What did Comcast Corp. v. Behrend actually hold?
Comcast Corp. v. Behrend, 569 U.S. 27 (2013), held that a class action was improperly certified under Rule 23(b)(3) because the damages model offered to support certification did not measure the harm from the single liability theory that had been certified. The plaintiffs, cable subscribers in the Philadelphia area, pleaded four theories of antitrust impact from Comcast’s clustering of its systems; the district court accepted one, overbuilder deterrence, and rejected the other three. Their expert’s regression compared actual prices with hypothetical but-for prices; the Supreme Court of the United States recorded that “[t]he model calculated damages of $875,576,662 for the entire class,” that it assumed the validity of all four theories, and that the expert acknowledged the model “did not isolate damages resulting from any one theory of antitrust impact.”
The reasoning is compressed into three sentences. “We start with an unremarkable premise. If respondents prevail on their claims, they would be entitled only to damages resulting from reduced overbuilder competition, since that is the only theory of antitrust impact accepted for class-action treatment by the District Court. It follows that a model purporting to serve as evidence of damages in this class action must measure only those damages attributable to that theory.” The Court added that “[c]alculations need not be exact,” citing Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 563 (1931), “but at the class-certification stage (as at trial), any model supporting a ‘plaintiff’s damages case must be consistent with its liability case, particularly with respect to the alleged anticompetitive effect of the violation.’” That last quotation is from a treatise, ABA Section of Antitrust Law, Proving Antitrust Damages: Legal and Economic Issues 57, 62 (2d ed. 2010).
The second half of the holding is procedural and often skipped. The Third Circuit had refused to consider Comcast’s attack on the model because doing so would reach the merits. The Supreme Court held that refusal “flatly contradicts our cases requiring a determination that Rule 23 is satisfied, even when that requires inquiry into the merits of the claim.”
Did Comcast hold that damages must be measurable on a class-wide basis?
No, and the majority opinion says why in a clause that decides most of the litigation about this case. Describing the standard applied below, the Court wrote that the district court “held, and it is uncontested here,” that to meet predominance the plaintiffs had to show both that antitrust impact was “capable of proof at trial through evidence that [was] common to the class rather than individual to its members” and that “the damages resulting from that injury were measurable ‘on a class-wide basis’ through use of a ‘common methodology.’” That requirement was a premise of the appeal rather than a question presented, so the Court had no occasion to adopt or reject it.
The dissent made the point explicitly, which is why it is quoted so often by plaintiffs. Justices Ginsburg and Breyer, joined by Justices Sotomayor and Kagan, wrote that “the opinion breaks no new ground on the standard for certifying a class action under Federal Rule of Civil Procedure 23(b)(3),” and that “[i]n particular, the decision should not be read to require, as a prerequisite to certification, that damages attributable to a classwide injury be measurable ‘on a class-wide basis,’” citing the majority’s own acknowledgment that the point was uncontested.
Both overstatements travel. The defense-side one is that Comcast requires a class-wide damages model in every Rule 23(b)(3) case. The plaintiff-side one is that Comcast is confined to its antitrust facts. Neither survives the text: the Court did not impose a class-wide measurability requirement, and the fit requirement it did state is not written in terms limited to antitrust.
What must a class damages model tie to the theory of liability?
It has to measure the harm the certified theory causes, and stop there. The failure in Comcast was not that the regression was unsophisticated but that it answered a broader question than the one the case had been narrowed to. The Court described the consequence in terms of what the model could establish rather than what it got wrong: “In light of the model’s inability to bridge the differences between supracompetitive prices in general and supracompetitive prices attributable to the deterrence of overbuilding, Rule 23(b)(3) cannot authorize treating subscribers within the Philadelphia cluster as members of a single class.”
Two practical consequences follow. First, a model can become non-conforming through no fault of its author, because the certified theory can narrow after the model is built; the four theories in Comcast were all pleaded, and three fell away at certification. Second, the defect is structural rather than evidentiary: a model that never attempted to separate the certified theory’s effect cannot be repaired by arguing about its coefficients.
What the model is built from is a separate subject, covered at Class-wide models, with the overcharge setting at Antitrust overcharge and inference from samples at Sampling and extrapolation. This Institute states no overcharge percentage and no class-wide total.
Do individualized damages defeat predominance?
Not on their own, in the circuits that have addressed it. The Second Circuit settled its own law in Roach v. T.L. Cannon Corp., 778 F.3d 401 (2d Cir. 2015), decided February 10, 2015. Vacating a denial of certification that rested on Comcast alone, the court stated the narrow reading directly: “Comcast held that a model for determining classwide damages relied upon to certify a class under Rule 23(b)(3) must actually measure damages that result from the class’s asserted theory of injury; but the Court did not hold that proponents of class certification must rely upon a classwide damages model to demonstrate predominance.”
Roach collected the other circuits, and the agreement is broad. The Fifth Circuit in In re Deepwater Horizon, 739 F.3d 790, 817 (5th Cir. 2014), described the “principal holding of Comcast” as the fit requirement and rejected the argument “that certification under Rule 23(b)(3) requires a reliable, common methodology for measuring classwide damages.” The Seventh Circuit in Butler v. Sears, Roebuck & Co., 727 F.3d 796, 799 (7th Cir. 2013), read Comcast as holding only “that a damages suit cannot be certified to proceed as a class action unless the damages sought are the result of the class-wide injury that the suit alleges.” The Tenth Circuit in In re Urethane Antitrust Litigation, 768 F.3d 1245, 1257 to 1258 (10th Cir. 2014), wrote that “Comcast did not rest on the ability to measure damages on a class-wide basis.”
That is a rule about predominance, not a license. Individualized damages remain a factor in the weighing, and the Supreme Court’s own sentence in Comcast, “Questions of individual damage calculations will inevitably overwhelm questions common to the class,” was written about a record where the plaintiffs had offered no conforming methodology at all. The distinction that matters is between a case where impact and the measure are common and only arithmetic differs by member, and one where establishing that a given member was harmed takes its own inquiry.
Where do the circuits actually disagree?
Over classes that contain members who suffered no injury, and how many of them a class may contain. The Ninth Circuit answered en banc in Olean Wholesale Grocery Cooperative, Inc. v. Bumble Bee Foods LLC, 31 F.4th 651 (9th Cir. 2022), decided April 8, 2022 in an opinion by Judge Ikuta, rejecting “the dissent’s argument that Rule 23 does not permit the certification of a class that potentially includes more than a de minimis number of uninjured class members” because “[t]his position is inconsistent with Rule 23(b)(3), which requires only that the district court determine after rigorous analysis whether the common question predominates over any individual questions, including individualized questions about injury or entitlement to damages.” The majority added in a footnote that its conclusion that courts must apply Rule 23(b)(3) case by case, “rather than rely on a per se rule that a class cannot be certified if it includes more than a de minimis number of uninjured class members, is consistent with the approach taken by our sister circuits,” and said that neither In re Asacol Antitrust Litigation, 907 F.3d 42 (1st Cir. 2018), nor In re Rail Freight Fuel Surcharge Antitrust Litigation (Rail Freight II), 934 F.3d 619 (D.C. Cir. 2019), “adopted a per se rule.”
The Olean dissent disagreed with that characterization. Judge Lee, joined by Judge Kleinfeld, headed the relevant part of the dissent “The majority’s rejection of a de minimis rule creates a circuit split,” wrote that “the number of uninjured class members should be de minimis—based on Rule 23’s language, common sense, and precedent from other circuits,” and concluded that “the majority opinion needlessly creates a split with other circuits that have endorsed a de minimis rule.” The First Circuit’s position in In re Asacol Antitrust Litigation, 907 F.3d 42, 53 to 54 (1st Cir. 2018), decided October 15, 2018, carries a qualifier that is easy to drop and that does most of the work: where any class member may be uninjured and thousands in fact suffered no injury, “[t]he need to identify those individuals will predominate and render an adjudication unmanageable absent evidence such as the unrebutted affidavits assumed in Nexium, or some other mechanism that can manageably remove uninjured persons from the class in a manner that protects the parties’ rights.” Asacol reversed an order granting certification. In In re Rail Freight Fuel Surcharge Antitrust Litigation (Rail Freight II), 934 F.3d 619, 625 (D.C. Cir. 2019), decided August 16, 2019, the district court had denied certification, stressing that the plaintiffs proposed no “further way”—short of full-blown, individual trials—“to reduce this number and segregate the uninjured from the truly injured”; the D.C. Circuit held that “[n]one of this was an abuse of discretion” and affirmed.
The Supreme Court had the question and let it go. In Laboratory Corporation of America Holdings v. Davis, 605 U.S. ___ (2025), the writ of certiorari was dismissed as improvidently granted in a per curiam order issued June 5, 2025. Justice Kavanaugh, dissenting alone, framed the question the Court did not decide as “[w]hether a federal court may certify a damages class pursuant to Federal Rule of Civil Procedure 23 when the class includes both injured and uninjured class members,” and said he would hold that “a federal court may not certify a damages class that includes both injured and uninjured members.” Until that is resolved the answer depends on the circuit.
Is the damages expert’s admissibility tested at certification?
Yes, though Comcast Corp. v. Behrend itself did not reach it. The Supreme Court of the United States noted at 569 U.S. at 32 n.4 only that the respondents contended the petitioners had forfeited a Daubert objection by not challenging their expert under the Federal Rules of Evidence below, and reasoned that such a forfeit “would make it impossible for petitioners to argue that Dr. McClave’s testimony was not ‘admissible evidence’ under the Rules; but it does not make it impossible for them to argue that the evidence failed ‘to show that the case is susceptible to awarding damages on a class-wide basis.’” The Ninth Circuit turned that into a rule in Olean, 31 F.4th at 665, reading it together with Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442 (2016): “Where, as here, a defendant did not raise a Daubert challenge to the expert evidence before the district court, the defendant forfeits the ability to argue on appeal that the evidence was inadmissible, but may still argue that the evidence is not capable of answering a common question on a class-wide basis.”
Tyson Foods also declined to make representative and statistical proof a special category. The Court wrote that “[a] representative or statistical sample, like all evidence, is a means to establish or defend against liability,” and that “[i]ts permissibility turns not on the form a proceeding takes—be it a class or individual action—but on the degree to which the evidence is reliable in proving or disproving the elements of the relevant cause of action. See Fed. Rules Evid. 401, 403, and 702.” Admissibility and Rule 23 remain separate questions the same expert report has to survive twice, which is why the admissibility grounds are treated on their own at Daubert challenges to damages experts and Critiquing a model.
Two boundaries on all of the above. What a cause of action permits as a measure of damages, and what Rule 23 requires in a given forum, are questions of law for counsel; this Institute describes what the measures measure, and publishes no figure. Questions about what a business is worth under a governing standard of value belong to the Business Valuation Institute at businessvaluationsinstitute.com, whether the underlying books and records are reliable to the Forensic Accounting and Fraud Institute at forensicaccountinginstitute.com, and what a patent, trademark or trade secret claim permits as a measure to the Intellectual Property Institute at iplitigationinstitute.com.