Understanding What Should Be Saved and What Should Not

With the exception of "electronically stored information" or ESI, the most significant buzzword spawned by the new e-discovery jurisprudence is "accessibility." The concept of accessibility is now codified as an important threshold of discoverability and may also trigger important cost-sharing discussions with the court.

Recently revised Fed. R. Civ. P. 26(b)(2) allows parties to object to the production of electronically stored information "from sources that the party identifies as not reasonably accessible because of undue cost or burden." Such a designation is not absolute protection, however, and the court may order production upon a showing of good cause.

The question of accessibility creeps up quickly, and the risk is more than just theoretical that the failure to raise the "accessibility issue" at the initial conference can result in waiver and, hence, a much more onerous and costly discovery burden. Fed. R. Civ. P. 26(f) requires parties to develop a proposed discovery plan that "indicates the parties' views and proposals concerning ... any issues relating to the disclosure or discovery of electronically stored information [and] what changes should be made in the limitations on discovery imposed under these rules."

Reading the two parts of Rule 26 together, if the parties are discussing production of emails — the most common focus of e-discovery — during the initial conference, they should at that time have a grasp of the extent to which certain custodians' emails are not "accessible" and divert demands for production to the less costly media, at least for the time being. This of course begs the question of what "accessible" means in the context of the rules.

The official commentary to the rule states, "[I]t is not possible to define in a rule the different types of technological features that may affect the burdens and costs of accessing electronically stored information." In other words, the rules do not define "accessible" in black and white terms.

Reliable shades of gray can be found in the case law, however. While not legally binding on courts outside of the Southern District of New York, the multiple opinions authored by Judge Scheindlin in Zubulake v. UBS Warburg are widely regarded as strong persuasive authority on these issues. The first widely cited opinion received attention because of its cost-shifting analysis, in which the court reinforced long-standing precedent requiring the producing party to bear the costs associated with discovery production, while creating a possible exception for data that reside in an "inaccessible" format. Zubulake v. UBS Warburg, LLC, 217 F.R.D. 309 (S.D. N.Y. 2003.) Whether cost-shifting should be allowed for inaccessible data would then depend on a seven-part analysis, which is now incorporated almost verbatim into the commentary to Rule 26.

The Zubulake court first defined accessibility summarily, stating, "Information deemed 'accessible' is stored in a readily usable format." Id. at 319. Recognizing that the question of "[w]hether electronic data is accessible or inaccessible turns largely on the media on which it is stored," the court then ranked the accessibility of five common media as follows:

• Active online data provided generally by magnetic disks, used in the most active stages of the electronic record's life, such as computer hard drives. Accessible.
• Near-line data that are stored on removable media configured to access speeds ranging from milliseconds to minutes, such as optical discs. Accessible.
• Offline storage/archives that is typically used for disaster recovery or archiving and involves manual intervention and is much slower than online or near-line storage, with access speeds ranging from hours to days. The difference between near-line and offline data is that offline data lack the coordinated control of an intelligent disk subsystem. Sometimes accessible, sometimes inaccessible.
• Backup tapes that often compress the data by sequential access, meaning that in order to read a "block of information" one must reconstruct/read the preceding "blocks." The data on backup tapes are not organized for retrieval of individual documents, but rather are created in case a disaster requires the reconstruction of the entire operating environment, a painstaking and expensive process that yields varying levels of success. Inaccessible.
• Erased, fragmented or damaged data that are only retrievable, if at all, using sophisticated forensic tools and professionals. Inaccessible.

As the above classifications indicate, the continuum of accessibility evolves over time, with data converting from "accessible" when first created to "inaccessible" either when deleted or transferred to disaster-recovery backup tapes. Because of this, it is absolutely critical to gain an understanding of when and how such progressions evolve, since there are many ways that data can be made inaccessible or, conversely, preserved unnecessarily as "accessible" and therefore an easier target for production. In deciding whether production is to be made, perhaps at the significant expense of the producing party, courts will consider the extent to which the party could have prevented the data from becoming inaccessible.

Specifically, the seven-part test that is essentially copied from Zubulake to the commentary to Rule 26 states, in pertinent part:

The decision whether to require a responding party to search for and produce information that is not reasonably accessible depends not only on the burdens and costs of doing so, but also on whether those burdens and costs can be justified in the circumstances of the case. Appropriate considerations may include: ...the failure to produce relevant information that seems likely to have existed but is no longer available on more easily accessed sources. (Emphasis added)

Under the federal analysis, therefore, parties who allow manual deletion, automatic deletion, archiving or transfer to backup media face the substantial risk that they may be required to bear the increased expenses associated with production later.

Consider this example: An employee is terminated for performance reasons and suggests to a coworker that in retaliation he may file suit alleging discrimination. Management is informed of this and documents the personnel file and conducts an investigation, finding the allegations to be meritless. Eighteen months later a lawsuit is filed. During this period the autodelete settings on the network server continued to delete emails that were more than 90 days old. In addition, two of the plaintiff's coworkers and one supervisor have since left the company and their email accounts, pursuant to company protocol, were removed from the network.

Though it did not intend to allow the loss of evidence, the company, now a defendant, finds that a substantial number of emails are no longer "accessible," although the inboxes would still exist on backup tapes created during the time that the accounts were active. Initial correspondence, disclosures and proposed discovery plans all name the coworkers and supervisor as potential persons with knowledge and the plaintiff contends that the most discriminatory disparaging remarks occurred via emails, including inappropriate jokes.

Under this scenario, a corporate defendant will have a difficult time deflecting demands for restoration of at least some of the backup tapes. In considering how to rule on a motion, the court will consider that the company was aware of possible litigation, and the extent to which it could have preserved the email accounts in an accessible format. If the court finds that it would have been relatively easy and inexpensive to preserve the emails on the network, it is likely to rule that the corporation bear the significant cost of restoring backup tapes to retrieve old emails.

The above facts arise in litigation with increasing frequency and many variations. In response to the developments in the case law and new rules, many companies are deploying sophisticated archiving hardware and software to capture all emails without burdening the active network. Such tools are likely to fall in the middle category of sources of "offline storage/archives" listed above in Zubulake.

Before deciding to protect the company with an "email vault," there are risks and benefits to be evaluated so that the cure does not exceed the harm posed by the problem. For example, an almost infinite virtual warehouse of emails that unnecessarily goes back for years and encompasses thousands of employees is not necessarily a good thing, should an adversary be granted a license to "fish" using electronic search tools. Opposition to such adventures is usually inspired by the need to pay attorneys to review all emails prior to production, but courts are hearing increasingly creative, inventive approaches to undercut such arguments, not the least of which is the proposed "clawback agreement" allowing for the return and exclusion from evidence of inadvertently produced evidence.

All of these concerns are manageable, but require careful consideration in consultation with counsel, the IT department and others responsible for litigation and legal compliance. The evaluation of "accessible" media should take place in the context of a larger discussion concerning document retention, litigation hold policies (which will include the suspension of automatic deletion) and the identification of those responsible for the proper implementation of policy, both proactive and reactive. The sooner this discussion takes place, the better.

This Alert was authored by Duane Morris eDiscovery Team member John J. Coughlin.

For Further Information

If you have any questions about this Alert or would like more information on eDiscovery, please contact John J. Soroko, John C. Herman, one of the other eDiscovery Team attorneys or the attorney in the firm with whom you are regularly in contact.

Disclaimer: This Alert has been prepared and published for informational purposes only and is not offered, nor should be construed, as legal advice. For more information, please see the firm's full disclaimer.