Electronic discovery involves the collection, review, and production
of electronically stored information--such as e-mail, word processing
documents, spreadsheets, and databases--in accordance with state or federal
discovery requirements. With the increased proliferation of electronic
data, electronic discovery is one of the most rapidly evolving areas of
Just one personal hard drive can contain 1.5 million pages of data, and one corporate backup tape can contain 4 million pages of data. Thus the magnitude of electronic data that needs to be handled in discovery is staggering. In most corporate civil lawsuits, several backup tapes, hard drives, and removable media are involved. Depending on the circumstances of a case, the costs of electronic investigation and production can be significant. Not surprisingly, some of the most intense arguments ensue over how to allocate the costs associated with e-discovery.
Under traditional discovery rules, each side typically bears the cost of producing its own documents.1 As one California court noted, "It is a well accepted principle that each party to litigation normally bears the ordinary burden of financing his or her own suit."2
In the e-discovery arena, Zubulake v. UBS Warburg,3 a 2003 decision, generally has provided guidance for analyzing cost-shifting arguments, even though the case--from the Southern District of New York--is only persuasive authority in jurisdictions other than its own. Zubulake sets forth a seven-factor balancing test to determine whether the requesting party or producing party should bear the costs of a discovery request. Under the Zubulake test, cost shifting for electronic discovery may be warranted based on the weight accorded to:
1) The extent to which a discovery request is specifically tailored
to discover relevant information.
2) The availability of the requested information from other sources.
3) The total cost of production compared to the amount in controversy.
4) The total cost of production compared to the resources available to
5) The relative ability of each party to control costs and its incentive
to do so.
6) The importance of the issue at stake in the litigation.
7) The relative benefits to the parties of obtaining the information.
Last year, however, California charted a new course in the cost-shifting debate with the Sixth District Court of Appeal's decision in Toshiba America Electronic Components, Inc. v. Superior Court of Santa Clara County.4 Toshiba is the first appellate court ruling providing direction on e-discovery cost-shifting issues in California state courts and could significantly affect the way litigators in the state approach cost-shifting arguments. Rejecting the Zubulake cost-shifting test, the Toshiba court created an alternative standard for determining who bears e-discovery costs in cases pending in California courts. In today's technology-driven environment, California lawyers handling discovery matters must be familiar not only with the case law in this area but also with the complexities of backup tape discovery.
The underlying action of Toshiba involved Lexar Media, Inc., filing suit against Toshiba America Electronic Components, Inc. (Toshiba), and its parent company Toshiba, Inc., for misappropriating trade secrets, breaching fiduciary duty, and unfair competition. Lexar moved for production of over 800 backup tapes and asked that Toshiba bear the costs for the data retrieval. After receiving an estimate that the tapes would cost $1.5 million to $1.9 million to produce, Toshiba asked Lexar to assist in or completely cover the costs. Toshiba claimed that paying for the costs on its own would result in an undue burden and expense because of the large volume of data and because some of the tapes had become obsolete, making the data accessible only through the use of specialized tools. Moreover, the staff members that performed the tape backup work for the company were no longer employed by Toshiba and were thus unavailable to assist with the discovery request.
Lexar refused to bear any production costs. Instead, Lexar filed a motion to compel and contended it should not be penalized because Toshiba chose to keep records in a format that made data retrieval difficult. Without comment, explanation, or a suggestion of a sampling protocol, the trial court ordered Toshiba to produce the backup tape data and to bear all production costs.
Toshiba appealed, claiming that Code of Civil Procedure Section 2031(g)(1) was an automatic cost-shifting provision that required Lexar to pay all or part of the costs of restoring and searching the tapes. In its ruling, the appellate court declined to apply the Zubulake seven-factor cost-shifting test. Instead, the court referred to Section 2031 and stated, "[O]ur Legislature has identified the expense of translating data compilations into usable form as one that, in the public's interest, should be placed upon the demanding party." Thus the appellate court, using Section 2031 as authority, held that Lexar must pay for translating the backup tape data into a usable form if the restoration costs for the backup tapes were found to be a "reasonable expense for a necessary translation." The court also indicated that the parties should consider using data sampling to determine the relevancy of the data contained on the tapes.
While seemingly resolving a major e-discovery issue, Toshiba raises several unanswered questions for California lawyers in future matters. For instance, what kinds of problems will continue to be associated with backup tape discovery? What does the phrase "reasonable expense for a necessary translation" mean? How can a California practitioner address these issues?
The Challenges of Backup Tapes
Many obstacles can prevent a company from recovering or restoring backup data, such as the data requested in Toshiba. Even if a company physically has all the necessary backup tapes, pinpointing the location of specific data can prove difficult, due to the complexity of company network infrastructures and corporate IT backup tape policies.
The recent Coleman (Parent) Holdings, Inc. v. Morgan Stanley & Company case5 illustrates that courts are not hesitant to impose significant sanctions for a party's deficient discovery practices relating to backup tapes. While performing discovery in-house, Morgan Stanley overwrote e-mails, failed to timely process and produce data stored on hundreds of tapes, and failed to produce e-mails and attachments throughout the discovery process. The court declared that Morgan Stanley "gave no thought to using an outside contractor to expedite the process of completing the discovery, though it had certified completion months earlier; it lacked the technological capacity to upload and search the data at that time, and would not attain that capacity for months."6 The court issued an adverse inference instruction, directing the jury to accept that Morgan Stanley helped defraud investors. In May 2005, relying in part on that instruction, the jury awarded $1.45 billion in total damages against Morgan Stanley.
When backup tapes are at issue, one of the biggest initial challenges is developing an understanding about a particular company's network infrastructure. While most networks are similar in theory, each company's network environment is unique. For example, a simple network is similar to a wagon wheel--the center of the wheel is the server, the spokes are cables, and the end points represent individual workstations. In this type of network, a single server processes all e-mail, accounting data, word processing, and Web surfing and hosting, among other functions. Depending on the organization's retention policy, the network server could be backed up and stored on a backup tape on a daily, weekly, and/or monthly basis.
In today's increasingly digital workplaces, a small or medium-sized office may have five to 30 network servers that may be backed up to a single backup tape or backed up independently by server or function. Furthermore, larger corporations typically operate by using hundreds or thousands of servers that may be segregated by function or task, business unit, or geography. Each of these servers may be backed up by a system of multiple backup tapes. To add further complexities, the server or network may have undergone significant upgrades or application changes during the time in which relevant data is being sought. With all of this stored data, identifying the specific backup tape that contains relevant information can be an overwhelming task.
Producing data from backup tapes is hardly a straightforward task. For instance, a company may have experienced a merger, an acquisition, or downsizing. In a merger or acquisition, the new company likely has a completely different network configuration. As with corporate downsizing, in a merger or acquisition situation it is not atypical for the IT department to lose resources. Thus, many of the key people who are knowledgeable about data locations, associated time periods, and the company's tape retention policy may no longer be with the company. This leaves counsel with mounds of data and no IT guides to offer insight about them.
The lack of backup tape logs can present further difficulties. In a perfect world, backup tape logs would keep track of what data is contained on a particular backup tape, specifying the time frame and data type (such as e-mail server, accounting server, human resources server, and the like). However, when backup tape logs are not properly kept or the individuals who knew where the logs are located are no longer with the company, problems arise. Some companies' IT departments do a commendable job of individually labeling each backup tape so that anyone retrieving the tape will know what data is contained on it. Unfortunately, however, when tapes are labeled merely with a tag number, a log must be found to identify what data and date range correspond to the tag and are included on the tape.
The use of storage area networks (SAN)--an alternative (and less expensive) approach to network storage--may add further complications to backup tape discovery. A traditional backup environment is configured so that backups are segregated by function or task. When it is necessary to pull tapes, a backup log provides a reference for determining which specific tapes are needed. For example, if there is a group of 6,000 backup tapes for an entire network, but tapes containing accounting data or customer service data are irrelevant to a particular case, IT experts can quickly and easily remove those tapes from the group. In a SAN environment, however, this process may not be so easy, because all the data may be lumped together, adding time and costs to the recovery of specific requested information.
When data is stored overseas, language and timing issues may arise. Also, as in the situation in Toshiba, the personnel that handled the tape backups may no longer work for the company. Without the expertise of the people best able to determine the most effective way of retrieving the specified data, the information may not only be expensive to reach but may sometimes be lost.
Indeed, locating relevant tapes is simply step one in the discovery process. Tapes must be "restored" in order to review the data contained on them. Many corporations may not be able to restore their backup tapes if they no longer have the tape drive hardware and software to read the tapes or if the backup tapes contain "legacy" data (that is, data created with applications that no longer exist on the company's servers). Corporations also may not have the requisite server capacity, or they may have outsourced the staff they need to perform the work. These problems do not mean that restoration is impossible, but a company may need third-party resources for assistance. During the restoration process, the requested data must be segregated from all other data that may also reside on the tapes. The segregated data must be reviewed, with items noted as privileged or for redaction, and ultimately produced to the opposing party or court.
With the potential complexities and costs associated with discovering
and restoring backup tape data, the Toshiba case serves as an important
reminder that counsel must remain focused on continuing changes in the
law of electronic discovery. In one of Zubulake's e-discovery opinions,7
the court, in addressing the general role of counsel in litigation, stated
that "[c]ounsel must take affirmative steps to monitor compliance so that
all sources of discoverable information are identified and searched."
This includes assuming responsibility for uncovering relevant documents
(with the assistance of an expert if necessary), evaluating discovery
requests for electronic information, and properly preserving this information.
Unmistakably, litigators have a clear-cut duty to locate, preserve, and
produce relevant electronic information. When faced with complex discovery
involving backup tapes, practitioners should take the following steps:
Become knowledgeable about the data. From hard drives
and backup tapes to PDAs and USB drives, electronic data can exist in
a variety of formats, locations, and volumes. Whether representing the
requesting or producing party, practitioners must become technologically
savvy to make the best argument for or against the "reasonableness" of
expenses that are "necessary" to translate electronic data compilations.
Thoroughly researching the accessibility of the data and how much it will
cost to recover, search, and produce the data in discovery will put counsel
in the best position to zealously argue for their clients. In addition,
counsel should know where the data is located. Is it stored overseas?
In multiple offices? Counsel also should pinpoint relevant time periods
and specify which tapes address those periods. Finally, attorneys should
become knowledgeable about the tape rotation policy of a company. For
instance, if a company has only preserved monthly backup tapes for the
time period at issue, they may not contain the information sought. In
fact, the data may only reside at the workstation where it was created.
Determine what data is necessary. After pinpointing
key data locations, counsel should ascertain which tapes may be pertinent.
Examining the backup tape logs for relevancy and segregating potentially
relevant data can help winnow down the amount of data. For example, if
one-third of the tapes represent accounting data and that data is not
subject to the litigation, the tapes are unnecessary. If backup logs are
unavailable, a personal inspection of the tapes can reveal whether they
are self-identifying. Alternatively, an electronic evidence expert can
restore the content of the tapes.
Revamp cost-shifting arguments. While the Toshiba decision
requires the requesting party to pay the reasonable expense of translating
a data compilation into usable form, the court noted that reasonableness
and necessity are factual issues. The opinion does not require the requesting
party to pay all the costs associated with retrieving usable data. Rather,
the requesting party is expected to pay only reasonable expenses when
it is necessary to translate electronic data compilations--such as those
contained on a backup tape--to obtain usable information. The trial court
will determine what is reasonable and necessary on a case-by-case factual
basis, leaving practitioners room to craft an argument one way or the
other. The Toshiba court further stated that if the demanding party does
not believe the translations are necessary or if it disagrees with the
producing party on the reasonableness of the data costs, it has the option
of seeking a protective order.
Carefully consider cost-shifting trends. The Toshiba
decision is not alone in requiring requesting parties to pay production
costs. Other courts have issued a variety of opinions shifting burdensome
or expensive discovery to the requesting party.8
Consider data sampling. Although the importance of
preserving any data that may be relevant in litigation cannot be overstated,
counsel should consider data sampling. For example, in McPeek v. Ashcroft,9
the court adopted a sampling approach for backup tapes, noting that "economic
considerations have to be pertinent if the court is to remain faithful
to its responsibility to prevent 'undue burden or expense'....If the likelihood
of finding something was the only criterion, there is a risk that someone
will have to spend hundreds of thousands of dollars to produce a single
e-mail. That is an awfully expensive needle to justify searching a haystack."
In a subsequent decision,10 after examining the likelihood of relevant
data being contained on each of the backup tapes, the magistrate ordered
additional searches of selected backup tapes likely to contain relevant
In Hagemeyer North American, Inc. v. Gateway Data Sciences Corporation,11 the court used a sampling approach to evaluate the applicability of cost-shifting. It instructed the defendant to recover responsive data from any five backup tapes of the plaintiff's choosing, and after that process the court would assess whether the cost of recovering e-mails from the remaining backup tapes would be proportionate to the likely benefit.
The Toshiba court also indicated that data sampling would be an appropriate test for determining necessity. When sampling is used, the requesting party chooses a small selection of data for the producing party to restore, search, and produce. The selected data usually exists on so-called inaccessible sources, such as backup tapes for which restoration and production costs may be significant. The court then evaluates the effectiveness of the sample search and fashions an appropriate production and cost order based on a cost-allocation test that the court deems appropriate. If data sampling will be helpful in a case, the practitioner should commence the process by meeting with opposing counsel to forge an agreement on sampling procedures, including the time period of the sample and the custodian. If necessary, the attorneys should seek assistance from the court and an electronic evidence expert.
The Toshiba decision is a pioneering opinion regarding cost allocation in e-discovery. Attempting to chart a course for California, the Toshiba court plainly stated that California law should apply in California e-discovery issues. However, e-discovery case law remains jurisdiction-specific, unsettled, and conflicting. Staying on top of various e-discovery standards and judicial decisions will put litigators in the best position to effectively manage the discovery of electronic data.