Please ensure Javascript is enabled for purposes of website accessibility

W. Lawrence Wescott: Reducing e-discovery burden – easier said than done

The high cost of e-discovery in many cases has spawned numerous strategies attempting to reduce its burden. However, as a decision last month by a federal court in New York indicates, such techniques don’t always reduce costs or complexity.

Proportionality as a goal

In 2010, the Sedona Conference® released its Commentary on Proportionality in Electronic Discovery (11 The Sedona Conference Journal 289 (Fall 2010)), which discussed methods of applying the principles of proportionality contained in Rule 26(b)(2) to “manage the large volume of ESI [electronically stored information] and associated expenses now typical in litigation.”

Sampling is a strategy that has been in use for many years. For example, in Rowe Entertainment, Inc. v. William Morris Agency, 205 F.R.D. 421, 427 (S.D.N.Y 2002), plaintiffs proposed to reduce the cost of email production by restoring a portion of the defendants’ backup tapes based upon date restrictions and sampling.

Sampling was also a key component of the court’s resolution of a privilege issue in In re Vioxx Products Liability Litigation, 2007 U.S. Dist. LEXIS 60299 (E.D. La. Aug. 14, 2007). While defendant Merck had sought review of privilege rulings on over 30,000 documents (comprising over 500,000 pages), the 5th U.S. Circuit Court of Appeals suggested that the U.S. District Court review a sample of 2,000 representative documents that Merck would select.

Phasing is a prioritization technique mentioned in the Sedona Conference proportionality commentary. The court in Tamburo v. Dworkin, 2010 U.S. Dist LEXIS 121510 (N.D. Ill. Nov. 17, 2010) cited the commentary in ordering the parties in a long-running dispute to develop a phased discovery plan. The parties were ordered to first focus on their Rule 26(a) initial disclosures, then to identify the claims most likely to go forward and concentrate their discovery efforts there, and then to prioritize discovery which was less expensive and burdensome.

Chen-Oster v. Goldman Sachs

One might think that the more such techniques are utilized, the greater the savings and reduction in complexity. However, at least in Chen-Oster v. Goldman Sachs & Co., 2012 U.S. Dist. LEXIS 130123 (S.D.N.Y. Sept. 10, 2012), that was not the case.

Chen-Oster involved an attempt by the plaintiffs to certify a class-action gender discrimination suit. The court noted that “[t]his case illustrates some of the practical difficulties in implementing three of the principles intended to reduce the burden of discovery generally, and electronic discovery in particular: phasing, sampling, and proportionality.” The plaintiffs’ objective was to obtain compensation, promotion and performance evaluation information contained in databases, as well as policy and complaint information not contained in databases.

The Chen-Oster court first addressed the impact of the Supreme Court’s decision in Wal-Mart Stores, Inc. v. Dukes, 131 S.Ct. 2541 (2011) upon the use of phased (or sequencing) discovery.

Dukes involved a class action by all Wal-Mart female employees employed since Dec. 26, 1998, who might have been subject to Wal-Mart’s pay and management track promotions policies and practices. The Supreme Court overturned the class certification order on the ground that the requirement of commonality, that the plaintiffs had suffered the same injury, had not been met. The court directed lower courts to engage in a “rigorous analysis” to determine whether the prerequisites for class certification had been met, which would admittedly “entail some overlap with the merits of the plaintiff’s underlying claim.”

In Chen-Oster, Goldman Sachs argued that discovery on general practices should be permitted, but discovery on employee data which would provide the basis for a statistical analysis should be denied. The court disagreed, holding that such a phased approach was not appropriate, as plaintiffs had already identified potential discriminatory policies. The statistical data sought could create the link between a discriminatory policy and discriminatory impact required by Dukes to support the finding of commonality.

In a footnote, the court observed that phased discovery could be appropriate in class actions; if, for example, there were doubts about numerosity, discovery limited to that issue could be dispositive. “Phased discovery, however, is less likely to be efficient with respect to the issue of commonality, especially after Dukes.”

Sampling also presented problems for the parties. Plaintiffs were concerned that sampling the data would permit them to discover statistically significant patterns of discrimination. They argued that “[t]he data is relevant in the aggregate to perform the applicable analyses to show patterns of statistically significant shortfalls or effects of challenged policies.”

Goldman Sachs argued that “sampling requires a homogeneous population in order to form the basis for a valid analysis, and the population here is heterogeneous because it includes employees from different divisions with many different job titles and responsibilities.”

The court suggested a way around the dilemma by use of stratified sampling: “The parties’ objections seem to assume random sampling across the entire universe rather than use of stratified sampling. The latter technique would take into account the heterogeneity of the population by dividing it into subgroups that are each homogeneous with respect to the relevant variables, after which a random sample would be drawn from each subgroup.”

However, in addition to lacking the “expertise to impose any particular sampling technique on the parties unilaterally,” the court also noted the potential burden upon Goldman Sachs to develop database queries and write programming code to retrieve the appropriate data — applicable regardless of the size of the population subject to the search.

Short-term fix

Nevertheless, both parties argued that a dump of the contents of all of the databases could be desirable. Plaintiffs argued for this alternative in light of their desire to prove statistical patterns of discrimination, which would be facilitated by analysis of all of the data, while Goldman Sachs acknowledged that a “data dump” would be less burdensome to them than developing the queries and code involved in a more targeted production.

Although there was not a “legal impediment to ordering production in that form,” the court court observed, “any short term savings from disclosure of the entire databases are likely to be offset by the costs involved in converting the mass of unorganized data provided into useable information.”

The court concluded that under the proportionality test of Rule 26(b)(2)(C), the needs of the case justified the discovery sought. The database information was central to the plaintiffs’ claims, the amount in controversy was “surely substantial,” Goldman Sachs had “ample resources” to respond in discovery, and the “plaintiffs seek to vindicate the civil rights of the class members, and thus further an important public interest.”

In evaluating the burden upon Goldman Sachs to produce the data, the court critically analyzed Goldman’s estimates to extract it. The court engaged in the sampling analysis discussed above as a potential means of reducing Goldman’s burden, but concluded that it would not necessarily do so.

The court basically ordered Goldman to produce the data requested, with the exception of an older PeopleSoft database which was more difficult to query than the newer PeopleSoft database. Goldman would not have to produce data from the older database as long as they agreed that the data therein was not materially different from the data in the newer database.

The sheer volume of data in many cases involving electronic discovery has spawned many strategies that attempt to reduce those quantities in a legally defensible manner. The Chen-Oster case highlights the fact that even the tried-and-true strategies may not always be effective, and the challenge remains to both litigants and the courts to continue to work together to find ways to manage the discovery process.

W. Lawrence Wescott II, Esq., a former IT manager and database development manager, is an e-discovery consultant. He is currently chair of the Technology Committee of the Maryland State Bar Association’s Litigation Section. He can be reached at [email protected]