Proportionality as Part of Scope

Proportionality as Part of Scope

A request can be relevant, can seek nonprivileged matter, and still fall outside discovery because it demands more than the case needs. Rule 26(b)(1) permits discovery of nonprivileged matter that is relevant to a claim or defense “and proportional to the needs of the case,” and the rule names the considerations that decide proportionality.¹ The test requires the parties, and eventually the court, to assess six factors:²

1. the importance of the issues at stake in the action;
2. the amount in controversy;
3. the parties’ relative access to relevant information;
4. the parties’ resources;
5. the importance of the discovery in resolving the issues; and
6. whether the burden or expense of the proposed discovery outweighs its likely benefit.

Proportionality has an older history than this location suggests. Before the 2015 amendments, the definition of scope made no explicit reference to proportionality, and a lack of proportionality instead justified limiting discovery that was otherwise relevant and nonprivileged.³ The earlier rule required a court to “limit the frequency and extent of discovery otherwise allowed” when the burden or expense of the proposed discovery outweighed its likely benefit, considering the needs of the case, the amount in controversy, the parties’ resources, the importance of the issues at stake, and the importance of the discovery in resolving the issues.⁴ The 2015 amendment moved those considerations, with some revisions, into the definition of scope.⁵ The most visible revision is the addition of the parties’ relative access to relevant information, which the older list did not contain.⁶

The move changes where a court asks the question. Rule 26(b)(2)(C)(iii) now directs a court to limit proposed discovery that is outside the scope permitted by Rule 26(b)(1),⁷ and the court must do so on motion or on its own initiative.⁸ The cost-benefit comparison therefore operates inside the definition of scope instead of in a separate limiting clause that applies afterward.

The reason for the move lies in who pays. Parties ordinarily cover their own costs of responding, including the time and expense of retrieving and reviewing information before it is produced, and a responding party that finds those costs prohibitive or unreasonable can ask for a protective order under Rule 26(c).⁹ On that motion the court may shift costs to the requesting party instead of simply disallowing the discovery.¹⁰ Writing the comparison into scope makes it part of the threshold test every request must pass, and one predictable consequence is more litigation over whether information is discoverable at all.¹¹

Many decisions quoted in the subsections below predate the amendment. They apply the older cost-benefit clause, or cost-shifting tests built on it, and their wording differs from the current rule. They still show how courts reason about stakes, burden, and benefit, but many cases may have come out differently under the amended text.¹² The first question to ask of any such decision is which version of the rule it applied. The subsections that follow take the six factors one at a time and then show how courts weigh them together.

Importance of the Issues

The first factor asks what the case will decide, not what the discovery will cost.¹³ Zubulake v. UBS Warburg LLC gives the working rule: when a case has the potential for broad public impact, public policy weighs heavily in favor of permitting extensive discovery.¹⁴ The court’s examples were toxic tort class actions, environmental actions, so-called “impact” or social reform litigation, cases involving criminal conduct, and cases implicating important legal or constitutional questions.¹⁵ Zubulake applied the idea inside a cost-shifting test, and its list of Rule 26 factors is the pre-amendment list,¹⁶ so the opinion is a source for the idea and the current rule is the source for the wording.

Dollars cannot capture every stake, and this factor covers what they miss. A ruling in an environmental action or a constitutional case can matter to people who never appear in the caption, and public policy then favors a fuller record. The next factor measures the money. Where a case has no public dimension, the first factor gives the requesting party nothing to add, and the request has to succeed on the other five.

Amount in Controversy

The second factor sets the cost of the discovery beside the money at stake.¹⁸ Cost alone settles nothing. A court reviewing a predictive coding protocol cautioned that cost “cannot be considered in isolation from the results of the predictive coding process and the amount at issue in the litigation.”¹⁹ The Zubulake court made the same point when it added the amount in controversy to its cost-shifting test. The addition balanced the factor that typically weighs most heavily in favor of shifting costs, the total cost of production, which “is almost always an objectively large number in cases where litigating cost-shifting is worthwhile,” and the comparison with the amount in controversy “may tell a different story.”²⁰

A raw figure shows why. In a later stage of the Zubulake litigation, UBS estimated that restoring responsive email held on backup and archival media would cost approximately $175,000, not counting attorney time to review the email, and the plaintiff moved to compel production at UBS’s expense.²¹ Whether $175,000 is large depends on what the claim is worth.

The comparison has to be relative because the responding party ordinarily bears the cost of complying.²² A request priced above anything the case could return forces the responder to spend heavily on an exercise the claim cannot justify, and a responder of limited means may settle to avoid the bill, which turns the expense of discovery into leverage detached from the merits. Measuring cost against the stakes screens out that leverage and leaves a requester free to pursue costly discovery when a great deal is at issue.

Because the comparison uses realistic exposure, the figure in the complaint’s demand cannot control it. Courts evaluating this factor must look beyond the often inflated value stated in the ad damnum clause, the part of a complaint that states the damages demanded.²⁵ The reason is how cheap the figure is to produce. A pleader pays nothing to write a large number, so if the demand fixed the stakes, any requester could unlock expensive discovery by inflating it.

When the record does not supply grounded figures, a court can build them. In Mancia v. Mayflower Textile Services Co., the magistrate judge developed a rough range for what was at stake. The judge admitted the estimate was rough but found it useful for determining the amount in controversy and what is “at stake” under the proportionality analysis, and described the goal as quantifying “a workable ‘discovery budget’ that is proportional to what is at issue in the case.”²⁶ Mancia analyzed proportionality under Rule 26(b)(2)(C), where the cost-benefit test sat before 2015. The same comparison is now made inside the definition of scope.²⁷

Relative Access to Information

The third factor directs the court to weigh “the parties’ relative access to relevant information.”²⁹ Discovery exists to give each side evidence it cannot reach on its own, and the factor asks who holds that evidence. When one party keeps the information in its own systems and the other can reach it only through discovery, the holder is poorly placed to complain about the effort of gathering what only it can find.

Labrier v. State Farm Fire and Casualty Co. illustrates the point. The information the plaintiff sought sat in State Farm’s systems, and the special master ordered State Farm to answer interrogatories instead of producing documents or letting the plaintiff search for the information, because State Farm argued that producing data fields and giving system access would divulge confidential, trade secret information.³⁰ The court approved the order and added: “A litigant cannot keep its own system secret and then refuse to gather the information itself.”³¹ The record also showed that State Farm could reach the data when it wanted to, because its own employee had quickly retrieved at least part of it when State Farm wished to use it.³²

The retailer’s request also runs into a separate limit, because discovery that can be obtained from another source that is more convenient, less burdensome, or less expensive is subject to limitation under Rule 26(b)(2)(C)(i).³³ Equal access leaves the factor with nothing to add. A resisting party cannot treat asymmetry as the end of the analysis either, because the factor is one of six, and the court still compares the burden with the benefit and the stakes.

The Parties’ Resources

The fourth factor weighs the cost of discovery against each party’s ability to bear it.³⁵ Zubulake illustrated the idea this way: discovery that would be too expensive for one defendant to bear would be a drop in the bucket for another. The court’s own example was UBS, which reported net profits after tax of 942 million Swiss francs, approximately $716 million, for the third quarter of 2002 alone.³⁶ The comparison is therefore a ratio of cost to capacity, and a price tag means something only when set against the party who must pay it.

Capacity matters because the responding party ordinarily pays.³⁷ The same bill that ends one party’s defense passes unnoticed by another, and ignoring that difference would let a well-funded requester price discovery beyond a modest opponent’s reach.

Size does not make every demand acceptable, however. In the preservation setting, one court asked whether a corporation that recognizes the threat of litigation must preserve every shred of paper, every e-mail or electronic document, and every backup tape. It answered no, because such a rule would cripple large corporations that are almost always involved in litigation.³⁸ The resources factor measures a burden against a party’s means, and it does not make a large organization the funder of any request.

Importance of the Discovery

The fifth factor asks how much the requested discovery will contribute to resolving the issues.⁴⁰ It differs from the first factor. The importance of the issues asks how much the case matters beyond the parties, while this factor asks how much a particular request matters to the case. A momentous case can generate a peripheral request, and a private dispute can generate a request that decides the outcome. The factor matters because every request spends the responding party’s time and money, and what the information adds to resolving the dispute is the only return on that spending.

Labrier shows the factor at full strength. The court held that the special master acted within his discretion “particularly in view of the centrality and importance of the information sought,” adding that it is “difficult to imagine any fact discovery more necessary to the prosecution and defense of the case” than the discovery the order covered.⁴¹

The opposite pole appears in Aguilar v. Immigration and Customs Enforcement. The plaintiffs there asserted Bivens claims that home searches deprived them of constitutional rights, and they sought metadata for the word processing and PowerPoint files they had received. The court found that they had failed to show that the “who” and “when” of document creation or modification was relevant to those claims.⁴² Their example was an undated ICE “Enforcement Operation Plan,” and they argued that learning who created and modified it, and when, would give them insight into how the defendants planned and executed the searches.⁴³ Metadata was not needed for searching either, because only about 5,200 pages, roughly one banker’s box, had been produced, and the plaintiffs should not have had significant difficulty sorting and searching the files they received.⁴⁴

The two decisions differ in the link between the material and the dispute. In Labrier the information sat at the center of both the prosecution and the defense. In Aguilar the metadata connected to the claim only through a chain of inferences, and the chain led away from the question the claim presented.

A third decision treats the factor as marginal utility. Quinby v. WestLB AG applied a cost-shifting test, and the court found that, given the low number of relevant e-mails, and even though they existed only on backup tapes, the marginal utility test was low and leaned in favor of cost-shifting.⁴⁵ Where the discovery adds little, the court has little reason to make the responding party bear the cost of it.

Burden Versus Benefit

The sixth factor asks whether the burden or expense of the proposed discovery outweighs its likely benefit.⁴⁷ Because clause (iii) of Rule 26(b)(2)(C) directs the court to limit discovery outside the scope Rule 26(b)(1) permits,⁴⁸ a request whose cost exceeds its worth falls outside the scope of discovery from the start.

A resisting party reaches for this factor first, because the cost of answering is a fact about the responder’s own files and systems. The party resisting production bears the burden of establishing lack of relevancy or undue burden,⁴⁹ and a party challenging the production of a document, including electronically stored information, has the burden of establishing that the document is not discoverable.⁵⁰ The reason lies in who holds the facts. Hours, systems, and dollars sit with the responder, so a court told only that a request is burdensome has nothing to put on the scale. The bare assertion that requested discovery is unduly burdensome is “ordinarily insufficient to bar production,” and in one decision the court was “not persuaded” that compliance was an undue burden where the objecting party gave no estimate of hours or monetary cost.⁵¹

Sampling in Oxbow Carbon & Minerals LLC v. Union Pacific Railroad Co. shows how a responding party can turn a burden objection into numbers. Oxbow first opposed the defendants’ request on grounds of undue burden and excessive cost. It then agreed to analyze a random sample of Koch’s digital records using agreed-upon search terms, to supply concrete numbers on how responsive Koch’s documents were to those terms and a basis for negotiating new search terms if necessary.⁵² The sample cost $57,198 and showed that 11.67 percent of the “hits” were actually responsive.⁵³ On that experience Oxbow estimated that processing, reviewing, and producing the rest of Koch’s documents would cost about $85,000 more, for a total of about $142,000, and it declined to produce them. The defendants renewed their motion to compel, and the magistrate judge then conducted a proportionality analysis.⁵⁴ The sample gave the court a cost, a responsiveness rate, and a projected total, which are the inputs the sixth factor compares.

The objection in Labrier never got past assertion. State Farm claimed that its payment data sat in a “picture format” reachable only through one system and that calculating labor depreciation for a claim would require downloading estimates “one at a time and recalculating each estimate manually.”⁵⁵ The court rejected the claim because State Farm’s own employee had used Xactware software at the start of the lawsuit to identify putative class members and to estimate the damages alleged. It found it incredible that no cost-effective way existed to match one database against another, since data sorting is what computers do in very short amounts of time, and it held that even claim-by-claim matching did not justify preventing the plaintiff’s access to critical information.⁵⁶ State Farm’s estimate rested on an extrapolation from materials filed in another case, which were not prepared by people with expertise in its systems, while the employees who did know the systems described no burdensome process.⁵⁷

Part of the burden State Farm described was of its own making, and the court declined to credit it. State Farm had resisted every option the special master offered except its own sample of 400 cases, and the court said that, to the extent answering interrogatories burdened State Farm, its intransigent approach had created much of that burden.⁵⁸ The court also noted that State Farm would need to gather information on an affirmative defense for its own case anyway, and that objections that gathering is too hard are not honored when the responding party must gather the same information to prepare its own case.⁵⁹ A cost the responder would incur regardless is not a cost the request imposes.

The benefit side counts results, not only cost. In a dispute over electronic review, a court rejected a proposed 40,000-document cutoff as a “pig in a poke,” explaining that where to draw the line depends on what the statistics show, because “proportionality requires consideration of results as well as costs,” and a cutoff that leaves “a tremendous number of likely highly responsive documents unproduced” does not work.⁶⁰ Relevance does not carry a request either, because the rule makes discoverable only matter that is relevant and proportional.⁶¹ Convolve, Inc. v. Compaq Computer Corp. illustrates the point: the court acknowledged that the requested bills of materials were relevant but declined to compel them because the time and expense required were out of proportion to their marginal value.

Tailoring is the requester’s half of the comparison. In Mack v. Great Atlantic & Pacific Tea Co., the plaintiff sought data for a four-year period when a lesser time would plainly have sufficed. She asked about nine job classifications, several clearly immaterial because she had already held, or did not want to transfer to, at least four and perhaps five of them. Her request reached 62 stores, a total the court found unreasonable given her unwillingness to accept a transfer outside the Springfield area and the decentralized nature of management.⁶² Time period, job classification, and geography are the dials, and each removes burden faster than it removes benefit when the excluded material has little connection to the claim. Overbroad requests also carry consequences for the requester. In the Chudasama litigation against Mazda, the court of appeals upheld the finding that Mazda violated Rule 26(g) yet held the sanction “wildly excessive,” and so not “appropriate,” given the Chudasamas’ overbroad discovery requests and the trial court’s inaction. An appropriate sanction must take into account the “needs of the case.”⁶³

The duty to think about cost therefore runs to both sides. A lawyer’s signature on a discovery request certifies that it is “neither unreasonable nor unduly burdensome or expensive, considering the needs of the case, prior discovery in the case, the amount in controversy, and the importance of the issues at stake in the action.” Boilerplate objections that a request is overbroad and unduly burdensome persist despite a litany of decisions holding them improper unless based on particularized facts.⁶⁴ A tailored request and a documented objection give the court a real ledger to weigh.

Weighing the Factors

Rule 26(b)(1) lists the six factors in one sentence and tells the court to consider them together in deciding whether a request is proportional to the needs of the case.⁶⁶ The rule assigns no ranking and no weights, so the preceding subsections describe six inputs to a single judgment and not six separate tests. In Oxbow, the court observed that the list of multiple factors in the proportionality calculus “necessarily makes answers to questions like this fact-dependent and case-specific.”⁶⁷

The test is flexible because the inputs vary from one dispute to the next. A $140,000 discovery bill is small beside a claim worth tens of millions of dollars and heavy beside a claim for damage to a $30,000 car, and the same request can seek unique evidence in one case and a duplicate of what the requester already holds in another. A formula such as a fixed ceiling on cost as a share of damages would have to ignore those differences, so the rule gives the court a list and leaves the weighing to its judgment. The price of that flexibility is predictability, and commentators have long asked where judges are to find criteria for deciding how much discovery a $10,000 personal injury suit deserves compared with a $10,000,000 product liability suit.⁶⁸

The allocation of proof matters as much as the factors. The party resisting production bears the burden of establishing lack of relevancy or undue burden, and one court, writing after the amendments, found nothing to suggest that a more restrictive view of the scope of discovery is now the norm.⁶⁹ A party that fails to particularize its objections gets little help from the court. The failure leads to one of two conclusions: either the party lacked a factual basis for the objections, which would violate Rule 26(g), or it had facts and waived them by not specifying them. Either conclusion could justify a ruling requiring the discovery regardless of cost or burden, because proper grounds for objecting were never established.⁷⁰

When the record lacks the figures the weighing needs, a court can create them. In Mancia, the magistrate judge ordered counsel to discuss the discovery already provided and the discovery still sought, to determine whether the plaintiffs’ legitimate needs could be met from sources that were non-duplicative, more convenient, less burdensome, or less expensive, and he instructed that the defendants carried the burden of giving a particularized factual basis for any claim of excessive burden or expense.⁷¹ The judge also suggested phased discovery, in which the most promising but least burdensome sources are produced first so that the plaintiffs can reevaluate their needs in light of what they receive.⁷² Sampling, as in Oxbow, works the same way for electronic sources, because the court weighs evidence of what the files hold instead of a guess about it.⁷³ Later sections take up sampling and cost shifting in detail. The decisions in this paragraph framed the comparison under Rule 26(b)(2)(C), but the techniques serve the comparison now made inside the definition of scope.⁷⁴

The court does not have to wait for a well-drafted objection before it acts. On motion or on its own, it must limit discovery that is unreasonably cumulative or duplicative, that can be obtained from a more convenient source, that the requester could have obtained earlier, or whose burden or expense outweighs its likely benefit.⁷⁵ The next section takes up those mandatory limits.