By: Kevin J. Lynch*
Introduction
Discovery plays a key role in our modern federal courts. Discovery can be costly and burdensome, but it also enables settlement, reduces informational disparities between parties, and clarifies issues for trial. Under the Federal Rules of Civil Procedure, discovery is intended to occur with limited intervention by the court, absent a dispute arising. However, in cases where a motion to dismiss is filed, judges are routinely asked to stay discovery while that motion is pending.[1] Because the decision whether or not to stay discovery in this situation is so consequential, this Article examines what judges are doing currently on motions to stay discovery and recommends prescriptions for what judges should do in order to exercise their discretion and promote the goals of the Federal Rules of Civil Procedure.
By far the largest driver of litigation costs is the cost associated with discovery,[2] and thus access to discovery can have significant effects on parties and the outcomes of cases. Discovery can impose significant costs on both parties and may even take up the court’s time if discovery disputes arise. Some people also take the view that our discovery system allows for “discovery abuse.”[3] However, despite these costs and burdens, discovery has many benefits as well. Discovery equalizes information asymmetries, thereby enhancing settlement prospects and also reducing surprises and gamesmanship at trial.[4] Engaging in discovery allows both sides to more fully assess the strengths and weaknesses of claims and defenses.[5] The current discovery system is also designed to proceed without the direct involvement of judges unless a dispute arises.[6]
Due to the important costs and benefits of discovery, decisions that affect the scope, timing, or availability of discovery are enormously consequential. For civil litigation in federal court, district and magistrate judges make many decisions about discovery that affect the cases before them. They decide the length and number of depositions that may be taken, compel or protect against the production of large numbers of documents and electronic data searches, serve as gatekeepers for expert witness testimony, and even decide whether the parties may take discovery at all until any motions to dismiss have been resolved. This Article focuses squarely on the last issue, by both developing a framework for understanding the principles and considerations that affect whether a particular judge will stay discovery in a case pending resolution of a motion to dismiss and providing recommendations for how judges should exercise their discretion to control the discovery process. The principal goal of judges should be to reduce or balance the costs and burdens of unnecessary discovery against those of undue delay. In determining which test should be applied to achieve those ends, my principal goal is to align the burden on the judges with the risk of error in deciding a motion to stay discovery.[7]
Rule 1 of the Federal Rules of Civil Procedure directs that the Rules “should be construed and administered to secure the just, speedy, and inexpensive determination of every action and proceeding.”[8] Although the Rules go into some (limited) detail as to how judges are to achieve those goals in specific situations, in other areas the Rules are silent. One such area of particular importance is whether a case should proceed, including the potentially expensive and time-consuming discovery process, when a motion to dismiss is pending. In this context, the goals create readily apparent tensions. A speedy determination would ordinarily involve moving quickly to discovery so that the ultimate resolution of the case would not be delayed if the motion to dismiss were denied, at least in part. At the same time, if a case might be dismissed entirely, or narrowed significantly, by the resolution of the motion to dismiss, then commencing broad-based discovery would add unnecessary expense to the case. This raises the question: when does a stay of discovery effectively stay justice? The Rules and appellate decisions make clear this issue is left to the discretion of judges at the trial level.[9] In exercising this discretion, courts might automatically stay discovery when a motion to dismiss is pending, never stay discovery in that circumstance (or disfavor it), or apply some kind of balancing test to weigh the competing interests and the potential harms due to delay or due to unnecessary discovery.
This Article examines how courts determine whether to stay discovery when motions to dismiss are pending,[10] a critical procedural issue that has heretofore received relatively scant academic interest.[11] Part II provides background on the costs and burdens of discovery, the various interests at stake, and the judicial role overseeing discovery. Part III presents the current state of the law by looking at the various standards that courts have explicitly applied when deciding motions to stay discovery. Part IV develops a framework for understanding and reconciling existing precedent on discovery stays, describing eight primary considerations. Part IV also lays out a prescription for judges to use in exercising their discretion in this context. Part V examines the broader issue of “discovery abuse” and specific cases where discovery is automatically stayed while also noting areas for further inquiry into this issue.
I. Background on Discovery and Its Role
in Federal Civil Litigation
A. Costs and Burdens Associated with Discovery
Discovery creates a variety of costs and burdens for the parties and for the court. Attorneys and their clients spend numerous hours preparing, reviewing, responding, and objecting to discovery requests, such as requests for production, interrogatories, and requests for admission.[12] Depositions are time consuming and costly for the witnesses and the lawyers, and they include costs for court reporters and videographers. Witnesses have their work and personal lives disrupted by the depositions and associated preparation and travel time. Expert witnesses can be a major driver of discovery costs, particularly when specialized expertise is required and the experts can command high hourly rates for services.[13] The parties and the court spend valuable resources on discovery disputes.
Almost anyone would agree that discovery is expensive, at least in complex cases. The commonly held view is that discovery is too expensive, and that efforts should be made to reform the rules governing discovery, or the practices of attorneys and judges, to reduce the costs and burdens.[14] Concerns about the costs and burdens of discovery are particularly acute in regard to e-discovery, where the potential volume of discoverable information can be quite large.[15] Obtaining empirical data on the cost of discovery is challenging,[16] but at least some evidence exists to challenge the assumption that discovery is too expensive and that it is largely a function of the value of the case.[17] This Article does not attempt to address the normative issues surrounding the appropriate scope and burden of discovery. Suffice it to say that discovery is typically one of the primary drivers of cost in federal civil litigation.
B. Incentives at Play Regarding Stays of Discovery
Discovery is costly and burdensome to all involved—plaintiffs, defendants, and courts. As a result, this creates strong incentives for avoiding what might ultimately become unnecessary discovery, particularly for defendants to delay incurring the expense of discovery. It is this incentive that most commonly leads to a motion to stay discovery.
Despite the efforts of many to streamline federal court litigation, complaints about undue delays persist.[18] Particularly if they can avoid the expense of discovery, defendants in federal court litigation already have many incentives to seek delay.[19] Such delay is particularly troublesome for plaintiffs, as the majority of motions to dismiss across the country are denied, at least in part.[20] However, plaintiffs may also wish to stay discovery to avoid spending their time and resources until they have clarity on how the judge will rule on a motion to dismiss.
Defendants are typically concerned with avoiding the costs and burdens of discovery when those costs may ultimately prove unnecessary if a motion to dismiss is granted. However, defendants might also want to avoid a stay of discovery so that the ultimate resolution of the case will not be delayed. This is particularly relevant when the litigation creates uncertainty for the defendant that it wishes to have resolved expeditiously.
Other parties, such as intervenors or even those not party to the lawsuit, may be affected by the decision to grant or deny a stay of discovery.[21] The public may also have an interest more broadly in a particular case, either in seeing a speedy or an efficient resolution of the case, or in avoiding unnecessary costs that may be passed onto the public. And courts, of course, have a strong interest in controlling their dockets and ensuring they can manage their cases effectively. Discovery is also not cost-free to the courts despite modern discovery rules designed to limit court involvement in discovery disputes.
C. Judicial Management of the Discovery Process
The Federal Rules of Civil Procedure do not explicitly mention stays, either generally or in the context of discovery. Instead, judges cite either to a court’s “inherent authority” to manage the cases before it[22] or, in the context of stays of discovery, to the broad language of Rule 26(c),[23] which authorizes “any order which justice requires to protect a party . . . from annoyance, embarrassment, oppression, or undue burden or expense.”[24] Courts have read the language of Rule 26(c) as enabling them to issue orders staying discovery upon a showing of “good cause” by the moving party or parties.[25] Courts cite to one of these sources of authority,[26] if anything, when issuing written decisions on motions to stay discovery.[27]
Stays of proceedings in federal court, including stays of discovery, are committed to the discretion of the trial court.[28] Courts have stated that the mere filing of a motion to dismiss, standing alone, is not sufficient grounds for staying discovery.[29] In some contexts, however, a stay is more likely, even automatic. The Supreme Court has specifically addressed discovery in the qualified immunity context, stating that discovery should not be allowed until the threshold question is resolved.[30] Furthermore, Congress explicitly requires stays of discovery when a motion to dismiss is filed in shareholder derivative suits.[31]
Neither the appellate cases nor the Federal Rules of Civil Procedure go into any detail regarding how a trial court judge should exercise discretion in deciding a motion to stay discovery when a motion to dismiss is pending. Appellate courts will occasionally overturn a decision on a motion to stay discovery, but those opinions are based on an “abuse of discretion” standard of review.[32] The Rules do not lay out any specific factors to be considered by judges, although many courts have incorporated a “good cause” requirement.[33] Also, because many stays of discovery may be decided in the absence of a written opinion, it is not clear what factors, if any, judges are considering in a significant number of cases.
II. Standards for Deciding Discovery Stays
Close inspection of written opinions from a range of federal courts reveals that no uniform standard for deciding motions to stay discovery exists. The standards that are applied are context-dependent to be sure. The standard applied may depend not only on the jurisdiction in which the case is filed, or on the judge hearing the motion, but also on the cause of action, the identity of the parties, the issues raised in the motion to dismiss, and the strength of the arguments in the motion to dismiss. The considerations that affect decisions on motions to stay discovery are discussed more fully below in Part IV.A.
Furthermore, rare is the appellate case discussing what substantive standard should be applied at the trial court level.[34] The lack of appellate case law is unsurprising because an order staying or denying a stay of discovery, by itself, is not an appealable final order.[35] Thus, the bulk of the case law discussing discovery stays is written by district judges or magistrate judges. Yet even this universe of cases only captures those in which the trial judge issues a written opinion or order. It is unclear how often judges simply grant or deny motions to stay discovery without issuing a written opinion explaining their decision.[36] To further complicate matters, the stay of discovery might occur not due to the filing of a motion to stay but rather during the scheduling conference and issuance of a scheduling order.[37] As a result of these factors, courts within a given circuit might apply different standards in analogous situations, and there can even be variation among judges within the same district court.
Despite this variability, the range of approaches used by federal trial judges appears to be logically constrained to a few key categories. A court might automatically stay discovery in a case once a motion to dismiss is filed. On the flip side, a court might never use a pending motion to dismiss as grounds for staying discovery. In between those two extremes, a judge might instead apply a balancing test to weigh the competing interests, and might even take a preliminary peek at the motion to dismiss as part of this balancing. Without looking into the strength of the motion to dismiss, a court can balance the interests of parties, the court itself, and others. By first preliminarily evaluating the motion to dismiss, the court can then more accurately weigh the likelihood of harm due to delay against harm due to unnecessary discovery costs. These possible approaches will now be discussed in more detail.
A. Auto-Stay
Outside of a few specific subsets of cases, notably shareholder derivative suits and qualified immunity cases, courts have been hesitant to state that a stay of discovery should automatically be issued while a decision on a motion to dismiss is pending. At the same time, when a court issues a stay of discovery it will often use language that might imply, even strongly, that a stay of discovery would always be appropriate when a motion to dismiss is pending.
1. Example—Shareholder Derivative Suits
The example of securities litigation is an interesting one in the discovery stay context, even if the absolute number of these cases is somewhat limited.[38] In 1995, Congress enacted the Private Securities Litigation Reform Act (“PSLRA”), amending both the Securities Act of 1933 and the Securities Exchange Act of 1934 to add provisions automatically staying discovery during the pendency of a motion to dismiss.[39] This reform has been referred to as “the culmination of extensive lobbying efforts by accountants, securities firms, and the high-technology industry to curtail what they perceived to be abusive securities class action litigation.”[40] Proponents of this change argued that “the transaction cost of filing such suits was minimal,” allowing the filing of “strike suits” based on theories of “fraud in hindsight.”[41] Thus the PSLRA was enacted to discourage such suits and limit their number.
2. Example—Qualified Immunity
The other set of cases where an auto-stay approach is followed is those in which a defendant raises a defense of qualified immunity. In this example, it was not Congress, but the Supreme Court, in Harlow v. Fitzgerald,[42] that decided that defendants should not be subjected to the burdens of discovery until such time as their qualified immunity defense was ruled upon.[43] Although the Court in Harlow was not faced directly with the question of whether discovery should be stayed pending a motion to dismiss,[44] the Court did make the broad statement that “[u]ntil this threshold immunity question is resolved, discovery should not be allowed.”[45]
The Supreme Court recently made a similar pronouncement in another qualified immunity case, Ashcroft v. Iqbal.[46] In Iqbal, the Court stated that Rule 8 “does not unlock the doors of discovery for a plaintiff armed with nothing more than conclusions.”[47] The Court further discussed concerns about the “heavy costs” of litigation and rejected the lower courts’ “careful case management” approach for limiting the burdens of discovery.[48] The Court also noted the concerns of Second Circuit Judge Cabranes, in his concurring opinion, that defendants could be subjected “to the burdens of discovery on the basis of a complaint as nonspecific as [Iqbal’s].”[49]
The Iqbal decision has received significant media and scholarly attention for its impact on pleading requirements under Rule 8(a).[50] In addition to that debate, the case also has potential consequences relevant to this paper, as Iqbal might impact judges deciding discovery stays as well.[51] What are courts saying about Iqbal in the context of deciding motions to stay pending resolution of motions to dismiss? Given the recency of the Iqbal decision, only a relatively small number of cases have directly cited the case in deciding motions to stay. A recent search of the “ALLFEDS” database on Westlaw produced twenty-seven cases—two appellate cases and twenty-five district court cases—where Iqbal was cited in this context.[52] Yet even in this limited number of cases, there are indications that at least some courts find that Iqbal supports granting stays of discovery.[53]
As I argue in Part IV.B below, however, I do not believe that judges are required to interpret Iqbal so broadly[54] that it require automatic stays upon the filing of motions to dismiss. Some courts already have resisted that view,[55] and they were wise to do so. The automatic stay approach is over-inclusive and therefore results in unnecessary costs and burdens associated with delay. These costs and burdens can be avoided if judges apply the appropriate balancing test, and the benefits of such an approach outweigh the additional burden imposed by the balancing tests.
3. Auto-Stay in Other Contexts
Judges might take a similar approach with types of cases they wish to discourage and automatically stay discovery in those cases. Perhaps, therefore, the basis of jurisdiction for a case can be used to help explain which standard was applied by the judge in deciding a motion to stay discovery, or could help to explain any differences in the frequency of motions to stay being granted or denied in those types of cases. The type of case at issue might also explain why courts sometimes make what appear to be very broad statements that discovery should not proceed while a motion to dismiss is pending. One such example is discussed below.
A rule requiring an automatic stay of discovery might be the one most likely to result in a uniform standard within a given jurisdiction, perhaps even across an entire circuit. One interesting case out of the Eleventh Circuit appears to state just such a standard, at least for Rule 12(b)(6) motions. In Chudasama v. Mazda Motor Corp.,[56] the court distinguished between pretrial motions turning on findings of fact, such as motions to dismiss for lack of personal jurisdiction, and those which should be decided before discovery.[57] The court stated:
Facial challenges to the legal sufficiency of a claim or defense, such as a motion to dismiss based on failure to state a claim for relief, should, however, be resolved before discovery begins. Such a dispute always presents a purely legal question; there are no issues of fact because the allegations contained in the pleading are presumed to be true. Therefore, neither the parties nor the court have any need for discovery before the court rules on the motion.[58]
This language would seem to indicate, at least in the context of motions to dismiss pursuant to Rule 12(b)(6), that a stay of discovery should always be granted. TheChudasama case has been followed by subsequent decisions in the district courts and even in the Eleventh Circuit itself,[59] but at other times either the decision was not mentioned in the context of a pending motion to dismiss,[60] or the district court distinguished Chudasama and found that discovery should not be stayed.[61] Thus, it is not clear that the Chudasama case requires an automatic stay of discovery, even in cases involving Rule 12(b)(6) motions to dismiss within the Eleventh Circuit.
These cases do not preclude other approaches and should not be read so broadly by the lower courts. An automatic stay rule, especially a blanket rule that applies to all types of cases regardless of other factors, would inevitably cause unnecessary delay in some subset of cases, imposing the burden of that delay on the courts (in terms of a clogged docket) and on the parties (principally on plaintiffs who risk spoliation of evidence and must risk irreparable harm or wait to be compensated for injuries). While such a bright-line rule is simple and easy to administer, it imposes substantial costs and inserts bias into the system, as will be discussed in more detail below in Part IV.B.2.
B. Balancing of Interests Without Prejudging
Where competing interests are at play, some form of balancing test provides the best means for judges to reach the result that best takes all sides into account. The competing interests at play in the discovery stay context involve the costs and burdens associated with unnecessary discovery (typically associated with defendants although plaintiffs bear these as well) and the costs and burdens of delay in both discovery and ultimate resolution of the case (typically associated with plaintiffs although defendants may also have incentives to avoid delay and reduce uncertainty). These two principal burdens can be weighed against each other to determine whether there is greater risk from delay or from unnecessary discovery, with a discovery stay granted or denied in a way that minimizes the risk. Other factors may also influence a judge’s decision, such as interests from the court itself, from the public, or from others not party to the litigation. Thus the typical[62] formulation of this approach is a five-factor balancing test weighing:
(1) the interest of the plaintiff in proceeding expeditiously with the civil action as balanced against the prejudice to plaintiffs if a delay; (2) the burden on defendants; (3) the convenience of the court; (4) the interests of persons not parties to the civil litigation; and (5) the public interest.[63]
In this Article, I will refer to this test as the “balancing of factors” approach.
Courts have put some gloss on these factors, although in general there is not a lot of guidance on how these factors should be weighed against each other. For example, potential prejudice to the plaintiff might result if there is a risk that evidence will be spoiled during any delay, particularly if the case involves significant witness testimony.[64] Some courts require that the burden must be an extraordinary burden, so a defendant cannot rely on a generic assertion that discovery will be burdensome, because that is always the case.[65] Instead, the defendant must show that the case is unusually complex, discovery will be unusually extensive, or something along those lines.[66] The convenience of the court is sometimes described as reflecting the ability of the court to manage its docket and ensure that cases proceed at an appropriate pace; thus, some courts are hesitant to grant discovery stays when it may be many months before a motion to dismiss is decided.[67] However, judges typically do not go into great detail on each factor, or how they are to be compared with the other factors.
C. Preliminary Peek
An alternative balancing approach is to weigh the potential harms if the discovery stay is needlessly granted or denied, not in the abstract but according to the likelihood that the motion to dismiss will be granted or denied. Thus, a judge will assign more weight to the burdens of discovery if she thinks the motion to dismiss is likely to be granted but more weight to the risks of delay if she thinks the motion to dismiss is likely to be denied. The amount of weight put on either end of the scale depends upon how confident the judge is that the motion to dismiss will ultimately be resolved the same way as this preliminary determination. Some courts use the phrase “preliminary peek” to describe this approach,[68] and I will adopt that terminology in this Article.[69]
If the stay is granted, the potential harm is that the motion to dismiss will be denied, in whole or in part, and then discovery will have been delayed, which would therefore delay the ultimate resolution of the case. This is a particular concern for a plaintiff who ultimately prevails, but either does not or cannot recover for harms incurred during the delay, such as irreparable harm which cannot be compensated for by money damages. Delays in discovery itself may also prove harmful, as documents are lost or destroyed, witnesses’ memories fade, or witnesses become unavailable. However, as discussed in Part II.B above, delay may prove harmful not just to plaintiffs, but also to defendants, third parties, or even the public. The court may also experience harm due to delay in the form of an increasingly clogged court.
On the other hand, the risk of potential harm when a discovery stay is not granted is that the case will ultimately be dismissed, making any discovery burden an unnecessary cost, or that the case will be partially dismissed such that the scope of discovery was unduly broad before the ruling on the motion to dismiss. This harm could impact not only the defendant forced to bear the expense of discovery, but also the plaintiff who bears his own cost of discovery for a claim that ultimately proves unsuccessful. The court also may incur costs in terms of managing discovery disputes, holding a status conference, or in other ways expending resources that ultimately could have been conserved.
Some examples are helpful to distinguish the differences between the “balancing of factors” and “preliminary peek” approaches. Both of these balancing tests consider roughly the same factors, with the preliminary peek test adding one additional factor: likelihood of success of the motion to dismiss. Thus, while the “balancing of factors” test might give roughly equal[70] weight to the harm from unnecessary delay and to the harm from unnecessary discovery costs, the “preliminary peek” approach will adjust the weight given to the considerations based on their likelihood of occurring. Suppose, for example, that a case involves a small number of witnesses who may be deposed or who would testify at trial and a reasonable number of documents that are potentially discoverable. Suppose also that the parties have sufficient resources such that they can bear the costs of discovery without affecting their finances greatly, and that the plaintiff can be fully compensated by money damages in the future. Under the “balancing of factors” approach, a judge might be expected to deny the discovery stay, finding that the costs imposed by discovery are not unreasonable and they are outweighed by the admittedly small risk of harm through spoliation of evidence and delay. However, a “preliminary peek” at the motion to dismiss might reveal that the case is very likely to be dismissed, and, under this approach, a judge might find that even the small burden imposed by discovery in this case would be unacceptable given the likelihood that it would be a waste of time and resources.
Suppose instead that discovery in our example case will be quite extensive, with numerous experts submitting testimony and dozens of witnesses to be deposed, but that most of the material evidence is reflected in hundreds of thousands of documents subject to discovery, with little risk that they will be lost or destroyed. Under a “balancing of factors” approach, the court will now weigh the significant costs and burden associated with discovery against a somewhat greater burden of delay, but still without significant risk of spoliation of evidence. Furthermore, the court can expect multiple discovery disputes to be presided over. Thus a judge might be expected to stay discovery, finding that before incurring the associated costs and burdens the motion to dismiss should be decided. Under a “preliminary peek” approach, however, the same judge might determine that the motion to dismiss merely quibbles around the edges and that the major claims are very likely to succeed. The only way the entire case could be dismissed is due to a challenge to standing, although standing seems clearly evident in the case. Thus, the judge might decide to allow discovery to proceed, despite its costs and burdens on the parties and the court, because such discovery is unlikely to prove wasteful and any risk is outweighed by the harms associated with delay.
The “preliminary peek” approach thus allows judges to refine their balancing in a way that allows them to minimize the risk of unnecessary costs and burdens in any particular case. But this approach does not come without a cost, as the additional factor in the balancing requires the judge to preliminarily review the motion to dismiss and reach a tentative conclusion regarding how it will be resolved. The “balancing of factors” approach is simpler than the “preliminary peek” approach, but both are more burdensome than a bright-line rule to either stay or deny a stay in all or a subset of cases, because the balancing tests require the judge to consider a variety of factors and exercise her discretion; the balancing tests also require the parties to submit briefs on the relevant factors. An automatic stay rule, on the other hand, either does or does not apply based on the characteristics of the case, and so it saves both the courts and the parties time and resources. The various burdens associated with these tests are discussed in more detail in Part IV below, along with the most relevant considerations that can be used to decide which test is most appropriate to apply in a given case.
D. Disfavored Stay
One final key category is that the court might never stay discovery when a motion to dismiss is pending. One might expect that this extreme position[71] is unlikely to be found; indeed, I have not found any court going so far as to say that discovery could never be stayed due to a pending motion to dismiss. However, courts will indicate that stays of all discovery are disfavored,[72] or that more than simply the filing of a motion to dismiss is required to stay all discovery.[73] Courts might cite a policy that stays of discovery are disfavored and then point out that no “compelling reasons” exist to stay discovery in a given case, without either balancing the interests or taking a preliminary peek at the merits.[74] This type of approach might be simpler than applying one of the balancing tests, but it is difficult to imagine a judge that would not, or should not, stay discovery in some cases for good cause.
III. A Framework for Analyzing Requests to Stay Discovery: Factors to Explain Tests Applied and Outcomes
Regarding Discovery Stays
Deducing a pattern to explain what standard will be applied to decide motions to stay discovery, both across different types of cases and across jurisdictions, is a difficult proposition. Jurisdictions vary greatly in what tests they apply under what circumstances, even within the same circuit. Courts in the same jurisdiction will apply different tests (either slightly or significantly different) in similar situations. However, the reported cases explicitly identify a number of characteristics that judges consider that might affect either which standard the courts apply or how they apply the standards. Based on my review of the available written opinions on discovery stays, I have also identified additional considerations that may impact how judges decide motions to stay discovery. Some of these factors relate to not just the test that is applied but how the competing interests should be weighed against each other. Eight of these principal considerations will be discussed in this Article:
(1) The type of case (including the judge’s familiarity with and views regarding it);
(2) The presence of the government as a party;
(3) The likelihood of settlement;
(4) Threshold versus ultimate issues;
(5) Partial versus complete motions to dismiss;
(6) The remedy sought;
(7) The complexity of the case (or specifically the scope of discovery); and
(8) Whether the motion to stay discovery is decided by a district or magistrate judge.
These considerations will be used to both analyze and understand what judges are doing currently on motions to stay discovery as well as to formulate my prescriptions for what judges should do in order to exercise their discretion and promote the goals of the Federal Rules of Civil Procedure. Some of these factors relate to the question of which standard should be applied, while other factors are more properly limited to consideration in a balancing test.
A. Type of Case
The type of case might be a proxy that judges use to assess whether the motion to dismiss is more or less likely to be granted. Some judges might be more likely to take an initial look at the motion to dismiss in particular cases where they suspect that the motion to dismiss might have merit. Thus, courts might not bother to take a preliminary peek in run of the mill contract dispute cases. In more unusual or complicated cases, or in those cases in which the judge does not have much experience with the parties, the judge would be more concerned about the costs of unnecessary discovery. Environmental cases are typically complex, and most judges do not have much experience with them, so they might be more inclined to look at the merits of the motion to dismiss in those cases. This same dynamic could also play out in cases where the judge has significant experience but thinks that the motion to dismiss is more likely to be granted, such as prisoner petitions or pro se cases.
The type of case might also be a proxy for the judge’s view toward particular classes of parties, and so the outcomes of motions to stay discovery might be explained partially by how the judge views the parties and how that influences her exercise of discretion. This consideration is not explicitly mentioned in the cases, although it seems plausible to expect that it may influence at least some judges. One related consideration that has been explicitly mentioned is the litigation history of the plaintiff.[75]
The type of case is of course already an explicit factor that leads to application of the auto-stay rule in the specific examples of qualified immunity and shareholder derivative suits discussed previously.[76] Thus another way to view this factor is as a case-by-case application of the same principle justifying the application of the auto-stay rule. For particular kinds of cases, judges are more concerned about the risks of harm due to unnecessary discovery costs. For those cases, judges will be more inclined to look more closely at the risk of such harm and more likely to ultimately grant a stay of discovery to protect against it.
B. Government as a Party / Public Versus Private Law
Where a government official is a defendant, particularly when immunity is raised in the motion to dismiss, courts are more likely to apply the auto-stay rule or to grant a stay based on a balancing of interests.[77] However, outside of the § 1983 context, when the government is sued the case is often decided on the administrative record, and so discovery does not become an issue. I have not found any cases discussing discovery stays where the government initiates the suit, such as in a civil enforcement action. These cases might provide interesting additional data, however, as to whether judges are more or less likely to defer to the government’s position on a motion to stay discovery. At least in the immunity context, the presence of the government or a government official as a party does affect the test that is applied.[78] The government as a party also has the potential to influence the result of any balancing test that is applied if judges are more likely to give greater weight to the burdens of discovery when imposed on government officials who have other important duties.[79] A closer examination of these public law cases is needed to shed light on the deference judges give to government parties when deciding motions to stay discovery.
C. Settlement
I have not found any cases discussing the likelihood of settlement as a factor influencing either the outcome of a balancing test or the test that is applied. However, many judges view facilitating settlement as one of their most important roles,[80] particularly the magistrate judges who typically deal with discovery disputes including motions to stay discovery. Judges might also see their role as protecting defendants from the harassment of discovery designed to lead to a quick settlement.[81] Settlement becomes more likely as discovery progresses and the basic facts of the case become more established and known to both sides.[82] The exchange of documents, deposition testimony, and expert reports might all cause one side or the other to become less confident in his or her chances of success and therefore more open to settlement to resolve the case on their terms. Additionally, the imminent costs to be incurred through the discovery process can also provide an incentive for parties to come to the table for settlement discussions. On the other hand, if discovery is stayed while a motion to dismiss is pending, the parties have little incentive to reach a settlement until the motion to dismiss is decided and they can reassess the strength of their respective positions.
Therefore it is entirely plausible that judges might take the probability of settlement into account when deciding motions to stay discovery. If the judge views the likelihood of settlement to be reasonably high, she might therefore be more likely to allow discovery to proceed in order to encourage an earlier settlement. If, however, the parties seem to be far from settlement, then a bit more discovery is unlikely to result in resolution of the case.
D. Threshold Versus Ultimate Issues
Courts sometimes look to whether a “threshold” or “preliminary” issue is at the heart of the motion to dismiss, such as when jurisdiction, venue, or immunity are preliminary issues.[83] Standing might also be a basis for a motion to dismiss that could cause courts to look more closely into whether a stay of discovery should be granted.[84] Where these threshold issues are part of the motion to dismiss, courts apply more scrutiny and give greater weight to the concern that discovery burdens would ultimately prove unnecessary.[85] Judges appear to be more likely to apply an auto-stay rule or take a preliminary peek at the merits of cases when the motion to dismiss raises a threshold issue.[86] However, where some fact discovery is needed for the plaintiff to establish jurisdiction, courts will typically allow at least limited discovery on that issue.[87]
Why does the presence of a threshold issue matter to judges? Threshold issues are the types of issues that can completely dispose of an entire case, unlike failure to state a claim motions under Rule 12(b)(6), which may dispose of the entire case but also may dispose of only a claim or a portion of a claim. Also, because these threshold issues are seen as prerequisites to reaching the merits, courts are more likely to stay discovery because discovery is primarily geared toward the merits of the case. To the extent that threshold issues require discovery before they can be decided, courts can and do limit discovery to only those threshold issues in order to carefully manage the process and decide the threshold matters without expending resources on merits discovery until it will clearly be relevant.[88]
E. Partial Versus Complete Motions to Dismiss
A court is unlikely to stay all discovery in a case when the motion to dismiss would only partially dispose of the case.[89] Thus, one would not expect to find a court that applies the auto-stay standard in all cases where a motion to dismiss is filed. My research, not surprisingly, has turned up no such cases. Staying all discovery in a case where a partial motion to dismiss was filed would almost certainly cause unnecessary delay with respect to those parts of the case not subject to the partial motion to dismiss.
Judges who take a more active role in managing discovery will sometimes stay discovery regarding those portions of the case which are subject to a motion to dismiss, provided those portions of the case are discrete and separable from the remaining portions of the case. It is not uncommon, however, for the same set of documents or the same witnesses to contain information relevant to more than one part of a case, and so it may not be possible to stay discovery with respect to only the parts of the case that are subject to the motion to dismiss.[90]
F. Remedy Sought
Judges might also take into account the remedy sought by the plaintiff and how that reflects on the various interests that might be affected by delay. When the plaintiff is seeking only money damages, the risk of harm due to delay is lessened.[91] My research thus far has not turned up any courts that look to whether there is a risk of irreparable harm, although this could be incorporated into either the preliminary peek or the balancing interests without prejudging approaches. Of course, the risk of irreparable harm can also be addressed through a preliminary injunction, although a more lenient standard is justified here where the plaintiff seeks not to enjoin action by the defendant but simply to proceed with discovery in the case during the pendency of a motion to dismiss. As discussed below, I do think that the interests at stake should be considered by judges deciding motions to stay discovery and that this factor justifies greater judicial involvement to ensure that the risks of harm due to delay are appropriately balanced against the risks of unnecessary discovery.
G. Complexity of the Case (Scope of Discovery)
One consideration that some courts clearly take into account is the complexity of the case and whether a more complex case is likely to have unusually burdensome discovery.[92] This consideration is explicitly part of both balancing tests, where the judge will look to the potential burdens of discovery in deciding whether or not the motion to dismiss should be decided before discovery commences.[93] Although this consideration is clearly important, particularly in complex cases, I have ranked it after other considerations because it appears to be less important in influencing the standard that judges apply on motions to stay discovery.[94] However, as discussed below, I argue that the complexity of the case should impact not just the outcome of the balancing test but also which balancing test should even be applied. Complex cases pose a greater risk of wasted resources should discovery prove unnecessary, and so this justifies greater judicial involvement to strike the appropriate balance.[95]
H. District Judge Versus Magistrate Judge
Magistrate judges are authorized pursuant to Title 28, Chapter 43 of the U.S. Code. Jurisdiction and powers of the magistrate judges are set out in 28 U.S.C. § 636.[96] Although parties must consent to have a magistrate decide potentially dispositive motions, such as motions to dismiss, magistrates are often given broad authority to manage the discovery process and to decide discovery disputes. However, as I have already argued, a decision whether or not to stay discovery can have very significant consequences for the parties and can even be outcome-determinative if one party is forced to settle or abandon its case or defense due to the burdens of discovery or the costs of delay.
Particularly for cases where the parties do not consent to have the magistrate decide dispositive motions, it would be potentially problematic to have magistrates “prejudge” the motion to dismiss when deciding whether or not to grant a motion to stay discovery. A magistrate may deem a motion to dismiss to be without merit, and therefore allow discovery to proceed even if it is particularly burdensome. Or a magistrate may deem a motion to dismiss likely to succeed, and therefore stay discovery even though it presents a significant risk of spoliation of evidence or would not impose any extraordinary burden on the parties. In either scenario, the magistrate’s views on the motion to dismiss may differ from the district judge’s views. Absent consent by the parties, such important decisions are better resolved using a balancing test that does not involve the magistrate prejudging the motion to dismiss. Indeed, some magistrate judges explicitly disclaim prejudging motions to dismiss when they decide motions to stay discovery.[97]
I have not uncovered any cases where a magistrate states that she applies a different standard than a district judge would apply, however, so it is not clear whether the presence of a magistrate judge has any impact on the test applied or the outcome on motions to stay discovery. Without conducting an empirical study, it is difficult if not impossible to say at this point whether courts set a different standard based on whether a magistrate or district judge is deciding the motion to stay discovery.[98] Conceivably the test might vary even within the same jurisdiction, depending on whether the district judge refers the motion to stay discovery to the magistrate or not. I will argue below, in Part IV.B.3, that magistrate judges should apply a balancing test that does not involve a preliminary peek unless the parties have consented to magistrate jurisdiction.
IV. A Suggested Approach for Judges Exercising Their Discretion on Discovery Stays
A. Existing Proposals for Reform
I am aware of at least one proposal to reform the Federal Rules of Civil Procedure that specifically addresses stays of discovery in the motion to dismiss context. The American College of Trial Lawyers and the Institute for the Advancement of the American Legal System created a Task Force on Discovery and Civil Justice that conducted research, surveyed trial lawyers, and evaluated potential solutions to the problems that cause the federal civil justice system to be expensive, cumbersome, and protracted.[99] One of the proposed “Pilot Project Rules” from the Task Force directly addresses the issue of discovery stays:
Upon the making of a motion directed to the personal or subject matter jurisdiction of the court or the legal sufficiency of one or more claims for relief, made together with an answer or at the time within which an answer would otherwise be due, the court, at the request of the moving party based on good cause shown, may stay initial disclosures and discovery in appropriate cases for a period of up to 90 days. The motion must be decided within that 90 day period.[100]
This proposal would make explicit the courts’ authority to stay discovery pending a motion to dismiss, would require good cause to be shown, and, most importantly, would require a resolution of a motion to dismiss within ninety days. While requiring a ruling on a motion to dismiss within ninety days would greatly streamline the federal judicial system and reduce the risk of prejudice due to delays while discovery is stayed, I have doubts about the practicality of such a rule. Many federal courts have caseloads that are already excessive, and Congress has not acted to address the issue by adding additional judgeships.[101] Additionally, there would be no consequences if a judge did not rule on the motion to dismiss within ninety days. Extending the stay or ending it, while of great significance to the litigants themselves, is not likely provide much incentive to judges to decide the motions any more quickly. And in some cases, particularly complex cases, we may not want the judge to rush a decision on a dispositive motion.
Beyond the impracticality of requiring decisions on motions to dismiss within ninety days, the proposal merely states that “good cause” must be shown in order for the court to stay discovery. This might affect a change in those jurisdictions that are most likely to apply an “automatic stay” type of standard, but otherwise would not provide any additional guidance to trial court judges as they exercise their discretion. I believe that a more important, and practical, aspect of reform involves specifying what standards—and under what circumstances—judges should apply in the exercise of their discretion.
B. My Proposal
The two primary benefits of my proposal are (1) efficiency and (2) transparency. The proposal promotes efficiency by laying out a framework to justify the standard to be applied by courts that balances the burdens of that standard against the potential costs and benefits of reaching the “wrong” result. By focusing greater attention on those cases that pose greater of from either a “false positive” or a “false negative,” my proposal promotes the efficient use of court and party resources. The proposal promotes transparency by encouraging judges to give explicit consideration to the factors affecting their exercise of discretion and to make their weighing of competing interests known both to the parties and to the broader public. The proposal remains reliant on the exercise of judicial discretion because judges have the experience and the flexibility to minimize the risks associated with unnecessary discovery costs and undue delay.
Like any standard applied by the courts, rules for deciding motions to stay discovery due to pending motions to dismiss can be either over-inclusive or under-inclusive.[102] Examine the issues from the perspective of a judge who wishes discovery to proceed unless a stay is necessary to protect a party under Rule 26(c). From that viewpoint, a standard is over-inclusive if it results in a stay of discovery in cases where discovery should have proceeded without delay, such as when the motion to dismiss is ultimately denied. A standard is under-inclusive if it results in discovery proceeding for some time, until the motion to dismiss is granted, making the time and resources spent on discovery a waste. Arguably there is a grey area in the middle,[103] where either (1) discovery should have proceeded, even though the motion to dismiss was ultimately granted, because, for example, there was a significant risk of spoliation of evidence or a significant risk of irreparable harm due to delay; or (2) discovery should not have proceeded, even though the motion to dismiss was ultimately denied, because for example the costs and burdens of discovery were so great that any risk of waste should not be tolerated.
Of course a standard might also turn out to be both over- and under-inclusive, depending on the specific cases it is applied to. A rule that allows discovery to proceed when the burden of discovery is minimal but not when it is extraordinary is under-inclusive in the simple case that is ultimately dismissed but over-inclusive in the complex case that ultimately survives the motion to dismiss. A rule that always stays discovery when only money damages are at issue but that never stays discovery when there is risk of irreparable harm would also be both over- and under-inclusive, when the money damages case proceeds or the irreparable harm case gets dismissed. Any balancing test risks getting the outcome “wrong” in any particular case, and those balancing tests may or may not have a bias toward over- or under-inclusiveness.
Any proposal for changes or an increase in uniformity regarding standards in discovery stay cases should be focused primarily on striking the appropriate balance between over- and under-inclusiveness in the standards. The goal is not necessarily minimizing the error, because solutions to do so may be either unworkable in practice or may create an additional burden on the courts and the parties. Thus the inquiry becomes how much error the standards should tolerate, how the error is allocated between the over- and under-inclusive side, and what circumstances justify changes in the standards applied to either decrease the amount of error or to reduce the burden on the courts and the parties.
As noted previously, the “balancing of factors” and “preliminary peek” tests currently applied by various courts largely utilize the same or similar factors in weighing whether or not to stay discovery, with the “preliminary peek” test using the additional factor of the likelihood of success of the motion to dismiss. This additional factor allows courts to reduce their error rate, such that they reach the “right” result more often. Courts applying the “preliminary peek” test are less likely to stay discovery when a motion to dismiss will fail or to allow discovery to proceed when the motion to dismiss will succeed. However this reduction in error rate comes at a cost because the court must take the additional step of prejudging the motion to dismiss.[104] Additional burden and error are introduced when different judges decide the motion to stay discovery and the motion to dismiss. Because the other factors that the two tests consider are largely the same, applying one or the other test should not materially impact the bias toward either over- or under-inclusiveness. Therefore, the decision of which test to apply amounts to a decision of how much risk of error should be tolerated. Thus a balancing of factors approach is appropriate when the risk is relatively low, such as when the burdens of discovery are not high or the risk of delay can be fully compensated later. The preliminary peek test becomes appropriate as the risks increase, such as when discovery is particularly complex or the risks of delay become unacceptable.
A blanket rule automatically staying discovery when a motion to dismiss is filed is not appropriate.[105] Such a rule would be unfair to plaintiffs who wish to proceed with discovery, would impair judges’ ability to effectively control their dockets, and would be inconsistent with the Federal Rules, which allow a judge to stay discovery under Rule 26(c) but require some showing of good cause. To the extent that the Supreme Court or circuit courts have made statements that would seem to support an automatic stay rule, those opinions are properly seen as limited by their context to cases involving immunity concerns or a particularly high risk of unnecessary and overly burdensome discovery.
Automatic stays of discovery are not even appropriate in limited circumstances, such as is already provided for in qualified immunity or shareholder derivative cases. The automatic stay in shareholder derivative suits is commonly viewed as an attempt by Congress to limit those suits, and correctly so.[106] The judge-made rule to stay discovery in qualified immunity suits similarly serves to create additional barriers to bringing an action against government officials.[107] Yet in both sets of cases, motions to dismiss are denied, resulting in unnecessary delays in discovery. An automatic stay provision encourages defendants to file motions to dismiss, increasing the likelihood that nonmeritorious motions are filed. A preliminary peek standard applied to those cases would allow judges to weed out the nonmeritorious motions and allow discovery to proceed while still minimizing the risks of unnecessary discovery in cases that either Congress or the courts have determined are of particular concern.
In addition to increasing the efficiency of the system by focusing judicial resources on the cases involving the most significant risk of harm to either or both of the parties, my proposal also brings additional benefits due to the increase in transparency. Transparency will be increased as judges make explicit, in writing, which tests they are applying and how they are weighing the competing interests at stake. This increase in transparency will be helpful to other judges who can rely upon the collective wisdom of their peers when exercising their discretion. The accretion of judicial precedent on this issue—particularly if the decisions are consistent with each other in applying the same test in similar situations—will lead to greater uniformity of decision making across the country. This uniformity will increase fairness and will also lessen any risk of forum shopping.
Transparency in decision making on motions to stay discovery will also provide tangible benefits to parties. When parties know what standards will be applied to motions to stay discovery, they can focus their efforts on identifying the relevant facts and presenting those to the court. Transparency in terms of how the standards will be applied to specific fact situations will also decrease the number of disputes over discovery stays, particularly in the extreme cases. As a result, plaintiffs will be less likely to oppose motions to stay discovery when there is no risk of irreparable harm and the costs of discovery will be great, because they will know that the judge is unlikely to allow discovery to proceed. Conversely, defendants will have less incentive to seek a stay of discovery if they know there is a risk of irreparable harm and their motion to dismiss will not entirely dispose of the case. Thus, particularly over time, both courts and the parties will expend fewer resources on clear-cut cases, allowing courts to spend more resources managing novel cases or those where the risks of harm are unusually high.
My proposal is that courts should adopt the standard, already in place in some districts, that stays of discovery are generally disfavored and only justified where the movant has demonstrated good cause. How the court determines what establishes good cause is context dependent. In some situations, the court should take a preliminary peek at the motion to dismiss and use the likelihood of the motion being granted or denied to balance the risks associated with a stay of discovery. However, outside of certain clearly-defined contexts, judges should avoid prejudging the motion to dismiss and should instead balance the competing interests without taking a preliminary peek at the motion to dismiss.
Under what circumstances should the judge take a preliminary peek at the motion to dismiss? The considerations discussed in Part III provide a guide for addressing this question. The flowchart emphasizes that a disfavored stay is the starting point for judges faced with a motion to stay discovery due to a pending motion to dismiss. Magistrate judges should only apply the balance of interests test, regardless of other factors of the case. District court judges should employ the preliminary peek test in those cases where the risk of error is particularly great, due to the complexity of the case, the risk of irreparable harm, or the need to determine threshold issues. In all other cases, district judges should apply the balancing of interests test. An automatic stay is only appropriate in the limited circumstances dictated by Congress or the Supreme Court—qualified immunity and shareholder derivative suits.
Figure 1: Process for Deciding Which Test to Apply on Motions to Stay Discovery
1. Factors Insufficient to Justify a Preliminary Peek
The following factors are ones that do not, by themselves, justify the additional burden of the preliminary peek test. Instead, these factors should be considered, if at all, in either balancing test and might weigh for or against a stay of discovery. Judges should be clear about the extent they rely upon these, or other, factors, for the benefit of the parties and the integrity of the judicial system. Explicit discussion of the applicable factors will provide clarity to the parties and remove uncertainty around whether a stay will be granted. This discussion will also provide guidance for parties in future cases to tailor their arguments such that the judges actually receive the most relevant information, thereby reducing judicial speculation as well as the risk of error.
A judge should not use the type of case, even if it is an unusual case for the judge to hear, as a basis for taking a preliminary peek at the motion to dismiss. Nor should a judge use her view of a particular class of cases as a factor in either balancing test that is applied. Thus, a judge should not apply her notion of which types of cases are usually meritorious (or not) as a basis for prejudging the motion to dismiss or for granting a stay of discovery. Instead, the court should balance the interests without prejudging, unless another consideration justifies taking a preliminary peek. If any additional classes of cases warrant application of an auto-stay rule, that decision should come from either Congress or from the appellate courts, given the conflict that the auto-stay rule creates with the discovery system under the Federal Rules of Civil Procedure.
If the party seeking the stay asserts that the motion to dismiss would completely resolve the case, then the judge should look to the motion to dismiss to confirm whether this is true, but should only consider the likelihood of success of the motion to dismiss if another factor in the case justifies the use of the preliminary peek approach. Similarly, for motions to dismiss that would only partially dispose of the case, the judge should balance the interests independent of any prejudging of the motion to dismiss. And furthermore, any stay of discovery that is granted should only affect discovery related solely to the disputed claims.
The status of the government as a party to the action should not affect which standard applies, except in the situation where the government party asserts qualified immunity as a defense. Many public law cases would involve either an administrative record, and thus limited or no discovery, or the government as a plaintiff. The standard judges should apply therefore depends not so much on whether the government is a party but on other considerations, such as the complexity of discovery or the ability to preserve evidence. A government plaintiff, however, should not be treated differently than any other plaintiff for the purposes of deciding a motion to stay, whether the government supports or opposes the stay.
The final factor, settlement likelihood, should not affect the standard that is applied, but it should be included in both balancing tests. If the court thinks that settlement is relatively likely once the parties proceed to discovery and begin to narrow the scope of issues, then a stay of discovery is less appropriate. Quick settlement will help reduce the burdens on courts and will help the parties avoid the costs of delay and uncertainty. And if the parties are indeed likely to settle the case, then the risk of error in deciding the motion to stay discovery becomes lower. Therefore, the additional burden created by the preliminary peek is not justified by the settlement likelihood but instead must be justified by another factor in the case. Both approaches can take account of settlement likelihood when weighing the costs and burdens of discovery against those of delay.
2. Factors Sufficient to Justify a Preliminary Peek
Because I argue that the presumptive test to apply in deciding motions to stay discovery should be a balancing of factors without prejudging the motion to dismiss, certain conditions must exist in a given case to justify application of the more intensive preliminary peek approach. Any one of the factors discussed below is sufficient, but not necessary, to justify application of the preliminary peek test. The only factor I have identified as necessary for the preliminary peek approach is discussed in the next Subpart.
Where the motion to dismiss involves threshold issues such as jurisdiction, venue, immunity, or standing, the judge should take a preliminary peek at the motion to dismiss to inform the balancing of harms that might result from a grant or denial of a stay. For those cases where jurisdiction is challenged, a preliminary peek at the motion to dismiss will also enable the judge to allow discovery on matters relevant to establishing whether jurisdiction is proper, so that any stay that is entered would apply only to matters unrelated to jurisdiction.
The remedy sought would also affect what standard should be required to show good cause. There is greater risk of harm due to delay when a plaintiff seeks equitable relief because delay might result in irreparable harm. In such a case, the party seeking to stay discovery, typically (but not exclusively) the defendant, would need to show that he is likely to prevail on the motion to dismiss and make the discovery unnecessary. This can be seen as the converse of the preliminary injunction standard, where the defendant’s actions are only enjoined if the plaintiff can show she is likely to prevail on the merits. Here, the defendant (typically) can only avoid discovery and thereby delay the case if he can show that he is likely to prevail on the motion to dismiss. Where a party is seeking only money damages, the interest is not irreparable,[108] and thus the judge need not take a preliminary peek and should instead balance the competing interests without prejudging the motion to dismiss.
Finally, the complexity of the case (as a proxy for the burden of discovery) is another factor that justifies a preliminary peek at the motion to dismiss. Although the complexity of the case is taken into account under both balancing approaches, the risk of error increases along with the complexity of the case. Once the case becomes sufficiently complex, it justifies the use of the court’s time to conduct a preliminary peek analysis. But for standard run-of-the-mill cases without extraordinary discovery needs, the court’s time is better spent applying the balancing of factors test.
3. One Factor Necessary to Apply the Preliminary Peek Test
The preliminary peek test should only be applied by a judge who will ultimately decide the motion to dismiss. While the previous factors discussed either were or were not sufficient to justify the application of the preliminary peek test, this final factor is not sufficient but it is necessary.
If a motion to stay discovery is being decided by a magistrate judge and the parties have not consented to have the magistrate decide dispositive motions, then the magistrate should not take a preliminary peek at the merits and should avoid prejudging the motion to dismiss, which must be decided by the Article III judge. However, if the parties consent to have the magistrate decide their case, then the magistrate may apply the preliminary peek approach, if another consideration justifies it. A district judge may apply either approach, as justified by other considerations.
Three main reasons support this conclusion. First, because the parties have not consented to have the magistrate decide dispositive motions such as a motion to dismiss, the magistrate should base rulings on a motion to stay discovery on factors independent of the merits of the motion to dismiss. This both protects the rights of the parties and also falls within the core competency of magistrate judges, who are primarily responsible for settling discovery and scheduling disputes in civil cases.[109] Second, the additional burden imposed on the court by evaluating the merits of the motion to dismiss is greater for a magistrate than for a district judge—preliminary review by a district judge can inform the later, ultimate decision, whereas the magistrate’s preliminary review would only assist resolution of the motion to stay discovery. Finally, the error reduction aspects of the preliminary peek approach are also diminished when different judges decide the motion to stay discovery and the motion to dismiss. Magistrate judges are more likely to have a different preliminary evaluation of the motion to dismiss than the district judge who will ultimately decide it, and so the benefit of the more burdensome test is reduced.[110]
V. Broader Issues
A. “Discovery Abuse”
Any discussion of the costs of unnecessary discovery naturally raises the issue of whether the discovery process can be used strategically to threaten the imposition of burdensome costs on a party in an effort to get it to agree to a less-favorable settlement. This type of situation is commonly referred to as “discovery abuse.”
Critics of the modern discovery system decry what they term “wild fishing expeditions” that intrude on individual privacy, impose high costs on litigants, and use the discovery process unfairly to pressure parties into settlement.[111] Concerns over “discovery abuse” are not limited to scholars and practicing attorneys, however, as judges also cite it as a concern or as a justification for their rulings.[112]
However, the view of “discovery abuse” as a common and serious problem is not shared by all.[113] Indeed, reliable data regarding the costs of discovery are difficult to come by,[114] and instead proponents of reform tend to point to the extreme examples when making their case.[115] My proposal allows judges to prevent discovery abuses on a case-by-case basis, but it resists the calls by some who seek broader reforms to prevent “discovery abuse.” Alarms raised over “discovery abuse” do not provide a sufficient basis for expanding auto-stay rules to a broader set of cases. Instead, by properly focusing judicial attention on the cases that present the greatest risk of “discovery abuse,” my proposal effectively deals with the extreme cases while retaining the default of allowing discovery to proceed in run-of-the-mill cases. As a result, my proposal need not address the underlying issue of how much discovery is appropriate in given case, leaving that to the parties and the judges to decide.
B. Existing Auto-Stay Rules
Uniquely compelling circumstances led to the adoption of auto-stay rules for cases involving qualified immunity defenses and shareholder derivative suits. In those cases, either the Supreme Court or Congress acted to address what was viewed as excessive risk from unnecessary discovery in particular classes of cases.[116] Lower courts should respect those decisions by enforcing automatic stay rules only in that limited subset of cases. Outside of these two contexts, judges should appropriately exercise their discretion to decide discovery stays on a case-by-case basis and resist any temptation to broaden the application of automatic stay rules.
The concerns that led Congress to enact the Private Securities Litigation Reform Act[117] are not present in other classes of cases, and certainly not in all civil cases in our federal courts. Similarly, the extreme facts presented by Iqbal,[118] or even by more ordinary qualified immunity cases,[119] are not present in the vast majority of cases. Instead, as Professor Hartnett has argued, a broad rule requiring a stay of discovery whenever a motion to dismiss is filed should be rejected.[120] My proposal would limit the impact of the Iqbal decision’s broad statements regarding discovery prior to resolution of any motions to dismiss, and it would empower judges to stay discovery when appropriate without unduly presenting a barrier to litigants seeking to enforce their rights through the courts.
C. Difficulties with My Proposal
Although my proposal does not solve all of the difficult issues related to discovery stays and motions to dismiss, it does represent a positive step forward over the current state of the law. The proposal still relies on district and magistrate judges to exercise their discretion, but such reliance is justified based on the experience that those judges have in managing the discovery process. Furthermore, this reliance on judicial discretion also allows the system to retain flexibility to address discovery stays on a case-by-case basis to reduce the risk of harm due to either delay or unnecessary discovery costs.
Although reliance on judicial discretion does sacrifice predictability, that concern will be lessened by the greater consistency and transparency under my proposal. When judges explicitly state what factors they are considering, and how different factors affect their weighing of competing interests, they thereby provide helpful signals to other judges and to parties regarding when a discovery stay is appropriate.
My proposal does not address other concerns related to the length of time taken for motions to dismiss to be resolved. It does not address the underlying costs of discovery or any methods for increasing the efficiency of that process.
One potential concern is that by encouraging judges to take a “preliminary peek” at a motion to dismiss, this will encourage a “lock-in” effect, such that judges are more likely to confirm their preliminary view of the motion to dismiss, regardless of whether a more considered evaluation later might change their minds.[121] The lock-in effect has been discussed in other areas of the law,[122] and to a greater extent in the economics literature.[123] While I believe that this concern does have some merit, nevertheless I think the preliminary peek is appropriate to use in those cases where the risk of harm is relatively high. Any potential lock-in effect is mitigated by the fact that, properly applied, the preliminary peek test does not require the judge to decide if the motion to dismiss should be granted, but rather simply allows the judge to “tip the scales” on the motion to stay based on a rough assessment of the strength of the motion to dismiss.
One final difficulty is that the Supreme Court explicitly rejected a “careful case management” approach taken by the lower courts to limit the intrusiveness of discovery in Iqbal. My proposal, arguably, represents an example of the careful case management approach. However, as discussed above, Iqbal does not preclude my proposal, and should not influence judges to more often apply an auto-stay standard when a motion to dismiss is filed.[124]
Overall, my proposal represents important progress toward improving judicial management of the discovery process in federal civil litigation. The framework my proposal provides will enable judges to rationally decide what standard is most appropriate based on the particular details of the case before them. The selective use of the preliminary peek approach will enable courts to reduce the risk of error where that risk is more significant while conserving scarce judicial resources when the risk of error is not as great. Greater transparency by judges in deciding to grant a discovery stay will also promote consistency and will provide useful guidance to the parties seeking or opposing a discovery stay in their own cases.
Conclusion: Areas for Further Inquiry
My investigation into this issue thus far has raised more questions than it has answered. How frequently are motions to stay discovery filed subsequent to the filing of a motion to dismiss? How frequently are discovery stays granted? Do these frequencies vary by jurisdiction, by judge, or by type of case? Are discovery stays sought at higher rates in cases or jurisdictions where motions to dismiss are more likely to be either filed or granted? Conversely, are motions to dismiss more likely to be filed in jurisdictions or in cases where discovery stays are likely to be granted? To address these and related questions, a closer quantitative analysis of court dockets is required. I intend to investigate these and other related issues.
A subsequent study will also delve deeper into the impact of the Iqbal decision, going beyond direct citations by judges in deciding motions to stay. Some studies have already investigated whether Iqbal has impacted the rates of motions to dismiss being filed and their being granted or denied.[125] One study by staff at the Federal Judicial Center indicates a general increase in rates of filing of motions to dismiss for failure to state a claim.[126] The same study did not find an increase in the rate of grants of such motions to dismiss.[127] My research will supplement this work by extending the analysis to cover impacts on the filing of motions to stay and their ultimate success. I have identified several questions for further study. Are parties more likely to file motions to stay due to pending motions to dismiss after Iqbal? Are courts more likely to grant discovery stays post-Iqbal? To what extent can differences in the likelihood of motions to stay being granted be used to explain any changes in the prevalence of motions to dismiss?
* Environmental Law Clinic Fellow, University of Denver Sturm College of Law, [email protected]. I would like to thank those who provided feedback on earlier drafts of this Article, especially Tammy Kuennen, Brittany Glidden, Lindsey Webb, Eric Franklin, Alan Chen, and Beto Juárez, as well as the participants in the 2010 Clinical Law Review Writer’s Workshop.
[1]. Reliable data regarding the frequency of cases in which a stay of discovery occurs are lacking. I intend, as discussed below in the Conclusion, to address some of the many empirical questions raised by this issue in the future. However, given the large number of published and unpublished opinions available by searching Westlaw or other commercial databases, it is reasonable to assume that motions to stay discovery are fairly common in cases where a motion to dismiss is filed.
[2]. Estimates for litigation costs associated with discovery vary, yet the estimates typically assert that more than half of all litigation costs are due to discovery. In 1999, the Administrative Office of the U.S. Courts estimated that discovery represented 50% of the litigation costs in an ordinary case, and up to 90% in cases where discovery is actively used. Judicial Conference Adopts Rules Changes, Confronts Projected Budget Shortfalls, Third Branch (Admin. Office of the U.S. Courts, D.C.) Oct. 1999, at 1, 2–3, available at http://www.uscourts.gov/News/TheThirdBranch/99‑10‑01/Judicial_Conference
_Adopts_Rules_Changes_Confronts_Projected_Budget_Shortfalls.aspx. The common view is that this percentage has only increased as more information is available through e-discovery. See, e.g., Shira A. Schiendlin & Jeffrey Rabkin, Electronic Discovery in Federal Civil Litigation: Is Rule 34 Up to the Task?, 41 B.C. L. Rev. 327, 341 (2000) (referring to an “explosion of electronic evidence”); see also Edward D. Cavanagh, Twombly, The Federal Rules of Civil Procedure and the Courts, 82 St. John’s L. Rev. 877, 887 (2008) (noting that courts have authority under the Rules to “rein in potentially expensive e-discovery”). Some even assert that the costs of litigation exceed the amount in controversy in all but the largest cases. See Lawyers for Civil Justice, Civil Justice Reform Grp. & U.S. Chamber Inst. for Legal Reform, Litigation Cost Survey of Major Companies 2 (2010).
[3]. See the discussion at infra Part V.A.
[4]. John S. Beckerman, Confronting Civil Discovery’s Fatal Flaws, 84 Minn. L. Rev. 505, 513–14 (2000).
[6]. See, e.g., Fed. R. Civ. P. 26(f), 37(a)(1).
[7]. A decision on a motion to stay discovery would be “wrong” or would create “error” when it either stays discovery in a case that ultimately survives the motion to dismiss (a “false positive”) or when it allows discovery in a case where the motion to dismiss is ultimately granted (a “false negative”).
[9]. Fed. R. Civ. P. 26(c)(1)(A); see, e.g., Brazos Valley Coal. for Life, Inc. v. City of Bryan, 421 F.3d 314, 327 (5th Cir. 2005).
[10]. Stays of discovery might occur or be requested outside of the context where a motion to dismiss is pending, such as when a motion for summary judgment has been filed or when some nondispostive matter like a denied intervention request has been appealed. However, for the purposes of this Article, when I refer to “discovery stays” or “motions to stay discovery,” I am limiting my discussion to instances where a motion to dismiss is pending unless otherwise noted.
[11]. Two recent articles address the subject of discovery when a motion to dismiss is pending, although they do not analyze the standards that judges use to decide whether or not to grant a stay of discovery. See Edward A. Hartnett, Taming Twombly, Even After Iqbal, 158 U. Penn. L. Rev. 473 (2010); David L. Noll, The Indeterminacy of Iqbal, 99 Geo. L.J. 117, 141–43 (2010) (mentioning a few district court decisions addressing discovery stays). A search of the JLR database on Westlaw reveals many articles discussing stays of discovery in the securities litigation context, discussed infra Part II.A.1, which is controlled by statute and not the Federal Rules of Civil Procedure, as well as a few other special cases such as antitrust or civil cases with related criminal proceedings pending. I have been unable to find any broad discussion of stays of discovery in federal civil litigation that examine the standards judges use to exercise their discretion. Even civil procedure textbooks largely do not cover this topic, and treatises on civil procedure provide only a cursory treatment. See, e.g., 8A Charles Alan Wright, Arthur R. Miller & Richard L. Marcus, Federal Practice and Procedure § 2046.1 n.15 (3d ed. 2010). It seems that the first time a young lawyer would encounter this situation is when it comes up in one of her cases, and in that situation the lawyer would not have any good resources to turn to that attempt to make sense of the variety of standards that judges apply, if any are explicitly applied at all.
[12]. See James S. Kakalik et al., Discovery Management: Further Analysis of the Civil Justice Reform Act Evaluation Data, 39 B.C. L. Rev. 613, 637 (1998) (“The average lawyer work hours per litigant is 232 hours, of which an average of 36%, or 83 hours, is spent on discovery, including discovery motions.”).
[13]. See Edward Brunet, Debunking Wholesale Private Enforcement of Environmental Rights, 15 Harv. J.L. & Pub. Pol’y 311, 314 (1992) (noting the inevitable increase in discovery costs in environmental litigation due to the need for expert witnesses in this area).
[14]. See generally Robert D. Cooter & Daniel L. Rubinfeld, Reforming the New Discovery Rules, 84 Geo. L.J. 61, 61–62 (1995) (proposing cost-shifting reform to discovery rules that internalizes discovery costs and promotes efficiency in the discovery process).
[15]. Scott A. Moss, Litigation Discovery Cannot Be Optimal but Could Be Better: The Economics of Improving Discovery Timing in a Digital Age, 58 Duke L.J. 889, 894 (2009) (noting that “e-discovery can cost tens or hundreds of thousands of dollars in even fairly typical cases”).
[16]. Judith A. McKenna & Elizabeth C. Wiggins, Empirical Research on Civil Discovery, 39 B.C. L. Rev. 785, 797 (1998).
[17]. See, e.g., Emery G. Lee III & Thomas E. Willging, Defining the Problem of Cost in Federal Civil Litigation, 60 Duke. L.J. 765, 769–76 (2010).
[18]. See, e.g., Brookings Inst., Justice for All: Reducing Costs and Delay in Civil Litigation: Report of a Task Force (1989).
[19]. Delays increase the costs associated with litigation, which discourages injured plaintiffs from filing suit. Additionally, the notion that “justice delayed is justice denied” rings true for many because delay reduces the value of winning a civil suit, increases the burden of losing, and imposes opportunity costs. Carrie E. Johnson, Rocket Dockets: Reducing Delay in Federal Civil Litigation, 85 Cal. L. Rev. 225, 230–31 (1997).
[20]. The monthly percentage of motions to dismiss granted compared to motions to dismiss filed was typically less than 40% and never more than 47% from January 2007 to October 2009. See Motions to Dismiss, Admin. Office of the U.S. Courts (Apr. 13, 2010), http://www.uscourts.gov/uscourts
/RulesAndPolicies/rules/Motions%20to%20Dismiss_042710.pdf.
[21]. An interesting example can be seen in some of the litigation that has arisen out of the oil spill in the Gulf of Mexico. An environmental group challenged the Department of Interior’s position that excluded individual offshore drilling projects from a complete review under the National Environmental Policy Act. A stay of discovery was sought by both the environmental group and the Department of Interior in order to give them time to narrow the claims through settlement discussions. Plaintiff’s and Federal Defendants’ Joint Motion for Stay, Center for Biological Diversity v. Salazar, No. 1:10-cv-00816-HHK (D.D.C. Aug. 17, 2010), ECF No. 29. Industry groups opposed the stay based on their desire to see the uncertainty created by the litigation resolved as quickly as possible. Intervenors’ Response to Joint Motion for Stay, Center for Biological Diversity v. Salazar, No. 1:10-cv-00816-HHK (D.D.C. Aug. 20, 2010), ECF No. 31. The court granted the stay until October 29, 2010, but this situation shows how the interests of parties who are not plaintiffs or defendants might be implicated in decisions on discovery stays. See Plaintiff’s and Federal Defendants’ Joint Motion for Extension of the Temporary Stay at 2, Center for Biological Diversity v. Salazar, No. 1:10-cv-00816-HHK (D.D.C. Sept. 28, 2010), ECF No. 40.
[22]. Justice Cardozo stated:
[T]he power to stay proceedings is incidental to the power inherent in every court to control the disposition of the causes on its docket with economy of time and effort for itself, for counsel, and for litigants. How this can best be done calls for the exercise of judgment which must weigh competing interests and maintain an even balance.
Landis v. N. Am. Co., 299 U.S. 248, 254–55 (1936).
[23]. See, e.g., String Cheese Incident, LLC v. Stylus Shows, Inc., No. 1:02-CV-01934-LTB-PA, 2006 WL 894955, at *2 (D. Colo. Mar. 30, 2006).
[24]. Fed. R. Civ. P. 26(c).
[25]. See, e.g., Wood v. McEwen, 644 F.2d 797, 801 (9th Cir. 1981); In re FirstEnergy S’holder Derivative Litig., 219 F.R.D. 584, 587 (N.D. Ohio 2004).
[26]. Conceivably, a court could also issue an order staying discovery in a case sua sponte, although I have not come across such a situation, at least in the general context. However, it is not difficult to conceive of situations where this might occur, such as when a magistrate judge issues a report and recommendation on a motion to dismiss or is faced with a motion to compel or for a protective order. The court’s own interest in avoiding presiding over discovery disputes, the public’s interest, or the court’s dim view of a case’s prospects for surviving the motion to dismiss might all support a court in issuing a stay of discovery of its own accord.
[27]. Stays of discovery also need not be granted after a written motion is filed with the court or even after an oral motion. Stays of discovery can effectively be determined by judicial approval of a scheduling order that either keys discovery deadlines off of resolution of the motion to dismiss, or simply sets deadlines sufficiently far out that the parties need not commence discovery until after the motion to dismiss has been decided. Stays of this kind will almost always be affected without a written opinion from the court explaining the reasons why a stay was or was not granted in a particular case. Such reasons could conceivably be discussed by the magistrate judge during the scheduling conference, although I expect that in most of those instances, only the judge knows what factors, if any, were considered in reaching her decision.
[28]. See, e.g., Jarvis v. Regan, 833 F.2d 149, 155 (9th Cir. 1987).
[29]. See, e.g., Tamburo v. Dworkin, No. 04 C 3317, 2010 WL 4867346, at *1–2 (N.D. Ill. Nov. 17, 2010).
[30]. Harlow v. Fitzgerald, 457 U.S. 800, 818 (1982); see also Ashcroft v. Iqbal, 129 S. Ct. 1937, 1953–54 (2009).
[31]. See Private Securities Reform Act of 1995, Pub. L. No. 104-67, § 101(a), 109 Stat. 737, 741 (codified as amended at 15 U.S.C. 77z-1(b) (2006)).
[32]. See Steinbuch v. Cutler, 518 F.3d 580, 588–89 (8th Cir. 2008). Only in rare cases will appellate courts overturn a stay of discovery. See, e.g., Clinton v. Jones, 520 U.S. 681, 693–96, 706–09 (1997) (failing to grant then–President Clinton immunity for unofficial conduct and finding that “a lengthy and categorical stay takes no account whatever of the respondent’s interest in bringing the case to trial”).
[33]. See, e.g., Ameritel Inns v. Moffat Bros. Plastering, L.C., CV 06-359-S-EJL, 2007 WL 1792323, at *2 (D. Idaho June 20, 2007); Feldman v. Flood, 176 F.R.D. 651, 652 (M.D. Fla. 1997).
[34]. Professor Hartnett has pointed this out as well: “Not only do district courts have broad and largely unreviewable discretion regarding the scope of discovery prior to a decision on a 12(b)(6) motion, but that discretion is frequently exercised in chambers, with scant (if any) explanation of the basis for the decision.” Hartnett,supra note 11, at 514. I have conducted my own preliminary research to find appellate cases that discuss both discovery stays and motions to dismiss in the same paragraph. Out of 5918 cases returned from a search of the ALLFEDS database in Westlaw using the terms “‘motion to dismiss’ & stay! /2 discovery” (as of March 9, 2012), only 626 cases were in the Courts of Appeals and 7 in the Supreme Court. I have not yet conducted an analysis to determine what portion of those appellate decisions resulted in the court overturning the district court’s decision as an abuse of discretion, but based on my initial review of the cases, I suspect it will be extremely low.
[35]. See, e.g., O’Donnell v. United States, 112 Fed. App’x 36, 37 (Fed. Cir. 2004) (finding that appeal of an order staying discovery pending resolution of a motion to dismiss does not fall under the collateral order doctrine for interlocutory appeals).
[36]. I am aware of at least two cases where this occurred, based on my experience in clinical practice, and I assume that motions to stay are frequently decided by other courts without issuing a written opinion. When no record of the judge’s reasoning exists, it is impossible to say what standards were applied. Whether a more complete record would indicate greater or lesser uniformity of standards is not known. It is theoretically possible to determine just how frequently a motion to stay is decided without a record of the judge’s reasoning, based on a review of docket entries in the PACER database, but such analysis is outside the scope of this Article.
[37]. The parties might agree to stay discovery until the motion to dismiss is decided. I am aware of at least one such case, in which I am an attorney of record. SeeJoint Motion to Vacate Scheduling Conference and Requirement that Parties Submit Confidential Settlement Statements at 2, WildEarth Guardians v. Pub. Serv. Co. of Colo., 805 F. Supp. 2d 1134 (D. Colo. 2011) (No. 1:09-cv-01576). The judge however might also decide a conflict between the parties in issuing a scheduling order, and this could be the equivalent of deciding a motion to stay. However, a judge is even less likely to issue a written order explaining his decision making with regard to a scheduling order than the same judge would be in deciding a motion to stay discovery.
[38]. According to the 2010 Annual Report of the Administrative Office of U.S. Courts, 1105 of 282,895 cases commenced during the twelve-month period ending September 30, 2010, were classified as “Securities, Commodities, and Exchanges” cases based on federal question jurisdiction. Admin. Office of the U.S. Courts, Judicial Business of the U.S. Courts 144 tbl.C-2 (2010), available at http://www.uscourts.gov/uscourts/Statistics/JudicialBusiness/2010
/JudicialBusinespdfversion.pdf. That amounts to less than one-half of one percent of all federal cases.
[39]. Congress did however leave courts with the discretion to lift the stay when necessary to preserve evidence or prevent undue prejudice to a party. 15 U.S.C. § 77z-1(b)(1), 78u-4(b)(3)(B) (2006).
[40]. Richard H. Walker et al., The New Securities Class Action: Federal Obstacles, State Detours, 39 Ariz. L. Rev. 641, 641 (1997).
[41]. Hillary A. Sale, Heightened Pleading and Discovery Stays: An Analysis of the Effect of the PSLRA’s Internal-Information Standard on ‘33 and ‘34 Act Claims, 76 Wash. U. L. Q. 537, 552–53 (1998) (internal quotation marks omitted).
[42]. 457 U.S. 800 (1982).
[43]. See id. at 818; see also Hartnett, supra note 11, at 511 (noting that the Supreme Court has not extended auto-stay beyond qualified immunity).
[44]. Indeed, that case involved “extensive discovery” and the appeal was made following a ruling on motions for summary judgment. Harlow, 457 U.S. at 802–06.
[46]. 129 S. Ct. 1937 (2009).
[49]. Id. at 1945 (citing Iqbal v. Hasty, 490 F.3d 143, 179 (2d Cir. 2007) (Cabranes, J., concurring)). Both Judge Cabranes and the Supreme Court noted that this concern was in the context of a defendant who was a high-ranking government official, entitled to assert qualified immunity, who was dealing with “a national and international security emergency unprecedented in the history of the American Republic.” Id. at 1953 (quoting Hasty, 490 F.3d at 179 (Cabranes, J., concurring)). This national security issue may well provide sufficient justification for lower courts to limit the applicability of the statements from Iqbal to other contexts that do not implicate terrorism.
[50]. See, e.g., Robin J. Effron, The Plaintiff Neutrality Principle: Pleading Complex Litigation in the Era of Twombly and Iqbal, 51 Wm. & Mary L. Rev. 1997 (2010); Adam Liptak, Case About 9/11 Could Lead to a Broad Shift on Civil Lawsuits, N.Y. Times, Jul. 21, 2009, at A10; Lyle Denniston, Analysis: New Obstacles to Wartime Challenges, SCOTUSblog, (Jul. 4, 2009, 4:09 PM), http://www.scotusblog.com/2009/07/analysis‑new‑obstacles‑to‑wartime‑challenges; see also Iqbal Project, Public Justice, http://www.publicjustice.net/Key
-Issues-Cases/Access-To-Justice/Iqbal-Project.aspx (last visited Mar. 9, 2010) (noting that some defendants now argue that many statements are too conclusory to satisfy the new pleading standards). As of March 2012, a search of the JLR database on Westlaw returns 219 entries that contain either “Twombly” or “Iqbal” in the title, and surely far more discuss those cases in some detail.
[51]. The Supreme Court’s pronouncements in Iqbal, particularly the statements about the high burden of discovery and the suspicion with which claims are viewed pending motions to dismiss, have the potential to (1) encourage judges to more often automatically stay discovery; (2) balance harms after a preliminary peek at the motion to dismiss; and (3) alter the balance among the five factors (discussed infra at Part II.B) for resolving a motion to stay without prejudging the motion to dismiss. An expansive reading of Iqbal could lead courts to place more emphasis on the burdens of discovery on defendants or on the potential for expending resources (by the defendants or the court) on discovery activities and disputes in a case that ultimately may be dismissed. This would have the effect of diminishing the interests of plaintiffs in proceeding expeditiously with their cases in the balancing of factors. Judges, particularly the magistrate judges who typically decide motions to stay, may be influenced by their own views of the likelihood of success of the motions to dismiss. Such assessments of the motions to dismiss would amount to courts essentially adding a sixth factor to the balancing test, perhaps doing so without acknowledging its potentially significant impact. Whether courts will ultimately alter their approaches to resolving a motion to stay—either explicitly citing to Iqbal or giving effect to the decision in an unacknowledged manner—remains to be seen. More time is needed for the effects of Iqbal to become fully apparent in lower court decisions and in briefing by litigants. The risk, however, is present and very real. And as I argue in Part IV.B, a better approach is to disfavor a stay unless a good cause can be shown.
[52]. The ALLFEDS database was queried using the search terms “iqbal /p stay /p discovery.” This search was last run on March 9, 2012, producing the twenty-seven relevant matches.
[53]. Indeed, the one circuit court decision available to date agreed that the district court properly stayed discovery while a motion to dismiss was pending, noting that “[i]n certain circumstances it may be appropriate to stay discovery while evaluating a motion to dismiss where, if the motion is granted, discovery would be futile.” Mann v. Brenner, 375 Fed. App’x 232, 239 (3d Cir. 2010) (citing Iqbal, 129 S. Ct. at 1954). One magistrate judge in the Southern District of Ohio has authored three of the twenty-one cases citing Iqbal in this context, and he has noted that “[t]he Supreme Court’s recent decision in [Iqbal] will undoubtedly give the Court more opportunities to consider” whether “to stay discovery during the pendency of some type of defensive motion.” Charvat v. NMP, LLC, No. 2:09-cv-209, 2009 WL 3210379, at *1 (S.D. Ohio Sept. 30, 2009). The same magistrate stated that Iqbal “breaks no new ground” on this issue. Id. Other courts have reached the conclusion that Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), and Iqbal do not require a stay of discovery every time a motion to dismiss is filed. See, e.g., Tamburo v. Dworkin, No. 04 C 3317, 2010 WL 4867346, at *2 (N.D. Ill. Nov. 17, 2010). However, the extent of the ultimate effect of Iqbal remains largely unknown at this time and will remain so until sufficient time passes for the lower courts to react to the decision.
[54]. For a contrary view, see generally Scott Dodson, New Pleading, New Discovery, 109 Mich. L. Rev. 53, 72–88 (2010) (arguing that presuit discovery should be limited in order to “counteract the information asymmetry and overscreening caused by Twombly and Iqbal”).
[55]. See, e.g., Solomon Realty Co. v. Tim Donut U.S. Ltd., Inc., No. 2:08-cv-561, 2009 WL 2485992, at *3 (S.D. Ohio Aug. 11, 2009).
[56]. 123 F.3d 1353 (11th Cir. 1997).
[58]. Id. (footnote omitted) (citation omitted).
[59]. E.g., Redford v. Gwinnett Cty. Judicial Cir., 350 Fed. App’x 341, 346 (11th Cir. 2009) (citing Chudasama and finding that the magistrate did not abuse discretion in staying discovery).
[60]. See, e.g., Moore v. Shands Jacksonville Med. Ctr., Inc., No. 3:09-cv-298-J-34TEM, 2009 WL 4899400, at *1–2 (M.D. Fla. Dec. 11, 2009) (citing Chudasamabut then applying the “preliminary peek” test to determine whether stay should be granted); McCabe v. Foley, 233 F.R.D. 683, 685 (M.D. Fla. 2006) (citingChudasama but saying that challenges to the legal sufficiency of a claim should “often” be resolved before discovery begins, and then applying the “preliminary peek” test).
[61]. See, e.g., Gannon v. Flood, No. 08-60059-CIV, 2008 WL 793682, at *1 (S.D. Fla. Mar. 24, 2008) (noting that Chudasama “does not indicate a broad rule that discovery should be deferred whenever there is a pending motion to dismiss”); S.K.Y. Mgmt. L.L.C. v. Greenshoe, Ltd., No. 06-21722-CIV, 2007 WL 201258, at *1–2 (S.D. Fla. Jan. 24, 2007) (distinguishing Chudasama because this case was “a straightforward commercial case” and instead conducting preliminary review of legal issues in motion to dismiss).
[62]. In some cases the court will not explicitly discuss all of these factors, but may instead only note a few factors that the court deems to be particularly relevant. It is not clear whether the court is actually considering, but dismissing as irrelevant, the other factors, or whether the court is only looking at the factors that it mentions in its written decision. Some courts have listed different factors as well:
In considering whether a stay of all discovery pending the outcome of a dispositive motion is warranted, a case-by-case analysis is required, since such an inquiry is necessarily fact-specific and depends on the particular circumstances and posture of each case. To assist in this determination, the Court is guided by the following factors, none of which is singly dispositive: the type of motion and whether it is a challenge as a “matter of law” or to the “sufficiency” of the allegations; the nature and complexity of the action; whether counterclaims and/or cross-claims have been interposed; whether some or all of the defendants join in the request for a stay; the posture or stage of the litigation; the expected extent of discovery in light of the number of parties and complexity of the issues in the case; and any other relevant circumstances.
Hachette Distrib., Inc. v. Hudson Cnty. News Co., 136 F.R.D. 356, 358 (E.D.N.Y. 1991).
[63]. Springfield Twp. v. Kuss, No. CIV. A. 93-1629, 1993 WL 430421, at *1 (E.D. Pa. Oct. 21, 1993); see also Avalonbay Cmtys., Inc. v. San Jose Water Conservation Corp., No. 07-306, 2007 WL 2481291, at *2 (E.D. Va. Aug. 27, 2007); In re Mid-Atlantic Toyota Antitrust Litig., 92 F.R.D. 358, 359 (D. Md. 1981); Golden Quality Ice Cream Co. v. Deerfield Specialty Papers, Inc., 87 F.R.D. 53, 56 (E.D. Pa. 1980).
[64]. Jackson v. Denver Water Bd., No. 08-cv-01984-MSK-MEH, 2008 WL 5233787, at *1 (D. Colo. Dec. 15, 2008) (staying case could result in delay and attendant “adverse consequences such as a decrease in evidentiary quality and witness availability”).
[65]. See, e.g., Hoxie v. Livingston Cnty., No. 09-CV-10725, 2010 WL 822401, *1 (E.D. Mich. Mar. 4, 2010) (“The wheels of justice would surely grind to a halt if discovery were stayed pending dispositive motions and based on such generic allegations of undue burden and expensive.”); Standard Bank PLC v. Vero Ins., Ltd., No. 08-cv-02127-PAB-BNB, 2009 WL 82494, at *2 (D. Colo. Jan. 13 2009) (“Parties always are burdened when they engage in litigation, whether the case ultimately is dismissed; summary judgment is granted; the case is settled; or a trial occurs. That is a consequence of our judicial system and the rules of civil procedure. There is nospecial burden on the parties in this case.” (emphasis added)).
[66]. S.K.Y. Mgmt. L.L.C. v. Greenshoe, Ltd., No. 06-21722-CIV, 2007 WL 201258, at *1 (S.D. Fla. Jan. 24, 2007) (distinguishing from other cases where there was a risk of “needless and extensive discovery”).
[67]. See Roueche v. United States, No. 09-cv-00048-WDM-BNB, 2010 WL 420040, at *2 (D. Colo. Feb. 1, 2010) (noting “the general interests of controlling the court’s docket and the fair and speedy administration of justice”); Simpson v. Specialty Retail Concepts, Inc., 121 F.R.D. 261, 263 (M.D.N.C. 1988) (noting that motions to stay “are not favored because when discovery is delayed or prolonged it can create case management problems which impede the Court’s responsibility to expedite discovery and cause unnecessary litigation expenses and problems”).
[68]. See, e.g., Feldman v. Flood, 176 F.R.D. 651, 652–53 (M.D. Fla. 1997) (“[I]t is necessary for the Court to ‘take a preliminary peek’ at the merits of the motion to dismiss to see if it appears to be clearly meritorious and truly case dispositive.”); Simpson, 121 F.R.D. at 263 (“It may be helpful to take a preliminary peek at the merits of the allegedly dispositive motion to see if on its face there appears to be an immediate and clear possibility that it will be granted.”).
[69]. An analogy to the test for a preliminary injunction is readily apparent. In both cases, the court looks to make a preliminary assessment of the merits when also considering the potential harm that might result if the requested relief is not granted. However, the discovery stay context is, appropriately, treated differently because of the nature of the interests at stake. First, in the discovery stay context, the risk of harm cuts both ways, or at least the court gives consideration to harm on both sides of the issue. Second, likelihood that the motion to dismiss will be denied is not necessary for discovery to proceed. Thus, the discovery stay example is closer to the “sliding scale” approach to preliminary injunctions favored by the Ninth Circuit, but even so, discovery stays present unique issues not involved when a preliminary injunction is sought. However, the similarities are not irrelevant, and indeed I argue in Part IV, infra, where there is a risk of irreparable harm due to delaying discovery (analogous to the requirement for irreparable harm for a preliminary injunction), then the court should apply a preliminary peek in order to reduce the risk of error.
[70]. Judges do not explicitly say whether they give equal weight to both sides of the balancing, and indeed it is not difficult to imagine that some or all judges might put more weight on one side of the scale or the other, based on their experience on the bench, views regarding the costs of discovery, or views toward particular types of cases or particular parties. But presumably they do attempt to roughly balance the competing interests at least somewhat evenly. However, because the costs of discovery do not share a common metric with the risk of spoliation of evidence or of irreparable harm, judges are essentially comparing apples to oranges, and thus the balancing involves a significant amount of judicial discretion that can be expected to vary from case to case and from judge to judge.
[71]. While extreme, this position would have a sound basis in the Federal Rules of Civil Procedure, which do not provide for a stay of all discovery, but instead allow for courts to issue a protective order upon a showing of good cause. Fed. R. Civ. P. 26(c)(1). However, the Rules do not specifically address any number of actions that courts take on a regular basis, and regardless, I agree with Justice Cardozo’s pronouncement that deciding whether to stay discovery is part of a court’s “inherent authority” to manage the cases before it. Landis v. N. Am. Co., 299 U.S. 248, 254–55 (1936).
[72]. See, e.g., Jackson v. Denver Water Bd., No. 08-cv-01984-MSK-MEH, 2008 WL 5233787, at *1 (D. Colo. Dec. 15, 2008) (“Generally, it is the policy in this district not to stay discovery pending a ruling on motions to dismiss.”); Steil v. Humana Health Care Plans, Inc., No. CIV. A. 99 2541 KHV, 2000 WL 730428, at *1 (D. Kan. May 1, 2000) (“The District of Kansas generally disfavors motions to stay discovery.”).
[73]. See, e.g., Turner Broad. Sys., Inc. v. Tracinda Corp., 175 F.R.D. 554, 556 (D. Nev. 1997) (“[A] pending Motion to Dismiss is not ordinarily a situation that in and of itself would warrant a stay of discovery.”).
[74]. See Steil, 2000 WL 730428, at *1 (“Absent some compelling reason, the Court will not stay discovery.”).
[75]. As an example, a judge in Florida cited the plaintiff’s status as a serial litigator, and one who regularly loses motions to dismiss, as part of the basis for granting a motion to stay discovery in the case. Moore v. Shands Jacksonville Med. Ctr., Inc., No. 3:09-cv-298-J-34TEM, 2009 WL 4899400, at *2 (M.D. Fla. Dec. 11, 2009).
[76]. See supra Part II.A.
[77]. See, e.g., Ashcroft v. Iqbal, 129 S. Ct. 1937, 1953 (2009); see also Dynamic Image Techs., Inc. v. United States, 221 F.3d 34, 38 (1st Cir. 2000) (noting that “uncontrolled discovery poses a special threat” when directed at federal government officials and that “compelling public policy reasons support stringent limitations on discovery pending resolution of threshold jurisdictional questions”).
[78]. See supra Part II.A.2.
[79]. Iqbal, 129 S. Ct. at 1953.
[80]. See generally Marc Galanter, The Emergence of Judge as Mediator in Civil Case, 69 Judicature 257 (1986); Marc Galanter & Mia Cahill, “Most Cases Settle”: Judicial Promotion and Regulation of Settlements, 46 Stan. L. Rev. 1339 (1994).
[81]. The Seventh Circuit has even referred to “settlement extortion—using discovery to impose asymmetric costs on defendants in order to force a settlement advantageous to the plaintiff regardless of the merits of his suit.” Am. Bank v. City of Menasha, 627 F.3d 261, 266 (7th Cir. 2010) (discussing one of the purposes of the Private Securities Litigation Reform Act). Yet on the flip side, defendants routinely use discovery tactically in Title VII employment discrimination cases to force settlements.
[82]. See, e.g., Maurice Rosenberg, Federal Rules of Civil Procedure in Action: Assessing Their Impact, 137 U. Pa. L. Rev. 2197, 2198 (1989). Although the utility of the discovery process in promoting settlement is disputed by some, see, for example, William A. Glaser, Pretrial Discovery and the Adversary System 91–101 (1968) (reporting that the available data do not clearly support the commonly-held view that discovery promotes settlement), a recent “natural experiment” that occurred in Taiwan provides evidence to support the settlement-encouraging role of the discovery process in common law systems. A study of settlement rates both before and after Taiwan implemented discovery practices comparable to those in the United States revealed that settlement rates increased over time following this reform. See generally Kuo-Chang Huang, Does Discovery Promote Settlement? An Empirical Answer, 6 J. Empirical Legal Stud. 241 (2009).
[83]. Twin City Fire Ins. Co. v. Emp’rs Ins. of Wausau, 124 F.R.D. 652, 653 (D. Nev. 1989).
[84]. See Tamburo v. Dworkin, No. 04 C 3317, 2010 WL 4867346, at *1 (N.D. Ill. Nov. 17, 2010) (holding that a stay of discovery is generally appropriate when a party raises a potentially dispositive threshold issue such as a challenge to standing (citing U.S. Catholic Conference v. Abortion Rights Mobilization, Inc., 487 U.S. 72, 79–80 (1988)).
[85]. Ellis v. Fortune Seas, Ltd., 175 F.R.D. 308, 312 (S.D. Ind. 1997) (holding that a threshold showing of jurisdiction is consistent with the district court’s obligation to control discovery).
[86]. This statement is not based on any formal analysis at this time, but only on my initial review of a non-representative sample of cases. A more formal analysis is required to answer such a question, as discussed infra Part V.C.
[87]. See, e.g., Aponte-Torres v. Univ. of Puerto Rico, 445 F.3d 50, 59 (1st Cir. 2006) (holding that a stay of discovery is appropriate where plaintiff had an adequate opportunity to conduct discovery); Chatham Condo. Ass’n v. Century Village, Inc., 597 F.2d 1002, 1011–12 (5th Cir. 1979) (holding that dismissal for lack of subject matter jurisdiction without ample opportunity for discovery should be rare).
[88]. See, e.g., Walk Haydel & Assocs., Inc. v. Coastal Power Prod. Co., 517 F.3d 235, 241–42 (5th Cir. 2008) (holding that the court has discretion in type and amount of discovery permitted).
[89]. See S.K.Y. Mgmt. L.L.C. v. Greenshoe, Ltd., No 06-21722-CIV, 2007 WL 201258, at *2–3 (S.D. Fla. Jan. 24, 2007) (denying defendant’s motion to stay discovery because even if motion to dismiss is successful, remaining claim would result in much of the same discovery); see also Sprague v. Brook, 149 F.R.D. 575, 577 (N.D. Ill. 1993) (“A plaintiff’s right to discovery before a ruling on a motion to dismiss may be stayed when the requested discovery is unlikely to produce facts necessary to defeat the motion.”). At least one court has even allowed discovery against defendants who asserted qualified immunity in a motion to dismiss because the defendants would also be fact witnesses in remaining claims not subject to the qualified immunity defense. See Seeds of Peace Collective v. City of Pittsburgh, No. 09-1275, 2010 WL 2990734, at *3 (W.D. Pa. July 28, 2010) (explicitly discussing Iqbal). However, the opposite view has been taken as well. See, e.g., Eggert v. Chaffee Cnty., No. 10-cv-01320-CMA-KMT, 2010 WL 3359613, at *3 (D. Colo. Aug. 25, 2010) (citing Iqbal for the proposition that discovery in a suit with multiple defendants should be stayed entirely, even if only some of the defendants asserted immunity).
[90]. See, e.g., Cohn v. Taco Bell Corp., 147 F.R.D. 154, 162 (N.D. Ill. 1993) (noting that a discovery stay is “particularly inappropriate in this case because . . . defendant’s motion [to dismiss] would not be dispositive of the entire case” and “all of the counts sound in the same alleged basic factual scenario”).
[91]. This is not to say that there is no harm due to delay. Particularly when the plaintiff does not have the same resources as the defendant, delay may add to the incentives for plaintiffs to settle or might make it impractical for the plaintiff to continue the lawsuit. In this way, defendants are sometimes able to wait out plaintiffs who lack sufficient resources. However, so long as the plaintiff is able to maintain the action and does not face undue pressures to settle for less than the case is worth, delays in cases dealing only with money damages do not prevent the plaintiff from fully recovering those damages. This is not true in cases where the harm due to delay is irreparable.
[92]. See, e.g., Computer Assocs. Int’l, Inc. v. Simple.com, Inc., 247 F.R.D. 63, 69 (E.D.N.Y. 2007); Hachette Distribution, Inc. v. Hudson Cnty. News Co., 136 F.R.D. 356, 358 (E.D.N.Y. 1991) (noting the “complexity of the action” as one factor to consider).
[93]. Roueche v. United States, No. 09-cv-00048-WDM-BNB, 2010 WL 420040, at *1–2 (D. Colo. Feb. 1, 2010) (considering burden of discovery in “balancing of interests” test); Feldman v. Flood, 176 F.R.D. 651, 652–53 (M.D. Fla. 1997) (considering burden of discovery in “preliminary peek” test).
[94]. One notable exception to this statement is in the Eleventh Circuit, where the court appears to make broad statements in support of an auto-stay rule, but lower courts pulled back from that in part because the Chudasama case was unusually complex and contentious. See S.K.Y. Mgmt. L.L.C. v. Greenshoe, Ltd., No. 06-21722-CIV, 2007 WL 201258, at *1 (S.D. Fla. Jan. 24, 2007).
[95]. Disentangling the complexity of a case from the “type of case” as discussed earlier is not necessarily easy. There may be certain types of cases that are more likely to be more complex, or less complex, for example. At the least, the type of case will likely have an impact on the type and amount of discovery involved. SeeMcKenna & Wiggins, supra note 16, at 791.
[96]. District judges may designate magistrates to “hear and determine any pretrial matter” with listed exceptions that include “motion[s] for injunctive relief, for judgment on the pleadings, for summary judgment” in addition to motions to dismiss for failure to state a claim. 28 U.S.C. § 636(b)(1)(A) (2006). Magistrate judges may, however, conduct hearings, including evidentiary hearings, and submit proposed findings of fact and recommendations for disposition in any of the listed exceptions. Id. § 636(b)(1)(B). Parties may consent to having a magistrate exercise jurisdiction. Id. § 636(c)(1).
[97]. See, e.g., Roueche v. United States, No. 09-cv-00048-WDM-BNB, 2010 WL 420040, at *2 (D. Colo. Feb. 1, 2010).
[98]. To the contrary, I am aware of at least one case where a magistrate judge explicitly applied the preliminary peek test. Tradebay, LLC v. eBay, Inc., No. 2:11-cv-00702-ECR-PAL, 2011 WL 6182039 at *6 (D. Nev. Dec. 13, 2011).
[99]. Inst. for the Advancement of the Am. Legal System, 21st Century Civil Justice System: A Roadmap for Reform: Pilot Project Rules 1(2009), available athttp://iaals.du.edu/images/wygwam/documents/publications/Pilot
_Project_Rules2009.pdf.
[101]. See, e.g., Press Release, Public Information Office, U.S. Courts for the Ninth Circuit, Justice Kennedy Joins Call for New Judgeships for Eastern California Court (Aug. 20, 2010), available at http://www.ca9.uscourts.gov
/datastore/general/2010/08/30/Justice_Kennedy_CAE_Remarks.pdf (calling attention to remarks from Justice Kennedy that the district needs more judgeships to be authorized by Congress).
[102]. The debate between rules and standards is an old one that I will not reinvent here. See e.g., Pierre Schlag, Rules and Standards, 33 UCLA L. Rev. 379 (1985). As I argue in this Article, I believe that a standard is more appropriate than a rule in deciding motions to stay discovery. This is consistent with Fed. R. Civ. P. 26(c), which allows judges to issue protective orders to prevent abusive discovery based on a showing of good cause. Bright-line standards are much more likely to result in consistently over- or under-inclusive results. Rules that always stay discovery in certain kinds of cases, such as shareholder derivative suits, will consistently err on the over-inclusive side because they will stay discovery in some number of cases where the motion to dismiss is denied (assuming at least some of the cases survive the motion to dismiss). Rules that never stay discovery in certain cases would be similarly biased toward the under-inclusive side because they would result in “wasted” discovery when the motion to dismiss is granted, although as noted previously no court has taken such a hard-line approach at this end of the spectrum.
[103]. The size of this grey area could even be quite significant, even for the great proportion of the cases, depending on the tolerance our society and our courts have for error in the discovery stay context.
[104]. Of course if the judge decides both a motion to stay discovery and a motion to dismiss at the same time, there is no additional burden in applying a “preliminary peek” approach. Yet deciding the motion to dismiss effectively moots the motion to stay discovery pending resolution of the motion to dismiss, so my analysis is focused instead on those cases where the judge decides a motion to stay discovery sometime before deciding the motion to dismiss.
[105]. See Hartnett, supra note 11, at 507–13 (pointing out that the Supreme Court has only made such a statement in the context of qualified immunity, and could not make a blanket prohibition on discovery during a pending motion to dismiss in a way that is consistent with the plain text of the Federal Rules).
[106]. See supra Part II.A.1.
[107]. See supra Part II.A.2.
[108]. This is not to say that delay cannot be a serious concern even when a plaintiff only seeks money damages, particularly when there are great resource disparities between the parties and one side can “wait out” the other, who is forced to give up on a potentially successful lawsuit. However, the risk of harm is lessened because theoretically there is no irreparable harm, and any concerns over resource disparities can also be addressed through application of the balancing test.
[109]. By this I do not mean that magistrate judges are incompetent to decide motions to dismiss, or even that they are necessarily less competent than district judges to make such rulings.
[110]. Of course district judges might change their minds between their preliminary assessment and ultimate ruling on a motion to dismiss, so the preliminary peek does not completely eliminate the risk of error in deciding a motion to stay discovery. However, it is reasonable to conclude that a magistrate judge and district judge are more likely to disagree than a district judge is to change her mind.
[111]. These are longstanding complaints. See, e.g., Maurice Rosenberg & Warren R. King, Curbing Discovery Abuse in Civil Litigation: Enough Is Enough, 1981 BYU L. Rev. 579 (citing conclusions of the Pound Conference and challenging the idea that “more is better” in the discovery context); Stephen N. Subrin, Fishing Expeditions Allowed: The Historical Background Of The 1938 Federal Discovery Rules, 39 B.C. L. Rev. 691 (1998). Debates over discovery abuse continue to this day. See generally John H. Beisner, Discovering a Better Way: The Need for Effective Civil Litigation Reform, 60 Duke L.J. 547 (2010); Rakesh N. Kilaru, Comment,The New Rule 12(b)(6): Twombly, Iqbal, and the Paradox of Pleading, 62 Stan. L. Rev. 905 (2010).
[112]. See, e.g., Roadway Express, Inc. v. Piper, 447 U.S. 752, 757 n.4 (1980) (“[M]any actions are extended unnecessarily by lawyers who exploit or abuse judicial procedures, especially the liberal rules for pretrial discovery.”); Herbert v. Lando, 441 U.S. 153, 179 (1979) (Powell, J., concurring) (noting that discovery is “not infrequently exploited to the disadvantage of justice”); Am. Bank v. City of Menasha, 627 F.3d 261, 265–66 (7th Cir. 2010) (referring to the prevention of “settlement extortion” as the purpose of staying discovery); Frank H. Easterbrook, Discovery As Abuse, 69 B.U. L. Rev. 635 (1989).
[113]. See, e.g., Linda S. Mullenix, Discovery in Disarray: The Pervasive Myth of Pervasive Discovery Abuse and the Consequences for Unfounded Rulemaking, 46 Stan. L. Rev. 1393 (1994). Additionally, the potential for discovery abuse is not limited only to plaintiffs, as defendants may also engage in tactical use of discovery to harass plaintiffs, bury them in paper, or drive up the costs of litigation. Such a situation can occur in the employment discrimination context, for example.
[114]. Reliable quantification of discovery costs is difficult to measure, and most studies that do exist are rather dated and “perhaps more illustrative of the difficulties of such research than of the actual cost picture.” McKenna & Wiggins, supra note 16, at 797; see also Emery G. Lee III & Thomas E. Willging, Fed. Judicial Ctr, Litigation Costs in Civil Cases: Multivariate Analysis 2 (2010) (“The point is obvious, but we state it for clarity’s sake: the model estimates presented in this section are only as good as the respondents’ reports of costs in the closed cases.”), available at http://www.uscourts.gov
/uscourts/RulesAndPolicies/rules/Duke%20Materials/Library/FJC,%20Litigation%20Costs%20in%20Civil%20Cases%20-%20Multivariate%20Analysis.pdf.
[115]. See David M. Trubek et al., The Costs of Ordinary Litigation, 31 UCLA L. Rev. 72, 82 (1983).
[116]. See supra Part II.A.
[117]. See Walker et al., supra note 40, at 641–42.
[118]. Ashcroft v. Iqbal, 129 S. Ct. 1937, 1951 (2009) (citing the response of law enforcement officials in the wake of the September 11 attacks).
[119]. Harlow v. Fitzgerald, 457 U.S. 800, 807 (1982) (discussing the need for qualified immunity to protect presidential aides).
[120]. See Hartnett, supra note 11, at 513.
[121]. The comparison to the preliminary injunction test also comes to mind here, where a finding regarding the likelihood of success on the merits might affect the ultimate decision on the merits as well.
[122]. See, e.g., Clayton P. Gillette, Lock-in Effects in Law and Norms, 78 B.U. L. Rev. 813 (1998).
[123]. See, e.g., S.J. Liebowitz & Stephen E. Margolis, Path Dependence, Lock-In, and History, 11 J.L. Econ. & Org. 205 (1995).
[124]. See supra Part V.B.
[125]. For raw data, see Motions to Dismiss, supra note 20.
[126]. Joe S. Cecil et al., Fed. Judicial Ctr., Motions to Dismiss for Failure to State a Claim After Iqbal 8–12 (2011), available athttp://www.fjc.gov/public/pdf.nsf/lookup/motioniqbal.pdf/$file/motioniqbal.pdf.
[127]. Id. at 12–16. The study did, however, find an increase in rates of dismissals for cases involving challenges to mortgage loans, although this trend may have been due to changes that occurred in the housing market at the same time. Id. at 12–13. For a critique of this study, see Lonny Hoffman, Twombly and Iqbal’s Measure: An Assessment of the Federal Judicial Center’s Study of Motions to Dismiss (Oct. 27, 2011) (unpublished manuscript), available athttp://ssrn.com/abstract=1904134
Lynch_LawReview_4.12
