Court Clarifies its own Decision in a Key E-Discovery Case

Robert B. Friedman and David J. Kessler
Executive Counsel August/September 2010

ANYONE WHO HAS MANAGED electronic discovery in modern litigation knows that no matter how carefully you imp-lement document preservation procedures and supervise collection efforts, electronically stored information (ESI) can be lost. It doesn’t take negligent or wrongful conduct.

Now, in an amendment to an important legal precedent, this seems to have been acknowledged by way of an explicit recognition that data loss can occur even when reasonable steps have been taken to preserve it.

A KEY REWRITE

Judge Shira Scheindlin, the author of the seminal Zubulake opinions that may have done more to define a party’s obligations to preserve in the digital age than any other opinion, initially created controversy when she revisited preservation in her recent opinion in Pension Committee of the Univ. of Montreal v. Banc of America, in January of this year. She acknowledges at the beginning of the opinion that perfection is not the standard, but then she borrows from the world of tort law and analyzes the concepts of negligence, gross negligence, recklessness, and willfulness in the discovery context, suggesting that these familiar standards provide the appropriate framework for evaluating discovery conduct.

No doubt deeply troubled by the plaintiffs’ deplorable conduct in this case, she spends little time discussing how and why specific mistakes occurred. Rather, having determined that the conduct was unacceptable, she focuses solely on the results of the conduct: Data was lost.

In so doing, the court appears to apply a strict liability, or at least a res ipsa loquitur approach. (“The thing speaks for itself.”) That is, the measures that were or weren’t taken to preserve the ESI are irrelevant. The loss of the data itself is evidence of spoliation.

However, on May 28, 2010, Judge Scheindlin amended her ruling on her own and without request by any party, by altering a single sentence. “Replace,” she says in her amendment, “… the failure to obtain records from all employees (some of whom may have had only a passing encounter with the issues in the litigation), as opposed to key players, likely constitutes negligence as opposed to a higher degree of culpability,” with “[T]he failure to obtain records from all those employees who had any involvement with the issues raised in the litigation or anticipated litigation, as opposed to just key players, could constitute negligence.”

It takes a careful read to see it, but this seemingly minor change helps resolve a significant internal conflict that runs through the opinion. It abandons the idea that any data loss must, at the least, constitute negligence by explicitly acknowledging that the data could have been lost without there being negligence.

The result is to take the focus of the inquiry off the question of whether ESI was lost, and on to the issue of why it was lost.

Was it, for example, due to insufficient efforts to preserve it, or would it have occurred regardless of what reasonable measures had been taken?

In the first version, it was the apparent absence of inquiry into the quality of the plaintiffs efforts to preserve that had caused the tension and potential for confusion.

THE PENSION COMMITTEE CASE

Pension Committee involved a group of investors who sought to recover $550 million from the liquidation of an off-shore hedge fund. In April 2003, the fund’s management filed for bankruptcy and the money was placed in receivership. In the spring and summer of 2003, several investors conferred with or retained counsel in preparation for a lawsuit against management.

By the fall of 2003, the investors had retained common counsel, and in February of 2004 they filed suit, asserting claims against several defendants under federal securities laws.

In June of 2004, the defendants filed several motions to dismiss, which caused the case to be stayed pursuant to the Private Securities Litigation Reform Act. In February 2007, the motions to dismiss were denied, and the stay was lifted. Discovery commenced in earnest, with requests for production by the defendants.

Subsequent depositions of the plaintiffs raised suspicions about how carefully the plaintiffs had preserved and collected relevant documents. The defendants raised these concerns with the court, requesting that the plaintiffs submit declarations detailing their discovery efforts. The court granted the request, declarations were submitted, and more depositions followed.

The defendants determined that for certain plaintiffs, at least 311 documents were wrongfully omitted. The defendants also discovered that the declarations that plaintiffs had submitted were “false and misleading and/or executed by the declarant without personal knowledge of its contents.”

As result, the defendants moved for dismissal. Although the district court declined to dismiss the case, it determined that sanctions were appropriate, and it crafted an adverse inference instruction.
THE COURT’S ANALYSIS

OF CULPABILITY

In Pension Committee, the court starts with the basic axiom that destruction of evidence is only sanctionable if the spoliator has a culpable state of mind, and unacceptable conduct lies along a spectrum of culpability, ranging from negligence to gross negligence to willful, wanton, or reckless conduct.

The court defines these categories by importing definitions from tort law. Negligence is conduct that falls below the standard established by law. Gross Negligence is failure to exercise even the care that a careless person would use. Willful, wanton, or reckless conduct is doing something when you know or don’t care that harm will follow.

Having described the categories of wrongful conduct that can occur, the court instructed that these definitions should be applied in the discovery context, with errors or omissions being identified and placed into one of the categories.

The important feature of these definitions is that they focus on the behavior of the actor, and whether the conduct was reasonable according to prevailing standards. They do not focus on the results of the conduct.

However, because of the amount of money at stake – a half billion dollars, according to the plaintiffs – and the undeniably poor conduct of the defendants, the court appears to import the theory of res ipsa loquitur into the spoliation analysis. The Pension Fund court initially looked at the outcome to determine culpability, reducing the spectrum of discovery conduct into two choices, regardless of circumstances: acceptable or unacceptable. If data is lost, then the conduct must have been unacceptable.

For example, the court stated that failure to preserve all relevant evidence, the failure to collect evidence, or the careless collection of evidence is “surely negligent and, depending on the circumstances, may be grossly negligent or willful.” This could be read to rule out the honest mistake.

But in the May 28 Amendment, Judge Scheindlin rejects the notion that the loss of data must entail negligence. As the sentence in the original opinion read, the loss of data from tertiary custodians would likely lead to negligence rather than the higher degrees of culpability, but it could be read to leave no room for even lesser degrees of culpability – that is, no culpability.

As amended, Judge Scheindlin’s opinion not only does away with the higher degrees of culpability in this scenario, but finds that even negligence is only a possibility. It depends on the actual conduct of the parties. This clarification is not only more in line with existing case law and doctrine, but also is more in line with her next sentence about taking appropriate (as opposed to all) measures to preserve ESI.

THE TAKEAWAY

Pension Committee reaffirms the core tenets of a party’s obligations to preserve (particularly in the Second Circuit). Parties must act reasonably or risk sanctions that will become more severe as their actions deviate from standard norms. The question then becomes: What actions should a party reasonably take to preserve potentially relevant information? Where does the marginal benefit of the preservation action outweigh the marginal costs?

Under this rubric, written hold notices to employees are generally reasonable. (The failure to issue one was gross negligence, in Judge Scheindlin’s view.) The cost of issuing them is low, and they alert custodians to avoid deleting or modifying the potentially relevant data. These notices are potentially a critical step in the typical decentralized management system, where employees manage their own files. In short, the bang for the buck in a hold notice is substantial.

Still, there are exceptions, notably where there is concern that the notice may alert particular employees who have the incentive to delete potentially relevant data (e.g. in the case of a supervisor accused of harassment or an employee accused of embezzling). In these circumstances, the reasonable preservation step may be to undertake other actions before issuing the hold – actions like capturing an employee’s data before notifying the employee at all about the dispute.

What counsel and parties cannot do is be willfully blind to their preservation obligations. Parties must perform a reasonable search to find and preserve potentially relevant data, and the more important the data, the more effort they must expend to preserve it.

This isn’t new. In its first set of principles, articulated more than five years ago, The Sedona Conference recommended that parties not be required to expend “heroic” efforts to preserve data or preserve deleted, residual or fragmentary data.

On the other hand, where data is material, substantial efforts are warranted. For example, in the 2007 Columbia Pictures v. Bunnell opinion, in the Central District of California, the IP addresses that were not being logged by the defendant Bunnell were deemed “essential” to the case and “extremely relevant.” As such, even though they existed in RAM for no more than six hours, the court ordered them preserved (especially given that it found the burden of doing so was not significant).

Pension Committee revisits the law of preservation that was so articulately laid out in the Zubulake cases. Initially some of Judge Scheindlin’s language could have been seen to require preservation of all data from all custodians regardless of their connection to the litigation. This reading has always been inconsistent with Judge Scheindlin’s Zubulake opinions and now is clearly at odds with her May 28 amendment.