Peeking under the tent of our nation’s largest and often most impactful cases reveals that judges often act like ringmasters: They dele­gate their authority to a wide array of magistrate judges, special masters, and settlement administrators. Some, like the American Bar Association, see this as a plus that promotes efficiency and cost savings. Critics, how­ever, contend that delegating judicial power, especially to private citizens, removes adjudication from public scrutiny, injects thorny ethical ques­tions about ex parte communications, and risks cronyism and high costs.

By constructing an original dataset of ninety-two multidistrict prod­ucts liability proceedings centralized over fourteen years, we introduce the first taxonomy of the diverse adjuncts working within them. Testing adjuncts’ effects with a multivariate analysis, we found that proceedings with special masters lasted 66% longer than those without, and appoint­ing any kind of adjunct meant that the proceeding was 43% less likely to end. Not only did justice take longer, it cost more: 74% of the adjuncts were not magistrate judges, meaning that the parties paid them.

Digging deeper, we interviewed some of the lawyers, judges, and adjuncts who participated in these proceedings. Attorneys’ experiences moved scholars’ concerns from law review pages to real life: Rather than improving justice, some adjuncts cajole parties through off-the-record discussions; repeat players tap one another for business; and plaintiffs’ outcomes may depend more on whether they picked an attorney with the inside track than their suits’ merits. Collectively, our findings support existing reservations about allocating judicial power to those in the pri­vate sector.

The full text of this Article can be found by clicking the PDF link to the left.


Shortly after he received the massive Opiate multidistrict litigation (MDL), the transferee judge appointed three special masters to do everything from meeting with the parties and mediating disputes to analyzing legal submissions, coordinating with other courts, and interpreting par­ties’ agreements. 1 Appointment Order at 1–3, In re Nat’l Prescription Opiate Litig., No. 1:17-MD-2804 (N.D. Ohio Jan. 11, 2018). Indeed, it was his special master who proposed a novel “negotiation class” as a means to unite cities and counties pre-settlement. 2 Francis E. McGovern & William B. Rubenstein, The Negotiation Class: A Cooperative Approach to Class Actions Involving Large Stakeholders, Tex. L. Rev. (forthcoming 2020) (manuscript at 22–23), https://papers.ssrn.com/abstract=3403834 (on file with the Columbia Law Review). The court compensated these special masters “at their current MDL rates” but did not disclose them, allowing each to file itemized fee statements under seal instead. 3 Appointment Order, supra note 1, at 5–6. Appointments included Cathy Yanni, an arbitrator who works for the for-profit arbitration company JAMS; the late Francis McGovern, who was a professor at Duke Law School; and David Cohen, a professional special master and former president of the Academy of Court-Appointed Masters. 4 E.g., David R. Cohen, Special Masters Versus Magistrate Judges: No Contest, Fed. Law., Sept. 2014, at 73, 73 [hereinafter Cohen, Special Masters Versus Magistrate Judges]. We were deeply saddened to learn of Francis McGovern’s recent death.

For more than two decades, judges presiding over mass-tort proceedings like these have parceled their authority out to a range of “judicial adjuncts”—non-Article III judges who are judicially appointed to perform judicial or administrative tasks within a specific proceeding and range from public magistrate judges to privately paid special masters, claims administrators, escrow agents, and settlement masters. 5 “Judicial adjunct” is a short-hand umbrella term that encompasses a wide array of positions that range from special masters appointed under Federal Rule of Civil Procedure 53 to magistrate judges appointed under Article I of the U.S. Constitution. See infra note 29 and accompanying text. For readability, we sometimes use the shorthand term “magistrate” rather than the full title “magistrate judge,” which is the proper title. We mean no offense by this. Often appointed and rarely studied, these adjuncts work behind the scenes, out of the spotlight, yet wield enormous power over the nation’s largest and often most significant cases.

Outsourcing judicial power to private adjuncts raises no shortage of questions about cost, transparency, horizontal equity, oversight, and accountability. On the defense side, companies pay private adjunct fees on top of their white-shoe lawyers’ hourly rates. 6 See Thomas E. Willging, Laural L. Hooper, Marie Leary, Dean Miletich, Robert Timothy Reagan & John Shapard, Special Masters’ Incidence and Activity: Report to the Judicial Conference’s Advisory Committee on Civil Rules and Its Subcommittee on Special Masters 42 (2000), https://www.fjc.gov/sites/default/files/2012/SpecMast.pdf [https://perma.cc/BXX3-M3TQ]. And on the plaintiffs’ side, contingent-fee attorneys eventually deduct adjunct fees as a litigation cost, which means that it is not typically lawyers’ bottom line that suffers but plaintiffs’ settlement amounts. 7 See infra section III.B.3. Deeper questions abound as well: If private adjuncts must depend on the attorneys for future appointments and income, can they be neutral? What effect might their appointment have on creating precedent and adhering to the rule of law?

Without addressing these questions, the American Bar Association (ABA) recently called for courts to appoint one type of adjunct, special masters, regularly in MDLs. 8 ABA, Resolutions with Reports to the House of Delegates: 2019 Midyear Meeting Res. 100, at 51 (2019), https://www.americanbar.org/content/dam/aba/administrative/house_of_delegates/ebook-of-resolutions-with-reports/2019-midyear-ebook-of-resolutions-with-reports.pdf [https://perma.cc/LGH7-8EDF] [hereinafter ABA Resolutions]; ABA, Summary of Action of the House of Delegates: 2019 Midyear Meeting 20–21 (2019), https://www.americanbar.org/content/dam/aba/administrative/house_of_delegates/summary-of-action/summary-of-action-midyear-2019.pdf [https://perma.cc/TE56-7YYB] (adopting the guidelines). Multidistrict litigation is a procedure that the federal courts use to coordinate cases with at least one common factual question before the same judge for pretrial handling. See 28 U.S.C. § 1407 (2018). In practice, however, few cases are ever remanded to their court of origin, which makes what happens in the MDL all the more important. See Elizabeth Chamblee Burch, Remanding Multidistrict Litigation, 75 La. L. Rev. 399, 400 (2014) [hereinafter Burch, Remanding Multidistrict Litigation] (“As of 2010, the Panel remanded only 3.425% of cases to their original districts.”). Creating an MDL means that the Judicial Panel on Multidistrict Litigation has decided that dispersed federal cases share enough factual similarities that transferring and centralizing them before the same judge for pretrial purposes promotes justice and effi­ciency. 9 28 U.S.C. § 1407(a). MDL has thus become the go-to mechanism for handling com­plex, high-profile mass torts, 10 See Burch, Remanding Multidistrict Litigation, supra note 8, at 417 (recognizing the expediency of settling complex cases as a key success of multidistrict litigation). such as cases against Juul over e-cigarettes 11 See Transfer Order at 1–2, In re Juul Labs, Inc., Mktg., Sales Pracs., & Prods. Liab. Litig., MDL No. 2913 (J.P.M.L. Oct. 2, 2019). and Johnson & Johnson over talcum powder. 12 See Transfer Order at 1–2, In re Johnson & Johnson Talcum Powder Prods., MDL No. 2738 (J.P.M.L. Oct. 4, 2016). The “transferee judge,” the judge who receives these cases, bears enormous responsibility: The vast majority of cases are resolved by motion or settlement within the MDL. Historically, fewer than 3% of all transferred cases are ever remanded to their court of origin. 13 Burch, Remanding Multidistrict Litigation, supra note 8, at 400–01.

The ABA’s report proposed that MDLs in particular could “benefit from specialized expertise” and that “[e]ffective special masters reduce costs by dealing with issues before they evolve into disputes and by swiftly and efficiently disposing of disputes that do arise.” 14 ABA Resolutions, supra note 8, at 56–57. But see Owen M. Fiss, The Bureaucratization of the Judiciary, 92 Yale L.J. 1442, 1466 (1983) [hereinafter Fiss, Bureaucratization] (“There is an undeniable freshness and richness to the judgments of a non-specialist. There is also less chance of capture by a special interest group.”). The ABA’s resolution thus urged judges to appoint special masters at “the outset of litigation” and permit them to do everything from overseeing discovery and pretrial litigation to conducting trials based on parties’ consent, allocating settlements, and administering claims. 15 ABA Resolutions, supra note 8, at 59–60. Failing to do so, it cautioned, “[r]egardless of the reason, . . . may disserve the goal of securing ‘a just, speedy, and inexpensive determination.’” 16 Id. at 59 (quoting Fed. R. Civ. P. 1). Yet neither this reproach nor the ABA’s claims about special masters’ utility included systematic empiri­cal support. 17 The ABA commissioned a study of the use of special masters. Data collection problems resulted in case studies that highlighted, but could not test, how special masters were used and what their impact was. The two case studies did not involve multidistrict proceedings. Barbara Meierhoefer, Special Master Case Studies 31 (2018), https://www.americanbar.org/content/dam/aba/publications/judicial_division/2018lc-specialmasterscasestudy.pdf [https://perma.cc/7A5V-N2LL] (noting that the response rate to the attorney survey pool was only 14%).

In contrast to the ABA’s position, Federal Rule of Civil Procedure 53 condones appointing special masters for issues “that cannot be effectively and timely addressed by an available district court judge or magistrate judge.” 18 Fed. R. Civ. P. 53(a)(1)(C). Put simply, it creates a presumption in favor of appointing mag­istrates in lieu of special masters. 19 Fed. R. Civ. P. 53 advisory committee’s note on 2003 Amendment (noting that “appointment of a master must be the exception and not the rule”); George C. Hanks, Jr., Searching from Within: The Role of Magistrate Judges in Federal Multi-District Litigation, 8 Fed. Cts. L. Rev. 35, 44 (2015). And though the Manual for Complex Litigation suggests that courts enjoy broad authority to designate experts, technical advisors, and special masters, it too cautions that these judicial adjuncts can increase the already high costs of complex cases, warns that “[t]ruly neutral experts are difficult to find,” and suggests that it may be hard to know early on whether appointments are warranted. 20 Fed. Jud. Ctr., Manual for Complex Litigation §§ 11.51–11.52 (4th ed. 2004), https://www.uscourts.gov/sites/default/files/mcl4.pdf [https://perma.cc/9P72-5PQV]. “Reference to a special master must be the exception and not the rule,” the Manual explains. 21 Id. § 11.52. Even the Federal Judicial Center’s Pocket Guide for Transferee Judges warns, “In a products liability MDL, it may be particularly difficult to appoint a completely disinterested special master with no prior relation­ship to any of the parties, since special masters are often practicing attor­neys and tend to have substantial experience with similar disputes.” 22 Barbara J. Rothstein & Catherine R. Borden, Fed. Jud. Ctr., Managing Multidistrict Litigation in Products Liability Cases: A Pocket Guide for Transferee Judges 38 (2011), https://www.fjc.gov/sites/default/files/2012/MDLGdePL.pdf [https://perma.cc/B3MD-YCLH]. In other words, special mastering is a business, and like most businesses, it requires repeat patrons.

Of course, the latest version of the Manual was written in 2004 and the Pocket Guide in 2011. The Federal Judicial Center last studied special masters twenty years ago. 23 See Willging et al., supra note 6. Considering all civil cases, not just complex ones, that study found that judges appointed special masters in fewer than two of every 1,000 cases (0.2%). 24 Id. at 3. Judges were slightly more likely to consider appointing a special master in cases involving patents, environmental matters, and airplane personal injuries. Id. But several years thereafter, one federal judge predicted “that we will see more and varied appointments of such adjuncts by the year 2020, to the benefit of the courts.” 25 Shira Scheindlin, We Need Help: The Increasing Use of Special Masters in Federal Court, 58 DePaul L. Rev. 479, 486 (2009).

It’s now 2020. This Article wades into this controversy to set aside the crystal ball and offer a fresh empirical look at all court-appointed adjuncts in products liability MDLs—not just special masters and magistrates. We built and hand coded an original dataset from all products liability proceedings centralized from 2004 to 2017 that closed by April 15, 2019—ninety-two proceedings that included class and nonclass settlements. 26 For a breakdown of outcomes, see infra section III.A.2. When a proceeding “closes,” it means that the entire proceeding and all of the actions within it have either concluded or been remanded to their courts of origin such that the transferee judge has nothing left to do. See, e.g., U.S. Jud. Panel on Multidistrict Litig., MDL Statistics Report – Docket Summary Listing 1 (2020), https://www.jpml.uscourts.gov/sites/jpml/files/Recently_Terminated_MDLs-January%201-September-15-2020.pdf [https://perma.cc/ND52-C3N4] (citing date closed). It is, to our knowledge, the first attempt that anyone has made to look under the tent and document courts’ interaction with and oversight of a growing industry that thrives upon settlements. 27 As Professor Deborah Hensler observed, there is a particular need for legal schol­arship on claims resolution facilities, or what we call settlement programs. See Deborah R. Hensler, Alternative Courts? Litigation-Induced Claims Resolution Facilities, 57 Stan. L. Rev. 1429, 1435–36 (2005) [hereinafter Hensler, Alternative Courts] (“[L]egal scholars have not delved deeply into the rules that govern claims resolution facilities or how those rules are implemented.”); Deborah R. Hensler, Assessing Claims Resolution Facilities: What We Need to Know, 53 Law & Contemp. Probs. 175, 176–78 (1990) (proposing “an agenda for research on claims resolution facilities” that could assist in providing better alternatives for the future). We join Professor Hensler in requesting more transparency as to what hap­pens within these settlement programs such that scholars can empirically assess the many substantive and procedural questions that remain unanswered. Our deep dive into appointments, costs, and proceedings’ duration ultimately prompts us to raise a caution flag about outsourcing judicial duties to private adjuncts.

In the existing academic literature, outsourcing has been both celebrated and condemned, often as the judicial role itself changes. Part I situates the debate amid judges’ shift from dispassionate arbiters to case managers to, now, supervisors. It also debunks some persistent myths about “burgeoning” caseload statistics along the way. Although surveying the existing literature on magistrates and special masters exposes a turf war between their various proponents, the primary purpose is to set the stage for our inquiry into how judges use these adjuncts in MDLs and high­light the remarkable scarcity of work on less traditional adjuncts like claims administrators, lien resolution companies, banks, and escrow agents.

Part II dives into the MDL ecosystem, placing our dataset within the world of multidistrict proceedings and debuting a taxonomy of the cast of characters who appear within it. Despite a call to arms over “a crisis in the courts,” 28 Mark A. Fellows & Roger S. Haydock, Federal Court Special Masters: A Vital Resource in the Era of Complex Litigation, 31 Wm. Mitchell L. Rev. 1269, 1287 (2005). this Part demonstrates that today’s judges appoint what we collectively dub “judicial adjuncts” with no greater frequency than they did in past years with a lighter workload. 29 This Article builds on Professor Linda Silberman’s term. Linda Silberman, Judicial Adjuncts Revisited: The Proliferation of Ad Hoc Procedure, 137 U. Pa. L. Rev. 2131, 2131–32 (1989) [hereinafter Silberman, Judicial Adjuncts Revisited] (using the term “judicial adjuncts” primarily to refer to masters and magistrates, who customize procedure for spe­cific cases). Adjuncts still appear at every stage: pretrial, settlement negotiations, and post-settlement. 30 See id. at 2135–36, 2139–40, 2169–70 (describing some of the roles played by judicial adjuncts in litigation). What we did discover, however, was a vast settlement support network comprised of banks, claims administrators, notice experts, certified public accountants, and lien resolution administrators. 31 See infra section II.B.3.

Part III moves from tallying and classifying these adjuncts to statistically analyzing their effects. In so doing, we are mindful that data and analytics can tell us only so much; justice cannot be measured solely by how long a proceeding lasts nor can one measure intangibles like the benefits or drawbacks of allowing adjuncts to operate behind the scenes. 32 See generally Miguel F. P. de Figueiredo, Alexandra D. Lahav & Peter Siegelman, The Six-Month List and the Unintended Consequences of Judicial Accountability, 105 Cornell L. Rev. 363 (2020) (finding evidence suggesting that the six-month list, which evaluates how long a motion has been pending on a judge’s docket, may prompt judges to make more errors). Nevertheless, these findings should help inform courts’ cost-benefit calculus, and two key results are highlighted here.

First, it was surprising to find that proceedings with special masters lasted 66% longer than those without. 33 See infra section III.A.4. Of course, this raised a causal which-came-first question: Did judges appoint special masters to larger proceedings anticipating that they would be more difficult (and thus take longer) to resolve? Or did the special masters themselves cause that delay? Using a duration model made it possible to screen some noise by control­ling for a proceeding’s outcome (settlements uniformly took longer), per­sonal injury claims (which likewise took longer), and the number of actions (the more actions, the longer the proceeding lasted). 34 See infra section III.A.4. We found that appointing a judicial adjunct of any kind made the proceedings con­tinue longer than they otherwise would, all else being equal. 35 See infra section III.A.4. Although adding another plaintiff’s action to a proceeding had almost no effect on how long it lasted, designating an adjunct meant that the proceeding was 43% less likely to end. 36 See infra section III.A.4. And for every additional adjunct appointed, there was a 12% decrease in the probability of a proceeding ending. 37 See infra section III.A.4.

These findings raise questions about whether efficiency claims can justify appointing judicial adjuncts. To be sure, there may be other benefits such as specialized expertise and,  perhaps  more  controversially,  the  ability  to  talk  with  parties  behind  closed doors, 38 See Silberman, Judicial Adjuncts Revisited, supra note 29, at 2147 (discussing the potential for abuse that arises from ex parte communication with adjuncts). but if judges rely on adjuncts primarily to ease their caseload and speed proceedings along, then they ought to reconsider.

Second, costs rise to the fore because parties are paying for both lengthier lawsuits and private adjuncts. But in considering adjuncts’ costs, we ran into a significant roadblock: Compensation information was either undisclosed or affirmatively sealed for 62.8% of private adjunct appointments. 39 See infra Table 15. Even though we couldn’t always identify the amounts charged, we were able to discern that plaintiffs alone bore the costs for 54% of private adjuncts, meaning that in over half of the appointments, defendants did not contribute. 40 See infra section III.B.3. Some of the payments that we could unearth ran into the millions. In the Actos proceeding, for instance, Special Master Gary Russo charged over $4.7 million, and Deputy Special Master Kenneth DeJean charged over $1.3 million. 41 Order Appointing Special Masters at 1, 6, In re Actos (Pioglitazone) Prods. Liab. Litig., No. 6:11-md-02299-RFD-PJH (W.D. La. Apr. 11, 2012); infra Table 16. We tabulated sixty-five docket entries to calculate the aggregate numbers for DeJean and sixty-four docket entries for Russo. To administer the Zyprexa settlement, Special Settlement Masters Ken Feinberg, Michael Rozen, Cathy Yanni, and John Trotter collectively charged over $9.4 million. 42 Order at 1, In re Zyprexa Prods. Liab. Litig., No. 1:04-md-01596 (E.D.N.Y. Jan. 29, 2007).

If proceedings with adjuncts cost more and last longer, why do judges appoint them? To piece together this puzzle, we supplemented our quantitative analysis with twenty-two semi-structured, confidential interviews. These interviews include conversations with special masters, magistrate judges, claims administrators, district judges, and plaintiffs’ and defense attorneys with a wealth of experience not only in our proceedings but also in MDLs spanning back to the 1960s. 43 For more on methodology, see infra Appendix. Interviews revealed two competing narratives, which Part IV highlights. In one version, courts outsourced to effectively manage complex cases behind the scenes and closely monitored those appointed. In the other, repeat players in both the bar and the private-adjunct sector came to mutually beneficial arrangements that exposed real-life problems over capture, self-dealing, bias, transparency, and ad hoc procedures. 44 See infra Part IV.

Shining light into this segment of the MDL world made it clear that privatization no longer exists on a parallel track with Article III courts—judges outsource their own power to private actors. As privatization occurs under the aura of the federal courts, it raises central questions about court access, cost, impartiality, and fairness, which intersect with existing literatures debating courts’ purpose, ad hoc rulemaking, and the judicial role.

Part V thus steps back to offer some larger institutional lessons and chart a path forward. As courts delegate authority to private adjuncts who, more often than not, plaintiffs must compensate, justice may come with a heavier price tag for those least able to afford it. Judges sometimes cite the enormous expense of complex proceedings as a whole as a reason to appoint an adjunct, but moving from the grandiose to the granular shows that those choices have real impact on individual plaintiffs. In the pelvic-mesh proceedings, for instance, one plaintiff’s settlement statement revealed a $6,200 fee for “settlement program expense allocation” plus a “Cathy Yanni Appeal Cost” of $2,000. 45 Confidential Claim Closing Statement (on file with authors). In the Ethicon proceeding, of all the private adjunct appointments, only Cathy Yanni’s settlement-master fee was sporadically disclosed, and it varied. To “appeal” whatever amount she awarded, a plaintiff always had to pay $2,000, but her initial review in one settlement cost $300 per claim plus $10,000 per calendar quarter. In another she charged $350 per claim, and in yet a third, claim review cost a flat $300. See Pretrial Order #335 (Order Appointing Cathy Yanni as Special Master for Private Settlement Agreements Between Ethicon & Certain Plaintiffs’ Counsel) at 3, In re Ethicon, Inc. Pelvic Repair Sys. Prods. Liab. Litig., MDL No. 2327 (S.D. W. Va. Mar. 15, 2019); Pretrial Order #253 (Order Appointing Cathy Yanni as Special Master for Private Settlement Agreements Between Ethicon & Certain Plaintiffs’ Counsel) at 3, In re Ethicon, Inc. Pelvic Repair Sys. Prods. Liab. Litig., MDL No. 2327 (S.D. W. Va. filed Apr. 28, 2017); Pretrial Order #236 (Order Appointing Cathy Yanni as Special Master for Private Settlement Agreements Between Ethicon & Certain Plaintiffs’ Counsel) at 3, In re Ethicon, Inc. Pelvic Repair Sys. Prods. Liab. Litig., MDL No. 2327 (S.D. W. Va. filed Sept. 8, 2016). Yanni’s contrasting charges, with some mesh plaintiffs having to pay more for administrative costs than others, illustrate some of the controversies these appointments entail. Orders are available at https://www.wvsd.uscourts.gov/MDL/ethicon/orders.html. Costs, attorneys’ fees, and manda­tory medical-lien holdbacks whittled a settlement offer of $195,000 down to less than $60,000. 46 Confidential Claim Closing Statement, supra note 45. As the New York Times reported, this plaintiff’s expe­rience was not an anomaly: The average pelvic-mesh plaintiff received “less than $60,000” despite horrific injuries. 47 Matthew Goldstein, As Pelvic Mesh Settlements Near $8 Billion, Women Question Lawyers’ Fees, N.Y. Times (Feb. 1, 2019), https://www.nytimes.com/2019/02/01/business/pelvic-mesh-settlements-lawyers.html (on file with the Columbia Law Review). Medical costs alone can wipe plaintiffs out financially, and one claims administrator estimated that at least 15% of mesh plaintiffs were bankrupt. 48 Telephone and In-Person Interviews with Products Liability MDL Actors (Dec. 1, 2019 through Mar. 1, 2020) (transcript on file with authors) [hereinafter Interviews]. Reducing costs—measured in both time and actual dollars—is thus crucial.

Appointing magistrate judges, whose salaries come from the general tax revenue, is often a better alternative. As salaried public employees, magistrates are further insulated from the capture, bias, and self-interest concerns that plague party-selected, party-compensated private adjuncts. 49 See infra Part V. They can (and have) performed the same work as special masters, and sometimes even claims administrators, for far less expense. 50 See infra section III.B. They are already ensconced within the federal courthouse and they possess the legitimacy of federal office. 51 See infra section I.A. Plus, they are accustomed to the ways in which judges review their decisions, and clear paths exist for correcting error. 52 See infra section V.B. Far from complaining about the workload, our interviews suggest, in harmony with previous findings, that many magistrate judges want and enjoy MDL work. 53 Interviews, supra note 48; see also Hanks, supra note 19, at 52–53. As products liability MDLs play a policing role that impacts public health, it makes sense to use the general tax base to com­pensate the public servants who help resolve these disputes and to ensure that the work itself is both public and reviewable. 54 In this Article, “policing” means investigating wrongdoing and holding companies civilly responsible for faulty products by seeking compensatory and, when warranted, puni­tive damages. Ensuring that this work occurs in the open addresses a recent criticism that MDLs lack transparency. See, e.g., Michelle Conlin, Dan Levine & Lisa Girion, Why Big Business Can Count on Courts to Keep Its Deadly Secrets, Reuters (Dec. 19, 2019), https://www.reuters.com/investigates/special-report/usa-courts-secrecy-lobbyist [https://perma.cc/R2YU-BT6H]; Jaimi Dowdell & Benjamin Lesser, These Lawyers Battle Corporate America​—and Keep Its Secrets, Reuters (Nov. 7, 2019), https://www.reuters.com/investigates/special-report/usa-courts-secrecy-lawyers [https://perma.cc/SA4V-SBWP]; Benjamin Lesser, Dan Levine, Lisa Girion & Jaimi Dowdell, How Judges Added to the Grim Toll of Opioids, Reuters (June 25, 2019), https://www.reuters.com/investigates/special-report/usa-courts-secrecy-judges [https://perma.cc/768U-KBHL]; Dan Levine, Court Let Merck Hide Secrets About a Popular Drug’s Risks, Reuters (Sept. 11, 2019), https://www.reuters.com/investigates/special-report/usa-courts-secrecy-propecia [https://perma.cc/7Z8H-PACV].