In a recent discovery order in Chen-Oster v. Goldman, Sachs & Co. – a putative class action in which plaintiffs allege that the Goldman Sachs defendants “engaged in a pattern of gender discrimination against female professional employees in violation of Title VII of the Civil Rights Act of 1964” and the NYC Human Rights Law – Southern District Magistrate Judge Francis ruled on which internal company complaints plaintiffs were entitled to.
The decision is instructive because it discusses what items are discoverable in employment discrimination cases in general, and so-called “pattern or practice” cases in particular.
Plaintiffs sought, among other things, discovery of “all internal complaints, made by all putative class members, that relate to compensation, promotion, or performance evaluation, regardless of whether such complaints expressly allege gender discrimination.”
They took issue with the fact that defendant “turned over only those complaints that contain ‘magic language’ indicating gender discrimination was at issue.” Plaintiffs argued that the complaints they seek are discoverable because
an alleged pattern or practice of discrimination creates an environment in which [a] female employee may experience … the disparate impact of a neutral policy or practice and may complain about the effect of that policy or practice … without asserting that her gender or discrimination are to blame. (Emphasis in original.)
Goldman Sachs opposed plaintiffs’ request for all complaints by female employees as beyond the scope of relevant discovery, arguing that many of these complaints “have nothing to do with gender concerns” and that the request amounts to a “fishing expedition into areas unrelated to plaintiff’s claims” including “discrimination claims based on factors other than sex.”
After reviewing the general principles of discovery contained in Federal Rule of Procedure 26(b), the court explained how these rules apply in the employment discrimination context:
[C]ourts typically apply more liberal civil discovery rules in employment discrimination cases, giving plaintiffs broad access to employers’ records in an effort to document their claims. As Title VII cases are particularly hard to prove in the absence of a proverbial smoking gun, such as a discriminatory comment made by a hiring official, discovery in these cases is necessarily broad. [T]he imposition of unnecessary discovery limitations [in employment discrimination cases] is to be avoided. Broader discovery is warranted when a plaintiff’s claims are premised on a pattern or practice of discrimination at the organization-wide level, as opposed to specific allegations of discrimination made against an individual supervisor.
Initially, the court rejected defendant’s reliance on precedent addressing discovery into areas “unrelated” to a plaintiff’s claims, since plaintiffs here “do not seek discovery of non-gender-related complaints, such as those alleging race or age discrimination.” Rather,
they seek discovery of all complaints made by female employees regarding compensation, promotion, or performance evaluation, including those complaints that appear to be based primarily on race or age, because they believe that these complaints are related to a disparate impact felt by female employees resulting from a pattern or practice of gender discrimination. In other words, the parties disagree about whether the complaints are gender-related, not whether discovery of non-gender-related complaints is also appropriate.
In assessing plaintiff’s request, the court turned to the law of retaliation to determine what qualifies as a “gender-related” complaint:
A plaintiff alleging retaliation as the basis for an employment discrimination claim must show that her employer was aware of her complaint and that it understood, or could reasonably have understood, that the plaintiff’s opposition was directed at conduct prohibited by Title VII. Courts must therefore determine whether an employer had sufficient notice that an employee’s complaint raised the issue of gender discrimination. [T]here are no magic words that must be used when complaining about the alleged discrimination in order to satisfy this requirement. However, if an employee does not use key words such as “discrimination” or “gender,” the complaint must lend itself to a reasonable inference of unlawful discrimination.
In applying these principles to this case, the court held that while complaints containing “buzzwords” (such as “sex discrimination”, “male”, “female”, “gender”, “glass ceiling”, or “women’s work”) are clearly gender-related and discoverable, such “buzzwords” are not “required at the discovery stage in a disparate treatment case.” Furthermore:
Given the broader scope of discovery in pattern-or-practice cases, together with the benefit of hindsight, complaints that were insufficient to alert Goldman Sachs to gender discrimination at the time they were made may be sufficiently related to gender to make them discoverable now. For example, complaints from a female high-level manager about having to perform secretarial work, while arguably failing to support a retaliation claim, could reasonably lead to evidence of gender discrimination and thus meet the broader relevance standard for discovery.
However, the court held that “plaintiffs’ demand for all complaints by female employees regarding compensation, promotion, or performance evaluation goes too far,” reasoning:
When a female employee complains that she “feel[s][she] should be paid more,” this may simply be an expression of personal job dissatisfaction unrelated to disparate treatment or gender discrimination. The likelihood that such a complaint would lead to anecdotal evidence of a disparate impact is outweighed by the burden of production.
It concluded by noting the appropriate balance lay somewhere between the parties’ positions:
The appropriate balance between these extremes rests on the concept of “disparity” at the heart of the present discrimination claims. Determining whether a disparity exists in the employment discrimination context inherently involves comparing at least two individuals: the complainant and another employee. When a female employee complains not simply that she “should be paid more,” as in the previous example, but rather that she “should be paid as much as or more than” somebody else, the nature of the complaint transforms into an allegation of disparity. And when that comparison involves employees of different genders, it is reasonable to conclude that gender-related disparity may be at issue. (Emphasis in original.)
The court threefore ordered defendant to “provide the plaintiffs any internal complaints regarding compensation, promotion, or performance review where a female employee who is a member of the putative class drew a comparison between herself or another putative class member and one or more of her male colleagues.” Such “comparisons are sufficiently gender-related that such complaints could reasonably lead to admissible evidence of value to the plaintiffs, and are thus discoverable.”
It also required defendant to provide plaintiff with unredacted names from complaints previously provided, reasoning that such discovery will enable plaintiffs “to contact potential witnesses and develop anecdotal evidence of the alleged gender discrimination.” Here, there was “no reason to burden the plaintiffs here with the task of randomly selecting employees to determine if they had gender-related grievances; a set of relevant complaints has already been produced, and the individuals identified in these complaints are the persons most likely to have information relevant to the plaintiffs’ case.”
The court additionally disallowed discovery of complaints by employees outside plaintiffs’ divisions, reasoning:
the likelihood that discovery of complaints by employees in the non-revenue divisions will be overly burdensome outweighs the scant possibility of uncovering admissible evidence. The plaintiffs have not demonstrated that these complaints are relevant to their allegations of a pattern or practice of discrimination allegedly impacting women in the revenue-generating divisions. It is unclear what additional value would be gained from anecdotal evidence derived from these complaints, as these employees are not the subject of the statistical pattern-or-practice analysis. There is no indication that putative class members were affected by policies or managerial decisions in the non-revenue generating divisions. And, according to Goldman Sachs, the compensation and evaluation models in the other divisions are significantly different from those in the revenue-generating divisions in which the plaintiffs worked.
Plaintiffs were, however, entitled to discovery of complaints by women within plaintiffs’ divisions who are not members of the putative class:
Complaints made by analysts or administrative assistants, for example, are more clearly and directly related to the plaintiffs’ case because these individuals have worked closely with putative class members. They may be able to provide anecdotes and information regarding their interactions with common managers, experiences utilizing the same internal complaint process, or the general culture of these divisions. Some employees may have first raised gender-related concerns when they were analysts, before they became members of the putative class as associates or vice presidents. And, although they are outside the class, these employees may also have felt, either directly or indirectly, the negative impact of a pattern or practice of discrimination against their putative class member coworkers. These complaints are reasonably likely to lead to additional evidence pertinent to the claims of the putative class employees, and Goldman Sachs has not argued that producing these complaints from the same divisions already under examination would be unreasonably burdensome.