Win the battle for access to e-data: the defense may try to limit discovery of electronically stored information by branding data 'not reasonably accessible.' Understanding the new rules for electronic discovery can give you an edge.

Amendments to the Federal Rules of Civil Procedure regarding the discovery of electronically stored information became effective December 1. The amendment to Rule 26(b)(2) creates two significantly different categories of electronically stored information: reasonably accessible and not reasonably accessible. A party from whom discovery is sought need not review or produce electronically stored information that is not reasonably accessible, even if it is otherwise discoverable within the meaning of the rule.

The amended rule provides that the court may nonetheless order discovery of information not reasonably accessible if "good cause" is shown. But it's likely that the requesting party would have to bear a part, or all, of the costs of that discovery. The far better course for plaintiff lawyers is to win the first battle--to succeed in convincing the court that the information sought is reasonably accessible.

Since 1970, when Federal Rule of Civil Procedure 34 was amended to include "data compilation," information stored electronically has been fair game during discovery. (1) As information has become increasingly digitized--more than 90 percent of all documents are now created and stored electronically, and more than 75 percent of documents created are never printed (2)--so too has discovery increasingly come to encompass electronic information. Requests for e-mail are standard practice, (3) and more esoteric types of electronic information--from databases, networks, and spreadsheets to voice mail--are regularly sought in discovery. (4)

Many legal issues arise during discovery because electronic documents have qualities that paper lacks, including greater volume, dispersal, and searchability; persistence; changeable content; underlying metadata; and environmental dependence (that is, the information is not intelligible absent a system to "translate"). Courts have struggled to apply the discovery rules to digital documents--in particular, the "proportionality test" set forth in Rule 26(b)(2)(iii), which says that the court may limit discovery if the "burden or expense of the proposed discovery outweighs its likely benefit." Because "the universe of discoverable material has expanded exponentially" (5) and the traditional method used to identify responsive paper records--piling them in a room and reading them--cannot be used for databases, (6) responding to requests for electronic data poses unique burdens.

Or so the responding parties--typically defendants--have convinced the courts. (7) The solution generally adopted by the federal courts has been cost-shifting, forcing the requesting party to bear all or part of the cost of cataloging, restoring, and processing electronically stored information. (8) The district courts have modified and expanded on the proportionality test to determine when and whether cost-shifting is appropriate in requests for electronic information. (9) But the new Rule 26(b)(2) takes a different approach.

Two-tier analysis

The amendment to Rule 26(b)(2) establishes a two-tier procedure and raises the bar for discovery of information "not reasonably accessible." Under the new rule, even if the responding party knows or has reason to believe that it has relevant information, it need not locate, review, or produce that information if it is not reasonably accessible. Instead, the responding party must only "identify, by category or type, the sources containing potentially responsive information that it is neither searching nor producing" (10) (tier one) unless "good cause" is shown (tier two).

Tier one. The requesting party may challenge the producing party's claim that certain information is not reasonably accessible and may request discovery on that issue. In such a case, a court might "requir[e] the responding party to conduct a sampling of information contained in the sources identified as not reasonably accessible; allow[] some form of inspection of such sources; or tak[e] depositions of witnesses knowledgeable about the responding party's information systems." (11) The responding party bears the burden of showing that the information is not reasonably accessible. (12)

Tier two. If the court agrees that the requested information is not reasonably accessible, the requesting party must show good cause for discovery. "Good cause" refers to the proportionality test of the existing rule (13) but is, apparently, not identical, for the advisory committee note lists seven factors for the courts to consider in determining whether good cause has been shown, including the "specificity of the discovery request," "predictions as to the importance and usefulness of the further information," and "the parties' resources." (14)

The committee also says that "a requesting party's willingness to share or bear the access costs may be weighed by the court in determining whether there is good cause." (15) Whether or not the requesting party is willing, the court may use its authority as set forth in Rule 26(b)(2)(C) to impose conditions for discovery, including "payment by the requesting party of part or all of the 'reasonable costs' of obtaining [the] information." (16)

A court's finding that certain data is reasonably accessible--or not--has significant practical consequences for the parties. Despite the importance of this determination, the rule does not clearly define "not reasonably accessible." Rule 26(b)(2)(B) says only that information becomes inaccessible "because of undue burden or cost" associated with producing it.

The advisory committee's introduction to the amendment provides some guidance. It lists examples of data that would be substantially burdensome or costly for the producing party to retrieve:

   backup tapes intended for disaster recovery
   purposes that are often not indexed, organized,
   or susceptible to electronic searching;
   legacy data that remains from obsolete
   systems and is unintelligible on successor
   systems; data that was "deleted" but remains
   in fragmented form, requiring a modern
   version of forensics to restore and retrieve;
   and databases that were designed to create
   certain information in certain ways and that
   cannot readily create very different kinds or
   forms of information. (17)

These examples are helpful in the near term but are likely to lose relevance quickly given the rapid pace of technological change, especially in the field of data-formatting and storage technologies.

Arguing for access

It will take some time for courts to develop a body of case law that fitly defines which electronic materials are reasonably accessible and which are not. Plaintiff lawyers will need to make strong arguments to prevent defendants from unfairly blocking discovery of information that is not overly burdensome or costly for them to produce. The following arguments may prove persuasive.

"Undue burden or cost" is a uniquely stringent test. A distinction must be made between "undue burden or cost" within the meaning of amended Rule 26(b)(2)(B) and "unduly burdensome" within the meaning of the proportionality test, or else the two-tier test would be meaningless. And because information deemed not reasonably accessible is specially protected (it need not be searched for responsive documents), the "undue burden or cost" test should be more demanding than the "unduly burden-some" standard.

For these reasons, if the producing party cites case law applying cost-shifting tests to particular types of data, (18) plaintiff attorneys should argue that these cases are not directly relevant to the tier-one analysis. On the other hand, cases holding that cost-shifting would be inappropriate for certain types of electronic data (19) could help persuade the court that the same categories are reasonably accessible within the meaning of the amended rule.

"Readily usable data" should be deemed reasonably accessible. Zubulake I sets up a two-tier cost-shifting test that is analogous enough to the two-tier test of amended Rule 26(b)(2)(B) to support the decision's relevance in a dispute over accessibility of electronic data. Zubulake I's cost-shifting test divides data into accessible and inaccessible formats. If data is accessible, "it would be wholly inappropriate to even consider cost-shifting." (20) For inaccessible data, a seven-factor proportionality test determines whether and to what extent the costs of production will be shifted to the requesting party. (21)

According to the Zubulake I court, "information deemed 'accessible' is stored in a readily usable format. Although the time it takes to actually access the data ranges from milliseconds to days, the data does not need to be restored or otherwise manipulated to be usable. 'Inaccessible' data, on the other hand, is not readily usable." (22)

Using these definitions, "whether electronic data is accessible or inaccessible turns largely on the media in which it is stored." (23) The court listed five categories of data. Three were considered accessible: active, online data; near-line data (that is, data on removable storage media that can readily be inserted into read/write devices, like optical discs, CDs, and DVDs); and offline storage archives.

Two types were considered inaccessible because the data was not readily usable as stored: backup tapes and erased, fragmented, or damaged data. The court ordered the defendant to produce all responsive e-mail in accessible format and, in a subsequent decision, ordered the plaintiff to bear a part of the costs of retrieving and processing responsive e-mail in inaccessible format. (24)

Zubulake I can be used to rebut any defense argument that active, online data, near-line data, or offline storage archives are not reasonably accessible within the meaning of amended Rule 26(b)(2)(B).

Trouble spots

Defendants are especially likely to object to production of certain types of electronic data on the grounds that it is "inaccessible." But well-crafted arguments can convince courts to require that information from these sources be turned over. These include backup files, data that requires "restoration," legacy data, and databases.

Backup files. Given that both the advisory committee and the Zubulake I court identified backup tapes as an example of data that is not reasonably accessible, obtaining backup data in discovery may seem a lost cause. But plaintiff lawyers should not give up the fight.

First, the committee limits its comments regarding backup tapes to those "intended for disaster recovery purposes that are often not indexed, organized, or susceptible to electronic searching." But "indexed" is a matter of degree. If backup tapes were completely unorganized, they would be nearly useless to the owner in the event of a business emergency. Consider conducting discovery to identify the procedures the responding party used to organize and store backup tapes.

Second, most tape-backup software automatically creates the equivalent of a table of contents or index. Retrieval of an index--as opposed to all the information contained on the tapes--is relatively simple; the index, at least, should be considered reasonably accessible. It might yield information that would justify pursuing discovery of the tapes themselves, even if some or all of the production costs are shifted to the requesting party.

Third, backup technologies have already improved since Zubulake I, which described the troublesome qualities of taped backup:

   'The disadvantage of tape drives is that they
   are sequential-access devices, which means
   that to read any particular block of data,
   you need to read all the preceding blocks.'
   As a result,' [t]he data on a backup tape are
   not organized for retrieval of individual
   documents or files [because] ... the organization
   of the data mirrors the computer's
   structure.' (25)

More and more often, businesses are opting for backup and recovery services delivered over the Internet or for disc-based backup, rather than tape drives. Both of these newer technologies overcome the sequential-access problem of tape drives. Both save data in a readily usable format. They should not be equated with older backup systems that use tape drives to store data.

"Restored" data. The court in Zubulake I concluded that data on backup tapes is inaccessible in part because it has been compressed for storage and must be "restored" to be readable. (26) The advisory committee introduction to amended Rule 26(b)(2) notes that a responding party might have difficulty reviewing or producing data that requires restoration or translation.

But restoration is not necessarily burdensome or costly. The WinZip utility compresses and organizes multiple compressed files into a single zip file. The individual files are restored with one easy click. This is not to suggest that it is worth arguing that the data stored on conventional magnetic backup tapes is readily usable or accessible. However, the objection that data must be "restored" should not simply be taken at face value.

What exactly that restoration entails should be analyzed. For example, the process of decompressing data might involve someone simply "babysitting" a computer--keeping an eye out to restart a failed or aborted decompression program--while the technical expert does other work. That restoration should be considered low on the scale of burden or hardship.

Legacy data. According to the advisory committee, legacy data--that is, data created by or stored in systems or architecture that the business no longer uses--is not reasonably accessible if it is "unintelligible" on the company's existing systems. But because data that is truly irretrievable is useless to the business, any legacy data related to operations or regulatory requirements must be accessible by some means.

If necessary, the plaintiff lawyer should ask how legacy data is retrieved. The answer may be that the legacy data is stored solely because of legal requirements and cannot be accessed absent the purchase of additional software. The court would be unlikely to require the defendant to do so, unless the plaintiff paid. But if the defendant needs to be able to retrieve and use the so-called legacy data, it is going to have some means of doing so. For example, it may have a limited license agreement for the application that created the legacy data, even if the business no longer uses that application to capture new data. If so, that data is still reasonably accessible.

Databases. Even before the amended rule was adopted, many defendants argued that producing databases is a burden, usually saying that they must be decoded or disassembled before the requesting party can use them. The advisory committee also adverts to this potential problem: Databases that "cannot readily create very different kinds or forms of information" from the kind or form for which they were designed are a problem in discovery. (27)

But that assumption can be proven wrong. For example, if the responding party claims that the particular form of information sought from a database--say, a report or listing of particular data entries--is not the form for which the database was designed, the requesting party's next question might be: Have you ever produced this type of report or list? If so, that document should be considered reasonably accessible.

Another approach is to reexamine the request for production. If the request does in fact require that information be created in a different form, the requesting party should examine whether the request can be reformulated to fit the design of the database while still obtaining relevant data.

If the database is designed to produce only a specific output, at least that data can be obtained. If that data can be replicated and produced in a usable format without being reprogrammed, the responding party should bear the cost of producing it; if significant time would be required to extract or manipulate the data, that cost could well be shifted, at least in part. (28)

The amendments to Rule 26(b)(2) will pose new hurdles to the discovery of electronically stored information. It is safe to assume that defendants will attempt to use the amended rule to avoid reviewing and responding to requests that they characterize as seeking inaccessible data. Plaintiff lawyers will need to be, or become, technologically savvy to make creative and compelling arguments for obtaining electronic evidence in their cases.

Notes

(1.) See e.g. Daewoo Elecs. Co. v. United States, 650 F. Supp. 1003, 1006 (Ct. Intl. Trade 1986).

(2.) Peter Lyman & Hal R. Varian, How Much Information? 2003 1 (U. Cal. 2003), www2.sims. berkeley.edu/research/projects/how-much-info-2003/printable_report.pdf (last accessed Sept. 25, 2006).

(3.) See generally Michael Marron, Student Author, Discoverability of Deleted' E-Mail: Time for a Closer Examination, 25 Seattle U. L. Rev. 895 (2002).

(4.) The Civil Discovery Standards, revised in August 2004 by the Litigation Section of the American Bar Association, lists 9 categories of data and 15 platforms that "should [be] consider[ed]" in preparing for electronic discovery. ABA, Civil Discovery Standards 57-58 (1999, rev. 2004), http:// abanet.org/litigation/discoverystandards/ 2004civildiscoverystandards.pdf (last accessed Sept. 25, 2006).

(5.) Zubulake v. UBS Warburg, LLC, 217 F.R.D. 309, 311 (S.D.N.Y. 2003) [hereinafter Zubulake I]. This employment discrimination case spawned five published opinions resolving electronic discovery disputes.

(6.) See generally Corinne L. Giacobbe, Student Author, Allocating Discovery Costs in the Computer Age: Deciding Who Should Bear the Costs of Discovery of Electronically Stored Data, 57 Wash. & Lee L. Rev. 257 (2000).

(7.) Indeed, many courts have "assumed that an undue burden or expense may arise simply because electronic information is involved," which, as the Zubulake I court noted, "makes no sense. Electronic evidence is frequently cheaper and easier to produce than paper evidence because it can be searched automatically, key words can be run for privilege checks, and the production can be made in electronic form obviating the need for mass photocopying." 217 F.R.D. at 318.

(8.) Id. at 316. As a general rule, cost-shifting is not considered appropriate for hours billed by attorneys and paralegals for document review before production. (Zubulake v. UBS Warburg, 216 F.R.D. 280, 290 (S.D.N.Y. 2003) [hereinafter Zubulake III].)

(9.) Courts use a variety of tests, but three often cited are Zubulake I, 217 F.R.D. at 322-23; Rowe Ent., Inc. v. William Morris Agency, Inc., 205 F.R.D. 421,429 (S.D.N.Y. 2002); and McPeek v. Ashcroft, 202 F.R.D. 31,34-35 (D.D.C. 2001).

(10.) Fed. R. Cir. P. 26(b)(2) advisory comm. n. at 45-46, www.uscourts.gov/rules/Reports/ ST09-2005.pdf, go to pp. 132-33 last accessed Sept. 25, 2006).

(11.) Id. at 46-47.

(12.) Id. at 47.

(13.) Fed. R. Civ. P. 26(b)(2) advisory comm. introduction at 41, www.uscourts.gov/rules/ Reports/ST09-2005.pdf, go to p. 128 (last accessed Sept. 25, 2006).

(14.) Fed. R. Cir. P. 26(b)(2) advisory comm. n., supra n. 10, at 47. The seven-part list in the advisory committee note is similar, though not identical, to the seven-part test for cost-shifting set forth in Zubulake I, 217 F.R.D. at 322-23.

(15.) Fed. R. Civ. P. 26(b)(2) advisory comm. n., supra n. 10, at 48.

(16.) Id.

(17.) Fed. R. Civ. P. 26(b)(2) advisory comm. introduction, supra n. 13, at 40.

(18.) See e.g. Portis v. City of Chi., 2004 WL 1535854 at ** 5-6 (N.D. Ill. July 7, 2004); Multi-technology Servs., L.P. v. Verizon S. W. f/k/a GTE S. W. Inc., 2004 WL 1553480 (N.D. Tex. July 12, 2004); Open TV v. Liberate Techs., 219 F.R.D. 474 (N.D. Cal. 2003).

(19.) See e.g. In re Bristol-Myers Squibb Sec. Litig., 205 F.R.D. 437 (D.N.J. 2002); see also Eggleston v. Wal-Mart Stores E., LP, 2006 WL 585152 (E.D. Va. Mar. 10, 2006).

(20.) Zubulake I, 217 F.R.D. at 320.

(21.) Id. at 322-23. Other influential cost-shifting tests do not incorporate this two-step approach. See Rowe, 205 F.R.D. 421; McPeek, 202 F.R.D. 31.

(22.) Zubulake I, 217 F.R.D at 320.

(23.) Id. at 318.

(24.) Cost-shifting was ordered in Zubulake III (216 F.R.D. 280).

(25.) 217 F.R.D. at 318 (quoting, respectively, Internetnews.com, Tape Drive (rev. June 21, 2002), http://inews.webopedia.com/TERM/ t/tape_drive.html (last accessed Sept. 25, 2006); Kenneth J. Withers, Computer-Based Discovery in Federal Civil Litigation, 2000 Fed. Cts. L. Rev. 2, 5, www.fclr.org/2000fedctslrev2.htm (last accessed Sept. 25, 2006)).

(26.) Id. at 319-20.

(27.) Fed. R. Cir. P. 26(b)(2) advisory comm. n., supra n. 10, at 40.

(28.) Compare Jones v. Goord, 2002 WL 1007614 (S.D.N.Y. May 16, 2002) with Open TV, 219 F.R.D. 474.

MARIAN K. RIEDY, a trial lawyer, and SUMAN BEROS, a technology expert, provide e-litigation consulting services through their Washington, D. C., firm, Riedy & Beros.