Contracting (out) rights.

Introduction

 I. How We Got Here
      A. The Gardner-Denver, Barrentine, and McDonald Trilogy
      B. Gilmer and the End of the Non-Waivability Principle
         1. Employment Cases
         2. Consumer Cases

 II. The Empirical Defense
      A. Fundamentally Flawed Data
      B. Neglected Societal Costs

 III. Contracting (Out) Statutory Rights

Conclusion

INTRODUCTION

On April 1, 2009, the Supreme Court ruled in 14 Penn Plaza LLC v. Pyett (1) that a union can bargain away a member's right to seek judicial relief for employment discrimination. The Court had already resolved that an employee could bargain away that right for herself, (2) and the recent decision only expanded the notion that statutory rights may be overwritten by contract. In the Supreme Court's current view, private dispute resolution through arbitration is preferable to litigation. The Court seems undeterred by the Congressional mandate that claims arising under, for example, Title VII of the Civil Rights Act of 1964, (3) the Fair Labor Standards Act of 1938, (4) the Americans with Disabilities Act, (5) the Age Discrimination in Employment Act, (6) and the Civil Rights Act of 1991, (7) are to be enforced by the federal judiciary. With the Court's approval, pre-dispute, mandatory arbitration provisions have become ubiquitous in contracts for employment and consumer goods, forcing employees and consumers to arbitrate, rather than litigate, their statutory claims.

What is the significance of this trend for the enforcement of federal laws and the vindication of the rights conferred by those laws? (8) A number of scholars have devoted empirical research to this question, many arguing that arbitration is not as bad as it seems if one looks at real outcomes. (9) There are at least two significant problems with these analyses. First, putting aside any methodological flaws in the individual studies, everyone agrees that the data available is extraordinarily limited, and even the degree of the limitation is unknown. This is because arbitration is a private, often confidential process, the initiation, outcome, and reasoning of which are generally invisible to the public and unavailable to social scientists. As others have suggested, drawing conclusions from the small fraction of available data is meaningless and misleading. (10)

A second curious aspect of the empirical literature, which has received less criticism, is its narrow focus on a single question: courts are expensive, so does arbitration provide the litigants with more bang for their buck? Even those who question the wisdom of mandatory arbitration generally adopt this analytic approach. Conceding that arbitration costs less overall, they raise concerns about the hurdle of arbitration fees, whether grievants win as often and as much as they would in court, and whether the arbitration process disproportionately rewards repeat players. (11) Even the theorists who direct their inquiries towards fairness evaluate it by comparing the interests of plaintiffs and defendants, ignoring the benefits of statutory enforcement beyond those accrued by individual parties. The trouble with this approach is that what is at issue is not simply private interests, but public rights. (12)

Measuring the economic utility of arbitration for isolated individuals might make sense if contract law were independent of, or superior to, statutory law, as the Court seemed to believe in Lochner. (13) Applying a contract-based approach to the adjudication of statutory rights, however, leaves out a key player: the legislature. Congress passed anti-discrimination laws because it wanted to end discrimination in the workplace, and Congress saw public adjudication of these claims as an important part of the fight against discrimination. (14) In a similar vein, Congress passed consumer-protection statutes because it recognized the wide disparity in bargaining position between corporate sellers and consumer purchasers and wanted to protect consumers from corporate overreaching. Empowering consumers to bring suit in federal court, where their claims could be aggregated into class actions, was seen as vital to leveling the playing field between consumers and corporations. (15) Congress added cost-shifting provisions in both the anti-discrimination and consumer protection statutes so that individuals could serve as private attorneys general and enforce those laws in court, where transparency and adherence to the law are matters of first principle.

Arbitration, on the other hand, is conducted out of public view and is less constrained by the letter of the law. (16) Governing statutes appear to play a minimal role in arbitrators' decisions. (17) Arbitration does not develop precedent, nor does it allow for appellate review of conflicting decisions. (18) Even when arbitrators do produce written opinions, the decisions by and large remain unpublished and unavailable; some arbitration services require written decisions, which they make publicly available, but even those decisions are often significantly redacted. (19) For the most part, arbitration decisions constitute a species of "secret law," known only to repeat players and a few others. (20) To the extent that we are considering wholesale acceptance of arbitration as a mandatory substitute for litigation, we must come to terms with the fact that we are sacrificing the public interpretation of public laws.

The sphere of privacy surrounding arbitration not only frustrates any attempt to study its social effects; it contravenes the public scrutiny and public education functions Congress intended for courts to serve. Privatizing the enforcement of statutory rights erodes those rights, as rights that are not enforced publicly vanish from the public's eye, making the public less educated about the laws governing society and probably less likely to recognize and correct the laws' violations. The opaqueness of the arbitration process also makes it easy for the protections that Congress carefully built into legislation to fall victim to unequal contractual bargains. Purchasing a cell phone routinely requires forfeiting the right to band together in a class action to challenge fraud. (21) As a condition of employment, individuals often unknowingly surrender their right to a jury trial. (22) We suggest that these sacrifices and their aggregate social costs deserve attention. In our view, the Supreme Court's embrace of mandatory arbitration reflects a return to a Lochner-like veneration for the freedom to contract unrestrained by public laws, and the studies measuring individual interests fail to grapple with this reality.

I. HOW WE GOT HERE

Before turning to our critique of mandatory arbitration, we situate our observations in the context of a far broader assault on the civil justice system. As other commentators have discussed, for approximately the past twenty years, the Supreme Court has engaged in a procedural and substantive revolution of erecting previously nonexistent barriers between ordinary individuals and the adjudication of their rights. (23) The Court's ringing endorsement of mandatory arbitration, as it pushes rights claims out of courts of law and into a contract-based dispute resolution system, is just one manifestation of this revolution, but there are many others. Newly-fashioned standing jurisprudence obstructs environmental and consumer cases. (24) District court judges suddenly enjoy license to dismiss any complaint whose bare allegations (25) strike the judge as insufficient to make out a "plausible" case. (26) Emerging preemption jurisprudence has foreclosed state-law claims in areas ranging from products liability to pensions. (27) At the same time as the Court has wiped away longstanding, state-law rights of action, it has flatly refused to recognize new implied rights in the most compelling cases and has stripped away remedies, rendering some rights "rights" in name only. (28) Finally, the Court's restrictive reading of fee-shifting provisions has left many lawyers who win substantial relief for fights-beating clients without compensation to sustain themselves and their work. (29) We focus on cases shunted out of public courts and into private arbitration.

A. The Gardner-Denver, Barrentine, and McDonald Trilogy

To trace the emergence of the Court's arbitration-favoring jurisprudence, we begin with a trilogy of cases--starting with the Court's 1974 decision in Alexander v. Gardner-Denver Co. (30) and its 1981 ruling in Barrentine v. Arkansas-Best Freight Systems, Inc., (31) and ending a decade later in McDonald v. City of West Branch (32)--that took the diametrically opposed view that arbitration was not well-suited to protect statutory rights Congress conferred on individuals. Each of the cases rejected the argument that union members had forfeited statutory claims because their employment claims were submitted to an arbitrator pursuant to their union's collective bargaining agreement with the employer and the arbitrator had ruled against the employee. These cases all turned, in part, on the idea that resolving statutory claims should be the province of courts, not arbitrators, because arbitration was inherently an inferior process for safeguarding rights guaranteed by statute. (33)

In Gardner-Denver, a unanimous Court rejected the employer's submission that the plaintiff was barred from filing a Title VII complaint because an arbitrator had decided a parallel claim, brought under a collective bargaining agreement, adversely to the employee. (34) The Court emphasized that "[a]rbitral procedures, while well suited to the resolution of contractual disputes, make arbitration a comparatively inappropriate forum for the final resolution of fights created by Title VII." (35) The Court rested its conclusion on several factors. One was that statutory claims were better entrusted to generalist federal judges rather than to arbitrators. As the Court saw it, "the special role of the arbitrator ... is to effectuate the intent of the parties rather than the requirements of enacted legislation." (36) This is problematic, the Court said, because when there is conflict between the contract and Title VII, "the arbitrator must follow the agreement." (37) Arbitrators are generally chosen by the parties "because they trust his knowledge and judgment concerning the demands and norms of industrial relations." (38) But "the resolution of statutory or constitutional issues is a primary responsibility of courts, and judicial construction has proved especially necessary with respect to Title VII, whose broad language frequently can be given meaning only by reference to public law concepts." (39)

The Court also expressed serious reservations about the adequacy of the arbitration procedures to protect statutory rights. (40) The Court noted that Congress had decided that private litigation was essential to the enforcement of Title VII. When an action is brought, "the private litigant not only redresses his own injury but also vindicates the important congressional policy against discriminatory employment practices." (41) The Court did not find the procedures available in arbitration sufficient. One particular concern was that "the factfinding process in arbitration usually is not equivalent to judicial factfinding. The record ... is not as complete; the usual rules of evidence do not apply; and rights and procedures common to civil trials, such as discovery, compulsory process, cross-examination, and testimony under oath, are often severely limited or unavailable." (42) Further, unlike courts, "[a]rbitrators have no obligation to the court to give their reasons for an award." (43) In sum, "it is the informality of arbitral procedure that enables it to function as an efficient, inexpensive, and expeditious means for dispute resolution," but these characteristics "make[] arbitration a less appropriate forum for final resolution of Title VII issues than the federal courts." (44)

The Court's 1981 decision in Barrentine echoed many of these concerns, but took Gardner-Denver one step further, emphasizing the nonwaivability of individual rights conferred by statute. Barrentine involved the Fair Labor Standards Act ("FLSA"). (45) The employer argued that the submission of an employee's wage claim to arbitration waived the employee's right to bring a separate action for the payment of wages under FLSA. In rejecting the employer's argument, the Court noted "[w]e have held that FLSA rights cannot be abridged by contract or otherwise waived because this would 'nullify the purposes' of the statute and thwart the legislative policies it was designed to effectuate." (46) Permitting a union to bargain away an individual employee's FLSA rights would subvert the goals underlying the statute. The Court justified its ruling by repeating many of the concerns raised earlier in Gardner-Denver, including that "many arbitrators may not be conversant with the public law considerations underlying the FLSA." (47) This question about the competence of arbitrators to decide public law claims was important because "FLSA claims typically involve complex mixed questions of fact and law," and "[t]hese statutory questions must be resolved in light of volumes of legislative history and over four decades of legal interpretation and administrative rulings." (48) The Court recognized that "arbitrator[s] may be competent to resolve many preliminary factual questions," but they "may lack the competence to decide the ultimate legal issue whether an employee's right to a minimum wage or to overtime pay under the statute has been violated." (49) The Court also repeated Gardner-Denver's reservations about the adequacy of arbitral procedures to protect statutory rights: "not only are arbitral procedures less protective of individual statutory rights ... but arbitrators very often are powerless to grant the aggrieved employee as broad a range of relief," including, for example, "liquidated damages, costs, or attorney's fees." (50)

The third case in this trilogy, the Court's 1984 decision in McDonald, held that submission of a civil rights claim to an arbitrator under a collective bargaining agreement did not waive a municipal employee's right to bring in court a claim under Section 1983. (51) Like Gardner-Denver and Barrentine, McDonald involved an employee who sought to assert a statutory claim after an arbitrator had rejected his contract claim. Summarizing its prior decisions in Gardner-Denver and Barrentine, the Court said that its rejection of a waiver-of-rights rule in those cases was "based in large part on our conclusion that Congress intended the statutes at issue in those cases to be judicially enforceable and that arbitration could not provide an adequate substitute for judicial proceedings in adjudicating claims under those statutes." (52) This concern applied with special force to Section 1983 actions because the statute's "very purpose" was "to interpose the federal courts between the States and the people, as guardians of the people's federal rights--to protect the people from unconstitutional action under color of state law." (53) The Court also repeated its prior conclusion that arbitration is a less desirable forum for resolving statutory claims, noting that (a) a specialist arbitrator may not have the "expertise required to resolve the complex legal questions that arise" in statutory cases; (b) the arbitrator "has no general authority to invoke public laws that conflict with the bargain between the parties"; and (c) "arbitral factfinding is generally not equivalent to judicial factfinding" because the record "is not as complete; the usual rules of evidence do not apply; and the rights and procedures common to civil trials, such as discovery, compulsory process, cross-examination, and testimony under oath, are often severely limited or unavailable." (54)

Taken together, this trilogy of cases signaled the Court's hostility to the suggestion that arbitration was a suitable replacement for litigation of statutory claims. Each decision underscored the Court's strongly held belief in the need for judicial vindication of statutory rights and the inadequacy of arbitral procedures to safeguard the fulfillment of the objectives embodied in these statutes. McDonald, however, proved to be the high-water mark for this view.

B. Gilmer and the End of the Non-Waivability Principle

Just seven years later, in Gilmer v. Interstate/Johnson Lane Co., (55) the Court reconceptualized arbitration of employment disputes altogether, so much so that it found the earlier trio of cases "far out of step with our current strong endorsement of the federal statutes favoring this method [arbitration] of resolving disputes." (56) So what had happened? How could a Court united in the belief that arbitration was inadequate to protect statutory rights repudiate that view only a few years later?

We suggest a number of reasons for this sea change. For one, the period between 1984 and 1991 brought about a seismic shift in the Court's composition. Chief Justice Burger and Justices Powell, Marshall, and Brennan all left the Court between 1984 and 1991, with Justices White and Blackmun following in 1993 and 1994, respectively. Thus, retirements and appointments reshaped the Court within a relatively brief time span.

For another, the seeds of discontent over the Court's hostility to arbitration of statutory claims had been sown, at least in part, in Barrentine, the only one of the trilogy not to have been decided by a unanimous Court. Chief Justice Burger, joined by then-Justice Rehnquist (who became Chief Justice in 1985, upon Burger's retirement), dissented in Barrentine, and their opinion foreshadows the Court's later movement away from nonwaivability. Justice Burger's dissent made two points, which together formed the cornerstone of the Court's new pro-arbitration jurisprudence.

The first was that, given the national policy favoring arbitration agreements, there was no reason why an employee, or an employee's union, could not in an arm's-length transaction agree to have statutory claims resolved conclusively in binding arbitrations The question in the dissent's view was not waiver of statutory claims, but consent (albeit not by the employee himself, but by a self-interested agent) to arbitrate statutory claims. (58) The fact that the agreement effectively waived the employee's right to file suit if disappointed with the arbitration award was untroubling to the dissenting Justices. The dissent simply dismissed the concern that arbitration procedures are inadequate to safeguard statutory rights. (59)

The dissent's second point was that FLSA cases involve "routine and relatively modest-sized claims," and thus could be removed from the docket of federal courts to stem the burgeoning litigation tide then swamping federal courts. (60) In the dissent's view, most employment cases are really little more than small claims actions, and like other small claims cases, they do not necessarily merit full-bore litigation in federal court. (61) Instead of "costly and time consuming" federal court litigation, arbitration provides "a swift, fair, and inexpensive remedy." (62) Here the dissenters were especially mindful of the burdens growing dockets were placing on federal courts, and feared that cases like Barrentine would further clog the pipelines. To drive the point home, the dissent ends with this call to arms: "This Court ought not be oblivious to desperately needed changes to keep the federal courts from being inundated with disputes of a kind that can be handled more swiftly and more cheaply by other methods." (63) These reasons, said the dissenters, counseled in favor of enforcing agreements to arbitrate statutory wage claims, even if the agreement was entered into by the employee's union and not the employee, and even if the employee never had access to a judicial forum to vindicate his statutory rights under FLSA. (64)

The arguments articulated in the Barrentine dissent took root in a number of the Court's post-McDonald cases, which quickly erased the stigma once attached to arbitration. The first wave of decisions, especially Southland Corp. v. Keating, (65) held that the Federal Arbitration Act ("FAA") was not, as had been previously thought, merely a procedural statute applicable only in federal court, but was instead a substantive statute that preempted state law and applied with equal force to actions filed in state court. (66) AS the Court put it, the FAA "withdrew the power of the states to require a judicial forum for the resolution of claims which the contracting parties agreed to resolve by arbitration." (67) Shortly after Southland, in Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., (68) the Court held that, when presented with a claim of arbitrability, the "first task of a court asked to compel arbitration of a dispute is to determine whether the parties agreed to arbitrate that dispute." (69) A court must make this determination by applying the "federal substantive law of arbitrability, applicable to any arbitration agreement within the coverage of the Act." (70) Thus arbitrability would no longer be a question of state law, but one of federal law, subject to the Court's control through the development of a federal common law of arbitration.

Once the Court declared that the Act had substantive content, all restraints were off. (71) The Court then dealt with lines of cases from both state and federal courts regarding employment and consumer contracts mandating arbitration. Without exception, the Court held that statutory rights--even core civil rights--could be relegated to arbitration by contract. We will first address the employment cases and then turn to the consumer matter.

1. Employment Cases

As noted above, until Gilmore, the presumption was that the Federal Arbitration Act did not apply to employment contracts. (72) The presumption was based on what was seen as a common-sense reading of Section 1 of the Act, which excludes from the Act's coverage "contracts of employment of seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce." (73) The turning point in the employment cases was Gilmer, (74) although, as a technical matter, the arbitration agreement at issue was not between Gilmer and his employer, a financial services company, but was a result of Gilmer's registration as a securities representative with the New York Stock Exchange. (75) The Exchange required all registered agents to agree to arbitrate "[a]ny controversy" between agents and their employers "arising out of the employment or termination of employment of such registered representative[s]." (76) Gilmer's employer terminated his employment when he turned sixty-two years old; Gilmer sued, claiming that his discharge violated the Age Discrimination in Employment Act of 1967 ("ADEA"). (77) The Court thus had to decide whether Gilmer's statutory ADEA claim could be subject to mandatory arbitration.

In answering yes to that question, the Court began by noting that the FAA was enacted "to reverse the longstanding judicial hostility to arbitration agreements that had existed at English common law and had been adopted by American courts," (78) and that the Act manifests a "'liberal federal policy favoring arbitration agreements.'" (79) The Court then said that it is "by now clear that statutory claims may be the subject of an arbitration agreement, enforceable pursuant to the FAA." (80) To support this contention, the Court steered clear of Gardner-Denver, Barrentine, and McDonald, and instead cited cases holding enforceable arbitration agreements relating to claims arising out of the securities, antitrust, and racketeering statutes. The thread that tied these cases to the ADEA was, in the Court's view, the fact that Congress had not "'evinced an intention to preclude a waiver of judicial remedies for the statutory rights at issue.'" (81) Accordingly, the burden was on Gilmer to show that Congress had forbidden waiver of his right to go to Court to enforce his ADEA claim, not on his employer to show that Congress permitted ADEA claims to be resolved through arbitration. This move marked a 180-degree turn in the Court's approach. In prior cases, the default was that Congress, by providing a direct right of action in federal court, and by failing explicitly to authorize arbitration as a litigation substitute, had decided that mandatory arbitration was not acceptable. (82) By shifting the burden to Congress to rule out arbitration, the Court opened wide the gates to it. (83)

Having shifted the burden, the Court quickly dispatched Gilmer's arguments. The Court first rejected the idea that litigation, and not arbitration, is essential to promote the social policies furthered by the ADEA. The Court asserted, without even a nod to its prior rulings, that "'[s]o long as the prospective litigant effectively may vindicate [his or her] statutory cause of action in the arbitral forum, the statute will continue to serve both its remedial and deterrent function.'" (84) The Court next rejected the idea that arbitration is structurally ill-suited to protect statutory rights. That idea, embraced by the Court a few years earlier, was now seen as "far out of step with our current strong endorsement of the federal statutes favoring this method of resolving disputes." (85) Although the Court recognized that judicial procedures are more expansive than those available in arbitration, an agreement to arbitrate "trades the procedures and opportunity for review of the courtroom for the simplicity, informality and expedition of arbitration." (86) The Gilmer Court concluded that arbitration is presumed to be equivalent to a court in all the ways that matter. (87)

Gilmer did not resolve, however, the most basic statutory interpretation question relating to the arbitration of employment matters--namely, whether the FAA applies to employment contracts. The Court resolved that issue in 2001, with its fractured ruling in Circuit City Stores, Inc., v. Adams. (88) Focusing on whether Section 1's exclusion of "contracts of employment of seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce," (89) put employment contracts out of the Act's reach, the majority held that the exclusion applied only to employment contracts with transportation workers, like seamen and railroad employees, but not more generally employees engaged in interstate commerce. (90) The majority brushed aside the broad text of the Act, finding that the phrase "engaged ... in commerce," along with the specific references to seamen and railroad workers, signaled Congress's intent that the provision should be "controlled and defined by reference to the enumerated categories of workers" and those similarly engaged in transportation. (91)

While the dissenters found many faults with the majority's reading of the Act, they concentrated much of their fire on the majority's refusal to grapple with the Act's legislative history, which "confirmed the fact that no one interested in the enactment of the FAA ever intended or expected that [section] 2 would apply to employment contracts." (92) Indeed, the history of the Act shows that "the potential disparity in bargaining power between individual employees and large employers was the source of organized labor's opposition to the Act, which it feared would require courts to enforce unfair employment contracts. That same concern ... underlay Congress' exemption of contracts of employment from mandatory arbitration." (93) In the aftermath of Circuit City, courts have routinely enforced boilerplate, mandatory arbitration provisions in employment contracts, even where there is clear evidence that, due to the disparity in bargaining power, the employee had no meaningful right to reject binding arbitration. (94) As various scholars, activists, and legislators have pointed out, inequality of bargaining power was not a significant issue in the business contexts for which the FAA was intended. (95)

Circuit City was not the Court's last word on the arbitrability of employment contracts. On April 1, 2009, the Court handed down 14 Penn Plaza LLC v. Pyett, (96) in which a 5-4 majority ruled that an arbitration clause in a collective bargaining agreement could waive union members' rights to pursue judicial relief for violations of the ADEA. (97) Although Justice Thomas's opinion for the majority denies overruling the Gardner-Denver-Barrentine-McDonald line of cases, at the very least it makes clear that they are on their last legs. (98) Penn Plaza may prove to be critically important because of the question of agency at play in the context of the collective bargaining agreement. Courts have already shown their willingness to enforce employment agreements entered into by individual employees, resting on the fiction that employees make a volitional choice (99) to bargain away their rights. (100) But it is another step altogether to find consent on the part of an employee when someone else bargained away his access to court. This is especially troubling because workers who join unions do so to enlist the unions' aid in matters of collective bargaining and resolution of contract-based disputes, not to cede control over their statutory rights. (101) The potential impact of Penn Plaza is great not only because of the more than sixteen million workers in the United States who are members of labor unions authorized to negotiate collectively on their behalf, (102) but also because of other situations in which an agency relationship may be inferred and rights waived.

2. Consumer Cases

To the extent that there was an analogue to the Gardner-Denver-Barrentine-McDonald line of cases in the consumer context, it was the Court's early decision in Wilko v. Swan, (103) which held that a mandatory arbitration agreement in a brokerage contract between a customer and a securities firm could not be enforced under the terms of the Securities Act. At the core of the Court's ruling was its concern that enforcing the arbitration agreement would run counter to Congress's directive providing a right of action, under the Securities Act that disputes be resolved in court. (104) The Court also expressed reservations about the ability of arbitrators to provide substantive and procedural protections Congress thought necessary to ensure fair adjudication of securities claims. (105)

Wilko's ultimate demise was presaged by a procession of cases in which the Court distinguished and criticized the decision while upholding mandatory arbitration provisions in the contexts of automobile dealer franchise arrangements, (106) antitrust claims, and securities claims arising under provisions of the Act not addressed in Wilko. (107) The chief case was Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., (108) which marked the first time the Court held that statutory claims, as well as contract claims, could be subject to mandatory arbitration. In the antitrust case of Mitsubishi, the Court, abandoned the "waiver of rights" theory developed in the employment context, claiming that arbitration agreements were nothing more than forum selection clauses, with no impact on claimants' substantive rights. (109)

When finally presented with a direct attack on Wilko in Rodriguez de Quijas v. Shearson/American Express, Inc., (110) the Court did not hesitate to overrule it. The Court saw Wilko as a relic of "'the old judicial hostility to arbitration'" (111) and said that, "[t]o the extent that Wilko rested on suspicion of arbitration as a method of weakening the protections afforded in the substantive law to would-be complainants, it has fallen far out of step with our current strong endorsement of the federal statutes favoring this method of resolving disputes." (112) Under the current view, "'arbitration is merely a form of trial to be used in lieu of a trial at law.'" (113) Since Rodriguez de Quijas, the Court has routinely found commercial contracts subject to mandatory arbitration, notwithstanding the fact that the claims arise under federal statutes that provide a right of action in federal court, and notwithstanding the presence of a state-law defense to arbitration. (114) The Court has also suggested that arbitration agreements may waive claims for remedies that would be available in court. (115) The difficulty for consumers is not simply that the Court will compel arbitration or will uphold remedy-stripping provision.

The Court seems willing to compel arbitration in cases where (a) the cost of arbitration might foreclose the ability of the consumer to vindicate his statutory right, or (b) the inability to bring aggregate litigation in arbitration will similarly defeat the plaintiff's claim. The key case concerning the potentially deterring costs of arbitration remains Green Tree Financial Corp.-Alabama v. Randolph. (116) The plaintiff, Larketta Randolph, purchased a mobile home with financing from the Green Tree Financial Corporation's Alabama subsidiary. (117) The loan agreement required Ms. Randolph to buy insurance to protect the Green Tree's interest in the mobile home in case of default. (118) The agreement also provided that all disputes arising from, or relating to, the contract, whether arising under case law or statutory law, would be resolved by binding arbitration. (119) Ms. Randolph brought suit against Green Tree, alleging that the loan agreement violated the Truth in Lending Act (120) by failing to disclose certain financing charges, and she later amended her complaint to add a claim under the Equal Credit Opportunity Act (121) for requiring her to waive statutory causes of action by compelling her to arbitrate all claims. (122) She brought the action on behalf of herself and other similarly situated borrowers. Green Tree filed a motion to compel arbitration, which the district court granted over Ms. Randolph's protest that she could not afford the high fees associated with the private process. (123) The court also denied Ms. Randolph's motion for class certification. (124) The Eleventh Circuit reversed, ruling that the arbitration provision was unenforceable it because it did not guarantee that steep filing costs, arbitrators' fees, and other arbitration expenses would not nullify Ms. Randolph's ability to vindicate her statutory rights. (125)

The Supreme Court reversed again, holding that Ms. Randolph had failed to meet her burden of showing that arbitration costs would be prohibitively expensive. (126) The Court recognized that Green Tree had drafted the agreement, which was silent as to costs, (127) and that Ms. Randolph cited materials reflecting the cost of arbitration in other, similar cases. (128) But that was not enough. According to the Court, "where, as here, a party seeks to invalidate an arbitration agreement on the ground that arbitration would be prohibitively expensive, that party bears the burden of showing the likelihood of incurring such costs" and "Randolph did not meet that burden." (129) The Court gave no indication of how Ms. Randolph could meet that burden, given that the arbitration agreement was silent as to costs, and the proceeding had not yet commenced. As the four Justice dissent pointed out, it made little sense to fault Ms. Randolph for failing to produce evidence of the costs of a proceeding not yet undertaken, and that burden should more properly be placed on Green Tree, which was "a repeat player in the arbitration required by its form contract," and had "superior information about the cost to consumers of pursuing arbitration." (130) As the dissent predicted, the only remedy for a consumer faced with a prohibitively expensive arbitration agreement is to return to court, post-arbitration, to challenge the allocation of costs--little help to a claimant who must sacrifice her statutory rights because she cannot afford the costs of arbitration in the first place. (131)

The Court's disposition of Randolph permitted it to avoid the other important problem presented in the case: whether an arbitration agreement is unenforceable where it precludes pursuit of a statutory claim as a class action. (132) The Court confronted that question three years later in Green Tree v. Bazzle, (133) but deadlocked on the right answer. (134) Bazzle involved two class actions by different groups of homeowners who alleged that Green Tree violated state lending laws. (135) The contracts were silent about the availability of class claims. (136) The company moved to compel arbitration; the homeowners contended that arbitration, if required, should be conducted on a class basis. (137) After a complicated set of procedural maneuvers, the court compelled arbitration, but the arbitrations proceeded on a class basis, and they resulted in a substantial award to the homeowners. (138)

After affirmance by the South Carolina Supreme Court, (139) the U.S. Supreme Court vacated and remanded, although it produced no majority opinion. Justice Breyer, writing for himself and Justices Scalia, Souter, and Ginsburg, concluded that because the arbitration agreement was silent as to whether class cases were permitted, the decision on that issue belonged to the arbitrator and not the courts. (140) Accordingly, the decision should be vacated and the case remanded to permit the arbitrator to make that threshold decision. (141) Justice Stevens concurred in that judgment, although he would have preferred simply to affirm the judgment of the Supreme Court of South Carolina. (142) He made clear that he joined Justice Breyer to avoid the Court issuing a ruling with "no controlling judgment." (143) Chief Justice Rehnquist, joined by Justices O'Connor and Kennedy, dissented, arguing that it is for the courts to determine the meaning of arbitration agreements, and finding that in this case, the South Carolina courts erred by not concluding that the contract excluded class action treatment of claims. (144)

What is most significant about Bazzle, however, is what the Court did not say. Not a single Justice voiced concern about the possibility that, on remand, the arbitrator would read the agreement, as did the dissent, to forbid class treatment of claims that in all likelihood would not be viable unless they were subject to aggregation. (145) Not surprisingly, in the wake of Bazzle, sophisticated employers and sellers modified their arbitration agreements explicitly to forbid class treatment of claims. (146) Equally unsurprisingly, because Bazzle does not suggest that contracting-out of procedural guarantees would be improper, many lower courts have upheld arbitration provisions that expressly forbid the use of class actions. (147)

II. THE EMPIRICAL DEFENSE

Having seen the Court take a sharp turn in favor of arbitration, the question we now explore is this: what are the implications of the Supreme Court's revised outlook? A great deal of ink has already been spilled on this topic, and we do not attempt to review or evaluate all the arguments and evidence that have been brought to bear. We do wish to make a few observations about the existing literature on mandatory arbitration of statutory rights, which, as we explain, does not provide empirical support for the chief defense of arbitration--namely, that it provides an adequate forum for the vindication of statutory rights at a lower price than conventional litigation.

A central feature of the Court's new pro-arbitration outlook is its expansion of the scope of the FAA from voluntary arbitration agreements between businesses, or between management and unions, to pre-dispute contract provisions drafted by businesses and imposed on individual employees, consumers, and other relatively weaker parties. (148) The Court's view that individual employees and consumers can waive their right to a judicial forum for the adjudication of statutory claims has rested on the assumption that arbitration provisions constitute merely "forum-selection clauses" that in no way compromise substantive rights. (149) The Court first announced this rule in Mitsubishi (150) and then extended it in Gilmer. (151) Since those decisions, a large body of literature has developed either challenging or shoring up the Court's constructed dichotomy between procedure and substance.

Following earlier theoretical and policy arguments, a growing number of scholars have undertaken empirical research comparing the outcomes of arbitration and litigation. (152) While the role of mandatory arbitration remains controversial, the empirical literature has largely clustered around a set of common conclusions. The prevailing wisdom is that arbitration is faster and cheaper than litigation, and that claimants actually win more often in arbitration than in litigation, though these winners get lower awards. (153) Defenders of mandatory arbitration assert that arbitration is shown to be generally better for everyone involved, except those few claimants who might have received large jury awards through litigation. (154) We question whether the existing empirical research so readily leads to this conclusion.

A. Fundamentally Flawed Data

As a preliminary matter, we question the validity of any generalizations about the comparable value of arbitration and litigation given the data upon which the assessments rely. Putting aside the methodological flaws of any individual study, the existing data sets are necessarily limited by the private nature of arbitration. Because arbitration is conducted behind closed doors, analyses of results reflect only those arbitrations that the parties and arbitral providers have chosen to make available. Of the decisions that are published, many are heavily redacted, making their characteristics difficult to assess and compare. (155)

That only a subset of arbitration decisions is available for analysis would be less of a problem if the publicly-available decisions were representative of those unavailable. Unfortunately, there is no basis for making such an assumption. On the contrary, the arbitral providers that make their decisions publicly available may be those with the least to hide. Richard Bales has suggested that studies relying on data from the American Arbitration Association ("AAA"), an organization that, for example, makes labor and employment decisions available online, will necessarily under-represent unscrupulous employers or egregiously unfair arbitration provisions, because the AAA is a relatively reputable organization, which as a matter of practice refuses to arbitrate under rules it deems unfair. (156) It is reasonable to surmise that, in addition to the self-selection bias of the better arbitral providers publishing decisions, there may also be an opt-out function by defendants who want the leeway to engage in bad conduct and keep their acts and liability shielded from public view. Arbitration awards against these rogue companies will never make it into the studies.

On top of the problems of the limited and possibly skewed data set of arbitrated cases, comparing arbitrated cases with those in litigation may be comparing apples to oranges. (157) As a matter of study design, there is the basic fact that no two cases are exactly the same, so it is difficult to control for all potentially relevant factors. Even if the sample sizes were large enough to smooth out differences in individual cases, the aggregate analysis would remain unreliable because of the possibility of systematic differences in the types of cases funneled through arbitration and through the courts. Possible differences between the sample sets could include types of plaintiffs, defendants, or claims. In the employment context at least, Stephen Ware suggests giving up on outcome comparison research altogether, because employers who use arbitration might have better (or worse) lawyers or human resources departments, more experience defending employment discrimination claims, better (or worse) reputations for how they treat their employees, or be more capable of paying large verdicts. (158)

Outcome-focused studies also neglect claims that are either settled before final disposition or never brought to arbitration or to court. (159) Whether one views the resolution of claims without adjudication as a good or bad outcome in itself, it necessarily makes dollar-based outcome studies incomplete. (160) Filtering mechanisms may disproportionately keep certain kinds of claims or parties out of certain fora. For example, employers that choose arbitration may be more likely to have internal dispute resolution systems in place, thereby screening out easily resolvable or non-meritorious cases. (161) Richard Bales suggests that potential defendants who use arbitration provisions in contracts may screen out cases through the use of "lopsided" arbitration agreements whose terms make arbitration so imbalanced as to make the pursuit of arbitration seem worthless to potential claimants. (162) This is not merely speculation. For example, Bales notes that Circuit City continues to use an agreement already ruled to be unenforceable, presumably because the corporation saves enough by "scaring off" litigants to make it worthwhile to re-litigate challenges to the provision. (163) Whatever the reasons for potential claimants not reaching arbitration, assessments of the accessibility of arbitration that do not account for the missing claimants suffer as a result.

Aside from the fundamental flaws in the data sets, the existing empirical studies fail to account for significant factors. In the next section, we turn our attention to the missing considerations.

B. Neglected Societal Costs

We suggest that existing studies defending arbitration as more efficient than litigation neglect significant externalities. The current studies emphasize decreased costs of money and time, and they demonstrate the comparable social utility of the fora by measuring relative win/loss rates. (164) The best of these studies recognize the distinction between costs to defendants and to plaintiffs, and between up-front filing fees, arbitrators' fees, and costs of process. As Christopher Drazhozal recently concluded, administrative costs and arbitrators' fees are more expensive in arbitration than are filing fees in litigation, so the argument that arbitration is the cheaper alternative necessarily depends on savings on attorney's fees and other costs of the process. (165) Even where claimants' costs and arbitrators' fees are limited by caps or absorbed by defendants, cost savings result largely from decreased process costs. (166)

The problem with drawing conclusions about social benefits from these decreased process costs, however, is that such analyses gloss over three critical points: (a) decreased process disproportionately benefits defendants; (b) a major source of decreased costs for defendants is not actually decreased process costs but decreased awards (which cannot be divorced from considerations of costs for claimants) (167); and (c) these outcomes have effects beyond the interests of the individual parties to the disputes.

To the extent that studies of arbitration focus on compensation to individual claimants, they neglect aggregate societal benefits of public litigation, such as the enforcement of public rights and deterrence of behavior Congress chose to ban. In other words, there are a number of social functions of courts and adjudicative processes that arbitration may not provide, or may not provide as well, and to claim arbitration is more efficient simply because it is cheaper, while failing to account for the lost social utility, makes little sense. (168) Adjudication by public courts produces the following societal benefits that are missing from arbitration: development of a consistent body of precedent, through reasoned opinions that interpret statutes as applied to facts and may be corrected by appeals courts, (169) democratic dialogue between courts and legislatures about correct interpretation of laws, (170) public education of potential bad actors regarding limits of the law, (171) potential victims who might bring suit to challenge bad conduct, citizenry with influence on their legislators, and customers who might influence bad actors by voting with their dollars, and the democratic value of jury deliberations and awards. (172)

All of the sources of decreased costs also decrease the deterrence of statutorily-prohibited behavior. Limited discovery cuts against the party with the burden of proof; the absence of a jury also tilts against the plaintiff. (173) The key reasons why the process of arbitration is considered to be less costly than the process of litigation--limited discovery, lack of appeal rights, the absence of the jury--all cut against the enforcement of laws.

Contrary to popular wisdom, faster, cheaper litigation is not simply a win-win situation for all concerned. (174) On the aggregate level, decreasing the time and money sunk into dispute resolution harms not only lawyers, as some suggest, but arguably also the public that benefits from the deterrent effects of these potential costs. (175) Costs for defendants may be driven down even further by additional provisions that only decrease potential plaintiffs' recovery amounts or likelihood of recovery but provide no benefit to plaintiffs, and arguably none to society as a whole: limited damages provisions, reduced statutes of limitations, and unavailability of class actions that aggregate small claims and make them worth pursuing. (176)

Certainly, decreased contributions of time and money will in many cases benefit individual plaintiffs. For example, decreased discovery or the lack of appeal opportunities may make adjudication less expensive for plaintiffs who would otherwise be outspent by defendants in litigation. But these options are available to litigants in judicial proceedings as well. Nothing in the Federal Rules forbids the parties from agreeing to limit discovery, to try a case to a judge or magistrate judge instead of a jury, or to forego an appeal. If, as some assert, the automatic sacrifice of these procedures makes arbitration more accessible than litigation, (177) that effect deserves consideration, as accessibility has a significant social value. Yet we need to clarify not only whether, in fact, arbitration increases accessibility, but also whether or not we have decided as a society that the increased accessibility justifies the sacrifices in the quality of justice dispensed. (178)

The most compelling argument for promoting arbitration is that it is available for low-wage workers and low-claim claimants who would otherwise have access to no forum at all, because courts are simply too expensive. Lawyers will not take their cases, the argument goes, because these individuals cannot pay out-of-pocket expenses, and their likely recoveries are too low to make a contingency fee worthwhile. Arbitration, on the other hand, can be navigated more easily pro se, given its informality. (179)

Putting aside our skepticism about the assumption that arbitration is simple and accessible for pro se litigants, and fairly adjudicated in spite of the imbalance in representation or lack thereof, we ask if there might be something wrong with a system of justice based on the premise that lawyers, full discovery, and other niceties of due process are out of reach for a certain class of society. Given the rough approximation of justice that arbitration provides, (180) are we ready to give up the adjudication of disputes based on the rule of law? Even if we were comfortable with individuals making this choice for themselves, do we want to permit them prospectively to bargain away public rights? (181)

III. CONTRACTING (OUT) STATUTORY RIGHTS

A number of commentators have criticized the Supreme Court's assumption that arbitration provisions are simply "forum-selection clauses" with no limiting effect on the application of substantive law. (182) Stephen Ware has argued that the Supreme Court's decisions have denied the ways in which arbitration reduces "mandatory" rules to "default" rules and thereby "jeopardizes mandatory rules of law." (183) He defines mandatory law as that from which deviation could not be an enforceable contract term, whereas default law is that from which terms of deviation would be enforceable, and argues that mandatory law ought not to be arbitrable. (184) What Ware does not dissect is the normative question of which laws ought to be mandatory, and which should be subject to contrary contract terms. (185) Perhaps even more troubling than the failure to recognize connections between procedural and substantive law, underlying the Supreme Court's embrace of arbitration is the suggestion that all laws are subject to contrary agreement between private parties: no law is mandatory.

A year after Circuit City, (186) writing for the Seventh Circuit in a decision enforcing an arbitration provision in a telephone company's publicly filed tariff, Judge Easterbrook declared, "[t]he Supreme Court has never held that any entitlement is outside the domain of contract, unless the statute forbids waiver." (187) Summarizing relevant cases, Judge Easterbrook included among "entitlement[s]" the right to a jury trial, the standard of proof beyond a reasonable doubt, the right to political expression, the right to attorneys' fees, and "an entire civil rights claim." (188) Putting aside the accuracy of Judge Easterbrook's characterization of the Supreme Court's history and the particular cases he cites, his opinion points to a very real and radically libertarian undercurrent in the Court's recent arbitration jurisprudence. (189)

The hints at this philosophy in Mitsubishi and Gilmer are particularly worthy of note. In response to concerns that forcing disputes into a private arbitral forum would harm public policies set out by the legislature, the Mitsubishi Court suggested that because Congress left the key to enforcement in private hands, private interests could trump public law values. (190) The Court highlighted that citizens are not required to bring antitrust suits, nor are plaintiffs required to seek judicial approval prior to settlement of their claims. (191) Yet the fact that the private attorneys general model gives way to rights-holders' individual interests and liberties does not justify a fully private conception of such rights. As the Mitsubishi Court recognized, a "prospective waiver of a party's right to pursue statutory remedies" would be void "as against public policy." (192) Though the Court's holding depends entirely on the notion that arbitration provisions are "forum-selection clauses" (193) rather than waivers of "substantive rights," (194) the Court fails to explain the substance behind this distinction. Had the Court addressed why a substantive waiver would be void, it might have been forced to confront the reality that pre-dispute arbitration clauses cause the same societal harms. (195)

Both prospective waivers of substantive rights and pre-dispute arbitration clauses decrease deterrence of socially undesirable behavior and undermine the normative force of mandatory public laws. The examples listed by the Court do not suggest otherwise, as the decisions not to pursue a claim or to settle it are made after injuries have occurred, the time when rights-holders tend to value their rights more highly than when faced with hypothetical injuries pre-dispute. (196) In this sense, rights-holders' interests will be more likely to line up with the enforcement of a statute after its violation. (197) Even if one agreed with Judge Easterbrook that individuals should be free to sell their fundamental rights for "something valued more highly," (198) this would still leave open the question of when that value should be measured: there is no reason to believe the pre-dispute value is accurate or socially optimal; on the contrary, it is less likely to be so because it is evaluated before the injury is defined. (199) While the Court claims not to have abandoned completely public policy exceptions to freedom of contract, (200) its refusal to take into account pre-dispute dynamics suggests otherwise.

The Court's pro-arbitration trend also reflects a profoundly libertarian bent in its denial of the role of social inequality in informing legislation. (201) In Gilmer, the Court pronounced that claims of unequal bargaining power are best resolved on an individual, case-by-case basis. (202) This assertion directly contradicted the Court's prior jurisprudence, as well as the history of the Federal Arbitration Act. (203) It also neglected the legislative purpose, present in anti-discrimination statutes among others, to protect a particular "class" of parties from another, more powerful class. (204) While it may be said that all law, whether statutory or based on common law, reflects value judgments, to suggest that all interactions between individuals carry the same social significance (205) neglects differences in societal position and throws into question the role of the legislature in correcting undesirable imbalances of power and resources.

At the outset of this Article we threw down the gauntlet by suggesting that the Court's new pro-arbitration jurisprudence marked a return to the days of Lochner, where liberty of contract was unrestrained by public law principles and the Court invalidated statutes that purported to interfere with the right of contract. We were not just trying to be provocative by raising the specter of Lochner. We believe that the Court's arbitration decisions come perilously close to reinvigorating that case.

The Court's key move towards Lochner was its re-conceptualization of arbitration in Mitsubishi. There the Court discarded the waiver theory developed in Gardner-Denver and its progeny, which posited that arbitration was ill-suited to safeguard statutory rights, and announced instead that a pre-dispute commitment to arbitration was simply a forum-selection decision. (206) The Court's analogy between mandatory arbitration and forum selection provisions is flawed, however, because forum selection rules are not choice of law rules. Although forum selection clauses dictate the forum in which the dispute will be resolved, they do not dictate substantive law, (207) especially where the substantive rule of decision is provided by a federal statute. To be sure, a forum selection clause may require a litigant to press a claim in a court of the other side's choosing; forum selection clauses may compel litigation in inconvenient court, or force a litigant to bring a claim in state court even though the claim might otherwise be brought in federal court. A forum selection clause may require a litigant to sacrifice the procedural advantages she would have in federal court, but regardless of the forum selected, the substantive law remains the same.

Arbitration is different. In arbitration, the substantive "law" that the arbitrator must apply is the contract, not the background federal statutes, (208) and courts have routinely upheld contracts that derogate or nullify statutory provisions, including the right to attorney's fees, certain damages, jury trials, and on and on. Thus, the Court's promise in Mitsubishi that in arbitration "a party does not forgo the substantive rights afforded by the statute" is, by now, hollow, and arbitration has become, with the Court's blessing, a way of voiding statutory rights by contract--just as in Lochner.

As hard as we have searched, we have yet to find a limiting principle in the Court's arbitration jurisprudence. Are there any rights so fundamental that they may not be bargained away? Judge Easterbrook, for his part, thinks there are none. As he put it, in language that eerily mimics Lochner: "One aspect of personal liberty is the entitlement to exchange statutory rights for something valued more highly." (209) Since Mitsubishi, the Court has consistently ruled that one is "entitled" to exchange one's statutory right to bring suit in court for arbitration, making what we believe is the dubious assumption that one obtains "something valued more highly" in the bargain. (210) Courts have also ruled that one is entitled to exchange her right to attorneys' fees, punitive damages, jury trials, appeals, discovery, and so forth as part of the same bargain--that is, the "bargain" to get a job, or the "bargain" to buy a cell phone. (211) The question, of course, is whether bargaining away substantive rights--the right to earn the minimum wage for one's labor, the right to work in an environment free of sexual harassment, or the right to be considered for a promotion regardless of one's race--rights hard-fought to obtain, are any different.

CONCLUSION

There is no question that litigation is expensive, but we remain puzzled as to why the solution to this problem should be arbitration. All the reasons arbitration is cheaper than litigation cut against the rights-holder or against the enforcement of laws. Commentators who argue that arbitration may be the only or best option for some rights-holders implicitly accept a deeply cynical conception of who is entitled to enjoy full remedies for a deprivation of rights. If the problem is a lack of counsel, that lack requires attention, as giving up on courts for certain segments of society is not a responsible solution. Mandatory arbitration provisions also block access to courts even for those who might be in the position to pursue litigation. In this way, enforcement of arbitration provisions implies resigning ourselves to the idea that judicial enforcement of the rule of law is just too expensive for society as a whole to afford. We think this sacrifice is too great.

(1.) 129 S. Ct 1456, 1474 (2009).

(2.) See Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 35 (1991).

(3.) 42 U.S.C. [section][section] 2000e-1 to -17 (2006) ("Title VII").

(4.) 29 U.S.C. [section][section] 201-19 (2006) ("FLSA").

(5.) 42 U.S.C. [section][section] 12101-213 (2006) ("ADA").

(6.) 29 U.S.C. [section][section] 621-34 (2006) ("ADEA").

(7.) 42 U.S.C. [section] 1981 (2006).

(8.) There is an ongoing, robust policy debate over the impact of the growth of predispute, mandatory arbitration on our civil justice system. See, e.g., Lewis L. Maltby, Private Justice: Employment Arbitration and Civil Rights, 30 COLUM. HUM. RTS. L. REV. 29 (1998); Rex R. Perschbacher & Debra Lyn Bassett, The End of Law, 84 B.U.L. REV. 1 (2004); Judith Resnik, Whither and Whether Adjudication?, 86 B.U.L. REV. 1101 (2006); Jean R. Sternlight, Creeping Mandatory Arbitration: Is It Just?, 57 STAN L. REV. 1631, 1648-53, 1671-72 (2005).

(9.) Theodore J. St. Antoine, Mandatory Arbitration: Why It's Better Than It Looks, 41 U. MICH. J.L. REFORM 783, 795-96 (2008).

(10.) See generally Taylor Lincoln & David Arkush, Public Citizen, The Arbitration Debate Trap 23 (2008), available at http://www.citizen.org/documents/Arbitration DebateTrap(Final).pdf (canvassing the empirical evidence on arbitration in depth and demonstrating why the evidence is insufficient to back up claims about mandatory arbitration's efficiency and fairness); David Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 STAN. L. REV. 1557, 1564-67 (2005) (describing differences between cases adjudicated in court versus cases resolved in arbitration); Stephen J. Ware, The Effects of Gilmer: Empirical and Other Approaches to the Study of Employment Arbitration, 16 OHIO ST. J. ON Disc. RESOL. 735, 753-57 (2001) (discussing studies comparing arbitration and litigation); id. at 756-57 (suggesting empirical literature may be futile); W. Mark C. Weidemaier, From Court Surrogate to Regulatory Tool: Re-Framing the Empirical Study of Employment Arbitration, 41 U. MICH. J.L. REFORM 843, 854-56 (2008) (describing how outcome-oriented research neglects "filtering" mechanisms that alter which cases end up in litigation or arbitration).

(11.) See, e.g., Myriam Gilles, Opting Out of Liability: The Forthcoming, Near-Total Demise of the Modern Class Action, 104 MICH. L. REV. 373, 430 (2005) (arguing that predispute, mandatory arbitration provisions which waive class actions threaten to undermine statutory guarantees in employment and consumer cases); Samuel Issacharoff & Erin F. Delaney, Credit Card Accountability, 73 U. CHI. L. REV. 157, 170-77 (2006) (arguing that predispute, mandatory arbitration provisions which waive class actions threaten to drive out low value claims, leaving corporations unaccountable); Sternlight, supra note 8, at 1651-52. A 2002 report by Public Citizen, The Costs of Arbitration, available at http://citizen.org/publications/print release.cfm?ID=7173, provides substantial empirical support for the argument that in consumer cases, costs of arbitration to the consumer substantially exceed the costs of litigation, even if the aggregate costs are lower.

(12.) David Schwartz argues that most arbitration scholarship legitimates mandatory arbitration, sometimes even contrary to the intent of the author, by focusing on the detailed application of the phenomenon, while ignoring "big picture questions" that are no longer in vogue. Yet he also suggests that hammering away at big picture questions makes writers irrelevant to the real world and foolish to academics, so perhaps writing about arbitration at all is worthless. See David S. Schwartz, If You Love Arbitration, Set it Free: How "Mandatory" Undermines "Arbitration", 8 NEV. L.J. 400, 420-21 (2007). We believe, however, that the big picture remains highly relevant, especially given the recent Penn Plaza decision and legislative proposals to exempt employment and consumer contracts from the reach of the FAA. See Arbitration Fairness Act of 2007, H.R. 3010, 110th Cong. One reason why we focus on mandatory arbitration is that, unlike some of the other ills that confront the civil justice system, this one could be remedied, in large measure, by congressional action.

(13.) In the famous case of Lochner v. New York, 198 U.S. 45, 62 (1905), the Supreme Court struck down a New York law that aimed to protect the health and welfare of bakery employees by restricting their working hours to a maximum of ten per day and sixty per week. The Court stated that whether the statute was "valid as a labor law, pure and simple, may be dismissed in a few words. There is no reasonable ground for interfering with the liberty of a person or the fight of free contract" to determine the hours of a baker. Id. at 57. The Court went on to hold that "the freedom of master and employee to contract with each other in relation to their employment, and in defining the same, cannot be prohibited or interfered with, without violating the Federal Constitution." Id. at 64. Lochner was formally overruled in West Coast Hotel Co. v. Parrish, 300 U.S. 379 (1937), and it is generally understood to represent since-discredited views, but many of the arguments made to support a broad application of the Federal Arbitration Act recall Lochner's libertarian underpinnings.

(14.) See, e.g., Alexander v. Gardner-Denver Co., 415 U.S. 36, 48 (1974) discussed infra Part I.A.

(15.) See, e.g., supra note 11 and accompanying text.

(16.) Judicial review of arbitration awards is exceptionally narrow and highly deferential. See, e.g., Major League Baseball Players Ass'n v. Garvey, 532 U.S. 504, 509 (2002) ("courts are not authorized to review an arbitrator's decisions on the merits" even if the fact-finding was "silly"); Brentwood Med. Assocs. v. United Mine Workers, 396 F.3d 237, 239 (3d Cir. 2005) (upholding arbitration award even though it was grounded on "glaring" mistakes of law); Rich v. Spartis, 516 F.3d 75, 82 (2d Cir. 2007) (holding that an arbitrator's ruling may be overturned only for manifest disregard for the law and only in the "severely limited" circumstance where the court finds that the "arbitrators knew of the governing legal principle yet refused to apply it").

(17.) Patricia A. Greenfield, How Do Arbitrators Treat External Law?, 45 INDUS. & LAB. REL. REV. 683, 692-93 (1992) (many arbitration decisions do not apply statutory law at all, or do so making conclusory judgments that neither interpret nor create precedent). But see Christopher R. Drahozal, Is Arbitration Lawless?, 40 LOY. L.A.L. REV. 187, 207-15 (2006) (arguing that current evidence is inconclusive as to whether arbitration stifles development of precedent).

(18.) See, e.g., Resnik, supra note 8, at 1154; Sternlight, supra note 8, at 1656-58 (2005); Harry T. Edwards, Where Are We Heading with Mandatory Arbitration of Statutory Claims in Employment?, 16 GA. ST. U. L. REV. 293, 297 (1999).

(19.) For example, the source information on Lexis indicates that AAA "irregular[ly]" releases to Lexis redacted Labor Arbitration Awards when the parties consent. See Lexis Nexis, http://w3.nexis.com/sources/scripts/info.pl?294112.

(20.) See, e.g., Lisa B. Bingham, Employment Arbitration: The Repeat Player Effect, 1 EMP. RTS. & EMP. POL'Y J. 189, 200 (1997).

(21.) See, e.g., Wither and Whether, supra note 8, at 1139 (reproducing Prof. Resnik's cell phone contract containing a class-action waiver clause); Jean R. Sternlight, As Mandatory Binding Arbitration Meets the Class Action, Will the Class Action Survive?, 42 WM. & MARY L. REV. 1, 12-16 (2000).

(22.) See, e.g., Jean R. Sternlight, The Rise and Spread of Mandatory Arbitration As a Substitute for the Jury Trial, 38 U.S.F.L. REV. 17, 35-38 (2003); Margaret L. Moses, Privatized "Justice", 36 LOYOLA U. CHI. L.J. 536, 640-41 (2005).

(23.) See generally HERMAN SCHWARTZ, ed., THE REHNQUIST COURT: JUDICIAL ACTIVISM ON THE RIGHT, 518-19 (2002); Andrew M. Siegel, The Court Against the Courts: Hostility to Litigation As an Organizing Theme in the Rehnquist Court's Jurisprudence, 84 TEX. L. REV. 1097, 1097 (2006).

(24.) The Court has set a high bar for environmental cases, see, e.g., Summers v. Earth Island Inst., 129 S. Ct. 1142, 1150-51 (2009); Lujan v. Defenders of Wildlife, 504 U.S. 555, 563-67 (1992); Lujan v. Nat'l Wildlife Fed., 497 U.S. 871, 887 (1990). But see Massachusetts v. EPA, 549 U.S. 497 (2007). These restrictive cases have also been applied to limit standing by consumers challenging the inadequacy of safety regulations promulgated to benefit consumers. See, e.g., Pub. Citizen, Inc. v. Nat'l Highway Traffic Safety Admin., 489 F.3d 1279, 1296 (D.C. Cir. 2007), supplemental opinion, 513 F.3d 234, 238-39 (2008) (per curiam).

(25.) See FED. R. CIV. P. 8, 12(b)(6).

(26.) Bell Atlantic Corp. v. Twombly, 550 U.S. 544, 556-57 (2007).

(27.) See, e.g., Riegel v. Medtronic, Inc., 128 S. Ct. 999, 1008 (2008) (preempting state law claims for medical devices specifically approved by the Food and Drug Administration); Geier v. Am. Honda Co., 529 U.S. 861, 874-75 (2000) (preempting state law claims for automobile safety devices); Ingersoll-Rand Corp. v. McClendon, 498 U.S. 133, 138-39 (1990) (holding that ERISA broadly preempts state law); Aetna Health Co. v. Davila, 542 U.S. 200, 209 (2004) (same). See generally THOMAS O. MCGARITY, THE PREEMPTION WAR: WHEN FEDERAL BUREAUCRACIES TRUMP LOCAL JURIES (2008); WILLIAM W. BUZBEE, ed., PREEMPTION CHOICE: THE THEORY, LAW, AND REALITY OF FEDERALISM'S CORE QUESTION (2008).

(28.) See, e.g., Corr. Servs. Corp. v. Malesko, 534 U.S. 61, 71-72 (2001) (refusing to find a federal right of action for a federal prisoner abused by guards at a private prison operating under contract with the federal government); Alexander v. Sandoval, 532 U.S. 275, 288-89 (2001) (refusing to find a right of action for minorities subject to discrimination under the terms of a federal contract); Blessing v. Freestone, 520 U.S. 329, 344-45 (1997) (barring actions by beneficiaries of federal benefits administered under contract with state agencies). The Court's remedy stripping of the Employment Retirement Income Security Act ("ERISA") has left every American who has a pension, health insurance, or other insurance, without a remedy even in the face of deliberate misconduct by the provider, notwithstanding that Congress authorized beneficiaries to bring suit. See 29 U.S.C. [section] 1132(a) (2006); see, e.g., Mertens v. Hewitt Assocs., 508 U.S. 248, 254-55 (1993); Great-West Life & Annuity Ins., 534 U.S. 204, 210 (2002); Aetna Health Inc. v. Davila, 542 U.S. 200, 207-08 (2004); see generally John H. Langbein, What ERISA Means by "Equitable": The Supreme Court's Trail of Errors in Russell, Mertens and Great-West, 103 COLUM. L. REV. 1317 (2003). Justice White's dissent in Mertens explained that these decisions result in the perverse "anomaly of ... leav[ing] those Congress set out to protect ... with 'less protection than they enjoyed before ERISA was enacted.'" 508 U.S. at 267 (citation omitted).

(29.) See, e.g., Evans v. Jeff D., 476 U.S. 1179 (1986) (approving the use of sacrifice offers in civil rights cases); Buckhannon Bd. & Care Home, Inc. v. W. Va. Dep't of Health & Human Res., 532 U.S. 598, 608-09 (2001) (discarding the catalyst rule in statutory fee cases); see also Sole v. Wyner, 551 U.S. 74 (2007) (applying Buckhannon).

(30.) 415 U.S. 36 (1974).

(31.) 450 U.S. 728 (1981).

(32.) 466 U.S. 284 (1984).

(33.) McDonald, 466 U.S. at 289; Barrentine, 450 U.S. at 745; Gardner-Denver, 415 U.S. at 56-57.

(34.) 415 U.S. 36 (1974).

(35.) Id. at 56.

(36.) Id. at 56-57.

(37.) Id. at 57.

(38.) Id.

(39.) Id.

(40.) Id. at 57-58. Here the Court relied on its earlier ruling in Wilko v. Swan, 346 U.S. 427 (1953), which rejected the idea that the efficiency of arbitration outweighed the statutory fight of judicial review in a case involving sales of securities. See also Scherk v. Alberto-Culver Co., 417 U.S. 506, 532 (1974) (Douglas, J., dissenting) ("An arbitral award can be made without explication of reasons and without development of a record, so that the arbitrator's conception of our statutory requirement may be absolutely incorrect yet functionally unreviewable, even when the arbitrator seeks to apply our law. We recognized in Wilko that there is no judicial review corresponding to review of court decisions. The extensive pretrial discovery provided by the Federal Rules of Civil Procedure for actions in district court would not be available. And the wide choice of venue provided by the 1934 Act would be forfeited. The loss of the proper judicial forum carries with it the loss of substantial fights.") (citations omitted). Wilko was expressly overruled by the Court in Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 480 (1989), which is discussed in some detail infra at nn. 110-114 and accompanying text.

(41.) Gardner-Denver, 415 U.S. at 45.

(42.) Id. at 57-58.

(43.) Id. at 58.

(44.) Id. One irony is that the Court thought that the only way that arbitration might provide a suitable forum for the resolution of an employee's Title VII claims would be to engraft all of the guarantees of federal court litigation--especially discovery and appellate review--onto the arbitration process. As the Court put it, "a standard that adequately insured effectuation of Title VII rights in the arbitral forum would tend to make arbitration a procedurally complex, expensive, and time-consuming process. And judicial enforcement of such a standard would almost require courts to make de novo determinations of the employees' claims. It is uncertain whether any minimal savings in judicial time and expense would justify the risk to vindication of Title VII rights." Id. at 59.

(45.) 450 U.S. 728, 730 (1981).

(46.) Id. at 740.

(47.) Id. at 743.

(48.) Id.

(49.) Id.

(50.) Id. at 744-45.

(51.) 466 U.S. 284.

(52.) Id. at 289.

(53.) Id. at 290.

(54.) Id. at 290-91 (citations omitted).

(55.) 500 U.S. 20 (1991).

(56.) Id. at 30 (citation omitted).

(57.) Barrentine, 450 U.S. 728, 746-47 (1981) (Burger, C.J., dissenting).

(58.) Id. at 747-48.

(59.) Id.

(60.) Id at 746.

(61.) Id.

(62.) Id. at 747-48.

(63.) Id. at 753.

(64.) See id. The empirical evidence supporting Chief Justice Burger's claim about the "inundat[ion]" of the courts was, it turns out, considerably overstated. See, e.g., Michael J. Saks, Do We Really Know Anything About the Behavior of the Tort Litigation System--And Why Not?, 140 U. PA. L. REV. 1147 (1992) (refuting the "myth" of litigation explosion).

(65.) 465 U.S. 1 (1984).

(66.) Id. at 14-15.

(67.) Id at 10.

(68.) 473 U.S. 614 (1985).

(69.) Id. at 625.

(70.) Moses H. Cone Mem'l Hosp. v. Mercury Constr. Corp., 460 U.S. 1, 24-25 (1983) (stating further that the Federal Arbitration Act evinces "a liberal federal policy favoring arbitration agreements" and that federal law governs the interpretation and enforcement of such agreements).

(71.) To illustrate the consequence of this doctrinal development, consider Doctors' Assocs., Inc. v. Casarotto, 517 U.S. 681 (1996), which involved a complex franchise agreement that obligated a Subway sandwich shop franchisee in Great Falls, Montana, to arbitrate any dispute with the franchisor (Doctors' Associates) in Connecticut under Connecticut state law. Casarotto argued that the agreement violated Montana law, which required form contracts containing arbitration agreements to disclose on the front page in underlined capital letters that the contract is subject to arbitration. The Court brushed that argument aside, finding the Montana law preempted by the Federal Arbitration Act, which the Court read as preempting any state law requirement that is applicable only to arbitration, because such provisions can undermine the strong federal policy favoring arbitration. Id. at 687. Thus, efforts by states to highlight that a contract contains a mandatory arbitration provision are presumptively invalid under the Arbitration Act. Nor will federal courts refuse to enforce arbitration agreements, even if the agreement appears to violate state law. In Buckeye Check Cashing, Inc. v. Cardegna, 546 U.S. 440, 443 (2006), the Court insisted that a mandatory arbitration agreement in a payday loan contract be enforced, even though the stated rate of the loan exceeded Florida's usery laws. While invalidity under state law might have provided grounds for setting aside an arbitral award, it did not provide a justification for setting aside the arbitration agreement. Id.

(72.) One reflection of that view was Justice Frankfurter's dissent in Textile Workers Union v. Lincoln Mills, 353 U.S. 448, 466-68 (1957), where the Court held that Section 301 of the Labor Management Relations Act of 1947 ("LMRA") implicitly granted courts authority to compel labor arbitration. Justice Frankfurter's dissent points out that the Court was forced to rely on Section 301 of the LRMA because Section 1 of the Arbitration Act exempts labor contracts.

(73.) 9 U.S.C. [section] 1 (2006).

(74.) 500 U.S. 20 (1991).

(75.) The Court's opinion does not address this question, but it is fair to wonder whether the securities firms that make up the Exchange preferred to have the Exchange impose this requirement rather than individual firms.

(76.) Id. at 23.

(77.) Id. at 23-24.

(78.) Id. at 24.

(79.) Id. at 25 (citation omitted).

(80.) Id. at 26.

(81.) Id. (citation omitted).

(82.) See, e.g., Barrentine, 450 U.S. at 738 & n.12; see also id. at 740 (pointing out that FLSA "permits an aggrieved employee to bring his statutory wage and hour claim 'in any Federal or State court of competent jurisdiction.' No exhaustion requirement or other procedural barriers are set up, and no other forum for enforcement of statutory rights is referred to or created by the statute.").

(83.) The courts do not even seem to be receptive to cases in which Congress has acted explicitly to forbid waivers. The Federal Credit Repair Organizations Act prohibits "[a]ny waiver by any consumer of ... any right of the consumer" under the Act, including the "right to sue." 15 U.S.C. [section][section] 1679f(a)-1679f(c) (2006). Nonetheless, some courts have held that the "right to sue" can mean bringing a claim in an arbitral forum and thus have upheld pre-dispute, mandatory arbitration provisions in cases under the Act. See, e.g., Gay v. CreditInform, 511 F.3d 369 (3d Cir. 2007). Just to drive home the irony here, in 2002, car dealers successfully lobbied Congress for an anti-arbitration provision that bars car companies from forcing them to arbitrate disputes; no similar provision was included to protect consumers from mandatory arbitration. See 15 U.S.C. [section] 1226(a)(2) (2006).

(84.) Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 28 (1991) (quoting Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 637 (1985)).

(85.) Id. at 30 (quoting Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 481 (1989)).

(86.) Id. at 31 (quoting Mitsubishi, 473 U.S. at 628). One does not need contract law to trade off process in favor of expedition. Litigants are free to stipulate away these rights in court. Nothing in the federal rules would forbid litigants from agreeing, with the Court's approval, to limit discovery, to dispense with a jury trial, and to waive appeal rights, or to take other measures that would provide "simplicity, informality and expedition." The fact that litigants rarely, if ever, choose to waive these rights in court suggests that, when given a meaningful choice, plaintiffs in employment and consumer cases value process over expedience. Indeed, empirical evidence shows that in entering into business-to-business contracts, businesses prefer litigation to arbitration and rarely enter into pre-dispute mandatory arbitration agreements. Theodore Eisenberg et al., Arbitration's Summer Soldiers: An Empirical Student of Arbitration Clauses in Consumer and Nonconsumer Contracts, 41 U. MICH. J.L. REFORM 871 (2008).

(87.) 500 U.S. at 30-34.

(88.) 532 U.S. 105 (2001).

(89.) 9 U.S.C. [section] 1.

(90.) 532 U.S. at 114-15.

(91.) Id. at 115. The majority had the unenviable task of trying to explain how its ruling that Section 1's exclusion should be read narrowly was consistent with its rulings in Allied-Bruce Terminix Cos. v. Dobson, 513 U.S. 265, 270-71 (1995) and Southland Corp. v. Keating, 465 U.S. 1 (1984), holding that Section 2 of the Act, which defined the Act's coverage by referring to contracts "evidencing a transaction involving commerce," reflected Congress's intent "to exercise [its] commerce power to the full." Id. at 112 (quoting Allied-Bruce, 513 U.S. at 277). Notwithstanding the Court's longstanding rule that, as a general matter, words in statutes should be read in pail materia, the majority's reading in Circuit City results in an interpretation of the Act that construes the word commerce in diametrically opposed ways. 532 U.S. at 115.

(92.) Circuit City, 532 U.S. at 128.

(93.) Id. at 132-33.

(94.) See, e.g., Margaret L. Moses, Statutory Misconstruction." How the Supreme Court Created a Federal Arbitration Law Never Enacted by Congress, 34 FLA. ST. U. L. REV. 99 (2006); Hardin v. First Cash Fin. Servs., 465 F.3d 470 (10th Cir. 2006) (upholding binding arbitration provision even though the employee did not sign the agreement and protested being bound by it, on the theory that the employee's conduct, that is, showing up for work, constituted acceptance of the provision).

(95.) See, e.g., Maureen A. Weston, Preserving the Federal Arbitration Act By Reining in Judicial Expansion and Mandatory Use, 8 NEV. L.J. 385, 392 (2007) (describing recent legislative findings that FAA "was intended to apply to commercial disputes between parties of generally similar sophistication and bargaining power") (quoting findings of Arbitration Fairness Act of 2007, S. 1782, 110th Cong. [section] 2(1)(2)); Sternlight, supra note 8, at 1636 (arbitration previously limited to business-to-business and management-union contexts, and Congress never conceived of it for "captive consumers or employees"); David S. Schwartz, Enforcing Small Print to Protect Big Business: Employee and Consumer Rights Claims in an Age of Compelled Arbitration, 1997 WIS. L. REV. 33, 75-77 (1997) (describing legislative history).

(96.) 129 S. Ct 1456 (2009).

(97.) 29 U.S.C. [section][section] 621 et seq (2006). The Court reserved that question in Wright v. Universal Maritime Serv. Corp., 525 U.S. 70, 77-79 (1998), recognizing the obvious tension between Gardner-Denver and Gilmer, but deciding the case on other grounds.

(98.) Compare Penn Plaza, 129 S. Ct 1456, 1469 n.8 (2009) with id. at 1476-81 (Souter, J., dissenting).

(99.) The evidence overwhelmingly suggests that individuals entering into consumer and employment contracts are generally unaware that they have agreed to waive their right to sue. See, e.g., Linda J. Demaine & Deborah R. Hensler, "Volunteering" to Arbitrate Through Predispute Arbitration Clauses: The Average Consumer's Experience, 67 LAW & CONTEMP. PROBS. 55, 73-74 (2004) (citing lack of information and the difficulty of deciphering arbitration clauses, most consumers do not knowingly waive right to sue); Christine M. Reilly, Achieving Knowing and Voluntary Consent in Pre-Dispute Mandatory Arbitration Agreements at the Contracting Stage of Employment, 90 CAL. L. REV. 1203, 1225 (2002) (demonstrating that employees "do not understand the remedial and procedural ramifications of consenting to arbitration" and that "[v]ery few are aware of what they are waiving....").

(100.) This view is, of course, at odds with the Court's concern in Gardner-Denver and other cases that labor unions serve an important equalizing function because of employers' massively superior bargaining power. It is also reminiscent of Anatole France's comment that "the majestic equality of the laws, which forbid rich and poor alike to sleep under the bridges, to beg in the streets, and to steal their bread." ANATOLE FRANCE, THE RED LILY 95 (Frederic Chapman ed., Winifred Stephens trans., New York, John Lane Co. 1910) (1894). The question whether pre-dispute, mandatory arbitration agreements are consensual in any meaningful way has been the subject of extensive commentary, most arguing that in the employment and consumer context, these provisions are ones of adhesion that are not subject to bargain. See, e.g., Richard M. Alderman, Pre-Dispute Mandatory Arbitration in Consumer Contracts: A Call For Reform, 38 HOUS. L. REV. 1237, 1246-49 (2001); Sternlight, supra note 8, at 1648-53, 1671-72; Charles L. Knapp, Taking Contracts Private. The Quiet Revolution in Contract Law, 71 FORDHAM L. REV. 761, 787-89 (2002).

(101.) See Barrentine, 450 U.S. at 734-35.

(102.) U.S. Dep't of Labor, Bureau of Labor Statistics, Union Member Summary (Jan. 28, 2009), available at http://www.bls.gov/news.release/union2.nr0.htm.

(103.) 346 U.S. 427, 438 (1953).

(104.) Id. at 434-35.

(105.) Id. at 437-38.

(106.) Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, 473 U.S. 614 (1985).

(107.) Shearson/Am. Express v. McMahon, 482 U.S. 220, 233 (1987) (observing that "the mistrust of arbitration that formed the basis for the Wilko opinion in 1953 is difficult to square with the assessment of arbitration that has prevailed since that time. This is especially so in light of the intervening changes in the regulatory structure of the securities law. Even if Wilko's assumptions regarding arbitration were valid at the time Wilko was decided, most certainly they do not hold true today for arbitration procedures subject to the SEC's oversight authority.").

(108.) 473 U.S. 614, 625 (1985).

(109.) The Court rejected the argument that statutory claims were not subject to mandatory arbitration. The Court observed that "[b]y agreeing to arbitrate a statutory claim, a party does not forgo the substantive rights afforded by the statute; it only submits to their resolution in an arbitral, rather than a judicial, forum. It trades the procedures and opportunity for review of the courtroom for the simplicity, informality, and expedition of arbitration. We must assume that if Congress intended the substantive protection afforded by a given statute to include protection against waiver of the right to a judicial forum, that intention will be deducible from text or legislative history." Id. at 628. The Court's ruling sparked a vigorous dissent by Justice Stevens, joined by Justices Brennan and Marshall. Justice Stevens pointed out that "[u]ntil today all of our cases enforcing agreements to arbitrate under the Arbitration Act have involved contract claims," and that "neither the Congress that enacted the Arbitration Act in 1925, nor the many parties who have agreed to such standard clauses, could have anticipated the Court's answer" that standard arbitration agreements encompassed statutory claims. Id. at 646-47. And the dissent went on to invoke the Gardner-Denver, Barrentine and McDonald trilogy to argue at length that the procedures in arbitration are inadequate to safeguard statutory rights. Id. at 647-57.

(110.) 490 U.S. 477 (1989).

(111.) Id. at 480 (citation omitted).

(112.) Id. at 481.

(113.) Id. at 480 (citation omitted).

(114.) See, e.g., Preston v. Ferrer, 128 S. Ct. 978, 987 (2008) (holding that the Federal Arbitration Act overrides state primary jurisdiction rules); Hall Street Assocs. v. Mattel, Inc., 128 S. Ct. 1396 (2008) (holding that arbitrator, not court, must decide whether a stay should be imposed pending arbitration); Buckeye Check Cashing, Inc. v. Cardegna, 546 U.S. 440, 448-49 (2006) (holding that a claim that a contract containing a mandatory arbitration provision is invalid under state law does not justify a failure to enforce the arbitration agreement, because the clause is severable); Green Tree Fin. Corp.-Ala. v. Randolph, 531 U.S. 79 (2000) (holding that claims arising under the Truth in Lending Act and Equal Credit Opportunity Act alleging failure to disclose finance charges in a consumer contract are subject to mandatory arbitration provisions, notwithstanding the plaintiff's claims that the costs of arbitration would preclude litigation and that plaintiff sought to maintain case as a class action); Doctor's Assocs. v. Casarotto, 517 U.S. 681 (1996) (holding Montana's requirement that form consumer contracts provide clear notice that they contain a mandatory arbitration provision was preempted by the FAA).

(115.) Mastrobuono v. Shearson Lehman Hutton, 514 U.S. 52, 58 0995) (noting that "the parties to a contract may lawfully agree to limit the issues to be arbitrated by waiving any claim for punitive damages," but concluding that the agreement at issue did not waive the claim).

(116.) 531 u.s. 79 (2000).

(117.) Id. at 82.

(118.) Id.

(119.) Id. at 82-83.

(120.) 15 U.S.C. [section][section] 1601-1667f (2006).

(121.) 15 U.S.C. [section][section] 1691-1691f (2006).

(122.) 531 U.S. at 79.

(123.) 531 U.S. at 84.

(124.) Id. at 83-84.

(125.) Id. at 84.

(126.) Id. at 90-92. The Court first held that Ms. Randolph could properly appeal from the district court's ruling compelling arbitration. Green Tree had contended that an order compelling arbitration was not a final order that was subject to an appeal; the Court disagreed. Id. at 88-89.

(127.) Id. at 82-83 & n.1.

(128.) Id. at 90-91 & n.6.

(129.) Id. at 92.

(130.) Id. at 96 (Ginsburg, J., dissenting).

(131.) Id. at 97. Just so there is no mistake on this point, costs in litigation involve expenses apart from simply filing fees, and, in arbitration, paying the quite substantial fees charged by private arbitrators. In anti-trust, securities, truth-in-lending, or other forms of consumer contract litigation, the cost of hiring experts--generally subject to fee shifting in a statutory case--are not necessarily subject to fee-shifting in arbitration, which is a serious setback to plaintiffs. See In re Am. Express Merchants' Litigation, 554 F.3d 300, 316 n. 11, 318 n.12 (2d Cir. 2009). In employment cases, the relevant statutes, as well as the Civil Rights Act fee provision, 42 U.S.C. [section] 1988 (2006), provide reimbursement for expert witness expenses.

(132.) At least one circuit had already answered that question by holding that an arbitration clause in a short-term loan agreement was enforceable even though it foreclosed the consumer's ability to pursue a class action. Johnson v. W. Suburban Bank, 225 F.3d 366 (3d Cir. 2000).

(133.) 539 U.S. 444 (2003).

(134.) Id. at 447.

(135.) Id. at 449.

(136.) Id.

(137.) Id.

(138.) Id.

(139.) Id. at 450.

(140.) Id. at 450-53.

(141.) Id. at 454.

(142.) Id. at 454-55 (Stevens, J., concurring in the judgment and dissenting in part).

(143.) Id at 455.

(144.) Id. (Rehnquist, C.J., dissenting). Justice Thomas filed a separate dissent, adhering to his longstanding view that the Federal Arbitration Act should not be read to apply to state court proceedings). Id. at 460; see also Doctor's Associates, Inc. v. Casarotto, 517 U.S. 681, 689 (1996) (Thomas, J., dissenting); Allied-Bruce Terminix Cos., Inc. v. Dobson, 513 U.S. 265, 285-97 (1995) (Thomas, J., dissenting).

(145.) Indeed, in his opinion, Justice Stevens observed that the South Carolina Supreme Court "has held as a matter of state law that class-action arbitrations are permissible if not prohibited by the applicable arbitration agreement," and that nothing in the Federal Arbitration Act "precludes" the ability of parties to prohibit them. 539 U.S. at 454.

(146.) Justice Stevens predicted this development in his questioning at argument; he asked Green Tree's counsel: "Does this case have any future significance, because isn't it fairly clear that all the arbitration agreements in the future will prohibit class actions?" Transcript of Oral Argument at *55, Green Tree, No. 02-634, 2003 WL 1989562. His prediction quickly came to pass. See, e.g., Note, The Supreme Court, 2002 Term: Leading Cases: III. Federal Statutes and Regulations: C. Federal Arbitration Act, 117 HARV. L. REV. 410 (2003); Jean R. Sternlight & Elizabeth J. Jensen, Using Arbitration to Eliminate Consumer Class Actions: Efficient Business Practice or Unconscionable Abuse?, 67 LAW & CONTEMP. PROBS. 75, 85-88 (Winter/Spring 2004); see also In re Am. Express Merchants' Litigation, 554 F.3d 300, 314 (2d Cir. 2009) (noting that, following Bazzle, the New York Stock Exchange modified its rule to provide that a "claim submitted as a class action shall not be eligible for arbitration under the Rules of the Exchange"). JAMS, a major arbitration company, initially took the position that it would not enforce restrictions on class action arbitration, but reversed itself as soon as its major commercial clients, including American Express, Discover Card and Citibank, objected. See Eric Berkowitz, Is Justice Served?, L.A. TIMES MAG., Oct. 22, 2006, at 1.20. On the other hand, in the aftermath of Bazzle the American Arbitration Association altered its rules expressly to permit class action arbitrations, and it is currently handling many class proceedings. See Am. Arbitration Ass'n, Class Action Cases, http://www.adr.org/sp.asp?id=25562 (last visited Feb. 6, 2009) (listing all class arbitration proceedings AAA has administered or is currently administering).

(147.) It would probably be an understatement to say that the law is, at least currently, in disarray on the validity of blanket class action waivers. Some circuits approve them generally. See, e.g., Anderson v. Comcast, Corp., 500 F.3d 66, 72 (1st Cir. 2007); Gay v. Credit-Inform, 511 F.3d 369, 381 (3d Cir. 2007); Carter v. Countrywide Credit Indus., Inc., 362 F.3d 294, 298 (5th Cir. 2004); Johnson v. W. Suburban Bank, 225 F.3d 366, 374-75 (3d Cir. 2000). And most circuits approve them with the caveat that they will be unenforceable if they preclude the claimant from bringing his claim. See, e.g., In re Am. Express Merchants' Litigation, 554 F.3d 300 (2d Cir. 2009); Lowden v. T-Mobile USA, Inc., 512 F.3d 1213, 1219 (9th Cir. 2008); In re Cotton Yarn Antitrust Litigation, 505 F.3d 274, 285 (4th Cir. 2007); Dale v. Comcast Corp., 498 F.3d 1216, 1224 (11th Cir. 2007); Livingston v. Assocs. Fin., Inc., 339 F.3d 553, 557 (7th Cir. 2003).

(148.) See Eisenberg et al., supra note 86, at 875 n.16 (collecting authorities).

(149.) The Court stated that by "agreeing to arbitrate a statutory claim, a party does not forego the substantive rights afforded by the statute; it only submits their resolution in an arbitral, rather than a judicial, forum." Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 26 (1991) (quoting Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 628 (1985)).

(150.) See 473 U.S. at 628.

(151.) See Gilmer, 500 U.S. at 26.

(152.) See, e.g., Bingham, supra note 20. A number of scholars have also debated the accessibility of arbitration as opposed to litigation, with many suggesting that low-income employees or consumers with relatively small claims are able to pursue arbitration in cases where they could not bring claims in court. See, e.g., Christopher R. Drahozal, Arbitration Costs and Forum Accessibility. Empirical Evidence, 41 U. MICH. J.L. REFORM 813 (2008) (summarizing research). While the accessibility literature concerns entry into the process, we consider it as part of the outcome research, because if arbitration truly does serve a large number or a particular sector of potential claimants, that social value may be understood as an output or outcome of the process. See discussion infra, Part II.B.

(153.) See Weidemaier, supra note 10, at 845-56 (summarizing studies).

(154.) See, e.g., Maltby, supra note 8.

(155.) See Sternlight, supra note 8, at 1658.

(156.) Richard A. Bales, Normative Consideration of Employment Arbitration at Gilmer's Quinceanera, 81 TUL. L. REV. 331, 349 (2006).

(157.) Bingham, supra note 20, at 199. See Sherwyn et al., supra note i0, at 1564-67 (describing differences between cases in arbitration and in courts); Ware, supra note 10, at 753-57 (same).

(158.) See Ware, supra note 10, at 756-58.

(159.) See id.; see also Weidemaier, supra note 10, at 848-54.

(160.) This critique has also been leveled against studies of litigation, but the difficulty of studying litigation is of course no basis for claiming the reliability of arbitration studies.

(161.) Weidemaier, supra note 10, at 850-51.

(162.) Bales, supra note 156, at 389-90.

(163.) Id.

(164.) See generally Drahozal, supra note 152.

(165.) Id. at 828-31.

(166.) Costs to plaintiffs have been lowered by caps on arbitrators' fees, and by provisions whereby companies assume responsibility for those fees. Id. at 829-30; Eisenberg et al., supra note 86, at 876.

(167.) See generally Drahozal, supra note 152.

(168.) See generally William M. Landes & Richard A. Posner, Adjudication As a Private Good, 8 J. LEGAL STUD. 235 (1979).

(169.) Greenfield, supra note 17, at 683 (demonstrating that many arbitration decisions do not apply statutory law at all, or do so making conclusory judgments that neither interpret nor create precedent); see also Weidemaier, supra note 10, at 868 (highlighting "risk that arbitrators will 'recast grievances in ways that downplay legal issues and that focus instead on more typically managerial concerns'") (quoting Lauren B. Edelman & Mark C. Suchman, When the "Haves" Hold Court." Speculations on the Organizational Internalization of Law, 33 LAW & SOC'Y REV. 941, 967 (1999)); Edwards, supra note 17, at 297 ("Imagine, for example, the impoverished nature of civil rights that would have resulted had all race discrimination claims in the United States in the 1960s and 70s been arbitrated rather than adjudicated."). But see Drahozal, supra note 17, at 207-15 (arguing that current evidence is inconclusive as to whether arbitration stifles development of precedent).

(170.) E.g., Civil Rights Act of 1991, Pub. L. No. 102-166, [section] 105 Stat. 1071 (amending civil rights statute in response to Supreme Court rulings including employment discrimination case of Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989)); see David Luban, Settlements and the Erosion of the Public Realm, 83 GEO. L. J. 2619, 2637 (1995).

(171.) See Sternlight, supra note 8, at 1661.

(172.) See generally Bales, supra note 156; see also Sternlight, supra note 8, at 1661-1665 (summarizing "public justice" objections to mandatory arbitration).

(173.) See Reginald Alleyne, Statutory Discrimination Claims: Rights "Waived" and Lost in the Arbitration Forum, 12 HOFSTRA LAB. & EMP. L.J. 381, 429-31 (1996) (discussing importance of jury awards' public function of deterring discriminatory conduct).

(174.) Bales, supra note 156, at 342-43 (describing advantages of faster resolution).

(175.) Schwartz, supra note 95, at 115 (suggesting waiver of liability makes discrimination less expensive, thereby decreasing the incentive for employers to take steps to avoid it).

(176.) Schwartz, supra note 12, at 413, 415; Richard A. Bales, The Employment Due Process Protocol at Ten: Twenty Unresolved Issues, and a Focus on Conflicts of Interest, 21 OHIO S. J. ON DISP. RESOL. 165, 188, 191 (2005).

(177.) See St. Antoine, supra note 9. Samuel Estreicher, Saturns for Rickshaws: The Stakes in the Debate over Predispute Employment Arbitration Agreements, 16 OHIO ST. J. ON DISP. RESOL. 559, 563 (2001); Maltby, supra note 8, at 60. But see Drahozal, supra note 152, at 840-41 (arbitration is less accessible where claims are only economical if litigated as class); Sternlight, supra note 8, at 1654-55.

(178.) Estreicher, supra note 177, at 563 ("In a world without employment arbitration as an available option, we would essentially have a 'cadillac' system for the few and a 'rickshaw' system for the many."). But see Sternlight, supra note 8, at 1654-55 ("[E]ven if it were true that mandatory arbitration is the Saturn, why do companies have the right to take away some consumers' and employees' Cadillacs?").

(179.) See St. Antoine, supra note 9, at 795-96; Estreicher, supra note 177, at 563; Maltby, supra note 8.

(180.) See, e.g., Greenfield, supra note 17. Arbitration awards are considered to be more frequent but lower than litigation awards, perhaps because arbitrators regularly "split the baby" instead of making all-or-nothing determinations based on legal merits; even defenders of arbitration recognize this may not be the best system. Sherwyn et al., supra note 10, at 1573, 1578 (suggesting fairness should mean not that one side wins, nor by how much, but that "employees prevail when they were discriminated against and awards provide the proper remuneration").

(181.) David Schwartz suggests that prospective agreements to waive statutory rights are particularly troubling, because individuals' interests are more likely to be in line with those of society after a dispute arises. Schwartz, supra note 95, at 119.

(182.) Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 26 (1991) (quoting Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 628 (1985)); see, e.g., Stephen J. Ware, Default Rules through Mandatory Rules: Privatizing Law through Arbitration, 83 MINN. L. REV. 703, 712-18 (1999).

(183.) Id. at 704.

(184.) Id. at 710.

(185.) Id. at 732 n.131.

(186.) Circuit City Stores, Inc., v. Adams, 532 U.S. 105 (2001).

(187.) Metro E. Ctr. for Conditioning & Health v. Qwest Commc'ns Int'l, Inc., 294 F.3d 924, 928 (7th Cir. 2002).

(188.) Id. at 928-29 (summarizing cases).

(189.) Id. at 929 ("One aspect of personal liberty is the entitlement to exchange statutory rights for something valued more highly.")

(190.) Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, 473 U.S. 614, 635-37 (1985).

(191.) Id. at 630-33

(192.) Id. at 637 n.19.

(193.) Id. at 629-31.

(194.) Id. at 628.

(195.) Even if arbitration has advantages, this does not justify the imposition of prospective mandatory arbitration provisions. For a discussion of why voluntary arbitration agreements after disputes arise would be preferable, see Sternlight, supra note 8, at 1654-55; Schwartz, supra note 95, at 116-119. While some might suggest that parties would never agree to arbitrate after-the-fact, we agree with Sternlight's view that if this is true, this only supports the notion that no one with full information and bargaining power would voluntarily waive access to a judicial forum. Jean R. Sternlight, In Defense of Mandatory Binding Arbitration (If Imposed on the Company), 8 NEV. L.J. 82, 83-84 (2007).

(196.) See Schwartz, supra note 95, at 114-16.

(197.) Id. at 119.

(198.) Metro E. Ctr. for Conditioning & Health v. Qwest Commc'ns Int'l, Inc., 294 F.3d 924, 929 (7th Cir. 2002).

(199.) It is after a concrete injury has occurred that the rights holder is more likely to seek advice as to the value of the rights being waived; free consultations or contingency fee arrangements make advice more available after the fact, and the presentation of the case in all its concrete details rather than a hypothetical problem makes the advice more accurate. See Schwartz, supra note 95, at 114-16.

(200.) Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, 473 U.S. 614, 637 n.191 (1985); Jeffrey W. Stempel, Pitfalls of Public Policy: The Case of Arbitration Agreements, 22 ST. MARY'S L.J. 259, 284-85 (arguing for abolition of public policy exception).

(201.) See Paul D. Carrington & Paul H. Haagen, Contract and Jurisdiction, 1996 SUP. CT. REV. 331, 344-45 (1996).

(202.) See, e.g., Sternlight, supra note 195; Schwartz, supra note 95, at 36; Katherine Van Wezel Stone, Mandatory Arbitration of Individual Employment Rights: The Yellow Dog Contract of the 1990s, 73 DENVER U. L. REV. 1017, 1050 (1996).

(203.) See, e.g., Weston, supra note 95, at 392 (describing recent legislative findings that FAA "was intended to apply to commercial disputes between parties of generally similar sophistication and bargaining power") (quoting findings of Arbitration Fairness Act of 2007, S. 1782, 110th Cong. 1(1)); Sternlight, supra note 8, at 1636 (2005) (arbitration previously limited to business-to-business and management-union contexts, and Congress never conceived of it for "captive consumers or employees"); Schwartz, supra note 95, at 75-77 (describing legislative history).

(204.) Stewart E. Sterk, Enforceability of Agreements to Arbitrate: An Examination of the Public Policy Defense, 2 CARDOZO L. REV. 481, 543 (1981); see St. Antoine, supra note 9, at 788-89 (acknowledging that freedom of contract bends to accommodate public policy goals, especially in fields with "customarily unequal participants"); Carrington & Haagen, supra note 201, at 334-39 (highlighting freedom of contract philosophy underlying pro-arbitration jurisprudence).

(205.) See Stempel, supra note 200, at 351 ("[T]here is a 'social' or 'public' interest in the average commercial dispute just as in a securities claim or a civil rights claim."). In support of his asserting that commercial contract claims reflect social values as fundamental as civil rights, Stempel highlights that the U.S. Constitution recognized the importance of property rights. Id. at n.431. This logic, too, leads us back to Lochner.

(206.) Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, 473 U.S. 614, 628 (1985); see also Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 26 (1991).

(207.) There are limits on the extent to which a state can enforce procedural rules that might impair the enforcement of federal rights. See generally Howlett v. Rose, 496 U.S. 356 (1990); Felder v. Casey, 487 U.S. 131 (1988).

(208.) See also Ware, supra note 182, at 718 (arguing that directing arbitrators to apply the "law that a court would apply" leaves open the question of what body of law that is, creating unanswerable conflicts of laws questions for a forum that lacks a common law on conflicts of laws).

(209.) Metro East, 294 F.3d at 929. Judge Easterbrook amplified his point in a later decision, which left to the arbitrator the question whether an arbitration agreement could waive a plaintiff's right to attorney's fees under the Fair Debt Collection Practices Act. Judge Easterbrook said: "no general doctrine of federal law prevents people from waiving statutory rights (whether substantive or procedural) in exchange for other things they value more, such as lower prices or reduced disputation. See Metro East, 294 F.3d at 928-29 (collecting authority). Whether any particular federal statute overrides the parties' autonomy and makes a given entitlement nonwaivable is a question for the arbitrator." Carbajal v. H&R Block Tax Servs., 372 F.3d 903, 906-07 (7th Cir. 2004) (emphasis added).

(210.) It could be said that person values the "something" more highly, even if society as a whole might disapprove, and Judge Easterbrook would defend that person's right to decide for herself, even if she is foolish. The obvious problem is that empirical evidence suggests people enter into "agreements" without an informed, voluntary evaluation process--the process by which people determine the values they assign. See, e.g., Demain & Hensler, supra note 99, at 73-74; Reilly, supra note 99, at 1225. Moreover, as we have endeavored to emphasize in this Article, the costs of these "choices" are not just individual but collective.

(211.) See id.

Kathryn A. Sabbeth & David C. Vladeck *

* Kathryn A. Sabbeth, is an Assistant Professor of Law at the University of North Carolina. David C. Vladeck is a Professor of Law at Georgetown University Law Center and directs the Institute for Public Representation. Prior to joining the Georgetown faculty, Professor Vladeck was an attorney with Public Citizen Litigation Group, a public interest law firm, where he participated in litigating some of the cases discussed in this Article, urging courts not to find that statutory rights could be subject to pre-dispute mandatory arbitration agreements.