Discovery Cost Shifting: Has Its Time Come?

by and

Vol. 99 No. 2 (2015) | The Mass-Tort MDL Vortex | Download PDF Version of Article
U.S. dollars in a jar

Since at least 1990, a few courts have shifted payment of discovery costs to the requesting party either under their general case- management authority or as part of a Rule 26 protective order. The Advisory Committee on Civil Rules rejected a proposal to provide explicit authority in the rules for cost shifting in 2000 in a closely divided vote. In 2006, Rule 26(b)(2) was amended to provide explicit authority to shift discovery costs if electronically stored information (ESI) was “not reasonably accessible.”

On Dec. 1, 2015, assuming Congress takes no action otherwise, Rule 26(c)(1)(B) will be amended to explicitly recognize a court’s authority to allocate expenses for disclosure or discovery. The Committee Note cautions that the recognition of this authority “does not imply that cost shifting should become a common practice. Courts and parties should continue to assume that a responding party ordinarily bears the cost of producing.”

Two experienced practitioners, John Vail, former vice president of the Center for Constitutional Litigation and proprietor of John Vail Law PLLC, and Alex Dahl, general counsel to Lawyers for Civil Justice and a shareholder at Brownstein Hyatt Farber Schreck, address this contentious issue.

Over the past two decades, have you witnessed a trend towards more or less discovery cost shifting in cases?

Vail: “Trend” would overstate the case. Certainly I notice keen interest among the cognoscenti of federal rulemaking — I confess to membership — fanned by representatives of American business interests.1 Flattered to be asked what I think, but being a footnote addict and recognizing that my own thoughts are highly subjective, I engaged in a crude empirical experiment that undoubtedly will make the fine researchers at the Federal Judicial Center cringe.

I ran a query in the Westlaw database of federal cases to see what had happened in the last two decades.2 It yielded 304 cases discussing cost shifting. In all the years prior to 1995, nine cases had discussed the issue.3 Curious about what was happening in state courts, I ran analogous searches4 in the database of state cases: four cases before 1995, 26 after. What’s it all mean?

In both state and federal courts the issue has received more attention in reported cases in the last two decades than in all preceding decades. But more is relative. In both sets of courts discussion occurred in a miniscule percentage of all cases. A brief review of the federal cases suggests that discussion occurs primarily about electronic documents and primarily in cases involving large business entities on both sides of the “v.” We know that those large cases are more common in federal than in state courts, a fact that could explain the much greater proportional attention in federal courts. We also know that, while discovery costs are modest in most federal cases, such cases are most likely to engender outsized disputes about discovery and outsized expenses.5

So, while I think more attention is being paid to the issue, I expect the attention says a lot more about a rarefied world of litigation between gargantuan entities than it does about litigation in general.

Dahl: Although there are notable exceptions, the judiciary remains largely closed-minded to the idea that the party asking for discovery should pay for some or all of its costs. This is interesting because we are talking about a very basic economic idea: the

person asking for a particular good or service is in the best position to decide what it’s worth. We accept this truth in almost every aspect of our daily lives without even thinking about it. By the age at which most children start receiving letter grades — around middle school or so — they understand that someone who wants a particular good or service must pay something for it — and that it’s up to the purchaser to decide whether the cost is worth paying. Around the world today, almost every transaction involves the consumer putting a value on a particular good or service by deciding whether it is worth the cost.

In the American legal system, however, we are entrenched in a practice of insulating parties from the costs of their requests. Instead, we rely on judges, who typically have less information than the parties, to determine whether the value of a discovery request to the requesting party is worth the burden it imposes on another, responding party (or, worse, on a nonparty). This system is not based on any thoughtful consideration of policy preferences, but rather is largely an accident of history.

The Federal Rules of Civil Procedure (FRCP) did not establish this practice and, in fact, the drafters of the original 1938 rules did not even consider this topic.6 There was no need because, at the time, business records were kept on paper, photocopying was not available, and the cost of providing discovery was minor. It is hard to imagine that, if the costs of discovery had been anything analogous to what they are today, the system would have developed in the same way.

The explosion in discovery costs over the last two decades — evidenced in part by the rise of a new multibillion dollar e-discovery industry — reveals the inherent flaw in the producer-pays system: that it fails to provide a meaningful mechanism for cost/benefit considerations. Our liberal discovery rules focus on what information parties are entitled to rather than whether the information is meaningful enough to the claims and defenses to justify the burden of obtaining it.

One result of our system is that liability costs in the United States are 2.6 times greater than the average level of the Eurozone countries.7 Nevertheless, most legal scholars, judges, and lawyers refuse to consider how profoundly helpful it would be to know how a requester values her own requests. The reflexive resistance to requester-pays rules makes raising the topic feel almost Copernican, as if the discussion itself violates a tenet that cannot be questioned. A typical first response is an accusation of trying to shut the courthouse door, which of course is not true and ignores the fact that a requester-pays system could and certainly should include safeguards to ensure access to justice.

Some legal experts who are willing to contemplate the topic of requester-pays rules think of them only as remedies for egregious discovery abuse, which they see as relatively rare occurrences that are properly policed by ethics rules. Intentional “impositional discovery” is a serious problem that exists because our producer-pays system invites (and rewards) overbroad discovery requests meant to impose economic pressure to force parties to compromise claims and defenses for reasons other than the merits. A default requestor-pays rule would end this kind of abuse, which is a sufficient reason to enact one. But the need for a requester-pays default goes well beyond its ability to remedy the worst behavior.

Determining whether discovery is worth the burden is an important calculation in every case. Even when discovery is not intentionally abusive, if it drives up the cost of litigation without producing information important to resolving the case on the merits, it injures the parties and our judicial system as a whole. The high cost of litigation deters people from asserting meritorious claims and forces parties to compromise meritorious defenses.

The FRCP should not encourage economically perverse litigation tactics, but rather should incentivize pursuit of discovery at the lowest cost and in the least burdensome manner possible to obtain the evidence necessary for a merits-based resolution by the fact finder. This is not just a matter of economics, it is fundamental to whether the judicial system can deliver on Rule 1’s guiding principle of just, speedy, and inexpensive determination of civil actions.

Under the rules going into effect this December, a judge must find a discovery request to be “proportional to the needs of the case,” taking into account six specific factors. In a case when the judge finds the request for admittedly relevant information not to be proportional, are there any drawbacks in offering an opportunity to obtain it but only if the requesting party pays for it?

Vail: In the commentary to the proposed rules there was hot debate about whether this proposed amendment changed anything about entitlement to discovery; the committee said no. If the committee is right, the presumption of Oppenheimer Fund, Inc. v. Sanders should be unaffected and a producing party should remain under a duty to rebut the presumption.8

Critics of the proposed rule contended that the amended rule would newly make proportionality a criterion of entitlement to discovery. That could narrow the applicability of the presumption of Oppenheimer and justify a judge’s conditioning entitlement on cost shifting. I think I am about to see my friends in the business community endorse my view of the proposed rule.

After finding that a party — and this is a good time to make explicit that I am talking here about discovery among parties, and not nonparty discovery subject to different cost assumptions under Rule 45 — possesses relevant, and assumedly admissible, evidence, cost shifting requires the requester to decide whether she will pay for evidence that should be made available without cost not simply presumptively under Oppenheimer but as part of the broader duty of the producer to give testimony.9

Cost sharing here would place an artificial barrier in front of a person who wants to put before a jury, the body constitutionally charged with defining its worth, relevant information that is in the hands of her opponent. Effectively, the rule will place the person in this position only when the opponent is an artificial person.10 Humans would be systematically disadvantaged. That seems bizarre.

Dahl: No, there are no drawbacks in this situation to allowing the requester — the party in the best position to decide what the discovery is worth to its claims or defenses — to obtain relevant discovery if it is willing to pay for it. This question points out the merits of a self-executing requester-pays rule. If our system required parties to pay for the discovery they request (subject to protections to ensure access to justice), judges would rarely if ever be confronted with whether to allow discovery in such situations. Rather, judges would understand from the request itself that the requester has already made the determination that this request is meaningful enough to the claims and defenses to justify the cost. It also relieves judges of having to ferret out whether the parties are using discovery as economic pressure on their adversaries.

Parties often seek additional relevant information after substantial discovery has already been produced in a first wave of discovery production. The adversarial process does not provide much help to a judge who is applying the proportionality test in such cases. In difficult-to-call cases where substantial matter has already been produced, should a judge consider shifting the costs of discovery?

Vail: Foreclosing consideration of any issue of this type is hard to square with a model that generally vests much discretion in trial judges to manage cases. So consider, yes. But the default should remain that the evidence should be produced and produced at the cost of the producer. Any consideration of cost shifting should come with appropriate skepticism and with clear recognition of the constitutional primacy of juries, not judges, to define the worth of facts. Judges should adhere to the general “policy promoting the admission of as much relevant evidence as reasonably possible,” and that policy is not advanced by cost shifting.11

Dahl: Of course. This is an easy example of when judges would benefit from the simple yet profound idea that the party making a claim or raising a defense is in the best position to know if particular information is worth the burden of obtaining it. Requiring the requester to pay for the additional discovery is the best way for the judge to ensure that parties have access to the information they need for the case while protecting the parties (and the judicial system) from inflated costs that at best are not worth it and at worst are purposeful attempts to force compromise for reasons of economic pressure rather than the merits.

This question illustrates how the current producer-pays presumption puts judges in an untenable position. How is a judge supposed to figure out whether a certain discovery request is worth the price when the person asking is not paying for it and the person who doesn’t want to provide the information says it is too expensive? Imagine a restaurant where you don’t pay for your food, but a third party is asked to decide whether the value you would obtain from the side dish you order is worth the price. Judges struggle to determine the proper scope of discovery without sufficient access to the underlying facts and with little or no knowledge about the costs involved. Even the most effective judicial managers cannot counteract the strong economic forces driving the inefficient outcomes under the current producer-pays paradigm.

A requester-pays default rule — with protections for parties who cannot afford necessary discovery — would encourage focused requests designed to obtain the information necessary for the just adjudication of the issues without the excessive costs currently experienced. It would encourage substantive assessment of cases before they are filed, create more realistic incentives to settle meritorious cases, and ease over-crowded court dockets by freeing judges from having to divine the value of particular information. It would do so while making it more likely that legal disputes will be resolved on the merits.

How great is the potential of a slippery-slope developing, where most judges quickly opt for cost shifting whenever a discovery dispute arises?

Vail: I don’t think judges are so reflexive that immediately they automatically will opt for cost shifting. I do fear, however, the corrosive power of many requests and time passing. As requests for cost shifting become more frequent in cases where debate about it might be appropriate, the idea will begin to pervade cases in which it never should be a consideration. Decisions in the big cases will be reported; decisions in the routine cases will not. The arc of the law will bend toward the reported cases, despite their rationales not being applicable to the routine.12

Dahl: This does not appear to be likely, which is unfortunate, because a requester-pays default rule (with appropriate safeguards to ensure access to justice) would have tremendous benefits for our judicial system. Our judiciary has been slow to recognize the fundamental change to litigation dynamics caused by the creation and storage of electronically stored information (ESI) and the power of requester-pays provisions to be a self-executing restraint on inefficient discovery that is not justified by its importance to the claims and defenses.

The use of the term “slippery slope” implies that there would be something wrong with judges applying basic economic principles to lawsuits. Yet there is nothing unfair or “punitive” about the idea that a litigant must bear the costs of litigating his claim. Indeed, conventional economic theory on prices as a mechanism for efficient allocation of resources is adequate justification for a “requester pays” rule. Prof. E. Donald Elliott explains:

Judges should not confuse allocating costs to those who request discovery with penalties. There is nothing punitive about requiring an economic actor to pay for resources that are consumed in an activity that he undertakes to make a profit. On the contrary, the philosophy behind a market economy is that resources will be used most efficiently if those who decide to consume them pay the marginal costs of production. For the same reasons that electricity will be wasted and over-consumed if government requires it to be supplied at a price below the marginal cost of production, litigation will be oversupplied, wasting societal resources, if those who initiate litigation pay only a small fraction of its cost.13

Requiring requesters to pay the tab for what they order will curb overbroad discovery, end the use of discovery as a weapon, and refocus attention on the merits of cases. Such a shift would accomplish more than any other single reform to control the scope and costs of civil discovery effectively and fairly.

Is it time to reexamine the presumption that the producing party bears the cost of discovery that was first laid down by the Supreme Court in Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340?

Vail: Only in the sense of rediscovering its constitutional roots and giving effect to the duty to provide evidence. Here’s what Oppenheimer says:

Under [the federal rules], the presumption is that the responding party must bear the expense of complying with discovery requests, but he may invoke the district court’s discretion under Rule 26(c) to grant orders protecting him from “undue burden or expense” in doing so, including orders conditioning discovery on the requesting party’s payment of the costs of discovery.14

Oppenheimer is a decision about the rules, so I can’t really say that new Rule 26 is inconsistent with it, regardless of whether new Rule 26 is a change in the rules. But I can and will say that increasing cost shifting is inconsistent with what I take to be the roots of Oppenheimer, something the case itself does not discuss.

“It is . . . beyond controversy that one of the duties which the citizen owes to his government is to support the administration of justice by attending its courts and giving his testimony whenever he is properly summoned.”15

That duty is a burden of implementing the right of access to justice,16 guaranteed by the First Amendment.17 The Supreme Court has recognized that this duty demands “personal sacrifice” and has not recognized cost as a “mitigating circumstance.”18

When litigation involved primarily humans, the duty to give testimony was readily applied. The rise and ubiquity of the modern corporation, which communicates primarily through documents, and the advent of electronic storage of documents complicate the application. In 1801 there were 317 corporations in the country, one for every 17,000 persons. Today there is one for each 50 persons. Many modern corporations are gargantuan, even by post-Word War II standards. In 1955 the Fortune Top 10 consisted of organizations like General Motors and U.S. Steel. Today, Walmart employs more people than all 10 of those corporations, together. If Walmart were one human, unquestionably I would be entitled to probe all its memory, stored in synapses. Because its memory is stored primarily in bytes, my ability to probe it is questioned. I think it is fair to place on the giants of the limited- liability world the burden of demonstrating why my ability to probe their memories should be circumscribed.

I don’t pretend that there are not responses to that challenge, that in some cases the burden cannot be borne. But I think stating the question that way illustrates that limitations on access should remain exceptional, lest we create systematic advantages to the benefit of artificial persons. No change in the deeper presumption of Oppenheimer is warranted.

Dahl: Yes. That presumption was not arrived at by a thorough examination of its public policy implications, and even if it had been, the world has changed dramatically enough to justify a re-examination. A requester- pays default rule (with appropriate safeguards to ensure access to justice) would encourage each party to tailor its discovery requests to the needs of the case by placing the cost-benefit decision onto the requesting party — the party in the best position to control the scope of those demands. This simple and largely self-executing rule would also encourage substantive assessment of cases before they are filed, create more realistic incentives to settle meritorious cases, and ease over-crowded court dockets. Such a rule would also discourage parties from using discovery as a weapon to force settlements without regard to the merits of a case.

Final Words

Vail: Fundamentally different views of the purpose of courts animate Alex’s and my responses. Alex begins by talking about “a very basic economic idea” applicable to a requesting party, whom he describes as a “person asking for a particular good or service.” He views courts as servants of the economy.

I view them, primarily, as servants of the polity. Their first function is providing a way for people to live together. Supporting a thriving economy is important to the polity, and tools of economic analysis can be useful for analyzing problems occurring in the administration of justice. But justice is the primary end. Courts look different when viewed through these different lenses.

Alex proposes a radical restructuring of the American justice system through adoption of a “self-executing requester- pays” rule. He criticizes the existing presumption of producer pays because its premises were not thoroughly examined and, even if they had been, things have changed. I suggested that the presumption is not an “accident of history” but a product of history, with deep roots in our experiences and jurisprudence. Change? It is not evident to me that things have changed in any systemic way.

Alex asserts certain truths to be self-evident: that there is an “explosion of discovery costs over the last two decades”; that “intentional ‘impositional discovery’ is a serious problem”; and that a requester-pays rule would “accomplish more than any other single reform” to make discovery efficient and fair. For the judicial system as a whole, I think the first two propositions are demonstrably untrue; I know of no evidence supporting the third. I have suggested that there could be some truth to all the assertions with regard to a certain small subset of disputes. But we never should make general rules to address special cases.

Dahl: The duty to testify is indeed crucial to civil liberty, and there is no dispute that fulfilling that duty can require significant sacrifice. But it does not follow that our system should impose discovery burdens that are not worth it. To the contrary, public respect for our judicial system — fundamentally, its source of authority — requires that courts not ask witnesses to shoulder burdens that have little or no value to the dispute. A fortiori, our system cannot maintain public respect when it suborns the imposition of significant discovery costs on other parties for the purpose of bringing economic pressure to bear on the outcome of legal disputes, and does so under the guise of societal responsibility. Requester-pays rules are self- executing protections against imposing too much upon, or abusing, the very witnesses whose duty to testify is so critical to justice.

Requiring a requester to pay for some or all the discovery it seeks is the most effective means of focusing litigation on the facts that matter to the claims and defenses in a case. It is based on the economic idea, almost ubiquitous in America and around the world, that a customer is in the best position to decide whether what she’s requesting is worth the cost of producing it.

Defenders of the status quo might find it useful to portray this idea as an attempt to circumscribe the right to obtain necessary information, particularly from large corporations, but that is neither true nor would it be the result. By reducing the epidemic of over-discovery, a requester-pays default rule (with safeguards to protect access to justice) would ensure attention to the merits of disputes, reduce the exorbitant expense of litigation, and free up precious judicial resources for better purposes.

See, e.g., U.S. Chamber Institute for Legal Reform, Public Comment to the Advisory Committee on Civil Rules Concerning Proposed Amendments to the Federal Rules of Civil Procedure (Nov. 7, 2013), available at http://www.instituteforlegalreform.com/uploads/sites/1/FRCP_Submission_Nov.7.2013.pdf.
2 “cost shifting” /P discovery & DA(aft 1995), run in the Westlaw allfeds database April 17, 2015.
3 “cost shifting” /P discovery & DA(bef 1995), run in the Westlaw allfeds database April 17, 2015.
4 As in notes 2 and 3, in the allstates database.
5 Emery G. Lee III & Thomas E. Willging, Fed. Judicial Ctr., Preliminary Report to the Judicial Conference Advisory Committee on Civil Rules (2009),, http://www.fjc.gov/public/pdf.nsf/lookup/dissurv1.pdf/$file/dissurv1.pdf.
6 Martin H. Redish, Allocation of Discovery Costs and the Foundation of Modern Procedurein The American Illness (F.H. Buckley ed., 2013).
7 David L. McKnight & Paul J. Hinton, International Comparisons of Litigation Costs: Canada, Europe, Japan and the United States (June 2013), available at http://www.instituteforlegalreform.com/uploads/sites/1/ILR_NERA_Study_International_Liability_Costs-update.pdf. .
Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340 (1978).
See my response to Question 5: “Is it time to reexamine …”.
10 See, again, my response to Question 5.
11 Johnson v. United States, 683 A.2d 1087, 1099 (D.C. 1996) (en banc).
12 See Nancy Gertner, Losers’ Rules, 122 Yale L.J. Online 109 (2012).
13 E. Donald Elliott, Twombly in Context: Or Why Federal Rule of Civil Procedure 4(b) is Unconstitutional, 64 Fla. L. Rev. 895, 955–56 (2012) (footnotes omitted); see also E. Donald Elliott, Managerial Judging and the Evolution of Procedure, 53 U. Chi. L. Rev. 306 (1986); E. Donald Elliott, Toward Incentive-Based Procedure: Three Approaches for Regulating Scientific Evidence, 69 B.U. L. Rev. 487 (1989) (Because regulating by incentives is more efficient than by judicial command and control, incentive-based procedure is the first-best solution.).
14 Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340, 348 (1978).
15 Blackmer v. United States, 284 U.S. 421, 438 (1932). Accord United States v. Calandra, 414 U.S. 338, 345 (1974) (“The duty to testify has long been recognized as a basic obligation that every citizen owes his Government.”).
16 Marbury v. Madison, 5 U.S. (1 Cranch) 137, 163 (1803) (Marshall, C.J.) (“The very essence of civil liberty certainly consists in the right of every individual to claim the protection of the laws, whenever he receives an injury. One of the first duties of government is to afford that protection.”).
17 Borough of Duryea v. Guarnieri, 131 S. Ct. 2488, 2494 (2011) (“This Court’s precedents confirm that the Petition Clause protects the right of individuals to appeal to courts and other forums established by the government for resolution of legal disputes.”).
18 Blair v. United States, 250 U.S. 273, 281 (1919).