This Sidley Update addresses the following recent developments and court decisions involving e-discovery issues:

a District of Columbia district court decision granting a motion to compel production of documents belonging to plaintiffs’ CEO where defendants showed the CEO possessed unique, relevant records and the request was proportional to the needs of the case a Middle District of Florida opinion vacating a magistrate judge’s order after finding its requirement of broad class discovery from defendants was not proportional to the needs of the case a Southern District of New York opinion denying defendant’s motion for spoliation sanctions where defendant failed to demonstrate that he was prejudiced by plaintiff’s failure to preserve emails and other records related to his employment a Northern District of California decision rejecting a motion to compel Twitter, Inc. to search and produce direct messages sent and received on Twitter by its employees because the messages were protected under the Stored Communications Act

1. In Oxbow Carbon & Minerals LLC v. Union Pacific Railroad Co., 322 F.R.D. 1 (D.D.C. 2017), Magistrate Judge G. Michael Harvey granted a motion to compel production of documents belonging to plaintiffs’ CEO where defendants showed the CEO possessed unique, relevant records and the request was proportional to the needs of the case.

In this antitrust action, plaintiffs, a group of related companies that mined and sold coal and petroleum coke, alleged that defendants, two railroad companies, had conspired to engage in anticompetitive conduct that caused plaintiffs to pay higher prices to ship their products. Id. at 3.

Defendants moved to compel plaintiffs to include the CEO and founder of the related mining companies as a document custodian whose records would be searched for material responsive to discovery requests. Id. at 4. Plaintiffs had refused to include the CEO as a custodian on the grounds that production of his documents would be disproportionately burdensome and duplicative of documents produced from 19 other custodians. Plaintiffs also argued that the CEO’s documents would be marginally responsive overall because he oversaw a conglomerate of companies, only some of which were involved in mining and shipping coal and petroleum coke. Plaintiffs initially estimated that including the CEO as a custodian would result in approximately 130 gigabytes of additional data to be filtered through the agreed-upon search terms, a process that would cost an estimated $250,000. In the alternative, plaintiffs requested that the costs be shifted to defendants.

At an initial hearing on the motion to compel, plaintiffs agreed to analyze a sample of the CEO’s documents using the preexisting search terms to provide a more concrete estimate of the responsiveness and burden of the request. Plaintiffs collected 467,614 documents from the CEO’s files. After removing duplicates and applying search terms, plaintiffs identified 45,639 document hits or 82,600 documents when including family members. Plaintiffs reviewed a random sample of 10 percent of the hits and determined that 1,300 of the 12,074 reviewed documents were responsive to the discovery requests. Id. at 4-5.

The processing and review of the sample cost approximately $57,000, and plaintiffs estimated that it would cost another $85,000 to process and review the remaining documents. This total cost of $142,000 was less than the original estimate of $250,000 because plaintiffs had anticipated reviewing 214,000 document hits when, in reality, the search terms produced only 82,600 documents including family members. Following the production of the responsive documents from the sample set, plaintiffs continued to oppose including the CEO as a custodian, refused to renegotiate the agreed-upon search terms and refused to provide defendants with data from the sampling necessary to evaluate the effectiveness of particular search terms.

Turning to the motion to compel and applying the proportionality factors listed in Rule 26(b)(1), Magistrate Judge Harvey concluded that including the CEO as a custodian was proportional to the needs of the case and granted the motion.

The importance of the issues at stake: Both parties agreed that the case involved serious allegations and had the potential to “broadly impact a wide range of third-parties.” Id. at 7. Accordingly, Magistrate Judge Harvey concluded that this factor weighed in favor of granting the request where it would produce documents relevant to the resolution of the case. The amount in controversy: Magistrate Judge Harvey found that the amount in controversy weighed in favor of granting the request because the proposed discovery would cost $140,000 in a case where plaintiffs were seeking to recover an estimated $150 million in damages. Id. at 7-8. The parties’ relative access to the relevant information: Magistrate Judge Harvey concluded that this factor weighed in favor of granting the request because there was no dispute that plaintiffs were in possession of the CEO’s relevant documents and that defendants had no other means of accessing the documents. Id. at 8. The parties’ resources: Magistrate Judge Harvey noted that this factor weighed in favor of granting the request because defendants did not object to the request on the grounds of an inability to pay. The importance of the discovery in resolving the issues: Magistrate Judge Harvey noted that plaintiffs had conceded that the CEO’s files contained unique, relevant documents. Moreover, after reviewing the production of the sample set documents, defendants argued that the CEO was a distinctive source of information regarding plaintiffs’ finances and business strategy. Magistrate Judge Harvey agreed that these issues would be critical to deciding whether the claimed damages were attributable to defendants’ conduct. Accordingly, this factor weighed in favor of granting the request. Whether the burden or expense of the proposed discovery outweighs its likely benefit: Plaintiffs relied heavily on this factor, but Magistrate Judge Harvey agreed with defendants that the burden of including the CEO as a custodian did not outweigh the benefits of the additional discovery. Id. at 9. The additional $85,000 cost to review and produce the CEO’s remaining documents was small compared with the $1.3 million that plaintiffs had already spent on discovery and even smaller compared with the tens of millions of dollars of claimed damages. Plaintiffs argued that the high rate of “false positives” weighed against discovery, but Magistrate Judge Harvey noted that the responsiveness rate for the CEO’s documents was not even the lowest among all custodians.

Finally, Magistrate Judge Harvey denied plaintiffs’ request to shift the costs of discovery on this issue to the defendants. While the “presumption under the Federal Rules of Civil Procedure is that the producing party bears the costs of complying with a discovery request,” Magistrate Judge Harvey noted that “the Court may shift a portion of the costs to the requesting party in the event that the discovery request would unduly burden the producing party.” Id. at 10 (quoting D’Onofrio v. SFX Sports Group, Inc., 254 F.R.D. 129, 134 (D.D.C. 2008)). In considering whether to shift costs, courts consider factors that are “essentially identical” to the proportionality factors under Rule 26(b)(1). Id. at 11. Applying these factors, Magistrate Judge Harvey concluded that the discovery request did not impose an undue burden or expense that warranted a reallocation of expenses.

2. In Nece v. Quicken Loans, Inc., 2018 WL 1072052 (M.D. Fla. Feb. 27, 2018), Judge Steven D. Merryday found that a magistrate judge’s order requiring broad class discovery from defendants was not proportional to the needs of the case and vacated the magistrate judge’s order.

In this putative class action, defendant solicited plaintiff 12 times by telephone after plaintiff inquired about obtaining a mortgage from defendant. Id. at *1. Plaintiff alleged that these calls violated the Telephone Consumer Protection Act (TCPA) because plaintiff’s phone number was listed on the national do-not-call registry. Judge Merryday denied defendant’s motion for summary judgment, finding that there were factual disputes regarding whether plaintiff revoked her consent to receive the telephone calls and whether defendant stopped calling plaintiff within a reasonable time.

As part of her discovery requests, plaintiff requested that defendant produce all documents and consumer communications relating to every do-not-call request that defendant received from a potential class member. Id. at *2. Defendant objected, arguing that the request was overly burdensome and irrelevant. In part because plaintiff did not limit the request to a particular timeframe, defendant estimated that responding to the request would cost millions of dollars and require collecting and reviewing at least 3 million emails. Plaintiff moved to compel, and the magistrate judge granted the motion in part, ordering that defendant produce records documenting do-not-call requests made by “similarly situated residential consumers” between September 2012 and June 2013.

Defendant then moved for clarification or reconsideration, again arguing that it would require “dozens of employees to spend months on document review” and cost “at least hundreds of thousands of dollars” to comply with the discovery request. As defendant explained, the identification of “similarly situated residential consumers” would require manually searching the company’s general email inbox and other locations to determine who revoked consent to phone calls. The “infinite variety of language” precluded using keyword searches to complete this step. Defendant would also need to obtain information from third-party telephone carriers to determine whether telephone numbers belonged to a residence. Defendant also asked for clarification because plaintiff interpreted the magistrate judge’s order to encompass all documents related to the do-not-call requests and not just the requests themselves. The magistrate judge denied the motion with little explanation.

Defendant next lodged objections with Judge Merryday to the magistrate judge’s discovery order and the denial of the motion for clarification. Id. at *3. By the time of the objections, defendant had already produced 12,000 pages of records relating to do-not-call requests from 450,000 phone numbers and individuals. Defendant asserted that full compliance with the discovery order as interpreted by plaintiff might require an additional 15,000 hours to complete.

Judge Merryday concluded that the plaintiffs’ discovery requests imposed on defendant a burden disproportional to the needs of the case. Judge Merryday noted that plaintiff had “avoided confronting the reality” that class certification was unlikely because “individualized issues often predominate in putative TCPA class actions.” In addition, defendant had already provided substantial discovery about the proposed class. Finally, vacating the magistrate judge’s order would obviate the need for a lengthy extension of discovery. For these reasons, Judge Merryday vacated the magistrate judge’s order and set a briefing schedule on a motion for class certification.

3. In International Business Machines Corp. v. Naganayagam, 2017 WL 5633165 (S.D.N.Y. Nov. 21, 2017), Judge Nelson S. Roman denied defendant’s motion for spoliation sanctions where defendant failed to demonstrate that he was prejudiced by plaintiff’s failure to preserve emails and other records related to his employment.

In this breach of contract action, defendant had worked for plaintiff and, after voluntarily resigning, went to work for a competitor. Id. at *2. Plaintiff filed suit to rescind unvested stock awards pursuant to the terms of the incentive compensation plan, which allowed plaintiff to cancel and rescind such awards if an employee went to work for a competitor. During defendant’s deposition, defendant testified that he worked on “strategic business paper[s]” for plaintiff that omitted any mention of his new employer as a direct competitor. Id. at *3. Defendant’s former supervisors also testified at deposition that they had not been asked to retain emails sent or received about defendant’s voluntary resignation.

Following these depositions, defendant moved to compel production of the strategic business plans and emails related to defendant’s departure. The magistrate judge overseeing discovery denied the request to produce the emails related to defendant’s departure because defendant failed to demonstrate the relevance of these files, but he ordered plaintiff to produce any strategic business plans that listed its competitors. In response, plaintiff informed the magistrate judge that it was unable to locate the strategic plans defendant described in his deposition and that neither of defendant’s supervisors had any recollection of such documents.

Plaintiff then moved for summary judgment on its breach-of-contact claims, and defendant cross-moved for spoliation sanctions. Id. at *4. Defendant asked that Judge Roman issue an adverse inference and impose other sanctions against plaintiff for failing to preserve electronically stored information (ESI), including the emails relating to defendant’s departure and the strategic business plans.

Beginning with the adverse inference instructions, Judge Roman denied defendant’s request because defendant merely alleged that plaintiff acted negligently in failing to preserve relevant emails and documents. Judge Roman noted that the Advisory Committee Notes to the December 2015 amendments to Rule 37(e) had rejected prior Second Circuit precedent authorizing such instructions on a finding of negligence or gross negligence. Following the amendments, the court may not issue an adverse inference instruction absent a finding that the party “acted with the intent to deprive another party of the information’s use in the litigation.” Id. at *5 (quoting Fed. R. Civ. P. 37(e)(2)). Even though the new standard under Rule 37(e) did not go into effect until after plaintiff filed suit, Judge Roman decided to apply Rule 37(e) retroactively where retroactive application would not be unjust or impracticable. In particular, Judge Roman noted that the issue of spoliation did not arise until after the new Rule 37(e) went into effect.

Judge Roman also denied defendant’s request for less severe spoliation sanctions under Rule 37(e). Rule 37(e) permits such sanctions where ESI is lost because a party failed to take reasonable steps to preserve it, and it cannot be restored or replaced through additional discovery. These sanctions require a finding of “prejudice to another party from the loss of the information.” Fed. R. Civ. P. 37(e)(1).

Judge Roman concluded that defendant failed to meet the prejudice requirement under Rule 37(e). Id. at *6. Even though the content of the allegedly spoliated ESI was fairly evident, defendant did not explain how he was prejudiced by its alleged loss. There was no suggestion in the depositions that the spoliated emails included discussions of whether plaintiff and defendant’s new employer were competitors. Moreover, the spoliation of the strategic business plans did not prejudice defendant because these documents were limited to particular markets, and the issue in the litigation was whether defendant’s new employer competed with plaintiff in any market. Id. at *7.

Finally, Judge Roman granted plaintiff’s motion for summary judgment, concluding that there were no material issues of fact and that plaintiff had shown, as a matter of law, that it was entitled to rescind the stock awards under the terms of the incentive compensation plan. Id. at *7-8.

4. In Shenwick v. Twitter, Inc., 2018 WL 833085 (N.D. Ca. Feb. 7, 2018), Magistrate Judge Sallie Kim denied a motion to compel defendant Twitter, Inc. to search and produce direct messages sent and received on Twitter by its employees because the messages were protected under the Stored Communications Act.

Plaintiffs alleged that defendants, including Twitter’s former CEO and current executives, violated the Securities Exchange Act of 1934 by misleading investors, resulting in an artificially inflated stock price. Discovery in this case focused on whether the executives intentionally concealed a change in metrics to show greater growth. As part of discovery, plaintiffs moved to compel Twitter to produce direct messages sent within Twitter among their employees.

The Stored Communications Act (SCA), 18 U.S.C. § 2701 et seq., prohibits a third-party electronic communications service (ECS) from producing electronic communications of its users. Id. at *2. The plaintiffs claimed that the court could compel Twitter as a party to provide information within its possession, custody or control, but the magistrate judge stated that Twitter as an entity (and ECS) was separate from the individual custodians who used and stored messages with Twitter. The magistrate judge therefore considered whether Twitter should be compelled to produce its employees’ direct messages sent within Twitter. As to two named defendants in the case, the magistrate judge observed that they were parties to the action, and as parties, they had to provide (and had agreed to provide) their direct messages pursuant to Fed. R. Civ. P. 34 requests for production. The other Twitter employees, however, were not named defendants and thus did not qualify as parties in this action. The magistrate judge found that Twitter did not require its employees to use direct messages for work communications, and those Twitter employees who did so had privacy rights regarding their communications that were protected by the SCA. For these reasons, the magistrate judge held that the court could not compel Twitter to produce its employees’ direct messages even though Twitter was the provider of the direct messaging service.