When should discovery come with a bill? Assessing cost shifting for electronic discovery.

• Author: Vainberg, Vlad

INTRODUCTION I. DEFINING THE PROBLEM: EARLY APPROACHES TO COST SHIFTING A. Electronic Discovery Expenses as a Cost of Doing Business B. The Rise of Electronic Discovery C. The Move Toward Cost Shifting II. COST-SHIFTING TESTS PRIOR TO THE 2006 E-DISCOVERY AMENDMENTS A. The Marginal Utility Test B. The Rowe Test and Discovery Protocol 1. The Eight-Factor Rowe Test 2. The Plaintiff-Friendly Rowe Discovery Protocol C. The Zubulake Test III. RULE 26(B)(2) AND THE 2006 E-DISCOVERY AMENDMENTS A. Rule 26(b)(2) Prior to the 2006 Amendments B. The 2006 Electronic Discovery Amendments 1. The Two-Tiered System 2. The Advisory Note to Rule 26(b)(2) 3. Trends from Practice IV. REACHING EQUILIBRIUM: WHAT IS THE OPTIMAL COST-SHIFTING PARADIGM? CONCLUSION INTRODUCTION

The plaintiff, a whistleblower suing his former employer for improperly eliminating his position, requests all documents, notes, memoranda, e-mails, and metadata (1) related to organizational restructuring from the employer's external hard drives. (2) The defendant is a relatively poor rural county that moves for a protection order, arguing that the requests are overbroad and would cost the county approximately $49,000 to produce, not including attorney review time. (3) The plaintiffs potential recovery in the case is estimated to be "significantly less" than $100,000. (4) How should a judge rule on the request?

The explosion of costly electronic discovery in the mid-1990s made this type of problem commonplace for district and magistrate judges, who in turn began exercising their authority under Federal Rule of Civil Procedure 26(c) to allocate some expenses to the requesting parties. (5) When a majority of the Supreme Court recently cited "sprawling, costly, and hugely time-consuming" discovery as a reason for its recognition of a heightened civil pleading standard in Bell Atlantic Corp. v. Twombly, (6) Justice Stevens retorted that Rule 26(c), among others, supplied a better tool for managing pretrial costs. (7) Providing an exception to the traditional discovery presumption that each party to a lawsuit bears its own discovery costs, (8) Rule 26(c) permits the district court to shift costs onto the party requesting discovery upon a finding of "good cause." (9) But what constitutes "good cause," and how should a court determine the appropriate amount of cost shifting? With little guidance from the Rules themselves (10) or the courts of appeal, (11) lower courts initially developed several analytical frameworks for analyzing the problem. These approaches can be roughly grouped into four categories: (1) the "marginal utility" test promulgated in McPeek v. Ashcroft, (12) (2) the Rowe test, (13) (3) the Zubulake test, (14) and, following the 2006 electronic discovery amendments to the Rules, (4) the application to cost shifting of seven factors outlined in the Advisory Committee's Note to Rule 26(b)(2), which were actually intended to guide the threshold question of whether certain discovery should be produced in the first place. (15)

This Comment analyzes the benefits and disadvantages of each cost-shifting approach in the context of electronic discovery (e-discovery). (16) It examines civil cases in which the court considered ordering the requesting party to bear some or all of the expenses of the responding party's technical search, restoration, and production of electronically stored information (ESI). (17) The Comment's scope is limited to cost-shifting disputes between parties to a lawsuit. (18) It also does not analyze cost-shifting orders meant to serve as a sanction for discovery violations.

Part I explores how the rise of costly electronic discovery in the 1990s led judges to consider cost shifting without developing robust analytical tests.

Part II traces the development of multifactor tests as a more sophisticated tool to handle expensive discovery requests. Although these tests were more systematic than the earlier approaches, they contained flaws. The marginal utility test in practice largely ignored the economic costs of each particular production. (19) The Rowe test employed a mechanical factor-counting approach that led to liberalized cost shifting in every reported case where it was applied. (20) It also accompanied a plaintiff-friendly discovery protocol, which seemed to authorize intrusive "fishing expeditions" so long as they were financed by the requesting parties. (21) The weighted-factor Zubulake test provided sound analytical underpinnings, but in practice it resulted in somewhat divergent decisions hinging on the least important factors. (22)

Part III evaluates the 2006 amendments to Rule 26(b)(2), which provided a multifactor test for the production of inaccessible data that many courts apply in determinations of cost shifting. This Part's qualitative analysis is supplemented with some broader observations from a survey of sixty-five published federal cases discussing cost shifting. (23) A significant majority of these cases involve individual plaintiffs requesting information preserved by corporate defendants on complex electronic networks and storage tapes. (24) This pattern is unsurprising. Judicial intervention into cost shifting is most necessary when there is a structural imbalance in the amount of discovery each party must produce, such that the requesting party has little incentive to negotiate mutual limits. (25) Because of this structural characteristic, the terms "requesting parties" and "plaintiffs" are used interchangeably throughout this Comment unless otherwise noted in the discussion of particular cases. (26)

My survey reveals that although courts have not uniformly applied Rule 26(b)(2), there appears to have been a decline in cost-shifting orders following the 2006 e-discovery amendments. (27) Although the sample of cases in the survey may not necessarily be representative of all cost-shifting opinions, (28) I posit that cost shifting is likely rarer now because the amended Rules make reasonably inaccessible data presumptively undiscoverable (29) and also emphasize negotiation among parties, limiting the need for judicial intervention. (30)

Part IV recognizes two troubling trends in cost-shifting cases: (1) the tendency of some courts to liberally shift costs in lieu of denying meritless discovery and (2) the possibility that wealthier parties' greater willingness to pay provides them with significantly upgraded access to discovery over poorer parties. It concludes that as discovery costs continue to spiral upwards, the optimal discovery paradigm would resort to cost shifting only when informational uncertainty makes the likelihood of uncovering critical information a very close call. The optimal approach would involve storage-tape sampling to determine the likelihood of uncovering relevant data, (31) followed by a combination of the two-tiered discovery structure in amended Rule 26(b)(2), as well as the factors in the Zubulake test, to guide judges in determining when ordering both discovery and cost shifting is appropriate.

1. DEFINING THE PROBLEM: EARLY APPROACHES TO COST SHIFTING

   In the mid-1990s, the traditional American paradigm of forcing each party to bear its own costs was undermined by one-sided and tremendous expenses associated with electronic discovery. As e-discovery became more common, the view that a producing party must automatically bear its associated expense as a cost of doing business became as outmoded as Commodore 64 computers. With individual plaintiffs able to coerce corporate defendants into settling because of the high cost of discovery, courts began developing fact-intensive balancing tests conditioning discovery on the requesting party's ability to pay for it.

   1. Electronic Discovery Expenses as a Cost of Doing Business

      When the Supreme Court reaffirmed the presumption that "the responding party must bear the expense of complying with discovery requests" in 1978, (32) complex discovery typically entailed scores of young associates reviewing boxes of documents in corporate warehouses. (33) Respondents who made their paper records available for inspection were able to limit plaintiffs' fishing expeditions to the extent of the plaintiffs' available manpower. (34) Courts typically refused to shift costs for expensive productions by stating that the defendants should have foreseen the cost when they chose to use expensive storage mechanisms: the so-called "cost of doing business" argument. (35) For example, in Delozier v. First National Bank of Gatlinburg, the defendant was ordered to pay for the photocopying of its records from microfilm because the defendant elected to save its records in that form. (36) In Daewoo Electronics Co. v. United States, the United States Court of International Trade held that "[t]he normal and reasonable translation of electronic data into a form usable by the discovering party should be the ordinary and foreseeable burden of a respondent in the absence of a showing of extraordinary hardship." (37) For the most part, courts did not even consider allocating costs.

      Even when courts analyzed cost shifting prior to the mid-1990s, the result generally remained the same. One of the earliest cases to consider cost shifting was Bills v. Kennecott Corp., an age discrimination action in which plaintiffs requested that their former employer provide printed computer records. (38) The defendant printed the data and moved the court for reimbursement of the $5411 cost. (39) Recognizing that the advisory notes to Rule 26(c) provided "no guidance" on determining what type of discoverable computer-stored information constitutes an undue burden, (40) the court set forth four relevant factors: (1) the total cost of production; (2) "the relative expense and burden" to each party in obtaining the data; (3) whether the requesting party would be substantially burdened by the expense; and (4) whether the responding party would benefit in any way from producting the data. (41)...