In FTC v. Boehringer Ingelheim Pharmaceuticals, No. 1:09-mc-00564-JMF (D.D.C. 10/16/12), U.S. Magistrate Judge John M. Facciola of the District of Columbia, well-known and respected for his opinions regarding e-discovery matters, held that responding to the FTC’s administrative subpoena would not be overly burdensome even though it involved the search of backup tapes.
While the court applied a test less favorable to producing parties than that found in Federal Rule of Civil Procedure 26(b)(2)(B) to measure what is “burdensome,” it had no real effect on the outcome of the case. What did influence the outcome was the FTC’s willingness to allow the defendant to focus on only a handful of tapes, and the court’s insistence that the FTC and the defendant seize upon that willingness to come to an informed agreement beneficial to both parties.
The opinion is important to practitioners because it, again, illustrates the effectiveness of focused and informed requests and responses, and that it demonstrates that backup tapes are not off-limits for searching, or immediately presumed to be inaccessible sources of data.
Why Backup Tapes Are Bad
It is no stretch to assert that the case that distilled the most important issues emerging in the new world of e-discovery at the turn of the century, and whose approach to those issues was codified in the December 2006 amendments to the Federal Rules, was Zubulake v. UBS Warburg. One of those issues was whether data on backup tapes was “inaccessible” and so did not have to be restored. The court’s approach to the issue has framed the way we think of backup tape restoration.
First, data on backup tapes is almost never “inaccessible,” from a technical perspective. Any data that was technically inaccessible — e.g., data that had been overwritten numerous times, or on a hard drive that was blown to pieces in an explosion — is simply irrecoverable. There is no debate as to how much a producing party need do to attempt to restore it since, by definition, it cannot be restored. Typically, data is considered “inaccessible” because it costs a lot of money to restore it.
But, just as the comedian Steven Wright observed that everywhere is within “walking distance” if you have the time, all data (save the truly, technically inaccessible) is “accessible” if you have the money. The real question is whether the benefit of restoration is worth the cost.
Restoration of backup tapes is an expensive way to access data because new data does not accrete to them over time, meaning that dozens, if not hundreds, of tapes will have to be restored, and the data on them processed and de-duplicated, simply to produce a complete set of data pertaining to one user, one project, one set of facts at issue. Backup tapes make a snapshot of data on a server at a single moment in time.
That means that if a tape is made on Day 1 (and remember that several tapes may be needed to back up a single server and that multiple servers — an email server and a file server, for example — may be at issue), and another on Day 2 (the next day, a week later), there will be a great deal of overlap in the data on the two tapes, but Day 1′s tape will contain files that had been deleted by Day 2, and Day 2′s tape will contain new data not resident on Day 1. So, to create a complete set of data — for an email user, for example — both tapes would have to be restored, the email boxes extracted, the email processed and de-duplicated, and one “composite” email box produced. Make that 10 or 20 users instead of a single user, make it all tapes from Day 1 through Day 50, and put two servers into play instead of one, and it’s very easy to see that the cost of data collection through this method quickly grows into astronomical numbers, and only because of the cost of capturing the incremental differential between tapes. New technologies have reduced these costs, but they are still very high.
Massive tape restorations were far more prominent when e-discovery first emerged, because users and entities simply did not think of data storage as important. Emails that users deemed “worth saving” were printed to paper and placed in folders. Human nature being what it is, it was the rare occasion when a “smoking gun” electronic file of any type was so saved.
So, when Laura Zubulake was terminated and went first to the Equal Employment Opportunity Commission and then to federal court, it was no surprise that the discovery she sought, i.e., all emails by and between the key players at UBS Warburg regarding her and her performance there, was to be found only on backup tapes.
The court in Zubulake did not rush to require UBS Warburg to search its backup tapes. It considered the high cost, demanded that Zubulake proffer what she expected to find on the tapes and, initially, granted UBS Warburg’s motion to have Zubulake bear the cost of restoration.
It was only when the dual stories emerged, one from the contents of the tapes showing Zubulake’s ill treatment by UBS Warburg, the other revealing spoliation by UBS Warburg, that the court decided to sanction USB Warburg and accord Zubulake the relief, i.e., the “adverse inference instruction,” for which the series of Zubulake opinions is now remembered.
The “undue burden” language of Rule 26(b)(2)(B), then, refers truly to the high cost of tape restoration created because numerous tapes need to be restored, and the data on them de-duplicated, to aggregate the incremental differences between the tapes, and not to any technical burden. The presumption is that such a method of producing e-discovery is unduly burdensome, which usually results in a successful motion by the producing party to be relieved of the obligation to restore and search the tapes, to shift the cost of production to the requesting party, or at least to split it.
Only when the producing party is seen as having brought the cost on itself, usually because it has been far less than forthcoming in its discovery productions, do courts, that still see tape restoration as burdensome, tend not to see it as “unduly” so.
The Facts and the Decision
The FTC moved to enforce an administrative subpoena served upon Boehringer Ingelheim Pharmaceuticals Inc. (BIPI). The thrust of its argument was that BIPI “did not conduct an adequate search of its electronically stored information because it refused to search backup tapes in accordance with the subpoena’s demands.”
BIPI put into place a litigation hold by having counsel identify 66 potentially relevant custodians who, in turn, were to identify potentially responsive files and preserve them by placing them in a “hold file” to be retained until the hold was lifted.
The problem with this implementation of the litigation hold was that a feature in BIPI’s email system that auto-deleted emails “90 days after the email was sent or received” was not disabled. In addition, the FTC alleged that “some relevant emails were inappropriately deleted manually or otherwise lost.”
The FTC’s initial proposed solution was to have BIPI search all electronically stored information (ESI) going back to 2003; that request included servers, hard drives and backup tapes. Given the volume of tapes and other sources, as well as that there was no simple, inexpensive way to extract the responsive email and e-docs of the targeted custodians, BIPI estimated that to search the backup tapes alone would cost more than $25 million dollars, with the cost of searching hard drives and servers being additional to that.
Before the court in oral argument, however, the FTC stepped away from its demand that BIPI search its servers (it also, apparently, abandoned its demand that BIPI search the hard drives of the 66 custodians, but this issue is not addressed directly in the opinion). As well, the FTC appeared to narrow its focus to the “relevant period” of “February through August 2008.” Finally, while initially the FTC had asked that 24 tapes from that period be searched, its position at oral argument was “that a search of four tapes would be sufficient.”
The FTC asserted that, even given its relaxed demands, BIPI still had failed to “conduct an adequate search of its records.” BIPI, of course, disagreed, and argued that conducting the additional searches requested by the FTC remained “unnecessary and unduly burdensome.”
The threshold question the court had to answer was what the appropriate standard of review was for the FTC’s request. BIPI asserted that Rule 26(b)(2)(B), under which “a 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,” although the court could still order production if the requesting party made a showing of “good cause,” governed. The FTC, on the other hand, argued that “its request for a search of the backup tapes should be analyzed under the standards of FTC v. Texaco, 555 F.2d 862 (D.C. Cir. 1977).”
Under that standard, a much stronger showing of “undue burden” is required to excuse a party from honoring an administrative subpoena. Instead, a court should enforce the subpoena (assuming the agency had jurisdiction to issue the subpoena) if “the demand is not too indefinite and the information sought is reasonably relevant.” Only if the producing party met its burden of showing that compliance with the subpoena threatened “‘to unduly disrupt or seriously hinder normal operations of a business’” would the court sustain a challenge to the subpoena.
The court found the Texaco standard applicable; although, as we go through its application to the facts, it appears that it would have reached the same result regardless of the standard imposed. The court noted that the FTC’s original request “would have consumed much of BIPI’s time and money.” Exercising judicial restraint, the court held back from ruling on that original request because “the scope of the search” the FTC requested had been “narrowed significantly” and, as a result, “in its modified form,” was “not unduly burdensome for BIPI” under Texaco. The narrowed search did not threaten “to unduly disrupt or seriously hinder normal operations of a business” as per Texaco. BIPI bore the burden that it would have done so, and did not meet its burden.
The court, however, was not finished with its analysis or in formulating the proper relief. While BIPI’s objection may have been overruled, that was only because, in its representations, the FTC had narrowed its request: Now the devil was in details, and the FTC had to specify exactly what it was requesting for the court to see just how narrow the request had become.
Any “burden” in conducting a search, the court noted, “is a function of how efficiently it is conducted.” Thus, the parties needed to meet and confer “to determine the appropriate method of searching the relevant backup tapes to render the process as efficient as possible.” If they believed they needed assistance, the court would ask a mediator who had previously worked with the parties “to meet with them again to guide their discussions as to how to conduct the search.” Should those steps not lead to agreement, the court would again take up the matter.
Reacting to Tape Restoration Requests
Following Zubulake and the enactment of Rule 26(b)(2)(B), the immediate reaction to requests for tape restoration seemed to be “undue burden.” In the instant matter, however, Facciola, demonstrating the careful thought and common sense that has helped him earn his reputation as a leading thinker in e-discovery matters, showed that the issues with tape restoration are no different than with any others.
The parties in the instant matter acted, at least initially, as parties typically (and unfortunately) do at the commencement of e-discovery, matters, with the requesting side asking for everything and the producing side saying it will cost a zillion dollars. Once, however, the court forced the parties to talk, and to articulate the issues to the court in an informed manner, the scope of the request narrowed considerably and, as it did, the objection to the request as originally made disappeared. In other words, as I have said so often in these columns, smart, informed answers win.
It is, unfortunately, the fact that very often when backup tapes come into play, it is in that Zubulake-type situation where large numbers of tapes have to be restored, at great cost and with a remarkable amount of de-duplication being conducted in data processing, solely to make composite email boxes and file shares of just a few users. In such cases, the cost will be considerable and the issue of whether the juice is worth the squeeze will have to be confronted by the court.
But in many, many cases, the set of tapes to be restored and searched can be narrowed to reduce costs considerably. Hopefully, the effect of the decision here will be to make parties and courts approach tape issues as they would any other ESI issue, so that they look for the smart, informed position that narrows the request so that both sides get what they truly need. •
Leonard Deutchman is general counsel and administrative partner of LDiscovery, a firm with offices in New York City, Fort Washington, Pa., McLean, Va., Chicago, San Francisco and London that specializes in electronic digital discovery and digital forensics.