Is Trial by Formula Kosher With SCOTUS?

Bloomberg Law’s combination of innovative analytics, research tools and practical guidance provides you with everything you need to be a successful litigator.

By Perry Cooper

Nov. 5 — The use of statistical evidence in class certification decisions will go under the microscope when the U.S. Supreme Court hears oral argument Nov. 10 in an employment law case.

The case involves allegations by 3,000 pork-processing workers at a Tyson Foods plant in Iowa that they were inadequately compensated for time spent changing into and out of protective equipment (16 CLASS 661, 6/12/15).

The plaintiffs' expert calculated the workers' average time putting on and taking off protective gear, despite the fact that actual changing time varied greatly among class members.

Tyson Foods Inc. argues that this kind of averaging constitutes impermissible “trial by formula” under Wal-Mart Stores Inc. v. Dukes, and can't be relied on to certify a class under Fed. R. Civ. P. 23.

“Statistical evidence should not be used to replace the fundamental requirement of our legal system that in order to sue and recover money, you need to prove that you were injured,” defense attorney Jessica D. Miller of Skadden, Arps, Slate, Meagher & Flom LLP in Washington, told Bloomberg BNA in a Nov. 6 e-mail.

“The question in litigation is not whether, on average, people were injured, but rather who was injured and if so, what compensation does he or she deserve,” she said. Skadden Arps filed an amicus brief on behalf of the Product Liability Advisory Council Inc. in support of Tyson Foods.

“The real problem with the class in Tyson is that all of the individual employees had vastly different working conditions—some of them worked overtime, some of them did not,” attorney Richard A. Samp of the pro-business Washington Legal Foundation, which filed an amicus brief in support of Tyson Foods, told Bloomberg BNA recently.

“The lower courts allowed the class to proceed on the false assumption that every employee worked the average amount of time,” he said. “Class actions were never intended to serve situations where individuals are not similarly situated.”

But the workers argue that the statistics are a reasonable approximation of the entire class's compensable time.

The plaintiffs were forced to use statistics because their employer didn't keep adequate records, attorney Jason L. Lichtman, of plaintiffs' firm Lieff Cabraser Heimann & Bernstein LLP in New York, said.

“It's hard to think of any modern class litigation where there isn't some use of statistical proof at some step of the process,” he said at an Oct. 14 webinar sponsored by the public interest firm Public Justice. “If the Supreme Court takes the bait and rules that litigants can't use any extrapolation, averaging or sampling, it will wipe out a huge chunk of modern class action practice.”

Lieff Cabraser filed an amicus brief on behalf of Public Justice in support of the class.

Average OK Under Wal-Mart?

The plaintiffs' expert videotaped 744 employees' routines to obtain average donning, doffing and walking times.

The expert calculated specific damages for each plaintiff, and damages weren't assigned to the small percentage of the class members who didn't work any overtime, Jocelyn D. Larkin said at the Oct. 14 webinar. Larkin is the executive director of the Berkeley, Calif.-based Impact Fund, which also filed an amicus brief in support of the class.

The trial court awarded the workers $5.8 million. But the judge made no plan for how that money would be distributed among the class members.

No state or federal jurisdiction has ever said plaintiffs need a precise measure of damages, Lichtman said. That has been settled law since the 1940s, when the Supreme Court held in another employment case that if employers don't keep the required time records, employees can present a “reasonable inference” of their time-worked, Mt. Clemens Pottery Co. v. Anderson, 328 U.S. 680 (1946).

“Usually a reasonable inference is shown through statistics,” Lichtman said.

But Samp said the Supreme Court rejected such use of random sampling in Wal-Mart Stores Inc. v. Dukes, 131 S. Ct. 2541 (2011) (12 CLASS 519, 6/24/11).

Wal-Mart involved a proposed class of over a million female Wal-Mart workers who alleged pay and promotion discrimination. The Supreme Court held that the workers' statistical evidence of discrimination failed to establish a nationwide pay or promotion pattern across all Wal-Mart's stores, and didn't tie all of the workers' claims together to satisfy commonality.

Commonality, a prerequisite to class certification under Rule 23, requires questions of law or fact common to the class.

“The point of a class action is to say that every single member of the class has been treated in an identical manner and therefore you can decide a legal issue that decides the case in one fell swoop and the whole case gets decided,” Samp said.

“But what you can’t ever do is to do what Wal-Mart very explicitly rejected,” he said.

“You can’t do a random sample of employees to see how many were discriminated against on the basis of sex and then try to say, ‘Well, our sample was random and 38 percent of employees were discriminated against on the basis of sex so we’re going to enter a class action judgment for 38 percent of the amount that is being claimed by the class.’”

Origin of ‘Trial by Formula.'

Tyson Foods' argument against “trial by formula” comes at the end of Wal-Mart, where the phrase is mentioned in the decision's final paragraph, Lichtman said.

The defense bar says trial by formula applies any time you use statistics, but that's not what Wal-Mart held, he said. The justices were worried about using a formula to replace trials entirely.

The Wal-Mart court describes its version of impermissible trial by formula, in the context of discrimination claims, this way: A sample set of class members is selected for whom liability for sex discrimination and resulting backpay would be determined by depositions. The percentage of the claims determined to be valid would be applied to the rest of the class, and that number of presumptively valid claims would be multiplied by the average backpay award in the sample set. That calculus would provide the entire class recovery, without further individualized proceedings.

“We disapprove of that novel project,” Justice Antonin Scalia wrote in Wal-Mart. “A class cannot be certified on the premise that Wal-Mart will not be entitled to litigate its statutory defenses to individual claims.”

To do so would run afoul of the Rules Enabling Act's prohibition of interpreting any law to “abridge, enlarge or modify and substantive right,” 28 U.S.C. § 2072(b), the court said.

Larkin called this discussion “super duper dicta.”

No-Injury Argument Out?

The second question before the court concerns whether a class action can be certified when the class contains members who weren't injured and have no right to damages.

Tyson Foods initially argued that the class improperly contained “no-injury” plaintiffs who didn't work any overtime and therefore weren't entitled to extra pay for donning and doffing time.

The business community has tried many times in recent years to get this question before the Supreme Court, most recently in the moldy washing machine cases where not everyone's defective washers grew mold.

The top court passed on the no-injury issue when these cases were up for certiorari review last term (15 CLASS 204, 2/28/14).

Larkin said having “non-injured” class members included at certification is the reality of class actions.

It's impossible to know at the time of certification that every member was injured, and it would involve an improper merits review at the certification stage, she said. But there are mechanisms at the end of a case to determine who was injured and to what extent.

Tyson Foods' merits brief has all but abandoned this no-injury argument. Now, the company argues that, while a class may contain uninjured members, the court can't award damages to such members. Everyone agrees on that point, Larkin said.

The company argues that, in the workers' case against it, the plaintiffs didn't provide a culling mechanism to separate out class members who weren't injured. But Larkin said Tyson Foods' argument assumes the jury verdict will be distributed pro rata to all class members, and the trial judge never ruled on how to distribute the funds.

The plaintiffs' bar is slightly relieved that Tyson Foods is backing off the no-injury argument, but Tyson's amici are still raising very vocal concerns about companies being scared into settling by overly ambitious class action lawyers, she said.


Samp said cases like this one against Tyson Foods get certified with the knowledge that, after certification, the defendant will be forced to settle.

Such forced settlements will all but disappear if the court rules in favor of Tyson Foods, as he confidently predicts, because cases where thousands of individual issues of fact predominate won't be certified.

Miller said the Supreme Court missed its opportunity to address the problem of class actions based on the complaints of a few atypical customers or employees in the washing machine cases. “But I'm hopeful that Tyson will start to turn this problem around,” she said.

Plaintiffs' attorneys, meanwhile, are hopeful the case won't be a home run for business.

“I'm pessimistic about the fate of the actual plaintiffs in this case, but cautiously optimistic that it won't be a big defense win on the statistics issue,” Lichtman said. “Every year there's a meteor headed towards our practice, but this time we have a very strong response.”

The middle ground for the court to take is also the easiest path, he said. The court could say the method used in this case was statistically invalid. It could throw out this expert's method without ruling more widely.

If the court rules against the plaintiffs on the original, and broadly presented no-injury question, which seems unlikely given the company's backtrack on the argument, “class litigation is over,” Larkin said.

But Larkin offered up a more “Pollyanna-ish” prediction: She is “fairly hopeful” that the liberal justices have come to recognize that the conservative justices' campaign to end class actions is on. Justice Elena Kagan, especially, understands the stakes, she said.

Larkin hopes the court, in this case and in the other three class cases already argued this term, will issue “more nuanced decisions this year that are not nearly as detrimental as some of the court's earlier opinions.”

To contact the reporter on this story: Perry Cooper in Washington at

To contact the editor responsible for this story: Steven Patrick at


Request Litigation on Bloomberg Law