Crashed Hard Drives: When in Doubt, Don’t Throw It Out

Originally published on New York Law Journal.

By Stephen M. Kramarsky

Just about every practicing litigator has a story about electronically stored information, or ESI: how it won the case or lost the case or turned up or disappeared. It’s fair to say that electronic discovery has entirely transformed litigation over the last 20 years and there is no shortage of court opinions and commentary on the subject. In the federal courts, a series of opinions written in 2003 and 2004 by Judge Shira A. Scheindlin in Zubulake v. UBS Warburg¹ planted the seed that has grown into the modern ESI discovery regime. Specifically, these opinions focused not only on the parties’ obligations to produce ESI, but to preserve it—even before litigation begins.

Since Zubulake, numerous courts have addressed how the discovery rules—which are general and apply to all types of materials—specifically apply to ESI. These opinions address when the obligation to preserve arises, what must be preserved, and when and how ESI must be searched and produced. As the law has developed, it has become clear that the obligation to preserve and produce ESI is broad, and a restrictive definition of ESI (email and electronic documents) is insufficient. ESI in the modern litigation landscape may include text messages, instant messages, voicemail, digital video, audio and images, and even social media postings. It may also include the metadata associated with those files and the media on which they reside, if that information is relevant to the case. An entire industry of ESI management products and consultants has sprouted and flourished in recent years to comply with the requirements of ESI preservation and production. But even now, a dozen years after Zubulake, new questions continue to arise.

In Dorchester Financial Holdings v. Banco BRJ S.A.,² the litigants recently presented the court with a question of first impression, at least in the Southern District of New York: May a litigant dispose of a hard drive that might contain relevant, discoverable information if that hard drive has “crashed?” If not, what is the appropriate sanction for doing so?

In Dorchester, plaintiff’s attorney had destroyed a hard drive that he claimed had “crashed” even though the hard drive contained documents related to the case. The attorney did not inform the defendant or the court until discovery began two years later. Defendant sought sanctions for spoliation of evidence, which the court granted. In coming to its decision, the court had to address several open issues of law, and in doing so provided some insight into how courts analyze spoliation, particularly when it involves rapidly advancing technologies such as forensic data recovery.

Facts

In 2011, Dorchester Financial Holdings sued the Brazilian bank Banco BRJ, S.A. for the second time, alleging that BRJ had committed breach of contract and fraud by failing to honor a $250 million letter of credit that it had issued to Dorchester a decade earlier in 2001. BRJ moved to dismiss the complaint, arguing that the New York court lacked personal jurisdiction. BRJ claimed that it had never done business with Dorchester and that the $250 million letter of credit was counterfeit. Dorchester opposed the motion, maintaining that the letter of credit was authentic.

While arguing that the court had personal jurisdiction over BRJ, Dorchester unveiled a new allegation. It asserted that, just days before issuing the letter of credit, BRJ had executed a contract with Dorchester that contained a choice of law provision requiring BRJ to adjudicate all disputes related to the letter of credit in New York. Dorchester attached a copy of the contract to its brief. BRJ responded that the contract, like the letter of credit, was counterfeit. Shortly thereafter, Dorchester amended its complaint to allege breach of the newly discovered contract and BRJ filed another motion to dismiss on the same grounds as before. The trial court dismissed the case, but Dorchester successfully appealed, and the case moved on to discovery phase.³

Thus it was years later, in February 2014, that BRJ served its first document requests on Dorchester. Naturally, the requests sought communications and documents, including metadata, related to the 2001 contract that Dorchester was suing under. In response, Dorchester produced some documents, but appended a surprising statement to the production: Dorchester did not possess any documents related to the letter of credit or the contract except for those it had already disclosed, because “[a]ny such data was destroyed by computer failure between 2001 and 2013.”⁴ This was the first time that Dorchester had mentioned that it had lost any documents to either BRJ or the court.⁵ BRJ responded to this admission by filing a motion for sanctions for Dorchester’s spoliation of evidence.

The court ultimately found⁶ that the actions of Dorchester’s attorney, T.J. Morrow, who was also a corporate officer, constituted spoliation and merited sanctions. In 2002, in connection with a previous, similar action against BRJ, Morrow had gathered all of Dorchester’s hard copy and electronic documents related to the dispute. Morrow saved and kept all of the electronic documents on his personal computer, including the 2001 contract and related emails. Morrow printed some of these electronic documents in August 2011, while Dorchester was opposing BRJ’s first motion to dismiss and filing its amended complaint, and later printed more in March 2012, during the appeal. Other documents related to the case were on the computer, but Morrow chose not to print them.

According to Morrow, shortly after printing these documents, his computer “crashed” sometime in March 2012. Although Dorchester’s case was open and active at that time, Morrow chose to not inform BRJ or the court. Instead, he had a family member with no formal computer training look at his computer. The family member told him that the computer “was basically gone and that [Morrow] needed to get a new machine.” He advised Morrow to transfer anything he “could take off of it.”⁷ Morrow never consulted a computer specialist, and instead opted to dispose of the computer, despite knowing that its hard drive had never been backed up or forensically imaged. Any documents that Morrow had not printed were lost, along with any metadata about their creation.

Fact-Finding and Spoliation

The fundamental goal of any court engaged in fact-finding is to do so accurately. But this goal is tempered by reality. In the modern world, courts have recognized that there may sometimes be too much data for human beings to examine piece-by-piece.⁸ And business realities may cause data to be lost or destroyed in the ordinary course, long before litigation is even seen as a possibility. Responding pragmatically to these considerations, most courts balance the cost of locating and producing a given class of ESI against its importance to the case. However, spoliation—the destruction of evidence—is another matter.

Under the spoliation doctrine, courts may award sanctions, in various forms, to punish parties who destroy evidence, and to provide relief to parties who might have been harmed by that loss of evidence. Spoliation sanctions are also prophylactic: They are designed to deter parties from destroying evidence in the future.⁹ The Second Circuit has identified three elements that a court must find before issuing sanctions for spoliation: (1) that the party controlling the materials had a duty to preserve them when they were destroyed; (2) that the materials were destroyed “with a culpable state of mind”; and (3) that the destroyed materials were relevant to and would support a party’s claim or defense.¹⁰ Recently, New York state courts have largely adopted the federal standards for spoliation, at least as to ESI.¹¹

Spoliation Analysis in ‘Dorchester’

The court in Dorchester began its spoliation analysis by identifying the three elements of spoliation—duty, culpability, and relevance. It then analyzed each element, and finished by determining the appropriate sanction.¹²

The court first determined that Dorchester was bound by a duty to preserve the documents and metadata at the time that Morrow’s computer crashed, which was during the appeal. Critically, the court held that “Dorchester’s duty to preserve did not end when the computer allegedly crashed,” though the court acknowledged that no other court in the Second Circuit had squarely addressed the issue.¹³ In support of its conclusion, the court asserted that it is common knowledge that experts can sometimes recover data from the hard drive of a computer that “crashes.” Therefore, the court held, “Dorchester had a duty to preserve Morrow’s computer after its alleged crash, and to make reasonable efforts to recover the data it contained.”¹⁴

Moving on to the second element, the court held that Dorchester had a culpable state of mind and had acted with at least gross negligence. Again, the court acknowledged that no court in the Second Circuit had addressed the issue at hand: whether a person is culpable for spoliation if they destroy a “crashed” computer that contains data that might still be recoverable with professional assistance. The court held that in Dorchester’s case, the destruction of the computer was the same “as [the] destruction of all data stored on the machine before the alleged crash.”¹⁵ The court identified three reasons supporting its conclusion.

First, the court found that at least some of the data on the machine likely could have been recovered by a computer specialist with present-day technological expertise. Second, the court recognized that spoliation is an equitable doctrine, and held that where there is a risk that a court will inaccurately determine the impact of the destruction of potential evidence on a party—because it is often uncertain exactly what evidence was destroyed—the court should place that risk of inaccuracy on the party who caused it by assuming that the data would have been recoverable. Third, the court warned that finding otherwise may incentivize future misconduct. Parties might let recoverable data perish, or even intentionally destroy it, if they are allowed to select which materials are preserved and used in a case—for example, by copying or printing select documents—and which are lost, without suffering repercussions.

Finally, for the third spoliation element—the relevance of the spoiled evidence—the court presumed that the lost data would have been relevant to BRJ’s defenses because Dorchester had acted with a culpable state of mind (at least gross negligence). Dorchester failed to rebut that presumption. In fact, Morrow’s own testimony supported it: He admitted that he had only printed documents that were beneficial to Dorchester’s case. Therefore, the final element was established.

Having held that BRJ had established the three spoliation elements, the only remaining question for the court was the appropriate sanction. The court considered the prejudice to BRJ, which it characterized as severe: “the potential scope of the evidence lost, and the pro-spoliator bias of the documents that remain.” The court surveyed the possible remedies: “further discovery, cost-shifting, fines, special jury instructions, preclusion, and the entry of default judgment or dismissal (terminating sanctions).”¹⁶ After ruling that it was bound to impose the least harsh sanction that would still provide an adequate remedy, the court held that the “substantial sanction” of a “mandatory adverse inference” was appropriate. In this case, that adverse inference will likely be fatal to Dorchester’s case, because the jury will be directed to infer that Dorchester “destroyed electronic evidence, including emails and metadata, favorable to BRJ’s claim that it did not participate in the transactions at issue.” In addition, the court ordered Dorchester to pay BRJ’s attorney’s fees and costs.

Conclusion

The decision in Dorchester is not especially surprising. Though the issue is apparently new to the federal courts, the First Department addressed it in 2013, when it upheld spoliation sanctions against a plaintiff who destroyed a purportedly broken computer that contained relevant documents after commencing the action.¹⁷ Regardless, these opinions are important reminders to attorneys that the state of the art is always advancing, and the courts are increasingly in touch with that. For instance, both of these opinions mention the importance of metadata and recognize that printed versions of documents do not preserve it. And in Dorchester, the court described in detail the advanced tools available for data recovery, and assumed that the parties understand that landscape as well. In this environment, the message to litigants is clear: When in litigation, it is far better to preserve potentially relevant information than to face the consequences.

Read more: