The Supreme Court’s decision last Term in Epic Systems Corp. v. Lewis1 is a vivid illustration of the declining power of workers in the U.S. political system. The opinion, authored by Justice Gorsuch, upheld the validity of employment contracts in which employees give up their right to collective litigation against their employer. It is reminiscent of a once-infamous labor law decision from the late 1920s, the Red Jacket case,2 in which Judge Parker of the Fourth Circuit protected the power of coal mine owners to forbid their workers from interacting with unions.3 These two cases, situated ninety years apart, present a useful comparison for tracing changes in worker power. Both cases address contentious issues about workers’ collective rights. They are similar, too, in terms of outcome and degree of engagement with on-the-ground labor issues. What is different is their political acceptability. In 1930, Judge Parker became the first Supreme Court nominee in decades to be rejected by the Senate,4 largely because of his decision in Red Jacket.5 Now, an opinion with a comparable outcome and analytical approach gets the support of a majority of the Court, without much of a reaction (so far) from Congress.
The legal question in Epic Systems involved a conflict between two federal statutes, the Federal Arbitration Act6 (FAA) and the National Labor Relations Act7 (NLRA). The FAA, enacted in 1925, provides that arbitration agreements “shall be valid, irrevocable, and enforceable, save upon such grounds as exist at law or in equity for the revocation of any contract.”8 The Supreme Court has interpreted this provision to allow for the enforcement of arbitration agreements for claims arising from federal statutes.9 There are exceptions, though. The FAA does not require enforcement of an agreement that waives a person’s substantive rights guaranteed by another statute,10 nor does it require arbitration of a statutory claim if the statute giving rise to that claim expresses a “contrary congressional command.”11 Predictably, a number of Supreme Court cases have examined whether various statutory claims can fit into these exceptions.12 Epic Systems is the latest installment in this series. It focuses on Section 7 of the NLRA, which guarantees that “[e]mployees shall have the right to self-organization . . . and to engage in other concerted activities for the purpose of collective bargaining or other mutual aid or protection.”13 The specific question raised by Epic Systems is whether Section 7 of the NLRA gives workers a substantive right to collective litigation, such that arbitration agreements that waive that right are unenforceable under the FAA.
In January 2005, Stephen Morris started working as a junior accountant at Ernst & Young (EY).14 About one year later, the company sent him an email with an attached arbitration agreement.15 Two provisions of that agreement would become particularly relevant in later litigation. First, the agreement stated that all employee claims relating to federal and state wage statutes would be resolved via arbitration. Second, the agreement specified that claims by individual employees could not be consolidated.16 According to the email, if Morris continued to work at EY, he would be bound by the agreement.17 He continued working.
Several years later, after leaving his job, Morris filed a class action against EY in federal court for, among other claims, violations of the federal Fair Labor Standards Act.18 The suit flatly contradicted the terms of his arbitration agreement: it was in court, and it combined the claims of several employees. Morris had an argument, however, for why the court should decline to enforce the arbitration agreement. A recent National Labor Relations Board (NLRB) decision, D.R. Horton,19 had held Section 7 rights to be substantive and therefore unwaivable under the FAA.20 Citing D.R. Horton, Morris argued that since the individualized arbitration clause of the arbitration agreement violated his Section 7 rights, the district court should refuse to enforce the agreement.21
Senior Judge Whyte of the Northern District of California was unconvinced. He noted that the NLRB in D.R. Horton had attempted to interpret the FAA, a statute it was not charged to administer, and that its interpretation was therefore not entitled to any deference.22 Moreover, in Judge Whyte’s view, the NLRB’s conclusion was simply incorrect; reading individual arbitration clauses to be unenforceable because of a conflict with the NLRA would be “inconsistent with the Supreme Court’s interpretation of the FAA.”23 He granted a motion by EY to compel arbitration.24
The Ninth Circuit vacated the order.25 Writing for the panel, Chief Judge Thomas26 reviewed the text of the NLRA and the relevant case law and concluded that “[c]oncerted activity — the right of employees to act together — is the essential, substantive right established by the NLRA.”27 Since the arbitration agreement violated the NLRA, and because the FAA “does not mandate the enforcement of contract terms that waive substantive federal rights,” the court would not enforce the agreement.28 Judge Ikuta, in dissent, called the decision “breathtaking in its scope and in its error.”29 Finding nothing in the text, legislative history, or purpose of the NLRA that created a substantive right to class-wide claims or expressed a congressional command opposed to arbitration, she would have enforced the arbitration agreement.30
The Supreme Court reversed.31 Writing for the majority, Justice Gorsuch framed the issue as a straightforward matter of resolving a potential contradiction between two federal statutes. On one hand, the Court had the FAA, in which Congress had expressed “a liberal federal policy favoring arbitration agreements.”32 On the other hand, the Court had the NLRA, which was silent about workers’ rights to collective litigation. “It is this Court’s duty to interpret Congress’s statutes as a harmonious whole,”33 Justice Gorsuch explained. “[A]biding that duty here leads to an unmistakable conclusion”34: the NLRA should be interpreted so as not to interfere with the enforceability of arbitration agreements under the FAA.35
Justice Gorsuch engaged in detailed statutory analysis of both the FAA and the NLRA. Starting with the FAA, he explained that the text and history of the Act made clear that courts should presumptively enforce arbitration agreements, even those that called for individualized proceedings.36 He acknowledged that the FAA did have a saving clause that called on courts not to enforce arbitration agreements that were illegal. He explained, however, that based on recent Supreme Court precedent, courts could not invalidate arbitration agreements based on “defenses that apply only to arbitration”37 or “interfer[e] with fundamental attributes of arbitration.”38 Using Section 7 of the NLRA (and Section 8, which prohibits employer interference with Section 7 rights) to overcome the FAA would violate this rule, since Section 7’s grant of the right to take collective action is in conflict with the traditionally individualized nature of arbitration.39
Next, Justice Gorsuch examined the NLRA. He supplied a number of reasons why Section 7 did not confer a right to collective litigation: the plain text of the act was silent about litigation, the ejusdem generis canon of construction counseled for a narrow interpretation of the catch-all term “other concerted activities,”40 and the idea of collective litigation rights being conferred by Section 7 would fit uncomfortably with the rest of the statutory structure.41 He pointed out that other statutes were much clearer in their grant of collective litigation rights, and observed that it would be anachronistic to construe Section 7 to confer class action rights, considering that Federal Rule of Civil Procedure 23 did not exist until 30 years after the NLRA was enacted.42
Finally, Justice Gorsuch resolved the potential conflict between the FAA and the NLRA. He noted that the Court had rejected every attempt in the past to overwhelm the FAA with another federal statute.43 In fact, the Court had previously allowed for arbitration of statutory claims even when those other statutes expressly allowed for collective litigation.44 Considering this precedent, the Court was compelled to find no conflict between the FAA and the NLRA, and to therefore hold the arbitration agreements enforceable under the FAA.45
Justice Thomas wrote a one-paragraph concurrence.46 He advanced an interpretation of the FAA in which the only types of arbitration agreements that are unenforceable are those that have some defect related to contract formation.47
Justice Ginsburg, in dissent, wrote that “[t]he Court today subordinates employee-protective labor legislation to the [FAA].”48 She placed the NLRA (and the Norris-LaGuardia Act,49 another foundational piece of New Deal labor legislation) in historical context. Both statutes were attempts by Congress to correct power imbalances between employers and employees.50 Justice Ginsburg drew special attention to “yellow-dog contract[s]” — employment agreements in which the employee promised not to participate in any union activities.51 One of the express purposes of the New Deal legislation, according to Justice Ginsburg, was to eliminate these types of agreements, and to protect employees’ “‘fundamental right’ to join together to advance their common interests.”52 With this historical backdrop in place, she made an affirmative argument for a broad interpretation of Section 7 of the NLRA, and countered each of the arguments that Justice Gorsuch had put forward for a narrow interpretation.53 Concluding that the right to collective action was squarely protected by Section 7, Justice Ginsburg examined the text of the FAA and the relevant case law and found that nothing “requires subordination of the NLRA’s protections.”54
In 2012, Professors Bruce Western and Jake Rosenfeld observed that “[o]ver the last few decades, . . . workers’ collective voice in the political process has weakened.”55 Epic Systems fits neatly within this historical narrative. The opinion can be contrasted with a highly controversial Fourth Circuit decision from the 1920s, Red Jacket, in which the court upheld an employment contract that prohibited workers from interacting with unions. The two cases follow a similar mode of analysis: they reach a defensible legal conclusion, but they leave unmentioned actual labor conditions. What makes the comparison between Epic Systems and Red Jacket particularly noteworthy is the drastic difference in how the cases were received by Congress. In 1930, the Red Jacket case was sufficiently toxic that its author, Judge Parker, became the first Supreme Court nominee in decades to be rejected by the Senate. Now, a roughly similar opinion gets the support of a majority of the Justices, without provoking an especially strong response from Congress.
In order to understand the parallels between Red Jacket and Epic Systems, it is important to explore Red Jacket’s factual background. The case arose from the coal fields of West Virginia. In the early 1900s, mines in that region functioned as “closed, martial societies”56 — workers lived in company houses on company land, were monitored by “private guards and deputy sheriffs,”57 and could only access mail, entertainment, and education that were approved by their employer.58 Most companies insisted on yellow-dog contracts.59 In early 1920, the Red Jacket Company, one of the largest mine operators in the area, decided to demonstrate its commitment to keeping unions out of the region by firing all of its workers who had joined the United Mine Workers of America.60 Conflict ensued, which quickly became violent and widespread; Professor Peter Fish, in his historical study of the episode, describes the subsequent years as a period of “[g]uerrilla warfare” across the state.61 Red Jacket and other mine operators sought an injunction against the union. Judge McClintic of the Southern District of West Virginia granted their request, and the Union appealed.62
In Red Jacket, Judge Parker upheld the lower court’s injunction against the union.63 The statutory text in question came from the Clayton Act,64 which said courts could not enjoin unions from recruiting members by “peaceful and lawful means.”65 The union argued that the injunction violated this rule, since it prevented the union from taking lawful steps to increase its membership.66 Judge Parker disagreed. For one thing, he wrote, the union’s argument was at odds with Supreme Court precedent: an opinion from ten years earlier, Hitchman Coal & Coke Co. v. Mitchell,67 had upheld an injunction that placed even more restrictions on union activity.68 More fundamentally, though, Judge Parker thought the union misunderstood what the term “lawful” activity meant. Persuading employees with yellow-dog contracts to join a union, even if done peacefully, still was not “lawful,” since it interfered with the mine owners’ rights to negotiate and to contract freely with their employees. He upheld the injunction, implicitly supporting the validity of the underlying yellow-dog employment contract.69
Three years later, when Judge Parker was nominated to the Supreme Court, the perception of Red Jacket as an anti-worker decision likely proved fatal to his confirmation prospects. During the Senate debates about Judge Parker, Senator William Borah gave a detailed, multi-hour lecture about the legal missteps and moral blind spots in the Red Jacket decision.70 He concluded that Judge Parker had gone further than anyone else, including the Supreme Court Justices, in shielding employers from union activity, and warned that confirming Judge Parker would be “in moral effect a decision of the Senate in favor of the ‘yellow dog’ contract.”71 Senator Robert F. Wagner, who would later serve as the lead author of the NLRA,72 joined in the attack. He criticized Judge Parker’s “failure to be aware of the fact that he was in the presence of an important problem” — the disempowerment of workers.73 The Red Jacket opinion, to Senator Wagner, could only have been written by a judge who was “incapable of viewing with sympathy the aspirations of those who are aiming for a higher and better place in the world.”74 It is impossible to attribute Judge Parker’s rejection to any single issue — for example, he was also opposed by the NAACP for having made statements like “[t]he participation of the Negro in politics is a source of evil and danger to both races”75 — but most scholars who have examined the episode have identified the Red Jacket decision as one of the primary points of Senate resistance.76
The Red Jacket decision and Judge Parker’s failed confirmation hearings shortly predated a historic — and, in many ways, short-lived — surge in worker political power. The early 1930s saw two landmark pieces of pro-labor legislation: the Norris-LaGuardia Act of 1932 and the NLRA in 1935. To some observers, this New Deal labor legislation signaled the beginning of a new, lasting, and muscular role for workers in industrial relations; Professors Walter Gellhorn and Seymour Linfield, writing in 1939, called the NLRA a “perfected . . . instrument to prevent intimidatory employer tactics.”77 The following decades, however, saw an ongoing process of “deradicalization.”78 Courts restricted the types of remedies available to workers under the NLRA,79 limited the reach of labor legislation under expansive theories of employers’ First Amendment rights,80 and allowed for private arbitration of statutory employment claims.81 There have been important exceptions to this pro-employer trend — for example, workers have successfully pushed for major antidiscrimination legislation82 — but the overall power differential between workers and their bosses remains relatively large. As Professor Kate Andrias recently put it, “American labor unions have collapsed[,]” and employment law “has not filled the void.”83
Which brings us to the 2017 Term: Epic Systems was one of two major decisions dealing with labor and employment issues,84 and it resembled Red Jacket in two important ways. First, the broad strokes of both cases’ outcomes were similar: the courts upheld the legality of employer tactics to prevent workers from banding together. Second, the authors of each decision chose not to engage with the substantive labor conditions that lay behind the cases. The Red Jacket opinion presented the yellow-dog employment contract as the product of an arm’s-length transaction; nowhere did Judge Parker mention that mine workers were often isolated from outside contact and dependent on their employer for necessities like housing.85 In Epic Systems, Justice Gorsuch similarly left unmentioned many of the employment problems that motivated the plaintiffs to file their claims. Justice Ginsburg’s dissent pointed out two examples. First, Justice Gorsuch did not discuss the frequency of wage theft86 — a catch-all term for employer practices such as failure to pay minimum wage,87 which is particularly common in low-wage industries and among workers who are women, people of color, or born outside the United States.88 Moreover, Justice Gorsuch did not grapple with the fact that problems like wage theft can often only be remedied through collective litigation,89 given the high cost of bringing a claim and the relatively low individual payout.90 As with Red Jacket, these omissions probably do not render the opinion “wrong”; for example, even taking substantive labor inequalities into account, Justice Gorsuch still could have found, based on the NLRA’s text and history, that Section 7 does not protect the right to collective litigation. But the omissions do affect the tone of the opinion, and suggest a lack of sympathy for workers.
Whereas Red Jacket provoked a strong reaction from the Senate — the first rejection of a Supreme Court nominee in more than three decades — the response to Epic Systems has been relatively subdued. A few Democrats tweeted in protest,91 and at least one bill was introduced in the Senate to amend the NLRA to explicitly confer the right to collective litigation.92 But it is difficult to find evidence of more meaningful pushback. Two proposed pieces of legislation from late 2017 and early 2018 targeted at rolling back the recent expansion of arbitration93 remain stuck in various congressional committees.94 According to the daily Congressional Record, the issue of arbitration agreements was discussed only once on the floor of the House or the Senate in the two months after Epic Systems was handed down, and that was a passing remark in a speech in which recently appointed Senator Tina Smith of Minnesota introduced herself to her peers.95 Of course, one explanation for this lack of a reaction from Congress is that, as of this writing, both houses are currently controlled by the Republican Party, which tends to align itself less explicitly with the interests of unions and workers.96 But even when Democratic Party politicians have had an opportunity to express concern about the Court’s treatment of workers — the confirmation process of Supreme Court nominee Brett Kavanaugh, for example — they have instead tended to focus on other issues.97
Scholars have already started debating what Epic Systems means for workers’ rights moving forward.98 It’s worth reflecting, too, on what the opinion tells us about how far, and in what direction, we have already come. The comparison between Epic Systems and Red Jacket is imperfect — they emerge from different times and places, the legal issues are distinct, and the political outcry over Red Jacket may simply have resulted from the fact that its author happened to be nominated to the Supreme Court. But the cases still demonstrate a shift. In 1930, an opinion that was supportable by the existing law but indifferent to working conditions was offensive enough that it kept its author off the Court. Nearly ninety years later, that type of opinion gets the support of a majority of the Justices, without causing too much of a stir.