Jan. 2, 2007
What Every Lawyer Should Know About the New E-Discovery Rules
Recent amendments to federal e-discovery rules require early discussions and even earlier preparation to manage electronically stored data. Look for court tests this year on key terms that could shift the power balance between litigants. By Geoff Howard
Safe harbor. Two tiers. Inaccessible data. Quick-peek and claw-back. Metadata. Form of production.
By now you should be aware that these are some of the terms addressed in recent amendments to the Federal Rules of Civil Procedure covering access to electronically stored information (ESI). But do you know if the new rules will alter the strategic balance in litigation? Or whether the new rules work for or against you?
Answering the first question is easy: The new rules will shift the power balance during discovery toward savvy lawyers and their clients on either side. Answering the second question is more difficult: How parties respond, and how courts interpret various terms and concepts in the amended sections, will determine whether the changes help or hinder your case.
Some parties?those with lots of ESI?may be able to eliminate much of the cost and burden associated with ESI discovery if they act now. But if they wait, they may have to pay unnecessary costs, or cede an information advantage to their opponents. The new rules offer only opportunity; they do not determine which party will turn that opportunity to its strategic advantage.
First, a disclaimer and a clarification: My practice focuses on representing large companies. Notice I do not call them "defendants." In fact, I have litigated ESI issues for plaintiff corporations many times. Beware also that the ESI plaintiff may be a defendant in a case. And even though my client's name has appeared south of the v. on pleadings, I have assumed the role of plaintiff many times because my client had properly preserved, collected, and produced its ESI. We suspected that the plaintiff corporation?which became the ESI defendant?had not done the same. So except for true David-and-Goliath situations (principally employment, some unfair competition, and securities litigation), the case caption does not always reveal who has the upper hand in an ESI issue.
History in the Making
The new rules address unique problems associated with ESI, including runaway discovery costs, case-management difficulties, and uncertainty about how and when to preserve and produce ESI. Depending on your perspective, these rules took shape almost instantly, or over an agonizingly long period of time. The first commercial email application for a prototype of the Internet appeared around 1988. By 1993, America Online had connected its proprietary email system to the Internet, launching global email. In 1997 the newly formed Discovery Subcommittee of the U.S. Judicial Conference's Civil Rules Advisory Committee met to discuss possible rule changes, including proposals to deal with problems posed by ESI.
That first meeting set off a comprehensive examination of e-discovery by the subcommittee. The outward process included conferences around the country conducted from 2000 to 2002 that included plaintiffs lawyers, the corporate defense bar, in-house counsel, vendors, engineers, professors, and judges. The advisory committee and its subcommittee continued to discuss and debate proposed amendments, issuing its first draft of changes in August 2004 and a revised set a year later. The final amendments took effect December 1, 2006.
In the years consumed by this rule-making process, the scope of e-discovery expanded exponentially, as did complaints about its cost and uncertainty. Yet the issues remain constant: What should be discoverable? When should it be discoverable? Which ESI is permissible to destroy? And the question that pervades the entire debate: Who should pay for e-discovery?
Under almost every revised rule, determining which side gains an advantage in ESI discovery hinges on what the responding party does before?and immediately after?it reasonably had notice of a claim or investigation. In short, responding parties have more control of their destinies than requesting parties. By their actions, smart responding parties also can reduce their discovery costs and their exposure to sanctions. A review of the relevant new rules will show how they can accomplish that?and also what happens if responding parties choose to ignore their opportunities.
Sanctions for Failing to Make Disclosures or Cooperate in Discovery
I'll start with Rule 37, which ostensibly deals with sanctions, because it shapes the strategic approach. Rule 37(f) precludes a court from imposing, "absent exceptional circumstances," sanctions for failing to provide ESI lost as "the result of the routine, good faith operation of an information system." The impact of the new rule lies in what it permits a company to do before litigation is reasonably foreseeable. It allows the responding party to destroy ESI?to shred, delete, and erase ESI. And it may allow the responding party to continue these activities during litigation if it has done them before a triggering event put the party on notice that litigation was "reasonably foreseeable." The trick is to prepare carefully before the triggering event to enable continuation of the activities after the triggering event.
A number of terms and phrases in the rule remain murky and await clarification through court interpretation. First, the ESI must be destroyed as part of an "information system." But does that mean any system that generates and stores ESI? Must the system have a specific retention policy integrated into it? The Committee Notes give incomplete guidance.
Second, Rule 37 applies only to information systems that are operated in a "routine" way. A responding party's prelitigation activities will create the benchmark for what qualifies as routine for that party. But does routine mean that maintenance of the information system must follow a predictable, repetitive schedule? Can a company create its own benchmark? Can the schedule change to accommodate external events, such as a litigation hold? Until the courts provide guidance, a distinct advantage will go to parties that design into their ESI retention systems both enforced predictability and reliable litigation holds (which prevent the automatic purging of related material). These parties can delete more data and potentially claim some protection under the rule for their actions.
Third, and perhaps most confusing, ESI destruction must occur in "good faith." The Committee Notes make clear that Rule 37 does not absolve a party of its common law duty to preserve relevant information. They specify that a party may not "exploit" the rule by routinely deleting information subject to its duty to preserve. But the precise standard that applies to the operation of information systems before and after a triggering event remains elusive?along with the question of which party bears the burden of proof in meeting that standard.
Despite its murkiness, revised Rule 37 offers an advantage to a party whose information system routinely manages ESI for legitimate business purposes. If the system includes a litigation-hold process, the party will have a powerful answer to spoliation claims that might have threatened to derail litigation in the past. Parties lacking such a system must deal with the consequences during discovery.
Limitations on the Scope of Discovery
Now let's consider what happens to the ESI that remains in the system when litigation begins. Rule 26(b)(2)(b) creates what some call the "two-tier" structure of ESI discovery. It operates to wall off some ESI by providing that the responding party "need not provide discovery of electronically stored information from sources that the party identifies as not reasonably accessible because of undue burden or cost." The rule essentially states that a responding party need not produce all the hard stuff?old data, backup tapes, legacy disks that no modern computer can read-if that would incur undue burden or cost. But collecting and producing any ESI imposes costs. So parties and courts must determine when the burden or cost becomes "undue."
Cases such as the Zubulake series of opinions?especially Zubulake v. UBS Warburg, LLC (I) (217 F.R.D. 309 (S.D.N.Y. 2003)) and Zubulake v. UBS Warburg, LLC (III) (216 F.R.D. 280 (S.D.N.Y. 2003))?provide some guidance. The Zubulake court articulated a seven-factor test for determining when to shift the cost of producing "inaccessible" data. Zubulake III at 284 explains that the test articulated in Zubulake I "was designed to simplify application of the Rule 26(b)(2) proportionality test in the context of electronic data ..." That court defined "inaccessible" in a technical way, to include compressed backup tapes and erased, fragmented, or damaged data that require restoration to use. (Zubulake I at 319.) The new rules largely adopt this test in the same context: cost shifting. But the new rules define "inaccessible" differently than did the Zubulake court, and they provide no specific guidance for when burden and cost rise to the level of "undue."
So the "undue burden" language in Rule 26(b)(2)(b) comes with no user manual. Interpreters may use some common law aids, such as the notion that a company cannot benefit from disorganized record keeping. (Rhone-Poulenc Rorer, Inc. v. Home Indemnity Co., 1991 WL 111040 (E.D. Pa., June 17, 1991).) And the courts could apply the Rule 26(c) test for generic discovery to determine an "undue burden or expense." But no court has addressed the question of accessibility as defined by Rule 26(b)(2)(b). Expect a series of fact-intensive court rulings that attempt to distinguish requested forms of ESI used in the "ordinary course of business" from those that might impose an "undue burden" to produce.
For responding parties, the accessibility test has the greatest potential for cutting costs and eliminating future discovery battles. But to gain the most benefit, responding parties should inventory their information systems before litigation to assess which ESI sources are "inaccessible." These assessments take time and must proceed in a comprehensive, orderly way. They may require detailed discussions among departments that don't collaborate easily?information technology, risk management, legal, finance, and the business units that routinely create and store ESI. But if responding parties wait until after they are sued, they may have trouble developing a credible, consistent, or sustainable litigation position (See the discussion of Rule 26(f) below.)
No doubt, some companies will ignore this opportunity or shove it to the bottom of the priority list. If they do, the new rules could offer a requesting party wide access to both accessible and inaccessible ESI. Furthermore, as fact-specific decisions emerge from the district courts, companies must have the commitment and resources to fine-tune their information systems and policies. Otherwise, an advantage seized once can quickly slip away.
Required Disclosures and Planning for Discovery
Rules 26(f) and 26(a) create a new disclosure and meet-and-confer mechanism relating to all information systems and all ESI. Under Rule 26(f), a responding party generally must discuss preserving and producing ESI no later than 90 to 120 days after a complaint has been served. A responding party must identify, at least by category, which ESI it need not produce because of undue burden or cost. It must identify all sources of ESI?accessible and inaccessible?in connection with its initial disclosures. The list of sources for the ESI includes backup tapes, cell phones, laptops, Blackberries, instant-messaging systems, and home computers.
The Rule 26(f) meet-and-confer provision leads directly to the Rule 16(b) conference, at which point the new rules encourage the court to deal with issues of preservation, cost, form of production, and accessibility. In this circumstance, litigation has barely begun, and already the court may have cemented in place an ESI discovery plan either beneficial or detrimental to the responding party. This is not the time to begin an ESI analysis. An unprepared responding party might even attract attention from its insurance carrier. Indeed, some have.
Production of Documents and ESI
Under Rule 34(b), a requesting party may specify the form of ESI production, including "native" forms and "metadata." Native refers to the form of data used in its ordinary course. For instance, a Microsoft Word document, in its native form, is nothing more than the original, unchanged Word version of that document. The data have not been cleaned or reformatted in PDF or TIFF form. Metadata refers to the embedded, largely invisible information about or behind the on-screen document, such as its revision history, authors, or cell formulas.
Courts have taken different approaches, mostly fact-specific, to determine whether a party must produce metadata. (See Hagenbuch v. 3B6 Sistemi Elettronici Industriali S.R.L., 2006 WL 665005 (N.D. Ill., Mar. 8, 2006).) Rule 26(f), however, requires a discussion of metadata from both a preservation and a production perspective early in the case.
Responding parties may object to the information form demanded in a Rule 34 discovery request and specify a different form of ESI production. Or, if no form is specified, the responding party may choose a preferred form. But, by default, a party "must produce the information in a form or forms in which it is ordinarily maintained" or "reasonably usable."
This standard default includes native data as one option, because parties typically maintain their ESI in native form. However, at least one court, after reviewing the new rules, decided that a responding party's obligation to produce documents in the "way they are ordinarily maintained" meant it must produce them with metadata intact. (Williams v. Sprint/United Mgmt. Co., 230 F.R.D. 640 (D. Kan. 2005).)
Sophisticated responding parties should consider their preferred form of production as one aspect of an overall ESI strategy?including accessibility, cost, disclosure, and leverage. The selection of form, in turn, affects the choice of retention, preservation, collection, review, and production technology (for both hardware and software).
A responding party will either promote or resist native and/or metadata production, and it will leverage the associated burdens and costs in developing its preferred argument leading into the Rule 16(b) conference. A responding party that ignores these choices could find itself on the wrong end of a court order requiring it to produce native ESI and metadata. The court order might impose extreme and unnecessary costs, and expose sensitive information that would not otherwise have been revealed.
Claims of Privilege or Protection
Rule 26(b)(5) requires a party receiving privileged information to return, sequester, or destroy such information. This rule intersects with new language in Rules 26(f) and 16(b) that encourages parties and the court to consider "quick-peeks" at information (allowing the opposing party to review documents and request some of them, which the responding party then reviews for privilege and production). It also allows for "claw-backs" (agreements to produce documents without a privilege review on condition that anything subsequently learned constitutes privileged material). Such arrangements existed before the new rules became effective, and no doubt they will continue in rare cases.
In most cases, however, a responding party would not knowingly produce privileged and other sensitive information, even if the requesting party agreed to return it when asked. So these new privilege rules will probably find few takers.
The effects of the new e-discovery rules will depend on case law and the preparedness of individual parties. Seizing the advantage will require initial effort and ongoing vigilance. Any party, and its counsel, can capitalize on the new rules, with perhaps only a right of first refusal bestowed on large ESI owners. The new rules offer only one certainty: Parties will litigate key words and concepts in the amended language early and often.
Geoff Howard is a partner in the San Francisco office of Bingham McCutchen. He is cochair of the firm's Complex and Class Action Litigation Practice Group.