Recently, some US courts have suggested that a litigation hold must be in place as soon as an entity marks any materials as work product, given the similar phrasing used to describe the preservation and work product protection triggers. “Reasonable anticipation of litigation” is the commonly accepted trigger for the duty to preserve discoverable evidence, and since Hickman v. Taylor1 the work product doctrine protects materials prepared “in anticipation of litigation” from discovery, but two very different conditions are described by these similar-sounding phrases. Conflating the two standards may be an easy, mechanical shortcut for a court faced with a spoliation motion, but doing so discards the appropriate balance between preserving relevant evidence and the costs thereof, and introduces more confusion into an already muddy area. Holdings from courts that an entity is subject to sanctions if it does not have a litigation hold in place when it first marks any documents as work product threaten to encourage, once again, over-preservation and thereby undermine some of the gains realized by the 2015 amendments to Federal Rules of Civil Procedure 26 and 37. Despite similar use of the phrase “reasonable anticipation” of litigation, an entity should not be required to issue a litigation hold as soon as it first marks “work product” on a document. 

Federal law on preservation affects many individual and corporate citizens every day, yet it is constantly in a state of flux and has evolved in fits and starts as a product of individual court decisions. The 2015 amendment to Rule 37(e) has done a good job of making the law on spoliation more consistent in US federal courts, but the Rules Committee dodged defining the trigger for preservation; it is still determined by the common law.2 Moreover, opinions conflating the triggers for work product protection and preservation obligations introduce confusion into an area that needs more clarity, not less. Litigants, particularly those not equipped with large in-house legal departments and counseled by sophisticated attorneys, risk being shut out of our federal courts by the increasing complexity of e-discovery requirements. A requirement that a preservation notice be issued in every case where a company lawyer marks a memo “work product” is exactly the kind of trap for the unwary that the 2015 amendments were designed to eliminate.

This article first explores some of the decisions that mix the preservation and work product protection triggers. We then discuss why courts should reject the false equivalence between these two triggers that some have embraced.  

Some federal courts have conflated the triggers for work product protection and document preservation

As noted, under US federal law, the duty to issue a litigation hold and the availability of the work product protection both nominally toggle on when a party “anticipates” litigation, and this has produced an easy equivalence between the two triggers in the eyes of some courts. Some cases have linked the triggers for work product protection and the duty to preserve, noting in dicta that the two are essentially the same.

In Crown Castle USA, Inc. v. Fred A. Nudd Corporation, the court determined the duty to preserve was triggered when plaintiff first marked communications with a work product legend.3 Defendant sought sanctions for spoliation of electronically stored information (ESI). Internal emails marked as work product suggested that plaintiff explored the possibility of bringing a claim against defendant’s insurance carrier eight months before filing suit. Without considering the inefficiencies associated with overpreservation, the court held that plaintiff’s obligation to preserve evidence arose eight months before suit was commenced, when several of plaintiff’s employees first considered filing a notice of claim and instituted a practice of labeling defendant-related communications as work product.

The US District Court for the Eastern District of New York reached a similar conclusion in an employment case, Siani v. State Univ. of N.Y. at Farmingdale, holding that if litigation was foreseeable for work product purposes, it was reasonably foreseeable for preservation purposes.4 Plaintiff alleged that defendant  intentionally deleted emails as shown by gaps in defendant’s production of electronic messages and that plaintiff accordingly was entitled to an adverse inference finding based on spoliation. Defendant argued that the duty to preserve only arose when it received notice that plaintiff had filed an Equal Employment Opportunity Commission claim on July 16, 2008. Plaintiff maintained that a duty to preserve actually arose five months earlier, when defendant hired a law firm “to obtain legal advice in connection with issues” related to plaintiff and first marked materials as work product. The court explained that if litigation was reasonably foreseeable for work product purposes, it was reasonably foreseeable for all purposes—including affirmative preservation obligations. Subsequent courts have relied upon the decision in Siani to explain that the duty to preserve evidence arises no later than the first assertion of the attorney work product privilege.5 

A similar analysis was applied by the Eastern District of Virginia in one of e-discovery’s more notorious cases: Samsung Elec. Co., Ltd. v. Rambus, Inc.6 Plaintiff alleged that four patents held by defendant were unenforceable. Defendant developed and licensed patents to companies that manufactured semiconductor memory devices, and had instituted a “Licensing and Litigation Strategy” aimed at several manufacturers, including plaintiff. Evidence showed that defendant’s licensing and litigation strategy included a pervasive document destruction program designed to dispose of discoverable documents and ESI. 

The issue before the court was whether defendant’s document destruction program constituted spoliation and warranted the imposition of sanctions. To determine if spoliation occurred, the court relied upon the work product protection test in evaluating when defendant anticipated litigation. The court held that defendant “clearly and convincingly” engaged in the spoliation of evidence while it anticipated litigation and when actually engaged in litigation. The court’s reliance on the work product protection test as a “helpful guide when assessing intentional spoliation” demonstrates how some courts have equated the trigger for the duty to preserve with the trigger for the work product protection. 

The 2015 amendments to the Federal Rules of Civil Procedure

The 2015 amendments to Rules 26 and 37 were adopted to emphasize proportionality in discovery and to reduce the incentives for companies to over-preserve ESI and other materials. When the Rules Committee took up the issue of spoliation and sanctions in 2011, it became convinced that American companies were preserving too much out of fear of harsh sanctions for merely negligent loss of ESI.

Accordingly, Rule 26 was revised to narrow the scope of discovery. The troublesome phrase that led many courts to abandon all limits on discovery—“reasonably calculated to lead to the discovery of admissible evidence”—was eliminated, and relevance was redefined to include an express requirement that discoverable material be “proportional to the needs of the case.” The complete rewriting of Rule 37(e) eliminated “death penalty” sanctions for mere negligent loss of data and sought to reward litigants who take “reasonable steps” to meet their preservation obligations.

Resist the false equivalence between work product protection and the duty to preserve

A rule that automatically triggers the preservation duty when an internal memo is first marked with a “work product” notation cannot be squared with the philosophy underlying the 2015 amendments to Rules 26 and 37. An entity may mark materials as work product in anticipation of litigation long before it reasonably anticipates litigation sufficient to justify the imposition of onerous preservation obligations. For example, in response to an employee comment about working conditions—that does not mention or suggest the prospect of litigation—an employer might initiate an internal investigation of the workplace. And after investigation, the legal department might draft a memo of its findings and mark it as work product. Although there is always a potential for litigation to arise, the company does not actually anticipate imminent litigation, but it has marked materials as work product. Under the cases described above, these actions alone trigger the duty to preserve. But this result is misguided. Without more reason to believe that litigation is imminent, such actions should not be construed as triggering affirmative preservation obligations because litigation is not “reasonably anticipated.” It is simply a fact that corporate legal departments and their outside counsel are often called upon to investigate conditions “in anticipation of litigation” at a point in time when there is no “reasonable anticipation” of actual litigation. Should we really want to force companies to institute litigation holds under those circumstances?

The point is further illustrated by imagining an in-house legal department being asked to provide guidance on a changing regulatory environment, perhaps in anticipation of a new law set to go into effect. The lawyers draft a memo analyzing the predicted impact of such change. It includes a discussion of potential liability that might be imposed under the new statutory regime and is marked as work product and attorney-client privileged. It would be absurd to require that the entity affirmatively preserve all related materials simply because the memo was marked as work product. Indeed, to require preservation at that time would lead to rampant overpreservation and magnify the likelihood of overlapping litigation holds—undermining the purposes of the 2015 amendments discussed above. Yet under the rule espoused in Crown Castle and Siani, such actions would be sufficient to trigger the duty to preserve. While conflating the two triggers appeals to some given the similar language used, that approach is misguided and results in inefficiency. 

In addition, requiring preservation at such a premature stage would be disproportionate to the civil justice system. The advisory committee’s comments to the Rule 37(e) amendment provide that the “rule recognizes that ‘reasonable steps’ to preserve suffice; it does not call for perfection.” As reflected in the 2015 amendments, we do not expect or require perfection from civil litigants—some data will always be lost—but the solution to negligence is additional discovery, not imposition of punitive spoliation sanctions. Because requiring preservation at such an early stage leaves entities vulnerable to charges of spoliation—which detract from the merits of the case and sully careers—imposition of such a requirement prompts an unattainable quest for perfection that is disproportionate to the system itself. Of course, trials remain available to resolve civil disputes even where the written records are imperfect, as they always have been. Thus, requiring affirmative preservation at the first moment any document is marked as work product simply does not align with the preponderance of the evidence standard of our civil litigation system.


Of course, in some instances it will be true that as soon as a company begins producing “work product” memoranda it will “reasonably anticipate” litigation, and in those cases it would be proper to require, under prevailing law, that it take steps to preserve evidence. But our purpose in writing this article is twofold: to urge courts not to resort to the easy equivalence between the two “anticipation” triggers and to urge litigants to argue strongly against the assertion that the two triggers are identical for all purposes.