The obligation to preserve potentially relevant documents when litigation is reasonably anticipated is well-settled and, through case law over the past few years, well-known. However, the scope of that obligation is not clear. Should the concept of proportionality apply to preservation obligations? And, if so, how do you apply it?
The issue of proportionality is of critical importance and, in some cases, dramatic significance. For example, in proposed class actions involving large numbers of employees over extended periods of time, the expense and effort involved in simply identifying potential custodians and preserving potentially relevant data can eclipse the real amount in controversy. To compound the problem, at the inception of a case, when preservation decisions have to be made, determining the scope of the issues involved and the individuals whose data must be preserved is rarely easy or clear. It is commonplace for cases ultimately to be dismissed on motion before discovery—but only after hundreds of thousands, or even millions, of preservation dollars are wasted.
Courts are split on whether, and if so, how, a party may limit the scope of its preservation efforts in a given case commensurate with the likely significance of the information and the amount at issue. U.S. District Judge Colleen McMahon of the Southern District of New York recently added her views to the debate in Pippins v. KPMG, No. 11 Civ 0377, 2012 WL 370321 (S.D.N.Y. Feb. 3, 2012). The court endorsed the concept of proportionality, but pointedly refused to grant KPMG relief from full preservation activities, as a result of KPMG's perceived lack of cooperation in the discovery process and failure to demonstrate that the value of preservation was outweighed by the costs. This case provides useful lessons in evaluating and applying proportionality analysis to preservation obligations.
Plaintiffs in Pippins brought their claims against KPMG as a "collective" action under the Fair Labor Standards Act and New York's Labor Law, on behalf of several thousand current and former employees in the role of audit associate. Plaintiffs claimed that audit associates were improperly classified as supervisors and thereby denied overtime wages to which they would be entitled. The court stayed discovery while it ruled on plaintiff's motion for conditional certification of a collective action. However, the obligation to preserve evidence was not affected by the stay, and KPMG faced the need to address the need to preserve over 2,500 hard drives of former employees who were putative "class" members. This effort alone, KPMG asserted, would cost over $1,500,000 ($600/hard drive).
KPMG sought to negotiate an agreement with plaintiffs whereby KPMG would preserve only a small percentage of these hard drives. KPMG argued that the stored information was either irrelevant or duplicative, and of low value, and that the expense of full preservation was simply not worth it. Plaintiff objected to that bald assertion, and asked KPMG either to submit five randomly selected hard drives for review or to submit to a Rule 30(b)(6) deposition as to their contents. However, KPMG, according to the court, hid behind the stay of discovery, refused the information requests and made a series of "take it or leave it" offers. When those were rejected, KPMG moved for a protective order either permitting it to preserve only a random sample of 100 hard drives, or requiring plaintiffs to pay for any additional preservation activities. KPMG argued that to preserve at a cost of $1,500,000 was more than was at stake in the potential litigation.
Magistrate Judge James L. Cott sided with plaintiffs. Pippins v. KPMG, No. 11 Civ. 0377, 2011 WL 4701849 (S.D.N.Y. Oct. 7, 2011). He ruled that, until the certification motion was resolved, every relevant former employee was a key custodian whose data must be preserved, and directed KPMG to preserve all hard drives until further order, or until the parties reached an agreement. The court noted:
Directing KPMG to continue its preservation in this case maintains the integrity of the materials contained on the hard drives by preventing their destruction. Given the finality that would result from granting KPMG's motion and what stands to be lost—especially as compared to the potential (if not likelihood) that the ongoing preservation can (and will) be limited through sampling once the Motion to Certify is resolved and discovery proceeds—it is not unreasonable that KPMG continue its preservation at this time.…
KPMG's ongoing burden is self-inflicted to a large extent. …KPMG did not provide Plaintiffs with the opportunity to learn of the hard drive's contents—such as by reviewing a handful of hard drives that counsel had vetted for privilege or created a log of contents—that might have enabled them to "propound targeted requests for specific files contained within the hard drives at lesser cost." Id. at *9.
KPMG appealed to McMahon of the district court, advancing the same positions even though the judge, in conditionally certifying the "class" action, had telegraphed her wish that the parties reach a compromise. KPMG's strategy failed. McMahon upheld Cott's order and issued a scathing rebuke of KPMG's tactics in the process. Significantly, the court ruled that proportionality limits are properly considered in weighing preservation obligations. However, she too found that KPMG's positions made it impossible to balance the hardship of preservation against the likely value of the information on the hard drives. Far from agreeing that KPMG was unduly burdened, the court stated:
Frankly, the only things that were unreasonable were: (1) KPMG's refusal to turn over so much as a single hard drive so its contents could be examined; and (2) its refusal to do what was necessary in order to engage in good faith negotiations over the scope of preservation with Plaintiffs' counsel, in purported reliance on an order of this Court [the stay of discovery] that it interpreted unreasonably. It smacks of chutzpah (no definition required) to argue that the Magistrate failed to balance the costs and benefits of preservation when KPMG refused to cooperate with that analysis by providing the very item that would, if examined, demonstrate whether there was any benefit at all to preservation. Pippins, 2012 WL 370321 at *10.
McMahon went yet still further, finding that the hard drives likely would contain probative evidence, because some of the plaintiffs stated that they commonly turned on their computers when they arrived at work in the morning, and turning them off at night—a modern-day punch clock. Because of KPMG's refusal to cooperate in getting the court and plaintiffs the information necessary to evaluate KPMG's claims of undue burden, it was "hoist on its own petard." Id. at *12.
The seminal decision of U.S. District Judge Shira Scheindlin of the Southern District of New York almost a decade ago in Zubulake v. UBS Warburg, 220 F.R.D. 212, 217 (S.D.N.Y. 2003) (Zubulake IV), made clear that the duty to preserve electronic evidence is triggered when a party reasonably anticipates litigation. As noted in Zubulake IV, the obligation is not limitless: It extends only to the relevant documents of "key players," which are those "likely to have discoverable information that the disclosing party may use to support its claims and defenses." Id. at 217-18. Courts have provided less guidance, however, on how to determine who are the "key players," and whether there are limits on the preservation obligation or, if there are, how to define and implement them.
The Pippins court ruled that all audit associates during the relevant time period were "key" players, as they could potentially opt into the certified class. It was exactly this ruling that KPMG had to anticipate in making preservation decisions, and the reason why it sought the protective order. Preserving data in cases involving substantial numbers of potential sources can entail astounding costs. (Imagine, for example, the preservation costs that may have been involved in the Wal-Mart putative class action, which sought to include up to 1.6 million former and current employees as plaintiffs.) The pressure to settle even the weakest of such cases can be overwhelming.
Judges see this dynamic at work, and have long wrestled with imposing reasonable limits. Rule 26 of the Federal Rules of Civil Procedure, which imposes a proportionality standard for disclosure, providing that the court will limit discovery if "the burden of expense of the proposed discovery outweighs its likely benefit." Fed. R. Civ. P. 26(b)(2)(C)(iii). The preservation obligation, however, often is triggered before a case is even filed (see Pension Committee of The University of Montreal Pension Plan v. Banc of America Secs., 685 F. Supp. 2d 456, 475 (S.D.N.Y. 2010)), and is not specified in the Federal Rules, which do not control until filing. See, e.g., Fed. R. Civ. P. 26(d), (f). While courts have relied by implication on the Federal Rules, as well as on inherent judicial powers, to impose preservation obligations and to sanction transgressions (e.g., Victor Stanley v. Creative Pipe, 2010 U.S. Dist. LEXIS 93644 (D. Md. Sept. 9, 2010)), they have been divided as to whether Rule 26's proportionality obligation should be similarly imported into preservation determinations.
Pippins came down squarely on the "proportionality" side. The court found that "[p]reservation and production are necessarily interrelated" and so "proportionality is necessarily a factor in determining a party's preservation obligations." Pippins, 2012 WL 370321 at *11. It "is at the very least relevant to a decision on a motion for a protective order, even if not degenerative of it." Id. That conclusion finds support in other courts as well as in publications of the influential Sedona Conference.
In Rimkus Consulting v. Cammarata, 688 F. Supp. 2d 598, 613 (S.D. Tex. 2010), for example, U.S. District Judge Lee H. Rosenthal found that whether preservation conduct is acceptable "depends on what is reasonable, and that in turn depends on whether what was done—or not done—was proportional to that case and consistent with clearly established applicable standards." See also Victor Stanley, 269 F.R.D. at 523 (an "assessment of reasonableness and proportionality should be at the forefront of all inquiries into whether a party has fulfilled its duty to preserve relevant evidence"); The Sedona Conference, "The Sedona Conference Commentary on Proportionality in Electronic Discovery," 11 Sedona Conf. J. 289, 291 (2010).
The determination to adopt the concept of proportionality was by no means a foregone conclusion, however. Other courts in New York and other jurisdictions have not been as ready to accept proportionality as a component of preservation obligations. While McMahon in Pippins cites for support to Orbit One Commons v. Numerex, 271 F.R.D. 429 (S.D.N.Y. 2010), Magistrate Judge James C. Francis there cautioned that concepts of reasonableness and proportionality "may prove too amorphous to provide much comfort to a party deciding what files it may delete or backup tapes it may recycle" and so do not provide a safe harbor for parties' preservation obligations. Indeed, the Orbit One court observed that "[p]roportionality is particularly tricky in the context of preservation" and that "[i]t seems unlikely, for example, that a court would excuse the destruction of evidence merely because the monetary value of anticipated litigation was low." Id. at 436 n.10. For that reason, the court ultimately advised the parties to "retain all relevant documents" until a more concrete definition of obligations is created by rule. Id.
The decision in Pippins provides support for those who seek to apply the concept of proportionality to preservation decisions. The decision also provides guidance to practitioners who read it closely.
First, Pippins highlights the importance of demonstrating that you are cooperating in the discovery process if you are seeking the application of proportionality in preservation. The decision also underscores the need to marry your e-discovery strategy with litigation acumen. This is not surprising, as most courts loathe discovery disputes, and have increasingly imposed cooperation mandates on parties. KPMG's determination to draw a line in the sand plainly cost it substantial good will before the court. While KPMG no doubt had its reasons for taking this stance, it put itself in the position of seeking the court's endorsement of KPMG's conclusions that full preservation was not worthwhile, while denying its adversary the opportunity to test those conclusions, and the court the benefit of evaluating competing arguments.
Second, Pippins shows that sampling can be very important. The Pippins court noted it could not determine the benefit of saving the hard drives against the hardship of cost because it was "devoid of information necessary to conduct such an analysis." Pippins, 2012 WL 370321 at *12. Sampling could efficiently reveal that only particular types of custodians, limited time frames or specific folders are likely to have any significant share of relevant information. Sampling can also give parties the ability to distinguish between useless and potentially meaningful data stores and provide comfort that unique evidence is not being destroyed. It was the lack of that sort of information that led to the rejection of KPMG's petition for limits or cost-shifting.
Third, Pippins underscores the importance of confronting hard preservation issues with your adversary head on. The court plainly thought that KPMG did the right thing in recognizing its significant problem and bringing it into the open before permitting unique evidence to be destroyed. Many a party has made matters more complicated by putting off key preservation decisions or making risky determinations not to preserve and hoping the issue just never comes up. That is an easy path to sanctions against the company and its counsel—including monetary sanctions, adverse inferences, orders of preclusion and dismissal and ethics charges. E.g., Qualcomm v. Broadcom, 2008 WL 66932 (S.D. Cal. Jan. 7, 2008), rev'd, 2008 WL 638108 (S.D. Cal. March 5, 2008).
Fourth, be prepared to present a detailed analysis of your data. Certainly, KPMG would have been better served by submitting some form of accounting of the contents of a random sample of hard drives. Of course, this whole process is made less harrowing with a pre-established litigation readiness program. There usually is not much time for a party to formulate and execute a game plan to determine its preservation obligations regarding "key players" and the scope of relevant information. It helps to be ready to go to the next step and collect data samples and analyze them.
Finally, if the other side does not cooperate in the process, get an early ruling that sets out the parties' obligations. The track record that you established with good faith meet and confer sessions can go a long way toward showing that you are the one being reasonable.
Pippins shows that if you plan, cooperate, sample, confront the issues and apply for relief early, you will allow the court to take what really is a courageous stand—issuing a ruling that evidence that would otherwise be considered within the preservation obligation may be destroyed—because in the grand scheme of things, it is not worth the effort and expense.
Wayne Matus is a partner, John Davis is counsel and Peter Ostrovski is an associate at Pillsbury Winthrop Shaw Pittman.
1. See, e.g., In re Facebook PPC Adver. Litig., No. C09-3043 JF (HRL), 2011 WL 1324516 (N.D. Cal. April 6, 2011) (granting motion to compel Facebook's participation in creation of ESI protocol). At least one court has dismissed a case where the failure to cooperate was sufficiently egregious. Lee v. Max Int'l, 638 F.3d 1318 (10th Cir. 2011).