Arbitration Nation: Data from Four Providers

Arbitration Nation: Data from Four Providers

Forced arbitration has long been controversial. In the 1980s, the Supreme Court expanded the Federal Arbitration Act (FAA), sparking debate about whether private dispute resolution was an elegant alternative to litigation or a rigged system that favors repeat-playing corporations. Recently, these issues have resurfaced, as the Court has decided a rash of cases mandating that lower courts enforce cby lass arbitration waivers in almost all circumstances. Critics argue that abolishing the class action insulates companies from wrongdoing, but businesses have predicted that pro se plaintiffs will flood the arbitral forum with their own low-value claims. The Obama administration responded to the Court’s FAA jurisprudence by regulating arbitration clauses in the employment, financial services, and health care fields. However, after the balance of power shifted in 2017, Republicans have repealed many of these rules.

Despite this policy-making frenzy, we know little about what happens inside the confidential world of arbitration. This Article sharpens our understanding of this pervasive and polarizing institution by reporting the results of an empirical study of 40,775 cases filed in four major arbitration providers between 2010 and 2016. It highlights three main points. First, a wave of reforms has made arbitration surprisingly affordable for consumers, employees, and medical patients.  Indeed, in leading arbitration providers such as the American Arbitration Association, the Judicial Arbitration and Mediation Services, and the Kaiser Office of the Independent   Administrator,  a majority of plaintiffs pay no arbitration fees. Second, enterprising plaintiffs’ lawyers—not pro se litigants—have taken advantage of arbitration’s open doors. In fact, some attorneys have filed class action-style cases, bringing dozens or even hundreds of related arbitrations against the same company. Third, although arbitration does indeed favor repeat-playing businesses, that is just half of the repeat-player story. Repeat-playing plaintiffs’ law firms also fare well. In fact, in a variety of settings, no variable affects win rates as dramatically as whether a plaintiff hires attorneys with arbitration experience.

The Article then uses these findings to propose reform. For decades, state lawmakers have tried to protect substantive rights by exempting claims from arbitration. Yet because the FAA prohibits state law from discriminating against arbitration, these efforts have failed. Accordingly, this Article urges policy-makers to reverse course and create incentives for plaintiffs’ lawyers to arbitrate. Specifically, jurisdictions should create an “arbitration multiplier”: a bounty for winning a case in arbitration. By encouraging skilled plaintiffs’ lawyers to capitalize on arbitration’s accessibility, this approach would counteract the corporate repeat-player advantage. In addition, because the multiplier would actually encourage arbitration, it would not be preempted.

 

 

Introduction

Epic Systems Corp. v. Lewis, the Supreme Court’s latest blockbuster opinion about the Federal Arbitration Act (FAA) and class actions,[1] began in the mid-2000s, when several groups of Ernst & Young employees sued the massive accounting firm for failing to pay them overtime.[2] However, the plaintiffs had signed employment agreements that contained arbitration clauses and class arbitration waivers.[3] Thus, although each employee sought between $2,000 and $29,000 in damages[4]—far less than the $200,000 necessary to prove their claims[5]—Ernst & Young sought to force them to arbitrate on an individual basis.[6]

What happened next illustrates the doctrinal chaos that has plagued federal arbitration law over the last decade. In March 2011, a federal district court in New York struck down Ernst & Young’s class arbitration waiver.[7] This judge noted that the cost of arbitrating one of these lawsuits dwarfed any one plaintiff’s possible recovery.[8] Thus, because “[o]nly ‘a lunatic or a fanatic’” would pursue a “low-value, high-cost claim on an individual basis,” the court reasoned that the class arbitration waiver immunized Ernst & Young from liability.[9] But then the legal landscape experienced a tectonic shift. In April 2011, the Court decided AT&T Mobility LLC v. Concepcion.[10] In a 5–4 opinion, the Justices held that the FAA preempted a state rule that deemed most class arbitration waivers in consumer contracts to be unconscionable.[11] In January 2012, the National Labor Relations Board (NLRB) tried to limit Concepcion to the consumer setting by finding that class arbitration waivers in employment agreements violate the National Labor Relations Act (NLRA).[12] Yet courts splintered over whether the NLRB’s ruling was correct.[13] For instance, the Second Circuit and a federal judge in California rejected the NLRB’s decision and enforced Ernst & Young’s class arbitration waiver,[14] but the Ninth Circuit invalidated the provision.[15] The Court granted certiorari in one of the Ernst & Young cases and consolidated it with two other matters into Epic Systems.[16]

In May 2018, with a 5–4 decision written by Justice Gorsuch, the Court reaffirmed the primacy of the FAA.[17] According to Justice Gorsuch, the statute’s command is simple: courts must “enforce arbitration agreements according to their terms—including terms providing for individualized proceedings.”[18] Although Justice Gorsuch acknowledged that “[t]he policy may be debatable,” he nevertheless concluded that “the law is clear.”[19]

Calling the normative stakes in Epic Systems “debatable” is an understatement. The Court’s docket was inundated with amicus briefs on both sides, and when these submissions discussed arbitration, they seemed to describe two different institutions. According to the defendants and their amici, extrajudicial dispute resolution benefits both parties. For example, the US Chamber of Commerce asserted that arbitration empowers regular people to act pro se and vindicate their own low-value causes of action:

For many employees with individualized complaints against their employer—whether involving wrongful discharge, unlawful discrimination, or other grievances—arbitration is the only viable means of recovery, because the employees’ claims are too small to justify the expense of court litigation or to attract a contingency-fee lawyer. . . . A small claim is more viable in arbitration because costs in arbitration are lower—and because in an arbitral forum, “it is feasible for employees to represent themselves . . . .”[20]

As the Chamber put it, “without individual arbitration, most of those claims could not be pursued at all.”[21] Likewise, the HR Policy Association contended that plaintiffs fare better in arbitration because its relaxed procedural and evidentiary rules prevent companies from landing a knockout blow via summary judgment.[22]

Conversely, the employees and their allies painted a darker picture. Some condemned arbitration as “really expensive” and “not particularly efficient.”[23] Others asserted that unrepresented individuals are unlikely to prevail against repeat-playing companies,[24] and that few plaintiffs “subject to arbitration agreements bring small—or indeed, any—claims.”[25]

These dueling views are part of an ongoing referendum on forced arbitration. Since 2010, the Court has decided no fewer than thirteen opinions interpreting the FAA.[26] In addition to making class arbitration waivers nearly bulletproof,[27] the Justices have repeatedly held that the FAA preempts state efforts to regulate arbitration. For example, West Virginia tried to exempt wrongful death lawsuits from the FAA,[28] and Kentucky limited the ability of agents acting under a power of attorney to contract away a principal’s right to access the courts.[29] These doctrines reflect concern that arbitration is not appropriate for certain claims, such as tort suits against nursing homes.[30] Nevertheless, for the Court, these state policy judgments are irrelevant, because the FAA overrides “any state rule discriminating on its face against arbitration.”[31]

The Court’s arbitration jurisprudence has divided commentators and lawmakers. In 2015, the New York Times ran three front-page stories on the expansion of the FAA.[32] Scores of law review articles have accused the Justices of illicitly “privatiz[ing] . . . dispute resolution”[33] and sounding the “death knell” for consumer and employment class actions.[34] The Obama administration restricted arbitration provisions in contracts between students and schools,[35] government contractors and their workers,[36] brokers and investors,[37] and nursing homes and residents.[38] In an even higher-profile maneuver, the Consumer Financial Protection Bureau (CFPB) banned class arbitration waivers in the financial services sector.[39] Yet after the 2016 elections, the pendulum swung back. President Trump repealed some anti-arbitration rules by Executive Order.[40] Republican-led agencies abandoned restrictions issued by their predecessors.[41] Finally, in October 2017, the Senate repealed the CFPB’s prohibition on class arbitration waivers.[42]

Despite this flurry of activity, we know little about what actually happens in the arbitral forum. After all, arbitration is private dispute resolution: a “black box . . . where the proceedings are confidential and non-precedential . . . .”[43] To be sure, some researchers have gained access to arbitration files, and used them to write useful empirical studies.[44] But the overwhelming majority are from the 1990s and the 2000s, and thus have not kept pace with the law’s rapid development.[45] Moreover, most of these studies simply tally up win rates, rather than use sophisticated econometric techniques that can identify how various factors affect arbitrators’ rulings.[46]

Recently, we published a pair of articles that tried to take our understanding of forced arbitration to the next level.[47] In the first, After the Revolution: An Empirical Study of Consumer Arbitration, we analyzed 4,839 cases filed by consumers in the American Arbitration Association (AAA) between July 2009 and December 2013.[48] Then, in Employment Arbitration After the Revolution—a shorter, invited symposium submission—we considered 5,883 employment arbitrations conducted by the AAA over the same period.[49] The centerpiece of both papers was our regression analyses. Even when we controlled for other variables, we found that when a plaintiff faced a company that had significant arbitration experience (an “extreme” repeat player), the plaintiff’s odds of winning were drastically reduced.[50]

However, like previous work in this genre, our articles only provided a snapshot of the private tribunal. Indeed, every paper to date (including ours) has focused on a single arbitration provider (almost always the AAA).[51] As a result, these studies shed no light on the thousands of cases in other institutions. Also, because the AAA refuses to handle certain matters, such as medical malpractice,[52] we know almost nothing about entire sectors within arbitration’s sprawling domain.

This Article fills these gaps by combining our own data collection with information provided to us by one of the journalists who wrote the 2015 New York Times articles about the FAA.[53] We analyze 40,775 consumer, employment, and medical malpractice arbitrations filed between 2010 and 2016 in four major arbitration administrators: the AAA, Judicial Arbitration and Mediation Services (JAMS), ADR Services, Inc., and the Kaiser Health Care Office of Independent Administration (Kaiser).

We highlight three descriptive points. First, arbitration has the capacity to facilitate access to justice. Cases move quickly through the system, and corporations pick up most of the tab. Second, arbitration is not currently living up to this potential. Although businesses are correct that more individuals are arbitrating after Concepcion, this uptick has been modest. Moreover, companies are wrong about who is bringing those claims. Plaintiffs’ lawyers—not self-represented consumers, employees, or medical patients—have been taking advantage of arbitration’s speed and relative affordability.[54] In fact, some attorneys have tried to create a simulacrum of the class action by initiating dozens or even hundreds of two-party arbitrations against the same defendant.[55] Third, concern that arbitration favors repeat-playing corporations is well founded. Indeed, businesses that arbitrate often in an institution perform particularly well within that institution. Nevertheless, this is just one-half of the repeat-player story. Arbitration favors repeat players on both sides. In a variety of different settings, serially arbitrating plaintiffs’ law firms also fare particularly well.

The Article then uses these findings to suggest a novel possibility for state lawmakers who are concerned that arbitration dilutes substantive rights. As noted, the FAA mercilessly preempts any state rule that exhibits distrust of arbitration.[56] Accordingly, rather than trying to exempt claims from the extrajudicial forum, the Article suggests that state lawmakers create rewards for plaintiffs’ lawyers to arbitrate. Specifically, jurisdictions should create a statutory “arbitration multiplier”: an extra bounty for winning a case in arbitration. This approach addresses the root of the arbitration drought, which appears to be a lack of incentives for lawyers to take these cases, rather than a lack of access to arbitration. Moreover, because having an experienced lawyer greatly increases the probability of a plaintiff win, it would help level the playing field between individuals and repeat-playing corporations. Finally, unlike the legions of failed state efforts to restrict arbitration, our proposal actually encourages private dispute resolution and thus exists in harmony with the FAA.

The Article contains three Parts. Part I provides background. It first traces the rise of the FAA and the emergence of arbitration providers such as the AAA, JAMS, ADR Services, and Kaiser. It then reviews the empirical literature on arbitration. It explains why much of this research, while valuable, has little bearing on contemporary issues. Part II first describes our methodology and then summarizes our results, focusing on costs, case length, filing levels, and win rates. Finally, Part III uses our data to explain why states should reward plaintiffs’ lawyers who prevail in arbitration.

I. Arbitration Nation

This Section lays the groundwork for our study. It starts by explaining how arbitration evolved from a disfavored mode of dispute resolution to “a phenomenon that pervade[s] virtually every corner of the daily economy.”[57] It then surveys previous empirical scholarship about arbitration.

A. The Forced Arbitration Controversy

Congress enacted the FAA in 1925.[58] Under the ancient ouster and revocability doctrines, courts refused to award specific performance of an agreement to arbitrate.[59] Section 2 of the FAA eradicated these anti-arbitration principles by making arbitration clauses “valid, irrevocable, and enforceable, save upon such grounds as exist at law or in equity for the revocation of any contract.”[60] By doing so, it put arbitration clauses “upon the same footing as other contracts.”[61] This was an uncontroversial objective. Indeed, as the Senate Report on the FAA declared, creating the infrastructure for a fast and economical alternative to the judicial system “appeal[ed] to big business and little business alike, to corporate interests as well as to individuals.”[62]

A year later, some of the lobbyists behind the statute formed the AAA.[63] This organization sold a particular brand of dispute resolution. Historically, arbitration had been collaborative and informal: merchants submitted a disagreement to an expert in their field, who settled it by applying industry customs.[64] Conversely, the AAA sought to create a stripped-down version of the court system.[65] Lawyers were welcome, and arbitrators often held law degrees.[66]

Although the AAA made arbitration more like litigation, there was no doubt that the two spheres were distinct. One of the biggest differences was that the substantive law in arbitration was supposedly a bit like gravity on the moon: it applied, but in diminished form.[67] For example, a 1948 Note in the Harvard Law Review examined a handful of awards from the AAA and concluded that arbitrators considered compelling but legally irrelevant factors, such as a party’s motives, when calculating damages.[68] Likewise, Soia Mentschikoff discovered in a poll of AAA arbitrators that “almost 90 per cent believed that they were free to ignore these rules whenever they thought that more just decisions would be reached by so doing.”[69] In 1953, in a decision that the US Supreme Court eventually affirmed, a New York district court cited the fact that arbitrators need not “refer[] to technical rules of law and procedure” and exempted alleged violations of the federal securities laws from the FAA.[70]

Nevertheless, as the twentieth century marched on, arbitration blossomed into an established mode of conflict resolution. New providers emerged, challenging the AAA. In 1971, Kaiser Permanente, one of the nation’s largest health maintenance organizations (HMOs), mandated that patients resolve medical malpractice claims within Kaiser’s own proprietary arbitration system.[71] Eight years later, former trial judge H. Warren Knight created JAMS, a for-profit enterprise.[72] Doubling down on the AAA’s original business model, JAMS marketed itself as the forum for complicated, high-stakes cases.[73] It recruited arbitrators who, like its founder, had served on the bench.[74]

Then, in the 1980s, the Court decided a rash of cases that pushed arbitration into the mainstream. Led by Chief Justice Warren Burger, who saw arbitration as a palliative for the so-called “litigation explosion,”[75] the Court backtracked from its view that arbitration was subordinate to the judicial system. For one, the Court required complex federal statutory claims—such as antitrust and employment discrimination—to be arbitrated.[76] The Justices did so because they believed that the choice between litigation and arbitration did not affect the ultimate result of a case:

By 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.[77]

In addition, in Southland Corp. v. Keating the Court announced that section 2 of the FAA applies in state court and preempts conflicting state law.[78] Under this muscular reading of the statute, Montana could not require arbitration clauses to be conspicuous,[79] California could not exempt wage disputes from arbitration,[80] and Alabama could not refuse to specifically enforce agreements to arbitrate.[81] As the Court explained, states could neither “singl[e] out arbitration provisions for suspect status”[82] nor regulate them under “laws applicable only to arbitration provisions.”[83]

Arbitration clauses spread throughout the economy. In rising numbers, employers, banks, and retailers began to opt out of the court system.[84] To accommodate the burgeoning market for dispute resolution, more arbitration providers opened their doors, including ADR Services, Inc.,[85] Alternative Resolution Centers (ARC),[86] and the National Arbitral Forum (NAF).[87]

This partial privatization of the civil justice system generated heated debate. Businesses and right-leaning scholars defended arbitration as quick and cheap.[88] This view resonated with certain members of the Court, who saw arbitration as an outlet for self-represented plaintiffs with small claims. For example, as Justice Breyer remarked in Allied-Bruce Terminix Companies, Inc. v. Dobson, the sleek and simple process was “helpful to individuals, say, complaining about a product, who need a less expensive alternative to litigation.”[89]

But a parade of scholars questioned whether plaintiffs received a fair shake outside of the court system.[90] In particular, they argued that arbitration was slanted toward the powerful corporations that routinely used it.[91] After all, because arbitrators bill by the hour, they have a pecuniary incentive to rule in favor of the frequently arbitrating corporations that might select them again in the future.[92] Similarly, some judges saw arbitration as a lawless vacuum that should not be entrusted with important rights: “an inferior system of justice, structured without due process, rules of evidence, accountability of judgment and rules of law.”[93]

In the 1990s, several arbitration providers responded to these critiques by self-regulating. As concern grew about the propriety of arbitrating medical malpractice claims, Kaiser commissioned a Blue Ribbon Panel to revamp its arbitration system.[94] In particular, the HMO agreed to pay the arbitrator’s fee as long as the plaintiff waived any objection that this might bias the arbitrator.[95] The AAA went even further, announcing that it would not administer any health care matter stemming from a pre-dispute arbitration agreement.[96] Finally, both the AAA and JAMS altered their approach to forced arbitration. Each institution announced that it would refuse to handle a case under any consumer or employment contract that did not provide for a fair alternative dispute resolution (ADR) process.[97] In addition, the AAA and JAMS adopted internal rules that tried to level the playing field by capping fees, authorizing broad discovery, and giving both parties a voice in arbitrator selection.[98]

Despite these progressive changes, a new flashpoint emerged in the arbitration war. Companies began using class arbitration waivers, requiring plaintiffs to arbitrate on an individual basis (rather than as part of a class).[99] In Discover Bank v. Superior Court, the California Supreme Court held that a class arbitration waiver was unconscionable where a credit card company had allegedly cheated its customers out of a small amount of money.[100] The court explained that because no single plaintiff would spend the resources necessary to pursue a low-value claim, the class arbitration waiver functioned as a “get out of jail free card” for corporate liability.[101] Several other courts applied this “Discover Bank rule” to strike down class arbitration waivers in similar circumstances.[102]

However, corporate lawyers had an ace up their sleeve. To address the objection that class arbitration waivers deterred small-dollar complaints, some companies began to offer rewards for plaintiffs to arbitrate those claims on an individual basis. AT&T, Sallie Mae, and Verizon promised to pay a bounty of between $5,000 and $10,000 and up to double attorneys’ fees to any consumer who won an award that exceeded the company’s settlement offer.[103] Although these defendants touted their class arbitration waivers as “pro-consumer,”[104] courts continued to deem them to be unconscionable.[105]

In AT&T Mobility LLC v. Concepcion, the Court granted certiorari in one of these cases and used it to transform the intersection of the FAA and class actions.[106] A group of consumers alleged that AT&T had violated a state consumer protection statute by overcharging them by $30.22.[107] A district court and the Ninth Circuit applied the Discover Bank rule and held that AT&T’s class arbitration waiver was unconscionable.[108] Nevertheless, Justice Scalia held that the Discover Bank rule flouted the “purposes and objectives” of the FAA and was thus preempted.[109] Justice Scalia opined that the statute’s overarching objective is “to facilitate streamlined proceedings.”[110] According to Justice Scalia, the Discover Bank rule undermined this goal by insisting on class arbitration, which is slower and more formal than conventional two-party arbitration.[111] Ultimately, he concluded that “[s]tates cannot require a procedure that is inconsistent with the FAA, even if it is desirable for unrelated reasons.”[112]

Since Concepcion, federal arbitration law has experienced a series of wild swings. The Court’s holding spread from the consumer milieu to high-stakes claims under employment law[113] and federal statutes.[114] But after a torrent of criticism by scholars and journalists[115] the Obama administration stepped forward. In 2012, the NLRB held that class arbitration waivers in employment contracts are impermissible waivers of the right to engage in “concerted activities” under the NLRA.[116] In 2014, President Obama signed Executive Order 13673 (EO 13673), which barred federal agencies from entering into agreements for more than $1 million in goods or services with firms that require their workers to arbitrate claims for sexual assault, harassment, or discrimination.[117] In 2016, the Department of Health and Human Services (HHS) announced that it would deny vital Medicaid and Medicare funds to nursing homes that use arbitration clauses in their admissions agreements.[118] Likewise, the Departments of Education (DoE) and Labor (DoL) restricted the use of arbitration by educational institutions[119] and investment companies.[120] Finally, in a much-publicized decision, the Consumer Financial Protection Bureau (CFPB) banned class arbitration waivers in the financial services sector.[121]

But after the inauguration of President Trump, the federal government reversed course. In March 2017, President Trump repealed EO 13673 with his own Executive Order.[122] In June, the HHS reconsidered its opposition to arbitration clauses in nursing home admission contracts, calling them “advantageous to both providers and beneficiaries because they allow for the expeditious resolution of claims without the costs and expense of litigation.”[123] Shortly afterward, the DoE and DoL also signaled that they would jettison the anti-arbitration rules they had inherited.[124] Finally in October, the Senate voted 51 to 50—with Vice President Pence casting the decisive ballot—to repeal the CFPB’s prohibition on class arbitration waivers.[125]

In sum, arbitration has evolved from a norm-driven method of settling disputes to a parallel justice system that is caught in a political tug-of-war. As we discuss next, one reason the field is so deeply unsettled is that there is little hard evidence about what happens behind the black curtain of the arbitral forum.

 B. Empirical Studies

This Subsection surveys prior empirical scholarship on arbitration. It explains that this literature, while valuable, is also limited.

In the late 1990s, Lisa Bingham published two pioneering studies of arbitration results.[126] First, Bingham examined 270 AAA employment awards from 1993 and 1994.[127] Bingham discovered that employees “won” (which she defined as obtaining an award of $1 or more) 71% of the time against “one-shotters” (employers that arbitrated a single time in her data).[128] Conversely, she found that employees prevailed merely 16% of the time against “repeat players” (firms that appeared twice or more).[129] Then, in a companion piece, she analyzed 203 AAA employment awards decided between 1993 and 1995.[130] She unearthed a similar dichotomy: employees prevailed in 67% of disputes against one-shotters, but in only 23% of arbitrations against repeat players.[131] In her second paper, Bingham also examined “repeat pairings” (matters in which an employer appeared at least twice before the same arbitrator).[132] She determined that employees won five of the twenty repeat pairings (25%).[133] In contrast, when there was no repeat pairing, employees won eighty-six out of 155 times (55%).[134]

Bingham mentioned several plausible explanations for this “repeat player effect.” For one, she explained that repeat-playing employers might face different kinds of lawsuits than one-shotters.[135] In particular, she noted that claims against repeat players often arose from personnel manuals (which cover low-level workers and tend to favor the employer), rather than individually negotiated agreements (which apply to executives and contain more concessions on the part of the company).[136] Thus, she observed that the difference in win rates could stem from the fact that repeat players enjoyed more contractual rights than other employers.[137] Similarly, she hypothesized that repeat players might have better lawyers or simply “learn to settle the cases they otherwise would lose.”[138]

Bingham’s work was a lightning rod. In 1997, the EEOC cited her papers for the proposition that “the more frequent a user of arbitration an employer is, the better the employer fares in arbitration.”[139] Similarly, in 2000, the California Supreme Court relied on her analysis to declare that there is a bias in favor of “‘repeat player[s]’ in the arbitration system.”[140] Nevertheless, other commentators criticized her for using a small dataset[141] and for “not control[ling] for some rather obvious variables that may explain positive employer outcomes.”[142]

A decade after Bingham’s articles, the Searle Civil Justice Institute at Northwestern University published the first in-depth look at consumer arbitration (the Searle Report).[143] The Searle Report examined 301 AAA awards from 2007,[144] 240 of which involved consumer plaintiffs.[145] The authors determined that in 128 such cases (53%), consumers “won some relief.”[146] They also discovered that plaintiffs with lawyers won 60% of the time, whereas pro se plaintiffs were victorious in just 45% of disputes.[147]

The Searle Report rejected the idea that arbitration is unduly hospitable to repeat-playing defendants. It offered two perspectives on the issue. First, it defined “repeat player” as Bingham did: as companies that appeared twice or more in the data.[148] The Searle Report labeled these “repeat(1) firms.”[149] It concluded that the consumer win rate was 52% against repeat(1) companies and 55% against one-shotters—a difference that was not statistically significant.[150] Second, the Searle Report evaluated “repeat(2) firms”: businesses that had informed the AAA of how to serve them with demands for arbitration (which implied that they had already appeared in a case administered by the organization).[151] The Searle Report conceded that plaintiffs prevailed 43% of the time against repeat(2) corporations and 56% of the time against other defendants.[152] Yet the Searle Report did not believe that the relative success of repeat(2) businesses was meaningful. Instead, it posited that repeat(2) firms won more because they “settle[d] meritorious claims and arbitrate[d] only weaker claims,” whereas one-shotters could not make these distinctions and thus “arbitrate[d] all claims, even meritorious ones.”[153] To support this claim, the Searle Report noted that although nonrepeat(2) firms settled 226 of 414 disputes (55%), repeat(2) firms resolved 133 of their 187 cases (71%) before the award stage.[154] Thus, the Searle Report concluded that any repeat-player effect was likely “due to case screening.”[155]

However, in 2011, Alexander J.S. Colvin used econometric analysis to reach a different conclusion in the employment context.[156] Colvin analyzed 1,213 AAA employment awards filed between January 1, 2003 and December 31, 2007.[157] First, he found that the employee win rate was 32% against one-shotters but just 17% against repeat players.[158] Second, he calculated that employees prevailed 23% of the time when there was no repeat pairing, but 12% of the time when there was.[159] Third, he performed a logit regression analysis, using employee win rates as a dependent variable, and (1) repeat employer, (2) repeat-employer-arbitrator-pairing, and (3) pro se status as independent variables.[160] He discovered that the mere fact that an employer was a repeat player reduced the likelihood of an employee victory by about 49% (p < 0.01).[161] Likewise, repeat pairings lowered the odds of an employee win by 40% (p < 0.05).[162]

Four years later, Colvin and Mark Gough revisited the AAA employment files.[163] Colvin and Gough examined 10,355 cases and 2,802 awards from between January 1, 2003 and December 31, 2013.[164] Rather than defining “repeat player” dichotomously, they used a continuous variable that increased each time a company appeared in their database.[165] Similarly, they abandoned the binary view of “repeat pairing” and instead coded cases to reflect each additional meeting between a firm and an arbitrator.[166] Also, they added a battery of new independent variables, including (1) the arbitrator’s gender, judicial experience, and membership in the National Academy of Arbitrators[167] and (2) whether a case was filed in either the reportedly pro-employee jurisdiction of California or the allegedly pro-employer state of Texas.[168] Then they ran one regression analysis using employee win as the dependent variable and another using the fact that the AAA had marked a case as “settled” as the dependent variable.[169]

Colvin and Gough found several points of interest. First, the odds of an employee victory declined by 45.6% when she was pro se, compared to when she was represented (p < 0.01).[170] In addition, for every additional case in which an employer was involved, the odds of an employee win fell by 0.3% (p < 0.001).[171] Likewise, with each repeat pairing, the odds that the employee would prevail decreased by 6.2% (p < 0.05).[172] Finally, although pro se plaintiffs were less likely to settle than their counterparts by a statistically significant margin, there was “no evidence that . . . repeat players are more or less likely than one-shot employers to settle cases prior to the final adjudicatory stage.”[173] In turn, this last point suggested that—contrary to the Searle Report—the divergence in win rates “between one-shot and repeat players is not simply a function of differences in settlement behaviors in the two groups.”[174]

Also in 2015, the CFPB released a study of 1,847 AAA financial services arbitrations filed between January 1, 2010 and December 31, 2012.[175] The CFPB observed that many of these cases involved both claims and counterclaims: companies sought to collect debts from consumers, but consumers “also brought affirmative claims against companies.”[176] The CFPB then focused on 341 awarded cases filed in 2010 and 2011.[177] It noted that these disputes fell into two camps: those that featured a challenge to an underlying debt and those that did not.[178] In debt-related cases, arbitrators awarded relief to seven of the sixty-six consumers (11%).[179] In other matters, twenty-five of ninety-two consumers were successful (27%).[180]

Finally, in 2015 and 2016, we published two papers.[181] For starters, in Consumer Arbitration, we surveyed 4,839 matters filed by consumers in the AAA between July 2009 and December 2013.[182] Then, in Employment Arbitration, we analyzed 5,883 disputes brought by employees in the AAA during the same period.[183] Both pieces began by investigating Concepcion’s impact on the volume of claims within the arbitral forum. In the consumer setting, we discovered that some plaintiffs’ lawyers had compensated for the abolition of the class action by initiating scores of two-party arbitrations on the same day against the same defendant.[184] However, we did not observe a similar phenomenon in employment cases.[185]

We then refined the tools that previous researchers had used to assess whether arbitration favors repeat players. First, we collected data not just on repeat-playing corporations, but also on repeat-playing plaintiffs’ lawyers.[186] Second, like Colvin and Gough (whose paper had not yet been published), we abandoned the binary approach of classifying participants in arbitration either as “repeat players” or “one-shotters.”[187] Instead, we sorted plaintiffs’ attorneys and companies into tiers based on how many times they appeared in the AAA disclosures.[188] These finer-grained categories included one-shotters, low-level repeaters, mid-level repeaters, high-level repeaters, and super repeaters.[189] We then conducted logit regression analyses to determine how these factors and other independent variables affected individual win rates (with a “win” defined as an award of $1 or more).[190] We reached the same basic conclusion in both consumer and employment cases: the mere fact that an individual faced a high-level or super repeater dramatically reduced the odds that the individual would prevail.[191] Specifically, we found that the likelihood of a consumer victory fell by (1) 79% against high-level repeaters (p < 0.001) and (2) 94% against super repeaters when compared to the reference group of one-shot businesses (p < 0.001).[192] In the same vein, we determined that the probability of an employee win declined by about 58% when the employee sued a either a high-level or super repeater, as opposed to a one-shotter (p < 0.05).[193]

 

Year Author Sample Key Findings
1997 Bingham 270 AAA employment arbitrations from 1993 and 1994 ·   Employees won 71% of time against one-shotters but only 16% of time against repeat players.
1998 Bingham 203 AAA employment awards from between 1993 and 1995 ·   Employees won 25% in repeat arbitrator-defendant pairings, but 55% when the arbitrator and the defendant had not crossed paths before.
2009 Searle Civil Justice Institute 301 AAA consumer awards from 2007 ·   Plaintiffs with lawyers won 60% of the time whereas pro se plaintiffs won just 45% of the time.

 

·   Certain repeat-playing defendants settled more often than one-shot defendants, suggesting that case selection is the root of the repeat-player advantage.

2011 Colvin 1,213 AAA employment awards from between 2003 and 2007 ·   Regression analyses estimate that the odds of an employee win decline by 49% against repeat players and 40% in repeat pairings.
2015 Colvin and Gough 2,802 employment awards from between 2003 and 2013  ·   For every additional case in which an employer was involved, the odds of an employee win fell by 0.3%.

 

·   With each repeat pairing, the odds of an employee win decreased by 6.2%.

·   Repeat players were not more likely than one-shot employers to settle cases before the award stage.

2015 CFPB 341 awarded AAA financial service arbitrations filed in 2010 and 2011 ·   Consumers fared worse when they challenged an underlying debt.
2015 Horton and Chandrasekher 4,839 AAA consumer arbitrations filed between July 1, 2009 and December 31, 2013 ·   Although Concepcion did not dramatically increase filing levels, some plaintiffs’ lawyers had filed class action-style cases in arbitration.

 

·   Regression analyses estimated that the odds of a consumer victory fell by 79% against high-level repeaters and 94% against super repeaters.

2016 Horton and Chandrasekher 5,883 AAA employment arbitrations filed between July 1, 2009 and December 31, 2013 ·   Filing levels had decreased since Concepcion, and there was no evidence of class action-style cases.

 

·   Regression analyses estimated that the odds of an employee victory fell by 58% against high-level and super repeaters.

Despite these glimpses into the arbitral forum, many questions remain unanswered. For example, only our After the Revolution papers shed light on filing levels before and after Concepcion.[194] This is unfortunate, because arbitration’s proponents and critics have wildly different views about how the Court’s watershed decision has affected private dispute resolution. Businesses and conservatives contend that the opinion is a “win” for individuals[195] because “consumer[s] fare[] better in an individual arbitration than as a member of a class.”[196] They predict that lay people will initiate waves of cases, seizing the opportunity to “present their claims themselves without the formalities of legal proceedings.”[197] But public interest organizations and liberals are skeptical. They assert that because few consumers and employees will spend the time, money, and energy necessary to pursue their own lawsuits, the only “realistic alternative to a class action is not 17 million individual suits, but zero individual suits . . . .”[198]

In addition, virtually all of the empirical pieces on arbitration focus on the AAA.[199] Yet as mentioned above, the AAA is just one of several leading arbitration providers.[200] Because each of these institutions has a different philosophy, target market, and slate of internal rules,[201] results from the AAA may not be generalizable. Also, the AAA does not administer several common case types, such as medical malpractice and certain debt collection disputes.[202] Accordingly, as we discuss next, we attempted to address these blind spots by greatly expanding the scope of our research.

II. Results

This Section discusses how we collected and cleaned our data. It then presents the results of our analysis of more than 40,000 arbitrations from four institutions.

 A. Methodology and Caveats

Section 1281.96 of the California Code of Civil Procedure requires entities that offer conflict resolution services to release information about every “consumer” arbitration that they have administered in the past five years.[203] The statute defines “consumer” arbitrations to include employment and medical malpractice cases.[204] To comply with the legislation, arbitration providers must list the type of dispute, the claim amount, the identity of the defendant, the identity of the plaintiff’s lawyer, the opening and closing dates, the prevailing party, the name of the arbitrator, the arbitrator’s fees, and the amount of damages awarded.[205] These disclosures have served as the basis for several empirical analyses of arbitration, including Colvin’s, Colvin and Gough’s, and our After the Revolution papers.[206]

We began by collecting disclosures from the third quarter of 2017. We focused on the AAA, JAMS, and ADR Services because they are well-known institutions, and because they publish data in Excel files on their websites.[207] We also included Kaiser because of the paucity of evidence about medical malpractice arbitration.[208] Unfortunately, Kaiser does not release case results in a downloadable format.[209] Thus, we worked with research assistants to harvest records from its website.

We then exploited a unique opportunity to expand the scope of our research. Because section 1281.96 only requires providers to divulge information from the past five years, our initial data collection was limited to disputes that had wrapped up in 2012 or later.[210] However, when the New York Times published its front-page stories on arbitration, it had collected section 1281.96 disclosures from the AAA, JAMS, ADR Services, and Kaiser that included cases decided as early as 2010.[211] One of the reporters was familiar with our After the Revolution pieces and was kind enough to forward these disclosures to us.[212] As a result, we were able to combine two datasets for each provider and examine filings from January 1, 2010 to December 31, 2016.[213]

We followed the same steps for each institution’s files. We began by dropping duplicate records. After that, we cleaned and refined the data.[214] For one, we classified cases as either “consumer,” “employment,” or “tort.”[215] In addition, we created variables related to repeat-player issues. As in the After the Revolution papers, we sorted both plaintiffs’ lawyers and defendants into tiers—one-shotters, low-level repeaters, mid-level repeaters, high-level repeaters, and super repeaters—depending on how often they appeared in a particular provider’s disclosures.[216] In addition, we flagged “repeat pairs”: cases that featured either a plaintiffs’ firm or defendant that had previously argued before the same arbitrator within the same institution.[217]

Our methodology differs in four major ways from our After the Revolution articles. First, in our previous papers, we tried to focus on cases in which a consumer or employee would have been a plaintiff in court by limiting our data to matters that were filed by individuals, (2) that requested $1 or more, and (3) in which the individual’s claim amount exceeded the company’s claim amount.[218] For this Article, we included all disputes, no matter which party initiated them or how much they sought in damages.[219] We did so because not all providers have reliable information on who set the arbitration machinery in motion and how much each party demanded. In addition, recall that the CFPB found that in the financial service industry, many cases involve crosscutting allegations in which a lender attempts to recover a debt and the borrower responds with her own claims.[220] As a result, disputes may not fall neatly into groups based on who would have been the plaintiff in the judicial system. Nevertheless, to sync up with the terminology we used in the After the Revolution papers, we will sometimes refer to individuals as “plaintiffs” and companies as “defendants.”

Second, we changed how we defined a plaintiff “win.” As noted, the After the Revolution papers followed the conventional rule that an individual who obtained $1 or more in damages has prevailed.[221] Conversely, in this Article, a victory occurs if either (1) the plaintiff was designated as the “prevailing party” or (2) the plaintiff recovered more than $1 in damages or attorneys’ fees, unless (3) the business was listed as the “prevailing party.” This is a more nuanced technique than the blunt “$1 or more” approach. It looks first and foremost to the arbitrator’s assessment of which side carried the day, and only employs the “$1 or more” metric as a fallback. This reduces the problem of counting nominal recoveries as “wins” and generous awards of equitable relief or attorneys’ fees as “losses.” It also helps accommodate our new case selection rubric, which features matters in which companies are asserting claims against individuals (for example, to collect a debt). Pegging an individual “win” to a positive damage recovery would treat cases as “losses” even when the arbitrator rebuffed a business’s attempt to obtain a remedy against an individual. Our three-step technique solves this problem by permitting an arbitrator’s nomination of the individual as the “prevailing party” to override an award of $0.

Third, the After the Revolution papers used a logit regression analysis for win rates,[222] but this Article uses a linear probability model (LPM). In our previous work, we chose a logit model to match up with Colvin’s paper.[223] However, the LPM’s regression coefficients are much more straightforward to interpret. Specifically, they tell us how a change in a particular independent variable is associated with a change in the plaintiff’s probability of winning (p). The regression coefficients of the logit model, on the other hand, explain how a change in an independent variable affects the natural logarithm of the odds of winning (or ln(p/(1-p)), a much less intuitive concept. Due to the added complexity of our multiprovider dataset, we opted for simplicity.[224]

Fourth, this Article includes slightly different variables than our After the Revolution papers. For one, we added a factor that we had not examined before: whether the arbitrator was a retired judge or merely a practitioner.[225] In addition, because each provider uses different procedures, we were not able to test the same variables all the time. For example, only the AAA reports whether plaintiffs chose in-person hearings or argued by telephone. Likewise, JAMS reports that some cases involve multiarbitrator panels, whereas other providers either do not offer this procedure or do not report when it has occurred. We will try to be clear how these differences might affect our conclusions.

 B. Results

This Subsection describes our results. For each provider, it first reports information about filing levels, fees, and case length. It then uses simple t-tests to examine whether statistically significant differences in win rates exist between different subgroups of the data. Finally, it uses a multivariate LPM regression analysis to determine the various correlates of the plaintiff win rate.

  1. The AAA

There are 27,332 observations in the AAA data: 14,691 consumer cases (54%) and 12,641 employment cases (46%). These matters yielded 6,229 awards: 4,570 in consumer matters (73%) and 1,659 in employment matters (27%). As Figure 1 reveals, these numbers reflect two huge filing spikes. We will say more about these salvos of claims in Part III.

The AAA epitomizes the rough-and-tumble model of alternative dispute resolution. As Table 1 demonstrates, the mean consumer case length is less than a year and the average arbitrator’s fees are a meager $3,797. Conversely, employment disputes generally last longer (a mean of 424 days) and generate a healthy average arbitral fee of $22,476. Notably, though, the plaintiff’s share of these fees is low, with a mean of $1,438 (in consumer matters) and $314 (in employment cases).

Table 1: AAA Awarded Cases: Overview
  Consumer Employment
Mean (SD) Median N Mean (SD) Median N
Case Length (Days) 255

 

(157)

216 4,570 424

 

(204)

396 1,659
Total Arbitrator’s Fees $3,797

 

($13,694)

$1,059 4,569 $22,476 ($24,658) $16,247 1,644
Plaintiff’s Share of Arbitrator’s Fees $1,438

 

($6,582)

$0 4,557 $314

 

($3,246)

$0 1,643
Notes:

(1)   One consumer case and fifteen employment matters do not reveal the total amount of the arbitrator’s fees. Also, twelve consumer disputes and one employment case lack data on the plaintiff’s share of the arbitrator’s fees.

Several variables exhibit a statistically significant relationship with win rate in the raw data. First, in consumer cases, plaintiffs with one-shot firms prevail 47% of the time, whereas pro se plaintiffs triumph in 29% of matters (p=0.000). Counterintuitively, though, repeat-playing firms have almost exactly the same success rate as pro se consumers. Second, 48% of plaintiffs facing one-shot defendants are victorious, compared to 25% of plaintiffs who encounter repeat-playing defendants (p=0.000).

Results in the employment context are slightly different. For one, as in the consumer context, there is a statistically significant difference in win rates between pro se employees (10%) and (1) employees with one-shot law firms (21%, p = 0.000) and (2) repeat-playing law firms (30%, p = 0.000). However, unlike consumer disputes, there is no statistically significant divergence in win rates based on the defendant’s repeat-player status. Finally, plaintiffs win more often when they appear before former judges (27%) than they do in front of practitioners (20%, p < 0.05).

Table 2: AAA Awarded Cases: Bivariate Analysis
  Consumer Employment
Win Rate N Win Rate N
Overall 33% 4,570 22% 1,659
Plaintiff is

 

Pro Se

29% 1,583 10% 469
Plaintiff Has One-Shot Firm 47%***

 

(p=0.000)

1,215 21%***

 

(p=0.000)

537
Plaintiff Has Repeat-Player Firm†† 28%

 

(p=0.359)

1,772 30%***

 

(p=0.000)

653
One-Shot

Defendant

48% 1,650 24% 406
Repeat-Player Defendant††† 25%***

 

(p=0.000)

2,920 21%

 

 (p=0.114)

1,253
Arbitrator is Practitioner 34% 4,231 20% 1,380
Arbitrator is Former Judge†††† 29%

 

(p=0.086)

323 27%*

 

 (p=0.013)

266
Notes:

(1) Sixteen consumer cases and thirteen employment matters are missing the identity of the arbitrator.

(2) †  T-test compares plaintiffs with one-shot law firms to pro se plaintiffs.

(3)  ††  T-test compares plaintiffs with repeat-playing law firms to pro se plaintiffs.

(4)  †††  T-test compares repeat-playing defendants to one-shot defendants.

(5)  ††††  T-test compares arbitrators who are former judges to arbitrators who are not.

(6)  * p < 0.05, ** p < 0.01, *** p < 0.001

 Table 3, which displays the results of our regression analyses, reinforces our findings in both After the Revolution articles that repeat-player status has a profound impact on results in the AAA. For one, consumers are much less likely to beat high-level, repeat-playing defendants than they are likely to prevail against the reference group of one-shot defendants. Specifically, the mere fact that a consumer faces a high-level repeater decreases the probability of a victory by 23.4 percentage points (p < 0.001). On the flip side, hiring a one-shot law firm increases the chances of a win (relative to proceeding pro se) by 6.9 percentage points (p < 0.01). Nevertheless, the probability of a win declines by 7.6 percentage points when a consumer has a high-level, repeat-playing firm (relative to proceeding pro se) (p < 0.01). We will say more about this puzzling phenomenon in Part III.

Even more than in the consumer setting, repeat-player issues dominate the employment results. When measured against the reference group of one-shot defendants, the probability of an employee prevailing falls by 9.6 percentage points against high-level repeaters (p < 0.05) and 14.9 percentage points against super repeaters (p < 0.001). On the opposite side of the lectern, every cohort of law firm outperforms the reference category of pro se employees. The likelihood of an employee victory rises by 6.7 percentage points for one-shot firms (p < 0.05), 12.2 percentage points for low-level repeaters (p < 0.001), 16.1 percentage points for mid-level repeaters (p < 0.001), 12.6 percentage points for high-level repeaters (p < 0.01), and 10.4 percentage points for super repeaters (p < 0.05).[226]

Table 3: AAA Awarded Cases: Regression Analysis

Linear Probability Model

(Clustered Standard Errors in Parentheses)

  Consumer Employment
Business Sophistication (Reference Category is Plaintiffs Facing One-Shot Businesses)
Low-Level Repeat Player -0.035

 

(0.022)

-0.021

(0.029)

     
Mid-Level Repeat Player -0.074

 

(0.045)

-0.073

(0.037)

     
High-Level Repeat Player -0.234***

 

(0.036)

-0.096*

(0.038)

     
Super Repeat Player -0.149***

(0.031)

     
Business-Arbitrator Familiarity (Reference Category is Plaintiffs Facing Nonrepeat Pairs)
Repeat Pair -0.061

 

(0.037)

0.082*

(0.031)

     
Plaintiffs’ Law Firm Sophistication (Reference Category is Pro Se Plaintiffs)
One-Shot Law Firm 0.069**

 

(0.024)

0.067*

(0.031)

     
Low-Level Repeat Player 0.035

 

(0.032)

0.122***

(0.031)

     
Mid-Level Repeat Player 0.067

 

(0.052)

0.161**

(0.059)

     
High-Level Repeat Player -0.076**

 

(0.028)

0.126**

(0.041)

     
Super Repeat Player †† 0.104*

(0.048)

     
Plaintiffs’ Law Firm-Arbitrator Familiarity (Reference Category is Defendants Not Facing Repeat Pairs)
Repeat Pair -0.002

 

(0.020)

0.036

(0.033)

     
Type of Case (Reference Category is Cases Coded as “Other”) †††
Accounting 0.125*

 

(0.054)

-0.218**

 

(0.075)

     
Aerospace/Defense †††† -0.016

 

(0.177)

     
Automotive †††† -0.032

(0.087)

     
Car Sale/Lease 0.100*

 

(0.045)

††††
     
Car Warranty 0.213**

 

(0.070)

††††
     
Construction †††† 0.159

(0.165)

     
Debt Collection 0.183***

 

(0.046)

††††
     
Education -0.087

 

(0.073)

0.127

(0.121)

     
Energy †††† 0.010

(0.047)

     
Entertainment/Media/Publishing †††† 0.385

(0.327)

     
Financial Services -0.024

 

(0.030)

0.013

(0.063)

     
Health Care †††† -0.097

(0.075)

     
Hospitality/Travel -0.017

 

(0.211)

-0.001

(0.084)

     
Insurance 0.436***

 

(0.093)

0.023

(0.109)

     
Legal Services 0.043

 

(0.083)

0.463*

(0.214)

     
New Home Construction 0.238***

 

(0.023)

††††
     
Pest Control 0.313

 

(0.189)

††††
     
Pharmaceuticals/Medial Devices †††† -0.224***

 

(0.055)

     
Real Estate †††† 0.151

(0.191)

     
Renovation/Home Improvement 0.248**

(0.072)

††††
     
Restaurant/Food Service †††† 0.057

(0.102)

     
Retail †††† -0.013

(0.065)

     
Sports †††† 0.739***

(0.051)

     
Staffing/Employment Agencies †††† 0.444*

(0.172)

     
Steel/Metals †††† 0.320

(0.376)

     
Technology †††† 0.042

(0.114)

     
Telecommunications 0.082

 

(0.052)

0.217

(0.155)

     
Textile/Apparel †††† 0.170*

(0.050)

     
Title Insurance -0.027

(0.179)

††††
     
Transportation †††† -0.160

(0.115)

     
Warranties (Other than Car) 0.206

 

(0.200)

††††
     
Other Controls
Arbitrator is Former Judge 0.009

 

(0.023)

0.033

(0.037)

     
     
Plaintiff Claim Amount in Millions 0.000

 

(0.000)

-0.000***

(0.000)

     
In-Person Hearing 0.039

 

(0.021)

0.074**

(0.022)

     
Documents-Only Proceeding 0.039

 

(0.021)

0.023

(0.057)

     
Individual Filed Case 0.369***

 

(0.023)

0.022

(0.044)

     
Length of Arbitration in Hundreds of Days 0.002

 

(0.065)

0.027***

(0.007)

     
Constant -0.478***

 

(0.077)

-0.106

(0.137)

     
N 4,554 1,659
adj. R2   0.228   0.079
Notes:

(1)     Standard Errors are clustered at the defendant level.

(2)     There are no super repeat-playing defendants in the consumer sample.

(3)     †† There are no super repeat-playing plaintiffs’ firms in the consumer sample.

(4)     ††† The “other” category includes cases coded by the AAA as “other” and any case type with ten or fewer observations.

(5)     †††† Not all case types are present in both the consumer and employment samples.

(6)     This specification includes dummies for the year of the arbitrator’s award.

(7)     * p < 0.05, ** p < 0.01, *** p < 0.001

  1. JAMS

As noted above, JAMS has been the AAA’s primary rival since the late 1970s.[227] Nevertheless, despite its role as a leading arbitration administrator, it has never been subject to empirical research.[228] This Subsection remedies that omission and reveals that there is also a strong repeat-player effect in the forum.

Our data contains 5,366 JAMS cases. Of these cases, 2,094 (39%) involved consumers and 2,463 (46%) featured employees. In addition, unlike the AAA, which refuses to administer medical malpractice claims stemming from predispute arbitration clauses,[229] JAMS handled 808 tort disputes (15% of all cases).[230]

These complaints generated 706 awards. Of these awards, 289 (41%) were in consumer cases, 310 (44%) involved employment causes of action, and 107 (15%) belonged to the tort category.[231]

JAMS lives up to its reputation as a more deliberate alternative to the AAA’s fast-and-furious method of resolving disputes. Consumer cases last for an average of 380 days, employment matters take an average of 504 days, and tort arbitrations wrap up in an average of 438 days. The mean amount of arbitration fees is $14,419 in consumer cases, $37,297 in the employment setting, and $27,583 in the tort context. However, as in the AAA, plaintiffs pay a mere sliver of these costs: a mean of $135 for consumers, $62 for employees, and $745 for tort claimants.

Table 4: JAMS Awarded Cases: Overview
  Consumer Employment Tort
Mean (SD) Median N Mean (SD) Median N Mean (SD) Median N
Case Length (Days) 380

 

 

 

(178)

356 289 504

 

 

 

(258)

441 310 438

 

 

 

(238)

390 107

Total 

Arbitrator’s Fees

$14,419 ($15,334) $9,090 280 $37,297 ($30,029) $31,225 310 $27,583 ($21,428) $26,310 107
Plaintiff’s Share of Arbitrator’s Fees $135  ($1,230) $0 280 $62

 

 

 

($803)

$0 310 $745  ($4,146) $0 107
Notes:

(1) Nine consumer cases are missing information about the total amount of the arbitrator’s fees.

Repeat-player status looms large in the bivariate data. For starters, in consumer cases, plaintiffs prevailed in 36% of cases involving one-shot defendants, but just 19% of the time against repeat players (p < 0.01). Likewise, in tort disputes, the plaintiff win rate was 67% against one-shot defendants but only 20% against repeaters (p < 0.05). In addition, plaintiffs reaped the benefits of legal representation. Pro se consumers were victorious in 6% of matters, but those with one-shot law firms succeeded in 38% of awards (p < 0.001) and those with repeat-playing firms won 31% of decisions (p < 0.001). Moreover, pro se employees were victorious in 10% of cases, which pales next to the 43% mark achieved by employees with repeat-playing attorneys (p < 0.001).[232]

Table 5: JAMS Awarded Cases: Bivariate Analysis
  Consumer Employment Tort
Win Rate N Win Rate N Win Rate N
Overall 21% 289 31% 310 24% 107
Plaintiff is

Pro Se

6% 123 10% 48 18% 22
Plaintiff Has One-Shot Firm 38%***

 

 

 

(p=0.000)

45 23%

 

 

 

(p=0.068)

100 21%

 

 

 

(p=0.793)

38
Plaintiff Has Repeat-Player Firm†† 31%***

 

 

 

(p=0.000)

121 43%***

 

 

 

(p=0.000)

162 30%

 

 

 

(p=0.313)

47
One-Shot

 

 

 

Defendant

36% 44 29% 99 67% 9
Repeat-Player Defendant††† 19%**

 

 

 

(p=0.008)

245 32%

 

 

 

(p=0.605)

211 20%**

 

 

 

(p=0.002)

98
Arbitrator is Practitioner 19% 59 38% 60 45% 22
Arbitrator is Former Judge†††† 22%

 

 

 

(p=0.537)

228 30%

 

 

 

(p=0.191)

250 19%**

 

 

 

(p=0.009)

85
Notes:

(1)    Two consumer cases are missing information about the identity of the arbitrator.

(2)    T-test compares plaintiffs with one-shot law firms to pro se plaintiffs.

(3)    ††  T-test compares plaintiffs with repeat-playing law firms to pro se plaintiffs.

(4)    †††  T-test compares repeat-playing defendants to one-shot defendants.

(5)    ††††  T-test compares arbitrators who are former judges to arbitrators who are not.

(6)    * p < 0.05, ** p < 0.01, *** p < 0.001

Our regression analyses reinforce the importance of repeat-player status within JAMS. For starters, the fact that a tort plaintiff faces a high-level, repeat-playing defendant, rather than a one-shot firm, reduces the probability of a win by 46.2 percentage points (p < 0.05).[233] Likewise, repeat-playing plaintiffs’ lawyers thrive when contrasted against the reference group of pro ses. For instance, in consumer cases, the probability of a win increases by the following margins: 31.2 percentage points for one-shot firms (p < 0.001), 27.5 percentage points for low-level, repeat-playing firms (p < 0.01), and a whopping 79.9 percentage points for super repeat-playing firms (p < 0.001). Employment matters are similar. In that niche, the chances of plaintiff success rise by 25.3 percentage points for low-level repeaters (p < 0.01), 25.7 percentage points for mid-level repeaters (p < 0.01), and 55.1 percentage points for high-level repeaters (p < 0.001).

Table 6: JAMS Awarded Cases: Regression Analysis

Linear Probability Model

(Clustered Standard Errors in Parentheses)

  Consumer Employment Tort
Business Sophistication (Reference Category is Plaintiffs Facing One-Shot Businesses)
Low-Level Repeat Player -0.067

 

 

 

(0.098)

0.046

 

 

 

(0.066)

-0.152

(0.252)

       
Mid-Level Repeat Player -0.066

 

 

 

(0.08)

-0.109

 

 

 

(0.060)

-0.302

 

 

 

(0.236)

       
High-Level Repeat Player -0.122

 

 

 

(0.078)

0.154

 

 

 

(0.271)

-0.462*

(0.198)

       
Super Repeat Player †† †††
       
Business-Arbitrator Familiarity (Reference Category is Plaintiffs Facing Nonrepeat Pairs)
Repeat Pair -0.040

 

 

 

(0.052)

0.016

 

 

 

(0.074)

-0.036

(0.111)

       
Plaintiffs’ Law Firm Sophistication (Reference Category is Pro Se Plaintiffs)
One-Shot Law Firm 0.312***

 

 

 

(0.063)

0.120

 

 

 

(0.070)

-0.071

(0.185)

       
Low-Level Repeat Player 0.275**

 

 

 

(0.066)

0.253**

 

 

 

(0.088)

-0.067

(0.089)

       
Mid-Level Repeat Player 0.253

 

 

 

(0.129)

0.257**

 

 

 

(0.085)

0.001

(0.165)

       
High-Level Repeat Player 0.168

 

 

 

(0.106)

0.551***

 

 

 

(0.134)

-0.077

(0.143)

       
Super Repeat Player 0.799***

 

 

 

(0.168)

0.529

 

 

 

(0.345)

††††
       

 

Plaintiffs’ Law Firm-Arbitrator Familiarity (Reference Category is Defendants Not Facing Repeat Pairs)
Repeat Pair -0.075

 

 

 

(0.103)

-0.016

 

 

 

(0.125)

0.632***

(0.047)

       
Other Controls
Arbitrator is Former Judge 0.013

 

 

 

(0.047)

-0.071

 

 

 

(0.067)

-0.174*

 

 

 

(0.082)

       
Multi-Arbitrator Panel 0.199

 

 

 

(0.198)

0.412

 

 

 

(0.320)

0.442

 

 

 

(0.398)

       
Length of Arbitration in Hundreds of Days 0.010

 

 

 

(0.015)

-0.014

 

 

 

(0.010)

0.030*

(0.000)

       
Constant 0.166

 

 

 

(0.119)

0.165

 

 

 

(0.130)

0.699

(0.376)

       
N 287 310 107
adj. R2   0.212   0.112   0.161
Notes:

(1)     The regression sample is smaller than the overall sample because some cases are missing data.

(2)     Standard Errors are clustered at the defendant level.

(3)     There are no super repeat-playing defendants in the consumer sample.

(4)     †† There are no super repeat-playing defendants in the employment sample.

(5)     ††† There are no super repeat-playing defendants in the tort sample.

(6)     †††† There are no super repeat-playing plaintiffs’ firms in the tort sample.

(7)     This specification includes dummies for the year of the arbitrator’s award.

(8)     * p < 0.05, ** p < 0.01, *** p < 0.001

  1. ADR Services

During our research period, plaintiffs filed 3,302 arbitrations in ADR Services. Of these arbitrations, 2,037 (62%) were tort cases, 1,132 (34%) were employment matters, and 133 (4%) revolved around consumer issues. Arbitrators issued 737 awards: 535 (73%) in cases brought by tort plaintiffs, 174 (24%) featured employees, and 28 (4%) involved consumers. Because there were so few consumer cases, we will focus on tort and employment disputes.[234]

ADR Services cases move quickly. Tort matters wrapped up in an average of 187 days and employment disputes concluded in a mean of 251 days.

However, unlike the AAA and JAMS, ADR Services has not enacted protocols to protect consumers, employees, or medical patients.[235] As a result, plaintiffs pay a greater proportion of the arbitrator’s fees: an average of $2,718 in the tort milieu and $2,534 in the employment setting.

Table 7: ADR Services Awarded Cases: Overview
  Tort Employment
Mean (SD) Median N Mean (SD) Median N
Case Length (Days) 187

 

 

 

(151)

144 516 251

 

 

 

(230)

198 170

Total

Arbitrator’s Fees

$8,554  ($9,231) $5,263 528 $12,033 ($13,085) $6,011 174
Plaintiff’s Share of Arbitrator’s Fees $2,718

 

 

 

($3,197)

$2,038 524 $2,534

 

 

 

($3,612)

$1,875 173
Notes:

(1)    Nineteen tort matters are missing information about case length. Seven more tort cases do not reveal the total amount of arbitrator’s fees. An additional four tort files do not disclose the plaintiff’s share of the fees.

(2)    Four employment matters are missing information about case length, and one case does not reveal the plaintiff’s share of the arbitrator’s fees.

Three variables in tort cases affected win rates in a statistically significant manner. First, plaintiffs with attorneys topped pro se plaintiffs, 68% to 42% (p < 0.01). Second, plaintiffs succeeded against one-shot firms in 74% of matters, but only beat repeat players 63% of the time (p < 0.05). Third, plaintiffs were victorious 75% of the time in front of practitioners, but a mere 63% of the time before former judges (p < 0.05).

The employment results were less revealing. The most interesting finding does not relate to employee success levels. Instead, it is the fact that just two of the 174 employees (1%) represented themselves.

Table 8: ADR Services Awarded Cases: Bivariate Analysis
  Tort Employment
Win Rate N Win Rate N
Overall 67% 535 59% 174
Plaintiff is

 

 

 

Pro Se

42% 36 50% 2
Plaintiff Has Law Firm 68%**

 

 

 

(p=0.001)

499 58%

 

 

 

(p=0.805)

172
One-Shot

 

 

 

Defendant

74% 160 57% 96
Repeat-Player Defendant††† 63%*

 

 

 

(p=0.021)

375 60%

 

 

 

(p=0.695)

78
Arbitrator is Practitioner 75% 146 58% 12
Arbitrator is Former Judge†††† 63%*

 

 

 

(p=0.015)

389 59%

 

 

 

(p=0.983)

162
Notes:

(1)    †  T-test compares plaintiffs with law firms to pro se plaintiffs.

(2)    ††  T-test compares repeat-playing defendants to one-shot defendants.

(3)    †††  T-test compares arbitrators who are former judges to arbitrators who are not.

(4)    * p < 0.05, ** p < 0.01, *** p < 0.001

 We limited our regression analysis to tort cases, where two metrics stood out. First, the fact that a plaintiff faced a high-level, repeat player reduced the likelihood of a victory by 26.7 percentage points (p < 0.05). Second, hiring a lawyer increased the probability of winning by 21.0 percentage points (p < 0.05).

Table 9: ADR Services Awarded Cases

Tort Regression

Linear Probability Model

(Clustered Standard Errors in Parentheses)

Business Sophistication (Reference Category is Plaintiffs Facing One-Shot Businesses)
Low-Level Repeat Player 0.010

(0.079)

   
Mid-Level Repeat Player -0.029

(0.051)

   
High-Level Repeat Player -0.267*

(0.062)

   
Super Repeat Player
   
Business-Arbitrator Familiarity (Reference Category is Plaintiffs Facing Nonrepeat Pairs)
Repeat Pair 0.035

 

 

 

(0.050)

   
Plaintiffs’ Law Firm (Reference Category is Pro Se Plaintiffs)
Plaintiff Has Law Firm 0.210*

(0.104)

   
Other Controls
Arbitrator is Former Judge -0.028

(0.042)

   
Length of Arbitration in Hundreds of Days -0.073***

(0.014)

   
Constant 0.728***

(0.127)

   
N 516
   
adj. R2   0.131
Notes:

(1)     The regression sample is smaller than the overall sample because some cases are missing data.

(2)     There are no super repeat-playing defendants in the sample.

(3)     This specification includes dummies for the year of the arbitrator’s award.

(4)     * p < 0.05, ** p < 0.01, *** p < 0.001

  1. Kaiser

Like JAMS and ADR Services, the Kaiser regime has never been investigated by empirical researchers. This lacuna is surprising for two reasons. First, Kaiser is one of the nation’s largest HMOs, boasting 11.8 million members, thirty-nine hospitals, 22,080 doctors, and an annual operating revenue of more than $70 billion.[236] Second, arbitrating medical malpractice claims has long been controversial, and Kaiser’s arbitration caseload is almost exclusively medical malpractice claims.[237]

Kaiser administered 4,775 cases during our research period. Of these cases, 4,659 (98%) were tort, while the remaining 116 (2%) were debt collection. Similarly, of the 1,008 awarded cases, 977 (98%) were tort. Due to the lack of consumer cases, we excluded them.[238]

Kaiser’s tort cases last for an average of 453 days. As noted above, thanks to the Blue Ribbon Panel’s recommendations, the system permits plaintiffs to shift most of the arbitrator’s fees to Kaiser.[239] Thus, although the average cost of the private judge was $14,635, claimants bore only an average of $368 of this sum.

Table 10: Kaiser Tort Awards: Overview
  Mean (SD) Median N
Case Length (Days) 453

 

 

 

(191)

413 980
Total

 

 

 

Arbitrator’s Fees

$14,635

 

 

 

($16,438)

$8,200 962
Plaintiff’s Share of Arbitrator’s Fees $368

 

 

 

($2,200)

$0 961
Notes:

(1)  Eighteen matters do not mention the total amount of arbitrator’s fees. In addition, one case does not disclose the plaintiff’s share of the fees.

Our bivariate analysis reveals that pro se plaintiffs fail spectacularly in Kaiser tort arbitrations. Out of the 459 self-represented plaintiffs, fewer than 1% prevailed. Conversely, one-shot law firms won 14% of cases and repeat-playing firms were victorious 36% of the time (both p < 0.001).

Table 11: Kaiser Tort: Raw Win Rates
  Win Rate N
Overall 16% 980
Plaintiff is Pro Se <1% 459
Plaintiff Has One-Shot Firm 14%***

 

 

 

(p=0.000)

144
Plaintiff Has Repeat-Player Firm†† 36%***

 

 

 

(p=0.000)

377
Arbitrator is Practitioner 14% 469
Arbitrator is Former Judge†††† 17%

 

 

 

(p=0.152)

511
Notes:

(1)  †  T-test compares plaintiffs with one-shot law firms to pro se plaintiffs.

(2)  ††  T-test compares plaintiffs with repeat-playing law firms to pro se plaintiffs.

(3)  ††††  T-test compares arbitrators who are former judges to arbitrators who are not.

(4)  * p < 0.05, ** p < 0.01, *** p < 0.001

As one would expect given the scarcity of pro se plaintiff victories, attorney representation dominates the regression results. The prospect of a win increases when a plaintiff selects a one-shot firm (by 12.6 percentage points, (p < 0.001)), a low-level repeater (by 25.2 percentage points, (p < 0.001)), a mid-level repeater (by 42.1 percentage points, (p < 0.001)), and a high-level repeater (also by 41.6 percentage points, (p < 0.001)). Oddly, a plaintiff’s choice to hire a super repeat-playing law firm does not impact the likelihood that she will prevail by a statistically significant margin (p = 0.08).

Table 12: Kaiser Awarded Cases

Tort Regression

Linear Probability Model

(Robust Standard Errors in Parentheses)

Plaintiffs’ Law Firm Sophistication (Reference Category is Pro Se Plaintiffs)
One-Shot Law Firm 0.126***

 

 

 

(0.030)

   
Low-Level Repeat Players 0.252***

(0.035)

   
Mid-Level Repeat Players 0.421***

 

 

 

(0.051)

   
High-Level Repeat Players 0.416***

(0.072)

   
Super Repeat Players 0.268

(0.151)

   
Plaintiffs’ Law Firm-Arbitrator Familiarity (Reference Category is Defendants Not Facing Repeat Pairs)
Repeat Pair 0.042

(0.065)

   
Other Controls
Arbitrator is Retired Judge -0.006

 

 

 

(0.021)

   
Length of Arbitration in Hundreds of Days 0.014

(0.007)

   
Constant -0.035

(0.042)

   
N 980
   
adj. R2   0.219
Notes:

(1)     †   Because there is only one defendant in the Kaiser sample, we used robust standard errors, rather than clustering them at the defendant level.

(2)     This specification includes dummies for the year of the arbitrator’s award.

(3)     * p < 0.05, ** p < 0.01, *** p < 0.001

III. Descriptive and Normative Implications

This Section brings our research to bear on the forced arbitration controversy. It begins by describing arbitration’s benefits and drawbacks. It argues that although arbitration is quick and economical, it is not currently picking up the slack left by the decline of the class action. Thus, to fix this problem, this Section urges state legislatures to create pecuniary incentives for plaintiffs’ firms to arbitrate.

 A. The State of Contemporary Arbitration

In the post-Concepcion era, there is vast disagreement about whether arbitration facilitates or thwarts claims. Businesses and conservatives argue that the class action is unnecessary because pro se plaintiffs can capitalize on arbitration’s speed and convenience to vindicate their own rights.[240] But the plaintiffs’ bar, public interest organizations, and liberals see this as unrealistic.[241] This Subsection explains that both sides are partially right. Arbitration is accessible, but it is also a pale substitute for the class action. Moreover, plaintiffs’ lawyers—not individual claimants—are in the best position to reap arbitration’s benefits.

  1. Time to Award

Arbitration has the potential to be an elegant shortcut to the court system. For starters, it is almost certainly faster than litigation. The US Department of Justice surveyed verdicts from state courts in seventy-five large counties in 2005 and found that the average disposition times were 26.6 months for jury trials and 20.8 months for bench trials.[242] More specifically, Michael Heise examined two random samples of jury trials from state courts and determined that the mean length of employment cases was 31.2 months and medical malpractice matters was 38.4 months.[243] In sharp contrast, the mean length of all awarded arbitrations in our dataset was less than eleven months.[244] Employment arbitrations concluded in an average of about fourteen months and our malpractice-heavy sample of tort matters clocked in at a mean of only twelve months. Thus, speed is one of private dispute resolution’s greatest virtues.

  1. Costs

Likewise, we discovered that the process is surprisingly affordable for plaintiffs. Combining all case types and all providers, the average plaintiff’s share of arbitrators’ fees was a manageable $1,114. This was true even in settings where overall costs were high. The best example is JAMS employment cases, where the mean arbitrator’s fee was a hefty $37,297, but plaintiffs only picked up an average of $62 of the tab. Moreover, the median fee in the sample of all awarded cases is $0, which means that the majority of plaintiffs who fully prosecuted their claims pay no fees.

  1. Filings and Representation

Despite arbitration’s virtues, it has not served as a surrogate for the class mechanism. Our data contains many Fortune 100 companies that once faced class actions, such as Amazon.com, Allstate Insurance, American Express, Anthem, AT&T, Bank of America, Cigna, Citigroup, Comcast, Fannie Mae, FedEx, Ford Motor Company, General Motors, Lockheed Martin, J.P. Morgan Chase, Microsoft, State Farm Insurance, UPS, Verizon, and Wells Fargo.[245] The Supreme Court’s FAA jurisprudence has funneled these massive lawsuits into the arbitral forum.[246] Yet contrary to predictions by businesses and their allies, Concepcion did not spawn a surge in arbitral filings.[247] Although the total number of arbitrations in all institutions rose after Concepcion, it did so modestly. Between January 1, 2010, when our research period begins, and April 27, 2011, when the Court handed down its decision, plaintiffs initiated 6,411 cases, or an average of 401 per month. By contrast, between May 2011 and December 2016, there were 34,364 filings, or a mean of 614 per month. A monthly increase of about 200 arbitrations does not compare to the thousands or
millions of complaints that were once bundled into a single class action.[248]

Moreover, contrary to assertions made by businesses and their allies, pro se plaintiffs have not led the charge into the arbitral forum.[249] Even as the total number of claims climbed after Concepcion, the volume of pro se filings decreased. For example, between January 2010 and April 27, 2011, self-represented plaintiffs in consumer cases filed an average of sixty-two complaints per month. Afterwards, the number fell to a mean of fifty-five. And although companies have singled out employment arbitration as a niche that is welcoming to self-represented plaintiffs,[250] the decline in pro se filings has been particularly sharp in that context. Before Concepcion, an average of thirty-six pro se employees initiated arbitrations each month; afterwards, the mean was just twenty-three. Figure 5 traces this decline in pro se filings.

Instead of pro se litigants, plaintiffs’ lawyers have driven the uptick in arbitrations after Concepcion. Recall that in Consumer Arbitration, we discovered that attorneys had started filing numerous related causes of action against the same company.[251] For example, a single plaintiffs’ firm initiated 1,094 consumer arbitrations on the same day against AT&T.[252] Likewise, a different group of lawyers initiated roughly 200 arbitrations over the course of a few weeks against Sallie Mae. Conversely, in Employment Arbitration, we found no evidence of this trend in AAA employment cases.[253]

Nevertheless, our research for this Article reveals class action-style cases in both the consumer and employment spheres. As Figure 6 depicts, a firm filed 1,354 employment cases against Macy’s in the AAA on September 26, 2013.[254] About a month later, the same lawyers sued the company an additional 231 times. And as Figure 7 demonstrates, there were similar (albeit much smaller) spikes in JAMS, as the same attorneys initiated several dozen employment arbitrations against Prospect Mortgage and consumer complaints against Asco Equipment.

 

As Figures 6 and 7 also demonstrate, class action-style cases are practically the only arbitrations brought against these mammoth firms. Indeed, there are few standalone claims. This makes sense: an attorney is much more likely to sink resources into investigating corporate wrongdoing if the facts can serve as the springboard for multiple lawsuits, rather than one. Thus, businesses are half correct when they argue that for plaintiffs with small claims, “the choice is ‘arbitration—or nothing.’”[255] The choice is actually a class action-style arbitration or nothing. If an individual is going to try to vindicate her rights at all, it is probably going to be in a wave of related complaints engineered by a plaintiff’s law firm.

  1. Win Rates

We found that the plaintiff win rate in arbitration is generally lower than its analogue in the judicial system. Although we are not aware of any studies of the outcome of consumer cases in court, plaintiffs prevail in 55% to 60% of non-employment-related civil trials in state court[256] and 85% to 89% of matters in small claims court.[257] By contrast, in our data, consumers were victorious in 33% of AAA cases and in 21% of JAMS matters. A similar pattern exists in the employment sphere. Employees succeed about a third of the time in federal court and half of the time in state court.[258] In arbitration, the employee win rate runs the gamut from 22% in the AAA to 31% in JAMS to 59% in ADR Services. Finally, although roughly 27% of medical malpractice plaintiffs recover damages in state court,[259] 16% of plaintiffs won in Kaiser arbitrations. Thus, plaintiffs win less frequently in the private forum.

Admittedly, this does not prove that arbitration is less hospitable to plaintiffs than the court system. Because of selection biases, we cannot draw strong inferences about how arbitration and litigation compare from win rates. For one, the case streams that feed the two forums diverge. Thus, evidence that plaintiffs prevail less often in arbitration does not necessarily mean that the process is skewed against them; rather, it could reflect differences in the nature or quality of claims that are subject to forced arbitration clauses.[260] Moreover, win rates are not always probative of whether the law favors one side or the other.[261] Litigants should be able to value, and thus settle, most disputes.[262] Accordingly, only extremely close cases—where the parties cannot agree on the range of possible outcomes—should proceed to a final ruling.[263] This means that any sample of arbitral awards or judicial verdicts will contain a disproportionate number of disputes that could easily go either way. In turn, no matter how the applicable rules are calibrated—whether they are pro-plaintiff or pro-defendant—this bias will not appear in any study of awards; instead, plaintiffs and defendants should each prevail about half of the time.[264] Even departures from this anticipated 50% success rate would not prove that the law is slanted one way or the other. Instead, it would suggest that one class of litigants has incentives not to settle (and thus will prosecute losing cases until the bitter end) or has better information about the odds of prevailing.[265] For these reasons, it would be unwise to use arbitration-to-litigation comparisons as the springboard for bold policy prescriptions.

However, contrasting the win rates of various cohorts within arbitration is more instructive. One of our most striking findings is that pro se plaintiffs struggle mightily. Indeed, as noted, the victory rate for self-represented individuals is (1) 6% in JAMS consumer disputes, (2) 10% in AAA and JAMS employment cases, and (3) less than 1% in Kaiser.

There are several plausible explanations for why pro ses do so poorly. First, they might be self-represented because their cases are too weak to attract a lawyer.[266] Second, the low victory rates among pro ses in Kaiser (and, to a lesser extent, in AAA and JAMS employment) might stem from the challenges that complex employment and medical malpractice claims pose for lay litigants.[267] Third, the Kaiser results might flow from the surprising prevalence of dispositive motions (which can be confusing for nonlawyers). According to the conventional wisdom, summary judgment in arbitration is “so rare as to be statistically insignificant.”[268] But in sharp contrast, we found that summary judgment motions are Kaiser’s primary line of defense against pro se patients. A staggering 377 of the 459 self-represented tort plaintiffs (82%) in our dataset lost at the summary judgment stage. Fourth, the wave of reforms in the 1990s may have made arbitration less hospitable to pro ses.[269] By encumbering arbitration with complex rules, they arguably made the process less transparent.[270] Even something as benign-seeming as the shared responsibility for choosing an arbitrator could reward repeat players for their familiarity with a potential decision-maker’s background, demeanor, and philosophy. This new system might help plaintiffs’ lawyers but it only magnifies the informational gap between corporations and the archetypical pro se individual “complaining about a product.”[271] Finally, as we discuss below, pro ses are less likely to settle disputes than represented plaintiffs. This means that they may arbitrate a greater percentage of weak cases all the way to the award stage, lowering their win rates.

  1. Repeat Players

Our research highlights the importance of experience within the arbitral forum. As we mentioned, scholars have long suspected that arbitration favors repeat-playing defendants.[272] Our data is consistent with this theory. In our regression analyses, when contrasted against the reference group of one-shot firms, the probability of a plaintiff win declines in AAA consumer cases against high-level repeaters, AAA employment matters against high-level and super repeaters, JAMS tort cases against high-level repeaters, and ADR Services tort arbitrations against high-level repeaters.

In addition, we discovered that this coin has a flip side. Arbitration’s proponents argue that even if the process is slanted toward repeat playing corporations, it might also tilt toward repeat-playing plaintiffs’ lawyers.[273] Although previous empirical studies have not been able to verify this claim,[274] our research brings it into sharp relief. In most providers and case types, the likelihood of a plaintiff win increases (relative to being pro se) when the plaintiff hires a law firm with several arbitrations under its belt. Specifically, there is a plaintiffs’ firm repeat-player effect in AAA employment cases with low-level, mid-level, high-level, and super repeaters, JAMS consumer cases with low-level and super repeaters, JAMS employment cases with low-level, mid-level, and high-level repeaters, and Kaiser tort cases with low-level, mid-level, and high-level repeaters. Thus, legal representation—especially with a firm that has arbitrated before—is paramount.

Finally, we investigated the idea that repeat-players win more often due to case selection. As noted above, the Searle Report suggested that repeat players enjoy higher win rates because they settle more than their counterparts and thus only fully prosecute cases they are likely to win.[275] Following Colvin and Gough’s lead, we ran simplified versions of our regression analyses using the fact that a case terminated before the award stage—rather than the fact that a plaintiff prevailed—as the dependent variable.[276] If case selection is the root of the repeat-player advantage, we would expect cohorts that enjoy higher probabilities of success on the merits to be more likely to settle than other groups.

Like Colvin and Gough, we unearthed little support for the case selection theory among repeat-playing defendants. Table 13 displays the results of our settlement regression analyses for the categories of repeat-playing companies that outperformed one-shot firms in our win rate regression analyses. It reveals that there is no statistically significant difference in the probability of settlement between one-shot firms and both AAA consumer high-level repeaters, AAA employment high-level repeaters, and JAMS tort high-level repeaters. Even more notably, compared to one-shot defendants, the probability of a settlement declines by 10.3 percentage points for AAA employment super repeaters (p < 0.001). The fact that these repeat players settle less often and yet enjoy higher win rates than one-shotters casts doubt on the case selection theory in this niche.

 

Table 13: Summary of Settlement Regression Analyses

 

 

 

All Providers: Repeat-Playing Defendants

AAA Consumer High-Level Repeat Players 0.026

 

 

 

(0.064)

AAA Employment High-Level Repeat Players -0.000

 

 

 

(0.023)

AAA Employment Super Repeat Players -0.103***

(0.013)

JAMS Tort High-Level Repeat Players 0.028

(0.048)

ADR Services High-Level Repeat Players 0.114***

(0.028)

Notes:

(1)     The full regression tables are available by request.

(2)     * p < 0.05, ** p < 0.01, *** p < 0.001

Conversely, the case selection story has more purchase in the context of plaintiffs’ firms. Table 14 demonstrates that most categories of plaintiffs’ attorneys that enjoyed higher win rate probabilities than self-represented claimants were also more likely to settle. Compared to pro ses, the chances of early termination of a dispute increased as much as 26.3 percentage points for super repeaters in AAA employment cases (p < 0.001) and 26.2 percentage points for high-level repeaters in Kaiser disputes (p < 0.001). Only one such group—low-level repeaters in JAMS consumer matters—did not settle more than pro ses by a margin that was statistically significant.

Table 14: Summary of Settlement Regression Analyses

All Providers: Repeat-Playing Plaintiffs’ Law Firms

AAA Employment Low-Level Repeat Players 0.146***

 

 

 

(0.021)

AAA Employment Mid-Level Repeat Players 0.181***

 

 

 

(0.025)

AAA Employment High-Level Repeat Players 0.212***

(0.031)

AAA Employment Super Repeat Players 0.263***

(0.015)

JAMS Consumer Low-Level Repeat Players -0.033

 

 

 

(0.026)

JAMS Consumer Super Repeat Players 0.093**

 

 

 

(0.028)

JAMS Employment Low-Level Repeat Players 0.190***

(0.030)

JAMS Employment Mid-Level Repeat Players 0.224***

(0.030)

JAMS Employment High-Level Repeat Players 0.243***

(0.034)

Kaiser Low-Level Repeat Players 0.183***

(0.019)

Kaiser Mid-Level Repeat Players 0.234**

(0.019)

Kaiser High-Level Repeat Players 0.262***

(0.020)

Notes:

(1)     The full regression tables are available by request.

(2)     * p < 0.05, ** p < 0.01, *** p < 0.001

  1. Summary

To conclude, our study renders a mixed verdict on the forced arbitration debate. The process’ speed and relatively low costs are encouraging. But despite these virtues, self-represented plaintiffs file few claims and rarely prevail on the merits or obtain relief in the form of a settlement. Together, these findings suggest that one way for policy-makers to ameliorate Concepcion’s harsh effects would be to inspire plaintiffs’ lawyers to pursue more arbitrations. Indeed, that would both leave fewer plaintiffs without counsel and increase the number of arbitration-savvy plaintiffs’ attorneys. In the next Subsection, we propose a step in that direction.

 B. The Arbitration Multiplier

This Subsection urges state legislatures to create an “arbitration multiplier”: a bonus for plaintiffs’ lawyers who prevail in arbitration. It first explores the advantages and drawbacks of this approach. It then explains why this rubric, unlike virtually every effort by states to regulate arbitration, would not be preempted by the FAA.

  1. Creating Incentives to Arbitrate

Dozens of federal and state statutes allow prevailing plaintiffs’ lawyers to recover their fees and costs.[277] For example, when an employee wins a case for unpaid wages or overtime, the Fair Labor Standards Act (FLSA) instructs courts to award “a reasonable attorney’s fee to be paid by the defendant, and costs of the action.”[278] Likewise, the California Consumer Legal Remedies Act—which was at issue in Concepcion—rewards lawyers who win consumer fraud claims with their fees and litigation expenses.[279] These fee-shifting statutes allow litigants to become “private Attorney Generals.”[280] They inspire plaintiffs to prosecute claims that are socially valuable, but which they might not otherwise bring “because the gains would not sufficiently further their [own] interests.”[281]

We envision the arbitration multiplier as a standalone state law that piggybacks on these statutes. It would permit an arbitrator who finds that a consumer, employee, or medical patient is entitled to fees or expenses under a fee-shifting law to augment the award by a factor of her choosing.[282] In addition, legislatures should make the multiplier nonwaivable to prevent defendants from contracting around it in the fine print. The multiplier would encourage plaintiffs’ attorneys to capitalize on arbitration’s speed and affordability by seeking redress for the small-dollar, widely-disbursed harm that once gave rise to class actions. In addition, by incentivizing lawyers to take cases, it would reduce the number of pro se claimants and allow more plaintiffs’ firms to gain experience within the forum.[283]

Our research suggests that plaintiffs’ attorneys would respond to such a regime. As noted, AT&T and Sallie Mae offer lump-sum payments and double attorneys’ fees for plaintiffs who arbitrate on an individual basis and recover more than their last settlement offer.[284] As we mentioned in Part III.A.3, attorneys have singled out these companies for class action-style arbitrations.[285] Accordingly, incentives like the multiplier seem to influence lawyers’ behavior.

A few points are necessary to clarify the scope of our proposal and address counterarguments. For starters, we acknowledge that the arbitration multiplier would be more complicated in the medical malpractice context.[286] Personal injury lawyers work on a contingency basis, so they collect their fees from the plaintiff’s judgment (not from the defendant). Thus, there are far fewer fee and cost-shifting statutes in the field. Moreover, payouts in medical malpractice cases are heavily regulated. As part of the tort reform movement in the late 1970s and early 1980s, several jurisdictions capped the amount of damages or fees that plaintiffs and their lawyers can receive.[287] Thus, states would need to find creative ways to import the multiplier into this realm.[288]

In addition, a skeptic might assert that the multiplier will trigger a tsunami of frivolous arbitrations. One data point suggests that this is a legitimate concern. As we mentioned, our regression analyses of AAA consumer cases reveals that high-level, repeat-playing law firms reduce the probability of a plaintiff win.[289] However, there is a simple explanation for this topsy-turvy conclusion: 58% of high-level repeater cases involve a bogus debt collection agency called World Law Group, which once filed numerous baseless arbitrations on behalf of its clients.[290] Yet one still might cite this perverse statistic and contend that arbitration is too accessible. Perhaps the last thing the law should do is subsidize the filing of meritless claims.

Nevertheless, we believe that the benefits of the multiplier outweigh these risks. There are checks on shakedown lawsuits. Some statutes allow two-way fee-shifting, permitting defendants to recover their fees and costs from vexatious plaintiffs.[291] In addition, extreme abuses can trigger governmental intervention. For example, in 2015, CFPB sued World Law Group, driving it out of business.[292] Moreover, even putting World Law Group to the side, there may be a benign explanation for the backwards-seeming effect of high-level, repeat-playing consumer attorneys. Recall that previous studies of debt collection arbitrations found that creditors prevailed in almost 90% of cases.[293] Many of the lawyers in the high-level repeater category seem to handle such matters. As a result, their low win rate may reflect the difficulty of winning such a dispute, rather than the fact that they have hijacked the arbitration process.

At the opposite pole, one might argue that the multiplier does not go far enough. Indeed, like most fee-shifting regimes, it would be merely permissive.[294] As a result, anyone who believes that arbitrators are hostile to plaintiffs or their lawyers might worry that arbitrators will routinely deny requests for augmented fees and costs. However, our data tell a different story. The AAA disclosures include information about attorney fee awards. Plaintiffs prevailed in 1,887 AAA consumer and employment cases, and arbitrators awarded fees a respectable 522 times (28%). Moreover, these fee awards were hardly stingy. Because of a few large judgments, the average amount of fees awarded in these 522 cases was a healthy $127,139, with a median of $16,723. Therefore, arbitrators are not shy about rewarding successful plaintiffs.

  1. Preemption

Finally, unlike decades of failed state attempts to regulate arbitration, the multiplier would not be preempted by the FAA. The FAA displaces state rules that try to exempt franchise,[295] wage,[296] and tort claims[297] from the private tribunal. Likewise, the statute trumps state doctrines that “impose[] burdens on arbitration agreements that do not apply to contracts generally.”[298] For example, Montana passed a law that invalidated arbitration clauses unless the drafter gave “[n]otice that a contract is subject to arbitration . . . in underlined capital letters on the first page of the contract . . . .”[299] In Doctor’s Assocs., Inc. v. Casarotto, the Court held that the notice requirement was inconsistent with the FAA, reasoning that “Congress precluded [s]tates from singling out arbitration provisions for suspect status.”[300] More recently, in Kindred Nursing Centers Ltd. Partnership v. Clark, the Justices held that the FAA preempts a Kentucky rule that refused to honor arbitration clauses in contracts signed by agents under a power of attorney unless the document expressly allowed the agent to waive the principal’s right to access the courts.[301]

However, the multiplier differs from these impermissible state laws in two important ways. First, every preempted state principle so far has been “enforcement-impeding”: it purports to invalidate all or part of an agreement to arbitrate.[302] As a result, these state rules clash with the text of Section 2 of the FAA, which makes arbitration clauses specifically enforceable as a matter of federal law.[303] Conversely, the multiplier is “enforcement-neutral”: litigants cannot use it to escape an obligation to arbitrate.[304] Enforcement-neutral laws are “not subject to preemption by the FAA.”[305]

Second, unlike the preempted state doctrines, the multiplier does not violate the spirit of the FAA. State regulation that seeks to protect plaintiffs from arbitration exhibits the very hostility to the process that Congress tried to eradicate by passing the statute.[306] Yet the multiplier does not regard arbitration with disfavor. To the contrary, it emphatically seeks to encourage arbitration. Thus, it would exist in harmony with the federal statute.[307]

Conclusion

For the last decade, federal arbitration law has been in flux. After Concepcion, lawmakers, agencies, judges, scholars, litigants, and journalists have weighed in on the future of forced arbitration. Yet these discussions have largely taken place in an empirical vacuum. We have tried to fill this void by surveying 40,775 arbitrations filed over the course of six years in four major institutions. Contrary to assertions by businesses and their allies, we find that pro se plaintiffs have not responded to the abolition of the class action by arbitrating their own claims. In addition, we discover that arbitration favors both repeat-playing defendants and repeat-playing plaintiffs’ lawyers. To compensate for the elimination of the class device and level the playing field between individuals and arbitration-savvy corporations, we propose that states create rewards for lawyers who guide consumers, employees, and medical patients to victory in the extrajudicial forum. As Justice Scalia explained in Concepcion, “the FAA was designed to promote arbitration.”[308] Creating incentives for plaintiffs’ lawyers to arbitrate is both good policy and dovetails with this master objective.

 

Andrea Cann Chandrasekher: Professor of Law, University of California, Davis, School of Law.

David Horton: Professor of Law and Chancellor’s Fellow, University of California, Davis, School of Law. Thanks to participants at the fourth annual Civil Procedure Workshop at Stanford Law School for helpful comments. Arbitration Nation is also the name of a popular blog run by Liz Kramer. We thank her for allowing us to use the name.

 

                    DOI: https://doi.org/10.15779/Z384F1MJ54

Copyright © 2019 California Law Review, Inc. California Law Review, Inc. (CLR) is a California nonprofit corporation. CLR and the authors are solely responsible for the content of their publications.

 

           [1].     See 138 S. Ct. 1612, 1632 (2018).

           [2].     See, e.g., Parties’ Joint Case Management Statement at 1, Ho v. Ernst & Young LLP, No. 05CV04867 (N.D. Cal. Mar. 11, 2009), 2009 WL 3662368 (mentioning several pending class actions that allege that the company incorrectly classified some employees as exempt from certain labor statutes).

           [3].     See, e.g., Reply Memorandum of Law in Support of Ernst & Young LLP’s Motion to Dismiss, or in the Alternative, Stay Proceedings and Compel Arbitration at 15, Sutherland v. Ernst & Young LLP, No. 10-CV-3332 (KMW) (S.D.N.Y. Aug. 20, 2010), 2010 WL 4022029.

           [4].     See, e.g., Sutherland v. Ernst & Young LLP, 768 F. Supp. 2d 547, 548 (S.D.N.Y. 2011); Morris v. Ernst & Young LLP, No. C-12-04964 RMW (N.D. Cal. July 9, 2013), 2013 WL 3460052, at *2.

           [5].     See, e.g., Sutherland, 768 F. Supp. 2d at 548.

           [6].     See, e.g., Defendants Ernst & Young LLP’s and Ernst & Young U.S., LLP’s Motion to Dismiss, or in the Alternative, Stay Proceedings and Compel Arbitration at 12, Morris v. Ernst & Young LLP, No. 12-cv-04964-RMW (HRL) (N.D. Cal. Jan. 11, 2013), 2013 WL 12190590.

           [7].     See Sutherland, 768 F. Supp. 2d at 548.

           [8].     See id. at 552 (“Sutherland would be required to spend approximately $200,000 in order to recover double her overtime loss of approximately $1,867.02.”).

           [9].     Id. (quoting Carnegie v. Household Intern., Inc., 376 F.3d 656, 661 (7th Cir. 2004)).

         [10].     563 U.S. 333 (2011).

         [11].     See id. at 351 (holding that the FAA preempts a California rule that deemed most class arbitration waivers in consumer contracts to be unconscionable).

         [12].     See D.R. Horton, Inc., 357 N.L.R.B. 2277, 2288 (Jan. 3, 2012) (holding that the employer’s forced arbitration agreement violated the NLRA by leading employees reasonably to believe they cannot file charges with the NLRB); see also Murphy Oil USA, Inc., 361 N.L.R.B. 774, 778 (Oct. 28, 2014) (reaffirming D.R. Horton).

         [13].     See, e.g., D.R. Horton, Inc. v. NLRB, 737 F.3d 344, 362 (5th Cir. 2013) (disagreeing with the NLRB); Owen v. Bristol Care, Inc., 702 F.3d 1050, 1053–55 (8th Cir. 2013) (same); Iskanian v. CLS Transp. Los Angeles, LLC, 327 P.3d 129, 142 (Cal. 2014) (same). But see NLRB v. Alternative Entm’t, Inc., 858 F.3d 393, 408 (6th Cir. 2017) (affirming the NLRB’s logic); Lewis v. Epic Sys. Corp., 823 F.3d 1147, 1159 (7th Cir. 2016) (same).

         [14].     See Sutherland v. Ernst & Young LLP, 726 F.3d 290, 297 n.8 (2d Cir. 2013); Morris v. Ernst & Young LLP, No. C-12-04964 RMW (N.D. Cal. July 9, 2013), 2013 WL 3460052, at *6.

         [15].     See Morris v. Ernst & Young, LLP, 834 F.3d 975, 983 (9th Cir. 2016).

         [16].     See Epic Sys. Corp. v. Lewis, 137 S. Ct. 809 (2017).

         [17].     See Epic Sys. Corp. v. Lewis, 138 S. Ct. 1612, 1632 (2018).

         [18].     Id. at 1619.

         [19].     Id. at 1632.

         [20].     Brief of the Chamber of Commerce of the United States of America as Amicus Curiae Supporting Petitioners in Nos. 16-285 and 16-300 and Respondents in No. 16-307 at *32–33, Epic Sys. Corp. v. Lewis, Nos. 16-285, 16-300, 16-307 (U.S. June 16, 2017), 2017 WL 2644688 (quoting Theodore J. St. Antoine, Labor and Employment Arbitration Today: Mid-Life Crisis or New Golden Age?, 32 Ohio St. J. on Disp. Resol. 1, 15 (2017)).

         [21].     Id. at *4.

         [22].     See Brief of HR Policy Association as Amicus Curiae in Support of Employer Petitioners and Respondent at *4–5, Epic Sys. Corp. v. Lewis, Nos. 16-285, 16-300, 16-307 (U.S. May 16, 2017), 2017 WL 2705659 (“Arbitration . . . provides a system in which employers generally have a diminished ability to use dispositive motions prior to an evidentiary hearing and where procedural defenses, such as limitations and jurisdiction, tend to be less likely to be accepted.”).

         [23].     Brief for Susan Fowler as Amicus Curiae in Support of Respondents in Nos. 16-285 and 16-300, and Petitioner in No. 16-307 at *3–4, Epic Sys. Corp. v. Lewis, Nos. 16-285, 16-300, 16-307 (U.S. Aug. 16, 2017), 2017 WL 4325881.

         [24].     See Brief of the Main Street Alliance, the American Sustainable Business Council, and Nick Hanauer as Amici Curiae in Support of Respondents in Nos. 16-285 & 16-300, and of Petitioner in No. 16-307 at *22, Epic Sys. Corp. v. Lewis, Nos. 16-285, 16-300, 16-307 (U.S. Aug. 16, 2017), 2017 WL 3588730.

         [25].     Id. at *21.

         [26].     See, e.g., Kindred Nursing Ctrs. Ltd. P’ship v. Clark, 137 S. Ct. 1421 (2017); DIRECTV, Inc. v. Imburgia, 136 S. Ct. 463 (2015); BG Grp. PLC v. Republic of Argentina, 572 U.S. 25 (2014); Am. Express Co. v. Italian Colors Rest., 570 U.S. 228 (2013); Oxford Health Plans LLC v. Sutter, 569 U.S. 564 (2013); Nitro-Lift Techs., LLC v. Howard, 568 U.S. 17 (2012); Marmet Health Care Ctr., Inc. v. Brown, 565 U.S. 530 (2012); CompuCredit Corp. v. Greenwood, 565 U.S. 95 (2012); KPMG LLP v. Cocchi, 565 U.S. 18 (2011); AT&T Mobility LLC v. Concepcion, 563 U.S. 333 (2011); Granite Rock Co. v. Int’l Bhd. of Teamsters, 561 U.S. 287 (2010); Rent-A-Center, W., Inc. v. Jackson, 561 U.S. 63 (2010); Stolt-Nielsen S.A. v. AnimalFeeds Int’l Corp., 559 U.S. 662 (2010).

         [27].     The Court revisited the issue of class arbitration waivers twice after Concepcion and before Epic Systems. See Am. Express, 570 U.S. at 238 (extending Concepcion’s mandate that judges enforce class arbitration waivers from the unconscionability context to a similar federal common law rule known as the “vindication of statutory rights” doctrine); DIRECTV, 136 S. Ct. at 471 (holding that the FAA preempted a state appellate court’s interpretation of a choice-of-law clause that led to the invalidation of a class arbitration waiver).

         [28].     See Brown ex rel. Brown v. Genesis Healthcare Corp., 724 S.E.2d 250, 293 (W. Va. 2011) (“[A]rbitration clauses in nursing home admission agreements . . . cannot be enforced to compel arbitration of a later negligence action against the nursing home . . . .”), cert. granted, judgment vacated sub nom. Marmet Health Care Ctr., Inc. v. Brown, 565 U.S. 530 (2012).

         [29].     See Extendicare Homes, Inc. v. Whisman, 478 S.W.3d 306, 328 (Ky. 2015) (“[T]he power to waive generally such fundamental constitutional rights must be unambiguously expressed in the text of the power-of-attorney document in order for that authority to be vested in the attorney-in-fact . . . .”), as corrected (Oct. 9, 2015), judgment rev’d in part, vacated in part sub nom. Kindred Nursing Ctrs. Ltd. P’ship v. Clark, 137 S. Ct. 1421 (2017).

         [30].     Brief for AARP, AARP Foundation, Justice in Aging, the National Consumer Voice for Quality Long-Term Care, and Nursing Home Ombudsman Agency of the Bluegrass as Amici Curiae in Support of Respondents at *19, Kindred Nursing Ctrs. Ltd. P’ship v. Clark, 137 S. Ct. 1421 (2017) (No. 16-32), 2017 WL 345127 (arguing that arbitration is not necessarily suitable “to resolve medical malpractice claims”).

         [31].     Kindred Nursing Ctrs. Ltd. P’ship, 137 S. Ct. at 1426; Marmet Health Care Ctr. 565 U.S. at 532 (“The [FAA’s] text includes no exception for personal-injury or wrongful-death claims.”).

         [32].     Jessica Silver-Greenberg & Robert Gebeloff, Arbitration Everywhere, Stacking the Deck of Justice, N.Y. Times (Oct. 31, 2015), https://www.nytimes.com/2015/11/01/business/dealbook/arbitration-everywhere-stacking-the-deck-of-justice.html [https://perma.cc/RJ2A-NG89]; Jessica Silver-Greenberg & Michael Corkery, In Arbitration, a ‘Privatization of the Justice System, N.Y. Times (Nov. 1, 2015), https://www.nytimes.com/2015/11/02/business/dealbook/in-arbitration-a-privatization-of-the-justice-system.html [https://perma.cc/PB4P-L6PJ]; Michael Corkery & Jessica Silver-Greenberg, In Religious Arbitration, Scripture is the Rule of Law, N.Y. Times (Nov. 2, 2015), https://www.nytimes.com/2015/11/03/business/dealbook/in-religious-arbitration-scripture-is-the-rule-of-law.html [https://perma.cc/6JXB-VVAE].

         [33].     Sarah Rudolph Cole, Blurred Lines: Are Non-Attorneys Who Represent Parties in Arbitrations Involving Statutory Claims Practicing Law?, 48 U.C. Davis L. Rev. 921, 923 (2015).

         [34].     See, e.g., Aaron Blumenthal, Comment, Circumventing Concepcion: Conceptualizing Innovative Strategies to Ensure the Enforcement of Consumer Protection Laws in the Age of the Inviolable Class Action Waiver, 103 Calif. L. Rev. 699 (2015); Andrew D. Bradt, Resolving Intrastate Conflicts of Laws: The Example of the Federal Arbitration Act, 92 Wash. U.L. Rev. 603 (2015); Brian T. Fitzpatrick, The End of Class Actions?, 57 Ariz. L. Rev. 161 (2015); Myriam Gilles & Gary Friedman, After Class: Aggregate Litigation in the Wake of AT&T Mobility v. Concepcion, 79 U. Chi. L. Rev. 623 (2012); Myriam Gilles, Class Warfare: The Disappearance of Low-Income Litigants from the Civil Docket, 65 Emory L.J. 1531 (2016); J. Maria Glover, Disappearing Claims and the Erosion of Substantive Law, 124 Yale L.J. 3052 (2015); David Horton, Federal Arbitration Act Preemption, Purposivism, and State Public Policy, 101 Geo. L.J. 1217 (2013); Christopher R. Leslie, The Arbitration Bootstrap, 94 Tex. L. Rev. 265 (2015); David L. Noll, Rethinking Anti–Aggregation Doctrine, 88 Notre Dame L. Rev. 649 (2012); Judith Resnik, Comment, Fairness in Numbers: A Comment on AT&T v. Concepcion, Wal-Mart v. Dukes, and Turner v. Rogers, 125 Harv. L. Rev. 78, 133 (2011); Maureen A. Weston, The Clash: Squaring Mandatory Arbitration with Administrative Agency and Representative Recourse, 89 S. Cal. L. Rev. 103, 116 (2015).

         [35].     See 34 C.F.R. § 685.300(e)–(f) (2018) (prohibiting schools that receive Title IV assistance from using forced arbitration clauses with their students).

         [36].     See Exec. Order No. 13,673 § 6(a), 3 C.F.R. §§ 283, 289–90 (2018), reprinted as amended in 41 U.S.C. § 3101 app. at 68, 70–71 (2012) [hereinafter EO 13,673] (prohibiting federal agencies from entering into certain contracts with companies that mandate arbitration for claims of sexual assault or harassment).

         [37].     See Definition of “Fiduciary,” 81 Fed. Reg. 20,946 (Apr. 8, 2016), amended by 82 Fed. Reg. 16,902, 16,905 (Apr. 7, 2017) (regulating class arbitration waivers in their contracts with investors).

         [38].     See Jessica Silver-Greenberg & Michael Corkery, U.S. Just Made it a Lot Less Difficult to Sue Nursing Homes, N.Y. Times (Sept. 28, 2016), https://www.nytimes.com/2016/09/29/business/dealbook/arbitration-nursing-homes-elder-abuse-harassment-claims.html [https://perma.cc/EN3U-REXZ].

         [39].     See Arbitration Agreements, 82 Fed. Reg. 33,210 (July 19, 2017).

         [40].     See Exec. Order No. 13,782, § 1, 82 Fed. Reg. 15,607 (Mar. 27, 2017).

         [41].     See Student Assistance General Provisions, 82 Fed. Reg. 49,155–56 (proposed Oct. 24, 2017); Carmen Castro-Pagan, DOL Defends Fiduciary Rule, Drops Anti-Arbitration Condition, Bloomberg News (July 5, 2017), https://www.bna.com/dol-defends-fiduciary-n73014461213 [https://perma.cc/S4YR-N4QN].

         [42].     Act of Nov. 1, 2017, Pub. L. No. 115-74, 131 Stat. 1243.

         [43].     Myriam Gilles, The Day Doctrine Died: Private Arbitration and the End of Law, 2016 U. Ill. L. Rev. 371, 409 (2016).

         [44].     See generally Consumer Financial Protection Bureau, Arbitration Study § 5.6 (2015) [hereinafter “CFPB Study”], http://files.consumerfinance.gov/f/201503_cfpb_arbitration-study-report-to-congress-2015.pdf [https://perma.cc/UQ43-YHP9] (reviewing 1,060 arbitrations filed by consumers against financial services companies in the AAA in 2010 and 2011); Mary Batcher, Amy Brachio, & Mridula Shrestha, Ernst & Young, Outcomes of Arbitration: An Empirical Study of Consumer Lending Cases 2 (2005) [hereinafter “E&Y Study”] (examining 226 cases filed between 2000 and 2004 in the National Arbitral Forum (NAF)); Searle Civil Justice Inst., Consumer Arbitration: Before the American Arbitration Association (2009) [hereinafter “Searle Report”] (analyzing 301 awards issued by the AAA between April 2007 and December 2007); Lisa B. Bingham, Employment Arbitration: The Repeat Player Effect, 1 Emp. Rts. & Emp. Pol’y J. 189 (1997) [hereinafter Bingham, Repeat Player Effect]; Lisa B. Bingham, On Repeat Players, Adhesive Contracts, and the Use of Statistics in Judicial Review of Employment Arbitration Awards, 29 McGeorge L. Rev. 223 (1998) [hereinafter Bingham, Statistics] (examining 203-case sample of arbitration awards under the AAA Employment Dispute Resolution Rules between January 1993 and December 1995); Alexander J. S. Colvin, An Empirical Study of Employment Arbitration: Case Outcomes and Processes, 8 J. Empirical Legal Stud. 1, 4 (2011) (surveying 3,945 employment arbitrations from the AAA that were filed between 2003 and 2007); Alexander J. S. Colvin & Mark D. Gough, Individual Employment Rights Arbitration in the United States: Actors and Outcomes, 68 Indus. & Lab. Rel. Rev. 1019, 1026–35 (2015) (considering 10,335 filings and 2,802 awards in AAA employment arbitrations from between 2003 and 2013).

         [45].     See supra note 44.

         [46].     The two exceptions are Colvin, supra note 44, at 17–18 (using a logit regression to examine the impact of several variables on the odds of an employee win) and Colvin & Gough, supra note 44, at 1026–35 (expanding on Colvin’s first piece by using more data and a different analytical approach). We say more about these articles infra Part I.B.

         [47].     See David Horton & Andrea Cann Chandrasekher, After the Revolution: An Empirical Study of Consumer Arbitration, 104 Geo. L.J. 57 (2015) [hereinafter Horton & Chandrasekher, Consumer Arbitration]. We followed Consumer Arbitration with David Horton & Andrea Cann Chandrasekher, Employment Arbitration After the Revolution, 65 DePaul L. Rev. 457 (2016) [hereinafter Horton & Chandrasekher, Employment Arbitration], which we wrote for the Twenty-First Annual Clifford Symposium: The Supreme Court, Business and Civil Justice, hosted by DePaul Law School.

         [48].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 63.

         [49].     See Horton & Chandrasekher, Employment Arbitration, supra note 47, at 461–62.

         [50].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110–11; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487–88.

         [51].     See supra note 44.

         [52].     See infra text accompanying note 101.

         [53].     See supra text accompanying note 32.

         [54].     See infra Part III.A.3.

         [55].     We used the phrase “arbitration entrepreneurs” in our After the Revolution papers to describe “plaintiffs’ lawyers who have tried to pursue numerous arbitrations against single defendants.” Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 63 n.38. Our previous work only found arbitration entrepreneurs in the consumer setting. See Horton & Chandrasekher, Employment Arbitration, supra note 47, at 471 (“[P]laintiffs’ employment lawyers do not string together related cases like consumer attorneys . . . .”). However, as we discuss in greater depth infra Part II.B, one of our most important discoveries in this Article is that the trend has now spread to employment cases as well.

         [56].     See supra text accompanying note 31.

         [57].     Aaron-Andrew P. Bruhl, The Unconscionability Game: Strategic Judging and the Evolution of Federal Arbitration Law, 83 N.Y.U. L. Rev. 1420, 1429 (2008).

         [58].     Pub. L. No. 68-401, 43 Stat. 883 (1925) (codified as amended at 9 U.S.C. §§ 1–14 (2012)).

         [59].     See, e.g., Kill v. Hollister, (1746) 95 Eng. Rep. 532 (K.B.) 532 (opining that “the agreement of the parties cannot oust this [c]ourt”); Vynior’s Case, (1609) 77 Eng. Rep. 597 (K.B.) 599 (refusing to specifically enforce an arbitration provision over a party’s objection).

         [60].     9 U.S.C. § 2 (2012).

         [61].     H.R. Rep. No. 68-96, at 1 (1924).

         [62].     S. Rep. No. 68-536, at 3 (1924).

         [63].     See, e.g., Frances Kellor, American Arbitration: Its History, Functions and Achievements 15–17 (1948) (describing how two prominent arbitration societies merged to create the AAA); Amalia D. Kessler, Arbitration and Americanization: The Paternalism of Progressive Procedural Reform, 124 Yale L.J. 2940, 2949–51 (2015) (explaining how advocates for the FAA eventually formed the AAA).

         [64].     See, e.g., Soia Mentschikoff, Commercial Arbitration, 61 Colum. L. Rev. 846, 857 (1961) (noting that trade associations either discouraged or prohibited attorney involvement in arbitration); Katherine Van Wezel Stone, Rustic Justice: Community and Coercion Under the Federal Arbitration Act, 77 N.C. L. Rev. 931, 970–71 (1999) (“The merchant guilds established arbitration tribunals because they felt that the courts were not sufficiently knowledgeable about commercial customs . . . .”).

         [65].     See, e.g., Kessler, supra note 63, at 2984 (noting that the AAA broke new ground by adopting “procedures and structures that transcended any particular dispute and any given professional or trade association”).

         [66].     See, e.g., Mentschikoff, supra note 64, at 857–59.

         [67].     Even the FAA’s draftsman, Julius Henry Cohen, cautioned that arbitration was “not the proper method for deciding points of law of major importance involving constitutional questions or policy in the application of statutes.” Julius Henry Cohen & Kenneth Dayton, The New Federal Arbitration Law, 12 Va. L. Rev. 265, 281 (1926) (commenting in addition that arbitration is “peculiarly suited to the disposition of the ordinary disputes between merchants as to questions of fact—quantity, quality, time of delivery, compliance with terms of payment, excuses for non-performance, and the like”).

         [68].     See Note, Predictability of Result in Commercial Arbitration, 61 Harv. L. Rev. 1022, 1026 (1948).

         [69].     Mentschikoff, supra note 64, at 861.

         [70].     Wilko v. Swan, 107 F. Supp. 75, 78 (S.D.N.Y. 1952), rev’d, 201 F.2d 439 (2d Cir. 1953), rev’d, 346 U.S. 427 (1953) (quoting the Handbook and Guide to Arbitration Under the New York and United States Arbitration Statutes, Chamber of Commerce of the State of New York (1932)).

         [71].     See, e.g., David Shieh, Note, Unintended Side Effects: Arbitration and the Deterrence of Medical Error, 89 N.Y.U. L. Rev. 1806, 1826–27 (2014).

         [72].     See, e.g., Anne S. Kim, Rent-a-Judges and the Cost of Selling Justice, 44 Duke L.J. 166, 173–75 (1994) (describing JAMS’s founding); Christopher R. Drahozal & Samantha Zyontz, Private Regulation of Consumer Arbitration, 79 Tenn. L. Rev. 289, 297 (2012) (noting that, unlike the AAA, JAMS is a “for-profit business”).

         [73].     See JAMS Fact Sheet, JAMS, http://www.jamsadr.com/files/Uploads/Documents/Corporate-Fact-Sheet.pdf [https://perma.cc/U5VJ-Z4JB] (last visited Dec. 21, 2018) (“JAMS panelists resolve multi-party, complex cases . . . .”).

         [74].     See Kim, supra note 72, at 175 (“Over the past decade, the number of retired judges for hire has increased exponentially.”).

         [75].     Warren E. Burger, Isn’t There a Better Way?, 68 A.B.A. J. 274, 275 (1982)

         [76].     See, e.g., Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 626 (1985) (compelling arbitration of claim under the Sherman Act); Gilmer v. Interstate/Johnson Lane Corp., 500 U.S. 20, 27 (1991) (compelling arbitration of claim under the Age Discrimination in Employment Act).

         [77].     Mitsubishi Motors, 473 U.S. at 628. The Court has repeated this exact quote several times. See, e.g., 14 Penn Plaza LLC v. Pyett, 556 U.S. 247, 266 (2009); Preston v. Ferrer, 552 U.S. 346, 359 (2008); Gilmer, 500 U.S. at 31; Circuit City Stores, Inc. v. Adams, 532 U.S. 105, 123–24 (2001); Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 481 (1989); Shearson/Am. Express, Inc. v. McMahon, 482 U.S. 220, 229–30 (1987).

         [78].     465 U.S. 1, 11 (1984) (“We see nothing in the [FAA] indicating that the broad principle of enforceability is subject to any additional limitations under State law.”).

         [79].     See Doctor’s Assocs. v. Casarotto, 517 U.S. 681, 687 (1996).

         [80].     See Perry v. Thomas, 482 U.S. 483, 491–92 (1987).

         [81].     See Allied-Bruce Terminix Cos. v. Dobson, 513 U.S. 265, 272 (1995).

         [82].     Doctor’s Assocs., 517 U.S. at 687.

         [83].     Id.

         [84].     See, e.g., Hill v. Gateway 2000, Inc., 105 F.3d 1147, 1150 (7th Cir. 1997) (holding that consumers were bound by an arbitration clause that arrived in the box with their computer); Badie v. Bank of Am., 67 Cal. App. 4th 779, 783 (Ct. App. 1998) (involving a bank that mailed its customers an arbitration clause in a “bill stuffer”); Michael H. LeRoy & Peter Feuille, Judicial Enforcement of Predispute Arbitration Agreements: Back to the Future, 18 Ohio St. J. on Disp. Resol. 249, 305 (2003) (“Mandatory arbitration quickly spread in the 1990s . . . to a wide variety of workplaces, including those with Spanish-speaking, low-wage, poorly educated, and minor workers.”).

         [85].     ADR Services was founded by Lucie Barron, who is a psychologist, not a lawyer. See About Lucie Baron, ADR Services, Inc., http://www.adrservices.com/about/about-lucie-barron [https://perma.cc/MJ6P-K7FZ] (last visited Dec. 21, 2018).

         [86].     See About ARC, Alternative Resolution Centers, http://www.arc4adr.com/about.html [https://perma.cc/ZN2N-UANL] (last visited Dec. 21, 2018).

         [87].     See, e.g., Thomas E. Carbonneau, Law and Practice of Arbitration 105 (5th ed. 2014) (describing the origins of NAF).

         [88].     See, e.g., David Sherwyn, J. Bruce Tracey, & Zev J. Eigen, In Defense of Mandatory Arbitration of Employment Disputes: Saving the Baby, Tossing Out the Bath Water, and Constructing a New Sink in the Process, 2 U. Pa. J. Lab. & Emp. L. 73, 105 (1999) (“[A]rbitration provide[s] the parties with an inexpensive, confidential, and fast resolution . . . .”); Stephen J. Ware, Paying the Price of Process: Judicial Regulation of Consumer Arbitration Agreements, 2001 J. Disp. Resol. 89, 89 (2001) (“[A]rbitration provides opportunities for a business to save on its dispute-resolution costs . . . .”).

         [89].     513 U.S. 265, 280 (1995); see also Peter B. Rutledge, U.S. Chamber Inst. for Legal Reform, Arbitration—A Good Deal for Consumers 6 (2008) (“By streamlining the dispute resolution process and reducing the costs associated with it, arbitration makes it easier for individuals to find an attorney willing to take their case or, alternatively, to represent themselves.”).

         [90].     See generally Paul D. Carrington & Paul H. Haagen, Contract and Jurisdiction, 1996 Sup. Ct. Rev. 331 (1996) (condemning the Supreme Court’s deleterious construction of arbitration law); 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 (1997) (criticizing the Supreme Court’s broad endorsement of adhesive arbitration agreements.); Jean R. Sternlight, Panacea or Corporate Tool?: Debunking the Supreme Court’s Preference for Binding Arbitration, 74 Wash U. L.Q. 637 (1996) (urging the Supreme Court to apply the FAA as Congress originally intended and to reject unfair binding arbitration).

         [91].     This dichotomy between “one-shotters” and “repeat players” originally appeared in Marc Galanter, Why the “Haves” Come Out Ahead: Speculations on the Limits of Legal Change, 9 L. & Soc’y Rev. 95, 97–101 (1974).

         [92].     See, e.g., Schwartz, supra note 90, at 60–61 (“[I]ndividual arbitrators have an economic stake in being selected again, and their judgment may well be shaded by a desire to build a ‘track record’ of decisions that corporate repeat-users will view approvingly.”); Sternlight, supra note 90, at 685 (“An arbitrator who issues a large punitive damages award against a company may not get chosen again by that company or others who hear of the award.”).

         [93].     Stroh Container Co. v. Delphi Indus., Inc., 783 F.2d 743, 751 n.12 (8th Cir. 1986).

         [94].     See generally Eugene F. Lynch et al., Blue Ribbon Advisory Panel on Kaiser Permanente Arbitration, The Kaiser Permanente Arbitration System: A Review and Recommendations for Improvement (1998), http://www.oia-kaiserarb.com/pdfs/BRP-Report.pdf [https://perma.cc/S2TG-NHN7] (proposing recommendations to align Kaiser’s structure, processes, and operations to achieve the goals of the Kaiser Permanente arbitration program). Kaiser took this measure after the California Supreme Court had held that Kaiser’s arbitration regime was such a sham that the HMO had committed fraud by representing that it was “a fair approach to protecting . . . rights.” Engalla v. Permanente Med. Grp., Inc., 938 P.2d 903, 910 (Cal. 1997). Kaiser then implemented the Panel’s recommendations by creating a sovereign administrative body (the OIA) to shepherd disputes through the process. See Office of the Independent Administrator, Kaiser, http://www.oia-kaiserarb.com/24/rules-and-info/information-about-the-arbitration-system [https://perma.cc/TW7M-67RB] (last visited Dec. 21, 2018).

         [95].     See Rules for Kaiser Permanente Member Arbitrations, Office of the Indep. Adm’r 5, 15 (2016), http://www.oia-kaiserarb.com/pdfs/Rules.pdf [https://perma.cc/BC2B-K4WA].

         [96].     Reynolds Holding, Arbitration Reform in Works: Assembly Panel to Propose Broad Changes in Maligned System, S.F. Chron. (Mar. 11, 2002), https://www.sfgate.com/news/article/Arbitration-reform-in-works-Assembly-panel-to-2867174.php [https://perma.cc/V39W-RSGG].

         [97].     See Consumer Due Process Protocol Statement of Principles, Am. Arb. Ass’n 1 (Apr. 17, 1998) [hereinafter AAA Consumer Due Process Protocol], https://www.adr.org/sites/default/files/document_repository/Consumer Due Process Protocol (1).pdf [https://perma.cc/UMU3-9X57]; Employment Due Process Protocol, Am. Arb. Ass’n (May 9, 1995) [hereinafter AAA Employment Due Process Protocol], https://www.adr.org/sites/default/files/document_repository/Employment Due Process Protocol_0.pdf [https://perma.cc/SJ7H-6ELL]; Consumer Arbitration Minimum Standards, JAMS (2009) [hereinafter JAMS Consumer Minimum Standards], https://www.jamsadr.com/consumer-minimum-standards [https://perma.cc/U5VJ-Z4JB]; Employment Arbitration Minimum Standards of Procedural Fairness, JAMS (2009) [hereinafter JAMS Employment Minimum Standards], https://www.jamsadr.com/employment-minimum-standards [https://perma.cc/KU8Q-LFPE]. For more on the origins of the AAA rules, see Drahozal & Zyontz, supra note 72, at 301–05.

         [98].     See, e.g., Am. Arb. Ass’n, Consumer Arbitration Rules 18, 33–34 (2014), https://www.adr.org/sites/default/files/Consumer Rules.pdf [https://perma.cc/6V7N-7RPJ]; Am. Arb. Ass’n, Employment Arbitration Rules and Mediation Procedures 14 (2017), https://www.adr.org/sites/default/files/employment_arbitration_rules_and_mediation_procedures_0.pdf [https://perma.cc/55R3-W84K]; JAMS, Streamlined Arbitration Rules & Procedures, https://www.jamsadr.com/rules-streamlined-arbitration [https://perma.cc/M66Y-7YJL]; JAMS, Employment Arbitration Rules & Procedures, https://www.jamsadr.com/rules-employment-arbitration/#fifteen [https://perma.cc/7ZKU-LLZK].

         [99].     See, e.g., Myriam Gilles, Opting Out of Liability: The Forthcoming, Near-Total Demise of the Modern Class Action, 104 Mich. L. Rev. 373, 375 (2005) (“[T]he waiver works in tandem with standard arbitration provisions to ensure that any claim against the corporate defendant may be asserted only in a one-on-one, nonaggregated arbitral proceeding.”); Jean R. Sternlight, As Mandatory Binding Arbitration Meets the Class Action, Will the Class Action Survive?, 42 Wm. & Mary L. Rev. 1, 6 (2000) (“Increasingly, potential defendants are drafting arbitration clauses that explicitly bar class actions . . . .”).

      [100].     113 P.3d 1100, 1110 (Cal. 2005).

      [101].     Id. at 1108 (quoting Szetela v. Discover Bank, 118 Cal. Rptr. 2d 862 (Ct. App. 2002)).

      [102].     See Kinkel v. Cingular Wireless LLC, 857 N.E.2d 250, 275 (Ill. 2006) (“[P]laintiff’s only reasonable, cost-effective means of obtaining a complete remedy [was] as either the representative or a member of a class.”); see also Lowden v. T-Mobile USA, Inc., 512 F.3d 1213, 1218–19 (9th Cir. 2008); Dale v. Comcast Corp., 498 F.3d 1216, 1224 (11th Cir. 2007) (applying Georgia law); Cooper v. QC Fin. Servs., Inc., 503 F. Supp. 2d 1266, 1285–90 (D. Ariz. 2007); Lozada v. Dale Baker Oldsmobile, Inc., 91 F. Supp. 2d 1087, 1105 (W.D. Mich. 2000); Leonard v. Terminix Int’l Co., 854 So. 2d 529, 535–39 (Ala. 2002); Powertel, Inc. v. Bexley, 743 So. 2d 570, 576 (Fla. Dist. Ct. App. 1999); Brewer v. Mo. Title Loans, Inc., 323 S.W.3d 18, 22–23 (Mo. 2010) (en banc), vacated, 563 U.S. 971 (2011); Fiser v. Dell Comput. Corp., 188 P.3d 1215, 1221 (N.M. 2008); Tillman v. Commercial Credit Loans, Inc., 655 S.E.2d 362, 373 (N.C. 2008); Vasquez-Lopez v. Beneficial Or., Inc., 152 P.3d 940, 949–51 (Or. Ct. App. 2007); Thibodeau v. Comcast Corp., 912 A.2d 874, 885–86 (Pa. Super. Ct. 2006); Herron v. Century BMW, 693 S.E.2d 394, 399–400 (S.C. 2010); Scott v. Cingular Wireless, 161 P.3d 1000, 1006–08 (Wash. 2007) (en banc); State ex rel. Dunlap v. Berger, 567 S.E.2d 265, 278–80 (W. Va. 2002); Coady v. Cross Country Bank, Inc., 729 N.W.2d 732, 746–48 (Wis. Ct. App. 2007).

      [103].     See Wireless Customer Agreement, AT&T, http://www.att.com/legal/terms.wirelessCustomerAgreement.html [https://perma.cc/KE43-ZJNL] (last visited Dec. 21, 2018) (promising to pay $10,000, double attorneys’ fees, and “any expenses (including expert witness fees and costs) that your attorney reasonably accrues for investigating, preparing, and pursuing your claim in arbitration”); Myriam Gilles, Killing Them with Kindness: Examining “Consumer-Friendly” Arbitration Clauses After AT&T Mobility v. Concepcion, 88 Notre Dame L. Rev. 825, 857–58 n.131 (2012) (describing Sallie Mae and Verizon’s provisions and collecting other examples).

      [104].     AT&T Mobility’s Brief in Support of Its Motion to Compel Arbitration and to Dismiss Action at 13, Francis v. AT&T Mobility, LLC, No. 2:07-cv-14921 (E.D. Mich. Jan. 25, 2008), 2008 WL 393982.

      [105].     See, e.g., Laster v. T-Mobile USA, Inc., No. 05CV1167, 2008 WL 5216255, at *14 (S.D. Cal. Aug. 11, 2008); Hall v. AT&T Mobility LLC, 608 F. Supp. 2d 592, 603–04 (D.N.J. 2009); Stiener v. Apple Comput., Inc., 556 F. Supp. 2d 1016, 1033–34 (N.D. Cal. 2008).

      [106].     See generally 563 U.S. 333 (2011).

      [107].     Id. at 337.

      [108].     Laster, 2008 WL 5216255, at *14; Laster v. AT&T Mobility LLC, 584 F.3d 849, 855 (9th Cir. 2009).

      [109].     AT&T Mobility, 563 U.S. at 352. Chief Justice Roberts and Justices Alito and Kennedy joined the majority. Justice Thomas “reluctantly” signed on but also authored his own opinion. Id. at 353 (Thomas, J. concurring).

      [110].     Id. at 344 (majority opinion).

      [111].     See id. at 347–39.

      [112].     Id. at 351.

      [113].     See, e.g., Morvant v. P.F. Chang’s China Bistro, Inc., 870 F. Supp. 2d 831, 840 (N.D. Cal. 2012) (rejecting the argument that “Concepcion was in the consumer context” and thus did not govern “class action waivers in the employment context”).

      [114].     See, e.g., Am. Express Co. v. Italian Colors Rest., 570 U.S. 228, 238 (2013) (applying Concepcion’s logic to federal antitrust claims).

      [115].     See supra text accompanying notes 32–34.

      [116].     See generally D.R. Horton, Inc., 357 N.L.R.B. No. 184 (2012); see also Murphy Oil USA, Inc., 361 N.L.R.B. No. 72 (2014), at *3 (reaffirming D.R. Horton). The NLRA guarantees employees’ rights to engage in “concerted activities for the purpose of collective bargaining or other mutual aid or protection.” National Labor Relations Act § 7, 29 U.S.C. § 157 (2012).

      [117].     See Exec. Order No. 13,673, § 6(a), 3 C.F.R. 283, 289–90 (2018), reprinted as amended in 41 U.S.C. § 3101 app. at 68, 70–71 (2012).

      [118].     See, e.g., Silver-Greenberg & Corkery, Less Difficult to Sue Nursing Homes, supra note 38.

      [119].     See 34 C.F.R. § 685.300(e)–(f) (2018) (precluding forced arbitration for certain allegations against schools that receive Title IV assistance under the Higher Education Act); Press Release, U.S. Dep’t of Educ., Education Department Proposes New Regulations to Protect Students and Taxpayers from Predatory Institutions (June 13, 2016), https://www.ed.gov/news/press-releases/education-department-proposes-new-regulations-protect-students-and-taxpayers-predatory-institutions [https://perma.cc/59GY-58RA].

      [120].     See Definition of “Fiduciary,” 81 Fed. Reg. 20,946, at *20,959 (Apr. 8, 2016), amended by 82 Fed. Reg. 16,902 (Apr. 7, 2017) (“[P]rohibit[ing] pre-dispute binding arbitration agreements with respect to individual contract claims.”).

      [121].     See Arbitration Agreements, 81 Fed. Reg. 32,830 (proposed May 24, 2016); see also Jessica Silver-Greenberg & Michael Corkery, U.S. Agency Moves to Allow Class-Action Lawsuits Against Financial Firms, N.Y. Times (July 10, 2017), https://www.nytimes.com/2017/07/10/business/dealbook/class-action-lawsuits-finance-banks.html [https://perma.cc/3M5C-JMF3].

      [122].     See Exec. Order No. 13,782, § 1, 82 Fed. Reg. 15,607, 15,607 (Mar. 30, 2017).

      [123].     Dep’t of Health & Human Servs., Ctrs. for Medicare & Medicaid Servs., Revision of Requirements for Long-Term Care Facilities: Arbitration Agreements (2017), https://www.federalregister.gov/documents/2017/06/08/2017-11883/medicare-and-medicaid-programs-revision-of-requirements-for-long-term-care-facilities-arbitration [https://perma.cc/RTL9-JHKF].

      [124].     See Student Assistance General Provisions, 82 Fed. Reg. 49,155, 49,156 (proposed Oct. 24, 2017) (delaying implementation of the DoE’s rule targeting arbitration clauses in student enrollment contacts); Jacklyn Wille, DOL Again Signals Death of Fiduciary Rule’s Arbitration Ban, Bloomberg BNA (Aug. 24, 2017), https://www.bna.com/dol-again-signals-n73014463627 [https://perma.cc/7HC4-Q39R] (discussing how the DOL “may be working to undo portions of the Obama-era rule, including the anti-arbitration provision”).

      [125].     See Act of Nov. 1, 2017, Pub. L. No. 115-74, 131 Stat. 1243, 1243.

      [126].     See generally Bingham, Repeat Player Effect, supra note 44; Bingham, Statistics, supra note 44.

      [127].     Bingham, Repeat Player Effect, supra note 44, at 206.

      [128].     Id. at 202, 209–10.

      [129].     Id.

      [130].     Bingham, Statistics, supra note 44, at 236.

      [131].     See id. at 238.

      [132].     See id.

      [133].     Id.

      [134].     Id.

      [135].     See Bingham, Repeat Player Effect, supra note 44, at 213.

      [136].     See id.

      [137].     See id. Bingham eventually determined that employees won only 21% of cases stemming from a personnel manual, compared to 69% of matters that arose from a negotiated agreement. Bingham, Statistics, supra note 44, at 239.

      [138].     Bingham, Repeat Player Effect, supra note 44, at 214.

      [139].     Equal Employment Opportunity Commission, Policy Statement on Mandatory Binding Arbitration of Employment Discrimination Disputes as a Condition of Employment, 1997 WL 33159163, at *6 (July 10, 1997).

      [140].     Armendariz v. Found. Health Psychcare Servs., Inc., 6 P.3d 669, 690 (Cal. 2000).

      [141].     See Sherwyn et al., In Defense of Mandatory Arbitration of Employment Disputes, supra note 88, at 144; Peter B. Rutledge, Whither Arbitration?, 6 Geo. J.L. & Pub. Pol’y 549, 566 n.84 (2008).

      [142].     Samuel Estreicher, Saturns for Rickshaws: The Stakes in the Debate over Predispute Employment Arbitration Agreements, 16 Ohio St. J. on Disp. Resol. 559, 566 (2001); see also Elizabeth Hill, Due Process at Low Cost: An Empirical Study of Employment Arbitration Under the Auspices of the American Arbitration Association, 18 Ohio St. J. on Disp. Resol. 777, 807–08 (2003) (claiming that some employers benefit from an “appellate effect” caused by their internal dispute resolution programs, which settle cases that plaintiffs are likely to win before the arbitration stage).

      [143].     See generally Searle Report, supra note 44. To be clear, the Searle Report was not the first study of consumer arbitration. Previously, the California Dispute Resolution Institute (CDRI) reviewed 2,175 consumer and employment cases from the AAA, ArbitrationWorks, ADR Services, ARC Consumer Arbitration, JAMS, and Judicate West. See Cal. Disp. Resol. Inst., Consumer and Employment Arbitration in California 14 (2004), https://www.mediate.com/cdri/cdri_print_Aug_6.pdf [https://perma.cc/646R-5DFZ]. However, the CDRI was unable to draw many firm conclusions due to gaps and ambiguities in its data. See id. at 18. Likewise, in 2005, Ernst & Young analyzed 226 arbitrations filed against lenders between 2000 and 2004 in the National Arbitral Forum (NAF). See E&Y Study, supra note 44. E&Y determined that consumers “won” (defined as recovering $1 or more) fifty-three of the ninety-seven awards (55%). See id. at 9 & n.11. Nevertheless, the E&Y Study was funded by the American Bankers’ Association, see id. at 2, which raised questions about its objectivity.

      [144].     Searle Report, supra note 44, at 2. One advantage of the Searle Report is that it reviewed the actual underlying AAA case files, rather than relying on data published by the AAA. See id. at xii.

      [145].     Id. at 68.

      [146].     Id.

      [147].     Id. at 72–73.

      [148].     Id. at 76.

      [149].     Id.

      [150].     Id. at 76–77.

      [151].     Id. at 76.

      [152].     Id. at 78.

      [153].     Id. at 80.

      [154].     Id. at 81–82.

      [155].     Id. at 82.

      [156].     See generally Colvin, supra note 44.

      [157].     Id. at 4.

      [158].     Id. at 13.

      [159].     Id. at 13–14.

      [160].     Id. at 17–18.

      [161].     Id.

      [162].     Id.

      [163].     See generally Colvin & Gough, supra note 44.

      [164].     Id. at 1027. Colvin and Gough describe their dataset as consisting of cases both “filed and terminated” within this time period. Id.

      [165].     See id. at 1029–30.

      [166].     See id. at 1030.

      [167].     See id. at 1030–31.

      [168].     See id. at 1026, 1030.

      [169].     Id. at 1029, 1032–34. Colvin and Gough defined “settled” as cases that were not listed on the AAA spreadsheet as “stayed, permanently stayed, withdrawn, or otherwise disposed of before an award was rendered.” Id. at 1029.

      [170].     Id. at 1034.

      [171].     Id. at 1032.

      [172].     Id. at 1033.

      [173].     Id. at 1032.

      [174].     See id. at 1036.

      [175].     See generally CFPB Study, supra note 44.

      [176].     See id. § 1.4.3, at 11.

      [177].     See id. § 5.6, at 32–33. The CFPB did not include disputes that had been initiated in 2012 because many were still pending or were marred by incomplete information when the CFPB completed its data collection in 2013. See id. § 5.6, at 32.

      [178].     Id. § 5.6.6, at 39.

      [179].     Id at 40.

      [180].     Id. at 39.

      [181].     See generally Horton & Chandrasekher, Consumer Arbitration, supra note 47; Horton & Chandrasekher, Employment Arbitration, supra note 47.

      [182].     Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 63.

      [183].     Horton & Chandrasekher, Employment Arbitration, supra note 47, at 461–62.

      [184].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 94–96.

      [185].     See Horton & Chandrasekher, Employment Arbitration, supra note 47, at 471–72 (“[P]laintiffs’ employment lawyers do not string together related cases like consumer attorneys.”).

      [186].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 104; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487.

      [187].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487.

      [188].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487.

      [189].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487.

      [190].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 99, 110; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 476, 487.

      [191].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110–11; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487–88.

      [192].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110–11.

      [193].     See Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487–88.

      [194].     See supra text accompanying note 44. Although the CFPB examined arbitrations that had been filed as late as the end of 2012, it largely focused on awards that preceded the Court’s decision in April 2011. See supra text accompanying notes 175, 177. Cf. Judith Resnik, Diffusing Disputes: The Public in the Private of Arbitration, the Private in Courts, and the Erasure of Rights, 124 Yale L.J. 2804, 2814–15 (2015) (noting that disclosures by arbitration providers reveal that “almost no consumers or employees ‘do’ arbitration at all”).

      [195].     Ted Frank, Class Actions, Arbitration, and Consumer Rights: Why Concepcion is a Pro-Consumer Decision, 16 Legal Policy Report (Feb. 19, 2013), http://www.manhattan-institute.org/html/lpr_16.htm#.VOZvjPnF-So [https://perma.cc/KFM5-B7FW].

      [196].     Alan S. Kaplinsky & Mark J. Levin, Some Thoughts on the St. John’s School of Law’s Analysis of Consumer Understanding of Arbitration Agreements, Consumer Fin. Monitor (Nov. 10, 2014), http://www.cfpbmonitor.com/2014/11/10/some-thoughts-on-the-st-johns-school-of-laws-analysis-of-consumer-understanding-of-arbitration-agreements [https://perma.cc/43NE-V73E].

      [197].     Hans. A. von Spakovsky, The Unfair Attack on Arbitration: Harming Consumers by Eliminating a Proven Dispute Resolution System, 97 Legal Memorandum 1, 8 (2013); Brief for Amicus Curiae Cal. Emp’t Law Council in Support of Affirmance at 17, Johnmohammadi v. Bloomingdale’s, Inc., 755 F.3d 1072 (9th Cir. 2013) (No. 12-55578), 2012 WL 6801859, at *17 (“No company would willingly enter into arbitration agreements if the price were to require class arbitration . . . . That would harm employees, because for the most common employment disputes—individual claims too small to attract a contingent-fee lawyer—the choice is ‘arbitration—or nothing.’” (quoting Theodore J. St. Antoine, Mandatory Arbitration: Why It’s Better Than It Looks, 41 U. Mich. J.L. Reform 783, 792 (2008))) [hereinafter Cal. Emp’t Law Council Brief].

      [198].     Resnik, Fairness in Numbers, supra note 34, at 128 (quoting AT&T Mobility LLC v. Concepcion, 563 U.S. 333, 365 (2011) (Breyer, J., dissenting)) (emphasis removed); Jean R. Sternlight, Tsunami: AT&T Mobility LLC v. Concepcion Impedes Access to Justice, 90 Or. L. Rev. 703, 704–05 (2012) (“In many contexts, if plaintiffs cannot join together in a class action, lack of knowledge, lack of resources, or fear of retaliation will prevent them from bringing any claims at all.”); John Campbell, Mis-Concepcion: Why Cognitive Science Proves the Emperors Have No Robes, 79 Brook. L. Rev. 107, 143 (2013) (noting that to pursue their own claims, plaintiffs must skip work, travel, and deal with “the very limited potential reward for the effort spent”); see also supra text accompanying note 30.

      [199].     See supra text accompanying note 44.

      [200].     See supra text accompanying notes 71–72, 85–87.

      [201].     See supra text accompanying notes 71–74, 94–98.

      [202].     See supra text accompanying note 96; Am. Arb. Ass’n, Notice on Consumer Debt Collection Arbitrations, https://www.adr.org/sites/default/files/document_repository/Notice on Consumer Debt Collection Arbitrations (1).pdf [https://perma.cc/9ZVM-R8RF] (last visited Dec. 21, 2018) (explaining that the company will not handle “consumer debt collection programs or bulk filings and individual case filings in which the company is the filing party and the consumer has not agreed to arbitrate at the time of the dispute, and the case involves a credit card bill, a telecom bill or a consumer finance matter”).

      [203].     See Cal. Civ. Proc. Code § 1281.96(a) (2017).

      [204].     See Ethics Standards for Neutral Arbitrators in Contractual Arbitration, Standard 2(e)(4) (2003), http://www.courts.ca.gov/documents/ethics_standards_neutral_arbitrators.pdf. [https://perma.cc/YX4V-AQBR].

      [205].     See Cal. Civ. Proc. Code § 1281.96(a)(1)–(11).

      [206].     See Colvin, supra note 44, at 2–3; Colvin & Gough, supra note 44, at 1026; Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 88; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 463. Admittedly, the AAA records are not perfect. See, e.g., Judith Resnik, A2J/A2K: Access to Justice, Access to Knowledge, and Economic Inequalities in Open Courts and Arbitrations, 96 N.C. L. Rev. 605, 649 (2018) (detailing a coding error in the AAA disclosures that the institution subsequently corrected).

      [207].     See Am. Arb. Ass’n, Consumer Arbitration Statistics, https://www.adr.org/consumer [https://perma.cc/2WPR-R2ED] (last visited Dec. 21, 2018); JAMS, Consumer Case Information, http://www.jamsadr.com/consumercases [https://perma.cc/9H94-B6CH] (last visited Dec. 21, 2018); ADR Servs., Inc., Consumer Arbitration Disclosures, https://www.adrservices.com/services/consumer-case-information [https://perma.cc/FRX3-WJEP] (last visited Dec. 21, 2018). Several other prominent providers, including Judicate West and Resolution Remedies, merely display information without permitting users to download it. See Judicate W., Arbitration Disclosures, https://www.judicatewest.com/Resources/ConsumerArbDisclosures [https://perma.cc/2D7J-JDTW] (last visited Dec. 21, 2018); Nat’l Arb. & Mediation, NAM Disclosures for California Consumer Arbitrations, http://www.namadr.com/consumer-cases-california/ [https://perma.cc/E4BK-DWHE] (last visited Dec. 21, 2018).

      [208].     See supra text accompanying note 202.

      [209].     See Office of the Indep. Adm’r, Sortable Disclosure Table, http://www.oia-kaiserarb.com/51/consumer-case-information/disclosure-table-about-arbitrations-received-in-past-five-years-sortable/sortable-dislosure-table [https://perma.cc/43YW-73BL] (last visited Dec. 21, 2018).

      [210].     See Cal. Civ. Proc. Code § 1281.96(a).

      [211].     See supra text accompanying note 32.

      [212].     See Email from Robert Gebeloff to David Horton (Nov. 3, 2015) (on file with author).

      [213].     Not every institution included matters from before 2010. Likewise, because we were working from spreadsheets that had been released in late 2017, including results from 2017 might skew the data toward disputes with short disposition times. Thus, we narrowed our research period to ensure a common denominator among our cases.

      [214].     The disclosures often refer to the same business or law firm by several different names. For example, even within the same provider’s spreadsheets, a company like American Express might appear as “AMEX,” “American Express,” “American Express, Inc.,” and “American Express Inc.” Likewise, the same plaintiffs’ firm could be described as “Smith, Roberts, & Jones, LLP,” “Smith Roberts & Jones, LLP” or “Smith, Roberts, & Jones.” Stata, the statistical software program we used to analyze the data, is hyper-literal, and thus would “read” each of those entries as unique entities. Accordingly, we took the time-consuming step of going line by line through the disclosures and standardizing them.

      [215].     Providers sort cases into multiple categories in the “Type of Dispute” column of their disclosures. For example, JAMS breaks down matters as either Automotive Franchise, Business/Commercial, Class/Mass Action, Construction, Credit, Debt Collection, Employment, Goods, Health Care, Health Care (Non-malpractice), Insurance, Insurance Issues, NA, Other, Other Banking or Finance, Personal Injury, Professional Liability/Malpractice, Real Estate, Real Estate/Real Property (Non-Construction), or Telecommunications. We consolidated case types to make the data manageable and to create a common denominator across providers. When we did so, we treated “consumer” as our default category for ambiguous descriptions (such as Class/Mass Action and Other).

      [216].     Following the blueprint from the After the Revolution articles, our first step was to calculate the “raw repeat-player score,” which is the number of times that each business or plaintiffs’ firm or business appeared within a particular provider. See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 463. We used the following definitions: (1) “one-shotters” arbitrated once in a specific institution; (2) “low-level” repeat players arbitrated more than once but not more than the median raw repeat-player score; (3) “mid-level” repeat-players arbitrated a number of times that was between the median raw repeat-player score and the seventy-fifth percentile raw repeat-player score, inclusive; (4) high-level repeat players arbitrated a number of times that was between the seventy-fifth and ninetieth percentile of raw repeat-player scores, inclusive; and (5) super repeat players’ arbitrated a number of times that exceeded the ninetieth percentile of raw repeat-player score. For example, Wells Fargo appeared in seven ADR Services cases and forty-one JAMS cases. Because the median raw repeat-player score in ADR Services is eight, Wells Fargo is a low-level repeat player with respect to that provider. In JAMS, though, Wells Fargo’s raw repeat-player score falls between the median and the seventy-fifth percentile. Thus, the bank is a mid-level repeat player in JAMS. We encountered two related complications with calculating raw repeat-player scores. First, some cases involved more than one plaintiffs’ firm. For instance, in Kaiser, 167 of the 4,775 disputes featured at least two firms working in tandem. Most of the time these firms had different raw repeat-player scores. Second, plaintiffs often sue several parties for the same alleged wrongdoing. As a result, some spreadsheets list multidefendant cases several times, as though they were separate disputes. To avoid double or triple-counting these matters, we consolidated them into a single representative entry. Often, though, each defendant boasted a different raw repeat-player score, which meant that there was no single metric for defendant repeat-player issues in these cases. We resolved both of these dilemmas by using the highest raw repeat-player score whenever a matter featured multiple plaintiffs’ firms or defendants. Our logic was that if frequently arbitrating corporations or lawyers enjoy an advantage, it would also benefit any party that happened to be aligned with it. Finally, because we aimed in this Article to examine the impact of repeat-playing plaintiffs’ firms, we added a tier to our analysis. The After the Revolution pieces only examined one-shot, low-level, mid-level, and high-level repeat-playing attorneys. See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 488–90. In this Article, we also created a category for super repeat-playing plaintiffs’ attorneys.

      [217].     There are two differences between our treatment of “repeat pairs” in this Article and the After the Revolution pieces. First, in both Consumer Arbitration and Employment Arbitration, we counted all cases in which the same company (or plaintiffs’ lawyer) crossed paths with an arbitrator. See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 104; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 485. For example, if Company A appeared before Arbitrator Z in Case 1, and Company A appeared before Arbitration Z in Case 2, we coded both Case 1 and Case 2 as involving repeat pairs. For this Article, we stopped labeling Case 1 as a repeat pair. Instead, we treated Case 2 and all subsequent interactions as repeat pairs, meaning that only disputes in which the defendant or plaintiffs’ attorneys have already appeared before the same arbitrator meet the criteria. We made this change because the repeat-pair variable is supposed to detect whether arbitrators bend over backwards to appease parties or lawyers that might choose them in future matters. Yet in the first case involving a particular party or lawyer, it is not clear how the arbitrator would know that the party or lawyer would reappear. Cf. David Sherwyn et al., Assessing the Case for Employment Arbitration: A New Path for Empirical Research, 57 Stan. L. Rev. 1557, 1570 (2005) (asking how an arbitrator “who decided in favor of the employer in the employer’s first case [could have] know[n] that this large employer would be a repeat player while other large employers would not?”). Second, in both previous articles, we sorted repeat pairings into groups based on how often a party or lawyer had argued in front of an arbitrator. See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487. For simplicity’s sake, we drop that technique for the repeat-pairing variable (but not for the other repeat-player variables) in this Article.

      [218].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 89; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 465.

      [219].     We did drop one consumer case type from the AAA files. The institution administers “individual score reviews” of standardized tests. ACT, Terms and Conditions 2, http://www.act.org/content/dam/act/unsecured/documents/Terms-and-Conditions.pdf [https://perma.cc/Q264-QBUB] (last visited Dec. 21, 2018). These disputes are not analogous to lawsuits in court, and so we excluded them.

      [220].     See CFPB Study, supra note 44, § 1.4.3, at 11.

      [221].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 63; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 462; E&Y Study, supra note 44, at 9 n.11; Bingham, Repeat Player Effect, supra note 44, at 207–08; Colvin, supra note 44, at 5; Colvin & Gough, supra note 44, at 1027–29.

      [222].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 107; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 487–88.

      [223].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 110–11; Horton & Chandrasekher, Employment Arbitration, supra note 47, at 483, 487–88.

      [224].     There is one substantive difference between the two regression models: eagle-eyed readers may notice that whereas Colvin’s and our prior papers report the impact of variables on win rates as differences in percentages, see supra text accompanying notes 161–162, 170–171, 192–193, this Article reports differences in percentage points. At the risk of saying the obvious, these are not the same thing. For example, if the probability of a win is 50%, a 25% decline would make the probability of a win 50%-12.5% = 37.5%. However, a 25-percentage point decline would make the probability of a win 25%.

      [225].     Cf. Colvin & Gough, supra note 44, at 1030 (also using “former judge” as an independent variable). For some institutions, such as the AAA, we relied on whether the institution had used the prefix “Hon.” or “Judge” before listing the arbitrator’s name (in which case we inferred that the arbitrator had judicial experience) or employed the suffix “Esq.” (in which case we concluded that arbitrator was a practitioner). Cf. id. (employing the same technique). Other providers did not provide these clues, and so we Googled each arbitrator’s biography.

      [226].     A few other findings are worth noting. First, in both consumer and employment cases, the plaintiff win rate is higher for in-person hearings than telephonic arguments. Relative to arguing remotely, the odds of a win increase by 4.1 percentage points for consumers (although this result is just shy of significance, p = 0.050) and 7.4 percentage points for employees (p = 0.001). In addition, in employment cases, we unearthed a statistically meaningful difference in win rates in disputes that involve repeat-defendant-arbitrator pairings. Surprisingly, however, it is the opposite of what one would expect. The fact that a defendant has previously appeared before the same decision-maker increases the odds of an employee win by 8.2 percentage points (p = 0.009).

      [227].     See supra text accompanying notes 72–74.

      [228].     As noted above, although the CDRI study tried to examine JAMS cases, it was hamstrung by data limitations. See supra note 143.

      [229].     See supra text accompanying note 96.

      [230].     We dropped one case that was coded as “Other.”

      [231].     It is not always easy to tell whether a JAMS case was decided on the merits or voluntarily withdrawn by the plaintiff. When an arbitrator grants a dispositive motion, the “result” variable in the spreadsheet sometimes says “dismissed,” rather than “awarded.” Accordingly, we included matters that were coded as “dismissed” in our awards subsample if the arbitrator named a prevailing party.

      [232].     In addition, when a practitioner serves as arbitrator, tort plaintiffs succeed 45% of the time, as compared with only 19% of the time when a former judge presides (p < 0.01).

      [233].     Two other findings leap out from the tort regression analysis. First, the odds of a plaintiff win increase by 63.2 percentage points when the plaintiff’s law firm has appeared at least once before the same arbitrator (p < 0.001). Second, the likelihood of a plaintiff victory deceases by 17.4 percentage points if the arbitrator is a former judge.

      [234].     There are two quirks in the ADR Services data. First, the institution did not start listing the name of the plaintiffs’ law firm until 2015. Thus, although we were able to determine whether plaintiffs had attorneys or were pro se, we could not investigate the impact of repeat-playing plaintiffs’ firms on win rates. Second, some defendants are identified only as “Respondent.” Our best guess is that these are individual defendants in tort cases brought by plaintiffs against both the individual and the individual’s insurance company. We chose to treat them as one-shot defendants because we assumed that they did not have previous arbitration experience.

      [235].     See supra text accompanying notes 97–98.

      [236].     See Kaiser Permanente, Fast Facts, https://share.kaiserpermanente.org/article/fast-facts-about-kaiser-permanente [https://perma.cc/ZX3Q-Y36E] (last visited Dec. 21, 2018).

      [237].     See Off. of Indep. Adm’r, Annual Report, at ix (2016), http://www.oia-kaiserarb.com/pdfs/2016-Annual-Report.pdf [https://perma.cc/Z5B4-9DCL] (“Ninety-two percent . . . of the cases involved allegations of medical malpractice.”).

      [238].     Kaiser differs from all other providers in one important respect: it features a single perpetual defendant. Thus, we could not create a defendant repeat-player variable.

      [239].     See supra text accompanying note 99.

      [240].     See supra text accompanying notes 20–21.

      [241].     See supra text accompanying notes 23–25, 198.

      [242].     See Lynn Langton & Thomas H. Cohen, U.S. Dep’t of Justice, Civil Bench and Jury Trials in State Courts, 2005, at 8 tbl.9 (Apr. 9, 2009), https://www.bjs.gov/content/pub/pdf/cbjtsc05.pdf [https://perma.cc/6RK4-9S8Q].

      [243].     See Michael Heise, Justice Delayed?: An Empirical Analysis of Civil Case Disposition Time, 50 Case W. Res. L. Rev. 813, 835 tbl.4 (2000).

      [244].     Admittedly, this is not an apples-to-apples comparison with Heise’s jury trial statistics, because not all of the cases in our sample would have been tried by a jury.

      [245].     For a list of Fortune 100 companies, see Fortune 500, Fortune, http://fortune.com/fortune500 [https://perma.cc/ZS8E-MZG2] (last visited Dec. 21, 2018). For examples of huge class actions against some of these companies, see Jury Trial Demanded at ¶ 29(a), Schumacher v. Amazon.com LLC, No. SACV11-01906 CJC (JPRx) (C.D. Cal. Dec. 9, 2011), 2011 WL 9372528 (pursuing claims on behalf of “tens if not hundreds of thousands” of consumers who purchased a Kindle and a specific type of USB cord from Amazon); First Amended Complaint at ¶ 55, In re Am. Express Merchants’ Litig., No. 103CV09592 (S.D.N.Y. Feb. 19, 2009), 2009 WL 955691 (alleging that American Express charged supracompetitive rates to over four million businesses); First Amended Class Action Complaint, Washkoviak v. Student Loan Mktg. Ass’n, No. 01-9282 (D.C. Super. Ct. Sept. 15, 2004), 2004 WL 3676117 (asserting a nationwide class action against Sallie Mae); Class Action Complaint, Strawn v. AT&T Mobility, Inc., 513 F. Supp. 2d 599 (S.D. W. Va. 2006) (No. 2:06-cv-00988), 2006 WL 3845321 (contending that Cingular, which eventually merged with AT&T Mobility, violated a state consumer protection statute and affected over one million customers).

      [246].     For two reasons, plaintiffs must bring virtually all complaints against these companies in bilateral arbitration. First, many of these firms use express class arbitration waivers. See, e.g., Conditions of Use, Amazon.com, https://www.amazon.com/gp/help/customer/display.html/?nodeId=508088 [https://perma.cc/2L3A-JT67] (last visited Dec. 21, 2018); Customer Agreement, AT&T, https://www.att.com/legal/terms.wirelessCustomerAgreement-list.html [https://perma.cc/9BMG-YNHJ] (last visited Dec. 21, 2018); Comcast Agreement for Residential Services, Xfinity, https://www.xfinity.com/corporate/customers/policies/subscriberagreement [https://perma.cc/2QFN-G9MR] (last visited Dec. 21, 2018); Microsoft Services Agreement, Microsoft, https://www.microsoft.com/en-US/servicesagreement [https://perma.cc/MEW8-ZSG8] (last visited Dec. 21, 2018); Wireless Customer Agreement, Verizon, https://www.verizonwireless.com/legal/notices/customer-agreement [https://perma.cc/8DDQ-ZAZB] (last visited Dec. 21, 2018); Garcia v. Wachovia Corp., 699 F.3d 1273, 1275 (11th Cir. 2012) (noting that Wells Fargo’s customers agree to “arbitrate[] on an individual instead of a classwide basis”). Second, even corporations that do not use express class arbitration waivers can capitalize on Supreme Court precedent that forbids arbitrators from aggregating claims absent extraordinary circumstances. See, e.g., Stolt-Nielsen S.A. v. AnimalFeeds Int’l Corp., 559 U.S. 662, 684 (2010) (“[A] party may not be compelled under the FAA to submit to class arbitration unless there is a contractual basis for concluding that the party agreed to do so.”); David Horton, Clause Construction: A Glimpse Into Judicial and Arbitral Decision-making, 68 Duke L.J. (forthcoming 2019) (manuscript at 28, 37) (on file with author), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3125841 [https://perma.cc/EGM2-CVCW] (last visited Dec. 21, 2018).

      [247].     See supra text accompanying notes 20, 195–197.

      [248].     Admittedly, it is possible that wrongdoing by these companies tapered off during our research period, reducing the volume of claims. But this seems unlikely: several of these companies were embroiled in high-profile scandals between 2010 and 2016. See, e.g., Lesley Fair, AT&T Gets $105 Million Wake-up Call About Mobile Cramming, Fed. Trade Comm’n (Oct. 8, 2014), https://www.ftc.gov/news-events/blogs/business-blog/2014/10/att-gets-105-million-wake-call-about-mobile-cramming [https://perma.cc/VG5D-JZ8D] (describing allegations that wireless companies “signed consumers up without their express consent for text message services offering trivia, ringtones, flirting tips, celebrity gossip, and the like”); Bethany McLean, How Wells Fargo’s Cutthroat Corporate Culture Allegedly Drove Bankers to Fraud, Vanity Fair (Summer 2017), https://www.vanityfair.com/news/2017/05/wells-fargo-corporate-culture-fraud [https://perma.cc/5BC7-U8UR] (“Wells Fargo’s own analysis found that between 2011 and 2015 its employees had opened more than 1.5 million deposit accounts and more than 565,000 credit-card accounts that may not have been authorized.”); Ben Protess & Azam Ahmed, Freddie Mac’s Former Chief May Face S.E.C. Action, N.Y. Times (Mar. 15, 2011), http://query.nytimes.com/gst/fullpage.html?res=980DE2D8113BF935A25750C0A9679D8B63 [https://perma.cc/694Z-P7VP] (describing a government investigation into disclosure practices at Fannie Mae).

      [249].     See supra text accompanying notes 20–21; 195–197.

      [250].     See supra text accompanying notes 20–21.

      [251].     See Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 94–96.

      [252].     See id. at 93.

      [253].     See Horton & Chandrasekher, Employment Arbitration, supra note 47, at 471–72.

      [254].     These cases were still pending when we wrote Employment Arbitration, and thus did not appear in our data.

      [255].     Cal. Emp’t Law Council Brief, supra note 197, at 17.

      [256].     See Thomas H. Cohen & Steven K. Smith, Bureau of Justice Statistics, Civil Trial Cases and Verdicts in Large Counties, 2001, at 1 (2004); Lynn Langton & Thomas H. Cohen, Bureau of Justice Statistics, Civil Bench and Jury Trials in State Courts, 2005, at 1 (2008).

      [257].     See, e.g., Arthur Best et. al., Peace, Wealth, Happiness, and Small Claim Courts: A Case Study, 21 Fordham Urb. L.J. 343, 358 (1994); Suzanne Elwell & Christopher D. Carlson, Note, The Iowa Small Claims Court: An Empirical Analysis, 75 Iowa L. Rev. 433, 508 (1990).

      [258].     See, e.g., Colvin, supra note 44, at 4–7 (synthesizing the results of various studies of employment arbitration outcomes in court).

      [259].     See, e.g., David Allen Larson & Dr. David Dahl, Medical Malpractice Arbitration: Not Business as Usual, 8 Y.B. on Arb. & Mediation 69, 91 (2016).

      [260].     See, e.g., Horton & Chandrasekher, Consumer Arbitration, supra note 47, at 77–78 (discussing selection biases in arbitration awards).

      [261].     See George L. Priest & Benjamin Klein, The Selection of Disputes for Litigation, 13 J. Legal Stud. 1, 4–5 (1984).

      [262].     See id. at 14–15.

      [263].     See id. at 15.

      [264].     See id.

      [265].     See id. at 24–25.

      [266].     Indeed, pro se plaintiffs also routinely lose in court. See, e.g., Stephan Landsman, The Growing Challenge of Pro Se Litigation, 13 Lewis & Clark L. Rev. 439, 442 (2009).

      [267].     See, e.g., Alan B. Morrison, Can Mandatory Arbitration of Medical Malpractice Claims Be Fair? The Kaiser Permanente System, 70 Disp. Resol. J. 35, 37 (2015) (noting that medical malpractice claims are “quite complicated [and] almost always require expert testimony”).

      [268].     Lewis L. Maltby, Employment Arbitration and Workplace Justice, 38 U.S.F. L. Rev. 105, 113 (2003); Zev J. Eigen & David Sherwyn, Deferring for Justice: How Administrative Agencies Can Solve the Employment Dispute Quagmire by Endorsing an Improved Arbitration System, 26 Cornell J.L. & Pub. Pol’y 217, 262 (2016) (“Most arbitration policies do not allow for motions to dismiss or summary judgment motions.”); Brief of HR Policy Association, supra note 22, at *5 (“Arbitration . . . provides a system in which employers generally have a diminished ability to use dispositive motions prior to an evidentiary hearing . . . .”).

      [269].     See supra text accompanying notes 94–98.

      [270].     Cf. Sarah Rudolph Cole, The Lost Promise of Arbitration, 70 SMU L. Rev. 849, 863–64 (2017) (describing the creeping formalization of arbitration and noting that the process now “resembles a bench trial or a hearing before an administrative law judge”).

      [271].     Allied-Bruce Terminix Cos. v. Dobson, 513 U.S. 265, 280 (1995).

      [272].     See supra text accompanying notes 91–92.

      [273].     See, e.g., Christopher R. Drahozal, “Unfair” Arbitration Clauses, 2001 U. Ill. L. Rev. 695, 751 (2001) (“Plaintiffs’ attorneys may represent numerous employees, franchisees, or consumers against corporate defendants, effectively becoming repeat players. Their better information will discourage arbitrators who might otherwise show favoritism toward corporations.”); Estreicher, supra note 142, at 566 (“[T]he emergence of an organized plaintiff’s bar . . . should drive down considerably any claimed systematic advantage for employers.”); W. Mark C. Weidemaier, Arbitration and the Individuation Critique, 49 Ariz. L. Rev. 69, 75 (2007) (“[M]odern plaintiffs’ firms, like other repeat players, aggregate claims.”).

      [274].     See, e.g., Bingham, Repeat Player Effect, supra note 44, at 198 (“The repeat player effect, if any, from lawyer representation is not yet measurable.”).

      [275].     See Searle Report, supra note 44, at 81–82.

      [276].     See supra text accompanying notes 173–174. Our simplified regression analyses only included our repeat-player scores for defendants and plaintiffs’ firms. We did not include the full range of independent variables from our win rate regression analyses because most of these factors were not applicable to cases that ended before the award stage. For instance, our exploration of AAA win rates controls for repeat pairs, whether the arbitrator was a retired judge, and the type of hearing (in-person versus telephonic and paper-based versus oral argument). See supra Part II.B.1. We dropped these variables in our settlement regression analyses because some disputes settled before an arbitrator was appointed and many disputes settled before the hearing. Finally, we defined “settlement” differently than Colvin and Gough. As noted above, they excluded cases that were listed by the AAA as “withdrawn” from that category. See id. However, as the CFPB Study observed, the AAA records do not always draw a bright line between the two categories, and many “withdrawn” cases may also have been settled. See CFPB Study, supra note 44, § 6.2.2. As a result, we treated a case as “settled” if it did not reach the award stage.

      [277].     See, e.g., 42 U.S.C. § 2000e–5(k) (2012) (providing that in employment discrimination cases, “the court, in its discretion, may allow the prevailing party . . . a reasonable attorney’s fee (including expert fees)”); 15 U.S.C. § 1692k(a)(3) (2012) (allowing prevailing plaintiffs to recover fees and costs for violations of the Fair Debt Collection Practices Act); see also Dan B. Dobbs, Awarding Attorney Fees Against Adversaries: Introducing the Problem, 1986 Duke L.J. 435, 436 (discussing the evolution of these laws). Some fee-shifting statutes seem to make fee and cost awards mandatory. See, e.g., Cal. Lab. Code § 1194 (West 2017); 43 Pa. Stat. § 260.9a(f) (West 2017). However, even these laws only authorize fee awards for “prevailing parties,” which gives the judge the freedom to decide whether a plaintiff has “prevailed.” See, e.g., Graciano v. Robinson Ford Sales, Inc., 50 Cal.Rptr.3d 273, 281–82 (Cal. Ct. App. 2006) (discussing the ambiguity in the “prevailing party” concept). Other fee-shifting statutes are expressly permissive. See, e.g., Colo. Rev. Stat. Ann. § 8-4-110 (West 2018); Fla. Stat. Ann. § 501.2105(3) (West 2018); 815 Ill. Com. Stat. Ann. § 505/10(a)(c) (West 2018); N.Y. Gen. Bus. Law § 198-a(1) (McKinney 2018); Ohio Rev. Code § 1345.09(f) (West 2018); Tex. Bus & Com. Code Ann. § 17.50(d) (Vernon 2018).

      [278].     29 U.S.C. § 216(b) (2012).

      [279].     Cal. Civ. Code § 1780(a), (e) (West 2018); see also Laster v. T-Mobile USA, Inc., No. 05CV1167 DMS (AJB), 2008 WL 5216255, at *1 (S.D. Cal. Aug. 11, 2008), aff’d sub nom. Laster v. AT&T Mobility LLC, 584 F.3d 849 (9th Cir. 2009), rev’d sub nom. AT&T Mobility LLC v. Concepcion, 563 U.S. 333 (2011), and amended in part, No. 05CV1167 DMS WVG, 2012 WL 1681762 (S.D. Cal. May 9, 2012).

      [280].     Associated Indus. of N.Y. State, Inc. v. Ickes, 134 F.2d 694, 704 (2d Cir. 1943).

      [281].     Thomas D. Rowe, Jr., The Legal Theory of Attorney Fee Shifting: A Critical Overview, 1982 Duke L.J. 651, 662 (1982).

      [282].     Admittedly, arbitrators might not always be able to use the multiplier when a plaintiff prevails on federal claims. In general, “a state statute providing [a p]laintiff with the remedy of attorney’s fees [can be] preempted by federal law.” Damiano v. Harleysville Ins. Co., No. 3:13-cv-07239-FLW-LHG, 2014 WL 12622874, at *5 (D.N.J. July 17, 2014). Thus, it is not clear that state law can enlarge a fee award under a federal statute. Cf. Hubbard v. SoBreck, LLC, 554 F.3d 742, 745 (9th Cir. 2009) (holding that the Americans With Disabilities Act’s restrictions on awarding fees to defendants overrode the California Disabled Persons Act’s regime of mandatory fees for prevailing parties); Kohler v. Presidio Int’l, Inc., 782 F.3d 1064, 1070 (9th Cir. 2015) (same). But see Jankey v. Song Koo Lee, 290 P.3d 187, 189 (Cal. 2012) (reaching the opposite conclusion). However, most of the authority in this area involves attempts by victorious defendants to recover fees under state law despite federal fee-shifting rules that are designed to encourage claims by plaintiffs. In this context, allowing state law to alter its federal counterpart might “create a ‘chilling effect’ . . . by discouraging claimants from vindicating their rights.” Grimes v. Kinney Shoe Corp., 938 P.2d 997, 999 (Alaska 1997). Conversely, the multiplier—which cuts in the opposite direction by magnifying the claim-facilitating aims of federal fee-shifting statutes—does not conflict with Congress’s ambitions. And in any event, some federal statutes expressly permit states to enact parallel legislation “that provides additional remedies not available under [federal law].” Butler v. DirectSat USA, LLC, 800 F. Supp. 2d 662, 672 (D. Md. 2011) (analyzing the FLSA). Finally, because arbitrators have broad power to “decide disputes and fashion remedies,” courts might not grant a petition to vacate an arbitral award on the grounds that the arbitrator impermissibly granted the plaintiff extra fees. Todd Shipyards Corp. v. Cunard Line, Ltd., 943 F.2d 1056, 1064 (9th Cir. 1991).

      [283].     To be clear, we see the multiplier as a second-best solution. Like other commentators, we believe that Congress should ban class arbitration waivers. See, e.g., Sarah Rudolph Cole, On Babies and Bathwater: The Arbitration Fairness Act and the Supreme Court’s Recent Arbitration Jurisprudence, 48 Hous. L. Rev. 457, 467–68 (2011) (proposing that lawmakers amend the FAA to achieve this goal). However, federal intervention is wishful thinking given the recent failure of the CFPB’s rule and the current political climate. See supra text accompanying notes 122–125.

      [284].     See supra text accompanying note 103.

      [285].     See supra Part III.A.3.

      [286].     It may also be less necessary. Because medical malpractice lawsuits tend to involve individualized facts, tort lawyers rarely combine them into a class action. As a result, it is not clear how much Concepcion has affected the volume of malpractice lawsuits.

      [287].     See, e.g., Ind. Code Ann. § 34-18-18-1 (West 2018) (capping awards of attorneys’ fees from certain public funds at 15% of the recovery from the fund); Yalango v. Popp, 644 N.E.2d 1318, 1321 (N.Y. 1994) (describing New York’s statutory formula for awarding attorneys’ fees in medical malpractice claims); Newton v. Cox, 878 S.W.2d 105, 108 (Tenn. 1994) (explaining that the state legislature has capped contingent fee awards in malpractice cases at a third of the “gross recovery of a malpractice claimant”).

      [288].     One possibility might be to force defendants to pay an additional 5% of any award to the plaintiffs’ law firm.

      [289].     See supra Part II.B.1–2.

      [290].     See 2013 District Court Judges Summer Conference, Affidavit of Michael B. Stein, Univ. N.C. School of Gov’t 4–5 (2013), https://www.sog.unc.edu/sites/www.sog.unc.edu/files/course_materials/Ex C Stein Affidavit.pdf [https://perma.cc/WX9A-PE3F] (describing the World Law Group’s tactics). In the AAA, the World Law Group won 5 of 527 awards (less than 1%).

      [291].     See, e.g., Christiansburg Garment Co. v. EEOC, 434 U.S. 412, 421 (1978) (allowing defendants to recover attorneys’ fees in Title VII cases “upon a finding that the plaintiff’s action was frivolous, unreasonable, or without foundation”); Cal. Lab. Code § 218.5 (West 2017) (same in labor code cases “if the court finds that the employee brought the court action in bad faith”); Tex. Prop. Code Ann. S 92.058(c) (West Supp. 1995); see also Lorraine Wright Feuerstein, Comment, Two-Way Fee Shifting on Summary Judgment or Dismissal: An Equitable Deterrent to Unmeritorious Lawsuits, 23 Pepp. L. Rev. 125, 153–56 (1995) (collecting other examples).

      [292].     See Kat Sieniuc, CFPB Reaches $107M Settlement in World Law Group Debt Scam, Law 360 (July 20, 2016), https://www.law360.com/articles/819450 [https://perma.cc/WNT5-7TC9].

      [293].     See CFPB Study, supra note 44, § 5.6.6, at 40.

      [294].     See supra text accompanying note 277.

      [295].     See Southland Corp. v. Keating, 465 U.S. 1, 11 (1984).

      [296].     See Perry v. Thomas, 482 U.S. 483, 492 (1987).

      [297].     See Brown ex rel. Brown v. Genesis Healthcare Corp., 724 S.E.2d 250, 293 (W. Va. 2011), cert. granted, judgment vacated sub nom. Marmet Health Care Ctr., Inc. v. Brown, 565 U.S. 530 (2012).

      [298].     Sparks v. Stone St. Capital, Inc., No. 3-02-CV-724-R, 2002 WL 1575404, at *6 n.7 (N.D. Tex. July 15, 2002).

      [299].     Mont. Code Ann. § 27-5-114(4) (repealed 1997).

      [300].     517 U.S. 681, 687 (1996).

      [301].     See Kindred Nursing Ctrs. Ltd. P’ship v. Clark, 137 S. Ct. 1421, 1425 (2017).

      [302].     Hiro N. Aragaki, Arbitration’s Suspect Status, 159 U. Pa. L. Rev. 1233, 1244 (2011).

      [303].     See supra text accompanying notes 60–61.

      [304].     Aragaki, supra note 302, at 1243.

      [305].     Id.

      [306].     See supra text accompanying notes 59–61.

      [307].     Businesses might contend that the multiplier is inconsistent with the FAA because it makes arbitration less desirable from their perspective. See, e.g., Cal. Emp’t Law Council Brief, supra note 197, at *17 (making similar arguments about a rule that prohibits class arbitration waivers); cf. AT&T Mobility LLC v. Concepcion, 563 U.S. 333, 350 (2011) (opining that the Discover Bank rule violates the FAA’s goals because “class arbitration greatly increases risks to defendants”). However, the Court has declared that the FAA does not eclipse “state rules governing the conduct of arbitration . . . which are manifestly designed to encourage resort to the arbitral process.” Volt Info. Scis., Inc. v. Bd. of Trustees of Leland Stanford Junior Univ., 489 U.S. 468, 476 (1989). Given how much companies prize class arbitration waivers, it seems unlikely that the multiplier would cause an exodus from arbitration.

      [308].     Concepcion, 563 U.S. at 345.

More in this Issue

Visiting Judges

Despite the fact that Article III judges hold particular seats on particular courts, the federal system rests on judicial interchangeability. Hundreds of judges “visit” other courts each year and collectively help decide thousands of appeals. Anyone from a retired Supreme Court Justice to a judge from the U.S. Court of International Trade to a district […]

The New World of Agency Adjudication

In 1946, the Administrative Procedure Act (APA) set forth the criteria for “formal” adjudication, requiring an administrative law judge to make the initial determination and the agency head to have the final word. That is the lost world. Today, the vast majority of agency adjudications Congress has created are not paradigmatic “formal” adjudications as set […]

Copyright Arbitrage

Regulatory arbitrage—defined as the manipulation of regulatory treatment for the purpose of reducing regulatory costs or increasing statutory earnings—is often seen in heavily regulated industries. An increase in the regulatory nature of copyright, coupled with rapid technological advances and evolving consumer preferences, have led to an unprecedented proliferation of regulatory arbitrage in the area of […]

Disenfranchisement in the US Presidential Nomination Process Through Caucuses and the Gatekeeping Role of Iowa and New Hampshire

This Note examines inequities in the presidential nomination process. The nomination process has developed such that African American and women voters, compared to white male voters, wield less influence over which candidates parties nominate. By examining data from recent elections and scholarship from the fields of law, political science, and economics, this Note illuminates the […]

Big Data and the Non-Horizontal Merger Guidelines

Data is increasingly valuable as a product, input, and market tool. Exclusive data may be the most valuable asset many firms possess. Yet, regulators in the United States often overlook the importance of data-related mergers, especially between firms that do not directly compete. This is in part because the Non-Horizontal Merger Guidelines (“NHMG”) are out […]