Selective Settlement and the Integrity of the Bellwether Process

Note - Volume 97 - Issue 4

I. Introduction

In this Note, I hope to address a problem that can potentially undermine the usefulness of the “bellwether trial” process in multidistrict litigation. Specifically, I call this the “selective settlement problem.” In multidistrict litigation (MDL), one tool often used to resolve disputes is the “bellwether trial” process. In the bellwether-trial process, a select number of cases are chosen from the overall MDL pool and set for trial. Ideally, by trying this representative sample of cases, lawyers on both sides are able to gain useful information (such as judicial determinations on trial motions and the outcomes of the bellwether cases themselves) for purposes of reaching a global settlement.

By engaging in strategic conduct, however, parties are able to “selectively settle” plaintiff picks in the bellwether-trial process. For instance, rather than trying a representative subset of cases, the defendant could choose to settle all of the cases that carry the most risk if taken to trial. In turn, only those cases that are most favorable to the defendant are tried. This allows defendants to manipulate and distort the information that results from this process, which destroys much of the intended benefit of the process.

In this Note, I suggest a solution. First, by using judicial authority to condition participation in the bellwether process on an agreement to try all cases that have been selected as part of the bellwether pool to a verdict, the defendant’s ability to manipulate the process could potentially be removed. Second, as a way to ease the concerns of both parties as a result of having to forgo their right to settle the bellwether cases before trial, the courts could use “high–low agreements.”[1] These high–low agreements can help guarantee that neither record-setting verdicts, nor non-liable verdicts, will disproportionately harm the parties involved in the bellwether process merely because they chose to participate.

This risk mitigation prevents the rest of the plaintiffs in the MDL from simply free-riding on the informational benefits of the bellwether process. While the bellwether plaintiffs would waive their right to settle, these agreements would help limit their risk by guaranteeing at least some recovery. Further, it would also lower the risk for defendants by ensuring that a runaway jury does not award a massive verdict against them. At the same time, this allows for the entire selected pool of bellwether trials to be heard. The outcome of these carefully selected cases would then provide the informational benefits that are largely the justification for this process.

Finally, by making the values of the high–low agreement public, in conjunction with the values of the verdicts and other information created during the course of the trial, all parties will be provided with useful information for purposes of future settlement. As will be explained in this Note, this solution can potentially solve one of the key flaws of the bellwether-trial process, shoring up an already vital tool in our system of multidistrict litigation.

II. The Problem of “Selective Settlement” in the Bellwether Process

Before diving into the problem and my proposed solution, it is important to first lay out some essential background information. At the center of the selective settlement problem lies the multidistrict litigation process and bellwether trials.

First, multidistrict litigation is a process (authorized by 28 U.S.C. § 1407) that allows for a centralized forum where cases filed in multiple districts can be consolidated.[2] Often, cases concerning a single issue (for instance, a products liability suit) span multiple states, encompassing the claims of thousands of plaintiffs. These cases typically exist in both state and federal court. The MDL allows cases filed in a federal court to be consolidated in a single federal district court. This increases efficiency in pretrial procedures and often facilitates settlement.[3]

This process is designed to “provide for the temporary transfer to a single district for coordinated or consolidated pretrial proceedings of civil actions pending in different districts which involve one or more common questions of fact.”[4] This consolidation is run by the Judicial Panel on Multidistrict Litigation, or the “MDL Panel.”[5]

This Panel determines when cases should be transferred, and it may do this “upon its own initiative” or if a party files a motion “in any action in which transfer . . . may be appropriate.”[6] So, the MDL is created by the Panel issuing an “order of transfer,” which then transfers the cases under 28 U.S.C. § 1407.[7]

These cases are then litigated in the MDL court for “coordinated or consolidated pretrial proceedings.”[8] The transferee court retains broad authority, but it is limited by the landmark case, Lexecon Inc. v. Milberg Weiss Bershad Hynes & Lerach.[9] In Lexecon, the Supreme Court held that the MDL statute “not only authorizes the Panel to transfer for coordinated or consolidated pretrial proceedings, but obligates the Panel to remand any pending case to its originating court when, at the latest, those pretrial proceedings have run their course.”[10]

In most respects, the overall purpose of MDL is to move towards an efficient resolution of the underlying claims. Typically, this means settlement rather than actually going to trial for each of the claims filed in the MDL. While there are many ways to accomplish this goal, one of the most popular methods is known as the “bellwether trial.”[11] Bellwether trials are used as a tool to help resolve MDL, because they (ideally) not only consolidate pretrial proceedings, but also provide information to all parties involved through the experience at trial as well as through the outcome of the cases. In the bellwether process, “A subset of cases from the pool of suits in the multidistrict litigation are selected for trial.”[12] In theory, a subset of cases is properly selected by the parties, and then these cases can be tried to a jury.[13]

In general, the idea is that these cases will be representative of the overall composition of the MDL.[14] These cases are typically composed of “plaintiff pick” and “defendant pick” cases, and they are supposed to represent the universe of claims ranging from bad to good. By picking “plaintiff” and “defense” cases, ideally a bias towards an overall plaintiff- or defendant-friendly series of bellwethers is avoided. By trying these cases, the final outcome then provides critical information, which can be used by both parties in deciding how best to proceed in subsequent litigation.[15]

For example, in a situation where there are five plaintiff picks and five defendant picks, if a majority of the picks resolve in favor of the plaintiff, it is likely that the defendant will be especially eager to avoid future litigation and attempt to negotiate a settlement. And if the results go the other way, it provides important information for both parties on the risks associated with future litigation, further pushing towards settlement. What is crucial though is that this information is representative of the overall universe of cases filed in the MDL.

A number of practical problems limit the effectiveness of bellwether trials. First, as mentioned earlier, Lexecon limits the ability of the transferee judge to retain cases in the MDL by prohibiting a transferee judge from simply assigning cases not initially filed in his court to himself. Without the judge having this power, it may be more difficult for all parties in the MDL to get a truly representative sample of plaintiff and defendant picks for the bellwether process, since the cases originally filed in the MDL court may not be representative.

This gives the transferee judge a number of options. If pretrial proceedings have completed, the case is typically remanded to the transferor court for future litigation as mandated by Lexecon. But this is not in line with the idea of efficient resolution in the MDL court. True, the pretrial proceedings were completed in the MDL court, avoiding duplicative discovery, for instance. But one of the benefits (or some say drawbacks) of the MDL is not only that it facilitates more efficient pretrial activities, but that it facilitates settlement. If this is the goal, then a remand of all transferred cases to the many district courts for trial would not be the most efficient outcome.

A second option is to “sit by designation” in the transferor court, but many practical problems limit this option as well.[16] Finally, and arguably ideally, the judge can have the parties who have not initially filed in his court agree to have their case heard in the transferee court. These are known as Lexecon waivers.[17] As an example of this option, in the GM MDL, the parties involved in that set of bellwether trials agreed to have their cases heard in New York in front of the MDL judge, Judge Furman.[18] By giving Lexecon waivers, the cases are able to be tried in the MDL court rather than be remanded to the transferor court, which allows for the production of information (be it trial packages or data useful in settlement negotiations) that can then be used to resolve all of the cases that are filed in the MDL.

But a second problem, which is the focus of this Note, can arise when a proper pool of bellwether cases is chosen, but the informational value of the pool is (potentially) manipulated by the defendants to avoid taking plaintiff pick cases to trial.[19] While this problem has not been extensively discussed in legal scholarship, groups like the Duke Center for Judicial Studies have recognized the existence of a risk that defendants may manipulate the bellwether process to their advantage.[20] For purposes of this Note, I am calling this strategic manipulation of the bellwether process “selective settlement.”

As the bellwether pool is selected, the plaintiffs essentially show their hand. By picking cases they feel will resolve favorably, they provide an easy opportunity for defendants to selectively settle the best cases before they ever go before a bellwether jury. Now provided with a pool of the best cases, defendants can then simply attempt to settle those cases, and do so confidentially.[21] By settling these cases confidentially, the informational benefits to the overall MDL are distorted. The only cases that typically go to trial are those which are defendant picks (or at least those picks that the defendant felt were not worth preemptively settling), meaning the information derived from the bellwether trials can become skewed towards the defendants. Because of this, only verdicts from defendant picks become public, while the settled plaintiff picks remain confidential. While it may still be possible for members of the leadership team, for instance, to know the confidential settlement values (and so use that information in a later settlement process), if the purpose of the bellwether-trial process is to derive information from the actual trial of these cases, then there is something lost by these confidential settlements.

For instance, the defendant could offer a settlement to the individual bellwether plaintiffs with the best cases, but on the condition that any settlement would be confidential. The defendant would want these settlements for two reasons. First, it would allow the defendant to “buy up” potentially high verdicts before they reach trial. Second, it would not publicly provide the plaintiffs with any potentially positive information, though the plaintiff’s lawyers involved with the settlement still will retain those settlement values. These settlements distort the informational benefits of the bellwether process in favor of the defendant and potentially undermine the main purpose of the bellwether process.end

This is not merely a hypothetical problem. To see how this could be a problem in real life, we next look at two real-world examples of the bellwether-trial process: the “GM Ignition Switch” and the Vioxx multidistrict litigations.

III. Confidential Settlement in Two Real-Life Examples: In re Vioxx Products Liability Litigation and the General Motors Ignition Switch Multidistrict Litigation

There are plenty of multidistrict litigations to choose from for the purposes of this Note, but two are most useful for illustration. The first set of cases is the recent GM Ignition Switch litigation.[22] These cases arose as a result of General Motors’ faulty ignition switches in roughly 800,000 cars.[23] Because of these ignition switch defects, the vehicle’s engine would shut off while the car was being driven, causing the driver to lose control and the airbags not to deploy.[24] Hundreds of cases were filed all over the United States arising from these accidents.[25]

The second case is In re Vioxx Products Liability Litigation (MDL 1657).[26] Vioxx was a non-steroidal anti-inflammatory drug sold by Merck.[27] But this drug led to a number of serious injuries and deaths, giving rise to thousands of suits in federal and state courts.[28]

Both of these cases share many things in common: large scale injury, widespread litigation, and most importantly, both involved a series of bellwether trials. The difference, however, lies in the way the defense in both cases acted during the bellwether process.

In the Vioxx case, the MDL transferee court selected six bellwether cases for trial, and all six were tried to a verdict.[29] This seems to have been a strategic decision by Merck. The defendants used the bellwether process as a way to gauge the value of their claims. This is exactly why the bellwether process exists in the first place. By trying both plaintiff and defendant picked cases, it is possible to get accurate information useful for settlement. Both plaintiff and defendant were provided with a microcosm of what further trials could produce, and so could predict the benefits and risks of settlement. Following these bellwether trials, Merck was able to negotiate its famous $4.85 billion settlement, resolving roughly 27,000 claims in federal and state court.[30]

All settlements are criticized, but at a minimum the bellwether process was allowed to do the work it was intended to accomplish in the Vioxx litigation. By allowing all of the bellwethers to go to trial, the selective settlement problem did not arise. But in the GM ignition switch litigation, a different pattern emerged. For the first round of bellwether trials, GM and the plaintiffs chose six bellwether plaintiffs.[31] But only one produced a verdict—for GM. Of these first six, one produced a verdict, one was dismissed by the plaintiff before trial, one was dropped by the plaintiff halfway through trial in dramatic fashion, and three more were confidentially settled.[32] In the face of these results, a second round of six bellwethers was announced.[33] But of the second round of bellwethers, only one has gone to trial, yet again yielding a result for the defendants.

This seems to be the prototypical example of the selective settlement problem. By settling the cases that it felt were most risky, GM avoided trying those cases that could have led to potential plaintiffs’ verdicts. Then, by allowing the cases GM deemed least risky to go to trial, GM was able to make favorable verdicts more likely. With the risky cases settled confidentially and the safe cases tried to verdict (for the defendant), the information produced by the bellwether process has skewed in favor of GM in a way it may not have if all twelve bellwether cases had gone to trial.

Considering these results, an obvious question presents itself: what is the value of the bellwether process? Certainly, it could be said that the cases are still being settled and that some information is being produced. Even confidential settlements produce some information, though this is largely restricted to the MDL leadership team. But as even the judge in charge of the GM multidistrict litigation, Judge Furman, stated: “It’s an expensive way of yielding data for settlement purposes.”[34] Judge Furman noted, however, that he was “not aware of, and the parties haven’t suggested . . . a better alternative.”[35]

The question, then, is not whether the bellwether process produces any information. Rather, it is whether the bellwether process is the most effective, efficient, and just way of yielding data for settlement purposes. If it is not, what is to be done? As is demonstrated by the Vioxx case, the bellwether process (free of confidential pre-bellwether-trial settlements) appears to have been an incredibly effective way to produce information that was later central to the settlement process. When allowed to produce the information it is designed to provide, the bellwether process can be an incredibly effective tool for providing resolution in the MDL. But it may not always be the defense’s policy (as in Vioxx) to try every bellwether case to verdict.[36]

The problem from the defendant’s perspective is that if you do not share the same concerns as the defendant in Vioxx (for instance, if you do not want a quick global settlement for purposes of finality), then you may not care to try every case simply to quickly pursue global peace. Rather, you may choose to selectively settle the bellwether cases least favorable to your client, while trying those cases that are most favorable. This yields a distorted set of information, most likely in favor of the defendant. What happens if, like in the GM ignition switch litigation, the defense adopts the opposite strategy and selectively settles the risky cases, allowing only those cases that are least likely to yield negative results and information to go to trial?

It seems that in the face of this possibility, blind deference to the fact that the bellwether process has worked in the past cannot be a justification for using it going forward. The bellwether process is a powerful tool, and I do not suggest it should simply be abandoned because of the flaw produced by selective settlement. I suggest that something should be done to fix this potential for manipulation if we hope to preserve and maximize the informational value of the bellwether process.

IV. Potential Solutions to the Selective Settlement Problem

Following from these examples, it seems that there is at least the potential for manipulation by defendants in the bellwether process. The problem of defendants settling cases before the plaintiff picks reach trial at least has the potential to undermine the entire point of the bellwether process. If bellwether trials exist to provide accurate informational benefits for the use of all parties in the MDL, and the information is distorted by these settlements, then either (1) the bellwether process should be abandoned or more cautiously used, or (2) the bellwether process should be reformed. A number of potential solutions have been offered, but none seem to fully address this problem.

For example, one suggestion discussed by the Duke Law School Center for Judicial Studies is that “[s]uch strategic behavior can be mitigated by . . . allowing plaintiffs to choose the replacement for any bellwether case that defendants choose to settle rather than take to trial, or allowing defendants to select the replacement for any bellwether case that plaintiffs choose to dismiss.”[37]

This does not really solve the problem. By providing defendants with a steady pipeline of good cases to settle, this would allow defendants to effortlessly identify and settle all of the good cases without paying any finality or search premium. By the end of this process, it could leave only those cases that defendants may actually be willing to litigate, which would potentially result in a different type of information distortion. The only way for plaintiffs to avoid this sort of behavior would be to either not try the cases they would ideally take to trial or refuse to participate in the bellwether process entirely.

Another possible solution is to require disclosure of all settlements of bellwether cases. This solution has more appeal. As discussed before, a large part of the problem with selective settlements is that they remove the public informational benefit associated with the bellwether process. They do this not only by settling, but by settling confidentially. By keeping the settlement values private, the informational value of the bellwether process is diminished and potentially distorted because only defendant picks go to trial. Thus, by requiring the settlement values be made public, the informational distortion is at least mitigated.

While I think this solution could go towards solving the problem, I am still not entirely satisfied with it. The MDL process is in most ways designed to move cases towards settlement. So, it is not proper to say that settling rather than trying the bellwether cases defeats the entire purpose of the MDL. But if part of the premise of the bellwether trial is that there is informational value in the cases actually going all the way to verdict and becoming public, then this value would be lost by selective settlement. As noted by Judge Furman, the bellwether process is probably not the cheapest or most efficient way to achieve resolution.[38] Just because the bellwether process may lead to resolution (even with the information distortions), this does not mean that it is the ideal method of getting there or that it should continue to be used.

Further, if settlement values are made public, defendants may no longer be willing to settle the cases before trial for the value they would have if the settlements had remained confidential. It seems likely that part of the value of the settlement for the defendants is not only avoiding a large verdict, but also preventing the information from getting out in the first place. If avoiding having verdicts or settlement values become public is largely the motivation underlying selective settlement, then removing this motivation may remove the incentive to settle bellwether cases before they go to trial.

This may also simply push the defendants to offer settlements at a lower value than they would have if the settlements were confidential. By essentially lowering the value of the settlement from the perspective of the defendant, it may cause the defendant to simply try the case if the plaintiff is no longer willing to take this lower value settlement. It could also, however, cause these artificially low-value settlements to be the only publicly available information that is used as the basis for settlement, in conjunction with the only verdicts being those the defendant picks to go to trial.

It would certainly be possible for a plaintiff to “discount” the value of these settlements to essentially guess what the defendant would have valued the risk of trying the case at if they had been able to keep the settlements confidential. But it seems possible that the informational value of these settlements could be even less useful than confidential settlements. This is because confidential settlements would more closely reflect what the defendants value the case at, and the information would be available at least to the plaintiff’s leadership team.

Further, with the possibility that bellwether plaintiffs are facing (1) the daunting prospect of trial, (2) potential financial difficulties, and (3) a high likelihood that even if they win at trial, they may have a long delay before receiving any compensation, they may still be willing to take these “low-ball” offers. These public low-ball offers, coupled with the trial of the defendant pick cases, would seem to create its own information distortion.

I suggest a different solution. Entering the bellwether process is a choice and a privilege for both plaintiffs and defendants. It is one of many tools the MDL judge has at her disposal to help resolve the cases filed within the MDL, but it is not a process that has to be used. If this is the case, then it seems that the courts can qualify any participation in the bellwether process. The bellwether process should not be available simply as a tool to distort the settlement process for either side. It is simply too costly for plaintiffs, defendants, and the courts to justify if it is not producing more useful information than would be available without the bellwether process.

To prevent this, the MDL judge should use the judicial authority granted by both the concept of “inherent judicial authority” and by the MDL statute itself.[39] To benefit from the bellwether process, both parties should be required to agree to not settle any cases that have been set for the bellwether process before trial. But this does not mean that both parties must suddenly be subject to immense risk. To mitigate the potential risks, both parties could enter into high–low agreements, which would limit the potential liability for the defendant and guarantee at least some recovery for the plaintiffs involved in the litigation. This could potentially satisfy both parties entering into the bellwether process. It would guarantee some recovery for the bellwether plaintiff (and remove some of the risk associated with giving up their ability to settle) and would limit risk for defendants by ensuring that they are only liable for a maximum amount of damages if they are unlucky enough to find themselves before a very plaintiff-friendly jury. Finally, the values of these high–low agreements could be required to be made public following the end of trial. Making these values public would allow for a more accurate and efficient settlement process following conclusion of the bellwether process.

In Part V, I work backwards from my proposed solution by describing the way high–low agreements can be used to improve the bellwether process and mitigate risk for both parties. In Part VI, I discuss the solution of using inherent judicial authority and the power granted by the MDL statute to require bellwether litigants to not settle their claims before trial. Further, I argue that the MDL judge can use this authority to require that the high–low agreement be made public following trial as an additional source of information, and I discuss the sources of judicial authority supporting these solutions.

V. The High–Low Agreement and the Bellwether Trial

High–low agreements are defined as “settlements in which a defendant agrees to pay the plaintiff a minimum recovery in return for the plaintiff’s agreement to accept a maximum amount regardless of the outcome of the trial.”[40] Essentially, high–low agreements are contracts between plaintiffs and defendants designed to mitigate the risk for both parties in any risky litigation. It is difficult to predict how cases will work out once they actually go to trial, and this is something that is worrisome for all parties.[41]

Bellwether trials are not the only trials that are vanishing. In general, the fear of going to trial has led the vast majority of parties involved in litigation to decide that the risks of trial outweigh the rewards.[42] As a result, settlement has dramatically increased at the expense of the trial.[43] But what if trial is not only for the purpose of resolving individual cases, but like in the bellwether process, is focused more on providing information for the resolution of an entire universe of disputes arising from a similar set of facts?

It seems that the high–low agreement may be an ideal tool for resolving the problem of selective settlement. First, it mitigates the risk for any bellwether plaintiff going into litigation. Bellwether plaintiffs open themselves up to the possibility that they will receive a non-liable verdict and walk away with nothing. While bellwether plaintiffs may receive a jackpot verdict, the unpredictability of juries and trial coupled with the near certainty of lengthy appeals (and potentially reversed verdicts) makes the bellwether process less appealing. But with a guaranteed minimum payout, bellwether plaintiffs can at least be assured that they will be paid something, and soon. This is because the terms of the high–low agreement will not only ensure a minimum–maximum level of payment, but also exclude the possibility of appeal, a delay of payment, and the need to incur additional litigation-related expenses.

For defendants, a high–low agreement doesn’t necessarily address every one of their concerns. But it does address an essential one. Namely, it avoids defendants being forced to pay a large public verdict because they are forced to try a riskier case. As a massive corporation in a situation with very bad facts, it seems highly likely that at least one jury will grant a huge verdict in favor of the plaintiffs.

But for a defendant the size of General Motors or Merck, often the dollar amount of a single verdict may not be the reason to avoid taking part in the bellwether process. The fear is less of the single large verdict, and more of the risk of many large verdicts that could result if all cases were to go to trial. As discussed above, the defendant does not want this information to be made public, providing information for the plaintiffs in settlement negotiations, and causing concern for shareholders who are nervously watching the course of litigation. It is likely that one of the main reasons the defendant chooses to settle (rather than try the case) is to avoid providing the informational benefits associated with allowing a plaintiff pick to go to trial.[44] If this is the case, then simply agreeing to a high–low agreement with the plaintiffs will not seem attractive enough to prevent defendants from selectively settling plaintiff pick cases. The defendant would rather pay money now than risk having to go to trial later, even if the risk of a large verdict is mitigated by the high–low agreement.

Finally, if this is the case, then the high–low agreement does not solve the problem for the non-bellwether plaintiffs. If defendants do not believe the terms of the high–low agreement outweigh the potential benefits of selectively settling plaintiff pick cases, then there will be no real incentive to join a high–low agreement. So, the bellwether process will still suffer from the information distortion at the center of this Note, and any subsequent settlement will thus be the result of distorted information.

Thus, high–low agreements in and of themselves are not a solution to the problem. For high–low agreements to be useful, they need to be used in conjunction with something else. Judicial engagement seems to be the last piece of the solution. As will be discussed below, judges should condition the participation in the bellwether process on an agreement by the parties to (1) not settle their cases once they enter the bellwether process, to (2) encourage or potentially require the parties to enter into a high–low agreement as a way to soften the blow of removing the right to settle, and to (3) require the value of these high–low agreements to be made public at the conclusion of trial. This would ensure for the bellwether plaintiff that even if her claims blow up at trial (as they did in the GM Ignition Switch case, for instance), she would receive a minimum recovery, while also protecting the defendant from paying an overwhelmingly large verdict. While this would certainly provide more of a benefit to the plaintiffs than the defendants, this stands to reason, since the reform is focused on the potential for strategic behavior of the defendants.

Making these high–low agreements public (rather than confidential) at the end of the trial would provide an additional informational benefit. Even if the high–lows were negotiated without any trials, information would still be made available. But once a jury hears a trial, the value of the high–low agreements in conjunction with the verdict of the jury would provide additional valuable information to be used by both parties for purposes of later settlement.

But at least one more problem exists, which is at the center of this potential solution: Do judges have the authority to impose these requirements? If they do not, then the problem may prove insolvable. But as will be shown, judges, and in particular MDL judges, retain broad discretion and authority to craft solutions for problems of this sort. As will be discussed below, this power appears broad enough to allow for a judicially crafted solution to the problem of selective settlement.

VI. The Judicial Authority to Craft Solutions to Injustice and Inefficiency in the MDL

Considering that multidistrict litigation takes up close to forty percent of the federal docket, it would be reasonable to assume that the boundaries of judicial authority within the MDL would be well defined.[45] Yet except for the brief MDL statute, 28 U.S.C. § 1407, and a vague collection of inherent judicial powers, it is unclear what the actual scope of judicial authority in the MDL is. What is clear is that the MDL judge has authority over “coordinated or consolidated pretrial proceedings” after the case has been transferred to the MDL, and “at or before” the pretrial proceedings are concluded, the case “shall be remanded . . . to the district from which it was transferred.”[46] Finally, under Lexecon, MDL judges cannot simply transfer cases to themselves after pretrial proceedings are done, and so the judge needs to receive a waiver from the parties if she hopes to hold on to them for trial.[47] Other than this, everything is rather vague. Lexecon provides an outer limit to judicial authority, but it is not very effective as a limiting principle when discussing the solution proposed in this Note.

But by looking to two sources of judicial authority in the MDL context, “inherent judicial authority” and the powers granted by the MDL statute, it is possible to establish that MDL judges have the power and the duty to innovate to solve the selective settlement problem.

 A. Inherent Judicial Authority and the Broad Powers of Judges

First, I hope to demonstrate that judges have what is broadly called “inherent judicial authority” to solve issues like the selective settlement problem. This is a power that is inherent to judges inside and outside of the MDL process. The principle of “inherent judicial authority” provides a useful tool for judges when attempting to regulate conduct within their courtrooms. For instance, since early in American history, the Supreme Court has stated that courts retain broad authority to regulate the ethical conduct of lawyers, and this power has been construed rather broadly as an “inherent judicial authority.”[48]

One of the most well-known examples of the use of inherent judicial authority is in the contingency-fee context. By invoking inherent judicial authority, judges have been able to modify numerous contingency-fee contracts, even when no party has complained about them. For instance, in Zyprexa Products Liability Litigation,[49] Judge Jack Weinstein invoked his inherent authority to regulate the conduct (particularly the ethical conduct) of lawyers in his court as justification for revising a contingency-fee contract, even though no party had challenged the validity of the contract.[50] Judge Weinstein cited to a wide range of authorities to establish his inherent authority,[51] including to a federal case holding that district courts have the power to adopt their own rules creating an attorney fee schedule for personal injury actions for seamen,[52] as well as to a state court case limiting attorney’s fees in personal injury actions involving minors.[53] By citing to these sources, Judge Weinstein asserted that the inherent authority of judges to regulate fees (and their discretion more generally) was broad.

Judge Weinstein further invoked the MDL statute itself (as I will do in the next section) and its purpose to promote “the just and efficient conduct of such actions.”[54] He argued that the MDL process was designed to assemble cases for coordinated proceedings to promote “just and efficient conduct,” citing to the MDL statute’s text.[55] Weinstein argued that the risk of excessive fees, in conjunction with the fact that multidistrict litigation is an “important tool for the protections of consumers . . . [and] must be conducted so that they will not be viewed as abusive by the public,” justified his actions.[56]

Expanding on Judge Weinstein’s line of reasoning, if the purpose of the MDL statute and the MDL process more generally is to promote “just and efficient conduct,”[57] and a judge has inherent authority to regulate the conduct within his courtroom, then strategic conduct that attempts to manipulate the bellwether process goes against these goals of efficiency and should not be allowed. Multidistrict litigation is an effective tool both for resolving mass disputes, and for “protection of consumers in our modern corporate society.”[58] The bellwether process is one tool used by MDL courts to help lead to a “just and efficient outcome” of litigation. To allow this sort of strategic behavior would be an inefficient use of court resources, and would certainly not be “just.” So, to protect the integrity of the multidistrict litigation process, judges should utilize their inherent authority to prevent strategic behavior in their courtrooms that would undermine the effectiveness of the process.

Judge Weinstein is not the only judge to invoke inherent judicial authority for fee caps, though his Zyprexa opinion seems to have blazed the trail for rewriting contingency-fee contracts in the MDL context.[59] For instance, Judge Frank in the Guidant MDL held “this Court has the inherent right and responsibility to supervise the members of its bar in both individual and mass actions, including the right to review contingency-fee contracts for fairness.”[60] Additionally, in the Vioxx litigation, Judge Fallon also invoked this “inherent authority” by saying “[c]ourts that have considered the issue have nearly unanimously concluded that the power to consider the reasonableness of contingent fees is inherent in a federal court.”[61]

But the use of inherent judicial authority is not limited to the contingency-fee context. For instance, federal courts have been held to have the power to punish for contempt,[62] the power to vacate a judgment based on fraud upon the court[63] (based on the theory that it is necessary “to the integrity of the courts”[64]), and the power to outright dismiss a lawsuit.[65] This is not an exhaustive list of inherent judicial powers.[66] Finally, in Chambers v. NASCO,[67] the Supreme Court recognized that these inherent judicial powers are flexible. In Chambers, the Supreme Court upheld a ruling that imposed attorney’s fees against a litigant by invoking the court’s inherent authority and power to sanction bad-faith conduct.[68]

While it is difficult to really get a grip on the outlines of inherent judicial authority, it has been seen as a powerful tool “available to judges to give vigor to [their] managerial role.”[69] This inherent judicial authority is not limited to the contexts where it has already been used, but is still a tool available for judicial innovation. For instance, in a recent court order arising from the Ethicon vaginal mesh litigation, an MDL court stated that federal district courts have the “broad power to manage litigation, especially complex litigation, for the purpose of affording the parties a ‘just, speedy, and inexpensive’ disposition of the action.”[70] Even more, the court specifically noted that courts have the “inherent authority and duty to identify, define, and resolve issues.”[71] By invoking this authority, the court was able to dismiss without prejudice a large number of cases of plaintiffs who had not yet undergone “revision surgery” to preserve their claims for the future (if they went on to have revision surgery).[72] This was done to promote “[j]udicial economy as well as the speedy, just, and fair resolution” of the cases in the MDL.[73]

The inherent authority of judges to solve problems involving litigation in their courtroom is broad and has been used innovatively in a vast array of instances. The broad power of judges to regulate attorney conduct, such as the ability to regulate contingency fees, coupled with the text and history of the MDL statute, suggests a wide amount of judicial discretion when it comes to the conduct of parties within the MDL. This principle can be extended to the idea that by virtue of taking part in the MDL process (and more specifically the bellwether-trial system), MDL judges may retain the authority to condition certain conduct on having the benefit of participation.

B. The Text and History of the MDL Statute Support the Argument that MDL Judges Retain Broad Authority to Regulate Conduct to Prevent Unjust and Inefficient Conduct

More narrowly, the text, case law, and legislative history of the MDL statute support the idea that MDL judges retain broad authority over the cases before them.[74] First, the text of the MDL statute specifically states that the MDL panel shall transfer cases pending in different districts to the MDL court to, among other things, “promote the just and efficient conduct of such actions.”[75] Little context is given in the short MDL statute, but it is clear that at a minimum the MDL process is designed to promote just and efficient resolution of cases filed in the MDL.

Under the terms of the MDL statute, cases should not be transferred to the MDL judge if the process would not promote just and efficient conduct.[76] This principle should extend to the bellwether process. The bellwether process is designed to provide accurate and reliable information for settlement. But if that information is distorted as a part of strategic behavior, then the bellwether process loses its benefits while still carrying with it the risks and costs of trial.

The legislative history of § 1407(a) also points to the idea that the statute was created to ensure that the cases that are transferred to the MDL are handled justly and efficiently.[77] It follows then that judges would have the authority to ensure that multidistrict litigation in their courts can be performed justly and efficiently. For instance, when speaking about the purpose of passing the original MDL statute in 1968, a house report noted “[t]he objective of the legislation is to provide centralized management under court supervision of pretrial proceedings of multidistrict litigation to assure the ‘just and efficient conduct’ of such actions.”[78] As another example, Dean Phil C. Neal of the University of Chicago Law School testified before the Senate Judiciary Committee that the powers of MDL judges should be construed to be as broad as those of a federal district judge presiding over an individual case.[79] However, outside of a few comments like these, a review of legislative history sheds little light on the authority possessed by the MDL judge.

While these comments arise in the context of avoiding duplicative discovery and other pretrial issues, if the purpose of the MDL legislation was to assure “just and efficient conduct” broadly, then providing rules that prevent defendants from manipulating the bellwether-trial process seems reasonable.

Andrew Bradt, a scholar who has studied the history of the MDL statute, states that his study of the “drafters’ papers [shows] that they did not intend the role of the MDL statute, or the powers it confers on judges, to be modest.”[80] Rather, the statute’s drafters and proponents intended to “endow the judges overseeing these litigations with plenary power to manage them and with the flexibility to innovate when doing so.”[81] Through his extensive study of the history of the statute, Bradt concludes that the statute was intended to “profoundly change the way the courts process what they believed would be the lion’s share of federal civil cases,” since “the statute’s history demonstrates that the judges who developed the statute did not intend for it to play a limited role or for MDL judges to feel hemmed in.”[82]

Professor Bradt’s analysis of the history of the MDL statute provides important support for the idea that MDL judges have tremendous discretion when it comes to the cases before them. The statute’s drafters did not intend these judges’ roles to be limited. Instead MDL judges should be deeply involved in resolving the cases in front of them in a just and efficient manner. As noted by Bradt, MDL judges were supposed to be able to exercise tremendous power and to innovate pursuant to that power. The solution to the problem of selective settlement should be seen as one of those innovations within the broad judicial discretion possessed by MDL judges.

From case law, it seems that there are almost no examples (with the exception of Lexecon) of courts finding a limit on the conduct of transferee judges, while there are numerous cases that construe the power of transferee judges broadly. As just one example of the broad authority of judges in MDL, consider Upjohn,[83] where a transferee judge was held to possess the discretion to require parties to share discovery with litigants that were not parties to the MDL, despite protective orders that had been issued by earlier transferor courts.[84]

With the goals of Congress, as established by the legislative history, in mind, as well as the lack of case law suggesting strict limitations on the power of transferee judges in the MDL, it seems that the powers of a judge within the MDL are broad when it comes to solving problems like the selective settlement problem. If the goal of the MDL is largely to ensure efficiency and justice in situations where hundreds, if not thousands, of claimants are trying to resolve their claims, then allowing judges broad authority to run the process seems reasonable. Comparing this legislative history to the fact that there appears to be only one clear limit on the power of the MDL judges, it would seem, at least from the present state of the law, that there is nothing stopping a judge from providing the sort of solution I have suggested.

Because of the broad authority of judges within the MDL, the solution suggested by this Note seems to be workable and well within the powers of transferee judges in the MDL. Participation in the bellwether process is a beneficial procedure (at least theoretically) for both plaintiffs and defendants. Nothing about the process is necessary for purposes of the MDL. Rather, it is simply a tool an MDL judge is provided with to help resolve the dispute. Judges are not then obligated to provide this sort of solution free of any condition.

So, by using the discretion and power retained by the MDL judges, it is possible to craft a solution to the problem of selective settlement. First, judges can require litigants in the bellwether process to give up their right to settle prior to trial. Litigants are free to not take part in the bellwether process, but there appears to be no reason why a judge cannot condition their participation on certain choices and waivers which guarantee the integrity of the bellwether process and the MDL as a whole. To encourage this, judges could strongly suggest that the parties enter into high–low agreements. Among the terms could be an agreement not to settle the case before trial as well. This high–low agreement would help balance the risks among plaintiffs and defendants, and by requiring the case to go to trial, the MDL as a whole receives the informational benefits that are the purpose of the entire bellwether process. Finally, the high–low agreement could include within its terms that its contents would be made public at the conclusion of the bellwether trials. These contents could be yet another important source of information to be used to resolve the claims filed in the MDL.

VII. Conclusion

The problem of selective settlement is real and has been recognized by both practitioners and academics. If the integrity of the bellwether process is to be preserved, it needs to be addressed. One obvious solution is to abandon the bellwether-trial process entirely. But the problem of selective settlement seems neither so great that it justifies abandoning the process, nor so great that it remains unsolvable.

The solution I have provided in this Note consists of three parts. First, MDL judges can condition participation in the bellwether-trial process on agreement by all parties not to settle their claims before they go to trial. Second, the parties could use high–low agreements as a way to minimize the risk, thus offsetting some complaints of parties who would hope to selectively settle rather than try the bellwether case. Third, in conjunction with the information generated by the bellwether trial itself (the value of the verdict, potential trial packages, and the like), the values of the high–low agreements could also be required to be made public. This, in addition to the value of the verdict, would provide useful information for all parties involved in the bellwether process for future settlement.

By approaching this problem in a careful but determined way, MDL courts can protect the integrity of the bellwether process and ensure that the most accurate information possible is provided to the MDL as a whole. Unless the selective settlement problem is addressed, the bellwether process could devolve into a tool for abuse by savvy litigants. But by exercising the judicial power to supervise this process granted by the MDL statute and inherently possessed by judges, the bellwether trial could remain an effective tool for resolving complex multidistrict litigation.

  1. .High–low agreement, Black’s Law Dictionary (10th ed. 2014) (“A settlement in which a defendant agrees to pay the plaintiff a minimum recovery in return for the plaintiff’s agreement to accept a maximum amount regardless of the outcome of the trial.”). Even more, it is possible that the contents of these agreements could be made public following trial. A high–low agreement, in conjunction with the verdicts received through the completed bellwether process, is valuable information that could be used in future settlement negotiations.
  2. .28 U.S.C. § 1407 (2012).
  3. .See id. § 1407(a) (“[C]ivil actions involving one or more common questions of fact . . . may be . . . consolidated pretrial . . . [to] promote the just and efficient conduct of such actions.”).
  4. .Act of Apr. 29, 1968, Pub. L. No. 90-296, 82 Stat. 109, 109 (1968) (codified as amended at 28 U.S.C. § 1407 (2012)).
  5. .28 U.S.C. § 1407(a).
  6. .Id. § 1407(c).
  7. .Id.
  8. .Id. § 1407(a).
  9. .523 U.S. 26 (1998).
  10. .Id. at 34 (emphasis added).
  11. .See, e.g., Eldon E. Fallon et al., Bellwether Trials in Multidistrict Litigation, 82 Tul. L. Rev. 2323, 2330 (2008).
  12. .Alexandra D. Lahav, A Primer on Bellwether Trials, 38 Rev. Litig. 185, 185–86 (2018).
  13. .Whether this is true in practice is another issue. See Fallon et al., supra note 11, at 2343 (describing the ideal bellwether-trial-selection process).
  14. .Id. at 2346.
  15. .In the past, bellwether trials have also been used to bind the parties formally following the trial. But this practice has fallen out of use as courts have become skeptical of this process. See, e.g., id. at 2331 (citing Cimino v. Raymark Indus., Inc., 151 F.3d 297, 318 (5th Cir. 1998)) (explaining that the bellwether verdict would only be used to encourage settlement).
  16. .See, e.g., Lahav, supra note 12 (manuscript at 11) (noting that trying the case in the transferor court sitting by designation could result in discontinuity of decision-making, biased decisions, or additional costs).
  17. .See Fallon et al., supra note 11, at 2357–58 (describing a party’s consent to hear cases transferred to the transferee court by the MDL Panel as Lexecon waivers).
  18. .Lahav, supra note 12 (manuscript at 11).
  19. .Certainly, it is possible that a plaintiff, rather than a defendant, would attempt to manipulate the bellwether process. For instance, plaintiffs could agree to settle their worst cases or “defendant picks” before trial for a small amount, thus causing only the “plaintiff picks” to go to trial, thus distorting the available information. But practical problems make this far less likely than the other way around. The incentives are different. Plaintiffs with “bad” cases who have agreed to take part in the bellwether process have no incentive to settle their cases for next to nothing, rather than go to trial at the attorney’s expense and take a risk.
  20. .See Duke Center for Judicial Studies, MDL Standards and Best Practices
    22–23 (2014) (considering the potential for parties to settle or dismiss a bellwether case to manipulate the takeaways from the bellwether process).
  21. .See, e.g., Emily Field, GM, Driver Settle Ahead of Bellwether Ignition Trial, Law 360 (Oct. 11, 2017), [] (announcing confidential settlement between plaintiff and defendant GM).
  22. .See, e.g., GM Agrees $900m Settlement for Faulty Ignition Switches, BBC (Sept. 17, 2015), [] [hereinafter GM Agrees] (giving a general background of the GM ignition switch scandal).
  23. 23.GM Starts Fixing Ignition Switches in 2.6 Million Recalled Cars, NBC News (Apr. 7, 2014), [].
  24. .See GM Agrees, supra note 22 (“[The defect] could shut down engines, disable power-assisted steering and brakes, and prevent airbags [from] working.”).
  25. .Jonathan Stempel, GM Settles Hundreds of Ignition Switch Lawsuits, Reuters (June 23, 2017), [] (discussing, among other things, the extent of lawsuits filed against GM following the deaths and injuries resulting from faulty ignition switches in GM cars).
  26. .In re Vioxx Prods. Liab. Litig., 650 F. Supp. 549, 549 (E.D. La. 2009).
  27. .Barbara Sibbald, Rofecoxib (Vioxx) Voluntarily Withdrawn from Market, 171 CMAJ 1027, 1027 (2004), [].
  28. .Snigdha Prakash, Merck Ordered to Pay $4.5 Million in N.J. Vioxx Case, NPR (Apr. 6, 2006), [] (discussing the general background of the Vioxx litigation, including the large number of cases filed—roughly 10,000—as well as the source and type of injuries resulting from taking Vioxx); Snigdha Prakash and Vikki Valentine, Timeline: The Rise and Fall of Vioxx, NPR (Nov. 10, 2007), (recounting the number of injuries and deaths caused by Vioxx).
  29. .In re Vioxx, 650 F. Supp. 2d at 552. There were also a number of state court cases outside of the six bellwether trials in the MDL. These cases were tried to a verdict in Texas state court (two verdicts for the plaintiff), New Jersey state court (two split verdicts and two defense verdicts), California state court (one defense verdict and one hung jury), and Alabama state court (one defense verdict). See Vioxx Jury Verdict Summary, Toledo Law, [] (summarizing Vioxx jury verdicts for all cases tried to a jury in state and federal court).
  30. .Alex Berenson, Merck Agrees to Settle Vioxx Suits for $4.85 Billion, N.Y. Times (Nov. 9, 2007), [].
  31. .Cara Salvatore, GM Ignition Switch Didn’t Cause Driver’s Death, Jury Finds, Law 360 (July 19, 2017), [].
  32. .Id.
  33. .Id.
  34. .Max Stendahl, GM Ignition Switch Judge Won’t Ditch Bellwether System, Law 360 (Feb. 23, 2016), [].
  35. .Id.
  36. .See, e.g., Berenson, supra note 30 (describing Merck’s agreement to pay $4.85 billion to settle roughly 27,000 lawsuits related to Vioxx usage).
  37. .Duke Center for Judicial Studies, supra note 20, at 22.
  38. .See Stendahl, supra note 34 (quoting Judge Furman that the bellwether process is “expensive”).
  39. .These two sources of judicial authority in the MDL context will be discussed more fully in Part VI.
  40. .High–Low Agreement, Black’s Law Dictionary (10th ed. 2014).
  41. .See generally Samuel R. Gross & Kent D. Syverud, Don’t Try: Civil Jury Verdicts in a System Geared to Settlement, 44 UCLA L. Rev. 1, 61–62 (1996) (stating that high–low agreements are particularly common in cases with large potential damages and uncertain liability); Richard Lorren Jolly, Between the Ceiling and the Floor: Making the Case for Required Disclosure of High-Low Agreements to Juries, 48 U. Mich. J.L. Reform 813, 814 (2015) (characterizing high–low agreements as a creative limit to unpredictable jury awards); Molly McDonough, High-Lows’ Ups and Downs, 91 A.B.A. J. 12, 13 (Aug. 2005) (explaining how high–low agreements insulate both parties in difficult-to-evaluate cases).
  42. .Taunya Lovell Banks, Civil Trials: A Film Illusion?, 85 Fordham L. Rev. 1969, 1972 (2017) (stating that between 1962 and 2005, jury trials have dropped from 5.5% to less than 1% of all case resolutions); John H. Langbein, The Disappearance of Civil Trial in the United States, 122 Yale L.J. 522, 524 (2012) (stating that 1.2% of federal cases and 0.6% of state cases are currently resolved by jury trial).
  43. .Banks, supra note 42, at 1972.
  44. .See Gross & Syverud, supra note 41, at 53 (describing a defendant–manufacturer who settled to avoid public trial).
  45. .See Elizabeth Chamblee Burch, Monopolies in Multidistrict Litigation, 70 Vand. L. Rev. 67, 72 (2017) (noting that “from 2002 to 2015, multidistrict proceedings leapt from sixteen to thirty-nine percent of the federal courts’ entire civil caseload”); Judith Resnik, Lawyers’ Ethics Beyond the Vanishing Trial: Unrepresented Claimants, De Facto Aggregations, Arbitration Mandates, and Privatized Processes, 85 Fordham L. Rev. 1899, 1913–14 (2017) (“By September 2015, the percentage had risen again. Of 341,813 federal civil cases pending, 132,788 were concentrated in 271 proceedings aggregated before a single judge.” Further, “almost 40 percent of federal civil cases” are part of MDLs as of 2015.).
  46. .28 U.S.C. § 1407(a) (2012).
  47. .See Lexecon Inc. v. Milberg Weiss Bershad Hynes & Lerach, 523 U.S. 26, 35 (1998) (holding that the “straightforward language” of § 1407 prevents judicial self-assignment of multidistrict litigation).
  48. .See, e.g., Ex parte Burr, 22 U.S. 529, 530 (1824) (explaining that the authority to regulate lawyers belongs with the courts). See generally Aimee Lewis, Limiting Justice: The Problem of Judicially Imposed Caps on Contingent Fees in Mass Actions, 31 Rev. Litig. 209 (2012).
  49. .424 F. Supp. 2d 488 (E.D.N.Y. 2006).
  50. .Id. at 492–93.
  51. .Id.
  52. .Schlesinger v. Teitelbaum, 475 F.2d 137, 141 (3d Cir.), cert. denied, 414 U.S. 1111 (1975).
  53. .Leonard C. Arnold, Ltd. v. N. Tr. Co., 506 N.E.2d 1279, 1282–84 (Ill. 1987).
  54. .Zyprexa, 424 F. Supp. 2d at 494 (quoting 28 U.S.C. § 1407 (2012)) (stating that cases shall be sent to the MDL court “for coordinated or consolidated pretrial proceedings . . . [to] promote the just and efficient conduct of such actions” (emphasis added)).
  55. .Id.
  56. .Id.
  57. .Id.
  58. .Id.
  59. .In re Zyprexa Prods. Liab. Litig., 424 F. Supp. 2d 488, 491 (E.D.N.Y. 2006) (adopting caps on contingency-fee contracts for attorneys engaged in multidistrict pharmaceutical litigation).
  60. .In re Guidant Corp. Implantable Defibrillators Prods. Liab. Litig., MDL No. 05–1708 (DWF/AJD), 2008 WL 682174, at *18 (D. Minn. Mar. 7, 2008) (citing multiple authorities cited by Judge Weinstein in Zyprexa, including Rosquist, Farmington Dowel, and Taylor).
  61. .In re Vioxx Prods. Liab. Litig., 650 F. Supp. 2d 549, 559–60 (E.D. La. 2009).
  62. .Ex parte Robinson, 86 U.S. 505, 512 (1873).
  63. .Hazel-Atlas Glass Co. v. Hartford-Empire Co., 322 U.S. 238, 250 (1944).
  64. .Chambers v. NASCO, 501 U.S. 32, 44 (1991).
  65. .Link v. Wabash R.R. Co., 370 U.S. 626, 630 (1962).
  66. .See generally Chambers, 501 U.S. 32 (providing a survey of inherent judicial powers).
  67. .Id.
  68. .Id. at 43, 48.
  69. .Judith A. McMorrow, The (F)Utility of Rules: Regulating Attorney Conduct in Federal Court Practice, 58 SMU L. Rev. 3, 22 (2005); see also id. at 8, 22. McMorrow has a number of very useful statements on the issue of inherent judicial authority. For instance, she states “Federal courts have the inherent power to control the proceedings in their courtroom, including the long-recognized (albeit conceptually unclear) inherent power to regulate attorneys.” Id. at 8.
  70. .Order Regarding Disposition of Non-Gynecare TVT Products Cases at 3, In re Ethicon, Inc., No. 2:12-MD-02327 (S.D.W.Va. Apr. 17, 2018) (citing as authority for this argument Fed. R. Civ. P. 1, 16, 26, 27, 42, 53, 83; Fed. R. Evid. 706 (court-appointed experts)); Manual for Complex Litigation (Fourth) §§ 10.1, 11, 20, 22, 23 (2004).
  71. .Id. (citing Fed. R. Civ. P. 16(c)(2)).
  72. .Id. at 3–5.
  73. .Id. at 3.
  74. .See In re Patenaude, 210 F.3d 135, 142–46 (3d Cir. 2000) (discussing how the MDL statute’s text, case law, and legislative history support judicial discretion and authority).
  75. .28 U.S.C. § 1407(a) (2012).
  76. .See id. (“Such transfers shall be made by the judicial panel on multidistrict litigation . . . upon its determination that transfers for such proceedings . . . will promote the just and efficient conduct of such actions.”).
  77. .H.R. Rep. No. 90-1130, at 2 (1968).
  78. .Id.
  79. .Multidistrict Litigation: Hearings on a Proposal to Provide Pretrial Consolidation of Multidistrict Litigation Before the Subcomm. on Improvements in Judicial Mach. of the S. Comm. on the Judiciary, 89th Cong. 13 (1966) (“[T]he cases concerned would be brought within the control of a single district and so the very same powers provided by the Federal Rules of Civil Procedure should permit all of the same kinds of steps to be carried out by the presiding district judge.”).
  80. .Andrew D. Bradt, “A Radical Proposal”: The Multidistrict Litigation Act of 1969, 165 U. Pa. L. Rev. 831, 839 (2017).
  81. .Id.
  82. .Id. at 912.
  83. .In re Upjohn Co. Antibiotic Cleocin Prod. Liab. Litig., 664 F.2d 114 (6th Cir. 1981).
  84. .Id. at 118.