, because individual work time varied and records were incomplete, the plaintiffs relied on a time-and-motion study to establish average compensable time. The Supreme Court rejected a categorical ban on such representative evidence, explaining that its permissibility turns on whether it would be admissible in an individual action.
That reasoning carries well beyond wage-and-hour law. Modern litigation routinely relies on statistical and aggregate evidence. Courts don’t demand individualized reconstruction of every factual detail before permitting adjudication. The law doesn’t require fragmentation where the core wrongdoing is common, and representative proof would be used in any event.
TransUnion LLC v. Ramirez, 594 U.S. 413 (2021), introduces a distinct concern—Article III injury-in-fact—but not a new predominance standard. In TransUnion, the court held that class members who recover damages must demonstrate concrete injury. While that requirement governs recovery, it doesn’t require courts to conduct individualized standing trials before resolving the common liability issues. Nor does it transform Article III into a predominance barrier.
Courts retain practical tools to address injury at the appropriate stage: careful class definitions, subclasses, claims processes, and phased proceedings. The existence of individualized injury questions doesn’t negate the predominance of common liability issues. To hold otherwise would effectively insulate uniform misconduct whenever its harms are widespread but individually variable.
Read together, Wal-Mart, Comcast, Tyson Foods, and TransUnion demand coherence—not uniformity. Plaintiffs must articulate a unified course of conduct, support it with common or admissible representative proof, and present a manageable path forward. When those elements align, predominance is satisfied, even if some individualized issues remain.
Applying predominance to require identical proof for every element cannot be reconciled with Rule 23(b)(3)’s plain language or with ordinary litigation practice. Even in single-plaintiff cases, courts rely on inference, sampling, circumstantial evidence, and aggregate proof. They don’t demand microscopic symmetry before permitting a claim to proceed. Rule 23 doesn’t impose a higher evidentiary burden simply because claims are aggregated.
Predominance ultimately asks a simple but powerful question: What will decide this case? If the answer is the defendant’s uniform conduct, proven through common evidence, then common questions predominate. Individual damages calculations or injury determinations don’t displace that reality; they follow it. When predominance is inflated into a demand for perfection, however, Rule 23(b)(3) ceases to function in the very cases that it was designed to address. In that situation, the result isn’t heightened fairness; rather, it’s the practical elimination of accountability for systemic wrongdoing.
Federal judges need not choose between rigor and realism. While the Supreme Court’s precedents require disciplined alignment of theory and proof, they don’t require impossibility. Where common questions form the backbone of the dispute—where they determine whether the defendant’s conduct was lawful—predominance is satisfied.
Rule 23(b)(3) asks whether common questions predominate, not whether they stand alone. The rule was crafted to permit the coherent resolution of systemic disputes. It was never intended to demand perfection where primacy will do.
Rule 23 doesn’t require perfection; it requires predominance—and predominance means what drives the case, not what varies at the edges.
Adam J. Levitt is a founding partner of DiCello Levitt, where he heads the firm’s class action and public client practice groups. He can be reached at alevitt@dicellolevitt.com.
Thank you to DiCello Levitt associate Andrea Shang for contributing to this column.
