An Essay in Honor of Robert Bone

Essay - Online Edition - Volume 102


This Essay, in honor of Robert Bone and his new book, Justifying Litigation Reform,[1] considers the impact of Bone’s thinking on the academy. It teases out three themes in Bone’s work: critical appraisal of the “day-in-court ideal,” an affirmative vision of the role of risk distribution in civil litigation, and a re-envisioning of the relationship between procedure and substance. His identification and development of these ideas have been profoundly influential on my work and that of many other scholars I know.

A particularly important part of Bone’s contribution to the field is his commitment to rigor in the normative analysis of the procedural regime, pushing colleagues to articulate with greater specificity and sophistication what often come off in procedure scholarship as mere policy preferences. In his own work, he provides a normative analysis that is characterized by a welcome commitment to rigor and clarity of expression. Bone’s scholarship has been consistently analytically rigorous, using clearly delineated blended methodologies (mostly Rawlsian/Dworkinian and formal Law & Economics), all accompanied by an inventive spirit. In his generous engagement with other scholars and his egalitarian approach to those who hope to engage in his field, Bone has been and continues to be a model to emulate.

I. Critiquing the Day in Court Ideal

My introduction to Robert Bone’s rigorous and thoughtful scholarship was through his 1992 article Rethinking the “Day in Court Ideal” and Non-Party Preclusion.[2] This was one of the first articles I read when I started researching class actions. A significant part of my scholarship has been a continuing dialogue with this piece, in which Bone excavated a largely ignored history of non-party preclusion. He challenged “the standard assumption that the day in court ideal gives individuals a strong right to participate personally in all forms of litigation that concern them.”[3] In this article, Bone gave numerous historical examples from the nineteenth century of persons not being allowed to participate in lawsuits that affected their rights (particularly financial interests).[4] He showed how the twentieth century doctrines of interest representation (particularly in class actions) and the process-oriented theories of litigation were at war with one another.[5]

In his characteristically trenchant style, Bone also dismantled the arguments against virtual representation and in favor of the “day-in-court ideal.”[6] First, he argued the “tradition” of the “day-in-court ideal” was an invented one, given the prevalence of non-party preclusion in the nineteenth century.[7] Second, he questioned whether litigant satisfaction was in fact promoted by participation in litigation.[8] Third, he assessed whether the foundation for the idea that the “day-in-court ideal” was the source of normative legitimacy has been adequately laid.[9] He then provided rigorous arguments on both outcome-oriented and rights-based grounds, critiquing the idea that individual litigation was a necessary predicate to a valid adjudication. He concluded:

If the point of adjudication is to produce quality outcomes, then the demands of dignity should be satisfied in most situations by outcomes meeting the quality standards. Under these circumstances, it is difficult to see what institutional value there could possibly be in guaranteeing participation beyond what is needed for quality decisions and why anyone should have a right to demand more.[10]

What was important in Bone’s work then remains important today and is at the center of his very sophisticated book. That is that individual adjudication—the individual’s day in court—is not the core of the adjudicative process. The core of the adjudicative process is rectitude: the correct application of the law to the facts.

II. Defending Statistical Adjudication

Bone’s insight that individual adjudication was not the center of the adjudicative enterprise historically is important because it frees up the system to consider all manners of alternative, inventive procedures that better promote rectitude. One of these procedures is statistical adjudication. In another article which had a profound influence on my thinking, Statistical Adjudication: Rights, Justice, and Utility in a World of Process Scarcity,[11] Bone took on the normative question of whether it is possible to justify sampling procedures in adjudication. His article on non-party preclusion opened the door to innovative techniques not reliant on the “day-in-court” requirement, and Statistical Adjudication laid out a potential next step, inspired by real-world use of sampling to resolve mass lawsuits.[12]

One of the strengths of Bone’s work, and especially of this article and his book, is his ability to provide a nuanced evaluation of novel procedures. Instead of just arguing in favor of sampling, he carefully laid out the arguments in favor of and against sampling, and then explained the conditions under which sampling met the requirements of a both utilitarian and process-based justification for civil procedure. He explains in the article the germ of one of the most important ideas in his new book: that sampling can lead to unequal distribution of the risk of error among plaintiffs.[13] This is because, to put it simply, some plaintiffs whose actual damages differ from the average will nevertheless be given the average award. A sampling regime undercompensates the highest value cases. At the same time, it may overcompensate the lower-value cases. Even with more complex statistical techniques, Bone explained, there will be errors because the very point of a statistical regime is to avoid costly factual evaluation of individual cases.[14] Yet, that is what may be necessary for accuracy. [15]

Bone takes up this issue again in his book. Here he lays out more clearly the tension between the individualistic approach to litigation and rectitude as it plays out in mass cases.[16] Because cases cannot be tried at once—and often through no fault of their own, some plaintiffs file later than others and are therefore tried later than others—the latecomers may obtain a lower amount than those tried early.[17] This systemic error is avoided by a sampling regime if delay costs are high.[18] In some contexts, such as mass torts, some differences in outcomes that would occur absent sampling are morally irrelevant, Bone argues.[19]

In other situations, such as large-scale debt collection cases, sampling is even more fraught because individual differences are not the result of chance. If someone had no debt, because they did not incur one, then they should not have to pay the debt of the debtor class on average.[20] Here I can offer a friendly amendment: the class action might help. In a case called Sykes v. Mel S. Harris & Associates LLC,[21] lawyers brought a class action against a law firm alleging that it was purchasing debts of dubious validity, robo-signing lawsuits against the debtors (although it was required to file verified complaints), and then using sewer service so that the defendants would default.[22] The firm then allegedly used the legal system to collect the debts by garnishing wages or the like.[23] The plaintiffs were able to certify a class action against the firm, and the Second Circuit held that neither commonality nor predominance stood in the way of certification.[24] This was a good result under Bone’s theory. While we do not know what the situation was in each individual case, it is clear that this abuse of the system distributed error in a biased and unequal way.

The class action is not a solution to the problem of debt collection suits that do not pay attention to the circumstances of the individual case and the availability of individualized defenses, such as if a plaintiff is listed as owing the wrong amount, the interest owed was wrongly calculated, or the statute of limitations had run.[25] There may be other technological solutions that even out the risk of error between parties on a larger scale or old fashioned ones such as the availability of free or low-cost counsel or online proceedings.[26] A recent study demonstrated that individuals who lived farther from the courthouse were more likely to default in eviction proceedings, for example, a problem that could be fixed with online proceedings.[27]

The idea of thinking of the distribution of the risk of error, rather than merely avoiding error, raises an important substance-specific point that is at the core of Bone’s book.[28] When a procedure improves the distribution of the risk of error, it is justified. But when a procedure creates greater distribution of error-risk, it is not. Whether a procedure improves or impairs the fair distribution of the risk of error depends on the context. The effect of any procedural regime, either on outcomes or process values litigants expect, depends on context. A mass tort plaintiff who has suffered a significant injury (and therefore a potential payoff) is very differently situated vis-à-vis the procedural rules from a debtor-defendant who is one of a mass of similarly situated debtors. This is not only because of the differences in substantive law between products liability law and consumer protection laws. It is also because of the structure of the legal profession (that is, the debtor will be unable to obtain representation on a contingency fee basis because they will have no recovery); characteristics of the populations affected; and the type of litigants who are participating in the litigation. All of these are exogenous to the procedural system, raising an important question: how much should the procedural system account for differences in the situation of litigants before they come to court?

Bone, as I understand it, thinks that this is beyond the ken of the court system. It is a system within the larger social structure and if that larger social structure is rife with unfairness, then procedure is not the pathway toward curing that unfairness. This is not to say that we cannot recognize that unfairness, but only that we should also recognize the limitations of the system as we consider reforms. That said, I think he would surely agree that in-court reforms such as online hearings, which address exogenous problems that put the fundamental principles of the system (such as equal distribution of the risk of error or rectitude) at risk, should be adopted. This is why understanding the normative underpinnings of our system of adjudication is so important.

III. Critiquing the Substance/Procedure Distinction

How wedded should the system be to the context-specific molding of procedure to substance? In an article called Class Certification and the Substantive Merits,[29] Bone and his co-author David Evans argued that for class certification decisions, where so much is at stake for defendants and plaintiffs, the court should evaluate the substantive merits on a “likelihood of success” standard as courts do for preliminary injunctions, before certifying a class.[30] When I first read this article, I was very much against the idea on the grounds that previewing the merits at the class certification stage seemed to put the cart before the horse. In a world where evaluating litigation risk is more of an art than a science, should a prejudgment of a case early on be outcome-determinative?[31] This is particularly a problem if the case is to be decided by a jury.[32] What if a case appears weak at first, but later turns out to be meritorious?

In retrospect, developments have vindicated Bone and Evans’s position. Increasingly over the last ten years, courts have required substantial discovery prior to class certification. While class certification is supposed to be only about the requirements of Federal Rule of Civil Procedure 23, when the rule inquiry overlaps with the merits the courts are supposed to review the merits to the extent of that overlap.[33] Consider the Supreme Court’s approach to Wal-Mart Stores v. Dukes,[34] the most important class action case of the twenty-first century. There, the Court held that the plaintiffs lacked common questions susceptible to common answers.[35] What were the common questions that plaintiffs claimed in that case?

The female workers at Wal-Mart claimed that an environment of sexism pervaded the company and that as a result managers exercised their discretion with respect to promotions and pay in a discriminatory manner.[36] This legal theory was a bold one that had not yet been fully tested in the courts.[37] If the theory was correct, the idea was, then the policy decision to permit an environment of sexism was a centralized decision that pervaded the company’s decision-making down to the discretion exercised by local managers. The class would succeed or fail as a whole based on whether the theory was valid.

The Court rejected the theory.[38] The company’s policy, it explained, was precisely not to have a policy: to devolve decision-making to the local level.[39] Discretion could not be the basis of a claim because “left to their own devices most managers in any corporation—and surely most managers in a corporation that forbids sex discrimination—would select sex-neutral, performance-based criteria for hiring and promotion that produce no actionable disparity at all.”[40] This claim is provably false. As Justice Ginsburg explained in her concurrence in part:

An example vividly illustrates how subjective decisionmaking can be a vehicle for discrimination. Performing in symphony orchestras was long a male preserve. Goldin & Rouse, Orchestrating Impartiality: The Impact of “Blind” Auditions on Female Musicians, 90 Am. Econ. Rev. 715, 715–716 (2000). In the 1970’s orchestras began hiring musicians through auditions open to all comers. Id., at 716. Reviewers were to judge applicants solely on their musical abilities, yet subconscious bias led some reviewers to disfavor women. Orchestras that permitted reviewers to see the applicants hired far fewer female musicians than orchestras that conducted blind auditions, in which candidates played behind opaque screens. Id., at 738.[41]

The point here is not to relitigate the question of implicit bias in workplace decision-making, but rather to consider the relationship between procedure and substance. The majority opinion in Wal-Mart had to take into account its evaluation of the merits of the plaintiff’s claim (the substantive law) in order to make a decision about commonality (the procedural law). That decision on the merits was contested and not fully briefed because of the procedural posture of the case. Wal-Mart illustrates how the two became intertwined.

The Supreme Court subsequently admonished courts that they ought to only consider the merits when necessary to the decision on class certification,[42] but the move had been made toward integrating substance and procedure. Increasingly before Wal-Mart, and certainly now, class certification has become so elaborate an inquiry that it is nearly a summary judgment motion, looking even more onerous and substantive than the proposal originally made by Bone and Evans.[43] Would it be better to be more up-front about what courts are doing? That was the great merit of Bone and Evans’s proposal: that rather than consider the merits sub-rosa, under the guise of some procedural decision that is meant to appear neutral on its face, the courts take ownership over their substantive decisions. Further, there may be meritorious class actions that under today’s regime are rejected, so that plaintiffs cannot obtain compensation at all for the wrong done them, on commonality or predominance grounds. If the case is indeed meritorious and would not otherwise be heard because of the small size of recovery or the inability of individuals to bring their claims, aggregation is the way to fill that breach in enforcement of the law while balancing the concerns over both fairness and distribution of litigation risk between parties.

Conclusion: Rethinking Reform

At the core of Bone’s book Justifying Procedural Reform are two ideas that are consistent with his previous body of work. The first is that the primary function of procedure is rectitude—that is, the correct application of law to the facts of the case.[44] This point opens up a broader crucial discussion about the functions of litigation and how procedural reforms should address them. Bone’s approach to this question is a model of rigor.

Here I will add a slight amendment to his claims. While I agree with Bone that rectitude is the manifest function of the system, I think the system also has latent functions which are just as important and are baked into our procedural structure. I have argued that the court system plays a somewhat broader role in our social fabric: that is, it reveals information critical to self-governance,[45] allows forms of individual participation by allowing marginalized people to assert their rights and obtain recognition by the state, promotes a thin form of equality between persons, promotes participation in self-governance through jury service, and serves a critical function of law enforcement.[46] I agree that the story the litigation system tells about itself is the one Bone describes. But I disagree to the extent that the argument about the distinction between primary and secondary, or manifest and latent, functions dictates the scope of procedural reform.

This observation, or perhaps friendly amendment, does not undermine and may even strengthen the second argument Bone makes, which is that given that perfect rectitude is impossible, the risk of error should be evenly distributed across litigants. The import of this argument for understanding modern civil litigation’s problems and possible solutions cannot be overstated. These insights form the core of proposals to use statistical adjudication and consider other changes to procedure that take greater account of the importance of the cause of action to the polity. They provide a guide to future rule-makers as they tackle new technologies and new social problems that appear in the courts. The focus on accuracy has often been erroneously understood to require individualized adjudication—this assumption is at the bottom of the “day-in-court ideal”—but in many situations it is a non-individualized approach to litigation (such as that offered by the use of statistics to adjudicate mass claims, class action collective adjudication, or multi-district litigation) that offers the best hope for rectitude. This intervention is of extraordinary importance to both the study of procedure and its practice.

  1. . Robert G. Bone, Justifying Litigation Reform (2023) (unpublished manuscript) (on file with the author) (hereinafter Bone Manuscript).
  2. . Robert G. Bone, Rethinking the “Day in Court” Ideal and Nonparty Preclusion, 67 N.Y.U. L. Rev. 193 (1992).
  3. . Id. at 199.
  4. . Id. at 210.
  5. . Id. at 218.
  6. . Id. at 236.
  7. . Id.
  8. . Id.
  9. . Id.
  10. . Id. at 281–82.
  11. . Robert G. Bone, Statistical Adjudication: Rights, Justice, and Utility in A World of Process Scarcity, 46 Vand. L. Rev. 561 (1993).
  12. . Id. at 566–67.
  13. . Id. at 621; see generally, Bone Manuscript, supra note 1.
  14. . Bone, supra note 11, at 612.
  15. . Id. at 586 (“In an asbestos case, for example, the amount of individual damages depends on a number of complex and often hotly contested variables, such as exposure to carcinogens other than asbestos and the degree to which they contributed to the injury, the degree of exposure to asbestos itself, the actual severity of the disease, the amount of future lost earnings, and the degree of mental anguish suffered by the plaintiff. If some of these variables are ignored because of cost, the resulting unexplained error, combined with sampling and measurement error, may well exceed individual trial error.”).
  16. . See Bone Manuscript, supra note 1, at 106–16.
  17. . Id. at 110.
  18. . Id. at 112–14.
  19. . Id. at 158.
  20. . See id.
  21. . 780 F.3d 70 (2d Cir. 2015).
  22. . Id. at 75–76.
  23. . Id. at 76.
  24. . Id. at 87.
  25. . Thanks to David Marcus for this point. See also Judith Resnik & David Marcus, Inability to Pay: Court Debt Circa 2020, 98 N.C. L. Rev. 361, 362–63 (2020).
  26. . For a discussion of such possibilities through administrative adjudication, see David Ames, Cassandra Handan-Nader, Daniel E. Ho & David Marcus, Due Process and Mass Adjudication: Crisis and Reform, 72 Stan. L. Rev. 1, 2 (2020). For other technological solutions see Avital Mentovich, J.J. Prescott & Orna Rabinovich-Einy, Legitimacy and Online Proceedings: Procedural Justice, Access to Justice, and the Role of Income, 57 L. & Soc’y Rev. 189, 190 (2023).
  27. . David A. Hoffman & Anton Strezhnev, Longer Trips to Court Cause Evictions, 120 PNAS, 1, 2 (2023).
  28. . Bone Manuscript, supra note 1, at 75, 79.
  29. . Robert G. Bone & David S. Evans, Class Certification and the Substantive Merits, 51 Duke L.J. 1251 (2002).
  30. . Id. at 1254–55, 1279. See also Robert G. Bone, Who Decides? A Critical Look at Procedural Discretion, 28 Cardozo L. Rev. 1961, 2015 (2007) (arguing that rulemakers should amend Rule 16 to allow judges to signal their views on the merits early in a litigation).
  31. . Alexandra D. Lahav, Symmetry and Class Action Litigation, 60 UCLA L. Rev. 1494, 1521–22 (2013) (stating that “[t]he problem with this standard is that a case may have a low likelihood of success on the merits and still destined to win, yet be unable to obtain class certification. In small claims actions too small to litigate individually, the case would be barred merely because the likelihood of success is low”).
  32. . Despite being an equitable procedure, class actions are decided by a jury where the substantive law would require a jury under the Seventh Amendment. Whether this approach is correct has never been fully tested. See Samuel L. Bray, Equity, Law, and the Seventh Amendment, 100 Texas L. Rev. 467, 471 (2022) (suggesting that the right to a jury trial does not attach to class actions).
  33. . Albeit, only to the extent of that overlap.
  34. . 564 U.S. 338 (2011).
  35. . Id. at 338–39.
  36. . Id. at 338, 356.
  37. . For a description of the arguments during the relevant period see Samuel R. Bagenstos, Implicit Bias, “Science,” and Antidiscrimination Law, 1 Harv. L. & Pol’y Rev. 477, 479 (2007).
  38. . Wal-Mart, 564 U.S. at 338, 355–56.
  39. . Id. at 355 (stating that the company’s policy “is just the opposite of a uniform employment practice that would provide the commonality needed for a class action; it is a policy against having uniform employment practices”).
  40. . Id. at 355.
  41. . Id. at 373 n.6.
  42. . Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 455, 459 (2013) (“Rule 23(b)(3) requires a showing that questions common to the class predominate, not that those questions will be answered, on the merits, in favor of the class.”).
  43. . Alexandra D. Lahav, Procedural Design, 71 Vand. L. Rev. 821, 841–46 (2018) (explaining how the class certification motion has been transformed from a preliminary stage question to a question requiring substantial discovery and development).
  44. . Bone Manuscript, supra note 1, at 64.
  45. . See Elizabeth Chamblee Burch & Alexandra D. Lahav, Information for the Common Good in Mass Torts, 70 DePaul L. Rev. 345, 349–51 (2021). For a nuanced discussion on what publicness and secrecy in courts can mean, see Judith Resnik, A2J/A2K: Access to Justice, Access to Knowledge, and Economic Inequalities in Open Courts and Arbitrations, 96 N.C. L. Rev. 605, 606 (2018).
  46. . See Alexandra D. Lahav, In Praise of Litigation 1, 2 (2017); Alexandra D. Lahav, The Roles of Litigation in American Democracy, 65 Emory L.J. 1657, 1704 (2016).