PARKINSON – FINAL CLASS ACTIONS AS FIRMS Alex Atticus Parkinson∗ Class action literature is fixated on principal-agent problems—a near-singular focus that has largely been tracked by the courts, which frequently treat the class action mechanism as a Pandora’s box of agency costs. The ascent of the “principal-agent framework” is unsurprising. The divide between class members and class counsel gives rise to a host of classical principal-agent problems: self- interested class counsel often have incentives to act contrary to the interests of class members, and class representatives are poor monitors of such behavior. However, the hegemony of this framework can misguide courts, which not only aggressively regulate genuine principal-agent problems, as they should, but also regularly constrain the size and scope of class actions without explaining why doing so mitigates agency costs. That said, the principal-agent framework rests on an often-assumed, important, and inarguable foundation: principal-agent problems arise because class actions separate ownership (litigation interests belonging to class members) from management (class counsel). This Article refracts the frame of analysis by one degree—from the principal- agent framework to Coasean firm analysis. Descriptively, the class action is a like firm. Class actions are, like firms, comprised of ownership interests that aggregate to avoid the higher costs of acting alone in the market, cede control to a manager to achieve efficiencies, and operate pursuant to a profit motive. Moreover, like a firm, there is a natural—almost organic—limit to the growth of a class action: it will expand until litigating outside the class action would prove less costly than additional “growth.” This Article’s “firm framework” is consonant with the most useful components of the principal-agent framework—those that rest on the separation of ∗ I am grateful to the following individuals for helpful discussions, suggestions, and comments: Daniel Abebe, Douglas G. Baird, Brian Fitzpatrick, William H.J. Hubbard, Aziz Huq, Alexandra Lahav, Saul Levmore, Anup Malani, Jonathan Masur, and William Rubenstein. All views expressed here are my own. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 741 ownership and management—yet breaks with the traditionalist account in important ways. I.   Introduction .................................................................. 741   II.   The Dominance of the Principal-Agent Framework .. 745   A.   Class Action Literature and the Principal- Agent Framework ................................................. 746   B.   Impact on the Courts ............................................ 751   1.   Regulating Internal Governance: Compensation Problems ................................. 751   2.   Regulating External Boundaries: Monitoring Problems ....................................... 755   C.   Refracting the Frame One Degree ....................... 759   III.   Class Actions as Firms ................................................ 761   A.   Parallels: Class Actions and Firms ..................... 761   1.   Inception: Separation of Ownership and Management .................................................... 762   2.   Purpose: Profit ................................................. 764   3.   Structure: Tripartite Organization ................ 769   B.   Framework: Selecting a Theory of This Firm ..... 771   1.   Managerialism ................................................. 772   2.   Coase ................................................................ 772   IV.   Doctrinal Entailments: Tiers of Scrutiny ................... 777   A.   Internal Governance: Principal-Agent Issues ..... 779   1.   Doctrinal Fit: Compensation .......................... 782   2.   Doctrinal Change: Adequacy .......................... 785   B.   External Boundaries: Make-or-Buy Decisions .... 790   1.   Doctrinal Fit: Numerosity .............................. 794   2.   Doctrinal Change: Commonality and Predominance .................................................. 798   V.   Exchanging Internal Scrutiny for External Deference ...................................................................... 807   VI.   Conclusion .................................................................... 821   I. INTRODUCTION What is the optimal size of a class action and what is the best way to monitor class counsel? These questions have motivated class action scholarship and litigation for decades. PARKINSON – FINAL 742 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 For answers, one must piece together scores of opinions and hundreds of treatises, commentaries, and articles—rare is the source that claims to advance a unified framework. Many argue that class actions should be smaller, infrequently certified, and monitored with a skeptical judicial eye. Indeed, extant literature is dominated by portrayals of plaintiffs’ attorneys as unscrupulous raiders and class actions as principal-agent problems waiting to happen.1 This near- singular focus is often tracked by the courts, many of which treat class actions as a Pandora’s box of agency costs.2 This “principal-agent framework” is not without foundation: the divide between class counsel and class members indeed gives rise to a host of principal-agent problems and accompanying agency costs. Self-interested counsel often has the opportunity and incentive to act in a manner that is contrary to the interests of the class, and class members are poor monitors of such behavior.3 However, this framework is incomplete: not all class action issues are principal-agent issues. Moreover, even where principal-agent problems arise, their attendant agency costs and the responses they demand differ. While some class action issues present pernicious principal-agent problems that demand 1 See generally, e.g., John C. Coffee, Jr., Rescuing the Private Attorney General: Why the Model of the Lawyer as Bounty Hunter Is Not Working, 42 MD. L. REV. 215 (1983); see also infra Section II.A. 2 See infra Section II.B. 3 Pauline T. Kim, Beyond Principal-Agent Theories: Law and the Judicial Hierarchy, 105 NW. U. L. REV. 535, 574 (2011) (noting that “principal-agent theories suggest that informational asymmetries allow agents to avoid scrutiny of their activities and thereby afford them greater discretion”); see Christopher R. Leslie, Cartels, Agency Costs, and Finding Virtue in Faithless Agents, 49 WM. & MARY L. REV. 1621, 1636 (2008) (“The lawyer acts as a faithless agent when she pursues her own interests at the expense of her client’s. This is a particular problem in class action litigation because there are so many putative principals . . . . This sometimes results in, for example, class counsel negotiating low settlements in exchange for defendants’ payment of relatively high attorneys’ fees.”). See generally Eric W. Orts, Shirking and Sharking: A Legal Theory of the Firm, 16 YALE L. & POL’Y REV. 265 (1998). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 743 interventionist regulation, others can be self-regulating.4 The status quo’s lone class action framework provides no means for differentiating along these dimensions; rather, to it all class action issues present principal-agent problems of equal agency costs. This blunt framework cannot provide precise guidance to courts of limited resources and specialized competencies applying Rule 23.5 Fortunately, portions of the status quo’s preferred framework can be salvaged to fashion a more complete alternative. The principal-agent framework rests on a foundation that is perhaps so obvious as to be left unstated: class actions separate ownership interests (belonging to class members) from management (class counsel). Proceeding from that point of stasis, this Article refracts the frame of analysis by one degree—from agency costs to transaction costs. Turning on Coasean analysis, this Article’s “firm framework” aims to provide guidance for discerning between class action issues that present principal-agent problems that demand regulation, and those that are self-regulating or otherwise ill-suited to judicial intervention.6 Specifically, the class action is—or, more accurately, should be evaluated and analyzed as—a firm.7 Like firms, 4 See infra Part IV. 5 See FED. R. CIV. P. 23. 6 This Article concerns itself with so-called “damages classes,” which are filed pursuant to Rule 23(b)(3) of the Federal Rules of Civil Procedure and typically involve the amalgamation of a number of plaintiffs with relatively low-stakes damages claims. This is in contrast to injunctive- relief classes, which are governed by Rule 23(b)(2). The former are considered “paradigm class actions.” Jules Coleman & Charles Silver, Justice in Settlements, SOC. PHIL. & POL’Y, Oct. 1986, at 102, 122. Furthermore, the firm framework advanced by this Article concerns putative damages classes, rather than settlement classes. The market- mimicking forces that apply to the former are rendered inapplicable in the latter, for which class counsel has minimal incentive to consider the marginal costs and benefits of adding an additional claim to the class action. See generally Brian T. Fitzpatrick, An Empirical Study of Class Action Settlements and Their Fee Awards, 7 J. EMP. L. STUD. 811 (2010). 7 See David L. Shapiro, Class Actions: The Class as Party and Client, 73 NOTRE DAME L. REV. 913, 917 (1998). PARKINSON – FINAL 744 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 class actions are collective enterprises that separate ownership and management, aggregate resources to avoid the higher costs of transacting on the market, cede control to a centralized manager to realize efficiency gains, and organize around a profit motive. Class actions also contain a structure that not only resembles a firm but also provides a useful architecture for balancing tiers of interests. Moreover, like a firm, there is a natural limit to the growth of a class action’s external boundaries—class actions expand to the point that litigating individually and outside the class would be less costly than litigating within the “firm.” The firm framework is more complete than the principal- agent framework. It also permits courts of limited resources and specialized competencies to channel their efforts in welfare-maximizing ways. Moreover, the firm framework is consonant with the most promising elements of the principal-agent framework; even if the size (or external boundary) of a firm is often self-regulating and influenced by market forces,8 class actions are, clearly, not immune to internal principal-agent problems. These differing phenomena—one relating to external boundaries, the other to internal governance—demand tailored responses. Where principal-agent problems are likely to persist unchecked by self-regulation or market mechanisms in the context of a firm—internal governance—courts should apply Rule 23 with a heavy hand. Courts embrace this prescription as applied to attorneys’ fees, which present a clear principal- agent problem that is neither self-regulating nor addressed by any market-mimicking mechanism in the class action context. Therefore, courts apply Rule 23 to better align the principal-agent relationship, and they do so in a manner that is confidently within courts’ core institutional competencies—indeed, drawing on corporate law, judges have considerable experience scrutinizing the internal 8 Though perhaps the size of a class action “firm” is not as “consciously” regulated as the size of a classical firm in the corporate context. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 745 workings of firms.9 By contrast, when presented with an issue concerning a class action’s size or scope—external boundaries—courts should apply Rule 23 with a light touch. The costs attendant to a class action’s size and scope are often determined by market-mimicking mechanisms and are occasionally self-regulating. This approach “fits” with the way that courts treat an ostensibly forgotten certification requirement: numerosity.10 However, the firm framework advises that courts apply the Rule 23 commonality and predominance requirements with less scrutiny than in the status quo.11 The Article proceeds in four parts. First, it outlines the principal-agent framework, tracking its rise from scholarship to the courts. Second, the Article advances the firm framework. Third, it canvasses a sampling of the doctrinal entailments that follow from the firm framework, highlighting where these prescriptions break with the principal-agent framework. Finally, the Article concludes with a stylized model illustrating how the firm framework’s prescriptions might be used to ensure that class counsel self- regulates via cost internalization. II. THE DOMINANCE OF THE PRINCIPAL-AGENT FRAMEWORK Class action scholarship overwhelmingly presents class actions as the classical example of a principal-agent problem. 9 See generally 1 BRENT A. OLSON, PUBLICLY TRADED CORPORATIONS HANDBOOK § 3:15 (2016); Stephen G. Christianson, Liability of a Director to a Corporation for Mismanagement, in 29 AM. JUR. PROOF OF FACTS 3d 133, § 9 (2016). 10 See FED. R. CIV. P. 23(a)(1) (stating that “[o]ne or more members of a class may sue or be sued as representative parties on behalf of all members only if: (1) the class is so numerous that joinder of all members is impracticable”). 11 See FED. R. CIV. P. 23(b)(3) (stating that a “class action may be maintained if Rule 23(a) is satisfied and if . . . the court finds that the questions of law or fact common to class members predominate over any questions affecting only individual members”). PARKINSON – FINAL 746 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 This literature base, which finds its origins in the canonical work of Professor John C. Coffee, Jr., has undeniably influenced the courts. Much of the federal judiciary relies on—often explicitly—the principal-agent framework when applying Rule 23. A. Class Action Literature and the Principal-Agent Framework In 1912, the judiciary adopted Equity Rule 38, which permitted an early form of aggregate litigation.12 This legal device was formalized in 1937 with the first embodiment of what is now Rule 23.13 The original Rule 23 appears more concerned with conserving judicial resources than vindicating litigants’ claims—it divided class actions into three categories, at least two of which would have done little to aid a plaintiff whose claim was too small to warrant individual action.14 The 1966 amendment to Rule 23 and its 12 Rules of Practice for the Courts of Equity of the United States, 226 U.S. 629, 659 (1912) (“When the question is one of common or general interest to many persons constituting a class so numerous as to make it impracticable to bring them all before the court, one or more may sue or defend for the whole.”). 13 FED. R. CIV. P. 23 advisory committee’s note to 1937 adoption (“It applies to all actions, whether formerly denominated legal or equitable.”). 14 As originally formulated, Rule 23 divided class actions into three categories. These categories were sorted depending on the character of the relief sought. First, Rule 23(a)(1), known as the “true” category of class action, was available where the right to be enforced was “joint,” a term without much meaning. One commentary has noted that the prototypical “true” class action at the time was one wherein the plaintiffs sought common injunctive relief. 7A CHARLES ALAN WRIGHT, ARTHUR R. MILLER & MARY KAY KANE, FEDERAL PRACTICE AND PROCEDURE § 1752 (3d ed. 2005). Second, Rule 23(a)(2), known as the “hybrid” category of class action, was available where the “object of the action” is relief implicating “specific property.” Id. One court explained: “If the rights of the individual plaintiffs are separate causes of action and they have no right to a common fund or to common property, the class action at bar is a ‘spurious’ one. If, upon the other hand, the individual plaintiffs having individual causes of action have also a right to a common fund or in common property, the class action may be ‘hybrid.’” Penn. Co. for Ins. on PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 747 focus on small-claim plaintiffs15 was, therefore, heralded as a “drastic” shift.16 This litigant-centric change notwithstanding, courts proceeded without a unified framework, identifying at least three distinct guiding principles: remedial justice;17 judicial efficiency;18 and resource conservation.19 Lives & Granting Annuities v. Deckert, 123 F.2d 979, 983 (3d Cir. 1941). Third, Rule 23(a)(3), known as the “spurious” category of class action, was available where “there is a common question of law or fact affecting the several rights and a common relief is sought.” 7A WRIGHT, MILLER & KANE, supra, § 1752. This last category is the most closely related category to modern conceptions of Rule 23(b)(3). Importantly, it was said that this class action category was “allowed as a matter of efficiency to avoid multiplicity of actions and the joinder of parties in these actions was subject to the discretion of the court.” Id. § 1752 (citing Knowles v. War Damage Corp., 171 F.2d 15 (D.C. Cir. 1948)). Clearly, a unifying concern for the original Rule 23 was judicial economy; neither true nor hybrid actions would aid the plaintiff whose individual claim was too small to pursue individual action. 15 See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (noting that “the Advisory Committee had dominantly in mind vindication of ‘the rights of groups of people who individually would be without effective strength to bring their opponents into court at all’”). 16 FED. R. CIV. P. 23 advisory committee’s note to 1966 amendment; 7A CHARLES ALAN WRIGHT, ARTHUR R. MILLER & MARY KAY KANE, FEDERAL PRACTICE AND PROCEDURE § 1753 (3d ed. 2016). 17 See Deposit Guar. Nat’l Bank v. Roper, 445 U.S. 326, 339 (1980) (“The aggregation of individual claims in the context of a classwide suit is an evolutionary response to the existence of injuries unremedied by the regulatory action of government. Where it is not economically feasible to obtain relief within the traditional framework of a multiplicity of small individual suits for damages, aggrieved persons may be without any effective redress unless they may employ the class-action device.”). 18 See Zahn v. Int’l Paper Co., 414 U.S. 291, 307 (1973) (“Class actions were born of necessity. The alternatives were joinder of the entire class, or redundant litigation of the common issues. The cost to the litigants and the drain on the resources of the judiciary resulting from either alternative would have been intolerable.”). 19 See Califano v. Yamasaki, 442 U.S. 682, 701 (1979) (noting that the “class-action device saves the resources of both the courts and the parties PARKINSON – FINAL 748 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 Professor Coffee broke the fog with a series of seminal pieces20 that collectively articulated the principal-agent framework that has come to define class action scholarship. In his first salvo, Coffee made a simple observation: “The potential for collusion, of course, is present in any class or derivative action, because an inherent conflict of interest exists between the attorney and the class he represents. The latter is interested in the size of the settlement; the former, in the size of his fees.”21 In addition to this collusion problem, Coffee later identified several related principal-agent problems,22 each characterized by the information asymmetry between class counsel and class members, which affords the former the opportunity to engage in “opportunistic behavior,” and renders the latter helpless to monitor such behavior.23 Indeed, as Coffee noted, “it is more accurate as a descriptive matter to view the attorney as an independent entrepreneur than as an agent of the client.”24 by permitting an issue potentially affecting [many parties] to be litigated in an economical fashion under Rule 23”). 20 See generally Coffee, supra note 1; John C. Coffee, Jr., Understanding the Plaintiff’s Attorney: The Implications of Economic Theory for Private Enforcement of Law through Class and Derivative Actions, 86 COLUM. L. REV. 669 (1986) [hereinafter Coffee, Understanding the Plaintiff’s Attorney]; John C. Coffee, Jr., Rethinking the Class Action: A Policy Primer on Reform, 62 IND. L. J. 625 (1987) [hereinafter Coffee, Rethinking the Class Action]; John C. Coffee, Jr., Class Wars: The Dilemma of the Mass Tort Class Action, 95 COLUM. L. REV. 1343 (1995). 21 Coffee, supra note 1, at 243 (“[T]o the extent the attorney regards only his own self-interest, he would prefer a $500,000 settlement out of which a $300,000 award of attorneys’ fees would be paid, to a $1,000,000 recovery out of which only a $200,000 fee would be paid.”). 22 See Coffee, Understanding the Plaintiff’s Attorney, supra note 20, at 679–80 (“[T]his system also creates the potential for . . . opportunism,” which “arises because, in economic terms, there are high ‘agency costs’ associated with class and derivative actions.”). 23 Coffee, Rethinking the Class Action, supra note 20, at 628 (“High agency costs characterize class action litigation, and permit opportunistic behavior by attorneys.”). 24 Id. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 749 Several normative and doctrinal entailments follow. Most obviously, class actions warrant judicial scrutiny. Indeed, the principal-agent problems that infect class actions are so severe as to be distinguished from garden-variety principal- agent problems in the corporate-governance context: [T]he critical decisions in litigation typically have lower visibility and require greater expertise to understand than in the case of the shareholder- manager relationship (where at least publicly reported financial statements and the financial press reveal much and supply a basis for comparison). [Moreover], no public market exists in the case of the attorney-client relationship to motivate the agent to serve the interests of the principal.25 From these beginnings, the principal-agent framework rose in prominence with law-and-economics scholarship. As but one example, Jonathan Macey and Geoffrey Miller illustrated26 how “collective action and free-rider effects allow the plaintiffs’ attorney in class and derivative cases to operate with nearly total freedom from traditional forms of client monitoring.”27 That, in turn, allows class counsel to undervalue class members’ claims in exchange for a quick and favorable settlement.28 Samuel Issacharoff has stated that “class counsel selected on the basis of an economic commitment to maximize financial returns to the class will 25 Id. at 629. 26 See generally Jonathan R. Macey & Geoffrey P. Miller, The Plaintiffs’ Attorney’s Role in Class Action and Derivative Litigation: Economic Analysis and Recommendations for Reform, 58 U. CHI. L. REV. 1 (1991). 27 Id. at 20. 28 See Christopher R. Leslie, De Facto Detrebling: The Rush to Settlement in Antitrust Class Action Litigation, 50 ARIZ. L. REV. 1009, 1010 (2008); see also Alon Klement, Who Should Guard the Guardians? A New Approach for Monitoring Class Action Lawyers, 21 REV. LITIG. 25, 42 (2002); Bruce L. Hay, The Theory of Fee Regulation in Class Action Settlements, 46 AM. U. L. REV. 1429, 1436 (1997). PARKINSON – FINAL 750 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 be especially likely to succumb to the cross-cutting incentives in any principal-agent relationship.”29 In the clearest attempt to tie this (very real) principal- agent concern to class size, some suggest that poor monitoring allows class counsel to move for the certification of oversized classes. This allows class counsel to seek exorbitant fees (generated by a large class), while externalizing the risk of failure—decertification because the class is too large and the claims too dissimilar—on those class members whose claims would have been vindicated in a smaller class.30 These observations have come to shape a well-established and predominant framework in the literature. William Rubenstein remarked that, “[a]lthough commentators could not agree on a cure, they were in accord regarding the diagnosis—the virus infecting the class action system was the principal-agent problem.”31 Class action scholars regularly acknowledge this consensus,32 raising the risk of an echo-chamber effect. For example, Professor Issacharoff observed that: 29 Samuel Issacharoff, The Governance Problem in Aggregate Litigation, 81 FORDHAM L. REV. 3165, 3183 (2013). 30 See Klement, supra note 28, at 35–36 (noting that poorly monitored class counsel have an incentive to oversupply potentially unrelated claims while “underinvesting” resources in meritorious claims). 31 William B. Rubenstein, On What a “Private Attorney General” Is— And Why It Matters, 57 VAND. L. REV. 2129, 2164 (2004). 32 See, e.g., id.; Myriam Gilles & Gary B. Friedman, Exploding the Class Action Agency Costs Myth: The Social Utility of Entrepreneurial Lawyers, 155 U. PA. L. REV. 103, 113 (2006) (“Meanwhile, back in legal academia, the dominant story was agency costs. Beginning with John Coffee’s pioneering work in the mid-1980s, law and economics scholars began to critically examine the powerful financial incentives of entrepreneurial class actions lawyers.”); Samuel Issacharoff, Governance and Legitimacy in the Law of Class Actions, 1999 SUP. CT. REV. 337, 375. See generally Morris A. Ratner, A New Model of Plaintiffs’ Class Action Attorneys, 31 REV. LITIG. 757 (2012) (casting Professor Coffee’s principal- agent approach as the “conventional understanding” in class action literature). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 751 [i]ronically, the very insight that opened the class action to a more sophisticated scholarly account of agency costs may also now serve as a set of blinders on the range of agency problems in representation. The approach to agency cost suggested by Coffee drew (and still draws) exclusively from the securities and corporate governance literature, inviting a comparison of the extent to which the governance mechanisms chosen in the private domain could help overcome principal-agent problems.33 Alexandra Lahav remarked that “[o]ne might ask why we should care about class action governance at all. Scholars viewing class actions from an economic perspective generally view the central problem in class actions to be an agent- principal problem, and have a correspondingly narrow view of the governance regime required to resolve it.”34 B. Impact on the Courts Courts regularly embrace the principal-agent framework, sometimes explicitly so. This spillover from scholarship to the courthouse manifests in two tracts of doctrine: first, courts aggressively regulate the internal governance of class actions, principally to resolve “compensation problems”; second, courts aggressively police the external boundaries of class actions, ostensibly to address “monitoring problems.” 1. Regulating Internal Governance: Compensation Problems Rule 23 devotes significant attention to compensation. Rule 23(g) provides that courts “may order potential class counsel to provide information on any subject pertinent to the appointment and to propose terms for attorney’s fees and nontaxable costs.”35 Rule 23(h), which explicitly concerns 33 Issacharoff, supra note 29, at 3185. 34 Alexandra Lahav, Fundamental Principles for Class Action Governance, 37 IND. L. REV. 65, 72 (2003). 35 FED. R. CIV. P. 23(g)(1)(C). PARKINSON – FINAL 752 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 fees, provides courts with several mechanisms for evaluating attorney compensation, from mandatory notice to objectors36 to review hearings.37 These provisions are intended to address the particular principal-agent dynamic that attends to determining the agent’s compensation.38 The Advisory Committee was careful to note that “[a]ctive judicial involvement in measuring fee awards is singularly important to the proper operation of the class-action process,”39 and it assigned judges considerable discretion in executing that role.40 Courts have not shied from the moment, applying considerable scrutiny to attorneys’ fees,41 and at times explicitly clarifying that the principal-agent framework motivates that scrutiny.42 For example, the Court of Appeals for the Third Circuit has remarked that “the principal-agent problem endemic to class actions” can “[create] a situation 36 See Edward Brunet, Class Action Objectors: Extortionist Free Riders or Fairness Guarantors, 2003 U. CHI. LEGAL F. 403, 408–09 (“By definition, the objector is a monitor [and class member], who is evaluating a proposed settlement and then investing resources to either improve the settlement terms or reject the settlement. . . . Objectors create an adversary contest, usually regarding the difficult process of settlement approval.”). 37 FED. R. CIV. P. 23(h)(3). 38 FED. R. CIV. P. 23(h) advisory committee’s note to 2003 adoption (“Fee awards are a powerful influence on the way attorneys initiate, develop, and conclude class actions.”). 39 Id. 40 See id. (stating that the “rule does not attempt to resolve the question whether the lodestar or percentage approach should be viewed as preferable”). 41 See generally Vaughn R. Walker & Ben Horwich, The Ethical Imperative of a Lodestar Cross-Check: Judicial Misgivings About “Reasonable Percentage” Fees in Common Fund Cases, 18 GEO. J. LEGAL ETHICS 1453 (2005). 42 See, e.g., Dechert v. Cadle Co., 333 F.3d 801, 803 (7th Cir. 2003) (Posner, J.) (“In the usual class-action case, in which the class representative’s stake is so small that as a practical matter the lawyer for the class completely controls the litigation, there is a danger remarked in numerous cases that the lawyer will negotiate a settlement with the defendant that gives the lawyer a large fee but the class a meager recovery.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 753 where the defendants and plaintiffs [counsel] can collusively settle litigation in a manner that is adverse to the class’s interests.”43 Several courts have even explicitly tied similar misgivings to the work of Professor Coffee and others.44 In practice, courts address the compensation problem by implementing fee structures that aim to mitigate agency costs. For example, courts have largely rejected the “pure” lodestar approach to calculating compensation—a method by which the court calculates the hours “reasonably expended” by class counsel, multiplies those hours by a “reasonable hourly fee,” and then has the discretion to increase that amount by a “reasonable” enhancement multiple.45 Prior to the rise of the principal-agent framework, courts often approved this fee structure and generally treated an enhancement multiple of four or five as reasonable.46 Over time, however, the pure lodestar fell into disfavor, in part because of principal-agent concerns; one Judicial Task Force assigned to evaluate the model found that it provided poorly monitored class counsel with an incentive to “pad” expenses.47 43 In re General Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d 768, 802 (3d Cir. 1995) (citing John C. Coffee, Jr., Understanding the Plaintiff’s Attorney: The Implications of Economic Theory for Private Enforcement of Law through Class and Derivative Actions, 86 COLUM. L. REV. 669 (1986)). 44 See, e.g., id.; Polar Int’l Brokerage Corp. v. Reeve, 187 F.R.D. 108, 118–19 (S.D.N.Y. 1999) (“Many courts and commentators have described the danger of a conflict of interest between plaintiffs’ counsel in a class action suit and their clients—the class members they represent—namely, the temptation of class counsel to sacrifice procuring value for the class in exchange for maximizing attorney’s fees.”). 45 2 JOSEPH M. MCLAUGHLIN, MCLAUGHLIN ON CLASS ACTIONS: LAW AND PRACTICE § 6:24 (13th ed. 2016). 46 See Gilles & Friedman, supra note 32, at 140 n.157; Walker & Horwich, supra note 41, at 1472 (“In our informal review, the multipliers ranged from about 1.0 to over 5.0, with a substantial number of multipliers in the 3.0 to 4.0 range.”). 47 See Court Awarded Attorney Fees: Report of the Third Circuit Task Force, 108 F.R.D. 237 (1985); see also Goldberger v. Integrated Res., Inc., 209 F.3d 43, 48 (2d Cir. 2000) (“[D]istrict courts [have] found that it PARKINSON – FINAL 754 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 Courts now typically permit one of two related fee structures: First, the percent-of-the-fund model, which awards class counsel a set share of the common fund obtained in settlement or following a judgment.48 While courts apply a litany of factors to evaluate whether a percentage is reasonable,49 awards often fall between 20 and 30 percent of the fund.50 Second, the lodestar-crosscheck, in which the court “can use the lodestar method to confirm that a percentage of recovery amount does not award counsel an exorbitant hourly rate.”51 This second approach has “been almost universally adopted by [the] courts.”52 The two approaches are, of course, not mutually exclusive, the latter supplementing the former. Moreover, in both regimes, class counsel is given an award and is typically compensated “for the costs and reasonable out-of-pocket expenses incurred in prosecuting the litigation,” including many discovery-related expenses.53 created a temptation for lawyers to run up the number of hours for which they could be paid.”); In re Agent Orange Prod. Liab. Litig., 611 F. Supp. 1296, 1306 (E.D.N.Y. 1985) (stating that the lodestar method “tends to encourage excess discovery, delays and late settlements, while it discourages rapid, efficient and cheaper resolution of litigation”). 48 See, e.g., Camden I Condo. Ass’n, Inc. v. Dunkle, 946 F.2d 768, 774 (11th Cir. 1991) (holding that “in this circuit, attorneys’ fees awarded from a common fund shall be based upon a reasonable percentage of the fund established for the benefit of the class”); Swedish Hosp. Corp. v. Shalala, 1 F.3d 1261, 1272 (D.C. Cir. 1993) (“[W]e conclude that percentage-of-the- fund is the proper method for calculating fees in a common fund case.”); Brown v. Phillips Petroleum Co., 838 F.2d 451, 454 (10th Cir. 1988) (citing Blum v. Stenson, 465 U.S. 886, 900 (1984)). 49 See 2 MCLAUGHLIN, supra note 45, § 6:24. 50 See id. (“Regardless of whether 25% is adopted as a formal benchmark, most courts have approved fee applications approximating that figure, but awards in the 20 to 30% range are not uncommon.”). 51 In re General Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d 768, 821 n.40 (3d Cir. 1995). 52 See Gilles & Friedman, supra note 32, at 139. 53 2 MCLAUGHLIN, supra note 45, § 6:24. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 755 2. Regulating External Boundaries: Monitoring Problems Courts concerned with principal-agent problems also oversee the degree to which class members are capable of monitoring class counsel. Lacking a tool as obviously keyed to this task as Rule 23(g) and Rule 23(h) are keyed to regulating compensation, courts often resort to a heavy- handed application of the Rule 23 certification requirements. For example, Rule 23(a)(2) requires that “there are questions of law or fact common to the class”;54 Rule 23(b)(3), which goes a step further, requires that for class actions principally seeking financial remuneration—so-called “damages classes”—“the questions of law or fact common to class members predominate over any questions affecting only individual members.”55 These requirements, when strictly applied, allow courts to police the size and scope of class actions (i.e., external boundaries). The judiciary has seized the opportunity, at times explicitly connecting class size to the monitoring problem endemic to principal-agent relationships.56 For example, one court remarked that agency costs “often can be far more severe in the class action context, primarily because classes tend to be large, dispersed and disorganized and therefore suffer from a collective action dilemma not faced by individual litigants,” which contributes to “significantly less monitoring of the attorney by the class and consequential [sic] higher agency costs.”57 This trend has been influenced, in part, by a slate of seminal opinions in which the Supreme Court restricted the size and scope of expansive class actions via these 54 FED. R. CIV. P. 23(a)(2). 55 FED. R. CIV. P. 23(b)(3). 56 See, e.g., Culver v. City of Milwaukee, 277 F.3d 908, 910 (7th Cir. 2002) (Posner, J.) (noting that, in light of principal-agent problems, the “class action is an awkward device, requiring careful judicial supervision”). 57 In re Auction Houses Antitrust Litig., 197 F.R.D. 71, 78 (S.D.N.Y. 2000). PARKINSON – FINAL 756 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 requirements. In Amchem Products, Inc. v. Windsor,58 a case that concerned the certification of a settlement class comprised of current and future asbestos-related claimants,59 the Court declined to certify the class “[g]iven the greater number of questions peculiar to the several categories of class members, and to individuals within each category, and the significance of those uncommon questions,” which collectively ensured that “any overarching dispute about the health consequences of asbestos exposure cannot satisfy the Rule 23(b)(3) predominance standard.”60 The Court articulated its concern with reference to “class cohesion,” a principle derived from Rule 23(b)(3).61 This cohesion concern appears to draw, at least in part, from the principal-agent framework,62 melding the judiciary’s historic concern with the compensation problem—a classic principal-agent problem—with a newfound concern for class size: The adequacy-of-representation requirement “tend[s] to merge” with the commonality and typicality criteria of Rule 23(a), which “serve as guideposts for determining whether . . . maintenance of a class action is economical and whether the named plaintiff’s claim and the class claims are so interrelated that the interests of the class members will be fairly and adequately protected in their absence.”63 58 521 U.S. 591 (1997). 59 See id. at 597. For more information on settlement classes, which involve the simultaneous certification and settlement of a class action in order to bind the class as a whole, see 2 MCLAUGHLIN, supra note 45, § 6:3. See generally Samuel Issacharoff & Richard A. Nagareda, Class Settlements under Attack, 156 U. PA. L. REV. 1649 (2008). 60 Amchem, 521 U.S. at 624. 61 Id. at 623. 62 See John C. Coffee, Jr., Class Action Accountability: Reconciling Exit, Voice, and Loyalty in Representative Litigation, 100 COLUM. L. REV. 370, 375 (2000). 63 Amchem, 521 U.S. at 626 n.20 (quoting Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 157, n.13 (1982)). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 757 Following Amchem, courts have regularly restricted the size and scope of class actions, in part to combat monitoring problems—the larger the class, the greater class counsel’s autonomy, creating a feedback loop in which the monitoring problem fuels the compensation problem.64 As one scholar observed, “[b]eginning with the hugely influential writings of Professor Coffee twenty-five years ago, and continuing through the Court’s decision in Amchem, many of the governance mechanisms are directed to the prospect of agents acting in self-regarding means.”65 More recent Supreme Court cases also augmented judicial scrutiny of class actions’ external boundaries.66 In Wal-Mart Stores, Inc. v. Dukes,67 the Supreme Court reversed an order certifying a massive class encompassing millions of plaintiffs because “Rule 23(a) ensures that the named plaintiffs are appropriate representatives of the class whose claims they wish to litigate” by “effectively limit[ing] the class claims to those fairly encompassed by the named plaintiff’s claims.”68 In Comcast Corp. v. Behrend,69 the Court affirmed Amchem’s core holding—that “Rule 23(b)(3)’s predominance criterion is even more demanding than Rule 23(a)”70—and announced a 64 The Court later confirmed this view in Ortiz v. Fibreboard Corp., where it stated the more lenient the application of the Rule 23 certification requirements, “the greater the likelihood of abuse.” 527 U.S. 815, 817 (1999). 65 Issacharoff, supra note 29, at 3183. 66 See Issacharoff & Nagareda, supra note 59, at 1681 (“After Amchem, the focus on potential intraclass conflicts metastasized. Every difference in the potential interests of class members was seemingly fair game for challenge. Whether on direct or collateral review, challenges to the adequacy of class representation in the intraclass conflict sense quickly came to center on the contention that the class was, in one way or another, too encompassing in its scope, such that subclasses were needed with separate class representatives and, even more importantly, separate class counsel.” (emphasis added)). 67 564 U.S. 338 (2011). 68 Id. at 349 (quoting Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 156 (1982)). 69 133 S. Ct. 1426 (2013). 70 Id. at 1432. PARKINSON – FINAL 758 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 scrutinizing application of that rule: in order for common questions to predominate, the class must establish that individual “damages are capable of measurement on a classwide basis.”71 Both Wal-Mart and Comcast contain language sounding in and drawing on Amchem, echoing its monitoring-related concerns.72 * * * Courts concerned with principal-agent problems do not limit themselves to addressing compensation problems, i.e., the traditional domain of the principal-agent framework. Courts also articulate a size-based concern for monitoring problems. As a result, courts frequently apply the Rule 23 certification requirements with a heavy hand,73 aggressively regulating not only the internal governance of class actions, but also their external boundaries. Of course, class actions do present genuine principal-agent problems—the compensation problem is no apparition. However, that well-founded concern risks mission creep. The status quo’s deception is that where there are class actions, there are axiomatically principal- agent problems—and where there are principal-agent 71 Id. at 1433. 72 Id. at 1434–35 (“For all we know, [plaintiff] cable subscribers in Gloucester County may have been overcharged because of petitioners’ alleged elimination of satellite competition . . . while subscribers in Camden County may have paid elevated prices because of petitioners’ increased bargaining power vis-à-vis content providers . . . . The permutations involving four theories of liability and 2 million [plaintiff] subscribers located in 16 counties are nearly endless.”). 73 See e.g., Ex parte Assocs. Fin. Servs. Co., 705 So.2d 836, 838–39 (Ala. 1997) (Hooper, J., dissenting) (“The class action is a unique action at law. Unlike a normal plaintiff, the plaintiff members of the class need not even appear at court. This aspect of the class action was enough to cause at its inception a great deal of controversy. It is critical that courts, in particular the highest court of this State, make sure that the rules surrounding the use of the class action be strictly enforced. The potential for abuse of procedure is accentuated in unique types of actions like the class action.”). See generally Robert H. Klonoff, The Decline of Class Actions, 90 WASH. U. L. REV. 729 (2013). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 759 problems, judicial intervention is warranted. That is half right. C. Refracting the Frame One Degree The principal-agent framework, which finds its zenith in the contemporary application of Rule 23, rests on a simple observation: class actions separate ownership interests (property rights in litigation)74 from the management of those interests.75 This observation does not exclusively buttress the principal-agent framework, and it certainly does not lead to the conclusion that all class action issues present principal- agent problems. The separation of ownership and management does not dictate when principal-agent problems arise, the magnitude of their attendant agency costs, or the manner in which those costs should be mitigated. Instead, the principal-agent framework is better viewed as one of several policy options. That is, while principal-agent problems are likely to arise and present significant agency costs in particular contexts,76 whether and how those costs should be regulated does not speak to how the class action mechanism as a whole should be theorized. 74 See Logan v. Zimmerman Brush Co., 455 U.S. 422, 428 (1982) (stating that it is “settled” that “a cause of action is a species of property”); see also Jeremy A. Blumenthal, Legal Claims as Private Property: Implications for Eminent Domain, 36 HASTINGS CONST. L. Q. 373, 373 (2009) (“A lawsuit is property. A plaintiff has a private property right in his claim of action—i.e., in the right to sue—and in his lawsuit once filed.”). 75 See Lee-ford Tritt, The Limitations of an Economic Agency Cost Theory of Trust Law, 32 CARDOZO L. REV. 2579, 2593 (2011) (“Accordingly, agency-cost theory attempts to both identify the costs which are inherent in dividing ownership and control of a corporation, and to minimize them through more efficient incentive schemes.”); Richard M. Hynes, Securitization, Agency Costs, and the Subprime Crisis, 4 VA. L. & BUS. REV. 231, 274 (2009) (“Agency costs are a result of the separation of ownership from control.”) See generally Michael J. Meurer, Law, Economics, and the Theory of the Firm, 52 BUFF. L. REV. 727 (2004) (noting that the principal- agent framework rests on a separation of ownership and management). 76 See infra Section IV.A.1. PARKINSON – FINAL 760 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 Comparing the principal-agent framework to an alternative illustrates why both are better suited as two among several policy options. Contra the status quo’s focus on “whether current practice optimally aligns incentives to serve the ends of compensating absent class members,” some scholars and courts posit that the “appropriate lens [is] deterrence.”77 For this group, the appropriate framework by which to analyze class actions is not the principal-agent relationship, but instead the optimization of social costs— Rule 23 should be applied in order to “provide optimal deterrence” of bad actors.78 Although it rests on solid ground and raises valid arguments, this “framework” falls prey to the same critique as its declared principal-agent adversary.79 Specifically, both rely on accurate observations—class actions can give rise to principal-agent problems, just as surely as they can augment deterrence—but neither illuminates how class actions should be systematically analyzed across cases and issues. Agency costs and deterrence are often both at play on opposite sides of the ledger. It is not obvious how a court presented with both principal-agent and deterrence arguments should proceed on the limitless questions that class actions and Rule 23 pose. These at-time warring approaches80 are policy options to be 77 Gilles & Friedman, supra note 32, at 105 (emphasis added). 78 Id. at 106–07. 79 See id. at 107 (arguing that “any goal of class member compensation must be utterly disregarded in favor of a separate and often competing objective: forcing companies to internalize the social costs of undesirable behavior”). 80 To crudely generalize, the principal-agent framework prefers smaller class actions, while the deterrence-centric framework prefers larger class actions, though the latter certainly recognizes overdeterrence as a risk. See id. at 155 (“If we measure class action rules and proposals by whether they increase deterrent value, do we not run the risk of overdeterring efficient behavior? After all, to deter optimally is not to deter maximally.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 761 balanced against one another. They do not guide judicial practice in all instances.81 III. CLASS ACTIONS AS FIRMS Rather than view the class action mechanism as little more than a cesspool of principal-agent problems, class actions should be evaluated and analyzed as firms. Class actions parallel firms in a number of ways, making the two apt for comparison. So viewed, the potentially deleterious effects of the status quo’s preferred framework are brought into focus. The principal-agent framework roughly maps onto the “managerial” or agency-cost theory of the firm, which narrowly focuses on regulation in nearly all instances.82 By contrast, the firm framework maps onto the Coasean or transaction-cost theory of the firm. It is worth noting that these theories of the firm are not mutually exclusive; likewise, the firm framework is not at war with the principal-agent framework. Rather, the former subsumes the most useful parts of the latter. A. Parallels: Class Actions and Firms Class actions and firms share at least three features: inception, purpose, and structure. It is noteworthy that the principal-agent framework recognizes these parallels as hallmarks of the class action mechanism. As such, this is not a deconstructive project; rather, it builds with an eye toward constructing a more complete theory. 81 See generally William H.J. Hubbard, Optimal Class Size, Dukes, and the Funny Thing about Shady Grove, 62 DEPAUL L. REV. 693 (2013). 82 See, e.g., William W. Bratton, Jr., The New Economic Theory of the Firm: Critical Perspectives from History, 41 STAN. L. REV. 1471, 1476 (1989). PARKINSON – FINAL 762 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 1. Inception: Separation of Ownership and Management Class counsel typically initiates a class action: more often than not, it is incumbent upon counsel to identify the alleged wrongdoing, locate named plaintiffs, determine the scope of the class, and—as is the case in individual litigation—pursue the action.83 It can be extremely costly for class counsel to wait for a potential lead plaintiff; for example, in the securities context “courts most commonly select lead counsel by appointing the lawyer who files the first complaint.”84 And, of course, class counsel direct which arguments are leveled on behalf of certification,85 frequently deciding which claims are brought in the first instance.86 Class counsel may opt to exclude certain claims for a variety of reasons, including “to avoid (or take advantage of) venue or jurisdictional limitations, to prevent removal to federal court of a state-court class action, to evade evidentiary issues that could be harmful to existing claims, or simply to drop claims as to which the plaintiffs had little hope of success.”87 Class members—including named representatives—typically do not participate in strategy or settlement discussions, leaving 83 See Jill E. Fisch, Aggregation, Auctions, and Other Developments in the Selection of Lead Counsel under the PSLRA, 64 L. & CONTEMP. PROBS. 53, 56 (2001) (noting that it is incumbent upon class counsel to “investigate potential causes of action, mobilize the plaintiff class, and bear the costs and risks associated with the suit”). 84 Id. at 56 (citing Elliott J. Weiss & John S. Beckerman, Let the Money Do the Monitoring: How Institutional Investors Can Reduce Agency Costs in Securities Class Actions, 104 YALE L. J. 2053, 2062 (1995)). 85 See Issacharoff, supra note 32, at 346 (noting that even in ostensibly “similar[]” asbestos-related class actions, counsel advanced “different bases for the requested class certification”). 86 Edward F. Sherman, “Abandoned Claims” in Class Actions: Implications for Preclusion and Adequacy of Counsel, 79 GEO. WASH. L. REV. 483, 483–84 (2011). 87 Id. at 483. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 763 the decision-making at key nexus points to class counsel alone.88 Class actions are unique, therefore, in their separation of ownership and management.89 Moreover, the ownership interest that a plaintiff holds in a typical claim is even more pronounced in the class context because class members typically possess a property interest in any common fund that results from judgment or settlement.90 And the division that attorneys impose between that ownership interest and management is even more pronounced in the class action context—an observation that the principal-agent literature emphasizes.91 These conditions give rise to the firm,92 as well as its accompanying efficiency-related benefits and agency- related costs.93 88 See Thomas E. Willging et al., An Empirical Analysis of Rule 23 to Address the Rulemaking Challenges, 71 N.Y.U L. REV. 74, 100 (1996). 89 See generally Hynes, supra note 75. 90 See Klier v. Elf Atochem N.A., Inc., 658 F.3d 468, 474 (5th Cir. 2011); 2 MCLAUGHLIN, supra note 45, § 8:15 (“It bears emphasis that class members have a property interest in settlement funds, including unclaimed funds.”); see also Ortiz v. Fibreboard Corp., 527 U.S. 815, 846– 47 (1999). 91 See Issacharoff, supra note 29, at 3184 (noting that “representation always introduces a distance between the interests of the principals and the decision making of the agents”); Lahav, supra note 34, at 93 (observing that “class members’ ownership of their claims is separated from control over those claims”). 92 See OLSON, supra note 9, § 2:2 (concerning “General control structure—Separation of ownership and control—Antecedents to modern corporate governance”). While this Article will proceed with “firm” as the dominant terminology, it has been observed that this word is of a somewhat nebulous character. See, e.g., George S. Geis, The Space Between Markets and Hierarchies, 95 VA. L. REV. 99, 106 (2009) (noting that one can simply “envision any firm as a basic collection of inputs and outputs”). The firm envisioned here resembles one of a classical, corporate nature. 93 See R. H. COASE, The Nature of the Firm, in THE FIRM, THE MARKET, AND THE LAW 33, 54 (1988) (noting that “it is the fact of direction which is the essence of the legal concept of ‘employer and employee’”). See generally Stephen G. Marks, The Separation of Ownership and Control, in ENCYCLOPEDIA OF LAW AND ECONOMICS 692 (1999). PARKINSON – FINAL 764 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 2. Purpose: Profit Firms organize to return a profit above what might be attained by contracting through market transactions. Class actions are formed with a similar purpose in mind. Consider the interests of the two stakeholders in a class action: class counsel (as the manager-agent) and class members (as the owner-principals). Class counsel, often described as a class entrepreneur,94 realizes significant efficiencies by aggregating claims. Pursuing a claim on behalf of a client entails an opportunity cost. While attorneys are typically compensated for costs incurred in litigation, they trade in a finite resource: their time and the opportunity cost of the out-of-pocket expenses that they must expend in pursuit of a favorable settlement or judgment. Were counsel to pursue exclusively individual claims—never aggregating—there would be a limit, or cap, on the number of claims an attorney could represent in his or her lifetime.95 Moreover, accounting for the dynamic costs of pursuing a mass of individual claims, rather than an individual mass claim, the impact is even more pronounced: Courts flooded with individual claims and constrained by their own limited bandwidth would further cap the number of claims that an attorney could hope to litigate. Furthermore, class counsel can realize declining marginal costs by aggregating claims. This extends beyond the traditional observation that adding a marginal class member is a virtually costless endeavor.96 Momentarily suspending 94 See generally, e.g., Gilles & Friedman, supra note 32. 95 Of course, this is true regardless of the size or amount of aggregation; opportunity-cost considerations are always present. However, those considerations militate in favor of aggregation, where possible and appropriate. 96 See, e.g., David Betson & Jay Tidmarsh, Optimal Class Size, Opt- Out Rights, and “Indivisible” Remedies, 79 GEO. WASH. L. REV. 542, 552 (2011) (assuming “that the marginal cost declines continuously from the first class member, who has the highest expected gross benefit, to the last class member n, who has the lowest”); Hubbard, supra note 81, at 697 PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 765 the costs that courts impose at the certification stage—a topic to which this Article will return at length97—there are certain costs inherent to litigating a class action. Some of these costs are spread evenly across each class member—for example, the cost of issuing notice.98 Others, however, are variable and spread unevenly across marginal class members. Consider, for example, a class of geographically dispersed plaintiffs that seek varying levels of relief for damages relating to a negligent toxin release in a metropolitan center. Class counsel must demonstrate not only that each type of injury in the class—each “claim species”—is attributable to the toxin’s release, but also that each individual in the class was actually harmed. Assuming that the class is comprised of many similar claims that are not precisely uniform— owing to subtle differences in damages, liability, and theories of causation as one moves further from the center of the toxin release—class counsel must marshal sufficient evidence to vindicate each claim species, even if the class action mechanism relieves them somewhat of the burden of marshaling evidence for each individual’s claim. At some undefined point, the marginal class member will introduce a new species of claim. (“Each individual added to the class thereby reduces the number of times the fixed cost is expended in individual litigation.”). 97 See infra Part V. 98 See Hubbard, supra note 81, at 696. PARKINSON – FINAL 766 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 For example, as represented in Figure 1 above, plaintiffs closest to the epicenter of the toxin’s release, representing “Claim Species 1,” perhaps came into direct contact with the toxin, resulting in fatal harm. Plaintiffs located further from the toxin release, representing “Claim Species 2,” may also have been harmed, but in a slightly different manner— perhaps wind moved the airborne toxin to the outskirts of the city, where it was inhaled, causing medium-term harm. Finally, “Claim Species 3” encompasses individuals who were harmed by the toxin but were also furthest from its release— perhaps the toxin seeped into the groundwater, triggering long-term health effects for those many miles from the epicenter. Vindicating each claim species is costly and demands substantive, decisional, and administrative resources.99 However, with the inclusion of a marginal claim 99 See id. at 703 (“As the size of the class grows, the likelihood that one has included individuals who do not properly belong to the class increases.”); Betson & Tidmarsh, supra note 96, at 557 (“It is possible, although not inevitable, that the marginal cost of adding some class members—determined by combining the marginal common costs attributable to that class member plus the marginal individual cost FIGURE 1: TOXIN-RELEASE "CLAIM SPECIES" PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 767 that represents a new claim species, i.e., adding a plaintiff with Claim Species 2 to a class theretofore comprised solely of claimants representing Claim Species 1, adding class members of the same claim species becomes virtually costless. This demonstrates a broader phenomenon: once a class reaches a certain size, the probability that the marginal class member introduces a dissimilar claim dwindles. Imagine a chimerical global class action that sought to vindicate every toxin-related claim, regardless of the alleged tortfeasor, toxin involved, theory of liability, harm caused, and so forth—the odds that an additional class member would add a new toxin- related claim are infinitesimally small, if not zero. The marginal cost of litigating class certification, therefore, initially swells, as class counsel soaks up every similar-but- unique claim species, and then recedes as additional class members are aggregated and the risk of uniqueness dissipates. Next, consider the interests of putative class members— the owner-principals. While class members typically do not themselves opt to aggregate100—a second topic to which this Article will return101—there is ample reason to believe that some might choose to do so if the transaction costs of aggregating (i.e., finding, communicating, and negotiating with one another) were sufficiently small. For those class members with comparatively small claims, the fixed costs of litigation likely preclude any alternative to aggregation.102 associated with that class member—exceeds the marginal benefit that the class obtains from adding that member. For instance, it might cost $500 to identify a consumer entitled to a $100 remedy.”). 100 See Hay, supra note 28, at 1437 (“First, there is no contractual relation between principal and agent. The class typically does not ‘hire’ class counsel. As a result, class members cannot contractually protect themselves. They cannot structure the counsel's fee in a way that ensures she will act in their interests.”). 101 See infra Part V. 102 See Carnegie v. Household Int’l, Inc., 376 F.3d 656, 661 (7th Cir. 2004) (noting that in light of expected payoffs and litigation costs, “[t]he PARKINSON – FINAL 768 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 Even for those individuals with comparatively large claims, aggregation presents efficiencies. For example, behavioral- science studies suggest that jurors are more likely to award damages when presented with multiple plaintiffs.103 There are also personal costs to pursuing a claim; one should not discredit the reputational costs, discovery burden, and time- consuming process of attending to the litigation that one avoids as an anonymous class member.104 Finally, individual litigants face considerable dynamic costs absent aggregation. Were many claimants to pursue individual litigation, there would be a costly race to the courthouse among those with similar claims. This might be driven by a desire to avoid waiting in line for compensation, particularly given the risk that a defendant that has harmed many individuals will become judgment proof.105 The negative feedback loop and dynamic consequences of this rush to the courthouse are, from the claimants’ perspective, significant. Would-be class members, unwilling to aggregate,106 would collectively drain the judiciary’s limited resources. As a result, the timeliness with which a randomly selected claim is resolved falls. Because of the time-value of realistic alternative to a class action is not 17 million individual suits, but zero individual suits”). 103 See Irwin A. Horowitz & Kenneth S. Bordens, The Consolidation of Plaintiffs: The Effects of Number of Plaintiffs on Jurors’ Liability Decisions, Damage Awards, and Cognitive Processing of Evidence, 85 J. APP. PSYCHOL. 909, 914–17 (2000). 104 See, e.g., Does I thru XXIII v. Advanced Textile Corp., 214 F.3d 1058, 1067–68 (9th Cir. 2000) (recognizing that one benefit of the class action mechanism is that it can provide a degree of anonymity to unnamed class members who reasonably fear, for example, workplace retaliation that would follow individual litigation). 105 This is analogous to the creditors’ bargain in bankruptcy, which posits that courts should interpret and apply bankruptcy priorities to replicate the hypothetical agreements that creditors—that did not, in fact, bargain ex ante—would have reached to avoid “a race to the courthouse” to claim priority on a debtor’s limited assets. See Thomas H. Jackson, Bankruptcy, Non-Bankruptcy Entitlements, and the Creditors’ Bargain, 91 YALE L. J. 857, 862 (1982). 106 See generally Hubbard, supra note 81. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 769 money,107 every would-be class member is in a worse position ex ante. 3. Structure: Tripartite Organization Many firms organize around a tripartite structure by which principals’ interests are balanced against agents’ interests by an intervening monitor, such as a board of directors.108 The monitor is tasked with reducing information asymmetry and ensuring that corporate managers do not act in a self-interested manner contrary to the interests of the owner-principals, such as shareholders.109 Typically these monitors are charged with reviewing corporate governance at a high level, rather than managing the day-to-day operations of the firm.110 This monitoring role carries a fiduciary duty of care.111 As one commentator observed, “[d]irectors have a fiduciary duty to their corporations, such that liability for negligence in permitting mismanagement is considered a betrayal of those who had reposed trust and confidence in him to perform the duties of a director.”112 Class actions parallel this tripartite structure—though typically to limited effect.113 Specifically, class representatives might monitor class counsel, balancing the interests of unnamed class-member principals against the potentially self-interested behavior of class-counsel agents. The potential for an efficacious monitoring device lies in Rule 23, which requires courts to ensure that “the representative parties will fairly and adequately protect the interests of the 107 See Jacob Nussim & Avraham Tabbach, Tax-Loss Mechanisms, 81 U. CHI. L. REV. 1509, 1532 (2014). 108 See Christianson, supra note 9, §3. 109 See Nicola Faith Sharpe, Questioning Authority: The Critical Link Between Board Power and Process, 38 J. CORP. L. 1, 5 (2012) (noting that “boards are assumed to be effective at monitoring managers in order to reduce agency costs and maximize shareholder wealth”). 110 See Marks, supra note 93, at 706. 111 See Smith v. Van Gorkom, 488 A.2d 858, 872–73 (Del. 1985). 112 See Christianson, supra note 9, § 5. 113 See generally Coffee, supra note 62. PARKINSON – FINAL 770 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 class.”114 Indeed, courts regularly impose a fiduciary-like duty on class representatives “to promote and protect the interests of the class he or she purports to represent.”115 However, courts widely differ in how these requirements are enforced. Nearly all courts abstractly state that a class representative must guard class members’ interests116 and do more than simply lend a lead name to the litigation.117 Many go further, strictly applying the Rule 23 adequacy requirement by imposing additional monitoring obligations on representative plaintiffs.118 Setting a high-water mark, some courts even require that class representatives 114 FED. R. CIV. P. 23(a)(4). 115 1 JOSEPH M. MCLAUGHLIN, MCLAUGHLIN ON CLASS ACTIONS: LAW AND PRACTICE § 4:27 (13th ed. 2016). For examples, see London v. Wal- Mart Stores, Inc., 340 F.3d 1246, 1254 (11th Cir. 2003); Sondel v. Nw. Airlines, Inc., 56 F.3d 934, 938–39 (8th Cir. 1995) (stating that “certified representatives and the class counsel assumed certain fiduciary responsibilities to the Class,” and that “the certified representatives may not take any action which will prejudice the Class’s interest, or further their personal interests at the expense of the Class”); Roper v. Consurve, Inc., 578 F.2d 1106, 1110–11 (5th Cir. 1978) (“[B]y the very act of filing a class action, the class representatives assume responsibilities to members of the class.”). 116 See Beck v. Status Game Corp., 1995 WL 422067, at *6 (S.D.N.Y. July 14, 1995) (“[A] class is entitled to ‘an adequate representative, one who will check the otherwise unfettered discretion of counsel in prosecuting the suit.’” (quoting Weisman v. Darneille, 78 F.R.D. 669, 671 (S.D.N.Y. 1978))). 117 See 1 MCLAUGHLIN, supra note 115, § 4:27; see also, e.g., Beck, 1995 WL 422067, at *4 (holding that adequacy requires more from the class representative than “simply lending his name to a suit controlled entirely by the class attorney”). 118 This is not a fanciful proposal unfamiliar to these parties or courts. Rather, the Private Securities Litigation Reform Act of 1995 requires representative plaintiffs to not only monitor, but also select class counsel in securities class actions. See 15 U.S.C. § 77z-1(a)(3)(B)(v) (“Selection of lead counsel: The most adequate plaintiff shall, subject to the approval of the court, select and retain counsel to represent the class.”); see also 5 JAMES WM. MOORE ET AL., MOORE’S FEDERAL PRACTICE § 23.192 (3d ed. 2016) (“The most adequate plaintiff, whom the court must appoint as lead plaintiff, must select and retain counsel to represent the class, subject to the approval of the court.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 771 demonstrate active involvement in the litigation, establish a thorough understanding of the legal theories underlying the claim, and exercise careful supervision of class counsel’s behavior.119 B. Framework: Selecting a Theory of This Firm The choice between frameworks matters: theory informs and guides practice. Professor Coase recognized as much when he lamented the then-dominance of theories of the firm that exclusively sought out problems and demanded regulation in turn: “One important result of this preoccupation with the monopoly problem is that if an economist finds something—a business practice of one sort or other—that he does not understand, he looks for a monopoly explanation.”120 Armed with only a hammer, everything looks like a nail.121 Simply acknowledging the parallels between class actions and firms, a frequent observation, is insufficient—one must adopt a theory of this firm. Because the dominant principal-agent framework itself rests on an unstated likeness between class actions and firms—the separation of ownership from management—it finds its reflection in the very framework that Professor Coase critiqued as myopically focused on regulation: managerialism.122 119 See 1 MCLAUGHLIN, supra note 115, § 4:29. 120 R. H. COASE, Industrial Organization: A Proposal for Research, in THE FIRM, THE MARKET, AND THE LAW, supra note 93, at 57, 67. 121 See id. at 68 (“This is, I think, a common situation. There is some unusual feature—in this case, large discounts. The conclusion is immediately drawn: monopoly. What people do not normally do is inquire whether it may not be the case that the practice in question is a necessary element in bringing about a competitive situation.”). 122 See generally Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. FIN. ECON. 305 (1976). PARKINSON – FINAL 772 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 1. Managerialism The principal-agent framework roughly maps onto the managerial theory of the firm, which centers on “corporate management groups at the large corporation’s strategic center.”123 According to this view of the firm, “management determine[s] the processes of production and distribution,” giving rise to considerable agency costs.124 The normative entailments that follow sound in a regulatory key. Because “managers enjoy too much discretion, especially with respect to decisions regarding investments and the distribution of corporate profits,” firms must be closely monitored and aggressively regulated.125 Like its class action counterpart, managerialism makes little effort to distinguish between potential principal-agent problems and the differing probabilities and magnitudes of those problems’ respective agency costs. Instead, principal-agent problems inherent to the firm give rise to agency costs that demand regulation in nearly all instances. Both the principal-agent and managerial frameworks ultimately advise against “centralizing economic activity within a firm” (or class action).126 Because “[l]arge and complicated corporations harbor plenty of dark corners, and managers have incentives to use this information asymmetry to take advantage of equity owners,” agency costs “can conceivably be reduced if the same assets are divided into many discrete firms” (or, in this context, lawsuits).127 2. Coase In The Nature of the Firm, Professor Coase theorized the firm as a quasi-organic entity that naturally swells and 123 Bratton, supra note 82, at 1476. 124 Id. 125 Manuel A. Utset, Towards a Bargaining Theory of the Firm, 80 CORNELL L. REV. 540, 542 (1995). 126 Geis, supra note 92, at 112. 127 Id. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 773 contracts of its own accord—often without regulatory intervention—depending on the relative costs of organizing within the firm versus contracting in the market.128 Within the firm, “market transactions are eliminated, and in place of the complicated market structure with exchange transactions is substituted the entrepreneur coordinator,” who “plays his part as a single cell in a larger organism.”129 Coase was particularly interested in ascertaining what dictates the optimal size of a firm—the inputs that determine the external boundaries of a firm that faces a make-or-buy decision. For example, a firm that produces widgets—made from bolts and blocks—must decide whether to make or buy bolts and blocks.130 For Coase, whether to make or buy turns on marginal analysis: the firm will expand to the point at which the internal costs (e.g., agency costs) of expanding equal the savings from avoiding market transactions (transaction costs).131 As Coase noted, “no theory which assumes that only one product is in fact produced [by a single firm] can have very great practical significance.”132 Professor Coase identified several natural limits to this growth—factors that weigh in favor of the decision to “buy,” rather than “make.” For example, “as a firm gets larger, there may be decreasing returns to the entrepreneur function” because “a point must be reached where the costs of organizing an extra transaction within the firm are equal to the costs involved in carrying out the transaction in the 128 See COASE, The Nature of the Firm, supra note 93, at 37. 129 Id. at 35, 37. 130 Id. at 52 (“[T]here may be a point where it is less costly to organize the exchange transactions of a new product than to organize further exchange transactions of the old product.”). 131 See id. at 42 (“A firm becomes larger as additional transactions . . . are organized by the entrepreneur, and it becomes smaller as he abandons the organization of such transactions. The question which arises is whether it is possible to study the forces which determine the size of the firm. Why does the entrepreneur not organize one less transaction or one more?”). 132 Id. at 52. PARKINSON – FINAL 774 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 open market.”133 Market competition also plays a role: “[T]he firm has to carry out its task at a lower cost than the cost of carrying out the market transactions it supersedes, because it is always possible to revert to the market” or “some other firm . . . can take over the task if its costs are lower.”134 As previously noted, the optimal external boundary of a firm depends on the marginal costs and benefits of the firm’s efficiencies and agency costs, respectively—“a firm will tend to expand until the costs of organizing an extra transaction within the firm become equal to the costs of carrying out the same transaction by means of an exchange on the open market or the costs of organizing in another firm.”135 Contra the managerial framework, a firm is “not necessarily a hierarchy in which authority determines terms by fiat,” but rather a system based on “a continuous process of negotiation of successive contracts” guided by market forces.136 Professor Coase’s theory of the firm is a superior analytical guide to class action practice than theories that map onto managerialism. To start, the firm framework is a better fit, tracking the roles and interests of class counsel and class members, respectively. Class counsel is much more than a hired manager—in fact, one tenet of the principal- agent framework is that agency costs arise because class counsel is not hired ex ante by the class in any traditional sense.137 Instead, class counsel is an entrepreneurial 133 Id. at 43. 134 COASE, Industrial Organization, supra note 120, at 63. 135 See COASE, The Nature of the Firm, supra note 93, at 44. 136 Bratton, supra note 82, at 1478. 137 See Alexandra D. Lahav, Two Views of the Class Action, 79 FORDHAM L. REV. 1939, 1948 (2011) (“The agent-principal problem is a crucial issue in the class context because neither the class as a whole nor its individual members exercise control over the lawyer. An individual client can threaten to fire the lawyer, but the class cannot. An individual client, particularly the corporate client, may be a repeat player. Class members are decidedly not. Individual clients can negotiate lawyer pay and may withhold pay or negotiate discounts, while class members cannot.”); Hay, supra note 28, at 1437 (observing that “there is no PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 775 manager that aggregates resources (individual claims) within a defined structure (the class).138 Class members are not vested owners in the traditional sense; rather, they function as shareholders in the class action, contributing (though not always by choice ex ante) marginal units of production to the firm (their claims). Moreover, the firm framework corresponds to the interests of both stakeholders. Class members often benefit by aggregating (in a firm), instead of pursuing litigation individually (on the market).139 Similarly, aggregation yields efficiencies for class counsel above and beyond what can be obtained by pursuing individual litigation.140 The similarities do not end at mere profit motive. Entrepreneur-managers aggregate resources in part, Professor Coase explained, to attain greater “coordination” and “control” over factors of production.141 That control, in turn, generates organizational efficiencies. Likewise, entrepreneurial attorneys aggregate individual claims in part to augment their own control, both over how the litigation will proceed142 and how the class action as a firm might be unwound (or liquidated) vel non in settlement.143 contractual relation between principal and agent,” and the “class typically does not ‘hire’ class counsel”); see also infra Part V. 138 See Coffee, Understanding the Plaintiff’s Attorney, supra note 20, at 683–84 (stating that “one better understands the behavior of the plaintiff’s attorney in class and derivative actions if one views him not as an agent, but more as an entrepreneur who regards a litigation as a risky asset that requires continuing investment decisions”); see also text accompanying note 24. 139 See supra text accompanying notes 100–107. 140 See supra text accompanying notes 94–99. 141 See COASE, The Nature of the Firm, supra note 93, at 37. 142 See Macey & Miller, supra note 26, at 3 (noting that, in contrast to typical litigation, “plaintiffs’ class and derivative attorneys function essentially as entrepreneurs who bear a substantial amount of the litigation risk and exercise nearly plenary control over all important decisions in the lawsuit”). 143 See generally id. PARKINSON – FINAL 776 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 Critically, the firm framework provides a way to distinguish between various class action issues, i.e., between principal-agent problems relating to internal governance and make-or-buy decisions relating to external boundaries. There are natural limits to the range of claims that even the most unscrupulous of class counsel will include in a class action. Just as Professor Coase observed that one corporation does not control every factor of production in the economy, no class action has ever attempted—or ever would attempt, even setting aside Rule 23144—to resolve every cause of action in a fanciful global case. There comes a point at which the class, of its own accord (i.e., at the direction of counsel), will cease to expand: when the cost of adding an additional claim (or marginal unit of production) exceeds the cost of pursuing that marginal claim individually (on the market). Professor Coase observed that: a firm will tend to be larger: (a) the less the costs of organizing and the slower these costs rise with an increase in transaction costs organized; (b) the less likely the entrepreneur is to make mistakes and the smaller the increase in mistakes with an increase in the transactions organized.145 The addition of unique and dissimilar claims that are unlike those already safely within the class comes at a steep cost. Class counsel must invest resources not only to demonstrate at the certification stage that the dissimilar marginal claim “belongs” in the class action (what this Article terms “shoehorning costs”), but also to convince the fact-finder at trial (or a defendant in settlement negotiations) that the claim is one of merit and value (“litigation costs”).146 That is, class counsel must marshal evidence and conduct sufficient discovery to vindicate each claim species, devote time and effort to those claims, and avoid distracting the limited decisional capacity of the fact-finder with spurious claims. As 144 See infra Part V. 145 COASE, The Nature of the Firm, supra note 93, at 45. 146 See infra Part V. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 777 Coase noted, “the costs of organizing and the losses through mistakes will increase with an increase in the . . . dissimilarity of the transactions.”147 Rule 23 alone does not dictate the size and scope of class actions; rather, the self-interested, profit-maximizing behavior of class counsel and class members can also help explain why class actions attain their size and scope. These stakeholders not only have an incentive to aggregate—in order to transact at a lower cost than in the market—but they also have an incentive to restrict the external boundaries of their firm’s size and scope at some outer limit. Even accepting that expansive class actions give rise to principal-agent problems, class counsel and class members will often self-regulate, mitigating the potential impact of accompanying agency costs. IV. DOCTRINAL ENTAILMENTS: TIERS OF SCRUTINY The firm framework distinguishes between issues that present principal-agent problems, which rightly demand regulation, and those that concern “make-or-buy” decisions, which are often self-regulating. Namely, courts should apply Rule 23 with a heavy hand when faced with issues that relate to the internal workings of a class action— compensation, notice, and the adequacy of class representation, to name a few. By contrast, courts should apply Rule 23 with a light touch when confronted with issues that relate to the external boundaries of a class action— numerosity, commonality, predominance, and so forth. This approach maximizes judicial efficiencies and capitalizes on judicial competencies. The firm framework’s two-tier approach provides the judiciary with a guide for discerning between principal-agent problems and make-or-buy decisions. While principal-agent problems are likely to emerge within a firm, there are natural limits to the external growth of a firm. In the class 147 COASE, The Nature of the Firm, supra note 93, at 45. PARKINSON – FINAL 778 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 action context, the former demands judicial intervention to correct for market failure, while the latter warrants deference. This is not an approach unfamiliar to courts. In corporate law, for example, courts provide considerable deference to business decisions.148 By contrast, courts have proven very capable at peering within a firm to govern the behavior of corporate agents.149 This two-tier approach to Rule 23 is not scholarly thought-play beyond the judicial ambit; rather, it finds textual support. Rule 23 introduces its core external- boundary provisions—primarily the Rule 23(a) commonality prerequisite and the Rule 23(b)(3) predominance requirement—with language that implies deference to litigants: “A class action may be maintained if Rule 23(a) is satisfied and if [Rule 23(b) is met].”150 The Supreme Court has stated that the “discretion suggested by Rule 23’s ‘may’ is discretion residing in the plaintiff.”151 That is, “[c]ourts do 148 See Cede & Co. v. Technicolor, Inc., 634 A.2d 345, 360 (Del. 1993) (stating that the business judgment rule is intended “to preclude a court from imposing itself unreasonably on the business and affairs of a corporation”); see also Stephen M. Bainbridge, The Business Judgment Rule as Abstention Doctrine, 57 VAND. L. REV. 83, 84–85 (2004) (“Although the separation of ownership and control in modern public corporations inevitably raises important accountability concerns, accountability standing alone is an inadequate normative account of corporate law. A fully specified account of corporate law must incorporate the value of authority—i.e., the need to develop a set of corporate governance rules and standards that enable corporations to adopt efficient decision-making systems and processes.”). 149 See Telxon Corp. v. Meyerson, 802 A.2d 257 (Del. 2002) (concerning a board’s attempt to entice its chairman to serve as CEO by purchasing all the stock of a company that he owned); see also In re The Walt Disney Co. Derivative Litig., 825 A.2d 275, 290–91 (Del. Ch. 2003) (concerning a CEO’s severance package). Note that the business judgment rule applies to corporate directors—it has an ambiguous relationship to corporate managers and officers. See generally Lyman P.Q. Johnson, Corporate Officers and the Business Judgment Rule, 60 BUS. L. 439 (2005). 150 FED. R. CIV. P. 23(b) (emphasis added). 151 Shady Grove Orthopedic Assocs., P.A. v. Allstate Ins. Co., 559 U.S. 393, 400 (2010). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 779 not maintain actions; litigants do.”152 By contrast, Rule 23 introduces its core internal-governance provisions in less deferential terms. For example, following a settlement, the reviewing “court must direct notice in a reasonable manner to all class members who would be bound by the proposal.”153 Similarly, “the court must consider” a number of enumerated factors in evaluating the adequacy of class counsel.154 This Part canvasses several doctrines related to internal governance and external boundaries, respectively, and discusses how those doctrines either are consistent with or require change under the firm framework. FIGURE 2: SORTING CLASS ACTION DOCTRINE Doctrines That “Fit” Doctrines to Change Internal Governance (Heavy Hand) Compensation Adequacy External Boundary (Light Touch) Numerosity Commonality/ Predominance A. Internal Governance: Principal-Agent Issues Courts should apply Rule 23 provisions concerning internal governance with a heavy hand. Principal-agent problems and market failures are particularly likely to thrive in this setting. The lack of ex ante contracting— particularly when coupled with the difficulty that class 152 Id. But see Hubbard, supra note 81, at 708 (construing this language to mean that, following Shady Grove, “a district court has no discretion to refuse to certify a Rule 23(b)(2) class action, so long as the prerequisites in the text of Rule 23(a) are met”). 153 FED. R. CIV. P. 23(e)(1) (emphasis added). 154 FED. R. CIV. P. 23(g)(1) (emphasis added). PARKINSON – FINAL 780 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 members face in monitoring and removing class counsel155— presents tempting opportunities to take advantage of the principal-agent relationship. Absent judicial intervention, the result is welfare sacrificing: class counsel may engage in a litany of inefficient and suboptimal behaviors that the market would ideally correct, were a market mechanism available. For example, an efficient and liquid market for class counsel would reduce the risk that an attorney might take excessive compensation or engage in collusive behavior; knowing that they would simply be terminated and replaced, class counsel’s incentives would align with class members’ interests. Similarly, if the litigation market were highly liquid and claims mobile such that removing one’s stake from a poorly performing class action were simple, there would be little risk of class counsel negotiating a “reverse-auction settlement”;156 class members would simply abandon any class action that undervalued their ownership interest. Of course, these mechanisms do not exist in the class action context, at least not in any efficient capacity. These market failures foster agency costs that warrant judicial intervention. Fortunately, courts are more than capable of overseeing the fact-finding necessary to review a 155 See supra note 137. 156 A reverse-auction settlement refers to a process by which a “defendant can play the plaintiffs’ attorneys off against one another, bargaining down the price of the settlement in exchange for ensuring the lowest selling attorneys that they will be the ones to get a fee out of the case.” WILLIAM B. RUBENSTEIN, NEWBERG ON CLASS ACTIONS § 13:57 (2016). Courts guard against the risk of a reverse-auction settlement with particular care. See, e.g., Reynolds v. Beneficial Nat’l Bank, 288 F.3d 277, 283 (7th Cir. 2002) (“Although there is no proof that the settlement was actually collusive in the reverse-auction sense, the circumstances demanded closer scrutiny than the district judge gave it.”); Bell Atl. Corp. v. Bolger, 2 F.3d 1304, 1310 (3d Cir. 1993) (“One of the risks flowing from shareholders’ difficulty in monitoring derivative litigation is that plaintiffs’ counsel and the defendants will structure a settlement such that the plaintiffs’ attorneys’ fees are disproportionate to any relief obtained for the corporation. Plaintiffs’ attorneys and the defendants may settle in a manner adverse to the interests of the plaintiffs by exchanging a low settlement for high fees.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 781 firm’s internal workings. In the corporate-law context, for example, courts regularly monitor a host of internal- governance matters, including executive compensation,157 internal collusion and deceptive practices,158 and decision- making transparency.159 In each of these broadly defined areas, information asymmetries and related principal-agent problems risk market failure,160 so the judiciary intervenes. 157 See, e.g., In re The Walt Disney Co. Derivative Litig., 825 A.2d 275, 290–91 (Del. Ch. 2003) (holding that plaintiffs sufficiently alleged that the directors’ dealings with the president’s employment with regards to compensation and termination fell outside the business judgment rule); see also Douglas C. Michael, The Corporate Officer’s Independent Duty as a Tonic for the Anemic Law of Executive Compensation, 17 J. CORP. L. 785, 786–87 (1992) (stating that corporate agents have a have a fiduciary “duty not to accept unreasonable compensation”); Randall S. Thomas & Harwell Wells, Executive Compensation in the Courts: Board Capture, Optimal Contracting, and Officers’ Fiduciary Duties, 95 MINN. L. REV. 846, 848–49 (2011) (“Today, courts have a stronger doctrine they can employ when called on to monitor abuses in executive compensation: the fiduciary duties of officers.”). 158 See generally In re Caremark Int’l Inc. Derivative Litig., 698 A.2d 959 (Del. Ch. 1996) (holding that there was not enough evidence of a sustained or systematic failure of the board of directors to exercise oversight to sustain a claim of breach of duty of care and good faith when the corporation failed to come into compliance with government regulation prohibiting collusive relationship between patient care providers.); see also, e.g., McPadden v. Sidhu, 964 A.2d 1262 (Del. Ch. 2008) (holding that the plaintiffs had sufficiently alleged that the board of directors negligently permitted a corporate officer to arrange an auction for one of the corporation’s subsidiaries that excluded a number of bidders and allowed the officer’s ally to purchase the subsidiary well below market value). 159 See, e.g., Gallus v. Ameriprise Fin., Inc., 561 F.3d 816, 823 (8th Cir. 2009) (holding that, particularly as it concerns fee negotiations, there is “a duty to be honest and transparent throughout the negotiation process”). 160 See Edward B. Rock, The Logic and (Uncertain) Significance of Institutional Shareholder Activism, 79 GEO. L. J. 445, 454 & n.29 (1991) (“Disciplining management remains necessary to ensure that the managers manage the corporation effectively and in the interests of shareholders. But discipline is a classic collective good: if it prevents managers from diverting profits from shareholders, all the shareholders benefit in the form of higher earnings and share prices, whether or not they contribute to the PARKINSON – FINAL 782 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 For example, judges fill gaps in contracts and corporate agreements for which exhaustive ex ante contracting might have been unavailable (or posed unduly burdensome transaction costs); courts also evaluate various contract terms, including those related to compensation.161 In the same vein, a vast body of agency law guides courts when determining whether corporate fiduciaries act in good faith vis-à-vis their principals’ interests.162 1. Doctrinal Fit: Compensation Given the dominance of the principal-agent framework,163 it is not difficult to identify a line of doctrine that coheres to the firm framework’s prescription that internal governance be aggressively regulated. For example, the regulation of fee structures “fits.” Courts aggressively review class counsels’ compensation164 for reasons that suggest an implicit adherence to the firm framework. As a baseline, compensation is a matter that concerns the internal workings of a firm. Rule 23 recognizes as much, assigning the court considerable power both to approve class counsel165 and calibrate compensation.166 This breaks with the American system’s general preference for freedom of discipline.”). See generally Bernard S. Black, Shareholder Passivity Reexamined, 89 MICH. L. REV. 520 (1990); Frank H. Easterbrook & Daniel R. Fischel, Voting in Corporate Law, 26 J. L. & ECON. 395 (1983). 161 See RESTATEMENT (SECOND) OF CONTRACTS § 204 (AM. LAW INST. 1981). 162 Generally speaking, a court may hold an agent liable for acting intentionally with a purpose other than the best interests of the principals, violating applicable corporate-governance laws, or intentionally failing to act in the face of a duty to act. See, e.g., In re The Walt Disney Co. Derivative Litig., 825 A.2d at 286; Stone v. Ritter, 911 A.2d 362, 370 (Del. 2006). 163 See supra Section II.B. 164 See supra Section II.B.1. 165 See FED. R. CIV. P. 23(g) (concerning “class counsel”). 166 See FED. R. CIV. P. 23(h) (concerning “attorney’s fees”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 783 contract167 and unencumbered settlement.168 Rule 23 suspends these background principles on the understanding that the internal dynamics of a class action foster a particular principal-agent market failure: the compensation problem. Absent regulation and lacking an obvious means of ex ante contracting, class counsel might negotiate an undeservedly high fee, siphon off funds rightly owed to class members, or “sell out” the value of the class members’ interests in a reverse-auction settlement. One can imagine several market-mimicking mechanisms that might resolve this principal-agent problem—in practice, however, each is unavailing. For example, one might rely on the reputational costs that class counsel incurs by raiding class members’ coffer. Presumably, such behavior would signal to the market—and future clients—that an attorney should be avoided.169 However, absent judicial intervention, this is simply question begging, because it is unclear how the 167 See Elizabeth E. Spainhour, Unsealing Settlements: Recent Efforts to Expose Settlement Agreements That Conceal Public Hazards, 82 N.C. L. REV. 2155, 2165 (2004) (noting that, traditionally, “notions of freedom of contract . . . support arguments in favor of settlement confidentiality” and “freedom of contract is a time-honored tradition that judges do not cursorily set aside”). 168 Rule 23(e), which provides that “claims, issues, or defenses of a certified class may be settled, voluntarily dismissed, or compromised only with the court’s approval,” is a notable exception to the judiciary’s general preference for quick settlement, embodied in the Federal Rules of Civil Procedure, which provides that “[s]ubject to Rules 23(e), 23.1(c), [and] 23.2 . . . the plaintiff may dismiss an action without a court order by filing: (i) a notice of dismissal before the opposing party serves either an answer or a motion for summary judgment; or (ii) a stipulation of dismissal signed by all parties who have appeared.” FED. R. CIV. P. 41; see also 7B WRIGHT, MILLER & KANE, supra note 14, § 1797 (“The importance of the Rule 23(e) requirements is underscored by the fact that a private settlement or compromise for which no approval is sought or notice given is not effective and may be ignored by the court.”). 169 See Hay, supra note 28, at 1438 & n.38 (arguing that “[i]n many settings, class counsel’s financial interests may not diverge substantially from the class members,” because, “[f]or example, perhaps the class counsel wants to develop a professional reputation for effectiveness, which encourages him to obtain large recoveries”). PARKINSON – FINAL 784 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 market will detect the attorney’s misconduct in the first instance. Moreover, this market-based approach presumes that future class members select the attorney with whom they work, and that they do so with something approaching complete information.170 Not only does that rarely take place, but it is in fact far more likely that defendants will ultimately select the class counsel with whom they prefer to deal.171 Market forces unavailing, judicial intervention is necessary. Therefore, courts carefully monitor and review compensation pursuant to Rule 23(g). While courts employ a variety of fee structures172—the optimal blend of which will be addressed in Part V—few courts outright fail to review of class counsels’ compensation. Courts also review class action settlements to ensure that the value that class counsel negotiates for their principals is reasonable and proportionate to the attorneys’ compensation.173 To aid in this process, courts draw on “objectors”—class members who detect improprieties in a proposed settlement and are awarded a bounty for their services—and often hold 170 See Lisa L. Casey, Reforming Securities Class Actions from the Bench: Judging Fiduciaries and Fiduciary Judging, 2003 B.Y.U. L. REV. 1239, 1268 (noting that “in the context of most class actions involving small claims but large stakes, reputational bonding is even more unlikely to affect agency costs” because “class members do not select counsel; rather, counsel initiates the litigation after finding one or more class members to represent the class”). 171 For a related concept, see id. (commenting on “the potential for a multimillion-dollar fee award [to] cause class counsel to risk (or ‘cash in’) her reputation in exchange for profits now”). 172 The vast majority of courts employ either a percentage-of-the-fund or lodestar-crosscheck model. In both regimes, class counsel is generally compensated for reasonable costs incurred during the course of pursuing the action. See supra notes 48–53 and accompanying text. 173 See In re Pet Food Prods. Liab. Litig., 629 F.3d 333, 350 (3d Cir. 2010); In re Warfarin Sodium Antitrust Litig., 391 F.3d 516, 534 (3d Cir. 2004); In re Gen. Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d 768, 785 (3d Cir. 1995). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 785 independent review hearings.174 While Rule 23(h)(3) provides that the reviewing court “may hold a hearing” after receiving post-settlement objections,175 in practice courts regularly apply this provision with a heavy hand, often reviewing primary objectors’ concerns in detail.176 2. Doctrinal Change: Adequacy Despite the ostensibly laissez-faire tenor of the firm framework, there are areas of class action jurisprudence for which it prescribes an even heavier hand than the status quo. For example, Rule 23(g) mandates that courts appoint class counsel that will “fairly and adequately represent the interests of the class.”177 This is an area of doctrine squarely within the firm framework’s tier of heightened scrutiny— 174 See, e.g., Reynolds v. Beneficial Nat’l Bank, 288 F.3d 277, 288 (7th Cir. 2002) (holding that a court can use its equitable powers to award objectors a fee for their services when they “produce an improvement in the settlement worth more than the fee they are seeking”). For information on objectors, see Brunet, supra note 36, at 408–09 (“Informational input from objectors regarding a proposed settlement could, in theory, improve the monitoring problem. By definition, the objector is a monitor, who is evaluating a proposed settlement and then investing resources to either improve the settlement terms or reject the settlement.”). 175 FED. R. CIV. P. 23(h)(3) (emphasis added). 176 See In re Pet Food Prods. Liab. Litig., 629 F.3d at 351 (“When the parties have not supplied the information needed for the court to determine whether the settlement is fair, reasonable, and adequate, the court may affirmatively seek out such information.”); Cotton v. Hinton, 559 F.2d 1326, 1331 (5th Cir. 1977) (“The Court should examine the settlement in light of the objections raised and set forth on the record a reasoned response to the objections including findings of fact and conclusions of law necessary to support the response.”). 177 See FED. R. CIV. P. 23(g)(4); see also FED. R. CIV. P. 23(g)(1) (providing that when appointing class counsel, the court “must consider,” inter alia, “the work counsel has done in identifying or investigating potential claims in the action; counsel’s experience in handling class actions, other complex litigation, and the types of claims asserted in the action; counsel’s knowledge of the applicable law; and the resources that counsel will commit to representing the class”). PARKINSON – FINAL 786 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 whether class counsel fulfills the duties owed to the class is a question of internal governance that falls within courts’ core competencies.178 Inadequacy of counsel plainly presents a principal-agent problem. An attorney’s initial endeavor as the entrepreneur- manager is to organize and constitute the firm; after certification, its secondary goal is to unwind or liquidate the firm on favorable terms, either by settlement or at trial. In the context of adequate representation, these two goals often conflict. Professor Coffee has noted that should one prioritize attaining class counsel that will most ably unwind the class action, that preference will typically undercompensate the initial counsel “who undertook search costs to discover a violation of law.”179 This risks “a reduced incentive for private attorney generals to seek out violations of law.”180 On the other hand, prioritizing a reward for the initial certification may entail sacrifices along the quality-of- counsel dimension at trial or in settlement negotiations.181 Both sides of this tradeoff are costly: one option risks undersupplying socially beneficial class actions;182 the other 178 See supra notes 157–159 and accompanying text; see also In re Gen. Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d at 801 (“Beyond their ethical obligations to their clients, class attorneys, purporting to represent a class, also owe the entire class a fiduciary duty once the class complaint is filed.”); Nick Landsman-Roos, Note, Front-End Fiduciaries: Precertification Duties and Class Conflict, 65 STAN. L. REV. 817, 838–39 (2013) (arguing that, in addition to the traditionally recognized fiduciary duties that class counsel owes to class members post- certification, certain “precertification” duties should attach at the moment the class action is filed). 179 Third Circuit Task Force Report on Selection of Class Counsel, 74 TEMP. L. REV. 689, 733 (2001) (quoting Professor Coffee’s statement to the Third Circuit Task Force). 180 Id. 181 Id. at 734. 182 See id. at 733 (“Even if the court ultimately decides to compensate losing firms for early contributions to the case, the ex ante risk of noncompensation may deter efforts that might be beneficial to the class.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 787 risks ineffective counsel,183 collusive bargaining,184 and reduced recoveries for class members.185 These concerns notwithstanding, courts occasionally approach the adequacy requirement with surprising deference to class counsel,186 rarely articulating what is required to satisfy this requirement, and instead outlining general principles, such as the abstract requirement that “counsel must be qualified, experienced, and generally able to conduct the litigation.”187 Consequently, district courts 183 See Linda S. Mullenix, Taking Adequacy Seriously: The Inadequate Assessment of Adequacy in Litigation and Settlement Classes, 57 VAND. L. REV. 1687, 1702 (2004) (noting that ineffective assistance of counsel often arises undetected in the class action context as a result of “inexperience” and class counsel’s incentive to minimize litigation costs by “underfinancing” the class claim). 184 See Issacharoff & Nagareda, supra note 59, at 1667; Susan P. Koniak & George M. Cohen, Under Cloak of Settlement, 82 VA. L. REV. 1051, 1170 (1996) (“All the critical findings made by a class action court— that the settlement was fair, class counsel adequate, and collusion absent—may be a product of class counsel’s negligence or fraud, either or both accepted without objection by the all-too-congenial defendant.”). 185 See Jay Tidmarsh, Rethinking Adequacy of Representation, 87 TEX. L. REV. 1137, 1171–72 (2009) (noting that because the “combination of fronted costs and expected attorneys fees typically makes class counsel the largest stakeholder in the class action” by several orders of magnitude, “the risk tolerance of the class counsel is often different from that of the class representative and class members” and, consequently, class counsel may act “to protect that investment in the case rather than undertaking riskier strategies that match up better with the risk positions of the class”). 186 See Mullenix, supra note 183, at 1702 (“[T]he cursory, presumptive, rubber stamping of the adequacy of class counsel usually is the result of plaintiff-drafted class certification orders. Once a judge determines to certify a class, many state judges typically invite class counsel to draft the court's certification order. Not surprisingly, plaintiff- drafted certification orders recite, in rote conclusory fashion, that the court finds class counsel adequate to represent the class.”). 187 RUBENSTEIN, supra note 156, § 3:54 n.2. The Advisory Committee’s comments to the 2003 revisions to Rule 23, which included Rule 23(g) concerning class counsel, treats this deference as an advantage. See FED. R. CIV. P. 23(g) advisory committee’s note to 2003 amendment (noting that Rule 23(g)(2) in particular “affords substantial flexibility”). PARKINSON – FINAL 788 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 rarely find class counsel inadequate.188 Instead, courts often operate under a presumption that class counsel is adequate, “reflexively” approving class counsel as presented.189 One commentator remarked that “one has to look long and hard to find cases in which class counsel have been deemed inadequate to represent the class.”190 This occasional deference is not without reason. Indeed, the firm framework advises a substantial degree of deference in the selection of counsel ex ante. However, deference ex ante does not preclude monitoring ex post. For example, several areas of corporate securities law enshrine reporting requirements that amount to a form of continuous adequacy review.191 These adequacy-like requirements are all the more pronounced when the manager or agent subject to review owes a fiduciary duty to the firm’s principals, as is the case in the class action context.192 Moreover, deference is not inevitable. As a textual matter, Rule 23 embraces continuous monitoring by providing for adequacy checks throughout the class proceeding. Even after a class action satisfies the 188 See Mullenix, supra note 183, at 1699. 189 Id. at 1699–1702. 190 Id. at 1699. 191 See SEC v. Tex. Gulf Sulphur, 401 F.2d 833, 850 n.12 (2d Cir. 1968) (“We do not suggest that material facts must be disclosed immediately; the timing of disclosure is a matter for the business judgment of the corporate officers entrusted with the management of the corporation within the affirmative disclosure requirements promulgated by the exchanges and by the SEC.”). 192 See In re General Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d 768, 801 (3d Cir. 1995) (“Beyond their ethical obligations to their clients, class attorneys, purporting to represent a class, also owe the entire class a fiduciary duty once the class complaint is filed”); In re Pet Food Prods. Liab. Litig., 629 F.3d 333, 351 (3d Cir. 2010) (“When the parties have not supplied the information needed for the court to determine whether the settlement is fair, reasonable, and adequate, the court may affirmatively seek out such information. . . . [W]e believe courts may find it necessary to drill down into the case and into the agreement to make an independent, ‘scrupulous’ analysis of the settlement terms.”) (citing In re Prudential Ins. Co., 148 F.3d 283, 317 (3d Cir. 1998)); see also Cotton v. Hinton, 559 F.2d 1326, 1331 (5th Cir. 1977). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 789 adequacy prerequisites at the certification stage,193 Rule 23(g) directs courts to independently review the adequacy of proposed counsel,194 and Rule 23(e)(2) demands a review of the adequacy of any proposed settlement.195 Rule 23 appears to allow courts to monitor the adequacy of class counsel at critical stages throughout the litigation in order to verify that counsel continually satisfies its duties.196 A heavy-handed approach to the adequacy of class counsel might entail leveraging these Rule 23 requirements in tandem with class actions’ tripartite structure. For example, several courts have required representative plaintiffs to act like a corporate board, constantly monitoring class counsel through reporting updates similar to those found in corporate law.197 This internal-governance requirement imposes an additional fiduciary obligation on both the class representative and class counsel.198 Rather than require that class representatives merely exhibit some knowledge of a complex class action,199 courts might invert 193 See FED. R. CIV. P. 23(a)(4) (requiring the representative parties to “fairly and adequately protect the interests of the class”). 194 See FED. R. CIV. P. 23(g)(4) (“Class counsel must fairly and adequately represent the interests of the class.”). 195 See FED. R. CIV. P. 23(e)(2) (“If the proposal would bind class members, the court may approve it only after a hearing and on finding that it is fair, reasonable, and adequate.”). 196 See FED. R. CIV. P. 23(g). 197 See supra note 115. 198 See In re Cendant Corp. Litig., 264 F.3d 201, 276 (3d Cir. 2001) (stating that when reviewing a lead plaintiff’s selection of class counsel “courts should consider: (1) the quantum of legal experience and sophistication possessed by the lead plaintiff; (2) the manner in which the lead plaintiff chose what law firms to consider; (3) the process by which the lead plaintiff selected its final choice; (4) the qualifications and experience of counsel selected by the lead plaintiff; and (5) the evidence that the retainer agreement negotiated by the lead plaintiff was (or was not) the product of serious negotiations between the lead plaintiff and the prospective lead counsel”). 199 See RUBENSTEIN, supra note 156, § 22:44 (“More experience or knowledge may be required on the corporate level. Securities suits involve legal issues and terms that a layperson may find confusing.”). PARKINSON – FINAL 790 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 the burden by requiring class counsel to provide updates to the class representative including, inter alia, a projection of going-forward costs, and a range of anticipated scenarios.200 One can imagine a reporting requirement roughly equivalent to a corporate-filing requirement, such as an “earnings” report, that provides the court with a baseline against which to measure counsel’s performance. The court may use this information to encourage class representatives to consider alternative counsel in the event that counsel’s performance appears inadequate.201 Within this framework, Rule 23 supplies the tools, and class actions the form, to remedy market failure via a heavy-handed adequacy requirement. B. External Boundaries: Make-or-Buy Decisions The firm framework advises that courts apply those considerable portions of Rule 23 that relate to external boundaries—the size and scope of class actions—with a light touch. Not only are there market forces that will often naturally constrain class action growth, but issues concerning external boundaries can extend beyond the judiciary’s core competencies. The composition of a class, its size, and the decision to add a group of similarly situated, yet subtly distinct individuals to the action are all business-like decisions made by the class entrepreneur. Determining 200 Courts might review these reports on the record for something close to an appellate court’s abuse-of-discretion review of a lower court’s findings of fact. See In re Diet Drugs, 582 F.3d 524, 538 (3d Cir. 2009) (reviewing fee award for abuse of discretion on the record). While this may still allow impropriety in the reporting itself, it nonetheless provides an additional level of monitoring. This ex ante monitoring may aid in preventing fee awards from being “locked in” by settlement ex post. Courts may further require that these reports comply with something similar to Generally Accepted Accounting Procedure (“GAAP”) principles, which may prove more susceptible to review. 201 See 7A WRIGHT, MILLER & KANE, supra note 14, § 1765 (“A favorable decision under Rule 23(a)(4) is not immutable, however. If later events demonstrate that the representatives are not adequately protecting the absentees, the court may take whatever steps it deems necessary under Rule 23(c) or Rule 23(d) at that time.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 791 whether to invest the considerable resources necessary to demonstrate that a group of geographically dispersed individuals were harmed by, for example, one of four distinct theories of antitrust impact,202 requires far more than assessing whether “one of these things is not like the others.” Rather, this is a complex and costly decision that requires a self-interested, profit-maximizing entrepreneur to weigh the costs and benefits of expanding the class. Courts exhibit considerable deference to comparable decision-making in corporate law. The business judgment rule, a paragon of judicial deference, establishes a presumption that “a corporation acted on an informed basis, in good faith and in the honest belief that the action taken was in the best interests of the company.”203 This bedrock of corporate law explains courts’ typical deference to the managerial decisions of directors.204 Consequently, even clear errors in judgment will not ordinarily result in liability.205 At least three considerations are cited in defense of the business judgment rule. First, courts are thought to lack the institutional ability to readily evaluate complex business decisions ex post.206 Augmenting this concern is the 202 See Comcast Corp. v. Behrend, 133 S. Ct. 1426, 1430–31 (2013). 203 Aronson v. Lewis, 473 A.2d 805, 812 (Del. 1984). 204 Paramount Commc’ns Inc. v. QVC Network Inc., 637 A.2d 34, 45 n.17 (Del. 1994). 205 See Brehm v. Eisner, 746 A.2d 244, 264 n.66 (Del. 2000) (stating that “directors’ decisions will be respected by courts unless the directors are interested or lack independence relative to the decision, do not act in good faith, act in a manner that cannot be attributed to a rational business purpose or reach their decision by a grossly negligent process that includes the failure to consider all material facts reasonably available”); Strassburger v. Earley, 752 A.2d 557, 582 (Del. Ch. 2000) (“The business judgment rule shields directors from liability for good faith business decisions, even those that turn out to be mistaken.”). 206 See Nixon v. Blackwell, 626 A.2d 1366, 1378 & n.14 (Del. 1993); see also Daniel R. Fischel, The Corporate Governance Movement, 35 VAND. L. REV. 1259, 1288 (1982) (“Courts . . . do not possess the experience, expertise, or information necessary to make complicated business decisions.”); Ryan Scarborough & Richard Olderman, Why Does the FDIC Sue Bank Officers? Exploring the Boundaries of the Business Judgment PARKINSON – FINAL 792 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 risk of hindsight bias—a business decision brought before a court is likely to have ended poorly, which may bias how the court views that decision with the benefit of hindsight.207 Second, market mechanisms obviate the need for judicial interference; those who make poor business decisions are likely to be sanctioned by their firm or lose profits.208 This self-regulating, market-based process is all the more useful where decisions are centralized and confined to a discrete cadre of identifiable agents.209 Finally, courts recognize that the specter of judicial review might deter socially desirable activity levels.210 Rule in the Wake of the Great Recession, 20 FORDHAM J. CORP. & FIN. L. 367, 374 (2015) (noting that, “as a matter of institutional competence, judges and juries are poorly positioned to assess the propriety of complex business decisions, . . . because they lack the specialized skill, knowledge, and judgment of actual businessmen”). 207 See Joy v. North, 692 F.2d 880, 886 (2d Cir. 1982) (professing a concern that “a reasoned decision at the time made may seem a wild hunch viewed years later against a background of perfect knowledge”); see also Johnson, supra note 149, at 456–57 (noting that “courts worry about ‘hindsight bias,’ the tendency to assign an erroneously high probability of occurrence to a probabilistic event simply because it ended up occurring” (internal quotation marks and citation omitted)). 208 See Johnson, supra note 149, at 457 (“[D]irectors who make faulty decisions already face the risk of sanction, both from stockholders who may vote them out of office and from the corrective, competitive pressure of product, labor, and capital markets; judges do not face these sanctions for faulty decisions.”); Michael P. Dooley, Two Models of Corporate Governance, 47 BUS. LAW. 461, 525 (1992) (stating that “the principal deterrent to managerial misbehavior” comes from “competitive forces in the product market, in the internal and external markets for managers and, ultimately, in the market for corporate control”). 209 See Paul L. Regan, The Unimportance of Being Earnest: Paramount Rewrites the Rules for Enhanced Scrutiny in Corporate Takeovers, 46 HASTINGS L. J. 125, 136 (1994). 210 See, e.g., Gagliardi v. TriFoods Int’l, Inc., 683 A.2d 1049, 1052 (Del. Ch. 1996) (noting that “the first protection against a threat of sub-optimal risk acceptance is the so-called business judgment rule”); see also Regan, supra note 209, at 134 (“[T]he business judgment rule promotes a policy which presumes that investors in a company—and perhaps society as a whole—benefit by encouraging talented people to pursue their managerial PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 793 These factors also advise against heavy-handed judicial interference in the class action context. First, courts may possess the ability to determine whether two claims are obviously dissimilar, yet lack the resources required to make close calls between marginal class members with subtle similarities and dissimilarities. Second, self-regulation and market forces can hold class entrepreneurs liable for poor decision-making regarding the size and scope of a class action. Class actions are risky propositions for class counsel—unlike billable work, compensation is often contingent, and counsel substantially bears the risk of failure.211 Class entrepreneurs who overreach in the size or scope of their class action not only incur the added marginal costs of growth,212 but also risk alerting the defendant to a critical flaw in the class composition or distracting the fact- finder with dissimilar claims. Even if courts were to vacate the field—entirely removing the risk of decertification— defendants would continue to highlight those class members who fail to state a claim or present suspect theories of liability.213 Finally, heavy-handed judicial regulation risks deterring socially desirable activity levels. When properly executed, class actions can compel actors to internalize some of the social costs of their behavior—behavior that, absent vision unfettered by the fear that courts, with the benefit of 20/20 hindsight, will find their decisions to have been flawed.”). 211 See Ferdinand S. Tinio, Annotation, Attorneys’ Fees in Class Actions, 38 A.L.R.3d 1384, § 8.5 (1971). 212 See generally Gilles & Friedman, supra note 32. 213 This omits instances in which defendants themselves attempt to augment the size of a class in order to obtain a “global peace.” See, e.g., Sullivan v. DB Invs., Inc., 667 F.3d 273, 285 (3d Cir. 2011). Of course, even in the most prominent example of such a “global peace” class action, Sullivan, the court was clear “that the predominance inquiry should be easily resolved here based on De Beers’s conduct and the injury it caused to each and every class member, and that the straightforward application of Rule 23 and our precedent should result in affirming the District Court’s order certifying the class.” Id. Moreover, as noted earlier, the firm framework applies to damages classes, not settlement classes, which, in many ways, do not resemble a classical firm. PARKINSON – FINAL 794 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 the class mechanism, would often go undetected and under- deterred given the small claims frequently at issue.214 In order to achieve this deterrent function, class entrepreneurs must be risk tolerant and have incentives to take risks.215 This is not to suggest that courts should vacate the field. Rather, courts should review the external boundaries of class actions much in the same way that they approach the business judgment rule. The firm framework calls for a deferential review, which can root out obvious outliers. Close calls relating to the size and scope of a class action should be accorded deference—a sharp break with much of status quo doctrine, which, if anything, pursues an opposite tact. 1. Doctrinal Fit: Numerosity The firm framework does not prescribe change in all instances. Consider, for example, numerosity. Rule 23(a)(1) provides that a class action cannot be maintained if it is too small.216 This requirement ensures that the principal 214 See Robert G. Bone & David S. Evans, Class Certification and the Substantive Merits, 51 DUKE L. J. 1251, 1309 (2002) (“The main purpose of the small-claim class action is deterrence, not compensation. Typically, class members have too little at stake to make individual compensation a significant private or public objective given the enormous social costs of class litigation. Instead, the class action empowers the class attorney as a private attorney general to internalize the social costs of defendant’s acts and deter wrongdoing.”). 215 See Theodore Eisenberg & Geoffrey P. Miller, Attorney Fees in Class Action Settlements: An Empirical Study, 1 J. EMPIRICAL LEGAL STUD. 27, 37–38 (2004); see also Tyler W. Hill, Financing the Class: Strengthening the Class Action Through Third-Party Investment, 125 YALE L. J. 484, 505 (2015) (“Without any reason to believe that they will receive a higher premium on their effort for a riskier lawsuit, rational lawyers will choose to pursue less risky suits. But these less risky suits often correspond to diminished enforcement benefits relative to suits that involve more risk.”). 216 See FED. R. CIV. P. 23(a)(1) (“Prerequisites. One or more members of a class may sue or be sued as representative parties on behalf of all members only if: (1) the class is so numerous that joinder of all members is impracticable.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 795 alternative to a class action—joinder—is “impracticable.”217 In practice, the numerosity requirement is all but forgotten. Courts regularly defer to class counsel when evaluating numerosity, rarely striking down a putative class action because it is too small.218 Occasionally courts point to Supreme Court dicta for the proposition that a class must contain at least fifteen members.219 As such, classes containing as few as thirty, forty, and fifty members are regularly certified.220 Even classes with as few as fifteen members, however, have been certified, Supreme Court dicta notwithstanding.221 Some courts simply “infer” numerosity, declining outright the possibility that an attorney would bother bringing forward a class that may be too small.222 217 FED. R. CIV. P. 23(a)(1). 218 See 1 MCLAUGHLIN, supra note 115, § 4:5 (“Indeed, numerosity ‘is rarely contested in class actions brought on behalf of shareholders or traders in publicly owned corporations. In class actions brought on behalf of securities traders, federal trial courts are quite willing to accept common sense assumptions in order to support a finding of numerosity.’” (quoting Zeidman v. J. Ray McDermott & Co., Inc., 651 F.2d 1030, 1039 (5th Cir. 1981))); RUBENSTEIN, supra note 156, § 3:13 (“[C]ertification denials on this basis are rare and courts are generally forgiving where plaintiffs are unable to do more than set forth commonsense assumptions.”). 219 See Gen. Tel. Co. of the N.W., Inc. v. EEOC, 446 U.S. 318, 330 (1980). For an application of this dictum, see, for example, Harik v. Cal. Teachers Ass’n, 326 F.3d 1042, 1051–52 (9th Cir. 2003) (“The Supreme Court has held fifteen is too small.”). 220 See 7A WRIGHT, MILLER & KANE, supra note 14, § 1762. 221 See, e.g., Jackson v. Danberg, 240 F.R.D. 145, 147–48 (D. Del. 2007) (sixteen members); Bublitz v. E.I. du Pont de Nemours and Co., 202 F.R.D. 251, 255–56 (S.D. Iowa 2001) (seventeen members); Gaspar v. Linvatec Corp., 167 F.R.D. 51, 55–57 (N.D. Ill. 1996) (eighteen members); Phila. Elec. Co. v. Anaconda Am. Brass Co., 43 F.R.D. 452, 463 (E.D. Pa. 1968) (“While 25 is a small number . . . it is a large number when compared to a single unit. I see no necessity for encumbering the judicial process with 25 lawsuits, if one will do.”). 222 Dean v. China Agritech, 2012 WL 1835708, at *4 (C.D. Cal. May 3, 2012); see also RUBENSTEIN, supra note 156, § 3:13 (noting that, “although the plaintiff has the burden of showing that joinder is impracticable, a PARKINSON – FINAL 796 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 This deference coheres to the firm framework. Numerosity is a prime example of an issue that concerns the external boundaries of a class action—literally, its size. While the size of a class action may present principal-agent problems,223 interventionist regulation of these problems is not necessarily the solution. Not only are courts ill-suited to determine whether a class should be larger,224 but class counsel and class members have compelling incentives to pursue a class action only if it is superior to available (or “practicable”) alternatives.225 Deference does not require courts to abandon the numerosity requirement altogether. Indeed, courts often employ the numerosity requirement as a screening mechanism to determine whether other provisions of Rule 23, typically those relating to internal governance, should be applied with greater scrutiny.226 One commentator notes the “interrelationship of the numerosity prerequisite and some of the other class-action requirements”: For example, the requirement in subdivision (a)(4) that the representatives adequately protect the interests of all members of the class is a peculiarly good-faith estimate of the class size is sufficient when the precise number of class members is not readily ascertainable”). 223 See supra Section II.B.2. 224 Courts lack the investigative and fact-finding resources to reach this conclusion sua sponte. At a certification hearing, there will be a predictable lack of any adversarial position on whether the class size is too small; no defendant will argue that the class should be larger, while no class counsel will argue that the class should not be certified because it is insufficiently numerous. Consequently, courts often defer to the judgment of class counsel. See, e.g., Gortat v. Capala Bros., Inc., 257 F.R.D. 353, 362 (E.D.N.Y. 2009) (stating that class counsel need only show “some evidence of the class members that, in effect, provides the Court with a reasonable estimate”); In re HealthSouth Corp. Sec. Litig., 257 F.R.D. 260, 274 (N.D. Ala. 2009) (stating that the court can rely on class counsel in “a finding of numerosity” and “make common sense assumption[s]” (citing Zeidman v. J. Ray McDermott & Co., 651 F.2d 1030, 1039 (5th Cir. 1981))). 225 See supra Section III.A.2. 226 See, e.g., Eisen v. Carlisle & Jacquelin, 391 F.2d 555, 570 (2d Cir. 1968) (Lumbard, J., dissenting). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 797 significant requirement when the class is very large. Furthermore, because of the mandatory notice requirement applicable in actions brought under Rule 23(b)(3) and the binding effect of the judgment on all class members, a few courts have suggested that concerns about size may have increased importance. Again, however, the concern these courts are expressing is not whether Rule 23(a)(1) has been satisfied but whether it is proper to bind all the members of the class and whether it is feasible to give them all proper notice.227 This is precisely what the firm framework prescribes. Courts should subject class actions that have been accorded greater external deference to greater internal scrutiny.228 This screening process is not unique to class actions or alien to courts—indeed, it is a familiar feature of antitrust law. Not only are smaller firms generally accorded a more latitude in the antirust context,229 but courts (and antitrust regulators) also rely on market share as a factor when weighing the degree of scrutiny that potentially anticompetitive behavior should receive.230 Though smaller 227 7A WRIGHT, MILLER & KANE, supra note 14, § 1762. 228 See infra Part V. 229 See STEPHEN G. BREYER ET AL., ADMINISTRATIVE LAW AND REGULATORY POLICY: PROBLEMS, TEXT, AND CASES 10 (7th ed. 2011) (noting that “Congress has granted small sellers an exemption from the antitrust laws, thereby allowing them to organize to deal more effectively with a large buyer”); ROBERT H. BORK, THE ANTITRUST PARADOX: A POLICY AT WAR WITH ITSELF 204–05 (The Free Press 1993) (1978). 230 See U.S. DEP’T OF JUSTICE & FED. TRADE COMM’N, HORIZONTAL MERGER GUIDELINES § 5.2 (2010). For examples of the judiciary’s use of the DOJ-FTC’s Merger Guidelines, see, for example, FTC v. H.J. Heinz Co., 246 F.3d 708, 721 (D.C. Cir. 2001) (stating that, “given the high concentration levels, the court must undertake a rigorous analysis of the kinds of efficiencies being urged by the parties in order to ensure that those ‘efficiencies’ represent more than mere speculation and promises about post-merger behavior”); FTC v. Staples, Inc., 970 F. Supp. 1066, 1082 (D.D.C. 1997) (“The Merger Guidelines, of course, are not binding on the Court, but . . . they do provide ‘a useful illustration of the application PARKINSON – FINAL 798 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 firms can certainly present antitrust problems, the costs of those problems are often too small to justify aggressive regulation. 2. Doctrinal Change: Commonality and Predominance By contrast, the firm framework prescribes a substantial overhaul to both commonality and predominance doctrines. As noted earlier, commonality provides that a class may be certified “only if there are questions of law or fact common to the class”;231 predominance requires that, for a damages class, “the court finds that the questions of law or fact common to class members predominate over any questions affecting only individual members.”232 Both directly implicate the scope of putative class actions, providing courts with a mechanism to restrict size. Taking advantage, courts apply often these provisions of Rule 23 with a heavy hand, ostensibly to root out agency costs. The Supreme Court articulated a particularly stringent understanding of the Rule 23(a)(2) commonality requirement in Wal-Mart, which concerned a class action comprised of more than 1.5 million current and former employees who alleged that their employer, Wal-Mart Stores, Inc., discriminated against women on the basis of sex by denying equal pay, promotion opportunities, and fair treatment in the workplace.233 The Ninth Circuit affirmed the district court’s certification order over an ardent dissent, which framed its objection as one that turned, in part, on class size: “No court has ever certified a class like this one, until now.”234 Seizing on that observation, the defendant’s of the HHI,’ . . . and the [C]ourt will use that guidance here.” (quoting FTC v. PPG Indus., Inc., 798 F.2d 1500, 1503 n. 4 (D.C. Cir. 1986))). 231 FED. R. CIV. P. 23(a)(2). 232 FED. R. CIV. P. 23(b)(3). 233 Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 342 (2011). 234 Dukes v. Wal-Mart Stores, Inc., 603 F.3d 571, 628 (9th Cir. 2010) (Ikuta, J., dissenting). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 799 principal objection before the Court was that it had a right to present individualized defenses with respect to each and every member of the class, revealing a lack of commonality and rendering any class-wide proceeding unmanageable.235 The Court agreed, determining that, because “it will be impossible to say that examination of all the class members’ claims for relief will produce a common answer to the crucial question why was I disfavored,” the putative class failed the commonality requirement.236 The dispute in Wal-Mart can be understood as one over how deferentially courts should approach commonality. For the dissent, “[e]ven a single question of law or fact common to the members of the class will satisfy the commonality requirement.”237 This standard requires class counsel to present an open controversy applicable to all class members,238 such as whether a defendant’s policy of managerial discretion discriminates on the basis of sex. For the majority, however, there could not be common questions without “common answers” to “drive the resolution of the litigation.”239 Consequently, commonality was defeated by the contention that the defendant should be able to present individualized defenses in the event that a jury was to initially determine that, “yes, as a general policy, managerial discretion does discriminate on the basis of sex.”240 This approach to commonality directs courts to apply the 235 See Wal-Mart, 564 U.S. at 365–66. 236 Id. at 352 (emphases altered). 237 Id. at 369 (Ginsburg, J., concurring in part and dissenting in part) (quoting Richard A. Nagareda, The Preexistence Principle and the Structure of the Class Action, 103 COLUM. L. REV. 149, 179 n.110 (2003)). 238 Id. (“A ‘question’ is ordinarily understood to be ‘[a] subject or point open to controversy.’” (quoting THE AMERICAN HERITAGE DICTIONARY OF THE ENGLISH LANGUAGE 1483 (3d ed. 1992))). 239 Id. at 350 (quoting Richard A. Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L. REV. 97, 132 (2009)). 240 Note that Wal-Mart did not foreclose this as a viable cause of action. See Samuel R. Bagenstos, The Structural Turn and the Limits of Antidiscrimination Law, 94 CAL. L. REV. 1, 26 (2006). PARKINSON – FINAL 800 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 requirement with a heavy hand.241 If anything, Wal-Mart suggests deference to the judgment of the defendant that certification should be avoided where class members hold “different jobs, at different levels of [an employer’s] hierarchy, for variable lengths of time, in [many locations], sprinkled across 50 states, with a kaleidoscope of supervisors.”242 This quasi-uniformity requirement provides courts with a ready tool for constraining the size of putative class actions. The Court’s interpretation of the Rule 23(b)(3) predominance requirement mirrors this approach. For example, in Comcast the Court reviewed a class of more than two million cable-television subscribers, alleging various violations of antitrust law and sought certification as a damages class pursuant to Rule 23(b)(3).243 Class counsel presented the lower court with an antitrust impact model— an econometric study that claimed to demonstrate whether and how an antitrust injury occurred244—illustrating four distinct ways that the defendant’s behavior had allegedly injured the class members. Each theory of antitrust injury was an outgrowth of a single theory of liability—the 241 See Wal-Mart, 564 U.S. at 357 (holding that the putative class had failed to “establish the uniform, store-by-store disparity upon which the plaintiffs’ theory of commonality depends”). At least one commentator has remarked that, after Wal-Mart, courts “conducting the commonality analysis . . . have focused on whether the defendant engaged in uniform conduct that had a substantially similar impact on all class members.” Anthony F. Fata, Doomsday Delayed: How the Court's Party-Neutral Clarification of Class Certification Standards in Wal-Mart v. Dukes Actually Helps Plaintiffs, 62 DEPAUL L. REV. 675, 687 (2013). 242 Wal-Mart, 564 U.S. at 359–60 (quoting Dukes v. Wal-Mart Stores, Inc., 603 F.3d 571, 652 (9th Cir. 2010) (Kozinski, C.J., dissenting)). 243 See Comcast Corp. v. Behrend, 133 S. Ct. 1426, 1429–30 (2013). 244 For a general explanation of the importance of these econometric impact models, see In re Linerboard Antitrust Litig., 305 F.3d 145, 153–55 (3d Cir. 2002); see also Ronald W. Davis, Standing on Shaky Ground: The Strangely Elusive Doctrine of Antitrust Injury, 70 ANTITRUST L. J. 697, 744 (2003) (stating that impact models demonstrate that the plaintiffs’ damages were the result of an injury “intended to be prevented by the statute or rule the plaintiff has invoked to establish liability”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 801 defendant, Comcast, injured customers by “clustering” its cable assets, concentrating a regional base of consumer operations to exclude competitors.245 On appeal, the Supreme Court trained its sights on the Rule 23(b)(3) predominance requirement.246 For the majority, the presence of multiple antitrust-injury theories was troubling. Even though each member of the class presented an identical theory of antitrust liability—clustering violates antitrust law—members of the class may have been injured by the same activity in one of four distinct ways.247 Therefore, individualized issues predominated.248 The dissent took issue with the majority’s application of the predominance requirement, pointing out the “well nigh universal” rule that individualized damages are no more than one non-dispositive factor among many in the predominance inquiry.249 Indeed, prior to Comcast courts rarely concerned themselves with individualized damages,250 245 See Comcast, 133 S. Ct. at 1430–31. 246 See id. at 1432–33. Note that the Court quickly dispensed of the question on which certiorari was granted: “Whether a district court may certify a class action without resolving whether the plaintiff class has introduced admissible evidence, including expert testimony, to show that the case is susceptible to awarding damages on a class-wide basis.” Id. at 1431 n.4. It did so by relying on Wal-Mart and noting that consideration of the merits at the certification stage, when necessary to pass judgment on the predominance inquiry, is well established. Id. at 1433 (citing Wal- Mart, 564 U.S. at 349–50). 247 See id. at 1434 (“For all we know, cable subscribers in Gloucester County may have been overcharged because of petitioners’ alleged elimination of satellite competition . . . while subscribers in Camden County may have paid elevated prices because of petitioners’ increased bargaining power.”). 248 See id. at 1434–35. 249 Id. at 1437 (Ginsburg and Breyer, JJ., dissenting). 250 See, e.g., In re New Motor Vehicles Canadian Exp. Antitrust Litig., 522 F.3d 6, 28 (1st Cir. 2008) (“Predominance is not defeated by individual damages questions as long as liability is still subject to common proof.”); Arreola v. Godinez, 546 F.3d 788, 801 (7th Cir. 2008) (stating that the “need for individual damages determinations does not, in and of itself, require denial of . . . certification”); Thorogood v. Sears, Roebuck and Co., 547 F.3d PARKINSON – FINAL 802 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 almost universally certifying class actions that presented common theories of liability coupled with individualized damages.251 Like Wal-Mart, this dispute can be viewed as one over the amount of deference that should be accorded to class counsel when reviewing the scope of a class action. While the dissent would have adhered to the old regime, which permitted class counsel to litigate class-wide liability at step one and then choose to absorb the cost of litigating individualized damages at step two,252 the Comcast majority hampered the ability of class counsel to self-regulate. Some courts now decline to certify putative class actions that risk de minimis 742, 748 (7th Cir. 2008); Chiang v. Veneman, 385 F.3d 256, 273 (3d Cir. 2004) (stating that it “has been commonly recognized that the necessity for calculation of damages on an individual basis should not preclude class determination when the common issues which determine liability predominate” (citing Bogosian v. Gulf Oil Corp., 561 F.2d 434, 456 (3d Cir. 1977); see also RUBENSTEIN, supra note 156, § 4:54 (citing FED. R. CIV. P. 23 advisory committee’s note to 1966 Amendment). 251 See, e.g., Beattie v. CenturyTel, Inc., 511 F.3d 554, 564 (6th Cir. 2007) (quoting In re Visa Check/MasterMoney Antitrust Litig., 280 F.3d 124, 139 (2d Cir. 2001) (holding that Rule 23(b)(3) predominance is satisfied “when liability can be determined on a class-wide basis, even when there are some individualized damage issues”)); Klay v. Humana, Inc., 382 F.3d 1241, 1260 (11th Cir. 2004) (“It is primarily when there are significant individualized questions going to liability that the need for individualized assessments of damages is enough to preclude 23(b)(3) certification.”); Yokoyama v. Midland Nat’l Life Ins. Co., 594 F.3d 1087, 1094 (9th Cir. 2010) (stating that “damage calculations alone cannot defeat certification”); see also 1 MCLAUGHLIN, supra note 115, § 5:23 (defining liability as the “core” of the predominance inquiry). 252 For examples of this process, see Mejdrech v. Met-Coil Sys. Corp., 319 F.3d 910, 911 (7th Cir. 2003) (“If there are genuinely common issues, issues identical across all the claimants . . . the accuracy of the resolution of which is unlikely to be enhanced by repeated proceedings, then it makes good sense . . . to resolve those issues in one fell swoop.”); In re Methyl Tertiary Butyl Ether (“MTBE”) Prods. Liab. Litig., 241 F.R.D. 435, 448 (S.D.N.Y. 2007) (“When liability can be resolved by a jury with a single decision that applies to the whole class, and the only individual question left to resolve relates to damages, class certification is warranted.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 803 discrepancies in individualized damages,253 even where individualized damages are readily addressed. For example, following Comcast some courts have declined to certify class actions alleging that groups of employees—all working for the same employer, yet for different hours—were impermissibly denied due pay because, in the event of certification, each employee would be owed individualized damages corresponding to the (different) amounts of time that each worked.254 Reading this uniformity-like standard into the predominance requirement poses the risk that Rule 23(b)(3) will be satisfied only when a class seeks uniform statutory damages.255 These twin doctrines are driven, in part, by a principal- agent concern.256 Notably, Wal-Mart and Comcast do not 253 See, e.g., In re Rail Freight Fuel Surcharge Antitrust Litig.-MDL No. 1869, 725 F.3d 244, 253, 255 (D.C. Cir. 2013) (noting that, before Comcast, “the case law was far more accommodating to class certification under Rule 23(b)(3),” whereas Comcast establishes, “[n]o damages model, no predominance, no class certification”); Smith v. Family Video Movie Club, Inc., 2013 WL 1628176, at *10 (N.D. Ill. Apr. 15, 2013) (denying certification of a Rule 23(b)(3) class by interpreting Comcast to hold that “damages must be susceptible to measurement across the entire class, and individual damage calculations cannot overwhelm questions common to the class”). 254 See, e.g., Smith, 2013 WL 1628176, at *10. 255 Consider, for example, Judge Richard Posner’s observation that a damages-centric predominance inquiry might increase the need for a statutory damages regime for all manner of cases, including commonplace products-liability claims. See Thorogood, 547 F.3d at 748 (stating that claims for statutory damages “might not require individual proof,” but claims for actual damages leave courts the difficult task of “determining the relief to which the individual class members are entitled”); see also Shabazz v. Morgan Funding Corp., 269 F.R.D. 245, 250–51 (S.D.N.Y. 2010) (“Any class action based on unpaid wages will necessarily involve calculations for determining individual class member damages, and the need for such calculations do [sic] not preclude class certification.”). 256 Importantly, both rely on Amchem, which explicated the Court’s information-asymmetry concern with large class actions. See supra notes 58–65 and accompanying text; see also Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 362 (2011) (“Rule 23(b)(3) . . . is an ‘adventuresome innovation’ . . . framed for situations ‘in which class-action treatment is not as clearly PARKINSON – FINAL 804 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 discuss blackmail settlements or ruinous corporate liability. Wal-Mart does, however, express an agency-cost concern (even if impliedly so).257 Meanwhile, Comcast voices concern for, inter alia, the costs that adjudicating individualized damages would impose on the courts.258 These are precisely the types of costs that class counsel, as class entrepreneurs, are well positioned to internalize.259 Were it more costly to litigate individual defenses in the aggregate—the alternative suggested by Wal-Mart260—profit-maximizing market participants would not so litigate. Similarly, were it a greater drain on the courts’ limited resources to dole out individualized damages in aggregate litigation—again, the presumed alternative in Comcast—rational actors would self-regulate and respond in kind. There are numerous alternatives to the heavy-handed approach in Wal-Mart and Comcast, respectively. For example, in the context of commonality, defendants might be permitted—indeed encouraged—to litigate individual defenses after class certification.261 Because plaintiffs bear called for.’” (quoting Amchem Products, Inc. v. Windsor, 521 U.S 591, 614 (1997))); Comcast Corp. v. Behrend, 133 S. Ct. 1426, 1432 (2013) (“Rule 23(b)(3)’s predominance criterion is even more demanding than Rule 23(a).”) (citing Amchem, 521 U.S. at 623–24). 257 See Wal-Mart, 564 U.S. at 354–60 (determining that the proffered statistical and anecdotal evidence presented by the plaintiffs is “worlds away from ‘significant proof’ that Wal–Mart ‘operated under a general policy of discrimination,’” meaning that the class would need to present far more rigorous studies and a greater number of anecdotal examples of workplace discrimination); id. at 364 (“Respondents’ predominance test, moreover, creates perverse incentives for class representatives to place at risk potentially valid claims for monetary relief.”). 258 See Comcast, 133 S. Ct. at 1434–35 (expressing concern that if the class action at bar were certified, the court would be required to review an “endless” number of “permutations involving four theories of liability”). 259 See supra Section III.A.2. 260 See Wal-Mart, 564 U.S. at 365. 261 See RUBENSTEIN, supra note 156, § 4:55 (“The general rule, regularly repeated by courts in many circuits, is that ‘[c]ourts traditionally have been reluctant to deny class action status under Rule 23(b)(3) simply because affirmative defenses may be available against individual PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 805 the burden of proof, class counsel would be required to internalize the cost of presenting evidence to rebut each individual defense.262 Class counsel would, predictably, avoid including claims in the class action for which the expected value of the claim (whether by judgment or settlement) is outweighed by the expected cost of litigation.263 Turning to predominance, and taking the facts presented in Comcast as an example, the class might be required to present individualized impact models for each species of claim relating to the defendant’s clustering.264 These models are members.’ This is particularly true given the range of procedural mechanisms available to courts to deal with potentially individualized affirmative defenses.”); see also Wal-Mart, 564 U.S. at 377 (“A finding that Wal–Mart’s pay and promotions practices in fact violate the law would be the first step in the usual order of proof for plaintiffs seeking individual remedies for company-wide discrimination. That each individual employee’s unique circumstances will ultimately determine whether she is entitled to backpay or damages . . . should not factor into the Rule 23(a)(2) determination.”). 262 See Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 361–62 (1977) (holding in a workplace-discrimination class action that the burden of proof initially lies with the plaintiff and “a district court must usually conduct additional proceedings . . . to determine the scope of individual relief,” at which point the defendant may raise individual affirmative defenses to “demonstrate that the individual applicant was denied an employment opportunity for lawful reasons”). 263 Which includes not only the costs of adducing evidence for that claim, but also the risk of distracting the jury or otherwise implicating the probability that other claims might prevail by engendering distrust, ill- will, and so forth. 264 For an example, consider Messner v. Northshore Univ. HealthSystem, 669 F.3d 802 (7th Cir. 2012). In the antitrust class action at issue, the court stated that the plaintiffs’ “ability to use common evidence to show impact on the class did not ultimately depend on assuming the uniformity” of antitrust impact. Id. at 818. While “such uniformity would certainly simplify matters,” “a lack of uniformity would only require [the plaintiffs’ econometrics expert] to do more [antitrust impact] analyses for each contract.” Id. at 819. The court held that the “ability to use such common evidence and common methodology to prove a class’s claims is sufficient to support a finding of predominance on the issue of antitrust impact for certification under Rule 23(b)(3).” Id. PARKINSON – FINAL 806 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 incredibly expensive.265 Again, because the plaintiff bears the burden of proof, class counsel would be forced to internalize these costs. While such deference might be thought to externalize considerable costs onto defendants, consider how the placement of the evidentiary burden impacts that assessment. Even in the context of blackmail settlements— which are of debatable theoretical266 and empirical267 grounding—defendants would be able to discount the plaintiffs’ potentially astronomical costs of litigating individualized defenses and damages by (credibly) threatening to require the class to meet its evidentiary burden (and the attendant litigation costs).268 Even under a deferential approach, class counsel might be compelled to internalize the costs of expanding the external boundaries of their class action ex ante.269 265 In re Am. Express Merchants’ Litig., 634 F.3d 187, 198 (2d Cir. 2011) (“[E]ven a relatively small economic antitrust study will cost at least several hundred thousand dollars, while a larger study can easily exceed $1 million.”). 266 See Alexandra D. Lahav, The Curse of Bigness and the Optimal Size of Class Actions, 63 VAND. L. REV. EN BANC 117, 127 (“[T]he class action is intended to force the defendant to account for all the damages it caused a group of claimants. The potential for enormous damages awards does not give the defendant a get-out-of-certification-free card. A rule against large class actions would merely encourage defendants to think big when they violate the law.”). 267 See Allan Kanner & Tibor Nagy, Exploding the Blackmail Myth: A New Perspective on Class Action Settlements, 57 BAYLOR L. REV. 681, 693– 95 (2005). 268 See infra Part V. 269 The longstanding debate over blackmail settlements is beyond the scope of this inquiry. See generally, e.g., Bruce Hay & David Rosenberg, “Sweetheart” and “Blackmail” Settlements in Class Actions: Reality and Remedy, 75 NOTRE DAME L. REV. 1377 (2000). Whether blackmail settlement is a genuine concern, it is not a germane concern. That is, the firm framework’s prescriptions do not spring blackmail settlements into existence where they did not once exist; rather, the framework might be said to augment the impact of this pre-existing background risk. More fundamentally, and to the extent that blackmail settlement is a genuine concern, it is inextricably linked to the risk that class counsel will PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 807 V. EXCHANGING INTERNAL SCRUTINY FOR EXTERNAL DEFERENCE Class actions are, of course, not literally firms. Among the differences between the two, one merits particular attention: the role of contracting and the ex ante process by which market participants come to transact within firms. Contracting allows markets to calibrate costs and ensures, or aims to ensure, that transactions cannot be conducted at a lower cost outside the firm.270 That dynamic breaks down in class actions because there is no ex ante contracting between sell out the value of class members’ claims. To see why, consider the validity of a defendant’s claim that it was blackmailed into settling for an amount nearing the full value of a putative class action versus a defendant’s claim that it was blackmailed into settling at a relatively small per-plaintiff value. But see Hay & Rosenberg, supra, at 1377–78 (arguing that the risks of “blackmail” and “sweetheart” settlements are “conflict[ing]” “mirror images” of one another). Both the “sellout” and “blackmail” settlement concerns turn on the assumption that foul play at the settlement stage will generate an outcome that is not reflective of the merits of the case and that reflects a detectably lower per-claim value. This concern is not without merit. External deference coupled with internal scrutiny may augment the risk that class counsel will sell out the value of the class members’ ownership interests in settlement. That said, several factors, each relating to the firm-centric framework’s complete treatment of class actions as firms, mitigate this concern. Indeed, the risk of undervaluing a claim at settlement—whether labeled as a blackmail or sellout settlement—simply begs the question of what mistakes courts should make. Courts are not well-suited to address mistakes concerning external boundaries. By contrast, courts have built up considerable expertise in monitoring the internal workings of both firms and class actions. In this way, even defendants might benefit from the firm framework’s intense scrutiny of the internal workings of the class action and, ultimately, settlement. More fundamentally, one may object that the firm framework speaks only to the incentives facing the class counselor who intends to litigate a class action to judgment. Not necessarily so. Importantly, the defendant should serve as a market-mimicking force that compels the settlement- minded class counsel to internalize—or at a minimum recognize—costs. See Kanner & Nagy, supra note 267, at 693–95. 270 See COASE, The Nature of the Firm, supra note 93, at 42–44. PARKINSON – FINAL 808 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 ownership and management.271 This descriptive shortcoming of the firm framework is not insignificant. Because the framework prescribes deference to the size and scope of a class action, unscrupulous class counsel may overreach and form classes comprised of members whose individual claims hardly resemble one another. Absent ex ante contracting or regulatory constraints, class counsel may capture claims that obviously do not belong. Relatedly, deference may also permit class counsel to bind unwitting class members who would be unwilling to contract ex ante. Fortunately, treating class actions as firms allays this descriptive shortcoming. That is, by embracing the firm framework and its doctrinal entailments, courts can exchange internal scrutiny for external deference, which will often yield results that resemble what might have taken place were ex ante contracting available. Internal scrutiny can compel class counsel to internalize the costs of including dissimilar claims in their class—particularly by carefully calibrating fee structures prior to settlement or judgment. Consider the relationship between one area of doctrine in which courts apply considerable scrutiny—compensation— and one in which courts apply very little deference—class size. Class size clearly relates to compensation, which is, in part, a function of total damages awarded and the cost of obtaining that award. Class counsel faces several costs when considering whether to increase the size of a class action. Some costs are spread evenly across each additional class member, regardless of the marginal claim’s similarity to the rest of 271 See supra note 137 and accompanying text; see also Lahav, supra note 34, at 92–95 (dismissing the notion that there might be any analogy between class action jurisprudence and “corporate governance” due to the lack of any contract-like market mechanisms in the former setting). For a related discussion of this issue in the settlement context, see Issacharoff & Nagareda, supra note 59, at 1654 (“Class actions further compromise litigant autonomy, for absent class members typically express their consent to a binding settlement not affirmatively but only tacitly, through their failure to withdraw from the class representation.”). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 809 the class—for example, the cost of issuing notice.272 Other costs, however, vary with the marginal claim’s similarity to the rest of the class.273 Some of these marginal costs are imposed by the heavy-handed application of Rule 23. Commonality and predominance requirements impose at least two costs: first, the risk of decertification; and second, the cost involved in making dissimilar claims “look like” the rest of the class in order to avoid that decertification risk. Together these “shoehorning costs” can take many forms. For example, shoehorning might involve attempting to conceal dissimilar class members from the court and defendant; or, it might involve the opposite. That is, class counsel may attempt to convince a skeptical court that a claim “belongs” by rebutting allegations of dissimilarity with high-skill, high- cost attorneys, exhaustive research and brief writing, and expensive expert studies showing that ostensibly dissimilar claims are in fact “like” the others. Separately, some marginal costs associated with including dissimilar claims in a class action are not attributable to Rule 23. These “litigation costs” resemble shoehorning costs, but would persist even if courts entirely vacated the field and no class action were ever decertified. Examples include identifying and locating unique claims in the first instance, convincing the defendant in settlement negotiations or the fact-finder at trial that these dissimilar claims are as meritorious as others, rebutting individualized defenses leveled against unique claims, and meeting the evidentiary burden for each claim species within the class. These litigation costs—costs that exert pressure on the 272 See supra text accompanying note 98. 273 Professors Issacharoff and Nagareda recognized, for example, that as a class grows so too do both “opportunity” and “risk”; specifically, as a class increases in size, so too does the risk “new class counsel can emerge, claiming—usually in a different court—to be the true champions of the class.” Issacharoff & Nagarada, supra note 59, at 1667. Though this particular issue, raised in the context of settlement classes and class objectors, is beyond the scope of this inquiry, it demonstrates a comparable market-based risk that class counsel must internalize. PARKINSON – FINAL 810 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 decision whether to include a marginal claim—are no less significant than shoehorning costs. These twin costs can be leveraged against one another. Exchanging external deference (thereby lessening shoehorning costs) for a heavy-handed review of fee structures can compel class counsel to internalize the costs of aggregation. As a preliminary matter, note that the firm framework does not require that courts ignore the commonality and predominance requirements; egregiously dissimilar claims can be quickly spotted by a court (or flagged by the defendant) and rooted out. For those marginal claims that are closer calls—some of which are sufficiently similar to the prototypical class member and should be included, some of which are not and should be excluded— exchanging internal scrutiny for external deference can compel class counsel to make the “right” decisions. Because compensation is a function of damages awarded (or obtained in settlement) and costs, one can restate the relationship: the damages that class counsel seeks are a function of, inter alia, expected compensation and cost. Consider four fee structures that are either grounded in or follow from extant case law: TABLE 1: CLASS-COUNSEL FEE STRUCTURES COMPENSATION MODEL VALUE OF MARGINAL CLAIM TO CLASS COUNSEL Percent-of-Fund Dx° (E) – Xn – Y Percent-of-Fund + Cost Dx° (E) – Xn+ Xn – Y + Y Percent-of-Fund Net of Cost (Dx° – Xn – Y) (E) Lodestar Crosscheck (E) Xn – Y *Where (Dn°) is damages owed per claimant; (E) represents attorney- reimbursement, be it a percentage assigned by the court or a calculated fee; (Xn) is the cost of litigation per claimant; and (Y) is the non-litigation cost of the marginal class claim (i.e., notice). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 811 The first fee structure is a subtle break with the status quo—it assigns class counsel a percentage of the common fund but requires the attorney to absorb all litigation costs.274 This is by far the least forgiving—and most interventionist—of the above fee structures. It allows the court to calibrate compensation by manipulating the attorney-reimbursement multiplier without regard to cost. The second and third fee structures are related versions of the status quo’s percent-of-the-fund model.275 The fourth is the lodestar-crosscheck model, which ties compensation to “the hours reasonably expended by the plaintiffs’ counsel on the litigation by a reasonable hourly fee in view of the geographical area.”276 Now consider how these fee structures impact the decision whether to include a marginal claim in a class action.277 For simplicity, assume that the prototypical class members who are assumed to “belong” in a class action cluster around either a common amount of damages owed (Dn0) or a common account of liability, yielding a uniform cost of litigation (Xn). This simplifying assumption is not wholly divorced from reality—the former tracks stringent applications of Comcast and the latter generally tracks Wal- Mart. Further, assume a class of 1,000 members, each of whom is owed $1,000 in damages. Class counsel has invested $100,000 in litigation costs to attain an 80 percent probability of success. Assume that status quo doctrine governs. Class counsel is presented with three potential plaintiffs, all or none of whom they may add to their class 274 Status quo percent-of-the-fund models compensate class counsel for reasonable costs incurred during the course of litigation. See supra notes 48–53 and accompanying text. 275 Available post-judgment filings do not make clear whether the attorney-reimbursement multiplier is typically calculated net of costs or prior to deducting the reasonable costs of litigation from the common fund. It is probable that courts varyingly apply both approaches. Certainly both approaches are available at the court’s discretion. 276 2 MCLAUGHLIN, supra note 45, § 6:24. Note that the lodestar- crosscheck model does not compensate for post-litigation costs (i.e., notice). 277 See Issacharoff & Nagareda, supra note 59, at 1699. PARKINSON – FINAL 812 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 action: Plaintiff 1 is a prototypical class member who should have been included in the first instance. Like the other class members, he is owed $1,000 in damages and adding him to the class costs nothing more than the cost of notice. Plaintiff 2 should be included in the class—though his claim is not identical to the prototypical class member’s, it is very close. He was injured by the same defendant in the same underlying course of events, but under slightly unique circumstances. He is owed $900 in damages, slightly less than the prototypical class member, and litigating his claim will cost an additional $500—both in shoehorning and litigation-market costs. Plaintiff 3 clearly should not be included in the class—he was injured by the same defendant but in an entirely unrelated matter. He is owed $2,000 in damages, double the prototypical class member, but litigating his claim in the class action will add $1,500 in costs, reflecting the considerable shoehorning and litigation- market costs that class counsel will incur. Which marginal claims, if any, will class counsel include in their class action? Applying the four fee structures to this stylized example yields the following results: Heavy Hand Model Plaintiff 3 (Should Not Be Included) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non-Lodestar Marginal Costs (i.e., Notice) (Y) Dc0 (2,000) X3 (1,500) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Plaintiff Dc0 (.7) Dc0 (.7)-X3-Y (Dc0-X3-Y) (.7) Dc0-(4)X3 Counsel Dc0 (.3)- X3-Y Dc0 (.3)- X3+X3-Y+Y (Dc0-X3-Y) (.3) (4)X3-Y Result Applied 2,000(.3)- 1,500-1 =-901 2,000(.3)- 1,500 +1,500 -1 + 1 =600 (2,000-1,500- 1) (.3) =149.7 4(1,500)-1 =4,999 Include in the Class? No Include Yes Include Yes Include Yes Include TABLE 2: WAL-MART/COMCAST PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 813 Several conclusions follow. First, class counsel’s decisions are clearly intertwined with the fee structure that a court adopts and how it regulates compensation. Indeed, each fee structure yields a unique result. Second, the status quo’s preference for a model that compensates class counsel for litigation costs yields a predictable result: class counsel rarely, if ever, turns down a marginal class member. This is the case even under the status quo’s stringent approach to predominance and commonality because class counsel is, if successful, compensated for shoehorning costs. Heavy Hand Model Plaintiff 2 (Should Be Included) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non-Lodestar Marginal Costs (i.e., Notice) (Y) Db0 (900) X2 (500) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Plaintiff Db0 (.7) Db0 (.7)-X2 - Y (Db0-X2-Y) (.7) Db0-(4)X2 Counsel Db0 (.3)- X2-Y Db0 (.3)- X2+X2-Y+Y (Db0-X2-Y) (.3) (4)X2-Y Result Applied 900(.3)- 500-1 =-201 900(.3) -500 +500 -1 + 1 =270 (900-500-1)(.3) =194.7 4(500)-1 =1,999 Include in the Class? No Include Yes Include Yes Include Yes Include Heavy Hand Model Plaintiff 1 (Prototypical) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non-Lodestar Marginal Costs (i.e., Notice) (Y) Da0 (1,000) X1 (0) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Plaintiff Da0 (.7) Da0 (.7) -X1- Y (Da0-X1 -Y) (.7) Da0-(4)X1 Counsel Da0 (.3)- X1-Y Da0 (.3)- X1+X1-Y+Y (Da0-X1-Y) (.3) (4)X1-Y Result Applied 1,000(.3) -1 =299 1,000(.3) -1+1 =300 (1,000-1) (.3) =299.7 4(0)-1 =-1 Include in the Class? Yes Include Yes Include Yes Include No Include PARKINSON – FINAL 814 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 Third, the lodestar-crosscheck model yields worrisome results. By compensating attorneys for litigation costs, this fee structure perversely incentivizes class counsel to add the exactly wrong marginal claims to their class. A considerable body of literature faults the lodestar-crosscheck model for encouraging class counsel to “pad the bills” by running up disingenuous litigation costs that do not, in fact, benefit the class.278 This, in turn, requires courts to incur considerable administrative costs in reviewing class counsel’s expense reports to ferret out bill padding.279 The above model suggests a secondary, but equally insidious, effect of the lodestar-crosscheck model: attorneys are encouraged to seek out dissimilar claims and expend considerable resources shoehorning those claims into class actions at the risk of every other members’ interests.280 By expending a suboptimal (excessive) amount on shoehorning costs, class counsel both increases the probability that a dissimilar claim is included in a certified class, harming defendants, and extracts greater wealth from the remaining meritorious claims, harming plaintiffs. Fourth, and most importantly, the modified—and heavy- handed—percent-of-the-fund model that does not 278 See Howard M. Downs, Federal Class Actions: Diminished Protection for the Class and the Case for Reform, 73 NEB. L. REV. 646, 667 (1994) (observing that the lodestar-crosscheck model encourages attorneys to “pad their hours and otherwise engage in unethical activities to enhance their fees”); Report of the Third Circuit Task Force, supra note 47, at 255. This observation has not escaped the attention of courts. See, e.g., Goldberger v. Integrated Res., Inc., 209 F.3d 43, 48 (2d Cir. 2000) (noting that “district courts found that it created a temptation for lawyers to run up the number of hours for which they could be paid”). 279 See Macey & Miller, supra note 26, at 4; 2 MCLAUGHLIN, supra note 45, § 6:24 (lamenting that the lodestar-crosscheck model involves “an inevitable waste of judicial resources” because it requires “district courts to engage in a gimlet-eyed review of line-item fee audits”) (quoting Goldberger, 209 F.3d at 49). 280 Note that, under the lodestar-crosscheck model, “the attorney can recover for time spent on issues on which he did not prevail.” ARNOLD S. JACOBS, 5E DISCLOSURE & REMEDIES UNDER THE SECURITIES LAWS § 21:77 (2015). PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 815 compensate class counsel for costs incurred yields partially favorable results. Unlike every other model, Plaintiff 3 is excluded when this fee structure is adopted, despite the fact that Plaintiff 3 ostensibly represents a greater marginal benefit to class counsel ($500) than Plaintiff 2 ($400). To state the obvious, costs matter. However, because of the combined shoehorning and litigation costs that Plaintiff 2 introduces, this fee structure leads class counsel to errantly exclude his claim. Most importantly, every fee structure except the heavy-handed approach reaches the wrong result for the entirely dissimilar claim of Plaintiff 3. Next, consider what happens when courts exchange external deference for internal scrutiny. At a high level, shoehorning costs are lessened, but litigation costs remain. Under the firm framework’s light-touch approach to the commonality and predominance requirements, courts will not decertify class actions that include subtly dissimilar claims. In turn, shoehorning costs decline. That said, class counsel must still incur marginal litigation costs—discovery must yield evidence to support each claim species, individual defenses must be rebutted, and the fact-finder (or defendant in settlement negotiations) must be persuaded that each claim species is meritorious. Assume a significant change in commonality and predominance doctrine in line with the firm framework’s prescriptions. Specifically, for this stylized example, doctrinal change halves net costs by drastically reducing shoehorning costs, yielding the following results: PARKINSON – FINAL 816 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 Light Touch Model Plaintiff 3 (Should Not Be Included) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non- Lodestar Marginal Costs (i.e., Notice) (Y) Dc0 (2,000) X3 (750) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Plaintiff Dc0 (.7) Dc0 (.7)-X3 -Y (Dc0-X3-Y) (.7) Dc0-(4)X3 Counsel Dc0 (.3)- X3 - Y Dc0 (.3)- X3+X3-Y + Y (Dc0-X3-Y) (.3) (4)X3-Y Result Applied 2,000(.3)- 750-1 =-151 2,000(.3)- 750 +750 -1 + 1 =600 (2,000-750-1) (.3) =374.7 4(750)-1 =2,999 Include in the Class? No Include Yes Include Yes Include Yes Include Light Touch Model Plaintiff 2 (Should Be Included) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non- Lodestar Marginal Costs (i.e., Notice) (Y) Db0 (900) X2 (250) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Db 0 (.7) Db0 (.7) Db0 (.7)-X2 -Y (Db0-X2-Y) (.7) Db0-(4)X2 Db 0 (.3)-X2 - Y Db0 (.3)- X2 - Y Db0 (.3)- X2+X2-Y + Y (Db0-X2-Y) (.3) (4)X2-Y Result 900(.3)-250- 1 =19 900(.3)- 250-1 =19 900(.3) -250 +250 -1 + 1 =270 (900-250-1)(.3) =194.7 4(250)-1 =999 Yes Include Yes Include Yes Include Yes Include Yes Include TABLE 3: DECREASE IN SHOEHORNING COSTS (PROPORTIONAL PER PLAINTIFF) PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 817 Table 3 shows that applying a light touch to the commonality and predominance requirements results in no change for three of the four fee structures. The heavy-handed percent-of-the-fund model, however, yields favorable results across all three plaintiffs. It is the only fee structure that achieves this ideal result. When the court approaches a class action’s external boundaries with deference, the attorneys operating under this scrutinizing fee structure self- interestedly exclude Plaintiff 3 from their class but include Plaintiffs 1 and 2. Functionally, by exchanging external deference for internal scrutiny, the court has employed class counsel as its agent for making the “right” decisions.281 281 See Samuel Issacharoff, Class Action Conflicts, 30 U.C. DAVIS L. REV. 805, 829 (1997) (“Although every system of principal-agent relations is fraught with difficulty, the best arrangement is one in which the attorneys function as partners of the class. The attorneys’ recovery should be tied to that of the class; to the extent the attorneys hope to prosper in the representation, that reward should be a direct product of what they return to the class. The optimal mechanism for creating this partnership is to establish a quasi-contractual relationship at the beginning of litigation in which the attorneys are provisionally awarded a percentage of the class’s recovery should they prevail.”). Light Touch Model Plaintiff 1 (Prototypical) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non- Lodestar Marginal Costs (i.e., Notice) (Y) Da0 (1,000) X1 (0) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Plaintiff Da0 (.7) Da0 (.7) -X1 - Y (Da0-X1 -Y) (.7) Da0-(4)X1 Counsel Da0 (.3)- X1 - Y Da0 (.3)- X1 + X1-Y + Y (Da0-X1-Y) (.3) (4)X1-Y Result Applied 1,000(.3) -1 =299 1,000(.3) -1 + 1 =300 (1,000-1) (.3) =299.7 4(0)-1 =-1 Include in the Class? Yes Include Yes Include Yes Include No Include PARKINSON – FINAL 818 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 This result holds when one relaxes the assumption that the decrease in commonality- and predominance-related costs is proportionate across each marginal plaintiff, even with a fixed decrease in shoehorning cost: Light Touch Model Plaintiff 3 (Should Not Be Included) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non- Lodestar Marginal Costs (i.e., Notice) (Y) Dc0 (2,000) X3 (1,250) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Plaintiff Dc0 (.7) Dc0 (.7)-X3 -Y (Dc0-X3-Y) (.7) Dc0-(4)X3 Counsel Dc0 (.3)-X3 - Y Dc0 (.3)- X3+X3-Y + Y (Dc0-X3-Y) (.3) (4)X3-Y Result Applied 2,000(.3)- 1,250-1 =-651 2,000(.3)- 1,250 +1,250 -1 + 1 =600 (2,000-1,250-1) (.3) =224.7 4(1,250)-1 =4,999 Include in the Class? No Include Yes Include Yes Include Yes Include Light Touch Model Plaintiff 2 (Should Be Included) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non- Lodestar Marginal Costs (i.e., Notice) (Y) Db0 (900) X2 (250) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Plaintiff Db0 (.7) Db0 (.7)-X2 -Y (Db0-X2-Y) (.7) Db0-(4)X2 Counsel Db0 (.3)-X2 - Y Db0 (.3)- X2+X2-Y + Y (Db0-X2-Y) (.3) (4)X2-Y Result Applied 900(.3)-250-1 =19 900(.3) -250 +250 -1 + 1 =270 (900-250-1)(.3) =194.7 4(250)-1 =999 Include in the Class? Yes Include Yes Include Yes Include Yes Include TABLE 4: DECREASE IN SHOEHORNING COSTS (FIXED ACROSS PLAINTIFFS) PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 819 This illustrates one cost of extant doctrine. When a court aggressively regulates the external boundaries of a class action, it not only engages in a costly activity for which its core competencies are ill-suited, but it also contributes little to the goal of keeping the “wrong” claims out of class actions. Its regulation is often redundant with the self-regulating internalization that class counsel would achieve if doctrine cohered to the firm framework’s prescriptions. Because the decision to include a marginal claim is a function of both the promised reward and the net cost of litigation, amplifying shoehorning costs while simultaneously scrutinizing compensation yields redundancy with respect to the “wrong” plaintiffs and perverse under-inclusion with respect to the “right” ones. Moreover, a light-touch approach to the commonality and predominance requirements does not result in a uniform marginal cost of zero across all claim species— litigation costs endure. The approach advised by the above model is no panacea. The example is highly stylized and one can certainly stylize scenarios in which class counsel includes a suboptimal claim in a class action. However, the impact of these contrarian Light Touch Model Plaintiff 1 (Prototypical) Damages Owed (Dx0) Costs of Litigating (Xn) Post-Litigation, Non- Lodestar Marginal Costs (i.e., Notice) (Y) Da0 (1,000) X1 (0) Y (1) Fee Structure Attorney Reimbursement (E) Percent of Fund E (30%) Percent + Cost E (30%) Percent Net of Cost E (30%) Lodestar Check E (4x) Award Plaintiff Da0 (.7) Da0 (.7) -X1 - Y (Da0-X1 -Y) (.7) Da0-(4)X1 Counsel Da0 (.3)- X1 - Y Da0 (.3)- X1 + X1-Y + Y (Da0-X1-Y) (.3) (4)X1-Y Result Applied 1,000(.3) -1 =299 1,000(.3) -1 + 1 =300 (1,000-1) (.3) =299.7 4(0)-1 =-1 Include in the Class? Yes Include Yes Include Yes Include No Include PARKINSON – FINAL 820 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 outcomes should not be overstated. It is difficult to imagine claims that would be less costly to litigate individually than in a class action. Moreover, alternative (suboptimal) scenarios are no different from the result that would be realized in the status quo. Within the firm framework, per the above model, class counsel are marginally less likely to make a suboptimal decision when they are required to internalize costs.282 One might object that this approach is half complete, failing to account for costs imposed on the defendant. After all, were class counsel to assemble a class action that optimized litigation-market costs, there would be little reason for the rational defendant to object to certification. It too would realize the administrative-efficiency gains to be had by aggregating what would otherwise be many individual claims into a single proceeding. That objection, however, elides the dominating role of a unique cost that operates only on defendants: liability costs. It is almost certainly the case that absent such costs defendants would generally prefer aggregation to the prospect of litigating thousands, or even millions, of individual trials. However, just as the behavior of class counsel is shaped by litigation and shoehorning costs, the behavior of defendants is shaped 282 It might be argued that, as such, the firm framework will have a perverse chilling effect on class actions. For example, cost-internalization may reduce the pool of sufficiently capitalized attorneys capable of pursuing class actions. This barriers-to-entry argument has been roundly dismissed in the market context. See, e.g., BORK, supra note 229, at 310– 29. Moreover, in the specific context of class actions, there are multiple solutions to this problem, including cost pooling, see Howard M. Erichson, The End of the Defendant Advantage In Tobacco Litigation, 26 WM. & MARY ENVTL. L. & POL’Y REV. 123, 131 (2001), and litigation financing, see generally Tyler W. Hill, Note, Financing the Class: Strengthening the Class Action Through Third-Party Investment, 125 YALE L. J. 484 (2015). Finally, it is not clear who stands to lose in a class-action market that, like most markets, requires cost internalization—other than under-capitalized attorneys. From class members’ perspective, little should change. Indeed, the very premise of the firm framework is that quality class actions will be brought because the expected value exceeds the (admittedly burdensome) ex ante costs. PARKINSON – FINAL No. 3:740] CLASS ACTIONS AS FIRMS 821 by litigation and liability costs. There are at least two liability costs that lead defendants to oppose certification, despite the prospect of administrative-efficiency gains. First, oftentimes claims pursued in a class action are unlikely to be brought as individual claims. Second, as behavioral-science studies indicate, aggregation increases the risk that a fact- finder will find a defendant liable.283 Returning to the stylized example with which this inquiry began, courts can compel self-interested attorneys to engage in market mimicking behavior without overly scrutinizing a class action’s external boundaries. When the complete range of the firm framework’s prescriptions are adopted in tandem—both external deference and heavy-handed internal scrutiny—class counsel do not need to be told by the court that they should not include an unrelated claim in their class action. VI. CONCLUSION Class action law is riddled with dualities. Rule 23, which might be read to invite a unifying framework,284 both permits class actions as an exception to the common-law prohibition on aggregate litigation and restrains that exception in important ways.285 Class actions themselves can be viewed as serving a multitude of unique, and at times competing, 283 See generally Irwin A. Horowitz & Kenneth S. Bordens, The Effects of Outlier Presence, Plaintiff Population Size, and Aggregation of Plaintiffs on Simulated Civil Jury Decisions, 12 L. & HUMAN BEHAV. 209 (1988). 284 Consider the Rule 23 superiority requirement: “A class action may be maintained if Rule 23(a) is satisfied and if: . . . the court finds . . . that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” FED. R. CIV. P. 23(b)(3); see also Betson & Tidmarsh, supra note 96, at 567–68 (noting that, “[i]n theory, optimal class size could be a relevant variable in determining whether the superiority prong in Rule 23(b)(3) opt-out class actions has been met”); id. at 573 (stating that “analysis of the superiority of a particular class action to other aggregation devices—an analysis that must consider the optimal size of the class—is a better way to ensure optimal aggregation than the indirect method of relying on opt-outs”). 285 See generally, e.g., Hansberry v. Lee, 311 U.S. 32 (1940). PARKINSON – FINAL 822 COLUMBIA BUSINESS LAW REVIEW [Vol. 2016 policy goals: ensuring that all plaintiffs, no matter how comparatively small their claims, have their rights vindicated in court;286 preserving the time, energy, and effort of resource-strapped courts;287 and deterring bad actors.288 There is, however, a consistency to this tension—a consistency defined by cost-benefit analysis. A unified framework might aid in balancing inputs and translating outputs into doctrinal prescriptions. Of late, the preferred framework has been to apply Rule 23 like a hammer, crushing all principal-agent nails that appear in its path. Some of these nails are real; others less so. The principal- agent framework is half right, and so should govern half of Rule 23. The firm framework provides a methodology for gaining purchase on which issues are properly defined as principal-agent problems that demand regulation and which issues are more likely to prove self-regulating. 286 See, e.g., Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (noting that “the Advisory Committee had dominantly in mind vindication of ‘the rights of groups of people who individually would be without effective strength to bring their opponents into court at all’” (quoting Benjamin Kaplan, A Prefatory Note, 10 B.C. L. Rev. 497, 497 (1969))). 287 See, e.g., Califano v. Yamasaki, 442 U.S. 682, 701 (1979) (noting that the “class-action device saves the resources of both the courts and the parties by permitting an issue potentially affecting [many parties] to be litigated in an economical fashion under Rule 23”). 288 See, e.g., Patrick A. Luff, Bad Bargains: The Mistake of Allowing Cost-Benefit Analyses in Class Action Certification Decisions, 41 U. MEM. L. REV. 65, 74 (2010) (“Class actions, in addition to serving individual needs, also serve the ‘public good’ as a deterrent of injury-causing behavior. Class actions deter defendant-wrongdoers by forcing those defendant-wrongdoers to internalize the social costs of their actions.”); Bone & Evans, supra note 214, at 1309 (“The main purpose of the small- claim class action is deterrence, not compensation.”).