I.    Introduction

¶ 1         As of December 1, 2006, Rule 37(f) of the Federal Rules of Civil Procedure ("Rule 37(f)" or "the Rule") will provide protection against rule-based sanctions for a party's inability to provide electronically stored information ("ESI") in discovery when that information has been lost as a result of the routine, good faith operation of an electronic information system.¹ This limitation reflects a concern that the threat of sanctions in those circumstances unfairly impacts primary conduct - the way in which users of electronically stored information manage their storage and retention of information.² While the frequency of the risk of sanctions in those circumstances is debatable - the reported cases are few - the issue to which the reassurance of Rule 37(f) is directed is the distortion of primary conduct resulting from that threat.³

¶ 2         A number of criticisms have been levied at Rule 37(f). For example, two typical criticisms are set forth in the Spring 2006 issue of the Northwestern Journal of Technology and Intellectual Property. The first, by the quasi-official historian of the E-Discovery rules process, Kenneth Withers, argues that Rule 37 falls "far short of that [which] the original proponents had wanted"⁴ and is unworthy of the label "safe harbor."⁵ The second, outlined by Nathan Drew Larsen, analyzes the efficacy of the Rule under the Rules Enabling Act and argues that it raises the specter of possible "collusion" among producers to "configure" their systems to destroy data.⁶

¶ 3         These observations - implying that the Rule is both ineffective and dangerous at the same time - miss the mark and deserve a response. In this article, I evaluate the quality and accuracy of the criticisms. Part II examines the Rule as promulgated by the Supreme Court. Part III traces the Rule's evolution and Part IV examines the common criticisms of the Rule and explains why they are unfounded. Part V concludes by explaining why the Rule satisfies appropriate and achievable objectives in a manner which will be useful to all courts facing similar issues.

II.    Rule 37(f)

¶ 4         Rule 37(f) was drafted by the Civil Rules Advisory Committee ("Advisory Committee") at its meeting in April, 2005⁷ after consideration of comments on an initial proposal made at Public Hearings held by the Committee in San Francisco, Dallas and Washington, D.C.⁸ The Advisory Committee is one of several committees formed by the Judicial Conference of the United States (the "Judicial Conference")⁹ to study potential changes in the rules of practice and procedure in specific fields¹⁰ and is empowered to submit rule proposals to the Judicial Conference Committee on Rules of Practice and Procedure (the "Standing Committee") for review and approval. Rule 37(f) was subsequently endorsed (along with the other proposed e-discovery Rules)¹¹ by the Standing Committee¹² and, after a favorable recommendation by the Judicial Conference, by the Supreme Court in April, 2006¹³ Assuming no complications, Rule 37(f) will come into effect in December, 2006.¹⁴

¶ 5         Requests for spoliation sanctions¹⁵ are often sought where an inability to produce ESI during discovery results from a failure to preserve a source of ESI before active discovery is undertaken. The Rule does not differentiate between losses of ESI which are caused by events that occur before institution of litigation and those that occur afterwards. It also applies only to motions seeking "sanctions," and not to "the kinds of adjustments frequently made in managing discovery if a party is unable to provide responsive information."¹⁶

¶ 6         Rule 37(f) is intended to serve as part of a solution to the practical problems of ESI preservation and production.¹⁷ The Advisory Committee chose not to enact detailed preservation rules but instead established a framework to encourage early voluntary agreement on preservation steps. Rule 26(f) establishes a "new paradigm"¹⁸ of mandatory discussion of preservation and production issues involving ESI, which will hopefully yield an accommodation on preservation steps satisfactory to both parties. If not, the Court is empowered to enter specific orders to guide the parties.¹⁹ A requesting party failing to identify or act on preservation issues will not be allowed to take a "gotcha" or ambush approach by waiting until too late to raise preservation issues.²⁰

¶ 7         Accordingly, the need to invoke Rule 37(f) will hopefully exist only in isolated instances, such as challenges to preservation steps taken or not taken concerning inaccessible sources of information in the absent of an agreement or court order.²¹ When invoked, Rule 37(f) will involve a two-step analysis.

¶ 8         First, a Court must determine if sanctions for the failure to produce lie pursuant to one of the provisions of the Federal Rules.

¶ 9         Second, a Court applying Rule 37(f) must determine whether, under all the circumstances, the failure to produce ESI in discovery resulted from circumstances which fall within the scope of the Rule. This will turn on whether the loss resulted from a "routine" and "good faith" operation of a covered "information system" and has not resulted in "exceptional circumstances."

¶ 10         An "information system" creates and manages electronically stored data, whether in the form of email, raw data, images or other electronic form. A wide variety of such systems are typically used to perform the range of important business and regulatory functions within an entity. As a normal or "routine" part of the operation of such systems, data is overwritten, deleted or modified by automatic or manually initiated processes, pursuant to the design or programming of the operating software.²² The use of the phrase "routine operation" in reference to such an information system emerged during discussions at the April 2004 Advisory Committee meeting²³ to describe the normal aspects of systems that might result in data losses and which should be presumptively subject to the safe harbor under discussion.²⁴ Serving in that role, the phrase was part of the initial Rule 37(f) proposal published for Public Comment²⁵ and ultimately survived the subsequent revisions as a basic concept in Rule 37(f).²⁶

¶ 11         The Committee Note to Rule 37(f) is neutral as to information technology, so it does not list specific examples of systems the Advisory Committee had in mind. However, the Standing Committee Report to the Judicial Conference mentions "programs that recycle storage media kept for brief periods against the possibility of a disaster that broadly affects computer operations" and "programs that automatically discard information that has not been accessed within a defined period" as examples of "routine" operations in current technology.²⁷

¶ 12         The fact that human involvement is part of the process does not make the operation causing the failure other than "routine," as is shown by the reference to the recycling of backup media, which is instituted by members of an IT function pursuant to a retention schedule.²⁸ The "routine" aspect of the operation is the physical recycling process. Similarly, the pressing of the "delete" key by users initiates a "routine" feature of the email system, incidental to its operation."²⁹

¶ 13         In other words, the issue under Rule 37(f) is whether the feature causing destruction is a normal part of that process, not whether some other approach could have been used. A feature added to an existing system solely for the purpose of deleting information needed in litigation would receive short shrift.³⁰

¶ 14         The key differentiating factor in applying Rule 37(f), however, is the presence or absence of "good faith" in the routine operation of the information system. It is only those losses resulting from "good faith" operations which are exempted from sanctions. It is clear that mere negligence or inadvertence in operating an otherwise reasonable information system does not bar a finding that the entity was acting in good faith. As a Member of the Advisory Committee stated during the April 2005 meeting of the Advisory Committee, "[G]ood faith lies at a point intermediate between negligence and recklessness. It assumes the party has a reasonable litigation hold, and did not deliberately use the system's routine destruction functions. 'If you know it will disappear and do nothing, that is not good faith...[t]he line is conscious awareness the system will destroy information.'"³¹ Rule 37(f) thus represents the state of mind which must exist to be exempt from sanctions.³²

¶ 15         Determining the application of Rule 37(f) requires an assessment as to whether the features causing the loss were operated without a disqualifying culpability.³³ In making the assessment, the issue should be whether the process involved was "reasonably calculated" to achieve its goals, not whether perfect results were achieved.³⁴

¶ 16         Rule 37(f) will not apply in those rare circumstances where the conduct of the party seeking relief from sanctions has resulted in such a degree of prejudice that it is unfair to restrict the power of the court. The Standing Committee Report to the Judicial Conference explained that the clause was added to provide "flexibility" in the event of "serious prejudice" resulting from the loss.³⁵

III.    EVOLUTION OF RULE 37(f)

¶ 17         Before turning to the specific criticisms of the Rule, it is helpful to understand its evolution during the five years preceding adoption in April, 2005.

¶ 18         The roots of Rule 37(f) can be fairly traced to the mini-conferences held by the Discovery Subcommittee of the Advisory Committee at its request³⁶ at Hastings³⁷ and Brooklyn³⁸ Law schools during 2000. Both mini-conferences featured panels of judges, lawyers and technical consultants with extensive experience in e-discovery and both identified issues about meeting preservation expectations as major concerns.

¶ 19         In December 2000, after reflecting on discussions at the Brooklyn Conference and my experiences as one responsible for litigation compliance,³⁹ I wrote to (then) Magistrate Judge John Carroll, Chair of the Discovery Subcommittee⁴⁰ to suggest adoption of a targeted safe harbor rule exempting the continued good faith operation of disaster recovery and other systems from rule-based sanctions unless a prior order was issued requiring their preservation.⁴¹ I also suggested that only willful violations of preservation orders should justify imposition of sanctions. As far as I can tell, this was the first explicit suggestion for a preservation safe harbor, which I amplified in other articles.⁴²

¶ 20         The ensuing period saw case law evolve on a number of basic e-discovery issues⁴³ in District Court opinions⁴⁴ and enactment of some local rules on specific aspects of e-discovery.⁴⁵ It was also during this period that the Sedona Conference®⁴⁶ was organized and bar associations, legal publications and litigation support vendors held scores of conferences on e-discovery.⁴⁷ A key milestone was the publication of the "prescient"⁴⁸ Law Journal article on e-discovery by Northwestern Law Professor Martin C. Redish.⁴⁹ Professor Redish famously noted that society and the rule-making process must be prepared to limit the search for truth to take into account other elements of the litigation matrix, such as the need to provide predictable standards of primary behavior.⁵⁰

¶ 21         In April 2003, after soliciting and receiving further input on the topic,⁵¹ the Discovery Subcommittee informed the Advisory Committee that "after more than three years of considering these issues, the Subcommittee believes that the time for more concrete action has arrived."⁵² The Subcommittee delivered a comprehensive set of proposals for consideration by the Advisory Committee in September 2003⁵³ and after preliminary review at its October 2003 meeting,⁵⁴ the Advisory Committee scheduled a two-day Conference on E-Discovery at Fordham University Law School (the "Fordham Conference") for February 2004 to discuss them.

¶ 22         The Fordham Conference was attended by a cross-section of the bench and bar with members of the Standing Committee of the Judicial Conference. Members of the technical and consulting community were also present. Participants were furnished with copies of the Subcommittee proposals in advance of the meeting.

¶ 23         In regard to preservation and spoliation issues, the Subcommittee recommended altering Rule 26 or adding a new Rule 34.1 to emphasize that parties need not suspend the "operation in good faith" of "disaster recovery or other [computer] systems" provided that one day's backup was retained. The proposal also stated that information in inaccessible form did not have to be preserved unless a court ordered the party to do so.⁵⁵ As a second and "complementary" step to these changes, the Subcommittee also recommended that a new Rule 37(f) be added to make it clear that no sanctions would be issued for a failure to produce ESI unless the party "willfully or recklessly" deleted or otherwise made ESI unavailable after having been served with a request that described it with reasonable certainty.⁵⁶

¶ 24         A separate panel discussion at the Fordham Conference was devoted to safe harbor issues, with the author participating.⁵⁷ Not surprisingly, the panelists were not in complete agreement either on the need for a safe harbor rule or on the specific components required should one be adopted.⁵⁸

¶ 25         The Advisory Committee met in April 2004⁵⁹ to decide whether to initiate the rulemaking process. The Discovery Subcommittee had met prior to the meeting and a revised set of e-discovery recommendations was before the Advisory Committee.⁶⁰ By the conclusion of the meeting, it had agreed to proceed and had settled on a proposed package of e-discovery amendments, including a proposed Rule 37(f) and related Committee Note.⁶¹ These materials were ultimately published by the Standing Committee (along with proposed amendments to the Bankruptcy, Criminal and Evidence rules) in August 2004 with public comments due by mid-February 2005.⁶²

¶ 26         Proposed Rule 37(f) provided, in relevant part, that ". . . a court may not impose sanctions under these rules on the [producing] party for failing to provide [ESI] if: (1) the party took reasonable steps to preserve the information after it knew or should have known the information was discoverable in the action; and (2) the failure resulted from loss of the information because of the routine operation of the party's electronic information system."

¶ 27         Some Members at the Advisory Committee meeting objected to attaching disqualifying consequences to the lower range of the "fault" spectrum (negligence)⁶³ and insisted that only willful or reckless conduct should deny protection under Rule 37(f). Accordingly, the Advisory Committee prepared and the Standing Committee approved inclusion of a footnote requesting comments "on whether the culpability or fault that takes a party outside any safe harbor should be something higher than negligence⁶⁴ and included an example "to focus comments and suggestions."⁶⁵

¶ 28         The Advisory Committee did not promulgate a separate rule spelling out preservation obligations. Instead, the proposed Committee Note described a three part analysis for use in evaluating the reasonable steps needed to preserve electronically stored information.⁶⁶ First, the outer limit of information to be preserved would be determined by the existing limits on the scope of discovery established by Rule 26(b)(1).⁶⁷ Second, given that amended Rule 26(b)(2) conditioned the discovery of inaccessible ESI on issuance of a court order, the Note observed that "[i]n most instances, a party acts reasonably by identifying and preserving reasonably accessible electronically stored information that is discoverable without court order."⁶⁸ The third factor was what the party then knew about the nature of the litigation, which "should inform its judgment about what subjects are pertinent to the action and which people and systems are likely to have relevant information."⁶⁹

¶ 29         The response to Rule 37(f) was mixed at the public hearings held in San Francisco, Dallas and Washington in January and February 2005. Representatives of public and private entities generally favored some form of safe harbor, noting that the fear of sanctions for inadvertent loss of ESI had created an unfair chilling effect⁷⁰ and had distorted business processes by encouraging over-retention of information.

¶ 30         However, some supporters of the safe harbor concept also expressed concern about the fact that, as written, the protection afforded by Rule 37(f) apparently could be lost by merely negligent or inadvertent conduct.⁷¹ A typical comment was made by a representative of The City of New York Law Department, which argued that it was unfair to subject a party to sanctions for the conduct of a low level employee who may negligently delete electronic information despite reasonable preservation efforts by City attorneys and management personnel.⁷²

¶ 31         Some opponents argued that there was no empirical proof of a need to act based on the paucity of reported decisions citing entities for anything other than willful misconduct.⁷³ Others argued that proposed Rules 26(b)(2) and 37(f) would encourage corporate entities to shift information from being reasonably accessible to becoming increasingly inaccessible, with the additional burden of access being imposed on the requesting party.⁷⁴ Some commentators also criticized the Committee Note for failing to adequately state the need to preserve inaccessible information for later review to determine if it should be produced.⁷⁵

¶ 32         In the end, after considering the matter from all angles, the Advisory Committee decided to simplify Rule 37(f) during their meeting of April 2005. The Committee dropped the reference to preservation obligations and court orders and adopted a concise rule limiting sanctions for losses resulting from the "routine, good faith" operation of an information system unless "exceptional circumstances" exist.⁷⁶ The intention expressed at the meeting was to use "good faith" as a culpability excluder which would operate midway between denial of exemption for merely negligent conduct and one which would require proof of willful or reckless conduct.⁷⁷

¶ 33         The Committee Note to Rule 37(f) was also shortened, thus deleting the three-part analysis of preservation obligations. New language was added to clarify that "[a] party's identification of sources of electronically stored information as not reasonably accessible does not relieve the party of its common-law or statutory duties to preserve evidence."⁷⁸ Finally, the affirmative role of good faith in determining how a party should react to those preservation obligations was added.⁷⁹

¶ 34         After some further minor modifications,⁸⁰ the amended Rule 37(f) was submitted to and approved by the Standing Committee at its June 15-16, 2005 Meeting⁸¹ in its final form.⁸²

IV.    CRITICISMS

¶ 35         Some have questioned whether Rule 37(f) was worth the effort. For example, in his survey piece in the Spring Issue of this Journal, Ken Withers argues that the final Rule, with its qualifications, is ineffective and unnecessary."⁸³ In a variation on that theme, another recent article criticizes Rule 37(f) for not providing sufficient "bright line" preservation guidelines, especially in regard to inaccessible information.⁸⁴ The authors of that article contend that proposed Rule 37(f) will have "little, if any, practical impact" because courts "will continue to have significant discretion to impose sanctions for a 'wrong guess' as to the precise moment when a litigation hold should be implemented."⁸⁵ Others argue that Rule 37(f) is too limited in its application.⁸⁶

¶ 36         These criticisms prompt several responses. First, it is certainly true that Rule 37(f) differs from the initial proposals which contemplated intervention to require suspension of routine operations of information systems only when and if a court ordered them for good cause.⁸⁷ In this regard, the original proposals were more akin to the safe harbor in the Private Securities Litigation Reform Act of 1995 (the "Reform Act") which provides (as one alternative) an exemption from civil liability for forward-looking statements accompanied by appropriate cautionary language.⁸⁸ This approach was ultimately rejected because of a concern it would lead to excessive resort to preservation orders.⁸⁹ Accordingly, the more flexible but less predictable approach based on retrospective "good faith" standard for preservation efforts was substituted by Advisory Committee.

¶ 37         It can be argued that the Committee approach is far more practical than the original suggestion. Rule 37(f) is not tied to any particular technology or any particular set of preservation practices.⁹⁰ New fact patterns will continue to evolve and Rule 37(f) will be available for use in circumstances where it is clear that reasonable efforts were made but losses to discovery nonetheless occurred. Courts will now have the flexibility to identify losses due to "routine, good faith" operations with the help of the Sedona Principles, especially Principles 5,⁹¹ 9,⁹² 11⁹³ and 12.⁹⁴ Additional help is on the way from that quarter.⁹⁵ Moreover, because the outcome of any particular challenge to preservation practices cannot be predicted, a premium will be placed on the development and implementation of good faith policies and practices..

¶ 38         Seen from this perspective, therefore, Rule 37(f) is analogous to the alternative version of the safe harbor under the Private Securities Act (and the equivalent SEC Rule) which provides exemption from liability for forward looking statements that are made in "good faith" on the basis of reasonable assumptions.⁹⁶

¶ 39         To some, the approach in Rule 37(f) suggests a return to the "pre-Zubulake" era, in which a number of courts recognized a broad duty to preserve back-up tapes without considering any limitations on g preservation obligations.⁹⁷ This is clearly a misreading of the impact of the amendments in general and Rule 37(f) in particular.

¶ 40         For example, when a producing party is faced with making plans for pre-discovery preservation, the focus on accessibility of information set forth in Rule 26(b)(2) is far from neutral in its impact. The clear intention of the Advisory Committee in adopting Rule 37(f) was that inaccessible sources of information need not be preserved when the party has a reasonable belief in the inaccessibility of the source and in the availability of the relevant information on active systems and when the party is acting in good faith.⁹⁸ If decisions are challenged by a motion for sanctions, the test of under Rule 37(f) should be whether the preservation process adopted and followed was "reasonably calculated" to achieve an appropriate result.⁹⁹

¶ 41         Moreover, once discovery commences, the mandatory requirement in Rule 26(f) of early discussion and resolution of identified preservation issues - a first for the Federal Rules - clearly requires affirmative action by the requesting party with knowledge of preservation steps undertaken. The producing party has the responsibility under Rule 26(b)(2)(B) to identify the sources of potentially responsive information which it does not intend to search or use for production. An unsatisfied requesting party seeking information must seek an appropriate order¹⁰⁰ or waive the right to seek sanctions later for non-production.¹⁰¹

¶ 42         Finally, the failure of individual users to perfectly comply with vague litigation hold instructions is entitled to protection under Rule 37(f). Delivery of a litigation hold notice under reasonable conditions should operate as a rebuttable presumption that appropriate steps were undertaken and that any loss despite that notice fits within the safe harbor.¹⁰²

¶ 43         A major criticism of Rule 37(f)¹⁰³ is that it addresses only sanctions under the federal rules, which generally do not apply prior to commencement of litigation.¹⁰⁴ Indeed, one Commentator refers to Rule 37(f) as providing only a form of "false harbor" because courts are likely to ignore Rule 37(f) and apply their inherent power to sanction.¹⁰⁵

¶ 44         Courts often cite their inherent powers to sanction when it is unclear that rule-based sanction authority exists.¹⁰⁶ Such authority is generally found in Rule 37(b)(2), which applies when a party fails to obey a discovery order or in Rule 37(c), which applies to a failure to make initial disclosures under Rule 26(a)(1) or to supplement or amend previous discovery.

¶ 45         However, it is unlikely that courts will deliberately invoke their inherent power in order to reach a different result on the same facts as would exist under Rule 37(f). Most courts understand that the limitation to the "rule-based" qualification was not based on anything other than the fact that the Advisory Committee and Standing Committee were simply respectful of the limits of rulemaking power.¹⁰⁷ Accordingly, the natural tendency of most courts will be to use Rule 37(f) as a "guidepost" or reference point in exercising their inherent powers. The power to sanction pre-litigation conduct is subject to the Supreme Court's admonition in Chambers v. Nasco that courts should act with great care and only when the Rules "are not up to the task."¹⁰⁸

¶ 46         A respect for carefully drafted Rule limitations in the face of more expansive inherent power is not unknown. For example, in the case of Brandt v. Vulcan, Inc., the Seventh Circuit refused to suggest that a court should exercise its inherent powers to sanction discovery misconduct where the District Court had concluded that it lacked power to do so because of limitations under Rule 37(b)(2).¹⁰⁹ And in Convolve Inc. v. Compaq Computer Corp., Magistrate Judge Francis relied upon the intitial version of proposed Rule 37(f) in absolving a party of the failure to undertake to preserve certain "ephemeral" information in the absence of a discovery order.¹¹⁰

¶ 47         A number of recent articles castigate Rule 37(f) as encouraging improper or illegal corporate conduct. A recent Student Note, for example, asserts that Rule 37(f) will "eliminate the possibility of judicial sanctions when parties evade discovery requests" because parties can "destroy potentially incriminating documents using a document retention policy."¹¹¹ Further, since "the rules are biased in favor of big companies," they will "make their electronic document storage system as inaccessible as possible."¹¹² Other articles predict that Rule 37(f) will cause "corporations" to "adopt comprehensive data deletion" based on "purported" good-faith desire to reduce data storage costs¹¹³ because "[the] subversion of document retention policies is shared by many in corporations across America."¹¹⁴

¶ 48         This type of ad hominem attack on private corporations is quite unfair. First, it ignores the fact that public entities have been equally strong in their support for a preservation safe harbor.¹¹⁵ Second, it also ignores the ample civil and criminal sanctions applicable to those who might decide to act in to avoid their obligations in discovery. As the Standing Committee Report noted in response to an analogous argument about Rule 26(b)(2)(B), "a party that makes information "inaccessible" because it is likely to be discoverable in litigation is subject to sanctions now and would still be subject to sanctions under the proposed rule changes."¹¹⁶

¶ 49         Generally speaking, an entity has the right to manage and dispose of its information. See The Sedona Guidelines: Best Practice Guidelines & Commentary for Managing Information & Records in the Electronic Age (Sedona Conference Working Group Series, September 2005), Guideline 3: "An organization need not retain all electronic information ever generated or received." Indeed, the Supreme Court endorsed this principle in Arthur Andersen LLP v. United States.¹¹⁷

¶ 50         However, corporate policies involving destruction of information sought in discovery must be both reasonable in purpose and reasonable as applied in a particular case.¹¹⁸ It is not, for example, reasonable to institute a policy to intentionally destroy information for the primary purpose of preventing a known future adversary from obtaining the information in litigation. See, e.g., Rambus, Inc. v. Infineon Technologies, 220 F.R.D. 264, 286 (E.D. Va. 2004) ("[E]ven valid purging programs need to be put on hold when litigation is reasonably foreseeable").¹¹⁹

¶ 51         Moreover, criminal sanctions may be implicated by deliberate attempts to withhold information in anticipation of certain types of proceedings. For example, in the wake of the Enron scandal, Congress stiffened existing criminal penalties and added new violations for one who knowingly alters or destroys documents with the intent to impede a federal investigation or proceeding or "in relation to or contemplation of such matter or case."¹²⁰

¶ 52         Nathan Drew Larsen argues that by focusing on losses which result from "routine" operations the Advisory Committee has inadvertently created an incentive in Rule 37(f) for producing parties to fail to invest in technology which is less destructive of data.¹²¹

¶ 53         This argument is a variation of the contention that Rule 37(f) will encourage producing parties to place information into inaccessible sources to avoid preservation and production obligations and is equally fallacious. It ignores the fact that the moving force in information technology is not litigation but business concerns.¹²² As the Standing Committee Report notes: "[M]any witnesses and comments [at the Public Hearings] rejected the argument that the rule would encourage entities or individuals to "bury" information that is necessary or useful for business purposes or that regulations or statutes require them to retain."¹²³ Moreover, the argument overstates the intended purpose of that phrase. A "routine" operation is merely part of process.¹²⁴ Its use in Rule 37(f) is not intended to freeze technological innovation but merely to indicate that the feature causing the loss is typically part of the system at that time.¹²⁵

¶ 54         The "good faith" standard of Rule 37(f) has also been criticized as too vague because parties will "continue to be subject to varying standards for determining when sanctions should be imposed."¹²⁶

¶ 55         There is, of course, some truth to that observation. However, Courts are often called upon to apply standards involving an assessment of "good faith" to determine exemptions from liability¹²⁷ or to assess rights and obligations.¹²⁸

¶ 56         In applying Rule 37(f) on a case-specific basis,¹²⁹ Courts will be employing the classic "excluder" function attributed to Professor Robert Summers.¹³⁰ Under this astute observation, "good faith" has no inherent meaning, but serves to exclude examples of bad faith conduct, which are identified on an ad hoc basis over time. For example, as the Committee Note makes clear, willful continuance of an operation involving destruction when preservation obligations are known are to be excluded from being treated as a "routine, good faith" operation of that system.¹³¹ In Mosaid Technologies v. Samsung Electronics, for example, the producing party made no attempt to justify its failure to modify the automatic deletion of e-mail for several years, thereby making it impossible to produce any of the technical email needed in the patent litigation. The destruction in that case was neither routine nor in good faith.¹³²

¶ 57         Assessing losses against a "good faith" standard presents a subjective issue, which may vary from case to case, since the personnel involved "may well have been guided by different mental states."¹³³ For example, a typical email management policy¹³⁴ requires that users identify and save email and attachments that meet business, regulatory or other criteria and delete the rest. Much as paper records are thrown away by employees making judgment calls about the necessity of retaining information and automatic deletion is often used to enforce the policy. This avoids the need for excessive storage and unnecessary review during discovery and also reduces the risk of inadvertent production of privileged information.

¶ 58         Thus, if information is lost by user or automatic action, the "good faith" criterion will help distinguish among those situations where sanctions should be applied and those where they should not.¹³⁵ There will be examples of lax or negligent oversight which may still qualify as "good faith."¹³⁶

¶ 59         The practice of recycling backup media presents a similar issue. These sources of ESI are not generally used for ongoing business purposes and can be difficult and costly to access.¹³⁷ Under some circumstances, a duty may exist to suspend their recycling and preserve them for potential discovery.¹³⁸ However, due to the complexity of the issue and without clear guidelines on when that duty is triggered, it is often difficult to consistently implement the duty to act.¹³⁹

¶ 60         One compelling indicia of "good faith" will be if there is evidence that the loss occurred during an operation pursuant to a reasonable policy in the absence of improper motive. In this regard, Rule 37(f) reflects the existing law in a majority of Circuits, which hold that a court should not issue an adverse inference instruction unless there is an affirmative showing of intentional destruction amounting to "bad faith."¹⁴⁰ One recent article, otherwise critical of Rule 37(f), acknowledges that the Rule will authorize the "general use of electronic document retention systems."¹⁴¹ To paraphrase Justice Stewart in another context, the advantage to Rule 37(f) is that courts will know "good faith" when they see it.¹⁴²

V.    CONCLUSION

¶ 61         Rule 37(f) represents a carefully considered policy judgment reconciling two sometimes competing policy considerations: the need for preservation of potentially responsive ESI for use in litigation and the need to minimize the unnecessary intrusion into productive use of the systems involved. Its impact is carefully targeted and depends upon court endorsement, which cannot be gamed in advance. Spoliation of potential evidence "occurs along a continuum of fault- ranging from innocence through the degrees of negligence to intentionality."¹⁴³ Rule 37(f) reflects a judgment that as to ESI lost through the routine operation of information system, only when fault lies along the higher ends of that spectrum are sanctions to be used.

Rule 37(f) should help to encourage public and private entities alike to act in good faith and undertake reasonable steps to accommodate preservation needs. Both requesting and producing parties have an interest in its effective use, which complements the other advances in the e-discovery amendments. It hopefully provides a template for those courts which face similar requests for sanctions under their inherent powers. As such, it more than meets the original intent of the proponents and should endure in its appropriate role for many years to come.

ENDNOTES