The predominance requirement in class actions is the Rule 23(b)(3) test that asks whether the legal and factual questions shared by the entire class outweigh the questions that would have to be answered member by member. A court applies it before certifying a damages class action, and it exists to separate cases where group treatment genuinely makes sense from cases where individual differences would fracture a single trial into hundreds of mini-disputes.
Where the Requirement Comes From
Rule 23(b) sets out three categories of class action. The first two cover situations involving a risk of inconsistent rulings across separate lawsuits, or cases seeking injunctive or declaratory relief that applies to the entire class. Neither requires predominance. Only the third category, Rule 23(b)(3), imposes it, along with a companion superiority requirement. That third category governs the vast majority of class actions seeking money damages, which is why predominance is the single most contested hurdle in class certification.
Predominance is not the same as the commonality requirement in Rule 23(a), and the difference matters. Commonality asks only whether there is at least one legal or factual question common to the class. Predominance asks whether those common questions dominate the case.
What the Standard Actually Demands
The Supreme Court in Amchem Products, Inc. v. Windsor described predominance as asking whether class members’ claims are “sufficiently cohesive to warrant adjudication by representation.” The questions that count are, in the Court’s words, “the legal or factual questions that qualify each class member’s case as a genuine controversy.” A shared desire for compensation does not count. A common interest in a favorable settlement does not count. What counts is the substance of the dispute itself: what the defendant did, whether the conduct was unlawful, and whether it caused harm to the group.
Amchem rejected a sprawling asbestos settlement class because individual differences among members (different products, different exposure periods, different injuries, different state laws) overwhelmed anything they shared. The predominance standard, the Court said, “is far more demanding” than commonality. Courts cannot soften it because a settlement looks fair or because the alternative is a flood of individual lawsuits.
When Common Questions Carry the Day
The strongest candidates involve a defendant’s uniform conduct that affected every class member in roughly the same way. An employer applying the same illegal pay policy to every worker. A manufacturer shipping a product with a standardized defect. A bank charging an undisclosed fee to every account holder. In each scenario, a single answer about the defendant’s behavior resolves the core liability question for everyone at once.
The Supreme Court sharpened this idea in Wal-Mart Stores, Inc. v. Dukes, explaining that the common question must be “capable of classwide resolution — which means 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.” When one answer resolves the central dispute, predominance is likely satisfied. When the answer depends on what happened to each individual, it probably is not.
Predominance does not require every aspect of the case to be identical. Minor variations in individual circumstances are expected. Courts also draw an important line between the fact of injury and the amount of damages. Differences in how much each member lost generally do not defeat predominance, because individual damage calculations can be handled after a common liability finding. But if the question of whether each member was actually harmed demands individualized proof, that is a merits question for every claim and can sink certification.
When Predominance Fails
Choice-of-Law Problems
One of the most reliable ways to defeat predominance in a nationwide class is to show that the laws of multiple states apply to different members. When a court would need to instruct a jury on consumer protection statutes from dozens of states, each with its own elements, defenses, and damage rules, the individual legal questions can overwhelm the common ones. Federal appellate courts have held that state-law variations “swamp any common issues” when the differences are substantial. Even divergence among a handful of states on a central legal question can force a proposed class into subclasses or defeat it entirely.
Individualized Injury Questions
A defective product that injured some users but not others illustrates the problem. The manufacturing flaw may be common, but causation is not. If proving injury requires looking at each user’s exposure, use, and medical history, those individual inquiries can outweigh the shared defect question. The same pattern appears in misrepresentation cases where reliance has to be shown person by person.
Proving Predominance with Evidence
Allegations are not enough at the certification stage. Plaintiffs typically gather internal corporate records, policy manuals, and communications showing that the defendant acted through centralized decision-making rather than scattered incidents. Expert witnesses, often economists and statisticians, build models to show that liability and damages can be measured on a group-wide basis. The quality of that expert work frequently decides whether a class is certified.
The Damages Model Has to Match the Liability Theory
In Comcast Corp. v. Behrend, the Supreme Court held that a plaintiff’s damages model must line up with the theory of liability that survives for class treatment. The plaintiffs advanced four antitrust theories but only one made it through to certification. Their damages expert measured harm from all four theories combined, without isolating the surviving one. The Court found that fatal, holding that “a model purporting to serve as evidence of damages in this class action must measure only those damages attributable to that theory.” The calculations need not be exact, but the model has to attempt to measure the right thing.
Statistical and Representative Evidence
In Tyson Foods, Inc. v. Bouaphakeo, the Court allowed workers to use a study of representative employees to estimate unpaid overtime across a wage-and-hour class, because the employer had failed to keep adequate time records and the same evidence would have been admissible in an individual lawsuit under the circumstances. The Court rejected “broad and categorical rules” either way. Whether representative proof works depends on the facts of each case, and models built on “implausible assumptions” will not hold up.
The Superiority Requirement Travels With It
Predominance alone does not get a (b)(3) class certified. The rule also requires a finding that a class action is the superior method for resolving the dispute fairly and efficiently. Courts weigh four factors:
- Whether class members have a strong interest in controlling their own separate lawsuits, which tends to be true when individual stakes are large enough to justify solo litigation.
- Whether related lawsuits are already pending, which can make a class action duplicative.
- Whether concentrating all claims in one forum is desirable or would create logistical problems.
- Whether the case would be manageable as a class action, considering how many individual issues might arise at trial.
When individual claims are too small to litigate alone, the classic “negative value” scenario where it costs more to sue than the claim is worth, superiority is usually straightforward. The class action exists for those situations. When individual stakes are high, superiority is harder to establish because each plaintiff has both the incentive and the means to go it alone. Predominance and superiority are separate findings, but in practice they rise and fall together: a case built on genuinely common questions is usually also the kind of case that belongs in a single forum.