by James M. Finberg and Kelly M. Dermody 1 |
| Discovery in employment discrimination class actions can be time consuming, burdensome, and expensive. Both plaintiffs and defendants have an interest in taking steps to ensure that the discovery proceeds as quickly, cheaply, and efficiently as possible. Three ways to do that are: (1) the early production of computerized payroll and human resources data; (2) use of 30(b)(6) depositions; and (3) a bifurcation of pre-class certification and post-class certification discovery. |
| A. The Importance Of Computer Data |
| Most employment discrimination class actions rise or fall on the statistical evidence.2 Accordingly, the production of computerized personnel and payroll data is critical in these cases. Once both sides have had access to this data and have an opportunity to analyze it (with the assistance from experts), they can sit down and discuss resolution of the case. In realty, no attorney knows whether his or her employment discrimination case is strong or weak until the statistician has spoken. |
| In view of this, the first and most important request that any plaintiff must make in an employment discrimination class action is for defendant company's computerized payroll and personnel records. Most major companies today keep both payroll and human resource information in computerized form. That information can be produced relatively quickly and cheaply. |
| In cases where plaintiffs from a few facilities are seeking certification of a class involving many more facilities, some defense counsel have attempted to make the argument that the defendant is not required to, and will not, produce computer data for facilities other than those facilities of the named plaintiffs. Many courts have rejected that argument. See, e.g., Diaz v. American Tel. & Tel., 752 F.2d 1356, 1362-64 (9th Cir. 1985) (employee bringing individual employment discrimination action entitled to discovery of statistical data regarding hiring and promotion practices throughout Western Region of defendant's company to establish defendant's discriminatory practices); Duke v. University of Tex. at El, Paso, 729 F.2d 994, 996-97 (5th Cir. 1984), cert. denied, 469 U.S. 982 (1984) (plaintiffs alleging classwide sex discrimination entitled to pre-certification discovery of promotion and pay records of departments other than the two departments where named plaintiffs were employed); Trevino v. Celanese Corp., 701 F.2d 397, 405-06 (5th Cir. 1983), reh'g denied, 707 F.2d 515 (5th Cir. 1983) (observing that "[t]he imposition of unnecessary limitations on discovery is especially frowned upon in Title VII cases" and vacating protective order prohibiting discovery into company's internal organization, computer facilities, personnel record keeping system, and company purging policies for internal personnel records); Zahorik v. Cornell Univ., 98 F.R.D. 27, 32-34 (N.D.N.Y. 1983) (plaintiffs entitled to pre-certification discovery of defendant university's affirmative action plans, university-wide reports on employment at university, relevant computerized information, statistical and biographical information on employees, and sex discrimination complaints against university); Barnhart v. Safeway Stores, Inc., 60 Fair Empl. Prac. Cas. [BNA] 751, 754 (E.D. Cal. 1992) (observing that "[i]t is this court's duty to permit the parties to give the district court all the information that the district court might find relevant in making a class certification determination"); Canty v. Philip Morris U.S.A., 18 Fair Empl. Prac. Cas. [BNA] 86, 88 (E.D. Pa. 1978) (plaintiff entitled to pre-certification discovery of readily available statistical information on hiring, promotions, and discharge on a nationwide basis, with regional and divisional breakdowns). In order to determine whether a class should be certified, there needs to be some evidence about whether or not there is a pattern and practice justifying certification. |
| In addition to being legally questionable, the failure to provide statistical information on a classwide basis early on is also very shortsighted. First, producing statistical information is almost always cheaper, less time consuming and less burdensome than producing hard copy information. A wise defendant can most likely make a tradeoff with plaintiffs' counsel that the defendant will produce computerized information and only minimal hard copy information. Second, if the defendant has a strong case, the company should be eager to show the statistics demonstrating the strength of its case to the plaintiffs, so that the company can quickly resolve the case on terms favorable to the defendant. Conversely, if the company indeed does have a problem, as reflected in its statistics, the company will be better off to settle a case early before the plaintiffs have overcome the hurdles of class certification and summary judgment. The terms of settlement will be much less favorable to the company later, as class members become part of the case (and thus feel entitled to monetary relief) and class counsel’s hand is strengthened. |
| B. Pre-Class-Certification Discovery |
| Whether a class is certified, and the geographic and temporal scope of the class certified, will have a tremendous impact on the shape of the trial and thus the discovery needed to prepare for trial. Accordingly, it is logical to bifurcate discovery in employment discrimination class actions into two phases: pre-class-certification discovery and post-class certification discovery. The pre-class-certification discovery should focus on computerized data, organization charts, forms, policy manuals, employment manuals, training manuals and other documents having general application, 30(b)(6) depositions, and samples of other types of information. Scorched earth document discovery should wait until after class certification. In addition, class certification depositions should be limited in scope, and not include such issues as damages or individual issues. |
| During the class certification discovery phase, the plaintiffs' attorney should endeavor to discover how the company is structured, who is involved in decision making, and how the decisions are made. The critical documents for plaintiffs to obtain during this phase include organization charts, copies of generic application forms, copies of generic evaluation forms, copies of all orientation materials, copies of training materials given to managers regarding human resources issues, the company's policy manuals and employee handbooks, civil complaints, and EEOC charges. The discovery at this stage should be aimed at getting an overview of how things work, not to analyze the specifics of each and every employment decision. |
| Pre-class-certification deposition discovery can be streamlined in the same fashion. By asking the company to designate the most knowledgeable people about its personnel practices pursuant to Federal Rule of Civil Procedure 30(b)(6), the plaintiffs' attorneys can get the company on record about the company's policies and practices. That, combined with depositions of a few other managers that focus on policies and practices, but not on the merits of individual claims, should eliminate the need for a large number of multi-day depositions. In Butler v. Home Depot, for example, the pre-class-certification depositions of Home Depot's managers, with few exceptions, took no more than one day each. |
| During the class-certification discovery phase, there is a tension between allowing enough discovery so that both sides can present sufficient evidence as to whether or not certification is or is not appropriate, and unduly burdening the parties. Focusing on the production of computer data and forms, manuals and documents of general applicability, and a limited number of depositions focused on policies and practices generally without exhaustive investigation of specifics, can achieve the former without causing the latter. |
| C. Merits Discovery |
| Once the class is certified, the parties must begin to focus on evidence that is needed to present the case to the jury. Because jurors want apples-to-apples comparisons between employees, the discovery in the merits phase should include hard copy and deposition discovery related both to members of the protected class who were treated adversely and to individuals of the non-protected class who were treated better. Such discovery is appropriate, even though the test in McDonnell Douglas v. Green, 411 U.S. 792, 803 (1973) need not be satisfied for each class member in a pattern and practice class action. See International Brotherhood of Teamsters v. United States, 431 U.S. 324, 359-60 (1977). |
| D. Personnel Files |
| Hard copy personnel files are the bane of an employment discrimination class action litigator's existence. If the trial of the class action involves statistical evidence and some anecdotal evidence to bring the statistics to life, see Teamsters, it should not be necessary to review of personnel files of all members of the class or of all non-class members who were treated better than class members. Inevitably, however, preparing an employment discrimination class action for trial will involve obtaining and reviewing hundreds of hard copy personnel files. |
| E. Informal Discovery |
| For plaintiffs, formal discovery has its place, but informal investigation can often be even more helpful. Such informal discovery can include discussions with former employees, government agencies (e.g. the EEOC and OFCCP), or other entities with which the company has worked in connection with its employment practices (e.g. college recruiters). Under the ABA Ethical Rules and the ethical rules of almost every state, communications with former employees of an entity are permitted. ABA Model Rule 4.2; Cal. Rule Prof. Conduct 2-100. Information received from former managers can be invaluable in developing a plaintiffs' case. |
NOTES |
| 1 Kelly M. Dermody is a partner with Lieff, Cabraser, Heimann & Bernstein, LLP of San Francisco, California. Ms. Dermody, together with her co-counsel at Saperstein, Goldstein, Demchak & Baller, represented the plaintiffs in Butler v. Home Depot, C94 4335 SI (N.D. Cal.). |
| 2 In Hazelwood School Dist. v. United States, 433 U.S. 299, 307-13 (1977), the United States Supreme Court found that statistical disparities of two or more standard deviations are sufficient to make a plaintiff's prima facie case. |