The E-Discovery Field Guide


The E-Discovery Field Guide

Adam R. Prescott, Jack Woodcock

Defendant’s Reckless Privilege Review Results in Denial of Clawback Request

Implementing appropriate review procedures for ESI is important not only to ensure that privileged documents are correctly withheld, but also to provide a defense for the return or “clawback” of inadvertently produced material later.  

The Case 

In Irth Solutions LLC v. Windstream Communications LLC, No. 2:16-CV-219, 2017 WL 3276021 (S.D. Ohio Aug. 2, 2017), Irth and Windstream entered a clawback agreement for the inadvertent production of privileged information, but which failed to establish a standard of care for preserving the clawback right. During discovery, Windstream produced 2,200 pages of documents, including 43 privileged documents. Six weeks later, Windstream produced the same 43 privileged documents again. When Irth refused to return the documents, Windstream filed a motion to compel.

The Court’s analysis turned on Irth’s conduct in screening for privileged documents and the reconciliation of the parties’ clawback agreement, which lacked a standard of care, with the requirements of Federal Rule Evidence 502(b), which provides that inadvertent disclosure of privileged information is not a waiver when “the holder of the privilege or protection took reasonable steps to prevent disclosure.”

On the latter question, the court surveyed how courts handle the interaction between private clawback agreements and Rule 502(b). The court rejected as inconsistent with Rule 502(b) the “no-fault” approach of some courts in which a private clawback agreement without a standard of care is strictly enforced and the return of privileged documents always is required. The court looked more favorably upon two alternative approaches followed by some courts: either (i) an inadvertent disclosure is not a waiver unless the production was “completely reckless,” or (ii) Rule 502(b)’s reasonableness standard applies as the default when the parties fail to define a different standard.

Ultimately, the court did not choose between the two approaches because Windstream’s conduct fell short under either test, including because counsel was unaware an employee of the defendant was in-house counsel, the same privileged documents were produced twice, numerous obvious privileged documents were produced, and the number of produced privileged documents was high relative to the total production size. As the opinion made clear, the court was skeptical that any privilege review was done, let alone one sufficient to prevent waiver.


Under Rule 502, litigants have a number of options to protect privileged information in discovery; they must choose wisely. Rule 502(b) applies by default for inadvertent disclosures of privileged or otherwise protected information. But parties may enter into different protections under Rule 502(e), or they may request that the court enter an order under Rule 502(d). If a separate agreement is entered into, the clawback agreement should track the issues in Rule 502(b), including setting a standard of care for avoiding a privilege waiver. If the parties want protections against waiver, as Irth shows, they may be safer not to seek it under Rule 502(e), but through a court order under Rule 502(d). See Zival v. AT&T Mobility, LLC, No. 08-cv-10310, 2010 WL 5065963 (S.D.N.Y. Dec. 6, 2010). In all situations, parties should take care to craft a reasonable review plan for ensuring that privileged material is identified and withheld, which often includes manually reviewing documents and using targeted searches. Diligence on the front-end can ensure that sensitive privileged documents are never produced, but also that such documents can be successfully clawed back later if some slip through the cracks.

ESI De-duplication: An Important, but Challenging, Early E-Discovery Question 

“De-duplication” refers to the process of comparing “digital fingerprints” of ESI and removing “duplicate” ESI from a dataset. Whether to de-duplicate ESI is an early question that arises during the ingestion and processing phase of e-discovery because multiple versions of e-mails, documents, or other ESI will exist within a company’s files. For example, a Word document may be saved in multiple places, and that document also may have been sent to, and saved by, other employees, resulting in more potential duplicates. The same situation arises with e-mails: if an e-mail is sent to a dozen individuals, each will have a copy of the e-mail in his or her inbox, along with copies of any attachments to the e-mail.

De-duplication is intended to remove identical copies of ESI from a dataset, with the goal of managing the size of the ESI collection and eliminating the expense of reviewing and producing identical files. Another benefit is that the process reduces the likelihood that identical documents are treated differently in discovery—such as one copy being redacted for privilege while another is produced without redactions due to human error. A risk of de-duplication, on the other hand, is that the responding party is culling the dataset to remove potentially responsive information that was maintained by the company in the ordinary course of business, and it is possible that relevant ESI or metadata will be lost in the process.

The case

Although technical e-discovery topics like de-duplication do not often find their way into court opinions, the issue did arise in Theidon v. Harvard University, 314 F.R.D. 333 (D. Mass. 2016). In that case, the defendant, Harvard University, proposed using a de-duplication process on its ESI in reliance on Federal Rule of Civil Procedure 26(b)(2)(C), which instructs a court to limit discovery if it determines that the discovery sought “is unreasonably cumulative or duplicative.” The plaintiff, a former Harvard professor, objected to that process on the basis that the metadata fields Harvard intended to use to identify duplicates were insufficient and would cause the withholding of documents that were not “complete” duplicates. In siding with Harvard, the court found that the only lost data from de-duplication the plaintiff had alleged was “a message flag indicating whether an email was read or unread,” which the plaintiff had not proven to be relevant. Harvard also offered to produce duplicates of documents as requested, which further eliminated any harm to the plaintiff. Thus, the benefits of de-duplication outweighed the minimal, if any, cost to the plaintiff in not receiving the metadata from all of the removed duplicates.


De-duplication is another e-discovery issue that should be discussed with experienced litigation counsel, as well as with a qualified e-discovery vendor that can provide additional technical guidance. If the decision is made to de-duplicate the production, there also are different options for how and when to go about it, and in the end, whether (and how) to de-duplicate documents is a case-specific and technical inquiry that a party should make after careful consideration and as part of its broader e-discovery strategy.

Court Strikes Down Broad Request for All Documents Previously Produced to Government Investigators without Subject-Matter Limitations 

On April 24, 2018, the U.S. District Court for the Northern District of California denied a motion to compel filed by a large group of Volkswagen franchise dealers that sought copies of documents that Bosch had produced to government investigators. The decision in the class action case In re Volkswagen “Clean Diesel” Marketing, Sales Practices, and Product Liability Litigation (MDL Case No. 2672), highlights the balancing act that courts perform in managing the scope of discovery when follow-on civil litigation overlaps with ongoing government investigations.

The Volkswagen dealers moved to compel Bosch to produce documents that Bosch had provided to state and federal agencies during their ongoing investigations into Volkswagen’s emission testing practices. Those government investigations overlapped to some degree with the claims in the class action litigation, but the scope of the investigations—and thus the scope of the documents produced by Bosch to the investigators—was confidential and not known to the Court or the dealers.

On the dealers’ motion to compel, the Court highlighted the scope of permissible discovery under Federal Rule of Civil Procedure 26(b)(1)— “Parties may obtain discovery regarding any nonprivileged matter that is relevant to any party’s claim or defense and proportional to the needs of the case . . .”—and refused to guess whether all of the documents produced by Bosch to investigators were relevant to the civil litigation. Thus, although it may have been less burdensome for both the dealers and Bosch if Bosch simply duplicated its entire government productions to the plaintiffs (as opposed to the dealers drafting new requests, and Bosch objecting and reviewing the documents again for relevancy), the court ordered the dealers to issue new subject-matter-specific requests to ensure that “the scope of discovery is not expanded beyond what is allowed by the Federal Rules.”


It is tempting for parties to draft document requests as broadly as possible in the hope that something useful will turn up. This decision, however, provides another example of a court limiting discovery even when such limitations may actually result in more costs for both parties. Although the dealers ultimately may get all of the documents they want, they will have to work a little harder to get them. And, as this case shows, burden is only one of the factors governing the scope of discovery.

U.S. District Court for the District of ME Adopts Amendment to District Court’s Local Rules Effective July 1, 2018

The U.S. District Court for the District of ME recently announced that several amendments to the District Court’s Local Rules will take effect on July 1, 2018. Of particular relevance here, the Local Rules will be amended as follows:

Local Rule 26 – Discovery and New Court Form to Request a Hearing Re Discovery Dispute:

Local Rule 26 has been amended to require the party seeking a hearing on a discovery dispute to confer with opposing counsel and agree on the relevant discovery materials that should be submitted to the Court with a new electronically filed form Request for Hearing Re Discovery Dispute, which will be available on the Court’s website. The Rule has also been revised to clarify that discovery dispute hearings conducted by the Court will be on the record, but will not be officially transcribed except on request of counsel or the Court. 

Although not major changes, any rule that clarifies and streamlines the process for presenting discovery disputes to the court has the potential of saving parties time and money. The change also is a reminder that even when disputes arise, cooperation remains a key tenet of discovery. It will be interesting to see how the changes play out in practice.