Document review often is the most expensive component of discovery in large, complex cases. Wouldn’t it be great if you could shift that cost to the party that requested the documents, along with the burden of performing the tedious, time-consuming review? Well, maybe you can. A federal magistrate judge in the N.D. Florida recently did exactly that.

In FDIC v. Brudnicki, No. 5:12-cv-00398, 2013 WL 2948098 (June 14, 2013) , the FDIC, as receiver for a bank, sued eight of the bank’s former directors, including one officer. The defendants moved to compel documents from the FDIC and sought sanctions against the FDIC “for delaying discovery.” Id. at *3. The FDIC had “agreed to produce responsive documents under a proposed protocol.” Id. at *4. The court observed, however: “The parties sharply disagree on the method of production and the ESI protocol.” Id.


The FDIC proposed a two-phased approach for its document production. The first phase wasn’t really in dispute. The FDIC had delayed that production only because the court had not yet entered “an appropriate confidentiality and non-disclosure order.” Id. The court entered such an order contemporaneously with its order on the defendants’ motion to compel, which resolved the sole impediment to the FDIC’s first phase production.

The FDIC’s proposed second phase was the crux of the dispute among the parties. The FDIC proposed the following for the second phase:

[T]he parties would collaborate in the process of culling relevant documents from the 2.053 terabytes of ESI on the DMS database by using agreed upon search terms and the names of key Bank employees. After relevant information has been identified the information will be uploaded to a database known as “Relativity.” The Defendants will then be given private and secure access to Relativity through the internet and be able to search, review and annotate all ESI on Relativity. Defendants will then be able to identify and request any document they want produced. The FDIC–R will then produce the requested documents subject to a privilege review and clawback agreement. The FDIC–R proposes to charge $.06 per page for any document Defendants identify from the Relativity database and then request the FDIC–R to produce. Additionally, the FDIC–R proposes to charge the Defendants $225 per gigabyte of data that is uploaded to the Relativity database. Id. at 4.

In other words, the FDIC proposed that for its two terabytes of ESI, comprising the vast majority of its document collection, the defendants themselves would review the collection in Relativity® after the application of agreed-upon search terms to cull the documents. And, the defendants would pay per-gigabyte processing and per-page production fees. (The FDIC apparently stopped short of requiring the defendants to collect the documents themselves.)

The FDIC reportedly has used the same protocol “in at least a dozen similar lawsuits brought by the FDIC–R against former officers and directors of failed banks.” Id. Evidently it’s good to be the FDIC—at least with respect to cost- and burden-shifting for large-scale document reviews.

What was the FDIC’s justification for shifting the cost and burden of the review to the defendants? “The FDIC–R suggests that using their proposed ESI protocol will expedite discovery, provide the Defendants with better access to the documents and will greatly reduce the enormous costs it would incur if it was required to utilize traditional methods of collection and review of ESI.” Id. (emphasis added). At bottom, the FDIC’s proposal would reduce the FDIC’s costs were it to have to conduct the document review itself. And how were those costs saving to be achieved? Not by eliminating the costs. Rather, by shifting the costs to the defendants who themselves would conduct the review.

Not surprisingly, the defendants objected “that the ESI protocol is an inappropriate cost shifting and ‘responsibility shifting’ scheme by the FDIC-R.” Id. at *5. The defendants also objected that “under the protocol they will be required to provide appropriate search terms in the first instance, and not the FDIC–R, thus placing on them the burden of locating and identifying relevant documents.” Id.

The court was largely unmoved by the defendants’ objections, concluding that, “for the most part, the ESI protocol proposed by the FDIC–R is appropriate, reasonable and consistent with principles of proportionality when dealing with ESI.” Id.

Regarding cost-shifting, the court observed that while some “courts have required a showing of inaccessibility for cost shifting, other courts have held that Rule 26(c) permits cost shifting as part of enforcing proportionality limits.” Id. The court offered two primary reasons for imposing the review and production costs on the defendants.

First, the court noted that this wasn’t a situation involving costly retrieval and searching of difficult-to-obtain data (e.g., backup tapes). Rather, the FDIC already had collected, processed, and loaded the data to a database. The court found that the costs the defendants would bear under the FDIC’s proposal were “nominal,” “pale[] in comparison to the hundreds of thousands of dollars that are typically incurred in collecting and retrieving ESI,” and were costs “which Defendants would have to pay if the documents were traditional paper documents and not ESI.” Id. The court also noted the countervailing benefits of the FDIC’s proposal “mak[ing] discovery less expensive, less burdensome and more expeditious.” Id.

Second, the court concluded that the application of the proportionality factors under Fed. R. Civ. P. 26(b)(2)(C) “strongly supports the ESI protocol proposed by the FDIC–R.” Id. at *6. The court observed that discovery requests subject to the ESI protocol were “far reaching, broad and do not call for the production of specific targeted documents.” Id. The court found that the documents to be reviewed and produced in the second phase of the FDIC’s proposed protocol “consist of largely marginal documents from a relevancy perspective.” Id. at *4. The court emphasized: “Notably, the phase I production constitutes the vast majority of the relevant documents in this case.” Id. Lastly, the court concluded that the costs the defendants would bear under the FDIC’s proposed ESI protocol constituted “an expense that these Defendants should be able to bear without impacting their ability to mount a vigorous defense to the claims in this case.” Id. at *6.

Regarding the defendants’ objection to having to propose search terms in the first instance, the court was unpersuaded. The court emphasized that the defendants were themselves former directors, including an officer, of the bank that was the subject of the litigation and thus “stand in the best position to know the names, titles and other information as to the custodians, the names of files and how the Bank’s documents are stored.” Id. at *6 n.12. The court also emphasized that the proposed protocol for developing search terms “should be a mutual collaboration and thus the Court expects the parties to work together, notwithstanding which side first proposes the search terms.” Id.

So does this decision open the door for parties to routinely shift document reviews and productions, and associated expenses, to opposing parties? Doubtful. It’s important to recognize the rather unique facts of this case. Here, the FDIC was acting as receiver for a bank of which the defendants were former directors and an officer. Why is this important? Two reasons—one the court noted and another it did not.

First, as the court observed, unlike in many (or perhaps most) cases, the opposing parties were not strangers to the responding party’s documents and information systems. As former directors and an officer—and given the FDIC was acting as receiver—the defendants here very likely were far more familiar with the bank’s documents, potential custodians, and information systems than the FDIC. So it very likely was more efficient for them to propose search terms and conduct the review themselves for the documents they wanted. This of course would not be the case where the opposing party has no familiarity with the responding party’s documents and information systems, and where the responding party is therefore in a far better position to identify responsive documents.

Second, the FDIC’s proposal for the second phase—loading a large number of documents to a review database without any preliminary review and giving the opposing parties direct access to search and review those documents—ordinarily would implicate serious privilege concerns. Sure, the FDIC ensured it could still assert privilege over any documents the defendants request for production and “claw back” those documents. But once reviewed by the defendants, the information in those documents is known to the defendants and likely can be used by them even without having the documents containing that information. Here, however, because the defendants were former directors and an officer of the bank whose documents were being reviewed, the FDIC likely wasn’t especially concerned about disclosure of privileged documents to the defendants, as that information likely already was known to the defendants. This of course would not be the case where the opposing party was never privy to privileged information before the litigation.

And it’s also important to keep in mind the court’s finding that the documents to be reviewed by the defendants were only marginally relevant, and that the FDIC was promptly producing the key documents to the defendants at its own expense. Though the court doesn’t say so outright, the court’s decision likely was motivated at least in part by a sense that as the parties requiring review and production of marginally relevant documents, the defendants should bear the burden and expense of that review and production.

This isn’t to suggest that the protocol approved by the court here is available only to the FDIC. There may be matters with sufficiently similar facts to justify the same approach—such as fiduciary duty litigation against corporate officers or directors, or employment litigation against former employees. At the very least, this decision suggests that litigants shouldn’t just assume that they always have to review their own ESI and bear the costs of that review and any resulting production. We know there are some circumstances where a court will impose that burden and expense on the other side—and at least for the FDIC, that reportedly is not uncommon.