In November 2012, the Judicial Conference of the U.S. Advisory Committee on Civil Rules proposed that Rule 37(e) be gutted and replaced with one that allows a court to sanction a producing party for failure to preserve "discoverable information that reasonably should" have been "preserved in the anticipation or conduct of litigation" only when the requesting party can show that the failure was "willful or in bad faith" and caused the requesting party "substantial prejudice," or that such failure "irreparably deprived" the requesting party "of any meaningful opportunity to present a claim or defense."

The proposed amendment, scheduled to take effect in June, would make it far more difficult for requesting parties to obtain sanctions for failure to preserve, as present case law allows for the imposition of sanctions when the failure to preserve is the result of gross negligence and, according to at least one court, "mere" negligence, in addition to willful conduct, and the requesting party need not show any "substantial prejudice" or irreparable harm. In this article, we will explore the wisdom of the proposed rule change.

Problems the Rule Change Seeks To Fix

As it explains in its lengthy notes, the advisory committee’s interest in revising Rule 37(e) arose from a panel discussion at its May 2010 conference at the Duke University School of Law.

"The committee’s focus on preservation and sanctions," the notes explain, "was sparked by the e-discovery panel at the Duke conference in May 2010. That panel discussion emphasized the large and growing burden of litigation holds in particular and preservation more generally. The panelists unanimously urged that the committee attempt to develop a rule to deal with these problems. Rule 37(e), adopted in 2006 to provide some solace about preservation sanctions risk, had not been sufficiently effective."

Later, in the draft committee notes to the actual revisions to Rule 37(e), the draft committee explained how it had learned of the problem.

"The committee had been repeatedly informed of growing concern about the increasing burden of preserving information, particularly before litigation has actually begun. The remarkable growth of information that might be preserved has heightened these concerns."

Lack of guidance, because of disagreements in the case law as to what had to be preserved, exacerbated the problem.

"Extremely expensive over-preservation may seem necessary due to the risk that very serious sanctions could be imposed even for merely negligent, inadvertent failure to preserve some information later sought in discovery."

The committee’s observation that different courts presented different standards for determining whether the loss of ESI to be preserved was sanctionable is certainly accurate.

In Pension Committee of the University of Montreal Pension Plan v. Banc of America Securities, No. 05 Civ. 9016, 2010 U.S. Dist. LEXIS 4546 (S.D.NY 1/15/2010), Judge Shira Scheindlin, author of the Zubulake v. UBS Warburg series of cases that, in many ways, introduced e-discovery to the bar, held that mere negligence may, and gross negligence or willful destruction of evidence certainly will, result in sanctions. She further held that the failure to issue a written preservation hold was per se gross negligence. Finally, she held that if a party allows for the destruction of ESI that should have been preserved, the court can posit that the ESI would have been helpful to the requesting party and so its destruction would be prejudicial to that party.

In Rimkus Consulting Group v. Cammarata, 688 F. Supp. 2d 598 (S.D.TX 2/19/2010), by contrast, Judge Lee Rosenthal held that U.S. Court of Appeals for the Fifth Circuit precedent required that to sanction a party for spoliation, a court had to find willful destruction of evidence, and that evidence prejudiced the requesting party. The court explicitly rejected Pension Committee‘s rationale that negligence, whether "mere" or "gross," was sufficient to warrant a finding of spoliation and sanctions for it, and that prejudice could be posited simply because the producing party allowed the ESI to be destroyed.

The contrast between the holdings in Pension Committee and Rimkus proves the committee correct that case law does not provide sufficient guidance to parties regarding the scope of preservation.

Moreover, it is clear that in choosing between the two positions, the committee wished to go with Rimkus. It did not point to Pension Committee directly, but did note that it very much wanted to make sure that the rules committee did not adhere to the standard set in Residential Funding v. DeGeorge Financial, 306 F.3d 99 (2d Cir. 2002), which sanctioned a producing party for "mere negligence" in allowing data that should have been preserved to be destroyed. Residential Funding is a U.S. Court of Appeals for the Second Circuit opinion that certainly controlled Pension Committee. By explicitly rejecting Residential Funding, the committee implicitly rejected Pension Committee.

Turning to the actual, practical problem that the committee believed created the need for the rules change, i.e., "the large and growing burden of litigation holds in particular and preservation more generally," caused by the "remarkable growth of information that might be preserved," it is unclear how the committee gathered the facts to determine that such was a critical problem, other than from complaints from counsel at or outside of committee meetings. I have been at conferences dedicated to e-discovery where counsel has voiced that complaint, but, as someone who is involved on a daily basis with the preservation, collection, processing, searching, review and production of e-discovery, I have not heard that complaint with regard to more than a few particular matters. Case law is by no means a good way of assessing the depth of the problem, as the problem may simply not make its way into the case law, for many reasons.

Two recent cases, for example, suggest that the problem is not as dire as advertised. In Pippins v. KPMG, 279 F.R.D. 245 (S.D.N.Y. Feb. 3, 2012), potential class-action plaintiffs demanded that the defendant preserve 2,500 to 4,000 hard drives at a cost of, by the defendant’s estimate, $1.5 million. The defendant moved to be permitted to preserve a far smaller group. The problem with the defendant’s argument, however, was that while the defendant claimed it could select a "representative sample," it refused to look at any of the drives, reasoning that discovery had been put on hold pending class certification. The court held, in essence, that the defendant could not have it both ways, refusing to look at the drives because discovery had not commenced but claiming at the same time that it "knew" its selection of drives would be representative.

The defendant having failed to prove that it could preserve a small subset of drives without losing any important data, the court denied its motion. The plaintiffs’ class certification ended up being denied; had the defendant acted more thoughtfully, it could have saved itself a great deal of money preserving drives that ended up not being used as evidence.

By contrast, the defendant in Vaughn v. LA Fitness, 285 F.R.D. 331 (E.D.PA. Aug. 16, 2012), convinced the court that the requesting party’s preservation demand was overbroad, such that the court held that if plaintiffs wished the data to be preserved and searched, they would have to pay for it. By the time the issue arose, the defendant had gotten involved in "active discovery management," a great number of documents had already been produced, the documents sought by the plaintiffs were in many ways duplicative of ones that had already been searched and had yielded little of value, and the cost to comply with the plaintiffs’ demands would run several-hundred-thousand dollars.

In the face of this record, the court concluded that the cost of preservation would be great and the benefit likely small. That being the case, the court held that if plaintiffs wanted to preserve and search the records, they could do so if they footed the bill.

Pippins and LA Fitness do not prove that the "problem" of over-preservation is not a real one, but neither do the complaints of counsel who, typically, represent producing parties that the problem is prevalent prove that it is. The problem may, indeed, be prevalent, but the committee did very little to justify empirically the need for such drastic change in Rule 37(e).


What is perhaps most striking about amended Rule 37(e) is what it does not address: when the duty to preserve arises and what it entails. Instead, the rule focuses solely upon limiting when sanctions can apply. Outside of obvious examples — an Oliver North "shredding party," for example — it is hard to know how to determine whether a failure to preserve was "willful or in bad faith." Going back to Zubulake, 220 F.R.D. 221 (S.D.N.Y. 2003) (Zubulake IV), Scheindlin discussed the many factors that could be seen as triggering the duty to preserve, such as an event that everyone recognizes will lead to litigation (e.g., the BP Deepwater Horizon explosion), formal notification from counsel or informal notification from the plaintiff that an action would be brought, administrative hearings on the same subject matter (such as the EEOC action that preceded the complaint in Zubulake) and, of course, the complaint itself.

In its notes, the committee explains that there are too many factors that could give rise to the duty to preserve to make listing them possible, and the committee did not want to create bright-line tests for when the duty is triggered. While certainly there are many factors, and the context in which they arise makes it impossible to provide bright lines, the committee’s solution is wanting, for two reasons.

First, there are many instances where statutes, rules or opinions provide guidance with regard to their specific subject matter by referencing factors with the caveat that the list is not exhaustive and the meaning of such factors must be understood in the context of the facts of any particular matter. Second, while, as just discussed, there are limitations built into any list of factors, if the point of amending Rule 37(e) is to provide guidance to those who must preserve ESI (and discovery in general), refusing to list such factors hurts, rather than helps, any attempt to provide guidance and regularize practices.

About 200 years ago, Voltaire wrote that the perfect was the enemy of the good. A list of factors to take into account when determining whether the duty to preserve has arisen would certainly not be perfect, but it would be good, and much better than not having such a list.

By focusing solely on when the remedy of sanctions can be employed, the proposed amendment to Rule 37(e) does very little to provide the guidance the committee claims it wants to provide. What constitutes a willful action or one done in bad faith is by no means clear.

Let us imagine a typical e-discovery scenario. A business is sued. The business has 50 computer users, each with a laptop computer, an email box on their email server and a file share on the file server. In addition, users can gain remote access to the servers from their home computers and save to those home computers. Users can also store email and e-docs to the C drives of their laptops.

The business determines that anywhere from 10 to 25 people may have ESI responsive to the discovery requests. They issue a litigation hold memo to all 25 people. They take no additional steps to preserve data, such as copying the data from the computers at the time they learn of the litigation.

Eighteen months pass and the legal wrangling focusing on the opening portions of the litigation is over. Discovery now must be produced. When the ESI is collected from all of the users, the emails and e-docs of several users reveal large gaps in time where it appears that no emails or e-docs exist. The users explain that they must have deleted these items, but assure counsel that none of them was responsive.

Was the failure to preserve willful or done in bad faith? Certainly, the actions taken to delete were willful, i.e., nothing was deleted by accident. If, however, willfulness or bad faith means the express intent to destroy potential discovery, such could not be proven by the deletions here. There is, however, another way to see the actions here as willful or in bad faith. The business knew it had a duty to preserve. It chose to try to discharge that duty by simply telling users not to delete anything for 18 months. It was highly unrealistic that such would be accomplished successfully; it was far more likely that during that long time period, files would be deleted.

By refusing to recognize that the steps it was taking to preserve would not do the job, the business was willfully choosing to put the money saved by not collecting data from the start — the best way to preserve it — ahead of the duty to preserve. Just as a drunken driver may not intend to hit his victim, but certainly does intend to drive home from the bar or party and not use a designated driver or call a cab. So the business here did not intend to lose data, but did intend to rely upon a means of preservation that was not a very good one over so long a period of time.

If asked why the business made that decision, its response would have been quite rational: Why bear the potentially unneeded expense of data collection when the case could have been dismissed before discovery even commenced. Would that mean the business did not act in bad faith? From one point of view, the answer is no. The business knew that trusting users not to delete data over an 18-month period was by no means the best way to preserve data, but chose that way anyway because it spared the business an upfront expense. To some, that is common sense, while to others, it is the essence of bad faith: acting as you see fit because to act otherwise would be to inconvenience you.

If determining whether an action was willful or done in bad faith is subjective, whether the loss of data caused the requesting party "substantial prejudice" or "irreparably deprived" the requesting party of a meaningful opportunity to make its case is simply unknowable. By definition, the contents of data destroyed cannot be known. The sole exceptions to that rule are unlikely events: the requesting party already had a copy of the ESI it sought, or the requesting party otherwise knew what those contents were (by reading the files, for example).

Those exceptions, however, account for less than 1 percent of all cases; in the other 99-plus percent of cases, the contents of the files go missing with the files. As we have discussed, in Pension Committee, Scheindlin resolved this is by drawing the inference that if the ESI was important enough for the producing party to destroy, it must have been important enough that its loss irreparably deprived the requesting party of important discovery. That inference may be dubious, but the advisory committee certainly did not come up with a better solution.

Thus, while a fair number of cases are devoted to the question of when the duty to preserve arises, in the typical case, the question is more about what one must do to preserve and whether failure to take action is bad faith or simple business common sense than it is about when one must start to preserve. The requesting party’s attorney will say the former, the producing party’s attorney the latter, and the question of how far a party need go to preserve ESI is no better answered than before Rule 37(e) was amended.


In his Dec. 7, 2012, blog, Ralph Losey, whose writings on e-discovery are informed and thoughtful, reported on Georgetown Law School’s Advanced E-Discovery Institute, held the previous week. A distinguished and impressive list of judges attended and spoke. The most shocking comments, Losey reported, came from Scheindlin, who expressed her disappointment with the proposed rules changes.

According to Losey, Scheindlin was disappointed that the rules committee did not have the "courage or vision to write a rule on preservation." As he notes, "Yes. Those were her exact words." She would have written a clear rule governing preservation, while the committee "only looked through the lens of failure" and addressed just sanctions (again, Losey emphasized that he was quoting her verbatim). What was needed, she emphasized, was "national guidance on preservation, instead of ad hoc case-law that varies by jurisdiction" (Losey’s words).

The shortcomings Scheindlin saw in the proposed amendment to Rule 37(e) are real and clear. It is, admittedly, not easy to provide clear guidelines as to when and how to preserve ESI. It is, however, more harmful than helpful to shred a rule that had not much of an effect to replace it with one that responds to a problem the evidence of whose existence is anecdotal at best, and with rules that focus on remedies for bad conduct, poorly defined, rather than on telling litigants what is good conduct. If teaching teenagers how to drive were done in the same way, i.e., don’t tell them to stop at the red light, just yell at them when they rush through the intersection, one can only imagine how much more dangerous driving would be. •