OPINION ON PROPOSED SETTLEMENT
MOTLEY, District Judge.INTRODUCTION
Plaintiffs filed this class action suit alleging various forms of gender discrimination by defendant Smith Barney, Inc. (“Smith Barney”) and two Smith Barney officers, James Dimon and Nicholas Cuneo. Plaintiffs also challenge the compulsory arbitration policies of defendants Smith Barney, the New York Stock Exchange (“NYSE”), and the National Association of Securities Dealers (“NASD”). Plaintiffs and Smith Barney reached a settlement providing revised arbitration procedures to resolve class members’ claims against Smith Barney and implementing four years of diversity initiatives at Smith Barney. For the reasons below, the court denies final approval, finding that the settlement is not fair, reasonable, and adequate.
BACKGROUND
On May 20, 1996, three plaintiffs filed this action on behalf of a nationwide class in the *250Southern District of New York.1 Plaintiffs’ complaint alleges that Smith Barney engaged in, and permitted, a long-term, company-wide pattern of gender-based discrimination, harassment, and retaliation in violation of various federal discrimination and employment laws (Title VII of the Civil Rights Act of 1964, as amended, 42 U.S.C. § 2000e et seq. (“Title VII”); the Equal Pay Act of the Fair Labor Standards Act (“FLSA”), 29 U.S.C. §§ 206-07, 215; and the Family and Medical Leave Act, 29 U.S.C. § 2601 et seq.), state and local discrimination laws (the New York State Human Rights Law, N.Y. Exec. Law § 296 et seq.; and the Administrative Code of the City of New York § 8-17 et seq.), and New York State common law.
Plaintiffs also challenge the defendants’ practice of conditioning their employment on compulsory arbitration of all employment-related claims, including discrimination claims. They allege that the securities industry’s Uniform Application for Securities Industry Registration-or Transfer, commonly known as the U-4 form, violates Title VII and, because NYSE and NASD are state actors, the Fifth Amendment Due Process Clause.
By September 1996, defendants had moved to dismiss pursuant to Fed.R.Civ.P. 12(b)(1) and 12(b)(6), defending the practice of compulsory arbitration and arguing that the arbitration provisions deprive the federal courts of subject matter jurisdiction over plaintiffs’ claims. Both parties now have fully briefed the motions, but the court has not decided them yet. Also in September 1996, plaintiffs began discovery focusing on the propriety of nationwide class certification. Discovery yielded comprehensive information regarding: (1) all employees in Smith Barney’s Retail Sales, Capital Markets, and Investment Banking Divisions during 1994 and 1995 (numbering over 30,000); (2) all complaints of sexual harassment, discrimination, or retaliation in those divisions, and their outcomes, from August 1993 to December 1995; and (3) all applications for Smith Bar-ne/s Financial Consultant Training Program during 1994 and 1995. Class counsel have retained a statistician/economist to assist in analyzing these data. See Stipulation of Settlement, §§ 1.5-1.6,1.8. (“Settlement ”).
On July 7-8 and 26-27, 1997, plaintiffs and defendants Smith Barney and James Dimon engaged in formal mediation, jointly retaining a mediation firm experienced in employment discrimination cases, including class actions. See Settlement §§ 1.9, 1.11. After mediation, the parties reached a settlement which they filed with the court on November 18, 1997. See Settlement. The court issued an order preliminarily approving the settlement, conditionally certifying the class, providing for notice, and scheduling a fairness hearing, see Order dated November 21, and then approved summary notice of pendency of class action, certification of class, settlement, and fairness hearing, see Order dated December 9, 1997. The class as preliminarily approved consists of “All women employed by Smith Barney Inc. in its domestic (including Puerto Rico) Retail Sales, Investment Banking or Capital Markets Divisions at any time between May 20, 1993 through and including November 21, 1997 (except women who become or have become or may be deemed to be employed by Smith Barney Inc. solely by virtue of Smith Barney Inc.’s merger or other combination with Salomon Brothers).” See id. at 2. The parties now move for a final order certifying the class, approving the settlement, and dismissing the claims.
DISCUSSION
I. Subject Matter Jurisdiction
The first question is whether the court has subject matter jurisdiction. While Smith Barney no longer asserts that the compulsory arbitration provision deprives plaintiffs of federal claims and eliminates subject matter jurisdiction, that does not resolve the issue. The court can only proceed *251to examine whether the parties’ settlement is fair, adequate, and reasonable, and to decide what post-settlement jurisdiction the court retains, if it has subject matter jurisdiction over plaintiffs’ suit. See Louisville & Nashville R.R. v. Mottley, 211 U.S. 149, 29 S.Ct. 42, 53 L.Ed. 126 (1908) (finding judicial duty to determine subject matter jurisdiction sua sponte); Tongkook Am., Inc. v. Shipton Sportswear Co., 14 F.3d 781, 786 (2d Cir. 1994) (“that [defendant] did not raise the defense of subject matter jurisdiction is not determinative, since subject matter jurisdiction cannot be waived and the issue can be raised at any time in the course of litigation”).
A valid, applicable compulsory arbitration provision would eliminate subject matter jurisdiction, barring plaintiffs’ suit from any non-arbitral forum. Compare State of N.Y. v. Oneida Indian Nation of N.Y., 90 F.3d 58 (2d Cir.1996) (finding subject matter jurisdiction where claim fell within exception to arbitration agreement) with Desiderio v. Nat’l Assoc. of Sec. Dealers, Inc., 2 F.Supp.2d 516 (S.D.N.Y.1998) (finding no subject matter jurisdiction where arbitration agreement was valid and applicable to plaintiffs claim). The court is not ruling on the merits of any party’s arguments regarding compulsory arbitration here.2 Yet it can only find subject matter jurisdiction to conduct the settlement inquiry if the validity or application of the arbitration agreement is sufficiently unclear that plaintiffs’ claims present questions properly in federal court.
In deciding whether an arbitration agreement precludes a federal forum, the court must consider four issues under Genesco v. T. Kakiuchi & Co., Ltd., 815 F.2d 840 (2d Cir.1987):
first, it must determine whether the parties agreed to arbitrate!;] ■ • • second, it must determine the scope of that agreement; third, if federal statutory claims are asserted, it must consider whether Congress intended those claims to be nonarbi-trablef;] ... fourth, if the court concludes that some, but not all, of the claims in the case are arbitrable, it must then determine whether to stay the balance of the proceedings pending arbitration. Id. at 844 (citation omitted).3
A. Presence of an Agreement to Arbitrate
The first Genesco issue is uncomplicated here, as the U-4 is an explicit arbitration agreement. It appears sufficiently knowing and voluntary to satisfy any requirements for waivers of Title VII protections, see, e.g., Laniok v. Advisory Comm., 935 F.2d 1360, 1365 (2d Cir.1991); Kristoferson v. Otis Spunkmeyer, Inc., 965 F.Supp. 545, 546-47 (S.D.N.Y.1997); Rangel v. El Paso Natural Gas Co., 996 F.Supp. 1093 (D.N.M.1998), and for waivers of judicial fora for Title VII claims, see, e.g., Nelson v. Cyprus Bagdad Copper Corp., 119 F.3d 756, 760-61 (9th Cir. 1997); Phox v. Allied Capital Advisers, Inc., Civil Action No. 96-2745(JR), 1997 WL 198115 at *1 (D.D.C. Apr. 14, 1997).
B. Scope of the Agreement to Arbitrate: Excluding Class Actions
In deciding the second Genesco issue, whether the agreement’s scope encompasses plaintiffs’ claims, the court must “ ‘construe arbitration clauses as broadly as possible.’” Oldroyd v. Elmira Sav. Bank, FSB, 134 F.3d 72, 76 (2d Cir.1998) (quoting Collins & Aikman Prod. Co. v. Bldg. Sys., 58 F.3d 16, 19 (2d Cir.1995)). This requires “ ‘resolving any doubts concerning the scope of arbitrable issues ... in favor of arbitration,’ ” Oldroyd, 134 F.3d at 76 (quoting Moses H. Cone Mem’l Hosp. v. Mercury Constr. Corp., 460 U.S. 1, 24-25, 103 S.Ct. 927, 74 L.Ed.2d 765 (1983)), because of the “strong federal policy favoring arbitration as an al*252ternative means of dispute resolution.” Oldroyd, 134 F.3d at 76.
Still, “ ‘[t]he scope of an arbitration clause, like any other contract provision, is a question of the intent of the parties,’ ” Haviland v. Goldman, Sachs & Co., 947 F.2d 601, 604 (quoting S.A. Mineracao da Trindade-Samitri v. Utah Int’l, Inc., 745 F.2d 190, 193 (2d Cir.1984)), and “ ‘federal policy alone cannot be enough to extend the application of an arbitration clause far beyond its intended scope.’ ” Haviland, 947 F.2d at 605 (quoting McDonnell Douglas Fin., Corp. v. Pennsylvania Power & Light Co., 858 F.2d 825, 831 (2d Cir.1988)). Rather, federal policy emphasizes respect for the parties’ expressed intentions and reasonable expectations regarding the scope of their arbitration agreements. While one of the “guiding principles of the law of arbitration ... requires full enforcement of arbitration clauses, ... [a] second principle protects parties from being compelled to arbitrate claims they did not agree to arbitrate.” State of N.Y. v. Oneida Indian Nation of N.Y., 90 F.3d 58, 59 (2d Cir.1996).
Apart from the plaintiffs’ and Smith Barney’s expressed present intent, the U-4 arbitration provision excludes class actions, as courts recognize. See, e.g., Nielsen v. Piper, Jaffray & Hopwood, Inc., 66 F.3d 145 (7th Cir.1995) (finding class action claim outside scope of arbitration agreement because of NASD and NYSE rules barring arbitration of such claims but recognizing court adjudication of class actions); Scher v. Equitable Life Assurance Soc’y of the United States, 866 F.Supp. 776 (S.D.N.Y.1994) (noting the propriety of a federal forum where plaintiffs signed U-4 forms because class action suits are not subject to NASD arbitration). Compelling arbitration of plaintiffs’ class action would frustrate their reasonable expectation that the U-4 did not cover class actions. Moreover, even if plaintiffs’ claims initially were inside the scope of the U-4, the settlement brings it outside, amending it to exclude plaintiffs’ class action. While the parties cannot waive a bar to subject matter jurisdiction, they can contract around an existing contractual bar to subject matter jurisdiction, as they did in the settlement.
C. Congressional Intent: Barring Prospective Agreements to Arbitrate Title VII Claims
1. Rebuttable Presumption of Validity for Agreements to Arbitrate
The third Genesco issue, where plaintiffs assert federal statutory claims, is whether Congress intended to limit agreements to arbitrate those claims. Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 111 S.Ct. 1647, 114 L.Ed.2d 26 (1991) (discussing claims under the Age Discrimination in Employment Act (“ADEA”)), shapes this inquiry as a rebuttable presumption that Congress allows such agreements: “statutory claims may be the subject of an arbitration agreement ... ‘unless Congress itself has evinced an intention to preclude a waiver of judicial remedies for the statutory rights at issue.’ ” Id. at 26, 111 S.Ct. 1647 (quoting Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 628, 105 S.Ct. 3346, 87 L.Ed.2d 444 (1985)). In rebutting this presumption, “the burden is on [plaintiff] to show that Congress intended to preclude a waiver of a judicial forum---- If such an intention exists, it will be discoverable in the text of the [statute], its legislative history, or an ‘inherent conflict’ between arbitration and the [statute’s] underlying purposes.” Gil-mer, 500 U.S. at 26, 111 S.Ct. 1647 (citing Shearson/Am. Express v. McMahon, 482 U.S. 220, 227,107 S.Ct. 2332, 96 L.Ed.2d 185 (1987)). See also, e.g., Oldroyd v. Elmira Sav. Bank, FSB, 134 F.3d 72, 76-78 (2d Cir.1998) (applying Gilmer to the third Gen-esco issue).
Careful analysis of legislative intent is particularly appropriate for the broad inquiry Gilmer requires into the statutory “text[,] ... legislative history, or ... underlying purposes.” Gilmer, 500 U.S. at 26, 111 S.Ct. 1647 (citation omitted). Where, as here, the issue is whether Congress intended to alter the balance between two broad, important statutory priorities (arbitration and civil rights), courts cannot assume the older arbitration statute takes priority. Thus, the court’s inquiry cannot insist that the newer civil rights statute must feature the sort of *253“ ‘unequivocal expression of congressional intent’” required when the issue is whether Congress intended to alter “the fundamental constitutional balance between the Federal Government and the States .” Atascadero State Hosp. & Cal. Dep’t of Mental Health v. Douglas James Scanlon, 473 U.S. 234, 238, 105 S.Ct. 3142, 87 L.Ed.2d 171 (1985) (requiring, to abrogate state sovereign immunity, clear statutory language) (citation omitted); see also Tafflin v. Levitt, 493 U.S. 455, 459-60, 110 S.Ct. 792, 107 L.Ed.2d 887 (1990) (requiring, to deny state courts concurrent jurisdiction, “ ‘explicit statutory directive, ... unmistakable implication from legislative history, or ... clear incompatibility between state-court jurisdiction and federal interests’ ”) (citation omitted). Accordingly, in legislative intent inquiries into arbitration of statutory rights, the Second Circuit has recognized that “we deal with two federal statutory schemes: one for encouraging voluntary arbitration of labor disputes and the other for eliminating discrimination in the workplace. These competing federal policies are not mutually exclusive; rather, they must be balanced.” Bates v. Long Island R.R. Co., 997 F.2d 1028, 1035 (2d Cir.1993) (balancing Railway Labor Act arbitration provisions and Rehabilitation Act antidiscrimination rights and procedures); accord Rosenberg v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 995 F.Supp. 190, 191-92 (D.Mass.1998).
Consequently, a majority of circuits, however, including the Second, only determine Congressional intent after substantive inquiry as to Congressional intent underlying the particular statute at issue. See Seus v. John Nuveen & Co., Inc., 146 F.3d 175 (3d Cir. 1998) (Title VII); Duffield v. Robertson Stephens & Co., 144 F.3d 1182 (9th Cir.1998) (Title VII); Oldroyd v. Elmira Sav. Bank, FSB, 134 F.3d 72, 76-78 (2d Cir.1998) (Financial Institutions Reform, Recovery, and Enforcement Act); Bercovitch v. Baldwin Sch., Inc., 133 F.3d 141 (1st Cir.1998) (Americans with Disabilities Act); Patterson v. Tenet Healthcare, Inc., 113 F.3d 832 (8th Cir.1997) (Title VII); Cole v. Burns Int'l Sec. Svcs., 105 F.3d 1465 (D.C.Cir.1997) (Title VII); Austin v. Owens-Brockway Glass Container, 78 F.3d 875 (4th Cir.1996) (Title VII).
The need for a meaningful legislative intent inquiry finds support not only in the text of Gilmer, but also in the fact that federal law limits waiver of many rights it grants. While individual rights are by default alienable, some are only partially alienable, and others are inalienable. See generally Guido Calabresi & A. Douglas Melamed, Property Rules, Liability Rules, and Inalienability: One View of the Cathedral, 85 Harv. L.Rev. 1089 (1972). Title VII aims to benefit all of society by eliminating systemic labor market inequality, which limits alienability of its protections because “rights that are relational and systemic are necessarily inalienable; individuals cannot waive them because individuals are not their sole focus.” Laurence H. Tribe, The Abortion Funding Conundrum: Inalienable Rights, Affirmative Duties, and the Dilemma of Dependence, 99 Harv. L.Rev. 330, 333 (1985); see also Margaret Jane Ra-din, Market-Inalienability, 100 Harv. L.Rev. 1849,1863 (noting that even a restricted view of inalienability considers it “a means of controlling externalities that prevent the market from achieving an efficient result”). The Title VII premise of systemic labor market disempowerment also supports inalienability. In settings of unequal bargaining power, waivers of individual rights may reflect not efficient decision-making, but inefficient “[ejconomic rents” providing windfalls to the parties in superior bargaining positions. Susan Rose-Ackerman, Inalienability and the Theory of Property Rights, 85 Colum. L.Rev. 931, 940-41 (1985).
The Second Circuit has found federally-granted rights inalienable in contexts presenting these concerns. Individuals cannot resolve personal claims in ways threatening others’ Title VII rights. See Newsday, Inc. v. Long Island Typographical Union, No. 915, CWA, AFL-CIO, 915 F.2d 840, 844-45 (2d Cir.1990) (invalidating arbitral reinstatement of sexual harasser). Firms, in turn, “may not require their employees to waive arbitration rights,” even if their bargaining power yields that outcome, where federal law aims to ensure such rights. See, e.g., Thomas James Assoc., Inc. v. Jameson, 102 F.3d 60, 66 (2d Cir.1996) (invalidating such waiver among NASD mem*254ber firms subject to SEC-approved arbitration rights regulations).
The Second Circuit has not ruled on the specific question of Congressional intent regarding prospective agreements to arbitrate Title VII claims. See Desiderio v. Nat’l Assoc., of Sec. Dealers, Inc., 2 F.Supp.2d 516 (S.D.N.Y.1998). Other circuits are split on such agreements, with some allowing them categorically, see, e.g., Alford v. Dean Witter Reynolds, Inc., 939 F.2d 229 (5th Cir.1991), others allowing them only subject to certain requirements, see, e.g., Cole v. Burns Int’l Sec. Svcs., 105 F.3d 1465 (D.C.Cir.1997), and others not allowing them at all, see, e.g., Duffield v. Robertson Stephens & Co., 144 F.3d 1182, 1998 WL 227469, No. 97-15698 (9th Cir. May 8,1998). The presence of this split, along with the Second Circuit’s recognition that some federally-granted rights are inalienable, requires the court to engage in a meaningful analysis of the extent to which Congress intended that Title VII rights be alienable and of the case law addressing that issue.
2. Imperfect Similarity Between the ADEA and Title VII
A minority of circuits have held that the Gilmer approval of prospective agreements to arbitrate ADEA claims forecloses rebuttal of the presumption that Title VII also allows such agreements. See Metz v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 39 F.3d 1482 (10th Cir.1994); Bender v. A.G. Edwards & Sons, Inc., 971 F.2d 698 (11th Cir.1992); Willis v. Dean Witter Reynolds, Inc., 948 F.2d 305 (6th Cir.1991); Alford v. Dean Witter Reynolds, Inc., 939 F.2d 229 (5th Cir.1991). These four circuits have applied Gilmer so as to declare arbitration agreements valid in contexts very different from Gilmer, an opinion that carefully considered the statute at issue (the ADEA), see Gilmer, 500 U.S. at 26-29, 111 S.Ct. 1647, and contract at issue, see id. at 33-35, 111 S.Ct. 1647 (distinguishing Alexander v. Gardner-Denver Co., 415 U.S. 36, 94 S.Ct. 1011, 39 L.Ed.2d 147 (1974), which barred collective bargaining agreements from making arbitration an exclusive forum for Title VII claims).
Certainly, Title VII and the ADEA are similar laws providing similar remedies for similar social problems, and it would be reasonable for Congress to make either ADEA and Title VII rights inalienable. Yet the presence of similar rationales for inalienability does not mean that Congress actually enacted identical degrees of inalienability. Title VII and the ADEA are different statutes that are not mirror images. In many ways, protections against age discrimination are not fully coextensive with remedies for race and gender discrimination.
Fox example, the Supreme Court held that the ADEA provided a right to a jury trial when Title VII did not because while “[t]here are important similarities between the two statutes,” in comparing rights under each, “it is the remedial and procedural provisions that are crucial and there we find significant differences.” Lorillard v. Pons, 434 U.S. 575, 584, 98 S.Ct. 866, 55 L.Ed.2d 40 (1978). Additionally, in apparent contrast to the strong Title VII disparate impact doctrine, the ADEA statutorily permits decisions based on a “reasonable factor other than age,” 29 U.S.C. § 623(f). This “suggests that decisions which are made for reasons independent of age but which happen to correlate with age are not actionable under the ADEA,” Equal Employment Opportunity Comm’n v. Francis W. Parker Sch., 41 F.3d 1073, 1077 (7th Cir.1994), and leaves (“ ‘considerable doubt as to whether a claim of age discrimination may exist under a disparate-impact theory’ ”), Gantt v. Wilson Sporting Goods Co., 143 F.3d 1042, 1048 (6th Cir.1998). While the circuits split on ADEA disparate impact claims, see, e.g., Dist. Council 37 v. N.Y. City Dep’t of Parks and Recreation, 113 F.3d 347, 351 (2d Cir.1997) (allowing ADEA disparate impact claims), the Supreme Court has declined to resolve this split, leaving ADEA and Title VII protections substantively different. See Hazen Paper Co. v. Biggins, 507 U.S. 604, 610, 113 S.Ct. 1701, 123 L.Ed.2d 338 (1993) (“we have never decided whether a disparate impact theory of liability is available under the ADEA ... and we need not do so here”).
The strong Title VII disparate impact doctrine is relevant as an example not just of a *255possible difference from the ADEA, but also of a possible incompatibility between Title VII and compulsory arbitration. To the extent that compulsory arbitration of Title VII claims disproportionately affects women and minorities, it could give rise to disparate impact claims against compulsory arbitration. Whether or not a disparate impact claim would win is irrelevant; the issue here is simply the conflict between compulsory arbitration of Title VII claims and the public policy against employment practices disproportionately burdening women or minorities. Arbitration is favored, but not to the point that it can run contrary to the public policy underlying Title VII. See, e.g., Newsday, Inc. v. Long Island Typographical Union, No. 915, CWA, AFL-CIO, 915 F.2d 840, 844-45 (2d Cir.1990) (invalidating arbitral reinstatement of sexual harasser).
Thus, despite the many similarities between the ADEA and Title VII, the outer perimeters of their procedural and substantive protections vary, making it perfectly plausible that Title VII limits alienability of its protections more than the ADEA does. More generally, as society expands the scope of antidiscrimination protections past the initial bar against racism, courts cannot cursorily assume that identical protections apply to different statutorily- or constitutionally-protected classes. See John J. Donohue III, Employment Discrimination Law in Perspective: Three Concepts of Equality, 92 Mich. L.Rev. 2583 (1994); cf. United States R.R. Retirement Bd. v. Fritz, 449 U.S. 166, 101 S.Ct. 453, 66 L.Ed.2d 368 (1980) (finding that, in contrast to race and gender discrimination, age discrimination only faces rational basis scrutiny under the Equal Protection Clause). Established interpretations of the ADEA and its legislative history thus do not preclude contrary interpretations of Title VII and its legislative history.
3. Congressional Intent to Limit Alien-ability of Title VII Protections
a. Arbitration-Specific Limitations: Adequate Forum for Statutory Redress
Just as Congress can bar arbitration agreements entirely, it can take the lesser step of raising the floor arbitral requirements above the FAA minima. Even absent a general “suspicion of arbitration,” Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 30, 111 S.Ct. 1647, 114 L.Ed.2d 26 (1991), courts may not find that all arbitral forums will meet the premise of judicial permission for prospective agreements to arbitrate statutory claims: “[s]o long as the prospective litigant effectively may vindicate ... statutory eause[s] of action in the arbitral forum, the statute will continue to serve both its remedial and deterrent function.” Id. at 28, 111 S.Ct. 1647 (quoting Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 628, 105 S.Ct. 3346, 87 L.Ed.2d 444 (1985)). Because it rests on these premises, “[ojbviously, Gilmer cannot be read as holding that an arbitration agreement is enforceable no matter what rights it waives or what burdens it imposes.” Cole v. Burns Int'l Sec. Svcs., 105 F.3d 1465, 1482 (D.C.Cir.1997).
Plaintiffs may arbitrate Title VII claims only in an arbitral forum that meets three requirements to ensure that the arbitration will not subvert the Title VII statutory scheme. First, the arbitration must meet certain standards of procedural fairness. See, e.g., Cole, 105 F.3d at 1468-69 (D.C.Cir. 1997) (requiring “reasonable right of access to a neutral forum” and meaningful review for Title VII claims); Ramirez-De-Arellano v. Am. Airlines, Inc., 133 F.3d 89, 91 (1st Cir.1997) (requiring notice, discovery, unbiased arbitrators, and meaningful review for FLSA claims); Patterson v. Tenet Healthcare, Inc., 113 F.3d 832. 837 (8th Cir.1997) (requiring “neutral arbitrators, adequate discovery” for Title VII claims). In a different statutory context, the Second Circuit has found similar requirements. Noting the limited discovery and relaxed evidence rules in arbitration under a Railway Labor Act collective bargaining agreement, the court held that “[ajbsent the same rights and pi’ocedures provided in federal court, arbitration should not be the sole forum for final resolution of civil rights claims.” Bates v. Long Island R.R. Co., 997 F.2d 1028, 1035 (2d Cir.1993) (Rehabilitation Act claim).
Second, the arbitration agreement cannot impose financial burdens on plaintiff access *256to the arbitral forum. Requiring plaintiffs to pay for access “would surely deter the bringing of arbitration,” running counter to Congressional intent underlying Title VII. Cole, 105 F.3d at 1468. This principle bars not only “steep filing fees,” but also “fee-shifting” agreements in which plaintiffs pay part of the arbitrator’s fee. Paladino v. Avnet Comp. Techs., Inc., 134 F.3d 1054, 1062 (11th Cir.1998) (Cox, J, concurring, for a majority of the court) (finding unenforceable an arbitration agreement imposing on plaintiffs a $2000 filing fee and possible liability for a portion of the arbitrator’s fee); cf. Rangel v. El Paso Natural Gas Co., 996 F.Supp. 1093 (D.N.M.1998) (finding unenforceable a “tender back” agreement requiring plaintiffs to return severance pay before arbitrating Title VII claims). Arbitrator’s fees are the responsibility of the employer because “it would undermine Congress’s intent to prevent employees ... from gaining access to a judicial forum and then require them to pay for the services of an arbitrator when they would never be required to pay for a judge in court.” Cole, 105 F.3d at 1484; see also Shankle v. B-G Maintenance Management of Colo., Inc., 1997 WL 416405, No. 96 N 2932 (D.Colo. Mar. 24, 1997) (ruling the same). Analogously, non-financial barriers can inhibit access impermissibly when the statute’s prescribed limitations period is “important to the effectuation of [its] policies.” Graham Oil Co. v. Arco Prods. Co., 43 F.3d 1244, 1248 (9th Cir.1995) (finding unenforceable an arbitration agreement shortening the limitations period for Petroleum Marketing Practices Act claims).
Third, the arbitration must allow remedies central to the statutory scheme. Arbitration need not provide exactly the same remedies as adjudication, but must provide remedies sufficient to satisfy statutory purposes. See, e.g., Roby v. Corp. of Lloyd’s, 996 F.2d 1353, 1365 (2d Cir.1993) (affirming arbitration after concluding that foreign arbitral forum’s different remedies met statutory purposes of deterrence because “the available remedies are adequate and the potential recoveries substantial”). No statutes require arbitrations to mimic all the procedural formalities of federal courts, but under Title VII, arbitrations cannot disallow remedies central to the Title VII scheme. Arbitration agreements prospectively waiving such central remedies are unenforceable. See, e.g., Paladino, 134 F.3d at 1059 (holding that for Title VII claims, arbitration must offer “all of the types of relief that would otherwise be available in court”) (quoting Cole, 105 F.3d at 1482). Arbitration cannot be the exclusive Title VII forum where it provides oply contract damages, not the broader range of Title VII monetary and equitable remedies. Pala-dino, 134 F.3d at 1060 (finding agreement “fundamentally at odds with the purposes of Title VII because it completely proscribes an arbitral award of Title VII damages”). Particularly important is that arbitration award attorney fees to prevailing parties; arbitration agreements providing otherwise are unenforceable as against the public policy of Title VII. See DeGaetano v. Smith Barney, Inc., 983 F.Supp. 459 (S.D.N.Y.1997). Other statutes similarly require that arbitration allow the statutorily-provided exemplary damages. See, e.g., Graham Oil Co., 43 F.3d 1244 (9th Cir.1994) (Petroleum Marketing Practices Act); cf. Bates, 997 F.2d 1028, 1035 (2d Cir.1993) (finding that collective bargaining agreement to arbitrate civil rights claims must provide “the same rights and procedures provided in federal court”).
While these requirements do not bar all compulsory arbitration agreements, they do provide plaintiffs potential federal challenges to their agreements. Gilmer upheld such an agreement by rejecting any presumption of arbitral inadequacy and finding that plaintiffs had not met their burden of proving such inadequacy in their specific situation. See Gilmer, 500 U.S. at 30-33, 111 S.Ct. 1647. This does not collaterally estop future plaintiffs from making such showings, however. See, e.g., Rosenberg v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 995 F.Supp. 190, 205-07 (D.Mass.1998) (finding that plaintiffs made such individualized showings against the same NYSE arbitration the Gilmer plaintiffs challenged unsuccessfully). Because the present plaintiffs could follow in the footsteps of the Rosenberg plaintiffs, whether or not successfully, their U-4 forms do not clearly eliminate the federal questions their claims present.
*257b. Prospectivity-Based Limitations:
No Compulsory Arbitration
In the years immediately following the Gilmer declaration that Title VII claimants are subject to compulsory arbitration agreements unless they prove contrary Congressional intent, plaintiffs failed to meet that burden and courts properly enforced their agreements. See, e.g., Mago v. Shearson Lehman Hutton, Inc., 956 F.2d 932 (9th Cir.1992); Maye v. Smith Barney, Inc., 897 F.Supp. 100 (S.D.N.Y.1995). Since 1995, however, plaintiffs in scattered cases have begun to make the necessary showings, leading courts to find that Congress did intend to bar prospective agreements to arbitrate Title VII claims. See Duffield v. Robertson Stephens & Co., 144 F.3d 1182 (9th Cir.1998); Rosenberg v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 995 F.Supp. 190 (D.Mass. 1998). The cases finding partial restrictions on compulsory arbitration of Title VII claims rest on similar legislative analyses that find Congressional intent to limit Title VII alien-ability. See, e.g., Cole v. Burns Int’l Security Svcs., 105 F.3d 1465 (D.C.Cir.1997); Paladino v. Avnet Comp. Techs., Inc., 134 F.3d 1054 (11th Cir.1998); DeGaetano v. Smith Barney, Inc., 983 F.Supp. 459 (S.D.N.Y. 1997). The failures of early post-Gilmer plaintiffs to meet their burdens of proof, see, e.g., Maye, 897 F.Supp. 100 (finding that plaintiff who argued only that he did not knowingly and voluntarily agree to arbitrate failed to meet his burden), do not collaterally estop all future plaintiffs from meeting that burden.
In amending Title VII with the Civil Rights Act of 1991 (“1991 Act”), Congress gave a qualified approval to arbitration of Title VII claims: “Where appropriate and to the extent authorized by law, the use of alternative dispute resolutions including ... arbitration is encouraged to resolve disputes.” Pub.L. 102-166, § 118, reprinted in notes to 42 U.S.C. § 1981 (emphasis added). In interpreting these two ambiguous qualifiers, the court must “ ‘assum[e] that Congress used two terms because it intended each term to have a particular, nonsuperflous meaning’ ” and thus that the two “set forth separate and distinct limitations on the eondi-tions and circumstances under which the arbitration process may be invoked to resolve Title VII claims.” Duffield, 144 F.3d at 1193 (quoting Bailey v. United States, 516 U.S. 137, 146, 116 S.Ct. 501, 133 L.Ed.2d 472 (1995)) (other citations omitted).
One key distinction between the two statutory qualifiers is that even where “authorized by law,” arbitration is not always “appropriate” to “further[ ] the purpose and objective of the Act.” Duffield, 144 F.3d at 1194. The legislative history of the 1991 Act clarifies that prospective agreements for compulsory arbitration of Title VII claims are what Congress found not “appropriate,” regardless of whether otherwise “authorized by law .” The historical backdrop is that § 118 was drafted before Gilmer, when Alexander was a clearer prohibition on compulsory arbitration of Title VII claims. See Duffield, 144 F.3d at 1194- 95. The House Committee Report, a preferred source of legislative history, see Garcia v. United States, 469 U.S. 70, 76, 105 S.Ct. 479, 83 L.Ed.2d 472 (1984) (citation omitted), states its rejection of prospective agreements to arbitrate Title VII claims in terms that could not be clearer:
The Committee emphasizes ... that the use of alternative dispute mechanisms is ... intended to supplement, not supplant, the remedies provided by Title VII.... [A]ny agreement to submit disputed issues to arbitration ... in an employment contract, does not preclude ... seeking relief under the enforcement provisions of Title VII. This view is consistent with the Supreme Court’s interpretation of Title VII in Alexander____The Committee does not intend this section to be used to preclude rights and remedies that would otherwise be available. H.R.Rep. No.40(I) at 97.
Congress seconded this point in the Conference Committee Report, see H.R. Conf. Rep. No. 755,101st Cong., 2d Sess. 26 (1990) (“any agreement to submit disputed issues to arbitration, ... in an employment contract, does not preclude ... seeking relief under the enforcement provisions of Title VII”), as well as in “rejecting] a proposal that would have allowed employers to enforce ‘compulsory arbitration’ agreements,” Duffield, 144 F.3d at 1195-96 (citing H.R.Rep. No. 40(1) at 104).
*258Whatever the occasional shortcomings of legislative history, here it makes clear that Congress intended to prevent otherwise-encouraged means of alternative dispute resolution from denying Title VII claimants their day in federal court. Congress did not weaken its statement of this principle by citing Alexander as support. Regardless of Gil-mer’s effect on Alexander, for courts examining what Congress meant to enact in 1991 ‘“the relevant inquiry is not whether Congress correctly perceived the state of the law, but rather what its perception of the law was.’ ” Duffield, 144 F.3d at 1196-97 (quoting Brown v. GSA, 425 U.S. 820, 96 S.Ct. 1961, 48 L.Ed.2d 402 (1976)). Moreover, because of the differences between the ADEA and Title VII, see supra Part I.C.2, Congress’s perception that Alexander supports limited alienability of Title VII protections was not inconsistent with Gilmer. See Rosenberg, 995 F.Supp. at 204, (“it was not at all clear that Congress needed to overrule a case interpreting the ADEA’s enforcement procedures in order to preserve access to federal courts under Title VII”).
Even if textual ambiguity and quirks of chronology leave some degree of uncertainty, this does not justify ignoring evidence of Congressional intent sufficiently persuasive for the Gilmer/Genesco inquiry. That inquiry does not require “unequivocal” evidence of intent, see supra Part I.C.l. (contrasting standards for legislative history inquiries), which is especially true because the statute at issue is Title VII. Both the 1991 Act’s legislative history and canons governing statutory interpretation of Title VII instruct that “ “when the statutory terms ... are susceptible to alternate interpretations, the courts are to select the construction which most effectively advances the underlying congressional purpose.’” Duffield, 144 F.3d at 1192-93 (quoting H.R.Rep. No. 40(1) at 88) (citing Dennis v. Higgins, 498 U.S. 439, 443, 111 S.Ct. 865, 112 L.Ed.2d 969 (1991) (noting that courts should interpret civil rights statutes broadly)).
The court now finds that plaintiffs could meet their burden of proving that in passing Title VII and its amendments, Congress intended to prohibit prospective agreements to arbitrate Title VII claims. Prior cases to the contrary, even if binding on particular points of law or on their parties, do not collaterally estop more detailed examinations of Congressional intent, which find support in the recent ease law. Thus, plaintiffs could show that their U-4 agreements do not validly compel Title VII class action arbitration by making proper arguments regarding particular arbitral fora, see supra Part I.C.3.a, or Title VII generally.
Accordingly, the court finds subject matter jurisdiction over plaintiffs’ claims. In the Genesco inquiry framework, there is clearly an arbitration agreement here (first Genesco issue), but its applicability (second issue) and validity (third issue) are uncertain as matters of fact and law. Whether the agreement bars plaintiffs’ claims from a federal forum thus remains a federal question without a clear answer, providing the subject matter jurisdiction necessary for judicial evaluation of the settlement the parties present.
II. Class Certification
A. Rule 23(a)
Parties seeking class certification must satisfy the four prerequisites of Fed. R.Civ.P. 23(a): numerosity, commonality, typicality, and adequacy of representation. Here, with notice by mail alone targeting over 22,496 women employed by Smith Barney, see Aff. of Michael Rosenbaum, Fairness Hearing Ex. B ¶ 3-4, numerosity is clear. See, e.g., Consol. Rail Corp. v. Town of Hyde Park, 47 F.3d 473, 483 (2nd Cir.1995) (presuming numerosity at 40).
Commonality and typicality “tend to merge into one another, so that similar considerations animate analysis.” Marisol A. v. Giuliani 126 F.3d 372, 376 (2d Cir.1997) (citing Gen. Tel. Co. of Southwest v. Falcon, 457 U.S. 147, 157 n. 13, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982)) (other citations omitted). Commonality looks to whether there are common issues of law or fact among plaintiffs, see, e.g., In re Agent Orange Prod. Liab. Litig., 818 F.2d 145, 166-167 (2d Cir. 1987), while typicality looks to whether the class representatives’ claims are typical of the class members’ in the sense of challeng*259ing the same practices and making similar legal arguments, see, e.g., In re Drexel Burnham Lambert Group, Inc., 960 F.2d 285, 291 (2d Cir.1992).
In a discrimination class action, the court cannot presume that all the plaintiffs’ allegations of discrimination against same statutorily protected group are common and typical. Rather, the court must find the plaintiffs’ discrimination claims sufficiently similar that “maintenance of a class action is economical and [that] the named plaintiffs claim and the class claims are so interrelated that the interests of the class members will be fairly and adequately protected in their absence.” Falcon, 457 U.S. at 157, n. 13, 102 S.Ct. 2364. See also Rossini v. Ogilvy & Mather, Inc., 798 F.2d 590, 596-598; Am. Fed’n of State, Cty., & Mun. Employees, AFL-CIO v. County of Nassau, 664 F.Supp. 64, 68 (E.D.N.Y.1987) (“‘Rule 23 does not require precise, mirror-image identity’ ” of claim and injury, nor that “thousands of potential class members ... be clones of these representatives.”) (citation omitted).
Among the present class, the representative plaintiffs’ claims are typical, and feature common issues of law and fact. The various plaintiffs allege essentially similar wrongdoing: a pattern of discriminatoiy treatment of women that results in entry barriers, sexual harassment, and glass ceilings, and that traces to a central failure by Smith Barney’s management and human resources department to prevent and remedy widespread discrimination. Class certification is appropriate if the claims entail common issues and the injuries trace to the central employer’s failings. This holds even where plaintiffs suffer varying forms of discrimination, see, e.g., Rossini, 798 F.2d at 598-599 (centralized failures to provide nondiscriminatory job criteria that resulted in different injuries due to different failures), where some plaintiffs hold managerial jobs while others do not, see id. at 594-596, where plaintiffs assert diverse legal causes of action, see Marisol A., 126 F.3d at 376-377, or where some plaintiffs apply for positions while others do not, see Grant v. Bethlehem Steel Corp., 823 F.2d 20, 24 (2d Cir.1987).
The adequacy of representation inquiry looks both to the ability of class counsel and to the potential for conflict of interest between the representative plaintiffs and the rest of the class. See, e.g., Marisol A., 126 F.3d at 378. Mary Stowell, one of the class attorneys from the firm Stowell, Friedman & Vernon, has documented her experience in employment discrimination litigation, including class actions lawsuits. See Decl. of Mary Stowell, Fairness Hearing Ex. A H 41. Class counsel also have proven themselves able thus far in this case with, for example, their discovery efforts, see Settlement, §§ 1.5 — 1.6, 1.8. (obtaining information on female employment and discrimination claims at Smith Barney and retaining expert assistance for data analysis), their motion practice to date (responding to dismissal motions and settlement objections), and their settlement negotiations. See Decl. of Mary Stowell, Fairness Hearing Ex. A.
No conflict among the plaintiffs prevents them from constituting a coherent class. Rule 23(a)(4) merely requires that plaintiffs not have “antagonistic” interests, In re Drexel Burnham Lambert Group, Inc., 960 F.2d at 291 (citation omitted), so “only a conflict that goes to the very subject matter of the litigation will defeat a party’s claim of representative status.” Krueger v. N. Y. Tel. Co., 163 F.R.D. 433, 443 (S.D.N.Y.1995) (quoting 7 Charles Alan Wright & Arthur R. Miller, Federal Practice & Procedure § 1768, at 639 (1972)) (other citation omitted). As in the commonality and typicality inquiry discussed above, Rule 23 allows for plaintiffs who have different relief preferences based on their particular injuries and situations.
B. Rule 23(b)
Class certification under Fed. R.Civ.P. 23(b)(2) is proper where injunctive or declaratory relief is appropriate against a defendant who has acted on grounds generally applicable to the class. See, e.g., Comer v. Cisneros, 37 F.3d 775, 796 (2d Cir.1994) (approving 23(b)(2) class “because the plaintiffs seek injunctive relief and they predicate the lawsuit on the defendants’ acts and omissions” that resulted in discrimination). Here, plaintiffs’ challenge to the compulsory *260arbitration clause seeks equitable relief from a commonly-faced practice, and the individual allegations also allege “central and systemic failures.” Marisol A., 126 F.3d at 378. This suffices for 23(b)(2) in spite of “the unique circumstances of each plaintiffs experience” and the fact that some in the class are no longer in a relationship with the defendant. Id. at 378.
Class certification under Fed. R.Civ.P. 23(b)(3) is proper where common questions of law or fact not only are present, but also “predominate over any questions affecting only individual members, and ... class action is superior to other available methods for the fair and efficient adjudication of the controversy.” While showing typicality and commonality under Rule 23(a) “goes a long way toward satisfying the Rule 23(b)(3) requirement of commonality,” Rossini, 798 F.2d at 598, in Rule 23(b)(3) “the predominance criterion is far more demanding,” Amchem Prods., Inc. v. Windsor, — U.S. —, —, 117 S.Ct. 2231, 2250, 138 L.Ed.2d 689 (1997). Here, the common questions making the representative plaintiffs’ claims typical for purposes of 23(a)(2) and 23(a)(3), see supra Part II.A., do predominate, given the salience of the challenge to the U-4 clause and the allegations of Smith Barney’s central and systemic failures to comply with Title VII.
The class also meets the 23(b)(3) “superiority” requirement. The degree to which the class meets the 23(a) requirements is relevant here: with a large number of plaintiffs making similar claims, consolidated litigation conserves judicial and party resources. There is little concern for the sort of prejudice to parties’ interests under the four factors listed as 23(b)(3)(A)(D), given the feasibility of opt-out, the appropriate job the settlement does reconciling extant litigation, and the efficiency justification for consolidation.
In appropriate circumstances, a class can be certified under both Fed.R.Civ.P. 23(b)(2) and 23(b)(3). The court must find that the class meets the threshold criteria for both, as it did above. The court also must find reconcilable the differences between (b)(2) and (b)(3) rights to money relief and opt-out notice, as it does below.
First, while (b)(2) envisions equitable relief, it does not per se exclude monetary relief as well. Money damages cannot be the predominant form of relief plaintiffs seek, see Eisen v. Carlisle & Jacquelin, 391 F.2d 555, 564 (2d Cir.1968), but can be part of an overall menu of relief that is predominantly equitable. See, e.g., Gelb v. Am. Tel. & Tel. Co., 150 F.R.D. 76, 78 (S.D.N.Y.1993); Ventura v. N.Y. City Health & Hosp., Corp., 125 F.R.D. 595, 601 (S.D.N.Y.1989). In employment discrimination claims, monetary relief for a(b)(2) class is typical. See, e.g., Grogg v. Gen. Motors Corp., 72 F.R.D. 523, 530-531 (S.D.N.Y.1976). The possibility of even sizeable Title VII damages will not defeat an otherwise proper (b)(2) class, whether because courts deem much of the monetary relief for discrimination to be equitable restitution or because of the predominance of the judicial finding of discrimination even in the absence of damages, as evidenced by the fact that plaintiffs can be “prevailing parties” without recovering damages.
Second, while neither due process nor the Federal Rules requires that (b)(1) or (b)(2) class members enjoy the notice and opt-out rights of (b)(3) class members, see Robertson v. Nat’l Basketball Ass’n, 556 F.2d 682, 685-686 (2d Cir.1977), it is within judicial discretion to find that circumstances justify such rights. The Second Circuit has so held in 23(b)(1)(B) limited fund class actions, see In re Joint Eastern & Southern Dist. Asbestos Litig., 78 F.3d 764, 778 (2d Cir. 1996); County of Suffolk v. Long Island Lighting Co., 907 F.2d 1295, 1304-1305 (2d Cir.1990), and other circuits have held that the discretion Suffolk recognizes naturally extends to (b)(2) employment discrimination actions. See, e.g., Eubanks v. Billington, 110 F.3d 87, 94-95 (D.C.Cir.1997) (collecting cases and holding that “the language of Rule 23 is sufficiently flexible to afford district courts discretion to grant opt-out rights”).
Opt-out rights are most appropriate where part of the relief is monetary, see id., or otherwise “uniquely individual to each class member.” Holmes v. Continental Can *261Co., 706 F.2d 1144, 1159 (11th Cir.1983) (finding, in a(b)(2) employment discrimination suit, that denial of opt-out rights was an abuse of discretion where relief became individualized). For equitable relief that implements a change regardless of the number of plaintiffs, such as the settlement’s diversity initiatives, opt-out may not be a meaningful concept. But because equitable remedies are flexible, they may take a shape that makes opt-out feasible, non-prejudicial to class members, and appropriate toward the goal of protecting the rights of non-representative claimants. Here, the new dispute resolution process for individual claims is just such an equitable remedy that justifies (b)(2) certification, but that also justifies opt-out rights for those who would decline the equitable remedy, a move which in no way prejudices the class members.
III. Review of Settlement
A. The Settlement
The settlement is an agreement among the class, defendant Smith Barney, and defendant James Dimon, see Settlement, page 1, that aims for final resolution of all class members’ claims of sexual harassment, discrimination, or retaliation related to their employment at Smith Barney, see Settlement, § 6.1. Toward that end, defendants agree solely for purposes of settlement that the court has subject matter jurisdiction over plaintiffs’ claims, see Settlement, § 3.1, that plaintiffs state claims on which relief can be granted, § 3.2, and that plaintiffs are a proper class under Fed.R.Civ.P. 23(b)(2) and 23(b)(3), see Settlement, § 5.1. The settlement contains two major categories of relief: a new dispute resolution program for individual resolution of class members’ claims; and a set of diversity programs and initiatives at Smith Barney.
1. Dispute Resolution Process
The settlement would “establish a new dispute resolution process ... as an exclusive forum ... [for] all alleged claims of sexual harassment, discrimination and retaliation by members of the proposed class against Smith Barney.”' See Settlement, § 2.1 (“DRP”). After attempts at mediation, see Settlement, § 7.13, three-arbitrator panels with a prescribed composition to ensure gender diversity, dispute resolution experience, and employment law experience, see Settlement, § 7.14(3), will resolve remaining individual claims. Any court of competent jurisdiction can review DRP arbitrations under the Federal Arbitration Act, 9 U.S.C. § 10 (“FAA”). See Settlement, § 7.14(12)(a). The court’s evaluation of the settlement, see infra Part III.C., will discuss relevant details of the DRP.
2. Diversity Programs and Initiatives
The settlement also “calls for the creation, implementation, and continuation of strong diversity programs and initiatives at Smith Barney.” See Settlement, § 2.1. Smith Barney is to spend $15 million over four years on “Diversity Programs and Initiatives,” see Settlement, § 8.1(1) (“DPI”), which the settlement defines as:
any program or other expenditure by the Firm which has as at least one of its primary goals ... promoting equal opportunity for all persons[,] ... providing education concerning ... [or] preventing or remedying any form of workplace harassment, discrimination, or retaliation^ or] ... fostering recruitment or mentoring of, or employment advancement or promotion opportunities for, women and minorities in the securities industry. See Settlement, § 8.1(2)
The DPI obligation increases if Smith Barney fails to make “reasonable efforts,” see Settlement, § 8.2(l)(f),4 to reach set percent*262age goals for gender diversity in various job categories, see Settlement, §§ 8.2(l)(a)-(e). Outside auditors will evaluate Smith Barney’s expenditures for compliance with the DPI obligation, and class counsel can inspect and challenge Smith Barney’s compliance to an appointed “Independent Diversity Advis- or,” and then to non-reviewable DRP arbitration. See Settlement, § 8.1(4). The settlement also provides for various internal Smith Barney reforms aiming to improve record-keeping, complaint, and review procedures. See Settlement, § 8.2(l)(g)-(u).
3. Attorneys’ Fees and Expenses; Incentive Payments; Resolution of Other Claims
The settlement requires Smith Barney to pay up to $13.2 million in attorneys’ fees and expense reimbursement, see Settlement, §§ 11.1, 11.2, 12.1,5 and up to $1.9 million in aggregate incentive payments for various class representatives based on their efforts on behalf of the class, see Settlement, § 12.2.6 It also resolves several existing individual claims of class representatives by allowing them into mediation or the DRP, see Settlement, § 12 .5,12.6, and requires as a precondition to the settlement taking effect that the parties dismiss the related Alvarez action in the federal district court for the Northern District of California, see Settlement, § 14.1(2).
B. Review Standards
Though there is a “general policy favoring the settlement of litigation,” Wein-berger v. Kendrick, 698 F.2d 61, 73 (2d Cir. 1982), the court must independently scrutinize the settlement to determine “whether the compromise is fair, reasonable, and adequate .” Weinberger, 698 F.2d at 73. This is not mere “rubber stamp approval,” but “an independent evaluation.” City of Detroit v. Grinnell Corp., 495 F.2d 448, 462 (2d Cir. 1974). The court making this evaluation has a “fiduciary” duty to the non-representative class members who were not party to the settlement agreement, Grant v. Bethlehem Steel Corp., 823 F.2d 20, 23 (2d Cir.1987), because “[i]nherent in any class action is the potential for conflicting interests among the class representatives, class counsel, and absent class members.” Maywalt v. Parker & Parsley Petroleum Co., 67 F.3d 1072, 1077 (2d Cir.1995).
1. Procedural Fairness
At a minimum, the settlement inquiry requires procedural fairness. The court must look to “the negotiating process by which the settlement was reached” and find that “the compromise [is] the result of arm’s-length negotiations and that plaintiffs’ counsel have possessed the experience and ability and have engaged in the discovery necessary to effective representation of the class’s interests.” Weinberger, 698 F.2d at 74 (citing Grinnell Corp., 495 F.2d at 463-466). Here, class counsel have shown sufficient experience and ability, see supra Part II.A. (discussing adequacy of class representation), and the third-party mediation process was an *263appropriate measure for arm’s-length negotiations, see Settlement, § 1.9-1.11.
2. Substantive Scrutiny: Grinnell Factors
In substantive scrutiny of a settlement, “[t]he primary concern ... ‘is the need to compare the terms of the compromise with the likely rewards to litigation.’” Weinberger, 698 F.2d at 63 (quoting Protective Comm. for Indep. Stockholders of TMT Trailer Ferry, Inc. v. Anderson, 390 U.S. 414, 424-425, 88 S.Ct. 1157, 20 L.Ed.2d 1 (1968)) (other citations omitted). The Second Circuit lists nine factors:
(1) the complexity, expense and likely duration of the litigation, (2) the reaction of the class to the settlement, (3) the stage of the proceedings and the amount of discovery completed, (4) the risks of establishing liability, (5) the risks of establishing damages, (6) the risks of maintaining the class action through the trial, (7) the ability of the defendants to withstand a greater judgment, (8) the range of reasonableness of the settlement fund in light of the best possible recovery, (9) the range of reasonableness of the settlement fund to a possible recovery in light of all the attendant risks of litigation[.]
County of Suffolk v. Long Island Lighting Co., 907 F.2d 1295, 1323-1324 (2d Cir.1990) (quoting Grinnell Corp., 495 F.2d at 463) (other citations omitted). Because the nine factors overlap here, the discussion will collapse them into four considerations.
a. Class Reaction to Settlement
The second Grinnell factor, the reaction of the class to the settlement, poses no problem here. The small number of objectors (12) and opt-outs (40) among the over 20,000 class members, see Decl. of Mary Stowell, Fairness Hearing Ex. A 1137, is a stark contrast to cases where a sizeable percentage of members object, which inherently calls the settlement into question. See, e.g., Pettway v. Am., Cast Iron Pipe Co., 576 F.2d 1157 (5th Cir.1978) (rejecting settlement opposed by 70% of affected class). Still, because of the court’s fiduciary duty to the entire class, it must conduct an independent inquiry that considers objections and the court’s own concerns. If the court finds the settlement unfair, unreasonable, or inadequate, it must reject even a settlement garnering overwhelming approval. See, e.g., Plummer v. Chem. Bank, 668 F.2d 654 (2d Cir.1982) (rejecting Title VII settlement "with only 25 objectors and 25 opt-outs, id. at 658, in class of “all the black officials, managers, and professionals then employed by [Chemical] Bank in New York City or who subsequently might become employed by the Bank in New York City,” id. at 657, which numbered over 500, id. at 659).
b. Stage of Proceedings and Amount of Discovery
The third Grinnell factor, the stage of the litigation and the amount of discovery completed, also reveals no problem with the settlement. Because much of the point of settling is to avoid litigation expenses such as full discovery, “it would be inconsistent with the salutary purposes of settlement,” Handschu v. Special Servs. Div., 787 F.2d 828, 834 (2d Cir.1986), to find that “extensive pre-trial discovery is a prerequisite to approval” of a settlement. Plummer, 668 F.2d at 660. Rather, discovery must be sufficient toward two ends. First, there must be sufficient discovery of facts for the court “intelligently to make such an appraisal” as its fiduciary duty requires. See, e.g., id. at 660 (rejecting settlement where no facts justified more favorable treatment of representative plaintiffs’ claims). Second, the pretrial negotiations and discovery must be sufficiently adversarial that they are not “designed to justify a settlement ... [, but] an aggressive effort to ferret out facts helpful to prosecution of the suit.” Saylor v. Lindsley, 456 F.2d 896, 899 (2d Cir.1972) (rejecting settlement where the limited discovery was largely contemporaneous with settlement and did not seek much “inculpatory” evidence, id. at 901).
Here, discovery was sufficient for these considerations. Unlike in Plummer, the representative plaintiffs’ more favorable treatment finds factual support in their affidavits supporting their incentive payments. See Fairness Hearing Ex. 4. Also, unlike in Say-*264lor, the discovery was prior to settlement and did seek inculpatory evidence. See Settlement, §§ 1.5-1.6, 1.8. (noting information on female employment and discrimination claims at Smith Barney). Even if additional discovery remains possible, see Objection of Evelyn Martinez, Fairness Hearing Ex. A, plaintiffs did unearth enough information useful for their settlement negotiations, as well as for individual arbitration claims. The court will not reject the settlement as so premature as to prevent intelligent appraisal of the parties’ claims and compromise terms. See, e.g., Handschu, 787 F.2d at 834; Grinnell Corp., 495 F.2d at 463-464.
e. Defendant Ability to Withstand a Greater Judgment
The seventh Grinnell factor, defendant ability to withstand a greater judgment, also poses no problem. The settlement does not guarantee that Smith Barney will pay any damages to the class as a whole and does not reach the limits of Smith Barney’s solvency with the costs it guarantees Smith Barney will pay (DPI spending, legal costs, incentive payments, etc.). However, a major aim of the settlement is to provide individual money relief through the DRP, which supports two reasons the settlement is appropriate given Smith Barney’s ability to withstand judgments. First, the guaranteed expenses of the diversity programs and payments to class counsel and representatives should not aim for the limits of Smith Barney’s financial resources, because that would jeopardize the amount remaining for individual money relief. Second, the current uncertainty as to that individual money relief means that the court cannot find it insufficient to Smith Barney’s financial resources.
d. Risk of Litigation: Uncertainty and Cost
Together, the remaining six factors address the extent to which settlements can provide less than the full set of remedies plaintiffs initially sought: “(1) the complexity, expense and likely duration of the litigation, ... (4) the risks of establishing liability, (5) the risks of establishing damages, (6) the risks of maintaining the class action through' the trial, ... (8) the range of reasonableness of the settlement fund in light of the best possible recovery, (9) the range of reasonableness of the settlement fund to a possible recovery in light of all the attendant risks of litigation.” Grinnell Corp., 495 F.2d at 463. Settlements sacrifice higher potential reliefto avoid the uncertainty and cost of prolonging the case. These factors recognize that the amount of relief a settlement can sacrifice is proportional to the amount of uncertainty and cost it avoids.
Plaintiffs would face substantial costs if they continue their suit without settling. Legal costs are already in the millions after only partial discovery, briefing of 12(b) motions, settlement pursuit, and similar preliminary activity. Continued prosecution likely would entail more extensive pretrial activity, such as motions to dismiss and depositions in multiple cities, as well as the actual class action litigation. The costs of delay accumulate not only for the lawyers and class representatives, but for the entire class. All class members who allege injuries would face delayed compensation of past injuries, delayed amelioration of continuing discrimination, and delayed redress of their grievances with the status quo arbitration process. A settlement that is smaller but more immediate may be in the class’s best interest, because additional relief via continued prosecution may not make up for the substantial costs of delay.
Uncertainty as to whether litigation would have yielded the full set of relief plaintiffs sought also supports pre-verdict compromise. Apart from the general factual uncertainty regarding proof of discrimination, there is also much legal uncertainty, with two issues in particular. First, regarding plaintiffs’ substantive claims, standards for harassment liability were unclear as of settlement in many ways. Second, the law on compulsory arbitration of Title VII claims is in great flux. Though the court notes the recent shift in which some plaintiffs have met their burden of proving that Title VII bars prospective agreements to arbitrate claims, it also notes that the Second Circuit and the Supreme Court have not yet ruled on the issue.
Whether any of this legal flux is favorable or unfavorable to plaintiffs’ claims, sufficient *265uncertainty remains to support short-of-windfall settlement that avoids the risk that legal standards will evolve unfavorably. When it is legally questionable whether the district court can grant plaintiffs the full equitable relief sought, plaintiffs may compromise for more limited relief as “respeeta-ble[ ] concessions” from the defendant. Handschu, 787 F.2d at 834.
C. Objections to the Settlement
1. Lack of Additional Relief
The settlement does .not provide a guaranteed damages fund, but instead lets class members win damages only via the new dispute resolution process, which entails uncertainty and burden. The settlement also does not provide for personnel changes at Smith Barney, such as the termination or discipline of any alleged perpetuators of the hostile work environments. See, e.g., Aff. of Edna Broyles ¶13.7
The lack of guaranteed damages is not fatal, as the settlement does provide relief targeted toward the plaintiffs’ complaints. The DPI and DRP, respectively, aim to redress the two systemic problems that brought the plaintiffs to court in the first place: widespread discrimination and mandatory seeurities-industry arbitration. Courts approve even settlements that fully waive damages claims where, as here, the settlements implement structural reforms to remedy the widespread civil rights violations plaintiffs allege. See, e.g., Majeske v. Fraternal Order of Police, Local Lodge No. 7, 94 F.3d 307 (7th Cir.1996) (approving Title VII settlement that altered promotion procedures but provided no backpay); Handschu v. Special Servs. Div., 605 F.Supp. 1384 (S.D.N.Y. 1985), aff'd, 787 F.2d 828 (2d Cir.1986) (approving settlement that reformed police investigative procedures but provided no damages).
Moreover, while the settlement fails to guarantee damages, it does not waive them, as in Majeske and Handschu. To the eon-trary, the DRP has plaintiff-friendly provisions that, in many ways, make damages more likely and less limited than in the status quo arbitration scheme. See infra Parts III.C.2.b,e. The uncertainty it leaves is standard in Title VII class actions. Even when a class proves Title VII liability, members often cannot receive damages until they prove injury in individualized proceedings. See, e.g., Teamsters v. United States, 431 U.S. 324, 361, 97 S.Ct. 1843, 52 L.Ed.2d 396 (1977). In the present settlement, unlike in Teamsters, the individualized proceedings require plaintiffs to prove liability as well as injury, but that is acceptable for a pre-liability settlement. Absent prima facie establishment of liability, a settlement likely will be less favorable to plaintiffs, and can provide only .a “fraction” of the relief that would be appropriate post-liability. See City of Detroit v. Grinnell Corp., 495 F.2d 448, 455 (2d Cir.1974). Just as a pre-liability settlement can accept less money, see id. at 455, n. 2, it can accept less certainty in its post-settlement individualized hearings, than would be appropriate post-liability.
The absence of personnel measures against alleged discriminators similarly is not fatal. A settlement may pass judicial muster even though some members prefer different remedies. The court’s fiduciary duty to class members entails not only protecting an objecting minority, but also protecting a non-objecting majority from a “veto” by dissenting members. Settlements can be approved over the preferences of original plaintiffs, because “a contrary view would put too much power in a wishful thinker ... to thwart a result that is in the best interests” of others. Maywalt v. Parker & Parsley Petroleum Co., 67 F.3d 1072, 1078 (2d Cir.1995) (citing Saylor v. Lindsley, 456 F.2d 896, 899-900 (2d Cir.1972)) (other citation omitted). Even where personnel changes adverse to non-parties are a plausible remedy, they are not a mandatory remedy whose absence torpedoes an otherwise *266fair, reasonable, and adequate settlement. See, e.g., Majeske, 94 F.3d at 309 (approving Title VII settlement improving future promotion prospects but not requiring demotions of prior promotees to make room for class members).
2. Challenges to the Dispute Resolution Process
a. Resolving More Claims than is Appropriate
The DRP encompasses more claims than those in the plaintiffs’ complaint, see Settlement, § 6.1, which necessitates particular scrutiny. While settlement necessarily implies the sort of give-and-take in which each party relinquishes something of value, there are limits on the extent to which settling plaintiffs can limit class members’ claims beyond the initial set of class claims. The Second Circuit has rejected settlements that go beyond those limits. See, e.g., Dunlop v. Pan Am. World Airways, Inc., 672 F.2d 1044 (2d Cir.1982); Nat’l Super Spuds v. N.Y. Mercantile Exch., 660 F.2d 9 (2d Cir.1981). There is no general rule against settlements that limit additional claims, however, and many settlements featuring such limitations earn Second Circuit approval. See, e.g., In re Baldwin-United Corp., 770 F.2d 328 (2d Cir.1985); Weinberger v. Kendrick, 698 F.2d 61 (2d Cir.1982); TBK Partners, Ltd. v. Western Union Corp., 675 F.2d 456 (2d Cir.1982); Steinmetz v. Toyota Motor Credit Corp., 963 F.Supp. 1294 (E.D.N.Y. 1997); Int’l Union of Elec., Elec., Salaried, Mach., and Furniture Workers, ALF-CIO v. UNISYS Corp., 155 F.R.D. 41 (E.D.N.Y. 1994); O’Brien v. Nat’l Property Analysts Partners, 739 F.Supp. 896 (S.D.N.Y.1990). These cases reflect that the propriety of settlement limitations on additional claims is fact-dependent.
One possible problem with the present limitation is that it broadly applies to all employment claims. The settlement thus restricts claims arising under different legal theories and different factual predicates than the original class claims. Some of the cases allowing limits on additional claims did so because, unlike in Nat’l Super Spuds, they limited only claims arising from similar facts. See, e.g., TBK Partners, Ltd., 675 F.2d at 460; In re Baldwin-United Corp., 770 F.2d at 336.
Three other factors, however, support the propriety of the present claim limitation, and distinguish it from those of Nat’l Super Spuds and Dunlop, and thus support its propriety. First, class members had notice and the opportunity to opt out of the settlement if they disapproved of the settlement package that includes the claim limitation. The class members in both Nat’l Super Spuds and Dunlop had no such opportunity, making the limitation on additional claims more troubling. See Nat’l Super Spuds, 660 F.2d at 16, 19 (finding no opportunity to opt out where notice did not address claim limitation); Dunlop, 672 F.2d at 1050 (noting that no opt-out rights exist in ADEA action by Secretary of Labor). Because opportunity to exit limits vulnerability to exploitation, courts have approved settlement claim limitations where class members had opt-out rights and sufficient notice. Weinberger, 698 F.2d at 70-71 (finding sufficient that class members be “fairly apprised ... [of] the relevant terms of the proposed settlement, and their options” to opt out or “to probe more deeply” into the terms); TBK Partners, Ltd., 675 F.2d at 461; Thompson v. Edward D. Jones & Co., 992 F.2d 187, 190-91 (8th Cir.1993); Steinmetz, 963 F.Supp. at 1299-1300; O’Brien, 739 F.Supp. at 901-02.
Second, in Nat’l Super Spuds, the settlement harmed a definable minority that held “unliquidated” claims distinct from the “liquidated” claims the class shared. Nat’l Super Spuds, 660 F.2d at 14. The need to protect the “few whose interests are being sacrificed for the benefit of the majority,” id. at 16, was the critical reason the court rejected the settlement: “An advantage to the class, no matter how great, simply cannot be bought by the uncompensated sacrifice of claims of members, whether few or many, which were not within the description of claims ascertainable by the class.” Id. at 19. Here, the settlement broadly limits diffuse and uncertain claims. It does not disadvantage any identifiable subset of the class; rather, it gives up something of probabilistic value to all class members. Unlike in Nat’l Super *267Spuds, there is little reason to fear that the claim limitation is an improper “sacrifice” of class members that merits judicial rejection of the settlement.
Third, the claim limitation does not waive any individual claims, but simply subjects them to a different, and more plaintiff-favorable, form of arbitration. All claims within the DRP scope already were subject to U-4 seeurities-industry arbitration, so all the settlement does is move them to a better arbi-tral forum. There is little prejudice from such a claim limitation, compared to the more typical limitation that extinguishes the claims altogether. See, e.g., Nat’l Super Spuds, 660 F.2d at 14.(settlement intended to bar unliq-uidated claims entirely).
b. Procedural Inadequacies in the DRP
The DRP includes several procedures different from applicable federal or state law, some of which arguably are disadvantageous to plaintiffs. See, e.g., Amicus Curiae Br. of New York State National Organization for Women (“NOW-NYS”), at 5-8. For example, the settlement’s DRP provisions create a rebuttable presumption against depositions, see Settlement, § 7.14(8), and limit joinder to claims from the same branch office, see Settlement, § 7.4(1). yet even if these and like provisions benefit defendants, other DRP procedures benefit plaintiffs. For example, the DRP imposes on plaintiffs no burden of proof of liability, see id., § 7.14(4), allows hearsay (which is particularly useful for plaintiffs given the abbreviated discovery), and disallows Smith Barney use of after-acquired evidence and evidence of sexual history, see id., §§ 7.14(10)(a)-(c).
Another positive for the class is the agreed upon panel of competent and experienced Title VII attorneys class counsel will establish to represent individual DRP claimants. See Addendum to Stipulation of Settlement, November 18,1997. In this court’s view, the most serious problem with Title VII litigation today is the paucity of competent and experienced lawyers available to plaintiffs in such cases. The settlement’s Title VII panel remedies this problem for DRP claimants.
At the least, the DRP seems more favorable to plaintiffs than the status quo seeurities industry arbitration, which makes it an advantageous settlement term. Disagreement as to the precise balance of plaintiff-friendly and defendant-friendly DRP provisions is simply disagreement as to the size of the bundle of relief the settlement provides. As noted above, the factors Grinnell outlines illustrate why that size will be less than the full bundle that the named plaintiffs initially sought and that many class members still would like.
c. Limits on Punitive Damages Recovery
The DRP imposes no limits on the size of punitive damages awards, but provides that for non-federal gender-based claims, plaintiffs must show clear and convincing evidence of their entitlement to punitive damages. Under New York law, which the settlement makes the starting point of substantive law for non-federal gender-based claims, see Settlement, § 7.9, punitive damages are unavailable for employment discrimination claims. See Thoreson v. Penthouse Int’l, Ltd., 80 N.Y.2d 490, 591 N.Y.S.2d 978, 606 N.E.2d 1369, rearg. den., 81 N.Y.2d 835, 595 N.Y.S.2d 397, 611 N.E.2d 298 (1993) (disallowing punitive damages in a sexual harassment suit under New York State Human Rights Law, N.Y. Exec. Law § 297(9)). A requirement of clear and convincing evidence for relief not available in the courts does not seem unduly burdensome, as NOW-NYS objects. See Amicus Curiae Br. of NOW-NYS, at 7-8.
d. Conflict of Interest of Proposed DRP Administrator
The settlement tentatively provides for the Duke University Private Adjudication Center (“DUPAC”) to serve as DRP Administrator. One objection asserts that Smith Barney’s service managing various Duke University financial matters constitutes a “business relationship with Duke University ... [that] creates a real or, at least, perceived conflict of interest” for DUPAC. See Aff. of Edna Broyles ¶ 15.
The court finds no impropriety in the choice of DUPAC as DRP administrator. Smith Barney’s financial services to Duke University create a very thin, tangential tie to DUPAC. That tie is insufficient for DU-*268PAC’s service as DRP Administrator to create a conflict of interest. Moreover, the duties of the DRP Administrator are, as the title suggests, essentially administrative, such as appointing three-arbitrator DRP panels from pools the parties provide, see Settlement, § 7.14(3), appointing mediators from pools the parties provide, see Settlement, § 7.13(3), and serving as a go-between for service of hearing complaints, answers, etc., see Settlement, § 7.14(1). The limited substantive discretion of the DRP Administrator prevents much possibility of any ap- ■ pearance of impropriety.
3. Inadequacy of the Diversity Programs and Initiatives (DPI)
The overarching question for the DPI is whether it can deliver on its promise “to promote equal employment opportunity, gender equality, a workforce rich in diversity and a workplace free from harassment, discrimination and retaliation” at Smith Barney. Settlement, § 2.2. The settlement leaves unclear both what new programs it will fund and to what extent existing Smith Barney expenditures can qualify under the vague definition of “diversity-related expenditure.” See id., § 8.1(2). The proposed programs are nowhere delineated or defined. It is possible to envision good-faith compliance with the settlement’s intent that the money support programs to implement change at Smith Barney. Yet it is also possible to envision that $15 million over four years for a company as large as Smith Barney is grossly inadequate if personnel training, for example, is included.
It would not be troubling to let the parties fill in the blanks if the court could ensure that the evolution of the DPI protects the interests of all class members. See, e.g., Objection of Joan Chasin, Fairness Hearing Ex. A (citing continued difficulties at the Garden City office). Moreover, the settlement denies the court retained jurisdiction over disputes regarding the DPI obligation. All such disputes begin with the parties, then proceed to the Independent Diversity Advis- or, and then end at a DRP arbitration panel, with no judicial review of any kind. See Settlement, § 8.1(4). This court and the Second Circuit previously have noted that “a federal court has jurisdiction to enforce a settlement agreement only if the dismissal order specifically reserves such authority or the order incorporates the terms of the settlement.” Scelsa v. City Univ. of N.Y., 76 F.3d 37, 40 (2d Cir.1996) (affirming district court denial of jurisdiction in Scelsa v. City Univ. of N.Y., 806 F.Supp. 1126 (S.D.N.Y. 1992)) (citing Kokkonen v. Guardian Life Ins. Co., 511 U.S. 375, 114 S.Ct. 1673, 128 L.Ed.2d 391 (1994)). The court cannot “reserve such authority” over settlement disputes where the parties to the settlement explicitly disclaim that authority and their right to return to court.
The court’s fiduciary duty to unnamed class members prevents it from risking an end to its settlement review before a centerpiece of that settlement is meaningfully well-defined. Unnamed class members are particularly vulnerable because they lack access to any forum except through class counsel. See Settlement, § 16.6 (“Only class counsel shall have standing and authority to bring any action to enforce this settlement.”). That standing limitation exacerbates the risk present in all class actions that “ ‘[t]he interest of lawyer and class may diverge, as may the interests of different members of the class, and certain interests may be wrongfully compromised, betrayed, or “sold out” without drawing the attention of the court.’” Plummer v. Chem. Bank, 668 F.2d 654, 658 (2d Cir.1982) (quoting Pettway v. Am. Cast Iron Pipe Co., 576 F.2d 1157, 1169 (5th Cir. 1978)) (other citation omitted). So long as the settlement’s touted diversity programs and initiatives remain admirable but amorphous aspirations rather than reasonably defined plans of action, the court’s fiduciary duty to the class remains unfinished business.
D. Attorneys’ Fees and Costs; Incentive Payments
Because the court is not approving the settlement, it would be premature to discuss the objections to settlement’s provision for attorneys’ fees and costs and for class representative incentive payments. There may be no later settlement; even if there is one, it *269may have different attorneys’ fees and incentive payments provisions; even if it provides the same fees and payments, its different terms may necessitate its own inquiry, distinct from what the court could conclude based on the present settlement.
CONCLUSION
For purposes of this settlement, plaintiffs constitute a proper class presenting claims properly in federal court. The settlement they propose, however, defers for future nonjudicial resolution any questions regarding what the as-yet-amorphous Smith Barney diversity programs and initiatives can and cannot include. The court cannot declare its duty to evaluate the settlement complete before the parties more meaningfully clarify what it does. Accordingly, the court denies final approval to the proposed settlement.
. This case initially was assigned to the Honorable Allen G. Schwartz, with Magistrate Judge Leonard Bernikow agreeing to the discovery schedule. In June 1997, the case was reassigned to this court. See Settlement, § 1.7. Class counsel also have represented plaintiffs who commenced a substantially similar action filed on October 28, 1996 in the Northern District of California, but who now are named plaintiffs in this case. Seeid.,%% 13,1.8-1.9.
. For example, a finding of proper subject matter jurisdiction does not foreclose a finding that plaintiffs fail to state claims on which relief can be granted, because Fed.R.Civ.P. 12(b)(1) (failure to state a claim) and 12(b)(6) (lack of subject matter jurisdiction) are not entirely coextensive. See Nowak v. Ironworkers Local 6 Pension Fund, 81 F.3d 1182, 1189-90 (2d Cir.1996) (denying 12(b)(1) motion while granting 12(b)(6) motion).
. The fourth Genesco issue is not relevant here because all of plaintiffs’ claims are employment-related, so the court will not discuss it.
. Smith Barney may show "reasonable efforts” sufficient to avoid the increased DPI obligation either by offering positions to the same percentage of women as apply, by offering positions to the same percentage of women as the target percentage requires, or by showing that differences in qualifications led to its failure to meet the target percentages. See Settlement, § 8.2(l)(f). Essentially, this makes the percentage goals provisions less an affirmative action plan than an agreement providing liquidated equitable relief for discrimination in which Smith Barney bears the burden of proof: if Smith Barney’s hiring statistics evidence a statistical gender disparity, it must prove a legitimate, nondiscriminatory reason for that disparity or face a *262specified penalty aiming to redress the problem by increasing Smith Barney’s diversity expenditures.
. The settlement allows class counsel to apply to this court for:, attorneys’ fees of up to $9.5 million ($6 million upon the effective date, $1.75 million after one year, and $1.75 million after two years); attorney's fees equaling 10% of class members' DRP damages (excluding attorneys' • fees and costs) up to a maximum of $2.5 million; and reimbursement of expenses of up to $900,000. See Settlement, §§ 11.1, 11.2. It also provides up to $50,000 in attorneys' fees and expenses for the law firm Altshuler, Berzon, Nuss-baum, Berzon & Rubin for their services to the class in connection with the Alvarez case. See Settlement, § 11.5. Additionally, it allows class representatives to apply to this court for reimbursement of up to $250,000 in attorneys' fees and expenses they incurred prior to their representation by class counsel. See id.., § 12.1.
. The settlement allows applications for incentive payments in the following amounts: $150,000 each for "Group 4” plaintiffs (Edna Broyles, Tracy Gibbs, Lori Hurwitz, Lorraine Parker, Roberta Thomann, Beverly Trice, and Cara Beth Walker); $75,000 each for "Group 3” plaintiffs (Jennifer Alvarez, Mary Ann Cabell, Marianne Dalton, Stanette Goepel, Patricia Hanlon, Lisa Mays, Lydia Klein, and Eileen Valentino); $50,-000 each for "Group 2” plaintiffs (Patricia Clem-ente, Bette Laswell, and Daniele Saccone); and $20,000 each for "Group 4” plaintiffs (Stephanie Rodruck, Simone Schwendener, Teresa Tedesco, Robin Tompkins, and Ardis Vinnecour). See Settlement, § 12.2.
. Ms. Broyles made a mix of timely and untimely objections. Because of the court’s fiduciary duty to the class, however, any untimely objections are not "waived" in the sense of eliminating judicial scrutiny. The court must consider all potential objections, raised or not, it finds substantial. Thus, and particularly because no objection garnered large numbers, the court’s analysis of objections does not depend on the identity of the objector(s), if any.