Twombly, pleading rules, and the regulation of Court access

Iowa Law Review

Volumn 94, Issue 3, 2009, Pages 873-936

  Bone, Robert G ^a,b  

^a Department of Law ^*

^b BOSTON UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 67650137170     PISSN: 00210552     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Review

References (338)

1. Tellabs, Inc. v. Makor Issues & Rights, Ltd., 127 S. Ct. 2499 (2007) (interpreting the pleading standard for securities fraud claims in the Private Securities Litigation Reform Act ("PSLRA"))
2. Erickson v. Pardus, 127 S. Ct. 2197 (2007) (per curiam) (interpreting the pleading standard in FED. R. CIV. P. 8 (a) (2))
3. Bell Ati. Corp. v. Twombly, 127 S. \Ct. 1955 (2007) (same).
4. Iqbal v. Hasty, 490 F.3d 143 (2d Cir. 2007), cert, granted sub nom. Ashcroft v. Iqbal, 128 S. Ct. 2931 (2008). The Court heard oral argument in the Iqbalcase on December 10, 2008.
5. Twombly, 127 S. Ct. at 1964-65.
6. Richard L. Marcus, The Revival of Fact Pleading Under the Federal Rules of Civil Procedure, 86 COLUM. L. REV. 433, 439 (1986) (coining the term "liberal ethos"). For examples of jurists reading Twombly as a watershed opinion effecting a major change in pleading practice, see Twombly, 127 S. Ct. at 1979, 1983 (Stevens, J., dissenting) (referring to the majority's decision as "rewrit[ing] the Nation's civil procedure textbooks")
7. Kendall v. Visa U. S. A., Inc., 518 F.3d 1042, 1047 n. 5 (9th Cir. 2008)
8. Iqbal, 490 F.3d at 155
9. A. Benjamin Spencer, Plausibility Pleading, 49 B. C. L. REV. 431, 433, 446-47 (2008)
10. Scott Dodson, Pleading Standards After Bell Atlantic Corp. v. Twombly, 93 VA. L. REV. IN BRIEF 135, 137-41 (2007), http://www.virginialawreview.org/inbrief/2007/07/09/dodson.pdf; Michael C. Dorf, The Supreme Court Wreaks Havoc in the Lower Federal Courts-Again, FINDLAW'S WRIT, Aug. 13, 2007, http://writ.news. findlaw.com/dorf/20070813.html.
11. See generally Lonny S. Hoffman, Burn Up the Chaff with Unquenchable Fire: What Two Doctrinal Intersections Can Teach Us About Judicial Power over Pleadings, 88 B. U. L. REV. 1217, 1235 (2008) (noting that "a majority of scholars" view Twombly as a "sea of change in the traditional pleading standard the Court has followed since Conley "). For a contrary view, see generally Allan Ides, Bell Adantic and the Principle of Substantive Sufficiency Under Federal Rule of Civil Procedure 8 (a) (2) : Toward a Structured Approach to Federal Pleading Practice, 243 F. R. D. 604 (2008) (concluding that Twombly was correcdy decided)
12. Douglas G. Smith, The Twombly Revolution?, 36 PEPP. L. REV. (forthcoming 2009), available at http://papers.ssrn.com/so13/papers.cftn?abstract-id=1147862 (arguing that Twombly's plausibility standard is consistent with notice pleading).
13. In his sharp Twombly dissent, Justice Stevens hints at the impact he believes the decision will have on the distribution of power. Twombly, 127 S. Ct. at 1989 (Stevens, J., dissenting) ("The transparent policy concern that drives the decision is the interest in protecting antitrust defendants-who in this case are some of the wealthiest corporations in our economy-from die burdens of pretrial discovery.").
14. See, e.g., Phillips v. County of Allegheny, 515 F.3d 224, 230 (3d Cir. 2008) ("Few issues in civil procedure jurisprudence are more significant than pleading standards, which are the key that opens access to courts."). Of course, concerns about court access extend well beyond the civil procedure examples mentioned in the text. In the 1970s, for example, a debate raged over the constitutionality of filing fees for certain types of cases. See, e.g., LAURENCE H. TRIBE, AMERICAN CONSTITUTIONAL LAW 1008-10 (1978)
15. Frank I. Michelman, The Supreme Court and Litigation Access Fees: The Bight to Protect One's Rights (pt. 1), 1973 DUKE L. J. 1153, 1165. Also, the Supreme Court's standing cases have evoked heated controversy for decades, with the critics complaining that restrictive standing doctrines impede socially desirable forms of public law litigation. See ERWIN CHEMERINSKY, CONSTITUTIONAL LAW: PRINCIPLES AND POLICIES 60-66 (3d ed. 2006) (discussing standing doctrine).
16. See N. Y. STATE COMM'RS ON PRACTICE & PLEADINGS, FIRST REPORT TO THE NEW YORK LEGISLATURE 75-76 (Albany 1848) (criticizing common law pleading)
17. CHARLES E. CLARK, HANDBOOK OF THE LAW OF CODE PLEADING 150-79 (1928) [hereinafter CODE PLEADING] (describing and criticizing code pleading rules). See generally Robert G. Bone, Mapping the Boundaries of a Dispute: Conceptions of Ideal Lawsuit Structure from the Field Code to the Federal Rules, 89 COLUM. L. REV. 1, 19-22, 101 n. 345 (1989) (describing the history and development of pleading rules).
18. Rule 11 of the Federal Rules of Civil Procedure imposes penalties on parties and their lawyers who file meridess suits either intentionally or without doing a reasonable pre-filing investigation. See 5A CHARLES ALAN WRIGHT & ARTHUR R. MILLER, FEDERAL PRACTICE AND PROCEDURE § 1331, at 470-74, § 1332, at 480-95 (3d ed. 2004) (discussing the 1983 amendments to Rule 11 and the accompanying controversy). Heightened or strict pleading requirements force plaintiffs (or defendants in an answer) to allege more factual detail in support of the elements which they must plead. See 15 U. S. C. § 78u-4 (b) (2000) (codifying the Private Securities Litigation Reform Act's strict pleading requirement)
19. Christopher M. Fairman, The Myth of Notice Pleading, 45 ARIZ. L. REV. 987, 994-97 (2003) (describing the development of judicially created strict pleading standards).
20. The Rules Enabling Act vests the authority to make procedural rules for the federal courts in the U. S. Supreme Court and creates a multi-stage rulemaking process with opportunities for public input. 28 U. S. C. § § 2072-2073 (2000). A proposed amendment to an existing Federal Rule of Civil Procedure or a new Rule is first considered by the Advisory Committee on Civil Rules. If the proposal meets with the Advisory Committee's approval, it is forwarded to the Standing Committee and then to the Judicial Conference, and eventually to the U. S. Supreme Court. If the Supreme Court approves the proposal, it is forwarded to Congress, which has the opportunity to veto it. If Congress does not exercise its veto within the prescribed period, the proposal goes into effect. See Robert G. Bone, The Process of Making Process: Court Rulemaking, Democratic Legitimacy, and Procedural Efficacy, 87 GEO. L. J. 887, 892 (1999) (describing the amendment process).
21. U. S. C. § 2072 (b). All types of fee-shifting involve one party paying all or a portion of the opposing party's attorney fees. The obligation to pay fees can be triggered by different conditions, such as a trial loss, failure to improve on a settlement offer at trial, or filing a meritless suit.
22. Hoffman, supra note 4, at 1222
23. see also Ettie Ward, The After-Shocks of Twombly: Will We "Notice" Pleading Changes?, 82 ST. JOHN'S L. REV. 893, 893 (2008) (counting more than 9400 citations of Twombly as of March 15, 2008).
24. Iqbal v. Hasty, 490 F.3d 143 (2d Cir. 2007), cert, granted sub nom. Ashcroft v. Iqbal, 128 S. Ct. 2931 (2008). The certified questions also raise substantive issues relating to the scope of a Bivens action against federal officials, so it is not certain that the Court will address the pleading issue.
25. See COMM. ON RULES OF PRACTICE & PROCEDURE, MEETING OF JANUARY 14-15, 2008, at 37-44 (2008), available at http://www.uscourts.gov/rules/Minutes/ST01-2008-min. pdf.
26. The AALS Civil Procedure Section chose a topic for its January 2009 annual section meeting, "The Changing Shape of Federal Pretrial Practice, " which was inspired in part by the Twombly decision. Ass'N OF AM. LAW SCH. 2009 ANNUAL MEETING PROGRAM 72 (2009), available at http://www.aals.org/documents/ 2009am/program.pdf (noting the impact of Twombly).
27. See, e.g., Spencer, supra note 4, at 431-33, 445-46 (arguing that Twombly breaks from the intent of the Rule 8 drafters).
28. Charles E. Clark, The Handmaid of Justice, 23 WASH. U. L. Q. 297, 304 (1938).
29. Bell Atl. Corp. v. Twombly, 127 S. Ct. 1955, 1962 (2007).
30. This history is taken from the summary in the Twombly opinion. Id. at 1961-62.
31. Id. at 1961.
32. Id.
33. Id.
34. Twombly, 127 S. Ct. at 1967.
35. Id. at 1962.
36. Id.
37. Id. at 1962-63 (quoting from the complaint).
38. See id. at 1964 (noting that section 1 requires more than parallel conduct).
39. Twombly v. Bell Atl. Corp., 313 F. Supp. 2d 174, 174 (S. D. N. Y. 2003).
40. Twomblyv. Bell Atl. Corp., 425 F.3d 99, 118-19 (2d Cir. 2005).
41. Twombly, 127 S. Ct. at 1955, 1963, 1974.
42. Id. at 1970-74.
43. See Hoffman, supra note 4, at 1225 ("The majority view among academics has been that robust efforts to regulate at the pleading stage are wrongheaded and inconsistent with the traditional pleading standard the Court has followed since Conley. ").
44. Conley v. Gibson, 355 U. S. 41 (1957).
45. Id. at 47.
46. Id.
47. Id. at 45-46.
48. Bell Atl. Corp. v. Twombly, 127 S. Ct. 1955, 1968 (2007) ("This 'no set of facts' language can be read in isolation as saying that any statement revealing the theory of the claim will suffice unless its factual impossibility may be shown from the face of the pleadings." (quoting Conley, 355 U. S. at 45-46)).
49. Id. at 1964-65.
50. Specifically, the Court reasoned that Conley's broad formulation was inconsistent with Rule 8 (a) (2) 's requirement that the plaintiff provide not only a "short and plain statement of the claim" but also a statement "showing that the pleader is entided to relief." Id. at 1964, 1965 &n. 3.
51. See, e.g., Lewis v. Marriott Int'l, Inc., 527 F. Supp. 2d 422, 424 n. 1 (E. D. Pa. 2007) ("[T]he applicability of Twombly to non-Sherman Act claims is unknown. ")
52. The Supreme Court, 2006 Term-Leading Cases, 121 HARV. L. REV. 305, 310 n. 51 (2007) ("[S]ome scholars view Twombly as primarily an antitrust case.").
53. See, e.g., Williams v. Gerber Prods. Co., 523 F.3d 934, 939-40 (9th Cir. 2008) (applying the plausibility standard to tort and breach-of-warranty claims)
54. Perez-Acevedo v. Rivero-Cubano, 520 F.3d 26, 29 (1st Cir. 2008) (stating that a § 1983 claim must "raise right to relief above a speculative level")
55. Davis v. Coca-Cola Bottling Co. Consol., 516 F.3d 955, 974 n. 43 (11 th Cir. 2008) ("We understand Twombly as a further articulation of the standard by which to evaluate the sufficiency of all claims brought pursuant to Rule 8 (a).")
56. Phillips v. County of Allegheny, 515 F.3d 224, 234 (3d Cir. 2008) ("[W]e decline at tins point to read Twombly so narrowly as to limit its holding on plausibility to the antitrust context.")
57. Sonnier v. State Farm Mut. Auto. Ins. Co., 509 F.3d 673, 675-76 (5th Cir. 2007) (extending the plausibility standard to civil contract disputes)
58. see also Kendall W. Hannon, Note, Much Ado About Twombly? A Study on the Impact of Bell Adantic Corp. v. Twombly on 12
59. (b) (6) Motions, 83 NOTRE DAME L. REV. 1811, 1814-15, 1834-38 (2008) (finding that reported cases apply Twombly to a wide range of different types of claims).
60. See, e.g., Robbins v. Oklahoma, 519 F. Sd 1242, 1247 (10th Cir. 2008) (calling the new formulation "less than pellucid")
61. Phillips, 515 F. Sd at 230 (describing Twombly's impact as "confusing")
62. Iqbal v. Hasty, 490 F.3d 143, 155, 157 (2d Cir. 2007) (noting the Supreme Court's "conflicting signals" and the "considerable uncertainty" created by Twombly).
63. The Court does make clear that it is not imposing a " probability" standard, perhaps suggesting diat plausible lies somewhere between possible and probable. See Twombly, 127 S. Ct. at 1965 ("Asking for plausible grounds to infer an agreement does not impose a probability requirement."). It is also clear that the Court does not mean to impose a standard as strict as the "strong inference" that the Private Securities Litigation Reform Act requires for scienter allegations in a securities fraud claim. 15 U. S. C. § 74u-4 (b) (2) (2000)
64. see Tellabs, Inc. v. Makor Issues & Rights, Ltd., 127 S. Ct. 2499, 2510 (2007) (holding that the PSLRA requires the plaintiff to allege sufficient facts so that "a reasonable person would deem the inference of scienter cogent and at least as compelling as any opposing inference one could draw from the facts alleged").
65. See Conley v. Gibson, 355 U. S. 41, 47 (1957) (holding that Rule 8 (a) requires that the complaint "give the defendant fair notice of what the plaintiff's claim is and the grounds upon which it rests")
66. 5A WRIGHT & MILLER, supra note 8, § 1215 (stating that Rule 8 (a) (2) "calls for simple, concise, and direct averments, authorizes alternative and hypothetical pleading, and eliminates any requirement of consistency of statement")
67. id. § 1216 ("[T]he pleading must contain something more by way of a claim for relief than a bare averment that the pleader wants compensation and is entitled to it."). As Justice Stevens in his sharp Twombly dissent recognized: [H]ad the amended complaint in this case alleged only parallel conduct, it would not have made the required "showing." Similarly, had the pleadings contained only an allegation of agreement, without specifying the nature or object of that agreement, they would have been susceptible to the charge that they did not provide sufficient notice. Twombly, 127 S. Ct. at 1979 n. 6 (Stevens, J., dissenting) (internal citation omitted).
68. See Iqbal, 490 F.3d at 157 (noting the "conflicting signals").
69. Conley, 355 U. S. at 45-46.
70. Twombly, 127 S. Ct. at 1973 n. 14 (noting that the creation of heightened pleading standards is for the formal rulemaking process or for Congress).
71. After the recent style amendments, Form 9-a model complaint for automobile negligence-appears as Form 11 and the specific date and location references in the original have been replaced with placeholders. FED. R. CIV. P. Form 11.
72. Erickson v. Pardus, 127 S. Ct. 2197, 2200 (2007) (per curiam). The lower courts held that Erickson had not adequately pleaded his Eighth Amendment claim, but the Supreme Court reversed. Id. at 2199-200. For more on this case, see infra note 68.
73. Pardus, 127 S. Ct. at 2198, 2200.
74. Id. at 2200 (emphasizing that "specific facts are not necessary").
75. Twombly, 127 S. Ct. at 1966-67 (arguing that the plausibility requirement is needed to avoid high discovery costs and to weed out meridess suits before the threat of these high costs pressures defendants to settle). It is easy to miss this important distinction if one focuses only on doctrine and the linguistic shift to "plausibility." Indeed, some courts and commentators have tried to subsume the "plausibility" standard under notice pleading by arguing that the Court merely specified what was required to give the defendant fair notice. See, e.g., Phillips v. County of Allegheny, 515 F.3d 224, 233-34 (3d Cir. 2008) (linking together "plausibility, " a Rule 8 "showing, " and notice)
76. Airborne Beepers & Video, Inc. v. AT&T Mobility LLC, 499 F.3d 663, 667 (7th Cir. 2007) ("[A]t some point the factual detail in a complaint may be so sketchy that the complaint does not provide the type of notice of the claim to which the defendant is entided under Rule 8.")
77. Smith, supra note 4, at 24 (arguing that Twombly simply sets out the requirements for adequate notice: "It is only where the complaint states a logically coherent theory of liability that a defendant is truly 'on notice' of the claim against it."). This effort might reflect an implicit realization that pleading standards are about fairness to defendants, not just system efficacy, fairness to plaintiffs, or efficiency. Infra notes 137-66 and accompanying text. Whether or not Twombly's plausibility standard can be justified in fairness terms, however, the Court's opinion clearly mentions case screening as an appropriate pleading function. See generally Marcus, supra note 4, at 452 (observing that the traditional notice-giving function requires very little by way of allegations when the defendant has access to broad discovery).
78. Twombly, 127 S. Ct at 1974.
79. Id. at 1967 (noting that the class consists of at least 90% of all subscribers in the continental United States).
80. Id.
81. Id. The named defendants were Verizon Communications Inc., AT&T Inc., Qwest Communications International, Inc., and Bellsouth Corp.
82. Infra notes 196-207 and accompanying text. The fact that defendants have a great deal of discoverable information and plaintiffs have very little gives plaintiffs settlement leverage, especially when information about the merits is asymmetrically distributed. In addition, the class action aspect coupled with treble damages creates a potential award sizable enough to support a credible trial threat even for a meritless suit, assuming a large enough error risk at trial.
83. Twombly, 127 S. Ct. at 1966-67.
84. Id.
85. Richard A. Epstein, Bell Atlantic v. Twombly: How Motions to Dismiss Become (Disguised) Summary Judgments, 25 WASH. U. J. L. &POL'Y 61, 81-92 (2007).
86. Twombly, 127 S. Ct. at 1970. Some readers might object to my calling this allegation "conclusory." In deciding a Rule 12 (b) (6) motion to dismiss, the rule is that the judge must assume that all plaintiff's allegations are true except those that are conclusory. The question therefore is why the Twombly plaintiffs' allegations of agreement are conclusory. The Court's answer is that a "fair reading" of the complaint shows that the plaintiffs themselves intended the allegations to be conclusions based on other allegations in their complaint. See id. ("Although in form a few stray statements speak direcdy of agreement, on fair reading these are merely legal conclusions resting on the prior allegations."). Had the plaintiffs good reasons to suspect a conspiracy apart from the existence of parallel conduct, their allegations of agreement might have sufficed, although the plaintiffs might also have had to plead those additional reasons.
87. See id. at 1969 (remarking that "Mr. Micawber's optimism would be enough" to bring suit).
88. The Twombly Court makes a point of noting that the allegations might be enough to support agreement in a different market setting. Id. at 1972 ("In a traditionally unregulated industry with low barriers to entry, sparse competition among large firms dominating separate geographical segments of the market could very well signify illegal agreement...."). This different setting would be associated with a different baseline.
89. Id. (concluding that there is a "natural explanation" for the parallel conduct alleged, given the distinctive history of the telecommunications market).
90. See, e.g., Iqbal v. Hasty, 490 F.3d 143, 156 (2d Cir. 2007) (citing Twombly's approval of Form 9 as a conflicting signal)
91. Dodson, supra note 4, at 141-42 (noting tension between plausibility pleading and Form 9).
92. Supra note 47.
93. FED. R. CIV. P. Form 9, para. 2. The complaint also alleges that the plaintiff suffered a broken leg and other losses as a result. The restyled Form 11 omits the particular date and place references. Supra note 47.
94. To be sure, the plaintiff might have been negligent himself, but contributory negligence is for the defendant to allege.
95. The Court's subsequent holding in Erickson v. Pardus, 127 S. Ct. 2197 (2007), easily fits this framework. In Erickson, a pro se plaintiff alleged that prison officials wrongfully withheld his hepatitis C treatment and that he would suffer irreparable and life-threatening damage to his liver as a result, in violation of the Eighth Amendment's prohibition on cruel and unusual punishment. Id. at 2198-99. The defendants argued that the complaint failed to allege that the plaintiff would suffer "substantial harm" distinct from whatever harm he would have suffered anyway as a result of his condition. Id. at 2199. The sensible baseline, given the prison regulations entitling inmates to treatment, is that the plaintiff would receive treatment for his serious illness. Withholding treatment deviates significantly from this baseline and in a way that correlates with substantial harm. After all, people suffering from serious illnesses usually get much worse when they do not receive medical treatment. Therefore, it is certainly plausible based on these allegations that the defendants' decision to withhold treatment would cause substantial harm to the plaintiff. The Twombly Court also approved its previous decision in Swierkiewicz v. Sorema N. A., 534 U. S. 506 (2002). The plaintiff in Swierkiewicz alleged that the defendants discriminated against him on the basis of national origin and age. Id. at 509. The Supreme Court held that the plaintiff could allege intentional discrimination in any way he saw fit and did not have to plead the four elements of a trial prima facie case outlined in McDonnell Douglas Corp. v. Green, 411 U. S. 792 (1973). Swierkiewicz, 534 U. S. at 510-11. Even though the Twombly Court's basis for distinguishing Swierkiewicz is problematic, see Bell Ad. Corp. v. Twombly, 127 S. Ct. 1955, 1973-74 (2007), the result is clearly consistent with my baseline interpretation of plausibility. The plaintiff in Swierkiewicz "detailed the events leading up to his termination, provided relevant dates, and included the ages and nationalities of at least some of the relevant persons involved with his termination. " Swierkiewicz, 534 U. S. at 514. He also alleged that he was more qualified than the younger person of French origin who replaced him. Id. at 508. The baseline for this type of case is a qualified employee continuing to be employed unless the position is closed or upgraded to a level beyond the employee's qualifications. Thus, a termination under the circumstances alleged deviates from the baseline and does so in a way that suggests a discriminatory motive.
96. As Part IV. B explains, it is important to consider not only the likelihood that a suit is meritless as opposed to meritorious, but also the cost of allowing a meritless suit to proceed compared to the cost of screening a meritorious suit. In Twombly, the costs are substantial on both sides. If the allegation of agreement is, in fact, correct, the dismissal would frustrate efforts to restore competition in a major market (although there are other options, such as action by the Federal Trade Commission). By the same token, if the allegation of agreement is incorrect, allowing the suit to proceed forward and secure an unjustified settlement might also disrupt competition in the telecommunications market.
97. I am not suggesting that the Court meant its plausibility standard to apply only to complex cases with a high risk of costly meritless suits. Some have suggested as much, but such an interpretation fits the language of Twombly rather poorly. See, e.g., Smith, supra note 4, at 18-20 (collecting sources suggesting a limitation to complex cases and criticizing this approach as a poor reading of Twombly). The Court condemns Conley's "no set of facts" formulation for all lawsuits, not just for complex suits.
98. The plaintiff's description must not only differ from the baseline but also correlate with illegal conduct in more than a trivial way. Moreover, it is possible that the requisite degree of correlation might vary to a limited extent with the risks and costs of meridess litigation and the effectiveness of judicial case management for the particular type of suit.
99. See Charles B. Campbell, A "Plausible" Showing After Bell Atlantic Corp. v. Twombly, 9 NEV. L. J. 1, 23 (2008) (suggesting that the appropriate pleading standard under Rule 8 (a) (2) requires factual allegations reaching all material elements necessary to recover under substantive law)
100. Smith, supra note 4, at 23 (arguing that the "central theme of the plausibility standard is logical coherence, " which means that the complaint cannot have "logical holes" or omit "essential elements"; liability must be "a necessary consequence" of the allegations).
101. A. Benjamin Spencer, Deconstructing Pleading Doctrine 15-16 (Wash. & Lee Pub. Legal Studies Research Series, Paper No. 2008-41, 2008), available at http://ssrn.com/abstract= 1272074.
102. See id. ("[T]he facts presented [in Twombly] were equivocal and thus the presumption of propriety was not overcome.").
103. Spencer suggests that dismissal is appropriate when lawful and unlawful explanations are "equally possible, " which might mean that one should compare the relative strength of different inferences that the allegations support. However, he cannot mean that an inference of illegality must be more likely than an inference of legality. That standard is even stronger than the very strict pleading requirement of the Private Securities Litigation Reform Act. See Tellabs, Inc. v. Makor Issues & Rights, Ltd., 127 S. Ct. 2499, 2510 (2007) (holding that the inference of scienter must be "cogent" and "at least as compelling as any opposing inference").
104. Professor Richard Epstein offers an interpretation of Twombly that, while useful in some respects, also misses the mark. He argues that the result makes sense given the fact that the plaintiffs relied on "public information, easily assembled and widely available, that can be effectively rebutted by other public evidence." Epstein, supra note 59, at 67. I agree with Professor Epstein that the defendants' conduct appears perfectly competitive when all the public information about the telecommunications market is considered, but I disagree that this turns the motion to dismiss into a "mini-summary judgment." Id. at 81. My baseline interpretation of Twombly has the advantage of explaining how the Court's holding extends beyond antitrust cases and cases involving public agency records or other publicly available information. I do agree with Professor Epstein, however, that pleading standards should balance error costs, a point I explore in more detail infra Part IV.
105. Hannon, supra note 40, at 1815 (finding no statistically significant effect on dismissal rates after Twombly except in civil rights suits).
106. The Court distinguishes between "factually neutral" and "factually suggestive" allegations, a distinction that fits the idea of a legal baseline relative to which allegations could be neutral or suggestive. Twombly, 127 S. Ct. at 1966 n. 5. Thanks especially to Alexi Lahav for helpful comments on my baseline idea.
107. Id. at 1974.
108. Erickson v. Pardus, 127 S. Ct. 2197, 2200 (2007).
109. See, e.g., Hoffman, supra note 4, at 1258-59 (expressing concern about what lower courts might do with Twombly). Another possible concern, which is relevant mostly for antitrust cases, is that judges untutored in economic theory will misapply economics at the pleading stage. This is certainly a risk, but it is also a risk at summary judgment and other stages of the litigation where the judge has power to influence the outcome. And it is arguably an even more serious problem when the jury deliberates. If we tolerate this risk at later stages, it is not clear why we should not tolerate it at the pleading stage as well.
110. See, e.g., Cash Energy, Inc. v. Weiner, 768 F. Supp. 892, 897-900 (1991) (explaining the trend toward higher standards of particularity)
111. Marcus, supra note 4, at 444-51 (discussing the revival of fact pleading).
112. Leatherman v. Tarrant County Narcotics Intelligence & Coordination Unit, 507 U. S. 163 (1993).
113. Swierkiewicz v. Sorema N. A., 534 U. S. 506 (2002).
114. Both Leatherman and Swierkiewicz affirm Conley v. Gibsons liberal notice pleading standard and hold that the lower courts have no power to create heightened pleading rules on their own. Swierkiewicz, 534 U. S. at 512-14; Leatherman, 507 U. S. at 168-69.
115. See, e.g., Fairman, supra note 8, at 1011-59 (discussing pleading requirements in a variety of different substantive areas).
116. Id. at 1062-64. In fact, the Supreme Court itself sent mixed signals in its 1998 Crawford-El opinion and its 2005 Dura Pharmaceuticals opinion. Dura Pharms., Inc. v. Broudo, 544 U. S. 336, 347 (2005) (holding that Rule 8 (a) requires securities fraud class action plaintiffs to allege loss causation with greater particularity)
117. Crawford-El v. Britton, 523 U. S. 574, 598 (1998) (rejecting strict pleading for a § 1983 complaint, but suggesting that courts might use the motion for a more definite statement or a Rule 7 (a) reply to require more factual detail).
118. See, e.g., Spencer, supra note 4, at 434-36, 461, 479-80 (making these criticisms)
119. The Supreme Court, 2006 Term-Leading Cases, supra note 39, at 311-12 (criticizing Twombly for inconsistency with the original drafters' intent). But see Campbell, supra note 72, at 27-32 (arguing that a proper understanding of Rule 8 (a) and 1938 Advisory Committee intent is consistent with Twombly). Also, it is common for critics to describe Twombly's plausibility standard as a throwback to "fact pleading" under the codes, which is tantamount to criticizing it for inconsistency with the 1938 drafters' intent.
120. CODE PLEADING, supra note 7.
121. Michael E. Smith, Judge Charles E. Clark and the Federal Rules of Civil Procedure, 85 YALE L. J. 914, 917-18 (1976).
122. See 5A WRIGHT & MILLER, supra note 8, § 1202 (" [The system of common law pleading was] excruciatingly slow, expensive, and unworkable. The system was better calculated to vindicate highly technical rules of pleading than it was to dispense justice.").
123. See CODE PLEADING, supra note 7, at 12-15, 21-23 (critiquing the technicalities of common law pleading).
124. Most code pleading provisions were similar to the New York Field Code section, which required that the complaint contain "a plain and concise statement of the facts constituting each cause of action without unnecessary repetition. " Id. at 138.
125. Id. at 159. Clark gave an example of a bicyclist who was injured when she struck a hole in a negligently maintained street. The bicyclist's suit was dismissed because, while she alleged the hole and the failure to maintain the street, she stated only that her bicycle "struck said defective, unsafe, and out of repair street, " not that it struck the hole in the street. Id. at 158.
126. Charles E. Clark, The Complaint in Code Pleading, 35 YALE L. J. 259, 285 (1926) (noting that a complaint should be structured around a legal theory, for otherwise it would "be a meaningless jumble")
127. Smith, supra note 90 (noting that Clark demanded more than a conclusory allegation of negligence).
128. Clark, supra note 16, at 316; accord Charles E. Clark, Simplified Pleading, 2 F. R. D. 456, 460-61 (1942) (stating that the pleadings need only give notice "of the general nature of the case and the circumstances or events upon which it is based").
129. See Marcus, supra note 4, at 439-40.
130. As for the existence of an agreement, the defendants had the information to know whether that allegation was true and could easily deny it.
131. 5A WRIGHT & MILLER, supra note 8, § 1202, at 87.
132. See CODE PLEADING, supra note 7, at 29-30 (distinguishing "issue pleading" at common law, "fact pleading" under the codes, and a new approach, "notice pleading, " that "is now urged for general adoption")
133. id. at 163 (observing that "notice pleading" requires only "a very general reference to the happening out of which the case arose").
134. See Marcus, supra note 4, at 445 (describing judges engaged in "something bordering a revolt" over the existing rule)
135. Stephen N. Subrin, Haw Equity Conquered Common Law: The Federal Rules of Civil Procedure in Historical Perspective, 135 U. PA. L. REV. 909, 976-77, 983-84 (1987) (describing disagreement over liberal pleading among Advisory Committee members and resistance to it after 1938).
136. See AM. BAR ASS'N, RULES OF CIVIL PROCEDURE FOR THE DISTRICT COURTS OF THE UNITED STATES WITH NOTES AS PREPARED UNDER THE DIRECTION OF THE ADVISORY COMMITTEE AND PROCEEDINGS OF THE INSTITUTE ON FEDERAL RULES, CLEVELAND, OHIO, JULY 21, 22, 23, 1938, at 220 (1938). Clark stated: I think there is no question that the rules are based on the theory of a rather general form of pleading. Whether they go as far as I believe they do, you may, of course, question. I want to point out that in this field there is a considerable room for variations between courts. After all, we made a generalized statement in the rules, a short and simple statement along the lines in which we believed, and to one judge that may require much more than it does to others. Id. Clark also explained that a judge who believed more factual detail was desirable could grant a motion for a more definite statement under Rule 12 (e), although he counseled against excessive use of this device. Id. at 223.
137. Id. at 220
138. see also Smith, supra note 90, at 917.
139. At one point, however, the Court does suggest distinctions between "the conclusory and the factual" and between "the factually neutral and the factually suggestive, " but these distinctions are very different from the code classifications. Bell Atl. Corp. v. Twombly, 127 S. Ct. 1955, 1966 n. 5 (2007).
140. See, e.g., Dioguardi v. Durning, 139 F.2d 774, 775 (2d Cir. 1944) (interpreting Rule 8 (a) (2) very liberally)
141. Marcus, supra note 4, at 445 (noting that Clark rejected moving toward more specific pleadings).
142. Conley v. Gibson, 355 U. S. 41 (1957). The complaint in Conley stated lots of facts and would have satisfied any pleading standard. Justice Black wrote the opinion for the Court. As a former trial lawyer, he was a strong proponent of jury trials and disliked procedural niceties. It is very likely that his strong language endorsing notice pleading was meant to ensure that as many cases as possible made it to the jury. See Emily Sherwin, The Story of Conley: Precedent by Accident, in CIVIL PROCEDURE STORIES 295, 308 (Kevin M. Clermont ed., 2d ed. 2008).
143. As the Court also stated: [T]he Federal Rules of Civil Procedure do not require a claimant to set out in detail the facts upon which he bases his claim. To the contrary, all the Rules require is a "short and plain statement of the claim" that will give the defendant fair notice of what the plaintiff's claim is and the grounds upon which it rests. Conley, 355 U. S. at 47 (quoting FED. R. CIV. P. 8 (a) (2))
144. Richard L. Marcus, The Puzzling Persistence of Pleading Practice, 76 TEX. L. REV. 1749, 1750 (1998). That debate was to heat up again in the 1980s, when concerns about a litigation crisis in the federal courts grew more intense. See supra notes 82-87 and accompanying text.
145. Some might argue that events in 1955 indicate the Committee's intent to adopt the most liberal version of notice pleading for Rule 8 (a) (2), but that would be a mistake. In 1955, die Advisory Committee, with Charles Clark still acting as Reporter, rejected an effort, spearheaded by the Ninth Circuit Judicial Conference, to amend Rule 8 (a) (2) by adding the phrase "facts constituting a cause of action, " so that the amended Rule would read: " (2) a short and plain statement of the claim showing that the pleader is entitled to relief, which statement shall contain the facts constituting a cause of action. " Claim or Cause of Action: A Discussion on the Need for Amendment of Rule 8 (a) (2) of the Federal Rules of Civil Procedure, 13 F. R. D. 253, 253 (1952)
146. see ADVISORY COMM. ON RULES FOR CIVIL PROCEDURE, REPORT OF PROPOSED AMENDMENTS TO THE RULES OF CIVIL PROCEDURE FOR THE UNITED STATES DISTRICT COURTS 18-19 (1955), available at http://www.uscourts.gov/rules/Reports/CV10-1955. pdf. The proposal was triggered by growing concern about the high costs of complex antitrust litigation in the 1950s. The problem with using this decision to infer committee endorsement of notice pleading is that the Committee's two-page explanation does not support the conclusion. Nowhere does it expressly refer to "notice pleading" by name. Rather, it defends Rule 8 (a) (2) on the ground that it eliminates the technical skirmishing of code pleading by allowing only a general statement, yet still requires "the pleader to disclose adequate information as the basis for his claim for relief as distinguished from a bare averment that he wants relief and is entitled to it." ADVISORY COMM., supra, at 19. Responding to criticisms of Dioguardi, 139 F.2d at 774, the Committee stressed that the Dioguardi decision "was not based on any holding that a pleader is not required to supply information disclosing a ground for relief." ADVISORY COMM., supra, at 19.
147. This is especially true considering that plaintiffs are routinely given a chance to amend their complaints to meet the applicable pleading requirement. Professor Spencer makes much of the stare decisis constraint, likening interpretation of the Federal Rules to interpretation of a statute. See Spencer, supra note 4, at 461-63. However, the Federal Rules are not statutes and are generally interpreted more generously, even though the precise scope of interpretive flexibility is a matter of some dispute. See 4 WRIGHT & MILLER, supra note 8, § 1029, at 154 (stating that the Federal Rules of Civil Procedure are to be construed generously "to further the cause of justice" and noting that this liberal interpretive principle is endorsed explicitly by Rule 1). Compare Joseph P. Bauer, Schiavone: An Un -Fortune-ate Illustration of the Supreme Court's Role as Interpreter of the Federal Rules of Civil Procedure, 63 NOTRE DAME L. REV. 720, 720 (1988) (arguing that the Court's role is more expansive when interpreting a Federal Rule than when interpreting a statute), with Catherine T. Struve, The Paradox of Delegation: Interpreting the Federal Rules of Civil Procedure, 150 U. PA. L. REV. 1099, 1119 (2002) (criticizing judges and commentators who support a more generous approach to Federal Rule interpretation).
148. This vision is sometimes referred to as the "liberal ethos." Marcus, supra note 4, at 439.
149. For a detailed account of the 1938 drafters' pragmatism and the influence of legal realism, see Bone, supra note 7, at 78-98.
150. Id.
151. Roscoe Pound, Some Principles of Procedural Reform, 4 ILL. L. REV. 388, 394 (1909).
152. See, e.g., Bone, supra note 7, at 101 n. 345 (discussing Pound's and Clark's pragmatic views on pleading).
153. The importance of deciding cases on the facts and evidence is also closely related to the legal realist beliefs of Charles Clark and others on the Advisory Committee. See Bone, supra note 7, at 80-89 (describing Clark's realist views)
154. see also Emily Sherwin, The Jurisprudence of Pleading: Rights, Rules, and Conley v. Gibson, 52 HOW. L. J. 73, 78-86 (2008) (discussing the effect of legal realism on the early twentieth-century pleading debate). Legal realism was a reaction to the formalism of the late nineteenth century, and moderate realists like Clark were very influential during the 1930s and 1940s. Bone, supra note 7, at 85. They stressed the importance of fitting law to social practice rather than deducing law from abstract concepts and principles. Id. Over the late nineteenth century, the idea of a "cause of action" became bound up with formalistic distinctions among primary rights, primary duties, and remedies-abstractions that the realists rejected as meaningless. Id. They redefined the cause of action in terms of operative facts rather than abstract rights. Id. In addition, the early twentieth century was a period of dynamic legal change, as the rise of sociological jurisprudence, and then legal realism, produced a more functional, policyoriented approach to common law rules and principles, and as the New Deal generated novel statutory claims. See Bone, supra note 7, at 78-98 (discussing the effect of legal realism and pragmatism on the early twentieth-century procedural reform movement)
155. Sherwin, supra (describing the debate between legal realists and legal formalists in the early twentieth century and its impact on the choice of federal pleading rules). Forcing a plaintiff to identify a cause of action and plead it precisely once and for all might have impeded the dynamic growth of the law. In any event, it must have made little sense to lock a plaintiff into a cause of action when discovery might reveal facts supporting something quite different. See Bone, supra note 7, at 101 n. 345
156. Sherwin, supra, at 83.
157. Judith Resnik, Failing Faith: Adjudicatory Procedure in Decline, 53 U. CHI. L. REV. 494, 508 (1986).
158. See id. at 508, 520
159. see also Daniel J. Meador, A Perspective on Change in the Litigation System, 49 ALA. L. REV. 7, 8-9 (1997). I do not mean to suggest, of course, that there were no complicated cases during the 1930s. Indeed, by the 1950s, judges in the Ninth Circuit were concerned enough about large antitrust suits to propose an amendment to the federal pleading rule. See supra note 109. There is no question, however, that today the large, complex suit is much more common, more complicated procedurally, and more salient as a focus of concern. Thanks to David Marcus for his suggestions on this topic.
160. See Meador, supra note 118, at 13-14 (noting the rise in recent decades of "mega-firms, national and international in scope, " and the increase in "incentives to litigate and aggressiveness in litigating"). When discussing the merits of a uniform code of pleading at the 1913 American Bar Association meeting, J. Hansell Merrill noted that few lawyers practiced interstate compared to the "much larger number whose practice is confined to a single state." AM. BAR ASS'N, REPORT OF THE THIRTY-SIXTH ANNUAL MEETING OF THE AMERICAN BAR ASSOCIATION, 38 A. B. A. REP. 44 (1913).
161. The automobile litigation crisis was felt in many different states and even prompted proposals to create administrative solutions modeled on the then-recently-enacted workmen's compensation statutes. See generally Robert G. Bone, Procedural Reform in a Local Context: The Massachusetts Supreme Judicial Court and the Federal Rule Model, in THE HISTORY OF THE LAW IN MASSACHUSETTS: THE SUPREME JUDICIAL COURT 1692-1992, at 403-04 (Russell K Osgood ed., 1992) (describing the automobile-litigation crisis in Massachusetts)
162. Noel T. Dowling, Compensation for Automobile Accidents: A Symposium, 32 COLUM. L. REV. 785 (1932).
163. See Meador, supra note 118, at 7-8, 12-14.
164. See supra notes 82-87 and accompanying text.
165. See supra notes 82-87 and accompanying text.
166. Clark, supra note 16, at 304.
167. In fact, Charles Clark seems to have tempered his more extreme notice pleading views in the late 1950s. See Marcus, supra note 4, at 451 & n. 113 (noting that Clark later described notice pleading as "helpful... but... [not] anything that we can use with any precision"). It is worth mentioning that Clark, who was a strong proponent of the court rulemaking process, might have objected to a stricter pleading rule being implemented by courts rather than through formal rulemaking. However, I believe there is a strong possibility that even Charles Clark, if he were living today, would approve the thin plausibility standard as a reasonable interpretation of Rule 8 (a) (2) in light of current litigation conditions.
168. Policy justifications for Rule 9 (b) 's heightened pleading standard include preventing unnecessary harm to the defendant's reputation, screening frivolous suits, protecting settled transactions, and preventing "fishing expeditions." See 5A WRIGHT & MILLER, supra note 8, § 1296, at 31-47. Sometimes the rationale is expressed in terms of providing "fair notice" to defendants, see id. at 39-45, but Rule 9 (b) is not required for tins purpose since Rule 8 (a) is supposed to do the notice-giving work. Furthermore, any adverse effect from a fraud allegation is a problem only if the allegation turns out to be wrong and the suit meritless. If the defendant is actually guilty of fraud, the reputation effect is deserved and any transaction based on the fraudulent representation should be rescinded. Thus, the stricter pleading standard for fraud must have to do with preventing unfounded allegations.
169. To be sure, the drafters, concerned about information access, included an exception in Rule 9 (b) for allegations of "[m]alice, intent, knowledge, and other conditions of a person's mind." FED. R. CIV. P. 9 (b). This exception indicates that the drafters, if they lived today, would take account of information access, as they should, when designing case-screening rules. How they would handle this factor depends on the pragmatic balance in light of the perceived severity of the frivolous suit problem. My point is simply that the drafters were alert to these problems and willing to use strict pleading to address them.
170. Bell Atl. Corp. v. Twombly, 127 S. Ct. 1955, 1967 (2007).
171. Id. at 1967 & n. 6.
172. Id.
173. Justice Stevens's dissenting opinion is a good example of the attitude that prevailed before Twombly.
174. The Court has restricted trial judge discretion in some special areas unrelated to pleading and case management out of fear that judges might use their discretion to make decisions that expedite resolution but adversely affect the public interest. For example, in order to avoid arresting the development of constitutional law, the Court until recently required district judges in § 1983 cases to decide the issue of constitutional violation before qualified immunity, even when a straightforward qualified immunity determination would easily dispose of the case. See Saucier v. Katz, 533 U. S. 194, 201, 210 (2001). Contra Pearson v. Callahan, 129 S. Ct. 808, 818 (2009) (abandoning Savcier 's strict rule and holding that district judges have some discretion in this regard).
175. See, e.g., Crawford-El v. Britton, 523 U. S. 574, 600 (1998) (noting that trial judge discretion can manage discovery and frivolous litigation effectively in § 1983 cases involving qualified immunity)
176. Leatherman v. Tarrant County Narcotics Intelligence & Coordination Unit, 507 U. S. 163, 168-69 (1993) ("[F]ederal courts and litigants must rely on summary judgment and control of discovery to weed out unmeritorious claims sooner rather than later.")
177. Pavelic & LeFlore v. Marvel Entm't Group, 493 U. S. 120, 127 (1989) (Marshall, J., dissenting) (stressing the trial judge's "primary responsibility for managing the cases before him" and the discretion Rule 11 gives to set an appropriate penalty).
178. See, e.g., Mass. Inst. of Tech. v. Abacus Software, 462 F.3d 1344, 1367 (Fed. Cir. 2006) (deferring to the district judge's case-management decisions because "a judge's discretion is at its broadest on matters of trial management")
179. Sims v. ANR Freight Sys., Inc., 77 F.3d 846, 848 (5th Cir. 1996) (lauding the district judge's "hard work and effective case management" skills even though his decision in the instant case caused a structural error)
180. Jaquette v. Black Hawk County, 710 F.2d 455, 463 (8th Cir. 1983) ("In almost all cases the key to avoiding excessive costs and delay is early and stringent judicial management of the case."). A caveat is in order: my claim that neither the Supreme Court nor the courts of appeals ever directly questioned the efficacy of trial judge discretion in managing cases before Twombly is based on extensive searches in the LEXIS and Westlaw databases.
181. See Robert G. Bone, Who Decides? A Critical Look at Procedural Discretion, 28 CARDOZO L. REV. 1961, 1967-70, 1974-75 (2007) (reviewing Federal Rules that feature discretion and explaining why it is so prevalent in the Federal Rules today)
182. Edward H. Cooper, Simplified Rules of Federal Procedure?, 100 MICH. L. REV. 1794, 1795 (2002) (emphasizing the prevalence of casespecific discretion in the Federal Rules)
183. Jonathan T. Molot, An Old Judicial Role for a New Litigation Era, 113 YALE L. J. 27, 37-46 (2003) (noting and criticizing the trend toward placing more case-management discretion in the hands of trial judges).
184. For a thorough analysis of these and other constraints, see Bone, supra note 135.
185. For more on the distinction between process-based and outcome-based theories, see Robert G. Bone, Agreeing to Fair Process: The Problem with Contractarian Theories of Procedural Fairness, 83 B. U. L. REV. 485, 508-16 (2003). A process-based approach is nonconsequentialist insofar as it evaluates procedure without regard to its consequences, whereas an outcome-based approach is consequentialist. However, an outcome-based approach can be tied to a nonconsequentialist metric for evaluating outcomes, as is the case for an outcome-based approach that relies on a rights-based theory of the substantive law. See infra notes 175-84 and accompanying text. For the distinction between consequentialist and nonconsequentialist theories in general, see Samuel Scheffler, Introduction to CONSEQUENTIALISM AND ITS CRITICS 1, 1-13 (Samuel Scheffler ed., 1988). Another point is worth mentioning. The analysis in this Section proceeds by deriving pleading and case-screening rules from broad normative theories of procedure. It is possible, however, to combine this analysis with a constructivist approach that develops procedural principles from an attractive normative account of settled practice. For an example of a constructivist approach, see Lawrence B. Solum, Procedural Justice, 78 S. CAL. L. REV. 181 (2004)
186. see also Bone, supra note 9, at 943-47 (arguing that the Advisory Committee should use a constructivist approach and illustrating it with the example of the small-claim class action).
187. See infranote 168 (describing and critiquing this tendency).
188. See infra Part IV. B.
189. Unless, of course, there are extrinsic circumstances that shed light on the probable reason for suit.
190. See, e.g., JERRY L. MASHAW, DUE PROCESS IN THE ADMINISTRATIVE STATE 158-253 (1985). Another version of the process-based approach anchors the participation right in legitimacy rather than dignity. See Solum, supra note 137. It is worth noting that much of the discussion of intrinsic process value in civil procedure draws on the "procedural justice" literature. This literature demonstrates through psychological experiments that people are more likely to feel they have been treated fairly by the process and the outcome, even an adverse outcome, when they are given an opportunity to participate. See, e.g., E. ALIAN LIND & TOM R. TYLER, THE SOCIAL PSYCHOLOGY OF PROCEDURAL JUSTICE 26-40, 61-83, 93-127 (1988) (discussing these empirical studies)
191. E. Allan Lind et al., In the Eye of the Beholder: Tort Litigants' Evaluations of Their Experiences in the Civil Justice System, 24 LAW & SOC'Y REV. 953, 967-87 (1990) (analyzing the relationship between tort litigants' fairness judgments and various objective and subjective factors). However, these psychological responses have no normative significance on their own. They might be relevant for a moral conventionalist if they counted as evidence of generally shared moral beliefs. However, psychological responses like these must be subjected to critical reflection before they can count as well-considered beliefs. In general, those who rely on procedural justice studies to support conclusions about fairness tend to assume that procedures are fair precisely because of the positive psychological reactions they elicit. But this is mistaken. If positive feelings are what count, then all positive feelings should be counted equally (even positive feelings about ice cream), but this quickly ends up sliding into some form of aggregate utilitarianism. See Bone, supra note 137, at 505-07.
192. I have argued elsewhere that the dignitary value underlies the strong right to a personal "day in court" that figures prominently in American civil procedure. Robert G. Bone, Rethinking the "Day in Court" Ideal and Nonparty Preclusion, 67 N. Y. U. L. REV. 193, 230 (1992). I have also raised questions about the coherence of this theory. Id. at 279-85.
193. In constitutional due process terms, we might say that the notice is not "reasonable" or the participation not "meaningful." See Armstrong v. Manzo, 380 U. S. 545, 552 (1965) (due process requires a hearing "at a meaningful time and in a meaningful manner")
194. Mullane v. Cent. Hanover Bank & Trust Co., 339 U. S. 306, 314 (1950) (due process requires notice "reasonably calculated to apprise interested parties of the pendency of the action and afford them an opportunity to present their objections").
195. See MASHAW, supra note 141, at 189, 198-99, 201-04 (deriving a requirement of comprehensibility or "thin rationality" from respect for dignity)
196. LAURENCE H. TRIBE, AMERICAN CONSTITUTIONAL LAW 663-64, 666 (2d ed. 1988) (arguing that due process provides an "institutional check on arbitrary government action" and identifying an intrinsic dignitary value to a hearing "as an expression of the rule of law, regarded here as the antithesis of power wielded without accountability to those on whom it focuses")
197. Richard B. Saphire, Specifying Due Process Values: Toward a More Responsive Approach to Procedural Protection, 127 U. PA. L. REV. 111, 120-24, 144-47 (1978) (arguing for "inherent dignity" as a foundation for due process and recognizing "the individual's right to be free from arbitrary governmental actions that threaten inherent dignitary values")
198. see also Robert S. Summers, Evaluating and Improving Legal Processes-A Plea for "Process Values, " 60 CORNELL L. REV. 1, 26-27 (1974) (arguing for "procedural rationality" as a "process value").
199. This is not the place to explore with care what constitutes a "non-arbitrary" decision. I employ the term in a loose way to refer to decisions that are backed by good reasons. For example, decision by random lottery is non-arbitrary by my definition as long as the decisionmaker has a good reason to use a lottery.
200. See supra notes 59-71 and accompanying text (describing the baseline idea).
201. See TRIBE, supra note 144, at 666, 743-44 (noting that "the right to be heard from, and the right to be told why,... express the elementary idea that to be a person, rather than a thing, is at least to be consulted about what is done with one" and arguing that a due process obligation to provide an explanation can be based on a dignitary value)
202. Frank I. Michelman, Formal and Associational Aims in Procedural Due Process, in DUE PROCESS: NOMOS XVIII, at 126, 126-27 (J. Roland Pennock & John W. Chapman eds., 1977) (arguing that "explanatory procedures, " including the giving of reasons for a decision, are "responsive to demands for revelation and participation " and can have value even when there is no opportunity to challenge the decision)
203. Edmund Pincoffs, Due Process, Fraternity, and a Kantian Injunction, in DUE PROCESS, supra, at 172, 175-79 (deriving a moral right to reasons from the Kantian categorical imperative not to treat a person as a mere means to an end)
204. Saphire, supra note 144, at 162-66 (arguing that due process protects inherent dignitary values of "revelation, explanation, and participation") Summers, supra note 144, at 26-27 (arguing that the process value of procedural rationality includes giving reasons, an obligation that respects persons as rational beings).
205. See, e.g., Goldberg v. Kelly, 397 U. S. 254, 264 (1970) (holding that procedural due process requires a hearing before the termination of a welfare recipient's benefits). This is, of course, based on the assumption that it is more difficult for an actor to act arbitrarily when she must justify her actions publicly. It is also worth mentioning that non-arbitrariness and reasongiving can be justified on outcome-based grounds, for example, as instrumental to the accurate enforcement of rights.
206. See Frederick Schauer, Giving Reasons, 47 STAN. L. REV. 633, 634, 637 (1995) (analyzing reason-giving and reason-avoidance in law).
207. See id. at 658 ("When the source of a decision rather than the reason behind it compels obedience, there is less warrant for explaining the basis for the decision to those who are subject to it.").
208. This last point, the existence of checks on impermissible decisionmaking, opens up the possibility that the fairness of a procedural system that allows filings without reasons might depend on how effectively the system screens impermissible filings in other ways.
209. See, e.g., Barzelis v. Kulikowski, 418 F.2d 869, 870 (9th Cir. 1969) ("A jury, however, does not have to give reasons for what it does.").
210. Another difference is that in the examples of no reason-giving, there are sound institutional grounds for acting without reasons. One can argue, for example, that the practice of jury verdicts, evidentiary rulings, or criminal sentencing works much better on balance without an obligation to give reasons. Therefore, even if there is a fairness value at stake in the abstract, it should not count because giving reasons is inconsistent with the effective functioning of the practice in question or it is outweighed by the benefits of decisionmaking without an obligation to state reasons publicly. By contrast, it is not clear why the practice of filing lawsuits would work better in general without an obligation to give good reasons for forcing the defendant to respond. Cf. Schauer, supra note 149, at 657-58 (arguing that reasongiving as an element of institutional design should be seen as contingent rather than necessary).
211. It is worth mentioning, however, that the U. S. Supreme Court has held that the Fourteenth Amendment's Due Process Clause does not prohibit a state from commencing a criminal prosecution without probable cause. Albright v. Oliver, 510 U. S. 266, 266 (1994). At first glance, this decision might suggest a negative answer to our question about a duty to act on and furnish good reasons. If the state can act without providing a good reason in the form of probable cause when initiating a criminal prosecution, perhaps the state can act without a good reason when initiating a civil suit as well. The Albright Court, however, did not hold that the state can initiate a prosecution without probable cause. It held only that any challenge to a state's decision to prosecute on the ground of probable cause had to be brought under the Fourth Amendment's seizure provisions rather than the Fourteenth Amendment's substantive due process provisions. And the Justices based their conclusions mainly on a theory of constitutional interpretation. I am grateful to my colleague David Rossman for bringing the Albright case to my attention.
212. See 5A WRIGHT & MILLER, supra note 8, § 1215, at 190-93.
213. The Twombly Court itself is not absolutely clear about the basis for its plausibility standard. On the one hand, the Court discusses the need for case screening at the pleading stage, suggesting that it might have in mind a case-screening rather than a fairness justification. Bell Atl. Corp. v. Twombly, 127 S. Ct. 1955, 1966-67 (2007). On the other hand, the Court introduces plausibility not in the course of discussing case screening, but rather in the course of dismissing the "possibility" standard as an improper interpretation of Conley and wrongheaded. Id. at 1968-69. Notably, the Court does not argue that the possibility standard is undesirable because of a high risk of meritless suits or high discovery costs. Instead, it argues that possibility simply is not a sensible standard in general because it tolerates "a wholly conclusory statement of claim" and allegations that do not even support "a 'reasonably founded hope' that a plaintiff would be able to make a case." Id. at 1969. While the Court does not refer here to fairness, a concern about fairness to defendants certainly could explain the force and generality of its criticism.
214. United States ex rel. Snapp, Inc. v. Ford Motor Co., 532 F.3d 496, 503 (6th Cir. 2008) ("[A] complaint need not provide an exhaustive roadmap of a plaintiff's claims, but it must be sufficient to 'give the defendant fair notice of what the... claim is and the grounds upon which it rests.'" (quoting Erickson v. Pardus, 127 S. Ct. 2197, 2200 (2007))). It is important to note, however, that the court does not explain what it means by "fundamental fairness."
215. See, e.g., Phillips v. County of Allegheny, 515 F.3d 224, 232 (3d Cir. 2008) ("[W]e understand the Court to instruct that a situation may arise where, at some point, the factual detail in a complaint is so undeveloped that it does not provide a defendant the type of notice of claim which is contemplated by Rule 8.")
216. Airborne Beepers & Video, Inc. v. AT&T Mobility LLC, 499 F.3d 663, 667 (7th Cir. 2007) ("[A]t some point the factual detail in a complaint may be so sketchy that the complaint does not provide the type of notice of the claim to which the defendant is entitled under Rule 8.")
217. cf. Robbins v. Oklahoma, 519 F.3d 1242, 1248 (10th Cir. 2008) ("This requirement of plausibility serves not only to weed out claims that do not (in the absence of additional allegations) have a reasonable prospect of success, but also to inform the defendants of the actual grounds of the claim against them.").
218. See, e.g., Bone, supra note 137, at 509-10 (critiquing process-based theories of procedure). Other scholars have expressed similar doubts, although on different grounds. See RONALD A. DWORKIN, A MATTER OF PRINCIPLE 101-03 (1985) (questioning whether there can be any moral harm from denying participation apart from the moral harm of an erroneous outcome)
219. Larry Alexander, Are Procedural Rights Derivative Substantive Rights?, 17 LAW & PHIL. 19, 31-36 (1998) (questioning the existence of substance-independent procedural values)
220. Larry Alexander, The Relationship Between Procedural Due Process and Substantive Constitutional Rights, 39 U. FLA. L. REV. 323, 325-26, 341-43 (1987) ("[B]ecause the procedure for applying a rule can always be viewed as part of the substance of the rule itself, a concern for procedure apart from substance verges on incoherence.").
221. To elaborate a bit, the question that a process-based proponent must answer is why respect for persons is not fully satisfied by an outcome for each individual litigant that is the best it can be given institutional constraints and the equal rights of other litigants. See Bone, supra note 137, at 509-10. In other words, since adjudication is primarily about producing good outcomes, why are parties entiled to anything more than those procedures that are justified on outcome-based grounds? Of course, there are background rights that all institutions must respect, such as the right against torture, but the case for process-based procedural rights extends further to encompass institution-specific rights, such as the right to control one's own lawsuit.
222. A skeptical reader might wonder whether the burden of discovery is substantial enough to trigger a moral obligation to give reasons. I believe the answer is yes, at least in some cases. Discovery can be very costly, especially when the plaintiff acts strategically to resist disclosure. Moreover, filing can also impose serious reputational and psychological harms. Another issue has to do with whether respect for persons extends to artificial persons like corporations. While I cannot explore this issue here, it is worth mentioning that the principle against the arbitrary exercise of state power extends generally, suggesting that its support is general too. I am indebted to Ian Farrell for alerting me to this point. Finally, one might object that the fairness argument fails to take account of important social interests. Suppose, for example, that plaintiffs suspect the defendant is using an unsafe method to transport nuclear waste, but the suspicions are not based on any concrete facts or evidence. Also, suppose that an accident could expose millions of people to serious levels of radioactivity. The possibility standard would permit this lawsuit, but the plausibility standard would bar it. One might argue that the powerful social interests at stake should allow plaintiffs at least to take some depositions. However, it is not at all clear that this is the right thing to do. Litigation is not the only institution for handling problems of this sort. Nuclear waste disposal, for example, is heavily regulated and there are public agencies capable of taking action. More generally, the hypothetical raises the complicated question of the proper use of adjudication as an investigative tool. Adjudication is most centrally about determining legal rights. It also has investigative value, of course, but this value alone is not enough to justify suit without some colorable legal claim. Furthermore, if the need for depositions is compelling enough on particular facts, it might be possible to make an exception. The fairness argument supports only a prima facie right to reasons. If litigation is somehow the only viable option to handle a serious social problem, the social interests at stake might outweigh the prima facie right. I am grateful to Patrick Woolley for prompting me to think about this issue.
223. See, e.g., Hoffman, supra note 4, at 1261-64
224. Spencer, supra note 4, at 433.
225. The Supreme Court has recognized a "fundamental constitutional right of access to the courts." Bounds v. Smith, 430 U. S. 817, 828 (1976). However, this constitutional right is limited and usually requires that the lawsuit involve a "fundamental interest." See TRIBE, supra note 6, at 1008
226. Michelman, supra note 6, at 1169.
227. See supra notes 141-43 and accompanying text. Or one might try to anchor it in the conditions for political legitimacy. But legitimacy in turn can be grounded in respect for persons. See, e.g., Frank I. Michelman, The Supreme Court and Litigation Access Fees: The Right to Protect One's Rights (pt. 2), 1974 DUKE L. J. 527, 539, 552-53, 556-57 (arguing that "all citizens by virtue of their citizenship have a 'fundamental interest' in access" to courts no matter what substantive right they allege and regardless of whether extrajudicial remedies exist).
228. One might argue that strict pleading is different from these examples because it denies access altogether and thus deprives a meritorious plaintiff of any opportunity to obtain relief. But this is incorrect. As I explain later, many plaintiffs who cannot sue because of strict pleading can still use the threat of filing as leverage for a pre-filing setdement. See infra notes 223-26 and accompanying text. To be sure, the pre-filing setdement is likely to fall short of full compensation, but that is true for post-filing setdements as well.
229. See supra Part II. B.
230. See infra notes 185-94 and accompanying text.
231. This is an application of the Kaldor-Hicks efficiency criterion. See generally Allan M. Feldman, Kaldor-Hicks Compensation, in 2 THE NEW PALGRAVE DICTIONARY OF ECONOMICS AND THE LAW 417 (Peter Newman ed., 1998) (describing Kaldor-Hicks efficiency analysis). It is quite common in the procedure literature to associate "efficiency" with reducing litigation costs and deterring frivolous suits, and "fairness" and "justice" with facilitating court access and expanding litigation opportunities for meritorious plaintiffs. See, e.g., Spencer, supra note 73, at 22 (discussing efficiency and entitlement to relief)
232. cf. Hoffman, supra note 4, at 1218-19 (juxtaposing efficiency and access to court). This formulation of the policy tradeoff can be misleading if it is not handled carefully. An efficiency metric is as concerned for plaintiffs as it is for defendants. It aims, after all, to achieve a socially optimal balance among all the competing objectives. In particular, efficiency, like fairness, is inconsistent with a single-minded commitment to reducing litigation costs and frivolous suits. Such a commitment gives insufficient weight to the error costs of screening meritorious suits. Efficiency and fairnessjustice support different metrics, of course, but the difference does not have to do with running roughshod over plaintiffs.
233. For a more detailed account of expected error and process costs and how to combine them to evaluate pleading rules, see ROBERT G. BONE, CIVIL PROCEDURE: THE ECONOMICS OF CIVIL PROCEDURE 125-49 (2003).
234. For example, both false-positive and false-negative errors dilute the deterrent effect of the substantive law, which increases social costs. Moreover, a very defective screening mechanism that allows lots of meritless suits in and keeps lots of meritorious suits out might lead to a loss of confidence in the court system, which could produce serious long-term costs if people resort to less peaceful means to settle disputes.
235. This is just an application of the general idea of expected utility or expected value. Formally, let r be the probability of an error and let c be the cost of the error. The expected error cost is r × c. For a more extensive explanation of expected value, see BONE, supra note 169, at 20-36.
236. See Alexander Volokh, N Guilty Men, 146 U. PA. L. REV. 173, 174-77 (1997). Another example involves environmental risk, where the cost of characterizing a toxic chemical as harmless (false negative) is significantly greater than the cost of characterizing a harmless chemical as toxic (false positive). See Talbot Page, A Generic View of Toxic Chemicals and Similar Risks, 7 ECOLOGY L. Q. 207, 219-20 (1978) (discussing false positives and false negatives in the context of environmental risk).
237. The objective function to be minimized can be expressed mathematically as follows. Let rFN be the probability of a false-negative error (screening of a desirable suit) and let cFN be the cost of a false-negative error. Let rFP be the probability of a false-positive error (filing and litigation of an undesirable suit) and let cFP be the cost of a false-positive error. Let s be the probability that whatever rule is in place will be invoked, thereby necessitating a motion, hearing, and the rest, and let cpc be the process cost expended to apply the rule when the rule is invoked. The goal of an economic analysis is to choose the rule or approach that minimizes the following expression: rFN Times; cFN + rFP × cFP + s × cpc -
238. For a numerical example, suppose 30% of all potential lawsuits are undesirable (e.g., meritless) and 70% are desirable (e.g., meritorious). Suppose that a strict pleading rule screens 60% of the undesirable suits and 20% of the desirable suits, and assume the cost of allowing an undesirable suit (false positive) is 20 and the cost of screening a desirable suit (false negative) is 10. Finally, suppose defendants file motions to dismiss in 30% of the cases and that it costs 5 to deal with a motion when one is filed. With these assumptions, one calculates the expected cost of the strict pleading rule as follows. The probability of a false-negative error is 20% (the fraction of desirable suits that are screened) × 70% (the fraction of potential suits that are desirable) = 0.14. Similarly, the probability of a false-positive error is 40% (the fraction of undesirable suits that are not screened) × 30% (the fraction of potential suits that are undesirable) = 0.12. Therefore, rFN = 0.14, cFN = 10, rFP = 0.12, cFP = 20, s = 0.3, and cpc = 5. The expected cost of the strict pleading rule is: 0.14 × 10 + 0.12 × 20 + 0.3 × 5 = 5.3. Compare this to the expected cost without strict pleading. Assuming no desirable suits are screened with very liberal pleading, the expected false-negative error costs are zero, and assuming all undesirable suits are filed and get past the pleading stage, the expected false-positive error costs are 0.3 × 20 = 6.0 Therefore, if the only choice is between a strict pleading rule and no strict pleading rule, the strict pleading rule wins, even though it screens 20% of desirable suits.
239. Unfortunately, proceduralists rarely explain their normative premises carefully, and this is true for those who criticize Twombly and strict pleading. Some Twombly critics frame the problem as a tension between achieving efficiency and guaranteeing access for meritorious plaintiffs to vindicate substantive rights. See, e.g., Hoffman, supra note 4, at 1218-19, 1224-25 (discussing the balance between access and efficiency)
240. Spencer, supra note 4, at 433, 479-83 (arguing that Twombly threatens open access values to achieve efficiency). Others simply note the risk of screening meritorious suits, apparently assuming that this alone is a sufficient ground for objection, See, e.g., Ward, supra note 11, at 911 ("Using motions to dismiss as a mechanism to control litigation costs... may... bar[] legitimate meritorious claims."). To make sense of these arguments, it is best to interpret them as invoking a right on the plaintiffs side capable of pushing against economic efficiency and supporting a priority concern for plaintiffs.
241. For a developed argument defending a rights-based theory of procedure, see generally DWORKIN, supra note 159.
242. A settlement can be unfair or unjust to a defendant even though the defendant gave his consent. Consent cannot validate a settlement when the background conditions that produce the consent are themselves normatively flawed. For more on this point, see Bone, supra note 135, at 1981-85.
243. There is a difference, however, between how a false negative and a false positive work in this argument, which might complicate the point a bit. A false negative denies access to an entitled plaintiff and thus absolutely prevents her from vindicating her substantive right through adjudication. A false positive, on the other hand, simply imposes a risk of unjustified liability; the defendant can still win in the end. To be sure, the innocent defendant is forced to defend a lawsuit, but it is not clear that this is a moral wrong if his legal entitlement is a right to be free from liability as opposed to a right to be free from suit. Cf. Mitchell v. Forsyth, 472 U. S. 511, 526 (1985) (distinguishing an immunity from a substantive defense, noting that "[qualified immunity] is an immunity from suit rather than a mere defense to liability"). Nevertheless, the defendant is also pressured to settle, and if the pressure is intense enough, it is not clear why paying the unjustified settlement is morally different from failing to gain access on plaintiff's side.
244. See generally DWORKIN, supra note 159 (making this point about procedural rights).
245. Section 1983 here refers, of course, to 42 U. S. C. § 1983, which provides the remedial vehicle for individuals to sue state officials who violate their constitutional rights. 42 U. S. C. § 1983 (2000).
246. This is captured doctrinally by the strict scrutiny test that requires a compelling state interest. See CHEMERINSKY, supra note 6, at 541-42.
247. See Anderson v. Creighton, 483 U. S. 635, 638 (1987) ("Permitting damages suits against government officials can entail substantial social costs, including the risk that fear of personal monetary liability and harassing litigation will unduly inhibit officials in the discharge of their duties.")
248. Peter H. Schuck, Suing Our Servants: The Court, Congress, and the Liability of Public Officials for Damages, 1980 SUP. CT. REV. 281, 305-15 (describing various ways that the risk of personal liability can chill government officials from acting in the public interest). But see Jack M. Beermann, A Critical Approach to Section 1983 with Special Attention to Sources of Law, 42 STAN. L. REV. 51, 95 (1989) (noting the absence of reliable empirical evidence to support concerns about chilling government officials).
249. This is one way to understand the Court's reasoning in Goldberg. Goldberg v. Kelly, 397 U. S. 254 (1970) (a procedural due process, not First Amendment, case)
250. see also Fuentes v. Shevin, 407 U. S. 67, 90 n. 22 (1972) (noting in a procedural due process case that "ordinary costs cannot outweigh [a] constitutional right" to a hearing and that "procedural due process is not intended to promote efficiency or accommodate all possible interests: it is intended to protect the particular interests of the person whose possessions are about to be taken").
251. See DWORKIN, supra note 159. This equality constraint needs more careful elaboration, but the simple formulation in the text is sufficient for our purposes. The general idea is that a procedural system should care more about outcome errors when the moral harm from those errors is more serious, and therefore invest more in preventing errors associated with more serious moral harms. However, it would be absurd if the equality constraint demanded that scarce process resources be funneled exclusively to those lawsuits that have moral stakes and away from those that do not. Professor Dworkin's theory, which derives ultimately from a background right to treatment as an equal, id. at 84-85, avoids this problem by assuming that all outcome errors produce moral harm no matter what substantive law is involved. The moral harm inheres in the denial of a substantive legal right itself (plaintiffs and defendants have mirror-image rights in Dworkin's theory). Id. at 80-81. The seriousness of this moral harm, however, depends on how the legal system as a whole values the type of substantive interest at stake, which in turn depends on how much the legal system is willing to invest to protect that interest. Id. at 93-98. The result is a requirement similar to the one I describe in the text: every procedural system must distribute the risk of error consistent with the best theory of moral harm embedded in the legal system as a whole. This means that the error-risk distribution must respect the relative weight of moral harm as between litigants in the same lawsuit and among litigants in different types of suits.
252. The discussion in Parts IV. B.3 and IV. C builds on my previous work in this area. See Robert G. Bone, Modeling Frivolous Suits, 145 U. PA. L. REV. 519, 529-37 (1997).
253. Unfortunately, this definitional question is all too often ignored in discussions of case screening, despite its pivotal importance. See Warren F. Schwartz & C. Frederick Beckner III, Toward a Theory of the "Meritorious" Case: Legal Uncertainty as a Social Choice Problem, 6 GEO. MASON L. REV. 801, 801 (1998) (noting the lack of careful attention to the question). Moreover, those commentators who do attend seriously to the issue vary in how they define the set of lawsuits that should not be brought, with some adopting definitions that include suits with substantial merit. See, e.g., David Rosenberg & Steven Shavell, A Model in Which Suits Are Brought for Their Nuisance Value, 5 INT'L REV. L. & ECON. 3, 3 (1985) (including in the category of nuisance suits some that have merit but that the plaintiff would not take to trial because the litigation costs exceed the expected recovery)
254. Schwartz & Beckner III, supra, at 816-19 (arguing on the basis of social choice theory for a 50% probability of success as the proper merits threshold).
255. This is, of course, the test for summary judgment, but assessed objectively on the assumption that all the information is available to the decisionmaker, not just whatever information the parties have been able to obtain.
256. As the threshold is set higher, the cases that enter the system have a greater chance of liability and thus a stronger claim on scarce private and public litigation resources. See, e.g., Keith N. Hylton, When Should a Case Be Dismissed? The Economics of Pleading and Summary Judgment Standards, 16 SUP. CT. ECON. REV. 39, 41 (2008) ("[E]arly dismissals, by eliminating low-merit claims before they become costly, offer benefits to society in comparison to late dismissals.")
257. see also Schwartz & Beckner III, supra note 186, at 816-19 (using social choice theory to justify setting the threshold at 50%).
258. I do not mean to suggest that a rights-based proponent would never set a threshold above zero. Process costs and false-positive error costs matter in a rights-based theory, and if these costs are sufficiently serious, a rights-based proponent can support a strictly positive screening threshold too. Before doing so, however, she would demand a high risk of very serious costs and reliable information confirming that risk.
259. One might cite Brown v. Board of Education, 347 U. S. 483 (1954), as an example. Of course, the judge has a chance to develop new law at the morion to dismiss stage when she evaluates the strength of the case in light of the legal standards. However, she would not have this chance if a screening device prevents filing, and even if she does, her decision might benefit from exposure to more information than is available at the pleading stage.
260. Bell Atl. Corp. v. Twombly, 127 S. Ct. 1955, 1966-67 (2007) (quoting Dura Pharms., Inc. v. Broudo, 544 U. S. 336, 347 (2005)).
261. Id. (emphasis added).
262. Id. at 1973 n. 14. It also fits what the Court says about the importance of judges not making predictions of likely trial success at the pleading stage. Id. at 1965 ("[A] well-pleaded complaint may proceed even if it strikes a savvy judge that actual proof of those facts is improbable, and 'that a recovery is very remote and unlikely.'" (citation omitted)).
263. See Bone, supra note 185, at 533 (providing a more in-depth discussion of how to define frivolous suits).
264. See Randy J. Kozel & David Rosenberg, Solving the Nuisance-Value Settlement Problem: Mandatory Summary Judgment, 90 VA. L. REV. 1849, 1851 n. 3 (2004) (citing empirical studies and noting a "general consensus that a problem exists" but also emphasizing the "paucity of empirical research substantiating its extent").
265. Twombly, 127 S. Ct. at 1967.
266. See generally Chris Guthrie, Framing Frivolous Litigation: A Psychological Approach, 67 U. CHI. L. REV. 163, 165 (2000)
267. Rosenberg & Shavell, supra note 186.
268. Twombly, 127 S. Ct. at 1967. The limits might be imposed by general rule, but that approach has problems as well. See Bone, supra note 135, at 2006-11.
269. Throughout the rest of the analysis, I use the standard economic model of litigation. S ee, e.g., BONE, supra note 169, at 30 (noting that the standard economic model is sometimes referred to as the Landes-Posner-Gould model)
270. ROBERT COOTER & THOMAS ULEN, LAW AND ECONOMICS 481 (1988). That model assumes that the parties are rational and make decisions by considering their expected benefits and costs. An expected benefit or expected cost is just the benefit or cost discounted by the probability it will materialize. See supra note 171 (explaining expected value). For example, the plaintiffs expected benefit from filing suit is the likely trial award discounted by the probability of success at trial, and her expected cost is the cost of litigating. A plaintiff will file suit if the expected benefit exceeds the expected cost. This can be expressed mathematically as follows. Let p be the plaintiff's probability of trial success, w be the likely trial award if the plaintiff is successful, and c be the plaintiffs anticipated cost of litigating through trial. The plaintiff will file suit if p × w > c. When this condition holds, the lawsuit has positive expected value and is known as a PEV suit. When the inequality is reversed, the lawsuit has negative expected value and is known as a NEV suit.
271. See BONE, supra note 169, at 19-20.
272. See, e.g., Rosenberg & Shavell, supra note 186 (constructing a formal model that incorporates this strategic interaction).
273. See Bone, supra note 185, at 537-41 (discussing several limitations of the argument). Indeed, there is no apparent reason to believe that the cost of filing an answer will exceed the cost of filing a complaint, which is a necessary condition in this model for a meritless filing to be profitable in the first place. Also, in cases where both sides know a suit is meritless, repeatplay defendants can benefit from building a reputation for fighting meritless suits, which deters future meritless filings. One might wonder why the defendant's response costs do not include more than the cost of an answer. The reason is that filing an answer forces the plaintiff to invest more in the lawsuit and a meritless plaintiff will not invest more unless she expects a large enough settlement to cover all her costs at a later stage of the litigation. Professor Lucian Bebchuk has modeled this dynamic. See generally Lucian Bebchuk, A New Theory Concerning the Credibility and Success of Threats to Sue, 25 J. LEGAL STUD. 1 (1996). He shows that if there are enough separate stages of a lawsuit and total litigation costs are divided among the different stages in the right way, it is possible for the plaintiff in an NEV suit to make a credible threat to take the case all the way through trial. Id. The analysis relies on backward induction and is somewhat complicated. Id. There is no need to explore it, however, because it is very unlikely to apply to a truly meritless suit or to produce substantial settlements even when it does. See Bone, supra note 185, at 539 n. 73 (arguing that a frivolous plaintiff will have a hard time presenting a credible threat at every stage of litigation). See generally Joseph A. Grundfest & Peter H. Huang, The Unexpected Value of Litigation: A Real Options Perspective, 58 STAN. L. REV. 1267, 1299-1305 (2006) (describing an option model of litigation that can explain negative-expected-value suits but does not clearly explain truly meritless suits that have tiny variance and that both parties know have no trial value).
274. The numbers work out as follows. With an error rate of 4%, the expected classwide recovery is 0.04 × $50 million, or $2 million, exactly equal to the expected litigation costs. Thus, anything in excess of 4% yields a positive expected value from taking the case to summary judgment and then on to trial. The result changes if the class attorney makes the decisions in her self-interest, because the attorney expects only a fraction of the total recovery. The typical class attorney fee is 25% of class recovery. See Theodore Eisenberg & Geoffrey Miller, Attorney Fees in Class Action Settlements: An Empirical Study, 1 J. EMPIRICAL LEGAL STUD. 27, 57-65 (2004). When the attorney expects 25% of the recovery and litigation costs are $2 million, litigation is cost-justified for any error rate in excess of 4% only if the expected recovery exceeds $200 million (0.04 & 0.25 × $200 million = $2 million). Alternatively, litigation is cost-justified when expected recovery is $50 million and litigation costs are $2 million only if the error rate exceeds 16% (0.16 × 0.25 × $50 million = $2 million).
275. Filing might also occur when the plaintiff is motivated by revenge or her attorney is overly optimistic about her ability to trick a jury. But it is difficult to imagine a level of optimism sufficiently high or a desire for revenge sufficiently strong to make it worthwhile to try a case that the attorney knows is meritless. Another way in which trial can become a credible threat for a meritless suit is if the plaintiff's attorney develops a reputation for taking meritless suits to trial. See Amy Farmer & Paul Pecorino, A Reputation for Being a Nuisance: Frivolous Lawsuits and Fee Shifting in a Repeated Play Game, 18 INT'L REV. L. & ECON. 147 (1998). But the conditions that make reputation work in this way can be difficult to sustain in practice.
276. In general, for a suit to be meritless, the plaintiff must have no admissible evidence of an essential element of her prima facie case, or the defendant must have a clearly decisive affirmative defense. In the former situation, it is difficult to see how the plaintiff could get past summary judgment without admissible evidence. In the latter situation, the defendant should be able to prevail on summary judgment if his evidence supports only one reasonable conclusion. In an antitrust conspiracy case like Twombly, for example, the plaintiff must offer evidence of conspiracy above and beyond parallel conduct in order to avoid summary judgment, and it is not clear what evidence a meritless plaintiff could provide. The plaintiff might have a better chance if the judge were less strict about admissible evidence and willing to exercise her discretion to deny summary judgment even when all the requirements are satisfied. See generally Jack H. Friedenthal & Joshua E. Gardner, Judicial Discretion to Deny Summary Judgment in the Era of Managerial Judging, 31 HOFSTRAL. REV. 91, 104-05 (2002).
277. Professor Paul Standi has developed a model that he claims shows how discovery cost asymmetries can produce meritless filings even when both parties know that the suit is meritless. Paul J. Standi, Balancing the Pleading Equation 30-52 (Univ. of Ill. Coll. of Law, Law & Econ. Research Paper Series, Research Paper No. LE08-018, 2008), available at http://papers. ssrn.com/sol3/papers.cfm?abstract-id=1266323. Professor Standi sharpens our understanding of the problem by clearly separating informational asymmetries from cost asymmetries and by proposing a bifurcated approach to pleading standards. His model, however, seems to assume that a meritless plaintiff can make a credible threat to litigate to summary judgment simply because the defendant cannot exit the lawsuit until the summary judgment stage. But this assumption ignores how the strategic imposition of costs bilaterally and sequentially can undermine the credibility of a meritless plaintiff's threats. See supra note 202. It is possible that the model assumes a cost-free stage for the plaintiff, but cost-free stage models have difficulty explaining the filing of truly meritless suits because there is not likely to be a cost-free stage sufficiently late in the litigation to support a substantial settlement. See Bone, supra note 185, at 539 n. 73.
278. See infra note 212-13 and accompanying text.
279. Bell Ad. Corp. v. Twombly, 127 S. Ct. 1955, 1967 (2007).
280. For a classic analysis of gatekeeping, although not in the context of litigation, see Reinier Kraakman, Gatekeepers: The Anatomy of a Third-Party Enforcement Strategy, 2 J. L. ECON. & ORG. 53, 53 (1986) (discussing "gatekeeper liability" as a regulatory device). In our situation, the gatekeeper is the judge, who prevents a meritless filing by checking for it at the pleading stage, and gatekeeping rules assist the judge in this task. See id. at 54 (defining a gatekeeper).
281. In a sense, all screening rules shape incentives. The difference is that a gatekeeping rule screens cases directly at the pleading stage while an incentive-shaping rule works indirectly by affecting the payoff.
282. Of course, a plaintiff who knows the suit is meritorious might try to persuade the uninformed defendant of that fact, and an informed defendant who knows the suit is meritless might do the same for the uninformed plaintiff. However, these efforts face serious obstacles because the other type-i.e., informed plaintiffs in meritless suits and informed defendants in meritorious suits-will mimic in an effort to trick the uninformed party into believing that they are the same and should receive a large settlement as well. Anticipating this, the uninformed party will discount the truth of any statements or representations made by the opposing side. See, e.g., Lucian Arye Bebchuk, Suing Solely to Extract a Settlement Offer, 17 J. LEGAL STUD. 437, 442 n. 8 (1988) (noting that parties cannot eliminate information asymmetry without incurring prohibitive costs)
283. Steven Shavell, Sharing of Information Prior to Settlement or Litigation, 20 RAND J. ECON. 183, 183-84 (1989) (analyzing voluntary information sharing where plaintiffs possess information unknown to the defendant).
284. Developing a fighting strategy by refusing to settle does not work as well when information is asymmetrically distributed. Cf. supra note 202 (mentioning the possibility of a fighting strategy for symmetric-information cases). If the defendant does not know whether suit is meritless, he cannot confine his fighting strategy to meritless suits, so he will sometimes end up fighting a meritorious suit by mistake and pay litigation costs as a result. If the plaintiff does not know whether her suit is meritless, the defendant has to worry that fighting might backfire if the meritless plaintiff is sufficiently confident that her suit is meritorious. Defendants sometimes fight in these situations-but not always-and fighting generates additional litigation costs.
285. In fact, a defendant who wants to settle meritorious claims has no incentive to offer anything less than a meritorious plaintiff would accept because he knows that only meritless plaintiffs would accept the lower amount. See Avery Katz, The Effect of Frivolous Lawsuits on the Settlement of Litigation, 10 INT'L REV. L. & ECON. 3, 9 (1990).
286. For example, some states have created medical malpractice panels to screen medical malpractice suits immediately upon filing. See Jean A. Macchiaroli, Medical Malpractice Screening Panels: Proposed Model Legislation to Cure Judicial Ills, 58 GEO. WASH. L. REV. 181, 188-97, 239-51 (1990). It is significant, however, that medical malpractice suits usually involve informed defendants and uninformed plaintiffs.
287. The analysis is more complicated. One must compare the filing cost with the expected gain from a meritless filing, which is the amount of the settlement discounted by the probability an offer is made. See Katz, supra note 213, at 13-15.
288. See supra note 184 and accompanying text (discussing the distributional implications of a rights-based approach).
289. In fact, most of the empirical studies indicating a possible problem focus on substantive fields where the defendants are likely to be the informed parties. See, e.g., Kozel & Rosenberg, supra note 195, at 1851 n. 3 (collecting studies, including ones done for securities fraud and medical malpractice cases). For example, securities fraud plaintiffs are likely to have difficulty obtaining the information necessary to allege scienter. Antitrust plaintiffs suffer from similar problems, as Twombly itself illustrates. Also, in medical malpractice cases, the doctor is the one with knowledge about what he did or did not do. And those civil rights suits involving constitutional rights that feature defendant's intent as an element also fit this profile. See, e.g., Crawford-El v. Britton, 93 F.3d 813, 817-18 (D. C. Cir. 1996) (involving a constitutional claim that turned on defendant's intent).
290. I have analyzed these incentives in some detail elsewhere. See. Bone, supra note 185, at 550-63. The strategic dynamics are complex but the intuition is as described in this paragraph.
291. A merits review at the filing stage benefits from the fact that the defendant is informed and can provide input, but the fact that the plaintiff is uninformed makes it more difficult to test the defendant's assertions. Many medical malpractice suits, for example, involve an informed doctor and an uninformed patient, and the informed doctor can provide information to a medical malpractice screening panel. See supra note 214 (discussing screening panels). A professional panel has the expertise and incentive to challenge the doctor's arguments as long as the panel is neutral (which is at least a questionable assumption), and this might offset to some extent the plaintiffs informational disadvantage.
292. Or if she does file, she is likely to obtain only a small nuisance settlement. See supra note 202 and accompanying text.
293. See, e.g., Epstein, supra note 59 (discussing publicly available information to support an inference of conspiracy in antitrust cases)
294. Geoffrey P. Miller, Pleading After Tellabs 11-19 (N. Y. Univ. Sch. of Law, Law & Econ. Research Paper Series, Working Paper No. 08-16, 2008), available at http://papers.ssrn.com/sol3/papers.cfm?abstract-id=1121396 (discussing the types of information that can support an inference of scienter in a securities fraud suit).
295. See supra note 216 and accompanying text.
296. See Miller, supra note 221, at 17.
297. There is precedent for a practice of extra-adjudicative settlement emerging in an environment of long delays and high litigation costs. In the late nineteenth and early twentieth century, repeat-play employers cooperated with brokers for injured employees to create a private scheme of aggregative settlement for workplace injuries. See Samuel Issacharoff & John Fabian Witt, The Inevitability of Aggregate Settlement: An Institutional Account of American Tort Law, 57 VAND. L. REV. 1571, 1584-96 (2004). Also, in the first half of the twentieth century, liability insurers developed a private scheme of scheduled settlement for automobile accident claims. Id. at 1603-10. At least some of these settlements appear to have been negotiated without the filing of a lawsuit. See id. at 1582-83. These developments are associated with special factors, such as correlated conduct and reputation markets, which might not exist in all cases. They are relevant nevertheless because they arose during periods of intense case congestion (a sharp spike in tort litigation in the late 1800s and a huge influx of automobile cases in the early 1900s) that should have greatly weakened any credible threat by plaintiffs to litigate their claims.
298. Consider the following simple and somewhat more formal analysis. Let R be the expected recovery by setdement or trial judgment if a meritorious plaintiff files suit and gets past the pleading stage. Let c be the plaintiffs cost of litigating to the point of recovery and let d be the defendant's cost. Assume that there are two types of cases. In type 1, all the information about wrongdoing remains hidden. In type 2, enough information is available from reasonably accessible sources to allow an investigating plaintiff to meet the pleading burden. Neither party knows for sure what type of case he has, but both estimate a probability, a, that the case is type 2. Finally, let b be the cost of a pre-filing investigation in a type 2 case. If b < α (R-c), the plaintiff would investigate if there were no pre-filing settlement and sue with probability a (assuming R - c > 0). Knowing that the plaintiff will investigate and sue with probability α, the defendant's loss without settlement is α (R + d). Thus, both parties are better off settling for some amount between α (R-c) -b and α (R + d). This much makes the point, but it is also worth mentioning that a more thorough analysis would take account of signaling and other strategic effects.
299. See supra note 225.
300. This is as compared with an inquiry that ranges more widely to engage the broad question of whether the suit meets some threshold level of merit.
301. Other creative approaches have been proposed, such as making summary judgment mandatory as a condition for a judicially enforceable settlement agreement and giving the defendant an option to bar settlement. See Kozel & Rosenberg, supra note 195, at 1860-79 (proposing mandatory summary judgment)
302. David Rosenberg & Steven Shavell, A Solution to the Problem of Nuisance Suits: The Option to Have the Court Bar Settlement, 26 INT'L REV. L. & ECON. 42, 49-50 (2006) (proposing an option to bar setdement). In addition, purely substantive reforms, such as caps on pain and suffering and elimination of punitive damages, have been proposed to reduce filing incentives for frivolous suits, but I do not address these alternatives here.
303. Fee-shifting conditional on a trial loss is not a good choice to deter meritless filings. It does generate deterrence benefits for some cases. See Farmer & Pecorino, supra note 204, at 149 (arguing that fee-shifting can reduce meridess filings when attorney reputation is a driving force behind nuisance suits)
304. Katz, supra note 213, at 17-19 (showing that the effect of adopting the English rule depends on whether it increases expected recovery for meritorious plaintiffs). But it is likely to deter a large number of weak but meritorious suits as well. See Steven Shavell, Suit, Settlement and Trial: A Theoretical Analysis Under Alternative Methods for the Allocation of Legal Costs, 11 J. LEGAL STUD. 55, 59-60 (1982).
305. The defendant does not know at the outset, but presumably would find out and seek penalties if die case goes through discovery. Therefore, the plaintiff must factor the penalty risk into her settlement calculations. See Katz, supra note 213, at 19-20 (discussing the implications of his informed-plaintiff model for refundable deposits or penalties).
306. Rule 11 currently uses tins approach. See FED. R. CIV. P. 11 (b). I have argued elsewhere that a penalty equal to the cost of discovery might encourage all plaintiffs to conduct reasonable pre-filing investigations, but this assumes courts can define reasonableness clearly enough to minimize errors and uncertainty at the enforcement stage. Bone, supra note 185, at 592.
307. There is another reason why a penalty system that targets only meridess suits might work better in cases where the plaintiff is the informed party. As a practical matter, trial judges are much more willing to impose penalties when the plaintiff and her attorney file knowing that the suit is meridess (or with reckless disregard for the obvious possibility). Penalties smack of punishment for a moral wrong, and knowingly filing is much more culpable than a negligent failure to conduct a reasonable pre-filing investigation.
308. To be sure, meritorious plaintiffs know they are likely to settle and not actually face a penalty proceeding, but the penalty risk reduces the settlement amount and thereby deters meritorious filings at the margin. See Katz, supra note 213, at 19.
309. This does not necessarily imply that the expected process costs of a penalty system would be high as well. While the cost of any particular penalty proceeding is likely to be large, the number of proceedings should be small if penalties are working properly. In that case, the defendant should expect few frivolous suits and therefore a low likelihood of success, which should produce few penalty motions and thus few proceedings. Still, expected process costs might be lower with strict pleading, because pleading-based dismissals do not implicate the attorney's reputation as directly and conspicuously and therefore are not likely to trigger the same degree of contentiousness.
310. See A. Mitchell Polinsky & Daniel L. Rubinfeld, Sanctioning Frivolous Suits: An Economic Analysis, 82 GEO. L. J. 397, 404-06 (1993) (modeling the effects of a system of sanctions, including the effects of costly sanction proceedings).
311. This is another reason why a penalty approach is better suited to informed-plaintiff than informed-defendant cases. The risk of chilling meritorious suits might be even greater when the plaintiff is uninformed and penalized for failing to conduct a reasonable pre-filing investigation if the vagueness of the reasonableness standard creates a large enough error risk.
312. See 5A WRIGHT & MILLER, supra note 8, § 1332 (discussing criticism of the 1983 amendments to Rule 11, citing the increase in satellite litigation adding substantial process costs as well as excessive screening of meritless suits).
313. See supra notes 203-04 and accompanying text.
314. For a skeptical view, see Charles Silver, " We're Scared to Death": Class Certification and Blackmail, 78 N. Y. U. L. REV. 1357 (2003) (analyzing and criticizing the argument that class action suits effectively blackmail defendants through settlement pressure).
315. See Robert G. Bone & David S. Evans, Class Certification and the Substantive Merits, 51 DUKE L. J. 1251, 1257 (2002).
316. See Kozel & Rosenberg, supra note 195, at 1853.
317. For a critique of one proposal citing, among other things, problems created by informational asymmetry, see Ted M. Sichelman, Why Barring Settlement Bars Legitimate Suits: A Reply to Rosenberg and Shavell, 18 CORNELL J. L. & PUB. POL'Y (forthcoming 2009), available at http://papers.ssrn.com/sol3/papers.cfm?abstract-id=986234 (criticizing Rosenberg's and Shavell's proposal to give defendants an option to bar settlements).
318. It might also be appropriate to impose penalties on a plaintiff or her attorney who does absolutely no pre-filing investigation without good reason. This conduct fails to show respect for the defendant and the court, and implicates moral culpability. See, e.g., Johnson v. A. W. Chesterton Co., 18 F.3d 1362, 1366 (7th Cir. 1994) (imposing sanctions for "reckless willingness to impose the burden of unwarranted litigation upon others"). However, extending penalties beyond this extreme case to an attorney who does some but not a reasonable investigation is more problematic, given the vagueness of the reasonableness standard. For this reason, I favor narrowing the current scope of subdivisions (b) (2) and (b) (3) of Rule 11.
319. See FED. R. CIV. P. 11 (c) (1). Also, Rule 11 's safe harbor provision should be eliminated. See FED. R. CIV. P. 11 (c) (2). And, of course, the penalties should be limited to the category of suits covered by current Rule 11 (b) (1).
320. See FED. R. CIV. P. 11 (c) (4) (requiring that a sanction under Rule 11 "be limited to what suffices to deter repetition of the conduct by others similarly situated").
321. A fee-shifting penalty forces the plaintiff to internalize the private costs of filing and litigating.
322. The current Rule 11 instructs the judge to choose the penalty that achieves general deterrence, but it is difficult to see how trial judges can do that effectively on a case-by-case basis.
323. See supra note 217 and accompanying text.
324. Judges routinely allow targeted discovery before deciding a motion to dismiss for lack of subject matter or personal jurisdiction, but rarely before deciding a Rule 12 (b) (6) motion. Nevertheless, a few courts before Twombly permitted discovery in cases subject to heightened pleading standards and with information in the exclusive control of the defendant, usually after finding the original complaint deficient and before the plaintiff filed an amended complaint. See, e.g., Cordero-Hernandez v. Hernandez-Ballesteros, 449 F.3d 240, 244 (1st Cir. 2006) (limiting discovery to those cases where the allegations make it likely that wrongdoing occurred and the information needed to plead with particularity is in the exclusive control of the defendant)
325. New England Data Servs., Inc. v. Becher, 829 F.2d 286, 290 (1st Cir. 1987) (holding that discovery may be warranted in certain fraud cases)
326. *15-16 (N. D. Ill. Apr. 10, 1996) (permitting a plaintiff to conduct some discovery before re-filing an insufficiently pled RICO claim)
327. see also Reints v. Sheppard, 90 F. R. D. 346, 347 (M. D. Pa. 1981) (agreeing with the policy that dismissal before discovery may be inappropriate for cases of heightened pleading specificity if the information is held exclusively by the defendant). It is worth mentioning that Federal Rule 27 allows a prospective party to take depositions before a lawsuit is filed. However, the prevailing view is that the right is limited to perpetuating testimony that might be lost before trial (such as a dying witness), although there are a few decisions to the contrary. See In re Bay City Middlegrounds Landfill Site, 171 F.3d 1044, 1047 (6th Cir. 1999) (allowing the deposition of a witness who was not dying). But see Reints, 90 F. R. D. at 347-48 (rejecting plaintiff's deposition requests)
328. Bowles v. Pure Oil Co., 5 F. R. D. 300, 302-03, 303 n. 2 (E. D. Pa. 1946) (rejecting plaintiffs deposition requests). Also, some states have special rules allowing pre-filing discovery, but these rules apply only in state court and are available only under limited conditions. See generally Lonny Sheinkopf Hoffman, Access to Information, Access to Justice: The Role of Presuit Investigatory Discovery, 40 U. MICH. J. L. REFORM 217 (2007) (discussing the range of pre-suit discovery tools).
329. *29 (N. D. Cal. Dec. 24, 2007) (finding no authorization in the Federal Rules for pre-dismissal discovery in the circumstances of the case).
330. Another possibility is to give discretion to the trial judge to stage discovery in increments, evaluating the strength of the case after each stage. To proceed to the next stage, the plaintiff would have to convince the judge that the case is strong enough to warrant further discovery given what has already been obtained. I am skeptical, however, that trial judges can make good decisions about discovery's benefits to be able to implement a staged approach like this effectively in a complex case. Moreover, additional discovery increases costs and enhances the plaintiff's settlement leverage.
331. See supra notes 128-36 and accompanying text.
332. Justice Stevens proposed something similar to this in his Twombly dissent. See Bell Atl. Corp. v. Twombly, 127 S. Ct. 1955, 1986-87 (2007) (Stevens, J., dissenting) (suggesting plaintiff should have been able to depose at least one representative of each defendant).
333. The rule also could require that the discovery be confined to those elements that the plaintiff has trouble alleging because of the informational asymmetry. To facilitate this, the plaintiff might be required to explain why she cannot make the necessary allegations without the information she seeks. See In re Graphics Processing, 527 F. Supp. 2d at 1032-33 (requiring a similar explanation).
334. The judge would be required to meet whatever level of confidence the strict pleading rule requires.
335. For one thing, defendants have incentives to conceal information or delay its disclosure when they know that the discovery tools available to plaintiffs are extremely limited and the time period for using them is short. The plaintiffs lawyer can anticipate these incentives and will try to counter them by strategically deploying the limited discovery available to her. Given this strategic interaction, it is difficult to predict exactly how effective pleadingstage discovery will be. However, it is promising enough to give it a try and see how it works.
336. For example, formal discovery is not supposed to begin until after the parties meet and confer pursuant to Rule 26 (f). FED. R. CIV. P. 26 (d) (1). The trial judge can grant an exception, but judges might be reluctant to do so without express authorization if they view routine exceptions as inconsistent with the purposes of the Federal Rules' elaborate discovery scheme.
337. *29 (N. D. Cal. 2007) (reviewing all the possible bases for pleading-stage discovery and concluding that allowing discovery "before an operative complaint is filed" is inconsistent with the timing and protective measures that constitute the Federal Rules' discovery scheme). Rule 27 allows pre-filing depositions but only to perpetuate testimony that might be lost before trial. FED. R. CIV. P. 27
338. see also supra note 249.