How the Roberts Court Has Changed Labor and Employment Law

Justices from both the conservative and liberal wings of the Roberts Court1 emphatically reject the suggestion that they act as outcome-oriented political actors rather than neutral arbiters of the rule of law. Chief Justice Roberts recently noted “that when you live in a politically polarized environment, people tend to see everything in those terms . . . [t]hat’s not how we at the court function, and the results in our cases do not suggest otherwise.”2 Similarly, Justice Kagan stated that “it would be very hard to tell a story of a politicized court. The 5-4 cases that we had were 5-4 in all kinds of different ways, often with lots of people doing unexpected things.”3

If all cases were of equal importance, the justices would have a point. Decisions that are 7-2, 8-1, or unanimous make up approximately half of all cases decided by the U.S. Supreme Court; by contrast, only about twenty percent of decisions are 5-4.4 But all cases are not equally important. Indeed, a remarkably consistent trend emerges if one examines the Roberts Court’s most significant labor and employment decisions. In all but one case, each of the conservative justices voted for outcomes that favor employers while each of the liberal justices voted for outcomes that favor employees.

Over the last sixteen years, the Roberts Court has reshaped the balance of power between employers and employees through these closely divided decisions, often in measurable ways. This article discusses how the Roberts Court has altered labor and employment law, focusing on the observable changes to the workplace.

I. Procedural Hurdles

The Roberts Court has made it more difficult for employees to get to court and easier to dismiss claims in court.5 These procedural changes—which largely do not address the underlying merits of a claim—have altered the dynamics in favor of employers.

A. Class Action Waivers

Many employment claims that historically would have proceeded to federal court are now litigated in private arbitration where employers are more likely to win and more likely to pay smaller awards if they lose.6 More than 60 million workers already have mandatory arbitration clauses in their employment contract, representing approximately 56.2% of the private, non-union workforce.7 By 2024, these figures are expected to surge by approximately half: more than 80% of the private sector non-union workforce will have a mandatory arbitration clause in their employment contract.8 One major reason why is the Roberts Court’s recent decision in Epic Systems v. Lewis9 where the Court’s five conservatives enforced an employment agreement that waived the right to pursue a class action (“class action waivers”).10

Epic Systems was not unexpected. Over the previous seven years, the Court’s five conservatives issued two previous decisions eliminating defenses that might render class action waivers unenforceable. First, in AT&T Mobility LLC v. Concepcion,11 the Court held that the Federal Arbitration Act (“FAA”) prohibited states from refusing to enforce arbitration agreements in consumer contracts that eliminated the availability of class actions.12 As a result, companies could require that consumers pursue claims individually in arbitration rather than collectively as a class, even though it is often irrational to do so given the small sum involved (the claim in Concepcion was for $30.22). Due to the U.S. Constitution’s Supremacy Clause,13 the FAA preempts any state laws that assert such contracts are unenforceable. Next, in American Express Co. v. Italian Colors Restaurant,14 the Court rejected the notion that class action waivers are unenforceable when they are negative value claims—i.e. when a plaintiff’s cost of pursuing the claim exceeds any potential recovery. The Court concluded that a plaintiff can still “effectively vindicate” negative value claims in arbitration, despite the fact that no rational plaintiff would do so.

The Court’s 2018 Epic Systems decision extended these principles to class action waivers in employment contracts. This time, the defense was that Section 7 of the National Labor Relations Act (“NLRA”)15 guaranteed the right to engage in concerted activity for mutual aid or protection, and a waiver of the right to pursue a class action in arbitration necessarily conflicted with this statutory right.16 The Court disagreed, finding that Section 7 did not encompass the procedural right to form a class action and therefore did not conflict with the FAA.17

As a result of the Court’s decision, employers may generally mandate workers waive the right to pursue class litigation in favor of individual arbitration. Currently, approximately 25 million workers have such class action waivers, and this figure is anticipated to grow substantially over the coming years due to Epic Systems.18

Studies show that these waivers are enormously effective in preventing claims in the first place—more than 98% of claims that one would expect to see in court are never filed in arbitration.19 This phenomenon is not surprising given the underlying dynamics of filing a claim in arbitration. Workers are far less likely to file small dollar wage and hour claims under the Fair Labor Standards Act (“FLSA”)20 because they cannot band together as a class. Further, attorneys are less likely to take these cases, perhaps in part due to lower anticipated recoveries and provisions that limit recovery when successful.21

Overall, Epic Systems permits employers to largely avoid class action claims by workers, provided such a waiver is included in an executed employment contract.

B. Class Action Litigation

The Roberts Court has also tilted the playing field in favor of employers for class action claims that proceed to court rather than arbitration. In Wal-Mart Stores, Inc. v. Dukes,22 the Court held that a class of approximately 1.5 million women, who were current and former employees of Wal-Mart, could not be certified because it could not meet the “commonality” prerequisite that applies to all class actions.23 In doing so, the Court made it more difficult for all sorts of class actions to be certified.

In order for any class action to be certified, the class must first meet four prerequisites: (1) numerosity, (2) commonality, (3) typicality, and (4) adequacy of representation.24 Then, the class must be one of three types identified in Rule 23(b): (1) a limited fund action, (2) a declarative or injunctive relief class, or a (3) class seeking monetary damages.25

In Wal-Mart Stores, Inc. v. Dukes, the class action plaintiffs introduced evidence that showed a persistent gap between what men and women employees earned at Wal-Mart.26 Specifically, women filled 70% of the hourly jobs but were only 33% of the management employees. They alleged that this wide disparity, among other things, demonstrated that Wal-Mart’s pay and promotion policies violated Title VII of the 1964 Civil Rights Act (“Title VII”) by discriminating against women. In response, Wal-Mart noted that it gives local supervisors discretion over employment matters so there was no uniform employment policy at issue; instead, there were individual employment decisions across different jobs, different levels of hierarchy, for different lengths of time, in 3,400 stores, across 50 states. As a result, they argued that there was no common question which could resolve the claim at issue, and therefore the class could not satisfy the “commonality” prerequisite for class certification.

The Court’s five conservative members agreed with Wal-Mart.27 The Court held that to meet the commonality requirement, the “common contention [] must be of such a nature that it is capable of classwide resolution—which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke.”28 Because the employment policy at issue involved “allowing discretion by local supervisors over employment matters,” it did not even raise a single common question necessary for all classes.29 The partial dissent countered that the single question was lying in plain sight—a uniform delegation of discretion over pay and promotion throughout all stores.30 Whether these questions predominate over individual questions was a determination for another day,31 but according to the dissent, there was no dispute that a single common question existed.

The impact of Wal-Mart Stores, Inc. v. Dukes has been undeniably favorable to employers. Clearly, the decision directly impacted Wal-Mart by avoiding billions of dollars in potential liability from 1.5 million plaintiffs. But, the decision also broadly impacted employment class actions. For example, “the year before the Dukes decision, the top 10 settlements totaled $346 million; in 2012, the first year after Dukes, that total plummeted 87 percent, to $45 million.”32 Anecdotal evidence also supports this view. As ProPublica noted:

Another measure, lawyers representing women and minorities say, is the drop-off in new employment discrimination class-action lawsuits being filed. Before Dukes, it was normal to see 25 or 30 such cases every year, said Jocelyn Larkin, executive director of the Impact Fund, a law firm/national litigation resource center based in Berkeley, Calif., which helped bring the Wal-Mart suit in 2001. Now, Larkin said, the number of new cases is closer to 10 or 12 a year.33

Finally, Wal-Mart Stores, Inc. v. Dukes may also be beneficial to employers by reducing their legal fees. A survey of 395 large companies revealed that they collectively spent approximately $642 million on legal fees to defend labor and employment class actions in 2019.34 To the extent that Wal-Mart reduced the number of class actions, it may also have reduced resources that otherwise would have been devoted to defending these class actions.35

B. Motions to Dismiss: From Conceivable to Plausible

Under the Roberts Court, claims that proceed to litigation are far more likely to be dismissed. While this helps all defendants, and theoretically harms employers who sue former employees, the overall impact clearly benefits employers who are much more likely to be defendants rather than plaintiffs.

For decades, a complaint could only be dismissed if there were no set of facts that could entitle the plaintiff to relief. In other words, a claim would proceed if it was at least conceivable. In 2007, the Roberts Court held that an antitrust claim must be plausible, rather than conceivable, raising the standard for stating a claim.36

The Court then substantially expanded and refined this concept two years later in Ashcroft v. Iqbal.37 The majority employed a two-step process: (1) judges must distinguish factual allegations from legal conclusions because only factual allegations are assumed to be true; and (2) in determining whether factual allegations plausibly set forth a claim, the court must use its “judicial experience and common sense,” relatively vague terms.38 The dissent—which included Justices Breyer and Souter, who voted in favor of the plausibility standard just two years earlier—argued that analyzing assertions in isolation to determine whether they were factual or legal incorrectly applied the rule.39

The impact of the change in standard has been noticeable, particularly for labor and employment claims. Theoretically, Iqbal should not have had much of an impact since judges routinely grant leave to amend, which allows the plaintiff to remedy deficiencies in a complaint. In reality, plaintiffs often fail to properly amend for a variety of reasons leading to a higher effective dismissal rate. For counseled employment discrimination cases, the effective dismissal rate was 7% higher right after Iqbal than before it, meaning that approximately 1 in 14 employment discrimination cases ended earlier without relief due to Iqbal.40 Similarly, there was a “significant increase” in the rate of dismissal for pro se litigants in employment discrimination cases.41

While Iqbal was not an employment law case, there is no doubt that it has significantly benefited employers by permitting them to dismiss claims more easily.

II. Substantive Claims

The Roberts Court has also generally made it easier for employers to prevail on substantive claims if those claims proceed to summary judgment or trial. The Court has largely limited the availability of relief under the Age Discrimination in Employment Act (“ADEA”), Title VII, and the Fair Labor Standards Act (“FLSA”).42

A. Raising the Standard of Proof: Gross and Nassar

By a 5-4 vote in two different cases, the Roberts Court made it more difficult for plaintiffs to prove that discrimination occurred under the ADEA and Title VII.43

First, in Gross v. FBL Financial Services, Inc., the Court held that a plaintiff bringing a disparate treatment age discrimination claim under the ADEA must prove that age was the “but-for” cause of the challenged adverse employment action rather than merely a “motivating factor.”44 For example, if an employer terminates an employee both because that employee is often late to work and because of that employee’s age, the employee will not have a viable claim unless the employee can demonstrate that age rather than (or in addition to) lack of punctuality caused the termination. Age must have a “determinative influence”45 on the termination decision. By contrast, if the Court had adopted the “motivating factor” standard, the employee would have a viable claim, subject to an employer’s affirmative defense, as long as age played some role in the termination. This difference makes it far more likely that the approximately 15,000 ADEA claims filed every year46 will result in a favorable outcome for employers. Indeed, AARP described the decision as “essentially gutt[ing the] ADEA.”47

Four years later, relying largely on Gross, the Roberts Court again found that “but-for” causation was the appropriate standard for Title VII retaliation claims. In University of Texas Southwestern Medical Center v. Nassar,48 by a 5-4 majority, the Roberts Court found retaliation claims require “but-for” causation, even though discrimination claims under Title VII only require that a protected class was a “motivating factor” in an employee’s adverse employment action.49 In other words, it is easier for an employee to prove discrimination than it is to prove retaliation for complaining about the discrimination. Nassar is arguably even more significant than Gross because there are approximately 30,000 Title VII retaliation claims filed with the EEOC every year.50

B. Limiting Supervisor Liability: Vance

The Roberts Court has also limited claims against employers by circumscribing the definition of a supervisor.51

An employer may be held vicariously liable under Title VII when a supervisor or coworker engages in discrimination against an employee.52 But the rules differ. For coworkers, the employer is only liable if it was negligent in controlling workplace conditions.53 This is a high standard for an employee to meet but reflects the fact that coworkers have limited control in the workplace over one another. Coworkers cannot typically change each other’s hours, pay, or other aspects of employment.54

A lower standard, however, applies for supervisory liability because supervisors have far greater control over an employee.55 If a supervisor’s harassment results in a tangible employment action (i.e. termination, reassignment, etc.), then the employer is strictly liable.56 Regardless of the employer’s intent, it will be liable for its supervisor’s unlawful activity. If there is no tangible employment action, the employer can escape liability for a supervisor’s harassment if it (1) exercised reasonable care to prevent and correct harassing behavior and (2) the plaintiff failed to take advantage of preventive or corrective opportunities that the employer provided.57

Overall, employers would much prefer to face claims brought against coworkers than supervisors because they are less likely to be held liable. But it was not always clear who was a “supervisor” and who was a coworker. The EEOC issued guidance explaining that a supervisor is either: (1) “an individual authorized ‘to undertake or recommend tangible employment decisions affecting the employee,’ including ‘hiring, firing, promoting, demoting, and reassigning the employee’; or (2) an individual authorized ‘to direct the employee’s daily work activities.’”58

By a 5-4 vote in Vance v. Ball State University, the Roberts Court rejected the second definition, holding that a supervisor is only someone who is empowered to take tangible employment actions against a worker.59 The majority thoroughly rejected the second definition—an individual who directs the employee’s daily work activities—describing it as “nebulous,” “vague,” and a “study in ambiguity.”60

Accordingly, employees who control day-to-day schedules and assignments but lack the ability to hire or fire are considered coworkers rather than supervisors, meaning employers are far less likely to be held liable for the actions of these individuals. While plaintiffs may still prevail by demonstrating that the employer was negligent in controlling these coworkers, this is more difficult to prove.

Vance undoubtedly tilted the playing field in favor of employers. One analysis conducted approximately one year after Vance concluded that 43 sexual harassment cases were dismissed because a worker did not meet the more restrictive definition of “supervisor.”61 While there does not appear to be any comprehensive study for all types of discrimination, it is safe to assume that many claims that might have proceeded to trial under the EEOC’s standard have been dismissed as a result of Vance. Further, settlement payouts—which are far more prevalent than trials62—are likely to be lower under the employer-friendly standard in cases where the parties dispute whether the relevant actor was a supervisor.

But it is also important to contextualize the limited importance of Vance and litigation more generally with respect to unlawful workplace behavior. The most common response to harassment in the workplace is to do nothing.63 The least common response is to take any formal action.64 Approximately 70% of individuals who experience harassment do not consult a supervisor, manager, or union representative about the harassing conduct.65 For certain types of harassment, the numbers are even more remarkable—unwanted physical touching is reported only 8% of the time.66 On average, 87%—94% of individuals who experience harassment do not file a formal complaint.67

This is not because every potential plaintiff knows the legal pitfalls associated with defining a worker as a supervisor rather than a coworker. Rather, it reflects the daily reality of pursuing a claim that may take years to resolve.68 Indeed, despite the fact that retaliation is unlawful, one study found that 75% of employees who spoke out against workplace mistreatment faced some form of retaliation.69 Other studies have found that sexual harassment reporting “is often followed by organizational indifference or trivialization of the harassment complaint as well as hostility and reprisals against the victim.”70 As a result, one researcher concluded that the most reasonable course of action for a victim to take is to avoid reporting harassment.71

Employers may conclude that this is all quite unfortunate but not particularly relevant to them. But adopting such a blinkered outlook is unwise, even if self-interest is the only concern. When a claim proceeds to trial, the jury is likely to be composed of others who may be sympathetic to a plaintiff based on their own negative yet unreported experiences in the workplace. Similarly, an employer who does not address harassment in the workplace may suffer severe reputational harm far costlier than any litigation when an undercurrent of resentment transforms into an outpouring of complaints on social media (see e.g., Google).72 In other words, just because an employer may not be legally liable for certain activity in the workplace does not mean that that activity is irrelevant for them.

C. FLSA Exemptions: Smithkline and Encino Motorcars

During the Great Depression, President Roosevelt signed the FLSA which established a minimum wage and overtime pay for certain workers who work more than 40 hours per week, and largely outlawed child labor.73 Exempt workers need not be paid the minimum wage or overtime.74 But, the vast majority of U.S. workers are governed by the FLSA—of the approximately 160 million wage and salary workers in the United States, only 22.5 million satisfy the main exemptions for Executive, Administrative, and Professional workers.75

The Roberts Court has limited the reach of the FLSA both for specific occupations and for all workers who believe they are covered by the Act’s overtime and minimum wage provisions.76

In Christopher v. SmithKline Beecham Corp.,77 a 5-4 majority held that pharmaceutical sales representatives qualify for the “outside salesman” exemption meaning that they are not entitled to the minimum wage or overtime. In doing so, the Court rejected guidance issued by the DOL, which would ordinarily be entitled to deference.78 Ultimately, the Court concluded that pharmaceutical sales representatives met the outside salesman exemption even though they do not sell anything—by law, only doctors may prescribe pharmaceutical drugs. It found that the legal prohibition on selling pharmaceutical drugs was a “technicalit[y]” that did not justify excluding sales representatives from the outside salesman exemption.79

This saved the pharmaceutical industry “billions of dollars in potential liability” for overtime.80 As a result of the ruling, many of the 65,000 pharmaceutical sales representatives81 in the U.S. are ineligible for this pay.

Six years later, the court issued Encino Motorcars LLC. v. Navarro, where it concluded by a now-familiar 5-4 majority that many of the nation’s 45,00082 car service advisors are similarly exempt under the FLSA because they are “primarily engaged in . . . servicing automobiles.” The dissent took issue noting car service advisors do not repair automobiles but instead suggest additional services and address complaints from car owners.83

In addition, the majority found that all FLSA exemptions should be given a “fair reading” rather than “narrowly construed,” as they had been under Supreme Court precedent for nearly six decades.84 This brief note essentially instructed the lower courts to review every misclassification claim under the FLSA with a more employer-friendly lens and it has already been cited by several courts of appeal.85

D. Time-Barred Claims: Ledbetter

In Ledbetter v. Goodyear Tire & Rubber Co.,86 a female supervisor alleged that she had suffered gender pay discrimination for approximately twenty years and sued under Title VII to recover. The supervisor, Lilly Ledbetter, claimed she received discriminatory evaluations early in her career, which set the course for her salary as time passed. As a result, Ledbetter was ultimately paid 15%-40% less than her fifteen male counterparts. Further, because the employer, Goodyear, kept salaries confidential, Ledbetter did not learn of the alleged pay disparity for years.

Goodyear argued that claims under Title VII must be filed with the EEOC within 180 days after the alleged unlawful employment practice occurred.87 Accordingly, even if Ledbetter could prove that she was discriminated against, she could only recover for discrimination that occurred in the final few months of her employment, not for the entirety of her two-decade career. Ledbetter countered that each paycheck she received reflected discrimination, and therefore each paycheck essentially restarted the clock for filing.

By a 5-4 vote, the Roberts Court sided with Goodyear.88 It concluded the pay-setting decision that occurred at the beginning of Ledbetter’s career was a “discrete act,” and therefore, “the period for filing an EEOC charge begins when the act occurs” regardless of whether Ledbetter had notice of the alleged discrimination.89 The liberal wing dissented, describing the majority’s interpretation as “cramped” and urging Congress to amend Title VII “to correct [the] Court’s parsimonious reading.”90

Less than two years later, a new Congress passed legislation effectively overturning the Court’s decision.91 President Obama signed the bill, which was the first piece of legislation during his administration.92 Accordingly, while this decision favored employers, its impact was limited by subsequent legislative action.

E. Expanding Protections for Sexual Orientation and Gender Identity: Bostock

One notable exception to the Roberts Court’s general pro-employer jurisprudence was Bostock v. Clayton County93 where the Court held that Title VII protects sexual orientation and gender identity. The Court interpreted Title VII’s prohibition on discrimination “because of . . . sex” to extend beyond discrimination based on biological differences between males and females.94 Rather, when an employer discriminates against an employee because that employee is homosexual or transgender, the employer has violated Title VII. Further, Justices Gorsuch and Roberts voted for this outcome, despite their traditional membership in the Court’s conservative wing.95 Still, despite these two arguably anomalous votes, the other seven justices voted as predicted—liberals for workers and conservatives for employers.

The ruling extends protections to approximately 3.6 million LGBT employees who live in states that did not previously prohibit discrimination based on sexual orientation and gender identity.96

III. Labor Law Issues

The Roberts Court has also altered labor law in ways that ultimately benefit certain unionized employers. Most significantly, the Court has made union dues optional for employees represented by public sector unions, almost certainly reducing their clout vis-à-vis public sector employers.97 The Court has also upheld mandatory arbitration provisions for statutory claims in collective bargaining agreements and limited union organizer access rights to farmworkers in California.98

A. Making Dues Optional in Public Sector Unions

The Roberts Court overturned a 41-year-old precedent allowing public sector unions in states with fair share fee laws to collect certain union dues from all represented workers.99 Previously, 5.9 million state and local employees in 22 states100 were required by law to pay agency shop fees, consisting of union expenditures on (1) collective bargaining, (2) contract administration, and (3) grievances.101 Employees, however, were never required to subsidize a public sector union’s political activity and could refrain from paying fees for such political activities.102

But, in 2018, the five conservative justices held in Janus v. American Federation of State, County, and Municipal Employees, Council 31 that any statute mandating that public sector employees pay any amount of union dues violated the First Amendment’s right to refrain from compelled speech.103 According to the Court, even grievances are of “substantial public concern,” and therefore a requirement to fund these services violates an employee’s constitutional right to not speak.104 As a result, public sector union dues are now optional in every state.

Unsurprisingly, this decision has led to a decline in union membership and revenues.105 In 2018 alone, two of the largest public sector unions, AFSCME and SEIU, lost 98% and 94% respectively of their agency fee payers (those represented workers who already objected to paying for political expenditures).106 This amounted to approximately 210,000 agency fee payers, who typically pay 75%-85% of full union dues.107 In New York alone, public-sector unions were projected to lose $100 million in agency fee revenue, approximately 10% of all dues.108

B. Waiving the Right to Bring Statutory Claims in Federal Court

By a similar 5-4 conservative majority, the Roberts Court also held arbitration clauses in Collective Bargaining Agreements (CBAs) that waive an employee’s right to bring statutory claims in federal court are enforceable.109

In 14 Penn Plaza LLC v. Pyett,110 the Court held that a union member was not permitted to pursue a statutory claim for discrimination in federal court under the ADEA because that member’s CBA had an arbitration clause which clearly and unmistakably mandated arbitration. As a result, unions and employers may agree that represented workers may not bring claims for alleged violations of Title VII, the ADEA, and other statutory violations in court. The Court did not decide what happens when a union fails to bring a statutory grievance in arbitration and the worker is therefore barred from seeking recovery.

14 Penn Plaza LLC permits arbitration in unionized workplaces where workers have far more power vis-à-vis their employer than workers in non-unionized workplaces. A union can always reject an arbitration clause in negotiations; it is much more difficult for an individual employee to do so as a condition of employment. Ultimately, while it is difficult to empirically assess 14 Penn Plaza LLC’s impact, it likely favors employers because employers tend to favor arbitration over litigation for cost, efficiency, and privacy reasons. There is no guarantee that a given union will agree to arbitrate statutory claims, but employers could not viably negotiate over the issue if not for the Roberts Court’s holding that these clauses are enforceable.

C. Access to Farmworkers in California

Earlier this year, the Court ruled a California regulation, granting access rights to union organizers for a limited duration, amounted to a per se physical taking under the Fifth and Fourteenth Amendments, therefore, requiring just compensation.111 The vote was divided along ideological lines, 6-3, with the Court’s newest member, Justice Coney Barrett, siding with the conservative wing.112

In Cedar Point Nursery v. Hassid,113 the Court analyzed a California regulation that grants unions a “right to take access” to an agricultural employer’s property for up to three hours per day for 120 days per year.114 The regulation was considered a “major achievement of the farmworkers’ movement led by Cesar Chavez in the 1970s, which had argued that allowing organizers to enter workplaces was the only practical way to give farmworkers, who can be nomadic and poorly educated, a realistic chance to consider joining a union.”115 Justice Roberts, writing for the majority, found that the regulation limited a “treasured” right of property ownership—the right to exclude—by permitting third party union organizers to enter an employer’s property without consent.116 In the majority’s view, this amounted to a physical appropriation of property, which requires compensation. The dissent concluded that the regulation does not appropriate anything but instead regulates property by temporarily limiting an owner’s right to exclude, which does not require just compensation.117

The implications of the ruling are significant, even if a bit unclear at this early juncture. Union organizers in California may still enter the property owner’s land, but the State of California must pay the owner just compensation every time this occurs, which could prove burdensome.

But the decision undoubtedly favors agricultural employers in California by making union organizing more difficult. In part because agricultural workers are excluded from the National Labor Relations Act,118 they are extremely unlikely to be unionized. Nationally, approximately 1% of agricultural workers are represented by a union.119 Further, even though the Court’s decision only directly impacts the California agricultural industry, California is home to approximately 410,000120 of the approximately 900,000121 agricultural workers in the United States. In short, Cedar Point Nursery makes it more difficult to unionize an already challenging sector, favoring employers.

IV. Conclusion

Justice Scalia believed that the justices’ voting patterns reflected preexisting judicial philosophy rather than partisan preference. In an interview with Piers Morgan, Scalia explained:

MORGAN: Do you think any of your colleagues act . . . from a politically motivated manner?

SCALIA: Not a single one of them . . . I don’t think any of my colleagues, on any cases, vote the way they do for political reasons. They vote the way they do because they have their — their own — their own judicial philosophy. And they may have been selected by the Democrats because they have that [] particular philosophy or they may have been selected by the Republicans because they have that particular judicial philosophy. But that is only to say that they are who they are. And they vote on the basis of what their own view of the law brings them to believe, not at all because — I mean to — the court is not at all a — a political institution. Not at all. I — I — not a single one of my colleagues. . .122

Supreme Court justices collectively cast 134 votes in the 15 cases discussed in this article. Those cases spanned civil procedure, constitutional law, and statutory interpretation. There is no unifying judicial philosophy—such as originalism or textualism123—that neatly explains why conservative justices would reliably vote in one manner and liberal justices in the opposite manner for these cases. Yet, if all one knew was that conservative justices favor employers and liberal justices favor workers, that person would have correctly predicted 132 of the 134 votes cast (98.5%).124

If judicial philosophy rather than political motivation explained the underlying dynamics, and we assume that judicial philosophy in the abstract is no more likely to favor employers than workers,125 then the Court’s collective votes are the equivalent of flipping a coin 134 times and getting heads 132 times. Statistically, this is virtually impossible, occurring just 1 out of every 2.4 x 1021 times. Even if one assumes that the conservative wing is one ideological bloc and the liberal wing another ideological bloc, rather than a collection of individual votes, the numbers are still startling. In the 15 cases noted here, the blocs voted in unison as one would expect 29 of 30 times. Using the coin flip analogy, this should occur 1 in every 34.6 million times.

Justices Souter and Stevens often voted with the liberal wing despite being appointed by Republican presidents. Focusing on the 11 cases mentioned in this article since they retired, one could have correctly predicted 96 of 98 votes based on the party that nominated the justice (with Democrats favoring workers and Republicans favoring employers), which is also virtually impossible, occurring just 1 out of every 6.53 x 1022 times. Alternatively, analyzing ideological blocs reaches substantially the same conclusion: the Democratic-appointed bloc voted for workers in each of the 11 cases, and the Republican-appointed bloc voted for employers in 10 of the 11 cases. Using the coin flip analogy, this should occur 1 in every 182,361 times.

Indeed, other studies analyzing voting patterns have demonstrated that each Republican appointed justice is “more favorable to business” than each Democratic appointed justice,126 which would be a remarkable coincidence if each justice was truly motivated by judicial philosophy.127

Justices on either divide of the Roberts Court may find it both individually and institutionally expedient to claim that they are merely applying an objective judicial philosophy. At least on labor and employment law, that theory does not appear explanatory. Instead, the sizable percentage of the public that believes that the Court is driven by politics128 may have a point, despite what the justices on that Court may claim.

The result of that pattern is a legal playing field that significantly favors employers relative to the rules that existed prior to 2005. Further, with the confirmation of Justice Coney Barrett, the Court now has six conservative justices, none of whom appear close to retirement based on their age.129 As a result, the Court is likely to continue issuing decisions, perhaps for decades to come, consistent with its recent practice.

a. Scott A. Budow is an associate at Kauff McGuire & Margolis LLP, where he practices labor and employment law. He is also an Adjunct Professor of Law at Fordham University School of Law.

1. The conservative wing of the Roberts Court has included Justices Roberts, Alito, Thomas, Scalia, Kennedy, Gorsuch, Kavanaugh, and Coney Barrett. The liberal wing has included Justices Stevens, Kagan, Souter, Sotomayor, Ginsburg, and Breyer.

2. Adam Liptak, Politics Has No Place at the Supreme Court, Chief Justice Roberts Says, N.Y. Times, (June 30, 2020) [].

3. Matt Sepic, Justice Kagan: U.S. Supreme Court More Complicated Than the Divided Public Knows, MPR News (Oct. 22, 2019, 9:00 AM), [].

4. Sarah Turberville & Anthony Marcum, Those 5-to-4 Decisions on the Supreme Court? 9 to 0 is Far More Common., Wash. Post (June 28, 2018), [].

5. Brianne J. Gorod, The First Decade of the Roberts Court: Good for Business Interests, Bad for Legal Accountability, 67 Case W. Rsrv. L. Rev. 721, 722 (2017).

6. Cynthia Estlund, The Black Hole of Mandatory Arbitration, 96 N.C. L. Rev. 679, 688 (2018). It is also possible that employers win more often because they are repeat players in arbitration—arbitrators may be inclined to favor employers likely to need arbitration services in the future over employees unlikely to need similar services. Repeat employers involved in multiple arbitrations have higher win rates and lower amount awards. It is not clear that these employers necessarily have superior outcomes because of any undue influence. It is plausible that their success may also be explained by greater resources and expertise in dispute resolution procedures. However, when arbitrators and employers are repeatedly paired together, “employees on average have lower win rates and receive smaller damage awards.” Alexander J.S. Colvin, An Empirical Study of Employment Arbitration: Case Outcomes and Processes, 8 J. Empirical Legal Stud. 1, 1 (2011).

7. Alexander J.S. Colvin, The Growing Use of Mandatory Arbitration, Econ. Pol’y Inst. (Apr. 6, 2018), [].

8. Unchecked Corporate Power, Econ. Pol’y Inst., [].

9. 138 S.Ct. 1612 (2018).

10. Id. at 1632.

11. 563 U.S. 333 (2011).

12. Concepcion provides a strong rebuke to those who suggest that the results in these cases merely reflect divergent judicial philosophies. The Court’s five conservatives voted to limit federalism by rejecting state defenses to federal law, while the four liberals voted to expand states’ rights by doing the opposite.

13. U.S. Const. art. VI, cl. 2.

14. 570 U.S. 228 (2013).

15. 29 U.S.C. § 157.

16. 138 S.Ct. 1612, 1619 (2018).

17. Id.

18. Colvin supra note 7.

19. Estlund supra note 6, at 696.

20. 29 U.S.C. § 203.

21. Id. at 702–703.

22. 564 U.S. 338 (2011).

23. Id. at 367.

24. Fed. R. Civ. P. 23(a).

25. Fed. R. Civ. P. 23(b).

26. 564 U.S. at 346.

27. Id. at 367.

28. Id. at 350.

29. Id. at 355.

30. Id. at 369 (Ginsburg, J., concurring in part and dissenting in part).

31. The partial dissent agreed that the class should not have been certified as a (b)(2) class and would have reversed the Ninth Circuit on this point. The dissent, however, also believed that the class may have been certifiable as a (b)(3) class and would have remanded that decision to the lower courts, allowing the suit to continue if plaintiffs could demonstrate that they satisfy this type of class.

32. Nina Martin, The Impact and Echoes of the Wal-Mart Discrimination Case, Propublica (Sept. 27, 2013, 9:53 AM), [].

33. Id.

34. Class Action Survey, Carlton Fields (2019), [].

35. But, Wal-Mart appears to have also led Plaintiffs “to file smaller regional class actions across the country to get around the notion that the problem with its nationwide class was primarily its size.” Michael Selmi & Sylvia Tsakos, Employment Discrimination Class Actions After Wal-Mart v. Dukes, 48 Akron L. Rev. 803, 830 (2015). If Wal-Mart led to a net increase of lawsuits by creating more regional class actions than it reduced national class actions, then litigation spending may have conceivably increased as a result of defending more suits overall.

36. Bell Atl. Corp. v. Twombly, 550 U.S. 544 (2007).

37. 556 U.S. 662 (2009) (a 5-4 split along familiar ideological lines).

38. Arthur R. Miller, From Conley to Twombly to Iqbal: A Double Play on the Federal Rules of Civil Procedure, 60 Duke L. J. 1, 23–24 (2010).

39. Iqbal, 550 U.S. at 688 (Souter, J., dissenting).

40. Alexander A. Reinert, Measuring the Impact of Plausibility Pleading, 101 Va. L. Rev. 2117, 2162 (2015).

41. Id. at 2151.

42. 29 U.S.C. §§ 621–634; 42 U.S.C. § 2000e; 29 U.S.C. §203.

43. Gross v. FBL Fin. Servs., Inc., 557 U.S. 167 (2009); Univ. of Tex. Sw. Med. Ctr. v. Nassar, 570 U.S. 338 (2013).

44. 557 U.S. 167, 180 (2009).

45. Hazen Paper Co. v. Biggins, 507 U.S. 604, 610 (1993).

46. Age Discrimination in Employment Act (Charges Filed with EEOC) (Includes Concurrent Charges with Title VII, ADA, EPA, and GINA) FY 1997 – FY 2020, U.S. Equal Emp. Opportunity Comm’n, (last visited Sept. 7, 2021) [].

47. Joe Kita, Workplace Age Discrimination Still Flourishes in America, AARP (Dec. 30, 2019), [].

48. 570 U.S. 338 (2013).

49. 42 U.S.C. § 2000e-2(m).

50. Charge Statistics (Charges Filed with EEOC) FY 1997 Through FY 2020, U.S. Equal Emp. Opportunity Comm’n, (last visited Sept. 7, 2021) [].

51. Emily Hobbs-Wright, Supreme Court Limits Definition of Supervisor for Employer Liability in Workplace Harassment Claims, Holland & Hart (June 25, 2013), [].

52. This is because Title VII defines “employer” to include “agents.” 42 U.S.C. § 2000e(b).

53. Hobbs-Wright, supra note 51.

54. Id.

55. Id.

56. Id.

57. Faragher v. Boca Raton, 524 U.S. 775, 807 (1998); Burlington Industries, Inc. v. Ellerth, 524 U.S. 742, 765 (1998).

58. Vance v. Ball State Univ., 570 U.S. 421, 451, (2013) (Ginsburg, J., dissenting) (citing EEOC, Guidance on Vicarious Employer Liability For Unlawful Harassment by Supervisors, 8 BNA FEP Manual 405:7651 (Feb. 2003)).

59. Vance, 570 U.S. at 431.

60. Id. at 431, 442.

61. Bryce Covert, EXCLUSIVE: 43 Sexual Harassment Cases That Were Thrown Out Because of One Supreme Court Decision, ThinkProgress (Nov. 24, 2014, 4:24 PM), [].

62. Theodore Eisenberg & Charlotte Lanvers, What is the Settlement Rate and Why Should We Care?, 6 J. Empirical Legal Stud. 111, 128 (2009).

63. U.S. Equal Emp. Opportunity Comm’n, Select Task Force on the Study of Harassment in the Workplace (2016), [].

64. Id.

65. Id.

66. Id.

67. Id.

68. Id.

69. Id.

70. Id.

71. Id.

72. Lauren Feiner, Google Employees to Launch Social Media Campaign on Sexual Harassment Policies After Previous Protests Brought ‘No Meaningful Gains for Worker Equity’, CNBC, (Jan. 14, 2019, 1:37 PM) [].

73. 29 U.S.C. § 203.

74. Id.

75. 29 C.F.R. § 541 (2016).

76. See Encino Motorcars LLC v. Navarro, 138 S.Ct. 1134 (2018).

77. 567 U.S. 142, 147 (2012).

78. The majority withheld deference because Smithkline Beecham Corporation dba GlaxoSmithKline could be subject to “massive liability” without appropriate notice, since the DOL’s interpretation was issued subsequent to much of the allegedly unlawful conduct at issue.

79. 567 U.S. at 165 n.23. The dissent took issue with this analysis, rhetorically asking “Where in this process does the detailer sell the product?” 567 U.S. at 173 (Breyer, J., dissenting) (emphasis in original).

80. James Vicini, U.S. Top Court Rules for Glaxo on Overtime Pay, Reuters (June 18, 2012. 9:31 AM), [].

81. Emma Court, Drug Sales Reps Stow the Sample Case and Turn to Zoom, Bloomberg (June 12, 2020), [].

82. 138 S. Ct. 1134; Colton Long & Noah Finkel, Encino Motorcars, LLC v. Navarro: SCOTUS Puts the Brakes on Faulty FLSA Construction Language, Seyfarth Shaw (Apr. 2, 2018), [].

83. Encino Motorcars, LLC v. Navarro, 138 S. Ct. 1134, 1148 n.7 (2018) (Ginsburg, J., dissenting).

84. Navarro, 138 S. Ct. at 1142.

85. See Carley v. Crest Pumping Techs., LLC, 890 F.3d 575, 579 (5th Cir. 2018) (“The Supreme Court recently clarified that courts are to give FLSA exemptions ‘a fair reading,’ as opposed to the narrow interpretation previously espoused by this and other circuits.”).

86. 550 U.S. 618 (2007).

87. Or, up to 300 days, depending on the State. See Time Limits For Filing A Charge, U.S. Equal Emp. Opportunity Comm’n, (last visited Sept. 7, 2021) [].

88. Ledbetter, 550 U.S. at 621.

89. Id.

90. Id. at 661 (Ginsburg, J., dissenting).

91. Lilly Ledbetter Fair Pay Act of 2009, Pub. L. No. 111–2, § 3, 123 Stat. 5, 5–6 (codified as amended at 42 U.S.C. §2000e–5(e)).

92. Megan Slack, From the Archives: President Obama Signs the Lilly Ledbetter Fair Pay Act, White House: President Barack Obama (Jan. 30, 2012, 1:41 PM), [].

93. 140 S. Ct. 1731 (2020).

94. Id. at 1737.

95. Id. Justice Gorsuch authored the majority opinion.

96. Christy Mallory, Luis A. Velasquez & Celia Meredith, Legal Protections for LGBT People After Bostock v. Clayton County, Williams Inst. (Aug. 2020), [].

97. Janus v. Am. Fed’n of State, Cnty., & Mun. Emps., Council 31, 138 S.Ct. 2448 (2018).

98. 14 Penn Plaza LLC v. Pyett, 556 U.S. 247 (2009).

99. Abood v. Detroit Bd. of Educ., 431 U.S. 209 (1977).

100. Daniel DiSalvo, Public-Sector Unions After Janus: An Update, Manhattan Inst. (Feb. 14, 2019), [].

101. Abood v. Detroit Bd. of Educ., 431 U.S. 209, 226 (1977).

102. DiSalvo, supra note 100.

103. 138 S. Ct. 2448, 2460 (2018).

104. Id. at 2477.

105. See sources cited infra notes 106–107.

106. Steve Delie, How the Janus Decision Changed America, The Hill (June 26, 2020. 8:00 AM), [].

107. Robert Iafolla, Mass Exodus of Public Union Fee Payers After High Court Ruling, Bloomberg (Apr. 5, 2019, 3:10 PM), [].

108. DiSalvo, supra note 100.

109. 14 Penn Plaza LLC v. Pyett, 556 U.S. 247 (2009).

110. Id.

111. Cedar Point Nursury v. Hassid, 141 S.Ct. 2063 (2021).

112. Id.

113. Id..

114. Id. at 2069.

115. Adam Liptak, Supreme Court Rules Against Union Recruiting on California Farms, N.Y. Times (June 23, 2021), [].

116. Cedar Point Nursery, 141 S.Ct. at 2072.

117. Id. at 2081 (Breyer, J., dissenting).

118. 29 U.S.C. § 152(3) (“The term ‘employee’ . . . shall not include any individual employed as an agricultural laborer . . . .”).

119. Erin Mulvaney, New York Farmworkers Get Labor Rights but Legal Battle Looms (2), Bloomberg (Jan. 7, 2020, 4:33 PM), [].

120. Philip L. Martin, Brandon Hooker, Muhammad Akhtar & Marc Stockton, How Many Workers are Employed in California Agriculture?, 71 Calif Agric. 30 (2017), [].

121. Agricultural Workers, U.S. Bureau of Lab. Stats., (last visited Sept. 8, 2021) [] (902,900 workers in 2019).

122. Transcripts—Piers Morgan Live, CNN (July 18, 2012, 9:00 PM), [].

123. One philosophical issue that might logically explain different outcomes is agency deference. Conservative justices are more likely to believe in a limited administrative state where agency decisions receive less deference for separation of powers reasons. This judicial philosophy understandably leads to outcomes that tend to favor employers, who tend to benefit from challenges to rules that have been imposed upon them by agencies. But, none of the cases mentioned here involved agency deference (in Smithkline Beecham Corp., the Court refused to grant agency deference due to inadequate notice, not because agencies should not be accorded deference out of separation of powers concerns). Further, a cynic could interpret this judicial philosophy as fundamentally reflecting a political desire to rein in the administrative state that “liberal” lawmakers built starting with FDR’s New Deal.

124. This does not include Twombly, which was only briefly mentioned and is far less significant than Iqbal. In Twombly, Justices Breyer and Souter cast votes for the pro-employer outcome, while subsequently dissenting when the Court further extended these principles in Iqbal. Including Twombly would have reduced the percentage slightly, from 98.5% to 97.2%.

125. There is no such thing as a judicial philosophy that explicitly favors workers or employers. See e.g., Noah Feldman, Democrats’ Misguided Argument Against Gorsuch, Bloomberg (Mar. 15, 2017, 7:30 AM), [] (“But the thing is, siding with workers against employers isn’t a jurisprudential position. It’s a political stance.”).

126. Lee Epstein, William M. Landes & Richard A. Posner, When It Comes to Business, the Right and Left Sides of the Court Agree, 54 Wash. U. J. L. & Pol’y 33, 36, 53 (2017) (noting that “the conservatives (all Republican appointees) on the Roberts Court are more favorable to business than the liberals (all Democratic appointees);” Appendix B, Closely Divided Cases).

127. The Court could also adopt a more transparent approach to how it hears cases in the first place. To hear a case, four of the nine Justice must vote to grant a writ of certiorari. Supreme Court Procedures, U.S. Cts., (last visited Sept. 8, 2021) []. These votes are not public. If the Court wanted to counter the widespread perception that it is acting in an outcome-oriented, activist manner, it could release a record of the votes on writs of certiorari. But, doing so may undermine its claim of even-handed neutrality and bolster the case of its critics if the votes on writs of certiorari substantially mirror the familiar ideological divide reflected in its final decisions. That would seem to indicate that Justices are selecting cases to hear because they are confident that their favored outcome will prevail before having even heard argument on the case.

128. Carl Hulse, Political Polarization Takes Hold of the Supreme Court, N.Y. Times (July 5, 2018), [].

129. The Court’s oldest conservatives, Justices Thomas and Alito, are seventy-three and seventy-one respectively. Justice Roberts is sixty-six, while Justices Gorsuch and Kavanaugh are in their fifties, and Justice Coney Barrett is approaching fifty.

The full text of this Article is available to download as a PDF.