The end of the ostrich defense: While it did not set new precedents or change the world, Zubulake v UBS Warburg signals a big shift--rational analysis is at least entering the realm of electronic records and the courts.

The information management world has been much abuzz of late over a series of interlocutory orders in the case of Zubulake v. UBS Warburg.

Since the filing of the case in 2002, federal judge Shira Scheindlin of the Southern District of New York has issued five opinions and orders, gaining her a reputation as the scourge of sloppy electronic records management. Scheindlin's July 20, 2004, order (Zubulake V) imposed hefty sanctions on the defendant, including the much-feared sanction of an adverse inference and an accompanying jury instruction, meaning the jury was told it could infer that the defendant destroyed potentially relevant evidence because the company feared the evidence would be unfavorable.

Zubulake continues to be an instructive case for those interested in the legal system's attempts to deal with records management issues in general and electronic records management issues in particular. In prior orders, Scheindlin considered the question of costs and difficulties of electronic discovery and the proper allocation of the costs between the parties. In so doing, she developed a test for determining the appropriateness of cost-shifting, taking into account prior authority such as the Rules of Civil Procedure and prior case doctrine, and ordered data sampling to be done in order to determine the potential relevance of e-mails located on backup tapes (Zubulake I). In a later order (Zubulake III), she applied that test and allocated costs for restoring backup tapes based upon the results of the data sampling done pursuant to her prior order. In the current phase of the case (as of the end of 2004), Scheindlin has had occasion to examine and rule upon the results of her earlier orders.

The Significance of the Case

Zubulake is not itself a particularly noteworthy or groundbreaking case. As noted by the judge herself, it is a routine gender discrimination suit involving no novel facts or questions of law. Nor are the discovery issues that have presented themselves novel; the discovery under dispute is routine discovery of e-mail on active servers, archives, and backup tapes, an issue likely to be familiar to most records and information managers, many of whom have had to undertake similar discovery at their own organizations. What is novel and instructive is the court's handling of the issues presented during the dispute and fact-finding process.

Subsequent to Zubulake III, the parties restored the backup tapes in question, and e-mail from them was recovered and given to the plaintiffs. During the restoration process, it was determined that some backup tapes had been destroyed or otherwise rendered unrecoverable. Although at least some of the e-mail contained on those tapes was still available from other tapes due to the redundant nature of the backup process, Zubulake contended that critical e-mail might be permanently unavailable.

Analysis revealed that some e-mails had been deleted from the active system after a duty to preserve them was attached. In some cases, this apparently resulted from miscommunication between counsel and UBS Warburg employees, but in other cases, no explanation of the destruction was offered. Zubulake claimed that this amounted to spoliation of evidence. She therefore sought a variety of sanctions, including restoration at the defendant's cost of still more backup tapes, re-deposition of certain key witnesses at the defendant's expense, and an adverse inference instruction from the court.

In an order dated October 22, 2003 (Zubulake IV), the court discussed at length several questions of interest:

When does the duty to preserve evidence attach? Since some of the missing e-mail apparently pre-dated the filing of the Equal Employment Opportunity Commission (EEOC) complaint at the center of the case, the court considered the time at which the duty to preserve the e-mail attached. The e-mails already produced had a considerable impact on this question, since in the court's mind they indicated that all of the relevant employees were worried about litigation well in advance of the actual filing, many even going so far as to tag e-mail with "attorney-client privilege" when in fact no attorney was involved in the e-mail thread. The court concluded that the duty to preserve was triggered in April 2001, five months before the filing of the EEOC complaint.

What should be preserved? The court considered the impact of a duty of preservation which would force retention of every paper document, e-mail, or electronic document in anticipation of litigation, and noted that "[s]uch a rule would cripple large corporations, like UBS, that are almost always involved in litigation." The court did, however, conclude that "[w]hile a litigant is under no duty to keep or retain every document in its possession ... it is under a duty to preserve what it knows, or reasonably should know, is relevant in the action, is reasonably calculated to lead to the discovery of admissible evidence, is reasonably likely to be requested during discovery and/or is the subject of a pending discovery request." Further, "[a] party or anticipated party must retain all relevant documents (but not multiple identical copies) in existence at the time the duty to preserve attaches, and any relevant documents created thereafter."

In considering the issue of short-term backup tapes and their continued use and recycling, the court concluded that their preservation in anticipation of litigation might not always be required, with one important caveat: "If a company can identify where particular employee documents are stored on backup tapes, then the tapes storing the documents of 'key players' to the existing or threatened litigation should be preserved if the information contained on those tapes is not otherwise available. This exception applies to all backup tapes."

How should it be preserved? The court did not attempt to impose any particular solution on litigants:

   In recognition of the fact that there are
   many ways to manage electronic data,
   litigants are free to choose how this
   task is accomplished. For example, a
   litigant could choose to retain all then-existing
   backup tapes for the relevant
   personnel (if such tapes store data by
   individual or the contents can be identified
   in good faith and through reasonable
   effort), and to catalog any
   later-created documents in a separate
   electronic file. That, along with a mirror-image
   of the computer system
   taken at the time the duty to preserve
   attaches (to preserve documents in the
   state they existed at that time), creates
   a complete set of relevant documents.
   Presumably there are a multitude of
   other ways to achieve the same result.

None of this is groundbreaking--it is merely a careful, detailed, and reasoned application of longstanding discovery rules to the area of electronic records. Nor is it particularly oppressive--the judge explicitly recognized the need to limit the duty of preservation and explicitly recognized the existence of alternative methods of complying with the rule.

Neither were the results earthshattering.

In analyzing Zubulake's motion for an adverse inference for spoliation of evidence, the court used a standard, three-part analysis in which a showing must be made that

1. the party having control over the evidence had an obligation to preserve it at the time it was destroyed

2. the records were destroyed with a "culpable state of mind"

3. the destroyed evidence was "relevant" to the party's claim or defense such that a reasonable trier of fact could find that it would support that claim or defense

The court specifically discussed the issue of "culpable state of mind": "In this circuit [United States Second Circuit], a 'culpable state of mind' for purposes of a spoliation inference includes ordinary negligence. When evidence is destroyed in bad faith (i.e., intentionally or willfully), that fact alone is sufficient to demonstrate relevance. By contrast, when the destruction is negligent, relevance must be proven by the party seeking the sanctions [citations omitted]."

Notwithstanding her conclusion that e-mails and backup tapes had been destroyed after the duty to preserve them attached, Sheindlin rejected the demand for an adverse inference. The court concluded that parts one and two of the test had been met: the duty to preserve had attached at the time the e-mails and tapes were destroyed; and the destruction was at least negligent, and in some cases grossly negligent or reckless, and thus culpable.

However, the court was unpersuaded that a showing of relevance had been made: "This corroboration requirement is even more necessary where the destruction was merely negligent, since in those cases it cannot be inferred from the conduct of the spoliator that the evidence would even have been harmful to him. This is equally true in cases of gross negligence or recklessness; only in the case of willful spoliation is the spoliator's mental culpability itself evidence of the relevance of the documents destroyed. [citations omitted]."

In light of this requirement, the court concluded that the relevance requirement had not been met: the e-mails produced thus far, while showing a clear pattern of improper conduct, did not show that the conduct was gender-related and the likelihood of proving this with further e-mails was deemed by the court to be low. Thus, an adverse inference was unwarranted. The court did, however, permit the re-deposing of key witnesses at UBS Warburg's expense for the limited purpose of exploring the issue of destruction of evidence and any newly discovered e-mails.

At this point it appeared that UBS Warburg had effectively met and parried Zubulake's discovery efforts. Some costs had been shifted to Zubulake, allegations of spoliation had resulted in minimal sanctions, and nothing conclusively supporting a gender discrimination claim had been uncovered.

The Roof Falls

The complexion of the case changed when the newly ordered depositions uncovered a long series of improprieties at UBS Warburg, which were cited in Zubulake V and included:

* failure of counsel to adequately inform and instruct all relevant UBS Warburg employees as to their duties regarding the preservation and turning over of all relevant evidence

* failure of counsel to request relevant information from key employees

* failure of counsel to inform themselves about how employees were maintaining relevant evidence, including e-mail

* failure by employees to produce relevant material, including e-mail, to counsel

* deletion of e-mails by UBS Warburg employees after having been instructed in writing and personally by counsel to retain them

* failure to safeguard backup tapes containing relevant e-mail

Some of these failures arose from classic information management miscommunications: In one instance, an employee told counsel that she maintained the relevant e-mails in an "archive." Counsel thought she meant a backup tape, while she meant only an e-mail folder on her computer. Others arose from simple failure to follow up on orders given: backup tapes were destroyed or overwritten because no one followed up to see whether and how the order was being complied with.

Still others had no innocent explanation: notwithstanding clear instructions from counsel (communicated, ironically, via e-mall), employees deleted relevant e-mail from their systems. The problem was compounded by the fact that many backup tapes were missing, notwithstanding a retention policy requiring their preservation and orders from counsel to preserve all relevant backup tapes.

Some of the e-mail was recoverable from backup tapes or other sources, thus proving the deletion from the active system. Other e-mail was apparently gone completely. The deletion of this e-mail was proven through two methods: testimony by witnesses during depositions and by reading other e-mail, some of which clearly referred to e-mail that had vanished from active systems or, in some cases, entirely.

The result of all of this was that Zubulake was only given some relevant material two years after it should have been produced and that some material--who knows how much?--could not be produced at all.

The court made a number of relevant observations about the discovery process and the duties of counsel:

* Counsel must actively oversee and direct the discovery and preservation process--merely issuing an order or memo is not enough.

* Counsel must meet with key players in the litigation to ensure they understand their role and duties.

* Counsel must take steps to protect relevant data.

* Counsel must be familiar with the client's document retention policies.

But in the final analysis, the failings were the responsibility of the client. The court therefore revisited its earlier decision on an adverse inference and concluded that in light of the newly discovered facts, an adverse inference and jury instruction was appropriate. This is an extreme penalty. In the course of rejecting Zubulake's demands for an adverse inference in Zubulake IV, Scheindlin noted: "The in terrorem effect of an adverse inference is obvious. When a jury is instructed that it may 'infer that the party who destroyed potentially relevant evidence did so out of a realization that the [evidence was] unfavorable," the party suffering this instruction will be hard-pressed to prevail on the merits. Accordingly, the adverse inference instruction is an extreme sanction and should not be given lightly [citations omitted]."

Lessons to be Learned

As of the end of 2004, the final out come of Zubulake had yet to be known--there may be still more developments which affect it. Indeed, we may never learn the details of the outcome as the specter of an adverse inference may spur UBS Warburg into settlement talks, and if the case settles, the details of the settlement will no doubt remain secret. We can, however, learn many things from the events that have thus far occurred:

Culpability: The judge's final determination that UBS Warburg ought to be sanctioned with an adverse inference for spoliation of evidence was predicated on a finding of it or its agents having a culpable state of mind. Yet, at no point did the judge make an explicit finding that anyone actually deleted or destroyed anything with the intent to prevent Zubulake from getting it. This might be inferred from the overall tone of Zubulake V, but nowhere is it stated. The point is that it is not needed. Enough errors by counsel and litigant, combined with enough prejudice to the other party, may rise to the requisite level of culpability, regardless of actual provable motives on the part of employees. Other organizations whose electronic records and discovery processes are in disarray would do well to con sider the implications of this.

Discovery Management: Many of the issues ultimately giving rise to sanctions were apparently the result of failure to communicate between counsel and UBS Warburg employees, as well as the failure of counsel to supervise discovery efforts. UBS Warburg is a large and distributed organization; close supervision of discovery in many locations, including some outside of the United States, in such a case may be a formidable job. Warburg is nut unique in this way; many other organizations are equally large and equally distributed. For all such organizations, communication and active involvement of counsel and supervisory personnel is a key factor in avoiding sanctions. Merely issuing a notice of litigation hold may well he inadequate and will be no defense if issues arise.

There is another lesson here as well: the line employees actually involved in a dispute such as this may not be the most reliable custodians when it comes to retaining things like e-mail. Procedures for ensuring that relevant material is captured and removed from their control early on may save the organization many headaches later.

Backup Tapes and Retention Policies: UBS Warburg was burned twice by its handling of backup tapes. Missing and poorly handled tapes provided part of the basis for sanctions for spoliation of evidence, while the backup tapes that did exist helped to prove the spoliation because they contained copies of deleted e-mails. The lesson is simple: handle backup tapes consistently. If the policy states that a year's worth of tapes will be kept, then a year's worth--not six or some other arbitrary number of months' worth--should be kept.

Another lesson is equally simple: less is better. The reason UBS Warburg was sanctioned for not having a full year's worth of backup tapes is because it had a retention schedule that said the company retained tapes for a year. Strictly adhering to the retention schedule makes life a lot simpler.

Judges and Lawyers: Sheindlin took the trouble to learn a lot about electronic records management during the course of this case and used that knowledge to analyze the facts in considerable detail. She was, no doubt, ably assisted in gaining this knowledge and doing this analysis by Zubulake's counsel. Nonetheless, the result was that she provided a very sound analysis of the failings of UBS Warburg's electronic records management and discovery processes, as well as sound and commonsense guidelines for future litigants.

The lesson here is that litigants can expect more of this judicial understanding. The orders in this case have been widely publicized, and the analysis and commentary are regarded in the legal community as cutting edge. To the extent that Scheindlin was aided or prodded in her analysis by Zubulake's counsel, it is likely that other lawyers will do the same, as will other judges.

The orders and analysis in this case provide a good roadmap for future litigants contemplating an allegation of spoliation of evidence, so if an organization's discovery responses are inadequate, they can be expected to be challenged. If an organization claims backup tapes contain nothing relevant, it should be prepared to prove it. It should be assumed that whatever e-mail is produced will be gone over with a fine-toothed comb and that material from active systems will be compared to what resides on the backup tapes, with discrepancies being brought to the court's attention.

Each of these lessons can and should be incorporated into the management of electronic records and the implementation of litigation discovery at all organizations. Careful consideration of the issues resulting in sanctions will reveal that the fixes are not, by and large, difficult to design or implement. The key is anticipation. We can learn from the past. Nowhere is this more true than here.