After the Supreme Court decided Wal-Mart Stores, Inc. v. Dukes, it was fashionable for some in the plaintiffs' bar to argue that the deciding factor was the size of the class. After all, it's not often that a class of 1.5 million people comes along. Such a sprawling class, as it is generally described, would almost by definition be unmanageable.
But most lawyers understand that the Wal-Mart decision turned not on class size but on the plaintiffs' theory of discrimination. Typically, plaintiffs seek class certification by relying on a common policy that affects all class members in the same manner. The Wal-Mart plaintiffs, however, sought certification of their claims of gender-based employment discrimination by arguing that the company's uniform corporate policy of granting discretion to local store managers to make pay and promotion decisions, coupled with an alleged culture of bias at the company, led to disparate treatment for female employees. They attempted to support their argument by statistical analyses showing disparities between the treatment of men and women in the workplace.
The Court, as we all know, rejected plaintiffs' theory, holding that a corporate policy that promotes non-uniform practices throughout the organization represents the opposite of a uniform employment practice that would support a finding of Rule 23(a)(2) commonality. The surprise in the case was the Court's pronouncement of a commonality standard applicable to 23(a)(2) that closely resembled the predominance standard of 23(b)(3). It was a higher standard for (a)(2) than any previously known, occasioned by the fact that, because plaintiffs' brought their case under Rule 23(b)(2) rather than (b)(3), predominance was not at issue, and commonality had to be confronted on its own terms. While the size of the class lurked ominously in the background, it did not factor in the Court's reasoning.
United States District Judge Rya Zobel recently applied Wal-Mart to a different kind of discrimination case without reference to the number of people included in the class. Plaintiffs in Download Barrett v Option One Mortgage Corporation represented a class of African-American borrowers who obtained mortgage loans from the defendants. They alleged that defendants' policy of giving mortgage brokers discretion to impose additional charges on a loan that were unrelated to the borrower's creditworthiness resulted in higher mortgage rates for African-American borrowers than for similarly situated white borrowers. Like the plaintiffs in Wal-Mart, the Barrett plaintiffs supported their argument with a statistical analysis purporting to prove the pricing policies' disparate impact.
Judge Zobel had certified the class months before the Supreme Court decided Wal-Mart, but on September 18, 2012, she issued an order decertifying it in light of the Supreme Court's decision. She explained that the Supreme Court “recognized that giving discretion to lower-level supervisors could be a policy giving rise to a disparate impact claim, but ruled that such a claim would only be common to members of a nationwide class if the plaintiffs identified a 'common mode of exercising discretion that pervades the entire company.'” Additionally, she cited the Court's rejection of the Wal-Mart plaintiffs' statistical analysies because they were performed on regional and national data, and could not establish the uniform, store-by-store disparity which alone could support a finding of commonality.
Because the Barrett plaintiffs did not point to any “common mode of exercising discretion,” their argument for commonality failed. As Judge Zobel explained:
[I]f plaintiffs claimed that Option One’s brokers uniformly exercised their discretion by considering specific attributes that produce disparate impact—such as “scores on general aptitude tests or educational achievements,” …–they would state a common question justifying class treatment. But without some such claim, the Supreme Court tells us, “demonstrating the invalidity of one [broker’s] use of discretion will do nothing to demonstrate the invalidity of another’s.” . . .Because plaintiffs do not claim that all of Option One’s brokers exercised their discretion in the same way, they do not raise a single question common to all plaintiffs in the class.
Similarly, because plaintiffs' statistical analysis looked only at aggregate data on a national level, rather than considering each broker individually, Judge Zobel held that it could not support a finding of commonality. “Instead, under Wal-Mart, plaintiffs can only show a common question if they can show a common disparate impact at the level of each individual broker, thereby raising the inference that the brokers shared a common mode of exercising their discretion.”
In decertifying the class, the judge was true to the Wal-Mart rationale and to the principles underlying class litigation. A nationwide claim based on a company's decentralized system that places discretion in local actors is, and should be, doomed to failure. Further, although both Wal-Mart and Barrett are discrimination cases, these principles should apply equally to other disciplines. And, finally, as Barrett demonstrates, what matters is not the size of the class, but the plaintiff's theory of the case.