Statistical and representative evidence sits at the intersection of class certification and merits analysis—and the Supreme Court's decision in Tyson Foods drew the line between permissible representative proof and an impermissible "trial by formula."
Doctrinal Framing
For decades after Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338 (2011), defendants argued that any class-wide damages methodology relying on averages or sampling violated the due-process requirement that each defendant have the opportunity to contest individual liability. Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442 (2016), answered that argument directly, holding that representative evidence may permissibly be the foundation for both liability and damages in wage-and-hour class actions—but the holding is narrower than it first appears, and counsel on both sides must understand its precise contours.
The Anderson v. Mt. Clemens Pottery Foundation
Statistical proof in FLSA litigation traces its lineage to Anderson v. Mt. Clemens Pottery Co., 328 U.S. 680 (1946). In Mt. Clemens, the Supreme Court held that where an employer has failed to maintain accurate records of compensable time—as § 11(c) of the FLSA requires—employees should not be barred from recovery simply because the precise quantum of uncompensated work cannot be established. The Court fashioned a burden-shifting framework: once an employee proves she performed compensable work for which she was improperly compensated and produces "sufficient evidence to show the amount and extent of that work as a matter of just and reasonable inference," id. at 687, the burden shifts to the employer to produce evidence negating the reasonableness of that inference. Failure to do so results in liability even if the damages award is "only approximate."
Mt. Clemens was born of necessity. The FLSA imposed record-keeping obligations on employers precisely to avoid the evidentiary gaps that arise when workers are paid off the clock. The Court's logic was that the employer's own statutory violation could not become a shield against liability.
The Tyson Foods Decision
Tyson Foods arose at a pork processing plant in Iowa where employees in the kill, cut, and retrim departments were required to don and doff protective gear—tasks that varied in duration depending on the employee's daily assignment. Tyson compensated some workers for donning and doffing but kept no records of how much time each individual spent. Plaintiffs sought Rule 23(b)(3) class certification for Iowa wage-law claims and FLSA collective-action certification.
To establish that each class member worked more than forty hours per week inclusive of donning-and-doffing time, plaintiffs relied principally on an industrial-relations expert's time study. The expert observed and videotaped donning and doffing activities, averaged the time, and produced estimates of 18 minutes per day for the cut and retrim departments and 21.25 minutes for the kill department. These estimates were added to each employee's time-card data. The jury awarded approximately $2.9 million.
The Supreme Court, in a 6–2 opinion authored by Justice Kennedy, affirmed. The Court's holding rests on two interlocking propositions.
First, whether representative evidence can be used in a class action depends on whether it could have been used in an individual action. If each Tyson employee had sued individually, each could have invoked Mt. Clemens to establish unpaid overtime through representative sampling when the employer had failed to maintain adequate records. That individual entitlement does not evaporate because the action was brought collectively. 577 U.S. at 455–56.
Second, the Court distinguished the "trial by formula" it had condemned in Dukes. In Dukes, the proposed class-wide back-pay methodology had been decoupled from any individualized determination of liability—a Seventh Amendment problem, not merely a manageability concern. In Tyson Foods, by contrast, the representative evidence went to a common question: the amount of time all employees spent on a single, shared activity (donning and doffing). The question was susceptible to class-wide resolution because the underlying activity—however varied in precise minutes—was meaningfully uniform. 577 U.S. at 453–54.
The Court expressly declined to "announce broad and categorical rules governing the use of representative and statistical evidence in class actions." Id. at 456. This makes Tyson Foods narrow: it authorizes representative proof in contexts where the employer's record-keeping failure created the evidentiary gap and where the common activity being measured is genuinely uniform enough to permit statistical inference.
When Statistical Proof Is Appropriate
Post-Tyson, courts have wrestled with the appropriate conditions for representative evidence. The following factors bear on admissibility and sufficiency:
1. Employer record-keeping failures. Tyson Foods carries the most force when the employer's own failure to keep records—a statutory violation under 29 U.S.C. § 211(c) for FLSA cases—necessitated the use of sampling. Where records exist and defendants can cross-examine each claimant's data, the Mt. Clemens burden-shift does not apply with the same force, and statistical averaging may face greater scrutiny.
2. Uniformity of the underlying activity. The Court emphasized that donning and doffing was a shared activity performed by all class members under materially similar conditions. Representative evidence is less defensible when the underlying conduct varies substantially across class members—as the Dukes Title VII cases demonstrated. Practitioners must be prepared to address how the measured activity compares across the class.
3. Reliability under Daubert. The time study in Tyson Foods was conducted by a credentialed expert using videotaped observations, peer-reviewed methodology, and disclosed data. Courts evaluate sampling methodologies under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). Defense experts will challenge sample size, selection bias, representativeness, and the statistical assumptions underlying extrapolation. Plaintiffs' counsel must retain an expert capable of defending those methodological choices.
4. The individual-action hypothetical. The Tyson Foods test asks whether each individual plaintiff could have used the representative evidence to establish her own claim. This test screens out scenarios where the evidence rests on the class mechanism itself rather than on independently cognizable individual rights.
Damages Distribution: The Unresolved Problem
The Tyson Foods Court left a significant problem for lower courts and practitioners: damages distribution. After a class-wide verdict, how are individual awards calculated when the representative evidence establishes only an average? If the study yields 18 minutes per day, but some employees actually donned and doffed in 10 minutes and others in 26, distributing a lump-sum verdict proportionately to timecard hours without any mechanism for identifying which individuals were actually underpaid may result in payments to plaintiffs who suffered no injury.
The Court acknowledged the problem but resolved it procedurally rather than substantively. It held that the district court could address any concerns about uninjured class members through scrutiny of the distribution plan at the damages phase, rather than at certification. 577 U.S. at 461. Justice Thomas in dissent argued that this deferred the constitutional problem rather than solved it.
In practice, this means:
- Plaintiffs' counsel should construct a detailed allocation methodology before trial, ideally one that uses the underlying time-study variance data to weight individual awards rather than treating all class members identically.
- Defense counsel should move in limine to require the allocation plan to be disclosed and tested before verdict; post-verdict challenges to allocation are harder to win and may be moot.
- Courts have increasingly required preliminary approval submissions in FLSA collective actions to address allocation before disbursement.
Practice Notes
Pleading. At certification, plaintiffs must demonstrate that the proposed statistical methodology is "sufficiently reliable and valid to reasonably demonstrate the element it is offered to prove" for the class as a whole. This showing typically appears in a Daubert-ready declaration from the sampling expert, filed with the Rule 23 motion.
Discovery. Request all payroll records, timekeeping systems, and any internal communications about how donning-and-doffing time was calculated. Evidence that the employer knew its time estimates were inaccurate strengthens both the Mt. Clemens burden shift and the case for higher damages.
Expert designation. Industrial-relations experts and labor economists capable of defending time-motion studies are essential. Budget for rebuttal of defense sampling experts who will challenge methodology at every step.
Decertification risk. Defendants in Tyson Foods situations routinely move to decertify after the close of expert discovery, arguing that variances in the time-study data have revealed material individual differences. The response is to show that the common question—whether the employer compensated for the activity at all—remains susceptible to common resolution even if damages vary.
Open Questions and Where the Law Is Moving
Tyson Foods left several questions unresolved that continue to generate circuit splits:
- Non-FLSA contexts. Lower courts have debated whether the Mt. Clemens burden shift and the Tyson Foods framework apply in state wage-and-hour class actions, ADA accommodation claims, and other class contexts where representative evidence is offered. Courts in California, New York, and Illinois have generally applied analogous burden-shifting frameworks.
- Uninjured class members and Article III standing. TransUnion LLC v. Ramirez, 594 U.S. 413 (2021), requires that each class member have Article III standing at the time of judgment. Defense counsel have argued that a class containing members who suffered no injury fails TransUnion scrutiny—a tension that Tyson Foods' deferred-distribution solution does not fully answer.
- Trial plan specificity. Some circuits now require plaintiffs to submit a detailed trial plan at certification showing how class-wide statistical proof and any individualized issues will be managed at trial. The absence of such a plan can defeat certification even if the underlying methodology is sound.
Closing
Tyson Foods preserved the viability of statistical proof in wage-and-hour class litigation, but it did not insulate any sampling methodology from scrutiny. The decision's durability depends on plaintiffs' ability to connect statistical methods to genuinely shared activities and on courts' willingness to grapple seriously with the damages-distribution problem the opinion left behind. For class counsel, the take-away is straightforward: invest early in a defensible statistical methodology, address distribution in advance, and be prepared to distinguish the class's evidence from the rote averaging that drew the Court's fire in Dukes.
Talk to Yates Anderson
If you are litigating a matter in this area — or weighing whether to — the working analysis above only goes so far. Request a case evaluation and a Yates Anderson attorney will respond within one business day.
Informational only. Not legal advice. No attorney-client relationship is created by reading this post. Consult a licensed attorney in your jurisdiction.