“At-will” employment is sometimes shorthanded as employers’ rights to fire employees (and employees’ right to quit) for a bad or arbitrary reason, or for no reason at all. Among the bad or arbitrary reasons that employers sometimes fire workers: something the worker said, or something they didn’t say. Employees have been fired, often without legal recourse, for criticizing their companies on social media, for running for office, or even for having a bumper sticker supporting a political candidate whose election the boss opposes. The freedom of speech that so many Americans valorize is in practical effect illusory for many American workers.
This report traces the legal rules governing freedom of speech at work. Following a summary that emphasizes the scope of the problem and gives examples, it begins by discussing the background rules of at-will employment, which establish that employers may generally terminate workers for what they say. This rule has its limits—for example, employers may not fire workers in contravention of a state’s explicit public policy—but judges tend to apply these exceptions in a patchy and inconsistent fashion. Further, because the First Amendment does not constrain private actors, private-sector workers cannot fall back on the constitution at all; even public-sector employers are often free to fire or discipline workers for their speech.
Beyond common law rules, the report also discusses federal, state, and local statutes that protect certain types of employee speech. These laws tend to apply only to specific subjects and manners of expression. For example, the National Labor Relations Act protects employees’ conversations about their working conditions—but only as long as those conversations occur at the right time, in the right place, and in the right manner. For example, among other limits, the NLRA protects only those conversations or meetings that occur during “nonwork time,” and the Trump NLRB has recently held that the NLRA does not protect employees’ use of their work-issued email addresses. Likewise, some states and localities forbid employers from retaliating against employees for their political views. But each of those laws has serious limitations in coverage, enforcement, or both. Worse, employers sometimes challenge even limited protections for workers’ expression on the grounds that those protections violate the employer’s own rights under the First Amendment.
Finally, some workers have meaningful contractual protections that curb the effects of the at-will doctrine, including as it applies to their speech and expression. But workers cannot achieve these protections without either individual or collective power, both of which have eroded for many workers over the last 80 years. The result is that one real source of protection for workers who speak out—collective bargaining agreements in which employers agree to discipline or fire workers only for good cause—are increasingly out of reach, especially for private-sector workers.
This report aims to help readers understand the legal landscape that effectuates the “freedom of speech” at work. It shows how employers have come to monopolize that freedom for themselves, and why workers experience speech control instead of speech freedom.
Americans are divided on any number of fundamental questions, but a recent poll shows broad support for free speech: “Seventy-one percent of Americans think that people should be able to say what they want without state or government censorship” (Gray 2016). But many U.S. workers are not really free to speak their minds, and it’s not the government censoring them but their bosses, even when the workers are off duty and even when they are outside of work. This paper discusses the reasons for—and limits of—employer control over employees’ expression.
In 49 U.S. states—every one except Montana—the default setting for private-sector, nonunion employment is “at-will,” which is often described as employment from which employees can be fired for a bad reason, an arbitrary reason, or no reason at all.1 One employment law casebook describes the at-will rule like this: “the employer [is] free to impose any conditions of employment, to discharge an employee at any time for any reason, and to effect the discharge in virtually any manner” (Rothstein, Liebman, and Yuracko 2015).2 The philosopher Elizabeth Anderson has emphasized the consequences of employers’ freedom under the at-will rule, characterizing private-sector employment as a form of “private government,” and a dictatorial one at that. Anderson rightly emphasizes that employers often have the power to fire employees because of their out-of-work speech and other off-duty activities—a reality of which many employees are unaware. Of course, private employers are not the same as government—Anderson points out that government can often impose much harsher penalties than employers can.3 Likewise, while workers often face practical barriers to changing jobs, it is still generally easier to change jobs than to immigrate to a new country. At the same time, “private governments impose a far more minute, exacting, and sweeping regulation of employees than democratic states do in any domain outside of prisons and the military” (Anderson 2017, 63). In other words, employers’ freedom to fire employees for bad or arbitrary reasons is in essence a freedom to control employees’ behavior in significant ways, even when they are off the clock.
The at-will default is nominally evenhanded, in the sense that employees are also free to quit their jobs for any reason, at any time, and in any manner. But, as other papers in the Unequal Power initiative show, workers operate under a host of constraints that limit their practical ability to quit. A weak social safety net, uncertainty about how easy it will be to find another job, and family obligations that prevent someone from moving for work can all leave employees stuck working for controlling or abusive employers—and the next employer might be just as bad (Edwards 2022). The rarity of full employment makes quitting and switching jobs problematic, especially for Blacks and workers lacking a college credential (Mishel 2022). And even when wages are cut workers do not quit to the extent that economic theory posits, indicating that employer power is pervasive (Naidu and Carr 2022).
Employees’ dependence on their employers for life’s necessities, coupled with the at-will default, means that employers can often exercise vast authority over their employees’ lives, including their speech and association. Consider the following examples:
- Unionized workers were compelled to choose between attending a Trump rally, using a day of paid time off, or staying home without pay; workers who chose to attend were apparently cautioned not to do or say anything that could be “viewed as resistance” (Allyn 2019). Of course, these choices might have been even tougher if the workers had not been protected by a union contract; in 2012, a group of nonunion mineworkers said they were ordered to “give up a day’s pay” in order to attend a Mitt Romney campaign event (Bradford 2017).
- Workers have been fired for expressing support for the political candidates not supported by their bosses. For example, a worker reported that she was fired because she had a Kerry-Edwards bumper sticker; her boss was a Bush supporter (Noah 2004). Juli Briskman, the marketing analyst who in 2017 communicated her disdain for Trump administration policies by flipping off a presidential motorcade, was fired, reportedly for violating a “company policy banning obscene content on social media” (Cassens Weiss 2018). Briskman sued, but her case was dismissed because Virginia, like nearly all states, is an at-will jurisdiction.
- Workers may also face pressure from their bosses to donate to or vote for a particular candidate. Mitt Romney famously encouraged this practice, telling business owners to “make it very clear to your employees what you believe is in the best interest of your enterprise and therefore their job and their future in the upcoming elections.” He noted that there was “nothing illegal about you talking to your employees about what you believe is best for the business, because I think that will figure into their election decision, their voting decision, and of course doing that with your family and your kids as well” (Phelan 2012). In fact, one in four private-sector employees said in a 2015 survey that they received political messages or requests from their bosses (Hertel-Fernandez 2018).
- A worker was required to take part in Bible study during working hours and was told he would be fired if he refused. This was likely illegal, but the worker, who had served time in prison, was “fearful that he wouldn’t be able to find other work,” so he “stuck with the weekly, hourlong Bible study sessions for six months” (Green 2018).
- During the COVID-19 pandemic, workers lost their jobs for raising questions about whether their workplaces were putting them (or members of the public) in danger (see, e.g., Stone 2020; Scheiber and Rosenthal 2020), and others were prohibited from revealing their own positive COVID diagnoses to their coworkers (Eidelson 2020).
On the other hand, there is some expression for which employers cannot legally fire their employees. For example, employers cannot fire employees for reporting discrimination to the Equal Employment Opportunity Commission (EEOC) or for discussing their respective salaries in the workplace breakroom; retaliating against employees in these situations would violate Title VII of the Civil Rights Act and the National Labor Relations Act, respectively. These examples reflect how the at-will default has been modified by federal, state, or local laws that protect from employer retaliation certain kinds of worker expression. Employees and employers can also modify the at-will default by contract, a reality reflected in most collective bargaining agreements—though only a small minority of American workers are covered by them.
These exceptions mean that whether an employer is constrained from firing an employee because of something the employee has said turns on three main questions. The first is whether the employee is protected by an individual contract or a collective bargaining agreement that overrides that at-will default, and instead sets out specific conditions that must exist before a worker can be disciplined or fired. The second is whether there is a legal rule that protects the employee by creating an exception to the at-will default and, if so, whether there is an enforcement mechanism that is available to the employee as a practical matter. And the third is whether individual workers have market leverage—for example, is the employee too difficult to replace, or will the employee’s coworkers go on strike?
Too often, employees will find that none of the answers to these questions works in their favor. Others will not even know which questions to ask and will instead simply decide to remain quiet, fearing that if they say something that angers the boss, they might lose their jobs. This course represents potentially a loss to workers themselves, as well as to their larger workplace and political communities (Eidelson 2020). Some of these larger losses are easy to recognize: workers who fear employer retaliation might not speak out about harassment they are experiencing; they might not learn that they have a shared workplace complaint that they could jointly raise with their boss; they might not blow the whistle about dangerous or illegal company practices; and they might not speak up about an idea that would make the workplace run more efficiently. Other harms can be harder to see; for example, Cynthia Estlund has persuasively argued that “the workplace is the location where people come together for purposive, cooperative activity and where they gain or lose much of their sense of community and of self-worth” (Estlund 1995, 108). That insight highlights the moral and political dimensions of workers’ freedom of speech.
A meaningful freedom of speech for workers could come from two places. First, workers could have enough individual or collective power to secure their employers’ commitments to discipline or fire workers only for good reasons—that is, only when the worker has meaningfully harmed the employer or the employee’s coworkers. And, as the next section of this paper discusses, some U.S. workers have done just that through unionization and collective bargaining, or even through individual negotiations. But the vast majority of workers do not have contractual protections and instead have to rely on a second pathway to freedom of speech: a combination of whatever power they can muster in the moment, their employers’ sense of propriety, and—importantly—a patchwork of legal rules that prohibit employers from taking action against employees in certain circumstances. The bulk of this paper is taken up with explaining some of these rules and illustrating why they do not add up to meaningful free speech for workers. That is, employers are often able to suppress their employees’ speech on and off the job in the private sector and, to a lesser extent, the public sector.
This paper begins by discussing workers who often do have meaningful protections against employment consequences for their speech, either because of contractual commitments or because they live in one of a small number of U.S. jurisdictions that have reversed the at-will presumption. Then, it turns to legal rules that take certain worker statuses and activities out of the at-will arena. It begins with the National Labor Relations Act (NLRA), which undergirds both unionized and nonunionized workers’ collective power by protecting private-sector employees’ “concerted activity,” including speech related to working conditions—though with significant limitations. Among those limits: the law protects only “concerted” activity involving multiple employees; it applies only to topics with a sufficient nexus to the workplace; and workers can lose protection if they pursue collective action outside approved pathways. In addition, the NLRA imposes few constraints on employers, who are free to require employees to listen to a near-constant barrage of anti-union messages as long as those messages do not rise to the level of coercion.
Next, the paper turns to particular topics that can place employees at heightened risk of employer retaliation, such as employee speech about their own working conditions, the employer’s misdeeds, or the employee’s religious, social, or political views and beliefs. Importantly, there exist legal protections that cover employees’ speech on these topics, though, as the paper discusses, these protections are full of holes. For example, a long list of statutes provides at least some protection for employee whistleblowing—provided employees tell the right audience about what they have discovered. Likewise, workplace protections such as those found in anti-discrimination law are typically coupled with protections against employer retaliation, though these protections also have limitations that can be counterintuitive.
This report focuses on types of employee speech that can have considerable social value in addition to reflecting employees’ own autonomy interests; often, that social value is reflected in the fact that Congress, state legislatures, or other government bodies have established at least a degree of protection from the harshness of the at-will default when workers engage in these types of speech. Conversely, this report does not discuss some other types of worker speech over which employers often exert significant control. For example, employees’ job duties often involve speech—for example, a fast-food restaurant might require its cashiers to attempt to upsell customers to a larger size meal. Additionally, some worker speech harms others; for example, harassing speech, which can trigger employer liability under some circumstances. But even when liability is not on the table, employees may sometimes find their coworkers’ speech to be upsetting or annoying, and employers may be called upon to mediate these disputes. This paper touches only briefly on these issues.
Some readers might be surprised that a report about workers’ freedom to speak without facing employment consequences has not yet mentioned the First Amendment. Because the First Amendment constrains only government, not private individuals or entities,4 the First Amendment is not implicated when a private employer fires employees for their speech. However, the paper does close with two observations about First Amendment law as it relates to worker speech. First, public employees have limited First Amendment protections against being fired by their government employers when they speak as citizens, though these protections rarely cover public employees’ speech about their working conditions. Second, private employers can invoke the First Amendment when the government regulates their speech, which means there is a risk that courts will strike down government regulation of employer speech even when the regulation is intended to promote employee freedom. For example, if Congress amended the NLRA to require employers to remain neutral about the possibility of employee unionization, courts might then conclude that the change violates employers’ First Amendment rights, even though it would also promote employees’ speech and association.5
For at-will employees, the threat of employment consequences for their speech is considerably greater than the threat of government censorship. There exists a list of legal rules intended to constrain employers, but a combination of employer-friendly exceptions and the difficulties of enforcing rights means that employers retain extensive practical ability to punish workers for speech of which employers disapprove. Freedoms that many Americans hold dear thus stop at the boss’s door.
II. Reversing the at-will default
What do the examples in the introduction to this report have in common? They each involve speech: an employer either pressured its employees to express or suppress certain views, or punished an employee after the fact because of something the employee said. But at a higher level of abstraction, we might say that these examples each reflect the “bad reason” prong of the standard at-will formulation.6 These employees either were fired for a bad reason, or they were told to comply with a capricious or unjust workplace rule on pain of being fired (for a bad reason). Put this way, employer control over employee speech is one manifestation of employers’ control over workers generally.
If employer control over employee speech is a symptom, and the at-will rule is the diagnosis, then the cure would be to change the at-will default to one in which workers can be disciplined or fired only for “just cause.” Or, as Cynthia Estlund put it, “[d]ue process, in the form of a ‘just cause’ requirement for discharge and a fair hearing, would both bolster the existing protections of highly valued employee speech and, incidentally, generate some protection for less exalted forms of speech that simply do not justify the extreme sanction of discharge” (Estlund 1995, 104; see also Estlund 1996).
What counts as just cause can be malleable, but two commentators described the concept like this: “Just cause…embodies the idea that the employee is entitled to continued employment, provided he attends work regularly, obeys work rules, performs at some reasonable level of quality and quantity, and refrains from interfering with his employer’s business by his activities on or off the job” (Abrams and Nolan 1985, 594). This standard does not equate to absolute protection for workers’ speech and association, but it does mean that workers cannot be disciplined or fired simply because they have expressed an opinion with which their boss disagrees. Instead, the employer will have to demonstrate that the employee has actually harmed the employers’ legitimate interests.7
For a relatively small number of private-sector workers in American workplaces, this concept is already a reality: they are covered by a contract that contains just-cause protections, or they work in one of the small number of U.S. jurisdictions that have legislatively changed the at-will default. This section briefly describes contractual and statutory just-cause protections. It offers several reasons that they can be a useful and durable source of employment protection for employees, especially when coupled with enforcement mechanisms that mean workers do not have to rely on expensive and uncertain litigation or individual arbitration to enforce their rights. This discussion will also provide a backdrop for the next section, which turns to legal protections that limit the extent to which employers can fire or discipline employees for specific categories of expression.
A. Contractual just-cause protections and employee speech
Some private-sector workers benefit from contractual protection against termination for arbitrary or bad reasons. Tenured college professors fall into this category,8 as do workers who have negotiated employment contracts that list specific grounds on which the employer can terminate the contract. But these are the exception rather than the norm; there is no requirement that an employment relationship be memorialized in a written contract, and most U.S. employees work without one. Other employees may benefit from “implied” contracts, which are created when the employer makes a unilateral promise regarding the employment relationship, but, again, these are relatively rare, especially because employers can draft employment documents to avoid making any contractual incursions on at-will employment.9 Finally some employees—though only about 6% of private-sector workers—are covered by collective bargaining agreements, which are the main focus of this section (BLS 2022).
Collective bargaining agreements commonly overturn the presumption of at-will employment by including just-cause protections from termination. These provisions can vary in their language but are often quite straightforward. For example, the United Auto Workers’ collective bargaining agreement with Ford Motor Company states simply that Ford “retains the sole right to discipline and discharge employees for cause, provided that in the exercise of this right it will not act wrongfully or unjustly or in violation of the terms of this Agreement” (UAW and Ford Motor Company 2019, 17). Further, collective bargaining agreements may require progressive discipline for lesser infractions and require that disciplinary infractions be wiped from an employee’s slate after a period of time. And they typically include procedural protections; for example, most collective bargaining agreements include a grievance process that a union or employee can use to challenge discipline imposed without justification, and labor law requires employers to allow unionized employees to have a union representative present during interviews that may lead to discipline.10
Two features make collectively bargained just-cause provisions especially meaningful. First, the employer typically has the burden to show that it had just cause. Second, primary responsibility for enforcing a collective-bargaining agreement rests with the union, rather than individual employees. Unions, unlike individual employee plaintiffs, then become repeat players with enforcement expertise developed over time. Moreover, they are usually considerably better resourced than individual employees, perhaps even employing an attorney who does nothing but handle grievances arising under the union’s collective bargaining agreements. Thus, whereas employers often violate statutory labor and employment law with impunity because of barriers to effective enforcement, the enforcement of just-cause collective bargaining agreements can and should be routine.
B. Statutory good-cause protections
A small number of U.S. jurisdictions have overturned the at-will presumption by statute. Montana is the only state to do so; in addition, Puerto Rico and the Virgin Islands have passed similar laws, as has New York City with respect to fast-food workers and Philadelphia with respect to parking attendants.11 While these laws are similar in that they restrict employers from terminating employees without a good reason, there are some important enforcement-related differences that bear on their effectiveness. This section briefly compares the Montana statute to the New York City one.
Montana’s law requires that private-sector employers have “good cause” to fire employees, with good cause defined as “any reasonable job-related grounds for an employee’s dismissal based on” the employees’ “failure to satisfactorily perform job duties”; “disruption of the employer’s operation”; “material or repeated violation of an express provision of the employer’s written policies”; or “other legitimate business reasons.”12 The law is enforced by employees through litigation in state courts or in arbitration,13 and employees can recover damages, subject to some important limits.14
New York City’s law prohibits fast-food employers from firing workers without just cause, which the statue defines as “the fast food employee’s failure to satisfactorily perform job duties or misconduct that is demonstrably and materially harmful to the fast food employer’s legitimate business interests.”15 Additionally, the law mimics many collective bargaining agreements by requiring employers to follow a written progressive discipline policy in non-“egregious” cases.16 Other provisions make it harder for employers to evade the law by, for example, using layoffs or hours reductions to disguise decisions to terminate without good cause.17 And, as to enforcement, “any person, including any organization, alleging a violation” may file a lawsuit or demand a specialized arbitration process,18 and the city’s law department can also undertake enforcement proceedings against repeat violators.19
Requiring that employers have good cause to terminate or discipline an employee can be a significant source of protection for workers. But not all good-cause provisions are created equal: as the preceding paragraphs describe, Montana’s law is weaker than New York City’s on substance, and it is also harder to enforce. In other words, New York City’s law comes closer to replicating the protections that usually come with effective union representation, which means it has the potential to do more than the Montana law to deter employers from treating employees arbitrarily or with bad cause.
The at-will default is a central reason that employers can fire workers because of their speech, even when that speech does not harm production. Contractual or statutory just-cause protections can be an effective way to get to the root of the problem, especially when their enforcement mechanism is well-designed.
III. Legal protections for (some categories of) worker speech
Another approach to protecting valuable speech by workers is for legislatures to enact (or state courts to recognize in their decisions) protections for specific categories of worker speech. For example, labor law protects workers’ rights to join together to speak out about labor conditions; employment discrimination law protects workers who oppose discrimination at work; and whistleblower laws protect workers who report corporate wrongdoing to the authorities. Where these statutes apply, employers may not leverage their economic power over employees to demand silence. However, as discussed below, there are pitfalls to this approach: these statutory protections can apply in unpredictable ways, and, even where they apply, enforcement can be difficult or ineffective.
This section is organized around several types of worker speech that have (or at least may have) social value but that many employers would like to suppress. The discussion that follows is not intended to be a comprehensive treatise on legal protections for employees but instead to illustrate two general ideas: first, that there exists tremendous variation in the extent to which law protects workers’ speech; and second, that even within protected categories of worker speech, it is easy for workers to fall through the cracks, losing statutory protection and reverting to the at-will default.
The section begins with a fairly detailed discussion of the NLRA and its protection for workers’ collective action, before moving on to other topics. The extended discussion of the NLRA is partially illustrative—an example of how even broadly drafted statutory protections can leave workers to navigate a confusing array of exceptions and qualifications. It is also intended to give readers a more complete sense of how the landmark federal statute that is aimed at increasing workers’ bargaining power functions to protect (or not) workers’ speech about their working conditions.
A. Workers’ rights to criticize their working conditions and organize for improvements in those conditions
While the COVID-19 pandemic has been disruptive for everyone, essential frontline service workers have had to cope with especially dangerous and challenging conditions, risking their health to ensure that essential services continued. While some employers moved quickly to make their workplaces safer, others took a more lackadaisical approach. Especially in the latter scenario, workers often responded by demanding that their employers do more to keep them safe and also to recognize the burdens of work during the pandemic by increasing pay, improving sick leave, and more.
But employers sometimes responded to these requests by firing the “troublemakers.” For example, when Ben Bonnema and one of his coworkers became concerned about ventilation and unmasked customers at the Trader Joe’s where they worked, they wrote a letter to the company’s chief executive officer (CEO) requesting specific improvements; Bonnema’s boss then fired him with a written termination notice that invoked Bonnema’s “at-will” status and claimed Bonnema’s suggestions were “not in line with company values” (Shammas and Knowles 2021).
Bonnema’s situation is not unique. Though particulars differ from case to case, it is distressingly common for employees who work together to try to improve their pay or other working conditions to be fired, disciplined, or reprimanded. In these scenarios, the employer’s response serves a double purpose: getting rid of a squeaky wheel, and sending a message to other employees about the costs of expressing discontent.
Bonnema’s situation had a happy ending: after he tweeted about being fired, a labor lawyer offered to help. Soon after, Trader Joe’s agreed to reinstate Bonnema and take other remedial steps. This remedy was available because the National Labor Relations Act guarantees workers, even nonunion workers, the right to act together to improve their wages and working conditions—exactly what Bonnema and his coworker did when they wrote to the CEO.
This section focuses on the NLRA and discusses statutory limits to the at-will rule that apply when private-sector workers try to improve their working conditions.20 However, there are also significant limitations to the NLRA’s protection, and workers are vulnerable to discipline or discharge when they act outside the boundaries of that protection.
One caveat: in discussing the boundaries of the NLRA protection for workers’ speech and association, this section tells only part of the story. Courts have also undermined the NLRA’s protection for workers’ collective action by limiting the remedies available when employers violate the NLRA,21 authorizing employers to counter employees’ collective action with solidarity-destroying tactics,22 and more.
1. Collective voice at work: The NLRA and its limits
The National Labor Relations Act is one of the most important protections for private-sector employees’ speech. The 1935 statute declares that “inequality of bargaining power between employees who do not possess full freedom of association or actual liberty of contract” burdens the economy, including by “depressing wage rates and the purchasing power of wage earners in industry.”23 The NLRA’s solution—employee organizing and collective bargaining—is rooted in employees’ abilities to communicate about their working conditions with each other, their employers, and the public. Thus, § 7 of the NLRA states that employees have the right to engage in “concerted activities” for “mutual aid or protection,”24 and § 8(a)(1) of the act makes it illegal for employers to “interfere with, restrain, or coerce”25 employees who are exercising their § 7 rights. Importantly, this provision applies to both union and nonunion employees, making the NLRA an important source of protection for covered employees’ organizing to improve their own working conditions.26
At the same time, the NLRA’s limitations are many and varied. The statutory language is limited to employees’ “concerted activities for…mutual aid or protection,” meaning that workers’ speech can fall through the cracks if it is not collective, if it is targeted at a concern that is not workplace-related, or if it is not action-oriented. And courts interpreting the NLRA have sided with employer interests on some critically important questions,27 blunting the statute’s potential to level the playing field between employers and employees. Taken together, these limitations also make it virtually impossible for workers to rely on their intuitions about when or how the NLRA might protect them.28
a. The ‘connectedness’ requirement
In general, the NLRA protects only “concerted activity,” that is, activity that involves at least two employees.29 This might look like Ben Bonnema and his coworker sending their letter, a strike or picket line, a petition expressing dissatisfaction with a new company policy, or two employees revealing their salaries to one another, perhaps because they suspect the employer is systematically underpaying workers of color. But an employee’s individual workplace advocacy usually will not qualify as collective action,30 though there are some exceptions.31
Surprisingly, some conversations between multiple employees also do not qualify as concerted if the board feels the employees were merely “griping.” This caveat might seem unimportant, but low-stakes workplace social interactions can set the stage for other forms of collective action. As Michael Oswalt put it, workers generally need to develop relationships and build trust before they will pursue collective action: “the subtle shift from ‘I hate this,’ to ‘We hate this’ [could] prompt a worker to think about starting a petition drive before barging into a supervisor’s office alone” (Oswalt 2017, 1002). And an employer who disciplines a worker for griping unmistakably sends a message about workplace control; it would take considerable bravery for a worker to engage in protective concerted activity after seeing a coworker fired for griping.32
b. The ‘for mutual aid or protection’ requirement
The NLRA’s second requirement focuses on employees’ goals, asking whether employees are trying to “improve terms and conditions of employment or otherwise improve their lot as employees.”33 Many topics obviously qualify—for example, collective action that relates to pay, workplace safety, or employer discrimination all meet this requirement.
Political issues that relate to working conditions are also in bounds: in the foundational decision Eastex v. NLRB,34 the Supreme Court agreed that the NLRA protected distribution of a union newsletter that urged workers to oppose a “right-to-work” amendment to the Texas Constitution, highlighted President Nixon’s recent veto of a minimum wage increase, and generally encouraged workers to “defeat our enemies and elect our friends” at the ballot box. The court deferred to the NLRB’s assessment that these political initiatives could have indirect effects on working conditions, writing that “[t]he Board was entitled to note the widely recognized impact that a rise in the minimum wage may have on the level of negotiated wages generally,” as well as that a constitutionally enshrined right-to-work law would be harder to change than a statutory one.35
This relatively broad approach to defining “mutual aid” means that employees’ collective action related to political topics can be protected by the NLRA as long as there is a tie between the topic and working conditions. Immigration law is a salient example, as illustrated by “Day Without Immigrants” protests that took place in 2006 and 2017.36 These protests were intended to highlight the importance of immigrant workers to the functioning of the U.S. economy; workers from all over the country missed work to participate, some with their employers’ permission and some without.
Following each protest, the NLRB’s general counsel (GC) analyzed whether employers had violated the NLRA when they fired workers for missing work to join the protests. (The board’s GC is its chief prosecutor; a GC’s views are important in part because they guide what cases the office will take and what arguments the GC’s office will make. However, the GC’s views do not bind the board or the public.) Importantly, both GCs (a Republican-appointed GC analyzing the 2006 protests and a Democratic GC analyzing those in 2017) agreed that workers’ advocacy on immigration policy did fall within the scope of the mutual aid or protection clause. But they disagreed about whether employers were nonetheless free to enforce their attendance rules when workers missed work to attend the protests. The first analysis (conducted by the Republican GC) concluded that employers could enforce their attendance rules because the demonstrations were not strikes over an issue that the employer could remedy.37 The second memo (issued by the Democratic GC) reached the opposite conclusion.38 Thus, while at least some forms of collective action related to immigration are protected by the NLRA, there is risk that that this protection will not extend to all tactics.39
While Eastex is relatively broad, some issues of great importance to workers can nonetheless fall outside the NLRA’s bounds. For example, the Fifth Circuit Court of Appeals held that workers who were members of an “outside” political advocacy group were not protected by the NLRA when they distributed the group’s literature at work, even though the group’s goals—ending mandatory workplace drug testing—related to a disputed issue at the workplace in question.40 And in another case, the NLRB decided that employees were not protected when they advocated for the employee stock option plan to acquire a majority share in the employer’s parent company, writing that the proposal did “not advance employees’ interests as employees but rather advances employees’ interests as entrepreneurs, owners, and managers.”41 This distinction between managerial interests and working conditions may also arise in a case against Google, which involves two worker-organizers who were fired after circulating a petition regarding the company’s potential collaboration with the U.S. Customs and Border Control agency.42 The case was pending as of this writing; it seems likely that the board will have to resolve whether the NLRA protects advocacy aimed at influencing an organization’s direction or choice of clients.
2. Concerted activity that loses NLRA protection
The discussion in the previous section illustrates how the NLRA provides meaningful-but-limited protection for workers’ collective action related to working conditions. But workers can lose NLRA protection if they choose the wrong place, the wrong time, or the wrong manner in which to express themselves. In particular, courts interpreting the NLRA have allowed employers to insist on loyalty, decorum, and productivity, reflecting what James Atelson has described as “deeply held judicial feelings about the contractual or status obligations of employees as well as the rights of property” (Atelson 1983, 5).
Because of these limitations, employees often have no recourse if they are fired for collective action that falls into any of the following categories:
- The message is deemed “disloyal, reckless, or maliciously untrue”
- The message is expressed using “opprobrious” language
- The employees use materials belonging to the employer, possibly even including the employer’s computer or email system, to express their message
- An employee solicits another employee to sign a union card during “working time”
- Employees picket in order to seek support from a “secondary” employer, that is, a company that does business with the workers’ employer.
This section briefly discusses these limitations, each of which can impede workers from effectively building solidarity with each other or with potential allies, such as consumers of the products the workers create. Despite the NLRA’s stated protections, there remain many situations in which courts defer to employer authority, allowing workers to be disciplined for their speech.
The disloyalty doctrine comes from a case known as Jefferson Standard,43 which arose out of a breakdown in collective bargaining negotiations between the union representing a group of technicians and their employer, a television and radio broadcaster. The employees picketed (but did not strike) for about six weeks, and then began to circulate a new handbill, which the Supreme Court later characterized as a “vitriolic attack.”44 It read:
Is Charlotte A Second-Class City?
You might think so from the kind of Television programs being presented by the Jefferson Standard Broadcasting Co. over WBTV. Have you seen one of their television programs lately? Did you know that all the programs presented over WBTV are on film and may be from one day to five years old. There are no local programs presented by WBTV. You cannot receive the local baseball games, football games or other local events because WBTV does not have the proper equipment to make these pickups. Cities like New York, Boston, Philadelphia, Washington receive such programs nightly.
Why doesn’t the Jefferson Standard Broadcasting Company purchase the needed equipment to bring you the same type of programs enjoyed by other leading American cities? Could it be that they consider Charlotte a second-class community and only entitled to the pictures now being presented to them?
The company fired 10 technicians because of this handbill. In the letter of termination, the employer did not contest the handbill’s factual statements; instead, it took the view that “[c]ertainly we are not required by law or common sense to keep you in our employment and pay you a substantial salary while you thus do your best to tear down and bankrupt our business.”45
The NLRB acknowledged that the technicians’ statements in the handbill either were true or at least were not known to be false.46 But the board nonetheless decided that the technicians lost NLRA protection because their conduct was “indefensible,” analogizing to union tactics such as sit-down strikes, sabotage, and strike-related violence.47 The Supreme Court agreed, writing that the “company’s letter shows that it interpreted the handbill as a demonstration of such detrimental disloyalty as to provide ‘cause’ for its refusal to continue in its employ the perpetrators of the attack. We agree.”48
There is much to criticize about the court’s reasoning, but the bottom line is that the Jefferson Standard rule continues to leave workers engaged in collective action vulnerable to discipline or termination by their employers. The recent case Miklin Enterprises, Inc. v. NLRB shows how.49 The case began when a group of employees placed posters on “community bulletin boards” in the employer’s restaurants. These posters showed two identical pictures of Jimmy Johns sandwiches, indicating that one was made by a healthy worker and the other by a sick worker. The text at the bottom of the poster read:
CAN’T TELL THE DIFFERENCE? THAT’S TOO BAD BECAUSE JIMMY JOHN’S WORKERS DON’T GET PAID SICK DAYS. SHOOT, WE CAN’T EVEN CALL IN SICK. WE HOPE YOUR IMMUNE SYSTEM IS READY BECAUSE YOU’RE ABOUT TO TAKE THE SANDWICH TEST.50
Following that and other speech along similar lines, the company then fired six employees who, as the Eighth Circuit put it, “coordinated the attack,” and issued warnings to three others.51
While the NLRB concluded that the workers had engaged in protected concerted activity because they did not have a “malicious motive,” the Eighth Circuit disagreed. In an en banc ruling—that is, a decision issued by the full Eighth Circuit rather than a panel of three judges—the court concluded that employee speech was “disloyal” if it “indefensibly disparaged the quality of the employer’s product or services,”52 characterizing the workers’ health-related statements as “equivalent of a nuclear bomb.”53
This rule can get in the way of the kinds of worker-consumer coalitions that can be critical to workers achieving their goals. An important way to generate community support is to show how working conditions affect consumers’ experiences in a business.54 This was the gist of the Jimmy Johns campaign—that making it difficult or impossible for workers to take time off work when they were sick was bad for workers and consumers alike. Yet this kind of argument will often focus on the product or service that the workers make or provide, meaning that workers will run the risk that decisionmakers will ultimately find the criticism was “disloyal”—causing them to lose NLRA protection and allowing them to be fired.
b. Opprobrious speech
Collective action sometimes takes the form of a spontaneous, outraged response to something the employer has just said or done. But with emotions running high, employees might use profane or insulting language, in which case an employer might insist that it has a right to terminate the employee for breaching workplace civility rules. As with disloyalty, there is not a bright-line definition of what counts as opprobrious—there is no list of words that are out of bounds. Instead, until recently, the NLRB addressed these cases by focusing on the setting and the circumstances of employee speech. However, in 2020, the Trump NLRB announced a new approach that is more deferential to employers regardless of setting, permitting them to prioritize workplace civility over employees’ concerted activity. This case, known as General Motors, will likely be reversed by the Biden board. Still, it is worth further discussion for two reasons: first, it reflects the instability of even relatively well-established legal protections for employee speech; and second, it illustrates how employers can weaponize workplace civility rules when they create upsetting situations and then punish workers for becoming upset.
Up until 2020, the NLRB used rules developed in three different contexts: picket-line speech,55 speech on social media,56 and workplace conversations with management.57 The particulars differed, but these cases generally recognized both that fraught workplace situations can lead to heightened emotions and intemperate comments, and also that managers cannot always be trusted to act in good faith—for example, an employer looking to fire a strike organizer might feign offense at language usually tolerated in the particular workplace. Under this approach, employees would generally lose the protection of the NLRA only if their speech crossed the line from heated to threatening or abusive.
However, the Trump board’s General Motors decision replaced these three separate approaches with a test that is considerably less protective of employee speech and more deferential to employer concerns.58 The case involved three suspensions of an employee, Robinson, who also served as a union committeeperson. (There were some disputes about the underlying events that led to these suspensions; this discussion relies on the board’s statement of those events.) In the first incident, Robinson swore at a manager during a dispute about whether the manager had gone back on an agreement to pay overtime to a group of employees participating in a training. In the second, a manager told Robinson that he was speaking too loudly during a meeting about subcontracting out work; Robinson reportedly responded by saying “’Yes, Master, Your Master Anthony,’ ‘Yes, sir, Master Anthony,’ ‘Is that what you want me to do, Master Anthony?,’ and also stated that the supervisor wanted him ‘to be a good Black man.’” In the third incident, Robinson said he would “mess [a manager] up” during a disagreement that arose in a meeting about shift and staffing changes, and later in the meeting played loud music that the NLRB described as “profane, racially charged, and sexually offensive.”59
Rather than analyzing the circumstances surrounding each exchange, the NLRB asked whether Robinson was fired because his employer objected to his profanity or instead because of his union affiliation or workplace advocacy.60 The reasoning in support of this change relied in part on “tension” between labor and anti-discrimination law; the problem, according to the board, was that the board’s existing rules tied employers’ hands, preventing them from responding to speech that could ultimately result in a hostile workplace.61 But the General Motors rule goes beyond racist or sexist harassment, conflating harassment with uncivil speech and using the specter of anti-discrimination law to empower employers to police employees’ legitimate grievances that are expressed using rude, profane, or obnoxious language.
Consider Robinson’s second suspension, which the company said was imposed because of Robinson’s response to a manager’s comment that Robinson should lower his voice. First, the context was apparently fraught. Robinson testified that management was proposing to subcontract work while also stonewalling Robinson’s requests for information about the subcontractors’ costs and “work hours and shifts for the bargaining unit employees.”62 And, to be sure, Robinson’s response to the request that he lower his voice was sarcastic—perhaps he thought the supervisor was “tone policing,” or perhaps he thought the supervisor would not have been so concerned about the volume of a white employee’s voice.63 Despite this, the NLRB used the phrases “racially offensive” and “racially charged” to characterize Robinson’s comments, obscuring the likelihood that Robinson was disciplined in part because he used sarcasm to indicate that he thought he was being treated poorly because of his race.
This chain of events reflects how deference to employers can disproportionately harm workers of color, notwithstanding the General Motors board’s stated concern for hostile work environments. First, workers of color are more likely that white workers to be subjected to bad treatment (including racist treatment) by their bosses, resulting in more occasions where an employee might (understandably) respond to a situation in anger. Second, managers are more likely to misperceive workers of color to be behaving aggressively as compared to white workers, and they may then impose discipline based on this misperception.64 And even managers who believe themselves to be enforcing civility rules in an even-handed manner may be acting on implicit biases. Thus, while workplace civility rules are sometimes justified based on workers’ rights to be free of abusive or harassing treatment while at work, they can also be weaponized against workers who are trying to respond to upsetting situations.
c. Collective action in the ‘wrong’ time or place
In addition to the substantive limitations on protected concerted activity, there are also limits on where and when that activity may occur. These limits fall into two general categories. The first category concerns employees who are trying to organize inside their own workplace: workers may engage in protected concerted activity during breaks and other downtime, but employers can also insist that “working time is for work.”65 The second category concerns workers who engage in collective action on the premises of a company that does not employ them, including labor organizers who work for unions, as well as “fissured” employees,”66 i.e., those who are not technically employed at the place where they spend their work days because of subcontracting and other contractual relationships. Both sets of rules reflect assumptions that prioritize employers’ property and managerial interests, leaving holes in the NLRA’s protections for worker speech.
i. Space and time limitations on organizing at work
It only makes sense for workplace organizing to take place at work. This organizing might take the shape of conversations between coworkers; handing out union cards and other written material; or wearing union buttons or T-shirts. The NLRA protects all of these activities—but with limitations. The rules consist primarily of a set of default presumptions that govern different types of activity. Under these defaults, workers can wear union T-shirts or insignia at any time;67 they can solicit their coworkers’ support for unionization and have other § 7-protected conversations only during nonwork time;68 and they can distribute union cards or other literature only during nonwork time and in nonwork areas of the workplace.69 But employers may impose more restrictive rules if they can prove “special circumstances” justifying them.70
Employers have even more freedom to control how workers use supplies that the employer owns. For example, employees generally have no right to use the office photocopier to produce flyers about an upcoming union rally, nor may they appropriate space on the office bulletin board to post one of these flyers. This rule might seem intuitive as applied to scarce or expensive supplies, but the board has also applied it to computers, servers, and even employees’ work-issued email addresses. (This is another issue on which the board has oscillated, with the Obama board holding that employees who had access to their employer’s email system had a presumptive right to use that system for concerted activity during nonwork time,71 only to be reversed by the Trump board in 2019.72)
There is a caveat that pertains to the solicitation and distribution presumptions as well as to the rule that employers can restrict the use of their own equipment: employers may not discriminate against union speech. At a minimum, this means that while an employer may broadly insist that “working time is for work,”73 it may not (for example) allow workers to express anti-union viewpoints while simultaneously prohibiting pro-union viewpoints. But the board has historically taken this nondiscrimination rule further, to reach employers who ban any union-related communication while allowing similar expression on other topics.74 Note, though, that even this broader nondiscrimination rule did not require parity between employers and employees: an employer that barred its employees from talking about unionization or any other cause over company email could still send its own anti-union message via email.
These space-and-time limitations have only become more significant during the coronavirus pandemic. For example, when the Trump board held that employees’ use of their work email addresses was not protected by § 7, it suggested that in-person communication would be “adequate” to allow workers to communicate with each other about their working conditions. That assumption was already flawed for workers who did not routinely share space with their coworkers; now, those consequences are multiplied because so many more employees work from home. Moreover, even workers who go into the office may find that in-person communication has become more difficult, especially where employers have closed breakrooms or changed schedules because of social-distancing concerns. Despite the NLRA’s broadly drafted protection for concerted activity, these employees may find themselves with few workable methods of communicating with each other about their working conditions.
ii. Workplace fissuring and NLRA protections for workers’ speech
The rules discussed in the previous subsection apply to workers whose collective action takes place on their own employers’ premises. But many people spend their working time at job sites with which they do not have an employment relationship. For example, consider janitors who clean a downtown building each night: the building’s management may employ the janitors directly, or it may have contracted for their services. If the latter is true, then another company formally employs the janitors and bears responsibility for paying wages and complying with employment law.75 “Fissured” arrangements like this one can contribute to worsening working conditions while also limiting employees’ recourse to NLRA-protected collective action to improve their situations.
Whether subcontracted employees have the same § 7 rights at their worksite as do traditional employees is another issue where the board has oscillated: the Obama board answered this question “yes”;76 the Trump board said the opposite,77 allowing property owners to exclude off-duty subcontracted employees from their worksite. The latter approach means that a property owner may bar subcontracted employees from coming to the worksite when they are not scheduled to work, effectively preventing them from talking to their coworkers on other shifts about working conditions.
Further, workers can be left unprotected when their collective action under § 8(b) of the NLRA counts as “secondary”—that is, picket lines or strikes aimed at influencing an entity other than the employer with whom there is a labor dispute. § 8(b) is mainly about restrictions on unions, but whereas unions that violate these provisions can face injunctions and money damages, participating workers can also lose the NLRA’s protection. Precisely what counts as prohibited secondary activity is beyond the scope of this paper, but it is worth noting that this prohibition can leave subcontracted employees with less protection than their peers in traditional employment relationships. For example, if hotel cleaners who technically work for a subcontractor become irate about harassment and low pay, they may want to picket in front of the hotel where they work, but if they run afoul of the byzantine secondary activity rules they could be fired without recourse. In contrast, housekeepers who are employed directly by the hotel would be on firmer footing. In other words, an employer’s decisions about how to structure its workforces have a direct bearing on workers’ rights to organize for better treatment.
The NLRA is a significant exception to the at-will default; it was intended to level the playing field between workers and employers by protecting workers’ ability to band together to demand better working conditions. But, as this section has discussed, there are significant gaps in the NLRA’s coverage, and employers can sometimes override the NLRA’s protections if employee speech is deemed too disruptive of employer interests. Thus, while of the various statutes discussed in this section the NLRA provides the most comprehensive protection for worker speech, it does not undo the effects of the at-will default: workers are still often vulnerable to employer domination when they advocate for better treatment from their employers.
B. Workers’ rights to report employer wrongdoing
One significant category of worker speech that employers might wish to suppress concerns the employer’s own wrongdoing. This wrongdoing might involve customers or regulators, as when a company suppresses information that its products might make people sick or that it is using dubious accounting practices to hide financial losses from investors. It could also involve employees’ complaints that the employer’s treatment of them is illegal—for example, that the workplace is unsafe, or that the employer has discriminated based on an employee’s race or gender, or that the employer has failed to pay required overtime.
The three subsections that follow discuss two main topics. The first two concern legal protections for employee whistleblowers and statutory anti-retaliation provisions that protect workers when they complain about their employer’s unlawful treatment of them or their coworkers. (This is not intended to suggest that these are the only legal protections that employees could invoke in these circumstances; for example, the NLRA will often protect workers who band together to report their employers’ unlawful treatment of them, and some state courts will treat retaliatory terminations of whistleblowers as wrongful discharges in violation of public policy.) The third subsection turns to employers’ ability to contract for employees’ future silence through nondisclosure agreements (NDAs). These agreements are an important reason that legal protections for some types of employee speech may not be enough to encourage employees to speak up. Even though NDAs are sometimes unenforceable—including when they purport to override nonwaivable rights to report corporate misconduct to relevant authorities—employees who believe their NDAs to be enforceable will be reluctant to breach them.
1. Whistleblower protections
There are dozens of whistleblower laws that apply to different kinds of employer conduct in different jurisdictions. These statutes usually protect employees under specifically defined circumstances, such as when they have reported a certain variety of corporate wrongdoing to the right audience, such as a government regulator. For example, the federal financial-fraud law known as Sarbanes-Oxley protects employees who provide information about suspected violations to law enforcement, members of Congress, or their own supervisors, as well as employees who participate in enforcement proceedings.78 Of course, Sarbanes-Oxley was drafted to guard against financial misdeeds, not to promote broader employee voice.79 Similarly, while state and federal law contains numerous anti-retaliation provisions—the Occupational Safety and Health Administration alone enforces 25 such provisions, which are found in statutes ranging from the Affordable Care Act to the Federal Water Pollution Control Act to the Taxpayer First Act80—it would be a mistake to think of them as mechanisms for general worker empowerment.
Still, whistleblower protections are useful where they apply, and their scope of application can be broad or narrow. For example, consider New Jersey’s whistleblower statute, the Conscientious Employee Protection Act. The specifics of the law are less important than a sense of its breadth: it reaches employer retaliation against employees who report a wide range of problematic or illegal practices by their employers to a list of listeners that includes the employee’s supervisor, the government, entities that contract with the employer, shareholders, and so on.81 In contrast, New York’s whistleblower statute was, until 2021, comparatively narrow; it covered only employees who reported violations of New York labor and employment law, substantial public health threats, and a few other specific categories of information. (New York recently amended its statutes to closely track the New Jersey statute.)
Whistleblowing is a category of speech where it is easy to see the connection between the employee’s interest and the public interest. However, whistleblower protections are inconsistent across jurisdictions and can turn on the subject of the report and the audience who receives it, making it hard for employees who are not advised by an attorney to know when they are on solid ground.
2. Anti-retaliation protections for employees who oppose workplace discrimination
Suppose an employee believes the employer has discriminated against a coworker based on that person’s race or gender. The employee might take action, for example, by filing a charge with the EEOC or a similar state agency. Alternatively, an employee might try to solve the problem internally, perhaps taking an employer up on its promise of an “open door policy.” What happens if the employer’s response is to retaliate?
If the employee takes action together with at least one coworker, then the NLRA likely protects them both. But employees who act without help from coworkers might still find protection. Anti-discrimination laws generally prohibit employers from retaliating against employees who report potential violations, either to the employer or an enforcement agency. Indeed, many federal, state, and local statutes intended to set baseline working conditions also prohibit employer retaliation against employees who report violations; the Fair Labor Standards Act, the NLRA, the Occupational Health and Safety Act, and their various state analogues are just a few examples. It would be impossible to canvas the various rules that apply to retaliation claims under all of these statutes; instead, this section uses Title VII as an example to illustrate how, even when they are backed by anti-retaliation protections, workers can still fall through the cracks.
Title VII prohibits employers from retaliating “because [an employee] has opposed any practice” that violates Title VII’s nondiscrimination provisions or “because [an employee] has made a charge, testified, assisted, or participated in any manner in an investigation, proceeding, or hearing” under Title VII.82 The first clause is referred to as the “opposition clause” and the second as the “participation clause.” (Near-identical provisions are found in the Americans with Disabilities Act and the Age Discrimination in Employment Act.83)
The participation clause protects employees who take part in EEOC proceedings (or those of a parallel state agency), including by filing a charge of discrimination or providing witness testimony. This protection is robust; the statute’s model of agency enforcement works only if employees are willing to come forward, and Congress and the courts have taken the view that when employees do come forward they should be protected (see Green 2014, 766). For example, the protection applies even if the agency ultimately decides not to take action based on a charge of discrimination.
But the opposition clause does not offer the same broad protection; instead, courts have interpreted the clause to cover only “reasonable” opposition. “Reasonableness” encompasses two important limitations: that the opposition be expressed in a reasonable manner, and that the complaining employee could reasonably believe the employer had violated Title VII.
With regard to the first limitation, many courts have held that employers may discipline or fire workers who express their opposition to workplace discrimination in a disruptive or insubordinate fashion, as well as those who merely “gripe” about discrimination;84 these cases have clear parallels to the Trump NLRB’s approach to protected concerted activity in General Motors—though perhaps with even more troubling effects on workers with marginalized identities. As Susan Carle points out, this rule gives employers an incentive to act badly, because they have “higher chances of prevailing in discrimination suits when their conduct is so infuriating that it causes employees to lose their temper” (Carle 2016, 186).
As to the second limitation, Title VII’s opposition clause protects employees only when they oppose employer conduct that they reasonably and in good faith believe violates Title VII. In evaluating whether an employee could reasonably believe that the employer’s conduct violates Title VII, courts assume that a “reasonable person” is aware of the vagaries of federal case law. In other words, it is not enough for an employee to have a genuine belief that an employer has discriminated in some fashion, because Title VII does not cover every incident of discrimination. The statute also fails to protect employees from retaliation when they advocate for their employers to do more than just refrain from violating Title VII: an employee who advocates for a workplace affirmative action policy or anti-bias training, for example, would not be protected from retaliation under the opposition clause (see Smith 2003, 555–56).
To see how the reasonableness requirement plays out, consider the Supreme Court’s decision in Clark County School District v. Breeden.85 The case arose out of a meeting between Shirley Breeden, her (male) supervisor, and another (male) employee, to review job applicants’ psychological evaluations. One evaluation revealed that an applicant had made a harassing remark to a coworker: “I hear making love to you is like making love to the Grand Canyon.”86 Breeden’s supervisor “read the comment aloud,” looked at Breeden, and said, “I don’t know what that means.”87 “The other employee then said, ‘Well, I’ll tell you later,’ and both men chuckled.”88 Breeden complained about this exchange to two school district assistant superintendents, after which she alleged she was punished by having her job duties revised and being transferred to a different job location.
The Supreme Court held that the district had not violated Title VII even if it did retaliate against Breeden for her complaint. It concluded that Breeden could not have reasonably believed that her coworkers’ discussion of the “Grand Canyon” remark violated Title VII, because Title VII prohibits only harassment that is so “severe or pervasive” that it “alter[s] the conditions of employment and create[s] an abusive working environment.”89 The exchange that prompted Breeden’s complaint, the court continued, was “at worst an ‘isolated inciden[t]’ that cannot remotely be considered ‘extremely serious,’ as our cases require.”90
In other words, it did not matter whether Breeden genuinely believed the conduct violated Title VII, whether she genuinely felt harassed because of the exchange, or whether the exchange that prompted her complaint was recognizable as harassment as that concept is colloquially understood. But most nonlawyers have little grasp of the vagaries of Title VII;91 they are more likely to have a very general sense of what that statute prohibits, gleaned from sources such as general news stories, representations in popular media, and the employer itself—and none of these sources is likely to provide detailed or accurate information about topics such as what conduct does and does not rise to the level of sexual harassment. In fact, one study of sexual harassment training materials that are supplied by employers to employees found that these materials “tend to suggest that relatively trivial slights could give rise to harassment-related liability” (Tippett 2018, 481)—precisely the type of belief that the Breeden court characterized as unreasonable, leaving the employee unprotected from employer retaliation.
Anti-retaliation provisions are an important source of protections for workers who invoke their rights under Title VII and other employment statutes. However, these protections are not absolute: workers who expect to be protected will sometimes find that they are not, either because they expressed themselves in a heated manner or because they opposed conduct that they saw as discriminatory but that was not actually covered by anti-discrimination law.
3. When may employers buy employees’ silence?
Much of this paper discusses statutory responses to employer power, that is, legal limits intended to prevent employers from exercising their leverage over employees in ways that harm the employees or society. But employers also contract for employee silence in advance, in the form of nondisclosure agreements. NDAs have legitimate uses, such as protecting company trade secrets, but their reach often extends far beyond this type of information, sometimes purporting to restrict employees from revealing anything learned in the course of their employment. These broad NDAs reflect employees’ unequal bargaining power but they can also perpetuate it in at least two ways: by making it harder for employees to criticize their current or former employers, and by limiting employees’ abilities to get new jobs in their fields (see Lobel 2018).
Nondisclosure agreements (and their close relatives, nondisparagement agreements) came to widespread public attention in the wake of the #MeToo movement, which is synonymous with disclosures of sexual assault and harassment by powerful—and often serial—offenders. By using a combination of industry influence and payoffs, numerous abusive bosses and their enablers were able to buy current and former employees’ silence, shielding sexual harassers and abusers from exposure and allowing them to go on to harm others. For example, consider Zelda Perkins, a former assistant of film producer Harvey Weinstein. Perkins signed a nondisclosure agreement when she left her job in 1998; she later explained that she had been harassed by Weinstein for years but quit after a colleague said Weinstein had sexually assaulted her (Garrahan 2017). In exchange for 125,000 pounds and other provisions aimed at preventing Weinstein from harming others, Perkins agreed to a very restrictive set of terms—remarkably, she was not even allowed to keep a copy of the agreement itself; nonetheless, Perkins broke the agreement in light of rape allegations made against Weinstein in 2017, but even then she feared “crushing legal and financial repercussions” (Perman 2018).
The NDA that Perkins signed was negotiated as she was leaving her job, and she was represented by counsel—but Perkins still felt acutely the imbalance between her power and Weinstein’s, later saying that she felt her credibility would be “destroy[ed]” if she rejected the agreement and reported Weinstein (Garrahan 2017).92 Other NDAs are imposed at the beginning of an employment relationship, before employees even know what information they might learn or whether or why they might want to disclose it. These agreements are often imposed as a condition of employment, and they can be drafted in exceedingly broad language. For example, one recent California case involved nondisclosure and nondisparagement agreements that collectively committed employees to keep confidential virtually all information about their work, including “information disclosed by the Company to Employee and information developed or learned by Employee during the course of employment with Company,” as well as “all information of which the unauthorized disclosure could be detrimental to the interests of the Company.”93 Signing these agreements, which were recently challenged in court, resulting in a settlement, was a mandatory condition of employment for all employees. Of course, when faced with the choice of signing or going elsewhere for work, many employees will simply sign without seeking legal advice or attempting to negotiate.
NDAs are not always enforceable. For example, an ex-employee alleged to have violated an NDA might later argue successfully that the agreement was signed under duress or that its enforcement would violate public policy. Further, agreements as broad as the one in the previous paragraph may violate state or federal law, either on their face or in specific applications. For example, whistleblower protections can override NDAs,94 and one court enjoined an NDA that purported to bar an employee from participating in an EEOC investigation.95 California courts have also treated very broad nondisclosure agreements as attempts to skirt legal limits on the use of noncompete agreements—the idea being that an NDA that makes it impossible for employees to bring their general knowledge and skills to a new employer will prevent an employee from landing a new job in the same field.96 And nondisclosure and nondisparagement agreements can violate the National Labor Relations Act,97 especially when they bar employees from criticizing their employer or disclosing information about their wages and working conditions either to each other or to “outsiders,” such as reporters. Finally, several states have responded to #MeToo disclosures by passing new laws limiting in various ways the use of nondisclosure agreements as they apply to allegations of sexual assault or harassment (see Spooner 2020, 355-63; Johnson 2019, discussing eight state statutes limiting the use of nondisclosure agreements in the context of harassment or assault). These laws cover not just NDAs that are imposed at the beginning of an employment relationship but also settlement or severance agreements like the one Perkins signed; other state laws take a similar approach to settlement agreements that conceal “public hazards.”98
Of course, even unenforceable NDAs can have effects on the employees who sign them. The legal reality that a given NDA is unenforceable may make little practical difference to employees who fear the repercussions if they break their word. Employees (and their potential future employers) may fear being sued by their previous employer, and employees may also fear repercussions from future employers. Thus, employees may have little power to resist agreeing to an NDA, and even an unenforceable NDA can be enough to silence employees.
C. Workers’ social media, political, and religious speech
What happens when workers disagree with the boss (or with each other) about sensitive topics, including politics and religion? Workplace conflicts over these issues can raise thorny problems, especially if one employee holds views that another finds hurtful or dehumanizing. But, as the introduction to this paper discusses, employers sometimes coerce employees’ political or religious speech in contexts that have nothing to do with maintaining workplace harmony; in these situations, the employer simply wishes to control employees’ views and beliefs for its own purposes. As the next two subsections discuss, there are some exceptions to the at-will rule that are sometimes relevant to these situations—but employees are often left unprotected.
1. Political expression
Several states protect employees from discrimination based on specific political activities such as voting or affiliating with a political party.99 Some of these statutes are of relatively recent vintage, but others date either to the 1700s or to Reconstruction, when they were passed in recognition of the likelihood that employers would use their control over workers’ livelihood to coerce their votes against anti-slavery Republicans.100 New York’s law is representative of many of these statutes in that it is limited to a specific set of activities—running for office, campaigning, or fundraising for a candidate, party, or political advocacy group101—and New York courts have dismissed cases involving employees who were fired for other types of political expression.102
A few other states, however, have statutory protections for “political activities” or “political opinions” more generally.103 California is one example; its labor code prohibits employers from threatening to fire employees in order to try to “coerce or influence” the employee into “adopting or following any particular course or line of political action or political activity.”104 Like many of the other state statutes discussed in this section, this law has not resulted in many reported decisions interpreting its meaning, but a 1979 decision by the California Supreme Court concluded that, in addition to partisan political activity, the statute covered political views and civil rights advocacy, including the plaintiffs’ participation in the gay liberation movement.105 Additionally, two states—New Jersey and Oregon—prohibit employers from compelling employees’ participation in meetings held for the purpose of allowing the employer to communicate its views about “religious or political matters.”106
In addition to laws specifically protecting employees from retaliation based on their political activity, a handful of states including California, Colorado, and North Dakota have broader protections for employees’ out-of-work “lawful activities” that could encompass politics. (New York has a more limited protection for off-duty “recreational activities.”)107 But even in those states, employees’ real-world protection is limited because of a combination of statutory exceptions and narrow interpretations by courts. For example, courts have narrowed nearly out of existence California’s law protecting employees’ off-duty activities.108
As with the NLRA and Title VII’s retaliation provision, courts may read an implied duty of loyalty into lawful activities statutes. Consider Colorado’s statue, which provides a damages remedy for employees who are fired for “engaging in any lawful activity off the premises of the employer during nonworking hours,” unless a statutory exception applies.109 Statutory exceptions include employment decisions that “relate . . . to a bona fide occupational requirement,” or that are “reasonably and rationally related to the employment activities and responsibilities of a particular employee or a particular group of employees,” or that are “necessary to avoid a conflict of interest with any responsibilities to the employer or the appearance of such a conflict of interest.”110 Echoing the Jefferson Standard decision from the NLRA context, courts have read these statutory exceptions to permit employers to fire employees whose lawful off-duty activities are “disloyal.” For example, a district court upheld an employer’s decision to terminate a baggage handler with 26 years of service because he published a letter to the editor that criticized his employer for replacing full-time employees with hourly contract workers.111 Writing a letter was plainly “lawful off-duty activity,” but the court went on to read an “implied duty of loyalty” into the statute, limiting its reach to off-duty activities that were “personally distasteful” to the employer but “legal and unrelated to an employee’s job duties.”112 Then the court concluded that the employee breached the duty of loyalty when he went straight to the public without first “attempt[ing] to solve his grievance through Delta’s grievance system.”113 Of course, in the context of a business decision affecting tens of thousands of Delta employees, making an individual internal complaint is obviously futile, and so no aggrieved employee would be likely to take that step.
Finally, there is the possibility that narrow, judge-made exceptions to the at-will rule will apply when employees are fired for their political speech. One of these exceptions, the tort of “wrongful discharge in violation of public policy,” reflects that “certain discharges that contravene well-established norms of public policy harm not only the specific employee but also third parties and society as a whole.” That means that employees can file a civil lawsuit when they are fired for a reason that is not just bad or arbitrary but also violates an established public policy. Given that description, one might envision broad protections for employee speech premised, for example, on the democratic ideal that each member of the polity should share equal footing to participate in public discourse on matters of shared concern.
The most well-known decision that reflects this sort of speech-protective approach to the wrongful discharge tort is Novosel v. Nationwide Insurance Co.114 In that case, the plaintiff alleged that he was fired from his job as a district claims manager because he refused to participate in his employer’s lobbying the Pennsylvania legislature to reform no-fault insurance law. The U.S. Court of Appeals for the Third Circuit agreed that the plaintiff stated a claim for wrongful discharge in light of “the importance of the political and associational freedoms of the federal and state Constitutions.”115
However, the Novosel decision has been criticized, and it reflects at best a minority approach. The main problem is that courts in wrongful termination cases generally require employees to point to “well-established” public policy, and most require a high level of specificity—for example, that the employee was discharged for refusing to violate a law or regulation or for invoking rights under a workers’ compensation program. But a list of courts have held that the First Amendment is not sufficient to demonstrate that there is a well-established public policy against political coercion by private-sector employers, because the First Amendment constrains only government actors.116 Edmondson v. Shearer Lumber Products reflects the dominant approach;117 there, the Supreme Court of Idaho upheld the termination of a 22-year employee because he opposed a political initiative spearheaded by the employer’s owner.
Freedom of political belief is at the core of the idea of freedom of speech. But when private-sector employees face political coercion, they can fall back on legal protections only to a limited extent. Statutory protections, where they exist, tend to focus on protecting employees from retaliation in connection with their own political activity—voting or running for office. These statutes generally do not speak to what is likely a more common scenario, in which employers simply push their own political views, and employees say nothing because they fear being fired if they express disagreement.
2. Religious expression at work
Conflicts over religious expression at work can arise both when religious employers try to impose their faiths on employees and when religious employees bring their faith to work in ways that offend their bosses or coworkers. For example, employees hired to work for a secular employer might be dismayed to be asked to participate in the boss’s Bible-study group. Conversely, a religious employee might want to take breaks at set times in order to pray or might want to put a religious message like “have a blessed day” into an email signature or keep an anti-abortion message in a designated workspace. Unsurprisingly, law has something to say about the extent to which employers can regulate religious expression at work. These rules seek to balance employers’ and employees’ interests—although the balance struck weighs heavily in favor of employers, especially when the issue involves accommodating an employee’s religious practice.
Employers’ legal obligations surrounding their own or their employees’ religious speech vary with the employer’s own mission and character. For example, an employer with a religious mission, like a church or a religious school, has a large amount of leeway to require employees to adhere to a set of religious views. Religious institutions have constitutionally protected “independence…in matters of ‘faith and doctrine,’” including “internal management decisions that are essential to the institution’s central mission.” Importantly, this independence reaches “the selection of the individuals who play certain key roles”118—that is, who will be hired to convey a religious message and who will be fired from such a role. This principle has resulted in a kind of supercharged at-will doctrine that applies to employees who qualify as “ministerial.”119 Ministerial employees generally cannot invoke the protections of anti-discrimination law,120 probably are not covered by the NLRA,121 and may be precluded from relying on other statutory employment protections as well.122 And even nonministerial employees who work for religious institutions can lawfully be required by their employers to remain a member in good standing of the employers’ faith—a rule that can effectively turn adherence to a religion’s rules for personal comportment into job requirements.123
Conversely, public employers are constrained by both the Establishment Clause and the Free Exercise Clause; they cannot require employees to participate in religious programs or to belong to a particular faith, nor can they discriminate against employees because of their religious exercise. However, public employers also must guard against employee religious expression that would give the impression that the government itself prefers some religious views over others. For example, a public employer cannot permit its librarians or groundskeepers to create public displays that give the impression of a religious endorsement,124 nor may it allow an employee addressing the public to use the opportunity to proselytize.125
For both public- and private-sector employers that have a secular mission, statutory law modifies the at-will default in two important ways.126 First, these employers may not discriminate based on religion or allow harassment of employees based on whether, where, or how they worship on their own time. Second, they must accommodate employees’ religious beliefs or practices, at least where doing so will not impose an undue hardship.127 A reasonable accommodation might involve an employer partially waiving a uniform requirement so that an employee can wear a religious head covering,128 allowing the employee to take breaks at set times to pray, or changing an employee’s work schedule to allow attendance at religious services. However, the catch is that employers can refuse to provide such accommodations if they would impose more than a “de minimis” cost.129
The reasonable accommodation requirement relates to employees’ bargaining power in several ways. First, it bars an employer from dismissing a request for a religious accommodation out of hand, which means that even employees who lack any form of individual bargaining power are supposed to be able to secure an employer’s agreement to no- or low-cost religious accommodations.
Second, an employee’s request for a religious accommodation might lead to a form of workplace bargaining. The EEOC and some courts have concluded that, when faced with an employee request for a religious accommodation, an employer must embark on an “interactive process”—a back-and-forth dialogue to determine whether there exists an accommodation that would meet the employee’s religious exercise needs without unduly burdening the employer.130 But, as Professor Shirley Lin has described in an analogous context, power imbalances between employers and employees can easily lead to a failure to accommodate, either because the employee does not ask for the accommodation in the first place or because the employer steamrolls the employee in the interactive process.131
Third, other employees may also be interested in the outcome of an accommodations process, including because they might be asked to bear some or all of the costs of the accommodation. For example, an employer might respond to an employee’s request to not be scheduled to work on the sabbath by requiring another employee to work every Saturday. Or an employee’s desire to display religious materials might be accommodated even though the display upsets a coworker.132 But the accommodation requirement does not require employees to take into account the views of the burden-bearing employees; whether they do anyway can depend on a host of factors including whether the prospective accommodation would conflict with a collective bargaining agreement,133 the employer’s own views about the desirability of the accommodation, the likelihood that either employee will quit, and so on. In other words, whether an employer agrees to an employee’s request for a religious accommodation depends to an extent on the scope of legal protections—but can also depend on the employee’s (and maybe coworkers’) bargaining power.
3. Employee speech on social media
As discussed briefly above, the NLRA reaches concerted activity that takes place over social media. This means that if two employees interact on social media about their working conditions (or even if one employee posts a workplace concern that is intended to be an overture to coworkers), the NLRA may protect them against employer retaliation. But in addition, several states prohibit employers from demanding access to an applicant’s or employee’s social media posting. These statutes are privacy protections rather than speech protections; they do not prohibit employers from acting on an employee’s social media content that they see because of the employee’s own privacy settings or because other employees show the employer their colleague’s posting. However, they prohibit employers from requesting an employee’s or applicant’s social media username or password, implicitly providing some protection for workers’ expression, at least for workers who have privacy protections in place on their accounts.134 In addition, some states also bar employers from requiring employees to promote the company on social media or to add the employer to a social media contact list.135
D. A note on the First Amendment and public employees
This paper focuses mainly on private-sector workers who cannot rely on the Constitution’s free speech guarantee. But what about public-sector workers? First, the good news: public employees are covered by many of the statues discussed above, such as the anti-retaliation provisions of Title VII and various whistleblower protections. Further, many public employees are covered by some version of just-cause, rather than at-will, employment. Just over one-third of public employees are covered by a union contract,136 and others may enjoy similar tenure or civil-service protections. These substantive protections are also backed up by constitutionally required procedural protections; public employees who have a reasonable expectation of continued employment are entitled to notice and an opportunity to respond before they can be fired, plus a more formal post-termination hearing.137 (These procedural protections do not apply to at-will public employees.)
The bad news is that public employees often cannot rely on the First Amendment to protect them against employment consequences levied in connection with on-the-job speech. Like private-sector employers, public-sector employers may want to control their at-will employees’ speech for reasons that range from plausible to capricious. One might expect that courts would engage in closer scrutiny of public-sector employers, but they often do not. Instead, the Supreme Court has concluded that “government has significantly greater leeway in its dealings with citizen employees than it does when it brings its sovereign power to bear on citizens at large.”138
Public-sector employers may exercise near-total control over how their employees do their jobs without violating the Constitution. This is true even when an employee’s job involves expression and even when the employer has previously delegated to the employee some degree of discretion to shape the message.139 Courts reach similar results when employees, speaking outside of their employment capacity, express views on matters that are not of “public concern.”140 Moreover, the court usually does not view public employees’ own working conditions as a matter of public concern, unless those working conditions touch directly on the provision of public service.141
Public-sector employees receive somewhat more First Amendment protection when they speak as citizens—at least when their speech involves a matter of public concern. In that event, the court balances the public employer’s First Amendment interests against the public employer’s need for “orderly . . . administration.”142 Though the burden of proof is on the employer, this standard is very deferential to the government as compared to how courts approach other kinds of First Amendment cases.143 Moreover, the court defines “disruption” broadly, allowing public employers to rely on predicted (not actual) disruption144 and treating even such nonevents as having to respond to a lawsuit as “disruptions.”145 Thus, even when the First Amendment applies to a public employee’s speech, courts will still often defer to public employers’ assertions that their reactions were justified by disruption.
This framework applies to public employees’ speech generally, but there are also two special contexts that are treated differently. First, the court has held in a series of cases that political patronage is mostly unconstitutional: as long as a job does not involve policymaking, public employers may not require membership in a political party or support for a particular candidate or incumbent as a condition of employment.146 Conversely, the court has upheld restrictions on public employees’ political campaigning established by the Hatch Act and similar state laws.147 In these cases, the court has deferred to legislative judgments that public employees’ overt support for a particular political party or candidate might undermine public faith in government.148
These two lines of cases are related to each other inasmuch as Congress passed the Hatch Act in part to prevent political coercion in the form of pressure exerted on public employees by their supervisors or coworkers to campaign on behalf of a particular candidate or party.149 But their mechanisms are quite different: whereas the political patronage cases constrain employers, the Hatch Act constrains employees. It also reflects a big difference between public and private employees: the state and local statutes discussed above—patchy and incomplete as they are—often protect private-sector employees’ rights to engage in some of the same activities that the Hatch Act forbids for public employees.
IV. How the Constitution can undermine legal protections for employees
This report mainly focuses on the limited extent to which law protects workers from speech coercion by their employers. But there is also another angle to consider: employers’ arguments that certain workplace regulations designed to protect employees unconstitutionally infringe employers’ own constitutional rights. These arguments have gained traction in recent decades, and our increasingly conservative federal judiciary has proven receptive to them. Thus, there is a real risk that the Constitution itself will become an effective tool to undermine statutory protections for workers’ speech.
For example, suppose Congress were to try to increase unionization, which would in turn likely increase the number of workers covered by just-cause provisions in collective bargaining agreements. Suppose further that Congress went about this by increasing unions’ access to employees while they are at work and by barring employers from making some of the borderline-threatening statements that are currently permitted. Both changes can be understood as speech- and association-enhancing; the first would give employees more information about unions, and the second would allow employees to make their choice in a less-constrained environment. However, a conservative federal judiciary might strike down both changes, the first as an uncompensated taking of employer property and the second as a violation of the employer’s First Amendment rights.
To see how employers’ constitutional litigation could undermine workers’ rights, it helps to have a grasp of the basic structure of a First Amendment claim. Recall that the rights contained in the Bill of Rights are protections against government intrusions, and not private intrusions. (This is also why private-sector employees cannot bring successful First Amendment claims against their employers.) But when employers are regulated by the government, they can argue that those regulations unconstitutionally infringe their liberty; likewise, public-sector employees can bring constitutional claims because their employer is the government.
It is not inevitable that the Constitution will undo worker-protective legislation, but our current conservative-leaning federal judiciary and Supreme Court are likely to be receptive to such arguments. For proof, one need look no further than the decisions in Janus v. AFSCME and Cedar Point Nursery v. Hassid.150 In Janus, the court held that public-sector employees had a First Amendment right to refuse to pay union representation fees, notwithstanding that the union had a duty to fairly represent them. And in Cedar Point Nursery, the court struck down as an uncompensated taking of property a California regulation giving union organizers access to growers’ property for three hours per day, 120 days per year, for the purpose of speaking with farmworkers. In both cases, the court adopted a formalistic approach to interpreting the Constitution that placed little value on workers’ interests in associating with unions.151
At-will employment leaves workers vulnerable to losing their livelihoods for arbitrary reasons, including that they have offended their employer. Workers have been fired because they criticized their working conditions; because their political views were different than their employer’s; and because they let off steam on social media. These workers can rely on legal protections that are intended to protect certain types of speech—sometimes. But legal protections are limited and patchy, making it impossible for workers to know when they will be protected and when they will not be, and this degree of uncertainty is only compounded by the difficulties of enforcing one’s rights in court, post-discipline or termination. Employees are left to speak up at their own peril, undoubtedly making silence seem the safer choice.
1. See Feinman 1976 for a discussion of the origins of the at-will rule.
2. See also Restatement (Third) of Employment Law § 2.01 (Am. Law Inst. 2015): “Either party may terminate an employment relationship with or without cause unless the right to do so is limited by a statute, other law or public policy, or an agreement between the parties, a binding employer promise, or a binding employer policy statement.”
3. At the same time, state power can become intertwined with employer power, such as when work is made a condition of parole or probation (Zatz 2020).
4. The principle that neither the Constitution’s Bill of Rights nor the 14th or 15th Amendments constrain private citizens is known as the “state-action” requirement. There are a limited number of situations in which courts will treat nominally private actors as government entities for constitutional purposes. For example, in Marsh v. Alabama, the U.S. Supreme Court held that the owner of a “company town” should be treated as a government actor for purposes of evaluating a religious proselytizer’s claim that she had a First Amendment right to stand on a sidewalk to distribute religious literature, just as she would in a regular town. 326 U.S. 501, 509 (1946). More recently, the court has construed exceptions to the state action requirement narrowly, meaning that there are few situations in which employees would be able to argue successfully that their private-sector employer should be required to comply with constitutional standards. For a discussion of the role that the state-action requirement has played in the development of U.S. work law, see generally Lee 2014.
5. In the early days of the NLRA, the National Labor Relations Board (NLRB) adopted a version of this rule. Then, in 1941, the Supreme Court addressed an employer’s contention that the First Amendment protected its right to express its views that its employees should not unionize. The court wrote that employers were free to express their “view[s] on labor policies or problems,” although it also wrote that “[t]he mere fact that language merges into a course of conduct does not put that whole course without the range of otherwise applicable administrative power. In determining whether the Company actually interfered with, restrained, and coerced its employees the Board has a right to look at what the Company has said as well as what it has done.” NLRB v. VA Elec. & Power Co., 314 U.S. 469, 478 (1941). Then, in the 1947 Taft-Hartley amendments, Congress adopted what is known as § 8(c) of the NLRA, which states that the “expressing of any views, argument, or opinion . . . shall not constitute or be evidence of an unfair labor practice under any of the provisions of this subchapter, if such expression contains no threat of reprisal or force or promise of benefit.” In 2013, the D.C. Circuit relied on § 8(c) to strike down a rule requiring employers to post a notice informing employees of their rights under the NLRA.
6. This section mainly focuses on situations in which the at-will rule is reversed because employees have statutory protections. However, there are also a small number of judge-made exceptions to the at-will rule. Professor Lea VanderVelde enumerated five common-law limitations on the at-will rule: “1) implied in fact [contract], 2) promissory estoppel, 3) covenant of good faith and fair dealing, 4) public policy considerations that override private orderings, and 5) torts of abusive or outrageous conduct, such as intentional infliction of emotional distress” (VanderVelde 2012, 374).
7. On the other hand, collective bargaining agreements (CBAs) also often waive workers’ rights to engage in collective action. For example, no strike clauses are commonly included in CBAs, so that employers are essentially trading better wages and working conditions for “labor peace.” See Emporium Capwell Co. v. Western Addition Cmty Org., 420 U.S. 50 (1975).
8. In private-sector colleges and universities, tenure is, at minimum, a contractual guarantee that a tenured professor will not be dismissed without cause; in the public sector, tenure is often backed by a statutory guarantee. In addition to the substantive aspect of just-cause protection, tenured faculty usually have a right to due process before they can be disciplined or terminated. See Neumann 2017.
9. Some courts have treated employee handbooks as establishing an employment contract. However, these claims are tenable only when a handbook promises continued employment; as employers (and their lawyers) realized that a handbook could be construed as a contract, they could and did redraft handbooks to include disclaimers and emphasize employment at will. See Porter 2008, 67, and Pincus and Gillman 1983, 1009.
10. NLRB v. J. Weingarten, Inc., 420 U.S. 251 (1975).
11. The New York law was seemingly passed at least in part in response to reports of fast-food employers firing workers for their expression or self-presentation. See Eidelson 2021 (discussing a Chipotle worker who reported being fired because she wasn’t smiling at work, who then worked “with union organizers and local officials” on New York’s law).
12. Mt. St. 39-2-903.
13. Montana courts have interpreted the good-cause statute to protect whistleblowers, though they have also expressed unwillingness to micromanage employers’ decisions and emphasized that employers retain discretion, especially when deciding whether to fire manager-level employees. See Krebs v. Ryan Oldsmobile, 843 P.2d 312, 316 (1992) (good-cause statute protects employees who report employer’s illegal activity to police); Moe v. Butte-Silver Bow Cty., 371 P.3d 414, 427 (Mont. 2016); Bird v. Cascade Cty., 386 P.3d 602, 609 (Mont. 2016). For a comprehensive discussion of Montana’s law, including its procedures and limitations, see generally Corbett 2005.
14. MT. Stat. § 39-20905 (allowing employees to recover up to four years of lost wages and benefits, subject to the employee’s obligation to mitigate damages, and allowing punitive damages if the employee can prove by clear and convincing evidence that the employer “engaged in actual fraud or actual malice” in firing the employee).
20. Occasionally, state law will duplicate the NLRA’s protection for specific types of activity. For example, the NLRA protects workers who disclose their wages to each other in order to advocate for fairer pay policies—but in addition, a few states have adopted statutes that also protect workers’ pay disclosure. E.g., VT. Stat. Ann. Tit. 21 § 495(a)(8)(B).
21. E.g., Hoffman Plastic Compounds, Inc. v. NLRB, 535 U.S. 137 (2002) (holding that undocumented workers are covered by the NLRA but ineligible for backpay remedies).
22. E.g., NLRB v. Mackay Radio & Tel. Co., 304 U.S. 333 (1938).
23. 29 U.S.C. § 151.
24. 29 U.S.C. § 157. The full text of § 7 states that “[e]mployees shall have the right to self-organization, to form, join, or assist labor organizations, to bargain collectively through representatives of their own choosing, and to engage in other concerted activities for the purpose of collective bargaining or other mutual aid or protection, and shall also have the right to refrain from any or all of such activities except to the extent that such right may be affected by an agreement requiring membership in a labor organization as a condition of employment as authorized in” § 8(a)(3).
25. 29 U.S.C. § 158(a).
26. The NLRA excludes some significant categories of workers, including agricultural workers, public-sector workers, some domestic workers, independent contractors, managers, and supervisors. Employees who are categorically excluded from the NLRA may be covered by another law; for example, many states have established collective bargaining rights for public-sector workers, and a few states have established collective bargaining rights for agricultural workers. In addition, railway and airline workers are covered by the Railway Labor Act.
27. This argument has been made forcefully by others. Key works in this vein include Klare 1978; Atelson 1983; Pope 2004; and Getman 2016.
28. Finkin (1985) evocatively illustrates how labor law can leave workers unprotected for arbitrary and unpredictable reasons.
29. See Fresh & Easy Neighborhood Market, Inc., 361 NLRB 151 *4 (2014) (concerted activity is “that which is ‘engaged in with or on the authority of other employees,’” including “’where individual employees seek to initiate or to induce or to prepare for group action’” (citing Meyers Indus., 268 NLRB 493 (1984) & Meyers Indus., 281 NLRB 882 (1986)).
30. Compare Alleluia Cushion Co., 221 NLRB 999 (1975) (employee engaged in protected concerted activity when he advocated for safety measures that would benefit all employees, even though he had not “discussed the safety problems with the other employees, solicited their support in remedying the problems, or requested assistance” in preparing a letter to regulators), with Meyers Indus., Inc., 268 NLRB 493 (1984) (overruling Alleluia Cushion, and holding that individual employee’s effort to address unsafe trucks was not protected concerted activity).
31. First, individual employees are acting concertedly when they make an appeal on behalf of a group, such as when workers discuss a problem together and then designate one member of the group to discuss the issue with the boss. Meyers Indus., 281 NLRB 882, 887 (1986) (“Meyers II”). Second, individual employees act concertedly when they attempt to initiate group activity or make a statement that implicitly seeks support from coworkers, even if the attempt falls flat. Id. Third, individual employees retain NLRA protection when they continue earlier concerted activity, as when one employee asserts rights under a collective bargaining agreement. NLRB v. City Disposal Systems, 465 U.S. 822 (1984) (holding that employee was engaged in concerted activity when he refused to drive a truck that he thought was unsafe, where he reasonably believed that the applicable collective bargaining agreement gave him the right to do so).
32. Even when single employees have not yet engaged in protected concerted activity—perhaps they have griped with their coworkers but have not yet moved toward action—an employer can still violate the NLRA if it fires or disciplines that employee in order to nip potential collective action in the bud. See Parexel Int’l, 356 NLRB 516 (2011).
33. Eastex, Inc. v. NLRB, 437 U.S. 556 (1978)
34. 437 U.S. 556 (1978).
35. Id. at 569–70.
36. Earlier examples also exist. In Kaiser Engineers, the NLRB agreed that a group of engineers was protected by the NLRA when the members wrote a letter to several legislators opposing a request by the company Bechtel that the Department of Labor authorize resident visas for engineers recruited outside the country. 213 NLRB 752 (1974) (enf’d Kaiser Engineers v. NLRB, 538 F.2d 1379 (9th Cir. 1976).
37. Ronald Meisburg, Memorandum GC 08-10, Office of the General Counsel NLRB, July 22, 2008.
39. For a thorough discussion of legal risks that confront workers and unions participating in Days Without Immigrants and similar protests, see Duff 2007.
40. NLRB v. Motorola, Inc., 991 F.2d 278, 285 (5th Cir. 1993) (“Employees acting as members of outside political organizations cannot demand the same § 7 rights as employees engaged in self-organization, collective bargaining, or in self-representation in disputes with management simply because the organization focuses on a workplace issue”).
41. Harrah’s Lake Tahoe Resort Casino, 325 NLRB 1244 (1992).
42. Second Amended Complaint and Notice of Hearing, Google & Communications Workers of Am., Case No. 20-CA-252802.
43. NLRB v. Local Union No. 1229, IBEW, 346 U.S. 464 (1953) (“Jefferson Standard”).
44. Id. at 468.
46. The board stated that the technicians “did not misrepresent, at least willfully, the facts they cited to support their disparaging report.” 94 NLRB 1507, 1511 (1951).
47. Id. at 1509–10.
48. NLRB v. Local Union No. 1229, supra, at 472.
49. 861 F.3d 812 (8th Cir. 2017).
50. Id. at 816.
51. Id. at 817.
52. Id. at 822.
53. Id. at 825 (quoting Diamond Walnut Growers, Inc. v. NLRB, 113 F.3d 1259 (D.C. Cir. 1997)).
54. The “bargaining for the common good” approach, for example, is based on the principle that unions can “use contract fights as an opportunity to organize with community partners around a set of demands that benefit not just the bargaining unit, but also the wider community as a whole.” This kind of approach looks beyond working conditions to consider how an employer affects the larger community in which it is situated, and therefore could involve criticism of the employer’s products or services.
55. In Clear Pine Mouldings, Inc., the board adopted an objective test that protects picketers’ speech unless “the misconduct is such that, under the circumstances existing, it may reasonably tend to coerce or intimidate employees in the exercise of rights protected under the Act.” 268 NLRB 1044, 1046 (1984) (citing NLRB v. W.C. McQuaide, Inc., 552 F.2d 519, 527 (3d Cir. 1977). “Coerce or intimidate” is a high bar: under this standard, the board had held that strikers do not lose the protection of the NLRA even when they use racist or sexist epithets, provided they do not overtly or impliedly threaten violence.
56. This line of caselaw is relatively new, and the board began by taking a flexible approach. Still, the board’s cases in this area tended to side with employers only where the employee’s speech was “egregious.” See, e.g., Pier Sixty LLC, 362 NLRB 505 (2015) (employee did not lose NLRA protection after venting on social media that his supervisor was “a nasty mother fucker,” in context of workplace where “vulgar language” was “rife”).
57. In Atlantic Steel, the board announced four factors to determine if an employee could be discharged for opprobrious speech in a conversation with management: “(1) the place of the discussion; (2) the subject matter of the discussion; (3) the nature of the employee’s outburst; and (4) whether the outburst was, in any way, provoked by an employer’s unfair labor practice.” 245 NLRB 814, 816 (1979). One might characterize the Atlantic Steel factors as a way of getting at whether the employee’s outburst was understandable under the circumstances: an employee’s profane outburst would be understandable in a workplace where profanity is common; or in response to an unfair labor practice; or because the employee believes the employer has made a consequential employment decision based on arbitrary or discriminatory criteria.
58. General Motors, 369 NLRB No. 127 (July 21, 2020).
59. Id. at *3.
60. Similar to the court’s reasoning in Jefferson Standard, the board relied on § 10(c) of the NLRA to conclude that the act did not change employers’ baseline “right” to fire employees for uncivil speech. The board held that it would apply the Wright Line standard, which is typically applied in “mixed motive” cases—those in which the employer argues that although it retaliated against an employee in part because of the employee’s § 7-protected activity, it would have made the same decision even in the absence of the activity. There are two important aspects to this change. First, under the new rule, the NLRB general counsel has to prove that the employer had “animus” against the § 7 activity, rather than simply proving that the employer took action against the employee because of protected concerted activity. Proving a mindset can be considerably more difficult than proving a sequence of events, making even strong cases harder to win. Second, the new standard gives employers more leeway to set rules governing employee speech—including speech that qualifies as protected concerted activity—although employers still may not apply stricter civility rules to § 7-protected speech than speech on other topics.
61. Id. at *10–11.
62. Id. This may have been a violation of the NLRA, which, as part of the duty of good faith bargaining, requires employers to provide economic information about subcontracting decisions that relate to labor costs.
63. “Tone policing” occurs when a listener dismisses the substance of a speaker’s questions or comments by focusing on the style of their delivery.
64. Studies show that people tend to perceive Black men as “larger, more threatening, and potentially more harmful in an altercation than a white person” (Lopez 2017). Although one cannot know for sure, this dynamic could have been present in the incident that led to Robinson’s first suspension. That incident began with a heated conversation in which Robinson accused a manager of going back on his word to pay overtime for certain employees. The manager testified that “Robinson’s behavior ‘was too aggressive not to allow…some sort of disciplinary action to occur,’” and that his “‘fight or flight mechanism kicked into high gear.’” A key question—though one without an answer—is whether the manager would have felt the same if Robinson had a different racial and/or gender identity.
65. This section focuses on the rights of an employer’s own employees, but other board- and court-constructed rules also impose strict limits on outside union organizers’ access to employer property. See, e.g., Lechmere, Inc. v. NLRB, 502 U.S. 527 (1992) (holding that in nearly all circumstances, employers may bar union organizers from the employer’s property). For further discussion of the relationship between property rights and the NLRA, see Pope 2004, 521–22.
66. Workplace “fissuring” occurs when “lead firms that collectively determine the product market conditions in which wages and conditions are set…become separated from the actual employment of the workers who provide goods or services,” and “the direct employers of low wage workers operate in far more competitive markets that create conditions for noncompliance” (Weil 2011, 34).
67. Republic Aviation v. NLRB, 324 U.S. 793 (1945).
68. Id. at 803; see also Peyton Packing Co., 49 NLRB 828, 843.
69. Stoddard-Quirk Mfg. Co., 138 NLRB 615, 619 (1962).
70. “Special circumstances” could relate to safety, business operations, or potentially other reasons. For example, the board accepted that special circumstances justified a rule prohibiting employees from wearing a union logo on a nonbreakaway lanyard that posed a safety risk. Sam’s Club, 349 NLRB 1007 (2007). Less plausibly, the Trump board accepted Wal-Mart’s explanation that it needed to ban employees from wearing large union buttons while they were on the selling floor for a combination of “customer satisfaction” and security reasons—specifically that the buttons might distract from the employee’s name badge. Wal-Mart Stores, Inc., 368 NLRB Bo. 146 *4 (Dec. 16, 2019). It is notable an employer’s public image can sometimes count as a “special circumstance” that justifies restricting employees from wearing union insignia; in one case, the board wrote that special circumstances exist where employees wearing union insignia or clothing would “unreasonably interfere with a public image that the employer has established as part of its business plan, through appearance rules for its employees.” Boch Imps., Inc., 362 NLRB No. 83 at *2 (2015), enf’d 826 F.3d 558 (1st Cir. 2016).
71. Purple Communications, 361 NLRB No. 126 (2014). This decision reversed an earlier decision that reached the same decision as Caesar’s Entertainment. Register Guard, 351 NLRB 1110 (2007). For an insightful discussion of the Purple Communications decision, see Hirsch 2015.
72. Caesar’s Entertainment d/b/a/ Rio All-Suites Hotel & Casino, 386 NLRB No. 143 (2019). The decision prioritized employer property rights over employees’ § 7 rights, reasoning that in-person solicitation and distribution were “adequate avenues of communication” and that employees could always supplement those avenues by using their personal email, phone, or social media accounts.
73. Republic Aviation, supra, at 803 n.10 (quoting Peyton Packing).
74. Compare Sandusky Mall Co., 329 NLRB 618 (1999), enforcement denied, Sandusky Mall Co. v. NLRB, 242 F.3d 682 (6th Cir. 2000) (board held employer could not prohibit union solicitation where it had allowed nonunion promotional activities on its premises), with Kroger Lt. P’Ship, 368 NLRB No. 64 (2019) (holding employer could eject union solicitors from premises while allowing “access for a wide range of charitable, civic, and commercial activities that are not similar in nature to protest activities). In First Amendment terms, the difference is that previous boards prohibited employers from engaging in content discrimination, whereas the Trump board viewed the nondiscrimination principle as reaching only viewpoint discrimination.
Even when the board took a broader view, circuit courts sometimes refused to enforce the board’s orders. For example, in 6 West Ltd. Corp. v. NLRB, 237 F.3d 767 (7th Cir. 2001), the court refused to enforce a board order finding that the employer had committed an unfair labor practice when it allowed employees to sell Girl Scout cookies and Christmas ornaments but barred employees from soliciting union membership. The court put its conclusion in terms of the employer’s freedom: “A restaurant in the United States of America should be free to prohibit solicitation on the premises that interfere with or bother employees or customers, and allow those solicitations which neither interfere with nor bother employees or customers.” Id. at 780.
75. It is possible for an employee to be “jointly employed”—for example, it can be appropriate to treat both an entity that signs employees’ paychecks and a client that directs their day-to-day work as employers of the employees.
76. New York New York Hotel & Casino, 356 NLRB 907, 916 (2011), enf’d. 676 F.3d 193 (D.C. Cir. 2012), cert. denied 133 S.Ct. 1580 (2013).
77. Bexar Cty. Performing Arts Ctr Fdn., 368 NLRB No. 46 (2019).
78. 18 U.S.C. § 1514A.
79. See generally Tippett 2007.
81. Even here, though, there is a prerequisite: for the statute to protect employees who disclose to a public body, employees must first report the problem to their supervisor.
82. 42 U.S.C. § 2000e-3(a).
83. 42 U.S.C. § 12203(a) (“No person shall discriminate against any individual because such individual has opposed any act or practice made unlawful by [the ADA] or because such individual made a charge, testified, assisted, or participated in any manner in an investigation, proceeding, or hearing under [[the ADA].”); 29 U.S.C. § 623(d) (“It shall be unlawful for an employer to discriminate against any of his employees or applicants for employment…because such individual…has opposed any practice made unlawful by this section, or because such individual…has made a charge, testified, assisted, or participated in any manner in an investigation, proceeding, or litigation under [the ADEA].”)
84. For discussion and critique of this rule, see Carle 2016 and Eisenstadt and Geddes 2017.
85. 532 U.S. 268 (2001).
86. Id. at 269.
89. Id. at 270 (quoting Faragher v. Boca Raton, 524 U.S. 775, 786 (1998)).
90. Id. at 272 (alteration in original).
91. In a study that is now nearly 25 years old, Professor Pauline Kim surveyed “330 unemployed workers in the St. Louis metropolitan area,” and found that they “consistently overestimate the degree of job protection afforded by law, believing that employees have far greater rights not to be fired without good cause than they in fact have” (Kim 1997, 110).
92. Describing that Perkins’s lawyers told her that “he and his lawyers would try to destroy my credibility if I went to court. They told me he would try to destroy me and my family.”
93. Hamilton v. Juul Labs, Inc., No. 20-cv-03710, 2021 WL 5331451 (N.D. Cal. Nov. 16, 2021).
94. E.g., Cal. Labor Code § 1102.5. For a discussion of how nondisclosure agreements can potentially violate whistleblower laws, see Moberly, Thomas, and Zuckerman 2014; Rose and Rush 2020; and Lobel 2017 (discussing whistleblower protections under the Defend Trade Secrets Act).
95. EEOC v. Astra USA, Inc., 94 F.3d 738, 745 (1st Cir. 1996).
96. E.g., Brown v. TGS Management Co., 57 Cal.App. 5th 303 (Ct. App. 4th Div. 2020) (concluding that sweeping nondisclosure provision violated state restriction on noncompete agreements, because its breadth would “plainly bar Brown in perpetuity from doing any work in the securities filed, much less in his chosen profession of statistical arbitrage”).
97. E.g., Relco Locomotives, Inc., 358 NLRB 229 (2012) (affirming that nondisclosure agreement violated the NLRA because it “barred employees from disclosing to any third party information concerning ‘compensation, payments, correspondence, job history, reimbursements, and personnel records’” without authorization from “‘Relco’s Chief Legal Officer or Chief Administrative Officer,’” among other provisions).
98. Fla. Stat. § 69.081; see also Hemel 2017.
99. See generally Volokh 2012 (cataloging state laws protecting political activity).
100. Id. at 297–98.
101. NY Labor § 201-d(2).
102. See Wehlage v. Quinlan, 55 A.D.3d 1344 (N.Y. App. Div. 2008) (dismissing political discrimination claim where plaintiff alleged she was fired because of her political affiliation because plaintiff had not engaged in any of the activities enumerated in the definition of “political activities”); but see Richardson v. City of Saratoga Springs, 246 A.D.2d 900 (N.Y. App. Div. 1998) (stating, without discussion, that an employee had stated a claim for political discrimination when plaintiff alleged he was fired after wearing a sticker supporting a political candidate).
103. See Volokh 2012, 313–20.
104. Cal. Lab. Code § 1102.
105. Gay Law Students Ass’n v. Pacific Tel. & Tel. Co., 595 P.2d 592 (Cal. 1979).
106. N.J. Rev. Stat. § 34:19-10; Or. Rev. Stat. § 659.785.
107. “Lawful activities” statutes are offshoots of “lawful products” statutes. The latter protect employees’ consumption of products such as cigarettes and alcohol—though employees sometimes attempt to invoke these laws in new contexts, albeit generally without success. See, e.g., McGillen v. Plum Creek Timber Co., 964 P.2d 18 (Mont. 1998) (lawful products law did not protect employee who placed prank classified urging prospective buyers of a used truck to phone the employee’s boss at home late at night). For further discussion of these statutes, see Sprague 2008, 412–16; Melzer and Barth 2020; Bodie 2017; Finkin 2006; and Pagnattaro 2004.
108. § 96(k) of the state’s Labor Code allows the labor commissioner to take assignments of claims for “demotion, suspension, or discharge from employment for lawful conduct during nonworking hours away from the employer’s premises.” Cal. Lab. Code § 96(k). Then, § 98.6 of the Labor Code states that an employer “shall not discharge” or discriminate against employees because of “conduct described in” § 96(k). Taken together, one might read these provisions as establishing broad, unqualified protection for employees’ off-duty “lawful conduct” protection. But California courts have read this language to establish no new substantive right and instead to have only created an enforcement mechanism with respect to rights “otherwise protected by the Labor Code.” Grinzi v. San Diego Hospice Corp., 120 Cal.App.4th 72, 86–87 (Cal.App.Ct. 2004).
109. Colo. Rev. Stat. Ann. § 24-34-402-5.
111. Marsh v. Delta Air Lines, Inc., 952 F.Supp. 1458 (D. Colo. 1997); see also Oransky v. Martin Marietta Materials, Inc., 400 F.Supp. 3d 1143 (D. Colo. 2019) (employer did not violate the law by firing a sales employee who protested potential oil and gas exploration in her community because the oil and gas company that was the subject of the protest was also a client of the employer). In Watson v. Pub. Serv. Co. of CO, a Colorado court of appeals observed that “No Colorado appellate opinion has approved the Marsh court’s analysis.” 207 P.3d 860 (Colo.App. 2008).
112. Id. at 1462 (emphasis added).
113. Id. at 1463.
114. 721 F.2d 894 (3rd Cir. 1983).
115. Id. at 899.
116. See, e.g., Grinzi v. San Diego Hospice Corp., 120 Cal. App. 4th 72, 81-82 (2004) (citing cases from various jurisdictions); see also Restatement (Third) of Employment Law, supra, § 5.02 (“Most courts do not recognize wrongful-discharge claims against private employers based on free-speech rights because the federal and most state constitutional free-speech protections constrain governments, and thus do not apply to private-sector employers”).
117. 75 P.3d 733 (Idaho 2003).
118. Our Lady of Guadalupe Sch. v. Morrissey-Berru, 140 S.Ct. 2049, 2060 (2020).
119. The category of “ministerial employee” is broader than the phrase suggests; in addition to clergy, it includes other employees of religious institutions who play a role in shaping or communicating religious doctrine, including teachers in religious schools.
120. Id. (“Under [the ministerial exception] courts are bound to stay out of employment disputes involving those holding certain important positions with churches and other religious institutions.”)
121. See NLRB v. Catholic Bishop of Chicago, 40 U.S. 490, 504 (1979) (construing the NLRA to exclude parochial-school teachers from the act’s coverage).
122. See, e.g., Shaliehsabou v. Hebrew Home of Greater Wash., Inc., 363 F.3d 299 (4th Cir. 2004) (applying statutory ministerial exemption in wage-and-hour case and stating that the Fair Labor Standards Act’s ministerial exemption was coextensive with the constitutional ministerial exemption).
123. 42 U.S.C. § 2000e-1 (stating that Title VII does not apply to religious organizations “with respect to the employment of individuals of a particular religion to perform work” for the organization); Corp. of the Presiding Bishop v. Amos, 483 U.S. 327 (1987) (gymnasium run by religious entities associated with the Latter Day Saints church did not violate Title VII when it fired building engineer who “failed to qualify for…a certificate that he is a member of the Church and eligible to attend its temples”).
124. See McCreary Cty. v. ACLU, 545 U.S. 844 (2005) (upholding injunction requiring county to take down a courthouse display that included the Ten Commandments, where the surrounding context suggested an intent to promote a religious viewpoint).
125. Cf. Greece v. Galloway, 572 U.S. 565 (2014) (observing that government may not “proselytize or force truant constituents into the pews,” but upholding the legislative prayer practices at issue in the case because they were noncoercive).
126. This section discusses Title VII, but many states have analogous statutes, which may offer religious employees stronger protections.
127. 42 U.S.C. § 2000e(j).
128. See EEOC v. Abercrombie & Fitch Stores, Inc., 575 U.S. 768 (2015) (observing that while employers may have general “no headwear” policies, “when an applicant requires an accommodation as an ‘aspec[t] of religious…practice,’ it is no response that the subsequent ‘fail[ure]… to hire’ was due to an otherwise-neutral policy.”).
129. The statute requires employers to provide accommodations that do not impose “undue hardship,” but the Supreme Court interpreted that language to reach accommodations that would impose “more than a de minimis cost” on the employer. Trans World Airlines, Inc. v. Hardison, 432 U.S. 63, 77 (1977).
130. See Flake 2019, 83–87 (describing EEOC guidance and court decisions about whether employers must engage in an interactive process when responding to religious accommodation requests).
131. Lin 2021, 1828, 1836 (describing the “bargaining-based assumptions” embedded in the interactive process required under disability accommodations law and critiquing the model for failing to account for power imbalances between workers and employers). As Lin also points out, these problems with a bargaining-based approach to accommodations are likely to be especially serious where the employee lacks power for reasons including position in the workplace hierarchy, or marginalized personal identities.
132. See Wilson v. U.S. West Comm’ns, 58 F.3d 1337 (8th Cir. 1995) (holding employer satisfied its accommodation obligation when it allowed a worker to display religious materials at work but directed her to stop wearing (or cover up) an anti-abortion pin that included a color photograph of a fetus, which had been the subject of a complaint by another employee).
133. See Hardison, supra, at 79 (“we do not believe that the duty to accommodate requires TWA to take steps inconsistent with the otherwise valid [collective bargaining] agreement”).
134. See, e.g., MD Lab. & Empl. Code § 3-712; Cal. Lab. Code § 980; 820 Ill. Comp. Stat. 55/10.
135. See, e.g., Or. Rev. Stat. § 659A.330; Me. Rev. Stat. 26 § 616.
136. BLS 2022.
137. See Bd. of Regents v. Roth, 408 U.S. 564 (1972); Perry v. Sindermann, 408 U.S. 593 (1972).
138. Enquist v. Or. Dep’t of Agric., 553 U.S. 591, 599 (2008); see also NASA v. Nelson, 562 U.S. 134, 152 (2011) (“Like any employer, the Government is entitled to have its projects staffed by reliable, law-abiding persons who will efficiently and effectively discharge their duties” (internal quotations omitted).).
139. Garcetti v. Ceballos, 547 U.S. 410 (2006).
140. Connick v. Myers, 461 U.S. 138, 146 (1983) (“When employee expression cannot be fairly considered as relating to any matter of political, social, or other concern to the community, government officials should enjoy wide latitude in managing their offices, without intrusive oversight by the judiciary in the name of the First Amendment.”).
141. Id. at 148–49 (lawyer who distributed to her colleagues in the Orleans Parish District Attorney’s office a questionnaire about their shared working conditions was not speaking on matters of public concern when she asked about her coworkers’ confidence in supervisors, office morale, grievance procedures, and transfer policy; conversely, question about whether employees felt “pressured to work in political campaigns on behalf of office supported candidates” was of public concern); see also Waters v. Churchill, 511 U.S. 661, 665 & 680 (1994) (O’Connor, J.) (plurality opinion) (criticisms of hospital obstetrics department as a bad place to work would not qualify as a matter of public concern).
142. Pickering v. Bd. of Educ., 391 U.S. 563, 569 (1968).
143. For example, one way that speech could disrupt public service is if a lot of listeners assemble at a government building to protest. A standard that focuses on disruption thus allows for what the Supreme Court has called a “heckler’s veto”—a form of de facto viewpoint discrimination in which government gives in to listeners’ angry reactions and silences a speaker rather than incurring the costs necessary to protect an unpopular speaker, and which the court has said is not a basis to restrict speech in other contexts. See Forsyth Cty. v. Nationalist Movement, 505 U.S. 123 (1992) (striking down parade permit fee that varied based on expected costs of maintaining public order).
144. Waters v. Churchill, supra, at 674.
145. Borough of Duryea v. Guarnieri, 564 U.S. 379, 390 (2011).
146. Elrod v. Burns, 427 U.S. 347, 367-68 (Opinion of Brennan, J.) (patronage dismissals violate the First Amendment, except in the case of “policymaking positions”); Branti v. Finkel, 445 U.S. 507 (1980) (assistant public defenders could not be fired based on their political affiliations because “whatever policymaking occurs in the public defender’s office must relate to the needs of individual clients and not to any partisan political interests”). The court has also held that it violates the First Amendment for public employers to fire employees because of their perceived political affiliation. Heffernan v. City of Paterson, 16 S.Ct. 1412, 1416 (2016) (employer violated police officer’s First Amendment rights by firing him based on misperception that the officer supported a challenger to the incumbent mayor). Finally, the court has held that it violates the First Amendment for public employers to demand that employees take a “loyalty oath” indicating that the employee is not, and has not been within five years, a member of a communist or subversive organization. Wieman v. Updegraff, 344 U.S. 183 (1952).
147. United Public Workers of Am. v. Mitchell, 330 U.S. 75 (1947) (upholding Hatch Act’s prohibition on executive branch employees taking “any active part in political management or in political campaigns”); U.S. Civil Service Comm’n v. Nat’l Ass’n of Letter Carriers, 413 U.S. 548 (1973) (upholding Hatch Act, and stating that Congress may bar executive branch employees from a list of political activities, including “holding a party office, working at the polls, and acting as party paymaster for other party workers,” and also “organizing a political party or club; actively participating in fund-raising activities for a partisan candidate or political party; becoming a partisan candidate for, or campaigning for, an elective public office; actively managing the campaign of a partisan candidate for public office; initiating or circulating a partisan nominating petition or soliciting votes for a partisan candidate for public office; or serving as a delegate, alternate or proxy to a political party convention”).
148. Letter Carriers, supra, at 565–67 (Hatch Act prevents “machine politics” and promotes “confidence in the system of representative Government,” among other purposes).
149. Id. at 566–67.
150. 141 S.Ct. 2063 (2021).
151. A thorough criticism of these cases is beyond the scope of this report. However, there exists a substantial academic commentary on both cases, as well as on the deregulatory Constitution more generally. See, e.g., Bowie 2021 and Andrias 2018.
Abrams, Roger I., and Dennis R. Nolan. 1985. “Toward a Theory of ‘Just Cause’ in Employee Discipline Cases.” Duke Law Journal 1985: 594–623.
Allyn, Bobby. 2019. “Pa. Workers Forced to Choose Between Watching Trump, No Pay or Using Paid Time Off.” NPR, August 17.
Anderson, Elizabeth. 2017. Private Government: How Employers Rule Our Lives (and Why We Don’t Talk About It). Princeton University Press.
Andrias, Kate. 2018. “Janus’s Two Faces.” Supreme Court Review 2018: 21–58.
Atelson, James B. 1983. Values and Assumptions in American Labor Law. University of Massachusetts Press.
Bodie, Matthew T. 2017. “The Best Way Is Always Through: Changing the Employment At-Will Default Rule to Protect Personal Autonomy.” University of Illinois Law Review 2017: 223–69.
Bowie, Nikolas. 2021. “Antidemocracy.” Harvard Law Review 135: 160–219.
Bradford, Harry. 2017. “Murray Energy Miners Allege They Had to Give Up Pay to Attend ‘Mandatory’ Romney Rally.” HuffPost, December 6.
Bureau of Labor Statistics (BLS). 2022. “Union Members—2021.” Press release.
Carle, Susan D. 2016. “Angry Employees: Revisiting Insubordination in Title VII Cases.” Harvard Law & Policy Review 10, no. 1: 185–227.
Cassens Weiss, Debra. 2018. “Judge Tosses Wrongful Termination Claim by Woman Forced to Resign After Flipping Off Trump Motorcade.” ABA Journal, July 2.
Corbett, William L. 2005. “Resolving Employee Discharge Disputes under the Montana Wrongful Discharge Act (MWDA), Discharge Claims Arising Apart From the MWDA, and Practice and Procedure Issues in the Context of a Discharge Case.” Montana Law Review 66, no. 2: 329–404.
Duff, Michael C. 2007. “Days Without Immigrants: Analysis and Implications of the Treatment of Immigration Rallies Under the National Labor Relations Act.” Denver University Law Review 85, no. 1: 93–151.
Edwards, Kathryn Anne. 2022. Worker Mobility in Practice: Is Quitting a Right, or a Luxury? Economic Policy Institute.
Eidelson, Josh. 2020. “Covid Gag Rules at US Companies Are Putting Everyone at Risk.” Bloomberg BusinessWeek, August 27.
Eidelson, Josh. 2021. “Most Americans Can Be Fired for No Reason at Any Time but a New Law in New York Could Change That.” Bloomberg BusinessWeek, June 21.
Eisenstadt, Leora, and Deanna Geddes. 2017. “Suppressed Anger, Retaliation Doctrine, and Workplace Culture.” University of Pennsylvania Journal of Business Law 20, no. 1: 147–209.
Estlund, Cynthia L. 1995. “Free Speech and Due Process in the Workplace.” Indiana Law Journal 71, no. 1: 101–51.
Estlund, Cynthia L. 1996. “Wrongful Discharge Protections in an At-Will World.” Texas Law Review 74, no. 7: 1655–92.
Feinman, Jay M. 1976. “The Development of the Employment At Will Rule.” American Journal of Legal History 20, no. 2: 118-35.
Finkin, Matthew. 1985. “Labor Law by Boz—A Theory of Meyers Indus., Inc., Sears, Roebuck and Co., and Bird Engineering.” Iowa Law Review 71, no. 1: 155–98.
Finkin, Matthew W. 2006. “Life Away from Work.” Louisiana Law Review 66, no. 4: 945–74.
Flake, Dallan F. 2019. “Interactive Religious Accommodations.” Alabama Law Review 71: 67–114.
Garrahan, Matthew, 2017. “Harvey Weinstein: How Lawyers Kept a Lid on Sexual Harassment Claims.” Financial Times, October 23.
Getman, Julius G. 2016. The Supreme Court on Unions: Why Labor Law Is Failing American Workers. ILR Press.
Gray, Alex. 2016. “Freedom of Speech: Which Country Has the Most?” World Economic Forum.
Green, Aimee. 2019. “Lawsuit: Oregon Construction Worker Fired for Refusing to Attend Bible Study.” Oregon Live, January 29.
Green, Matthew W. Jr. 2014. “What’s So Reasonable About Reasonableness? Rejecting a Case Law-Centered Approach to Title VII’s Reasonable Belief Doctrine.” University of Kansas Law Review 62, no. 3: 759–812.
Hemel, Daniel. 2017. “How Nondisclosure Agreements Protect Sexual Predators.” Vox, October 13.
Hertel-Fernandez, Alex. 2018. Politics at Work: How Companies Turn Their Workers into Lobbyists. Oxford University Press.
Hirsch, Jeffrey M. 2015. “Worker Collective Action in the Digital Age.” West Virginia Law Review 117, no. 3: 921–59.
Kim, Pauline T. 1997. “Bargaining with Imperfect Information: A Study of Worker Perceptions of Legal Protection in an At-Will World.” Cornell Law Review 83, no. 1: 105–60.
Klare, Karl E. 1978. “Judicial Deradicalization of the Wagner Act and the Origins of Modern Legal Consciousness, 1937–1941.” Minnesota Law Review 62: 265–339.
Johnson, Jeffrey. 2019. “Non-Disclosure Agreements and Arbitration Clauses in the #MeToo Era.” JD Supra, August 29.
Lee, Sophia Z. 2014. The Workplace Constitution from the New Deal to the New Right. Cambridge University Press.
Lin, Shirley. 2021. “Bargaining for Integration.” New York University Law Review 96: 1826–1901.
Lobel, Orly. 2017. “The DTSA and the New Secrecy Ecology.” Business, Entrepreneurship, & Tax Law Review 1, no. 2: 369–82.
Lobel, Orly. 2018. “NDAs Are Out of Control. Here’s What Needs to Change.” Harvard Business Review, January 30.
Lopez, German. 2017. “Study: People See Black Men as Larger and More Threatening than Similarly Sized White Men.” Vox, March 17.
Melzer, Andrew, and Whittney Barth. 2020. “Whether Employees Can Be Fired for Participating in Peaceful Protests.” Illinois Law Review Online 2020: 221–34.
Mishel, Lawrence. 2022. The Legal “Freedom of Contract” Framework Is Flawed because It Ignores the Persistent Absence of Full Employment. Economic Policy Institute.
Moberly, Richard, Jordan A. Thomas, and Jason Zuckerman. 2014. “De Facto Gag Clauses: The Legality of Employment Agreements that Undermine Dodd-Frank’s Whistleblower Provisions.” ABA Journal of Labor & Employment Law 30, no. 1: 87–120.
Neumann, Richard K. Jr. 2017. “Academic Freedom, Job Security, and Costs.” Journal of Legal Education 66: 595–605.
Naidu, Suresh, and Michael Carr. 2022. If You Don’t Like Your Job, Can You Always Quit? Pervasive Monopsony Power and Freedom in the Labor Market. Economic Policy Institute.
Noah, Timothy. 2004. “Bumper Sticker Insubordination.” Slate, September 14.
Oswalt, Michael M. 2017. “The Right to Improvise in Low-Wage Work.” Cardozo Law Review 38, no. 3: 959–1040.
Pagnattaro, Marisa Anne. 2004. “What Do You Do When You Are Not at Work? Limiting the Use of Off-Duty Conduct as the Basis for Adverse Employment Decisions.” University of Pennsylvania Journal of Labor & Employment Law 6, no. 3: 625–84.
Perman, Stacy. 2018. “#MeToo Law Restricts Use of Nondisclosure Agreements in Sexual Misconduct Cases.” Los Angeles Times, December 31.
Phelan, Jessica. 2012. “Mitt Romney Tells Employers to ‘Make It Very Clear’ to Employees How to Vote.” The World, October 18.
Pincus, S. Richard, and Steven L. Gillman. 1983. “The Common Law Contract and Tort Rights of Union Employees: What Effect After the Demise of the At Will Doctrine.” Chicago-Kent Law Review 59, no. 4: 1007–39.
Pope, James Gray. 2004. “How American Workers Lost the Right to Strike, and Other Tales.” Michigan Law Review 103, no. 3: 518–53.
Porter, Nicole B. 2008. “The Perfect Compromise: Bridging the Gap Between At-Will Employment and Just Cause.” Nebraska Law Review 87, no. 1: 62–121.
Rose, Rachel V., and Joshua M. Russ. 2020. “Whistleblowing, Separation Agreements, and the Enforceability of Contractual Limitations.” Health Lawyer 32, no. 5: 61–6.
Rothstein, Mark A., Lance Liebman, and Kimberly A. Yuracko. 2015. Employment Law: Cases and Materials. 8th Edition. Foundation Press.
Scheiber, Noam, and Brian M. Rosenthal. 2020. “Nurses and Doctors Speaking Out on Safety Now Risk Their Job.” New York Times, April 9.
Shammas, Brittany, and Hanna Knowles. 2021. “A Trader Joe’s Employee Called for Stronger Coronavirus Measures. The Company Fired Him.” Washington Post, February 28.
Smith, Terry. 2003. “Everyday Indignities: Race, Retaliation, and the Promise of Title VII.” Columbia Human Rights Law Review 34, no. 3: 529–74.
Spooner, Rachel S. 2020. “The Goldilocks Approach: Finding the ‘Just Right’ Legal Limit on Nondisclosure Agreements.” Hofstra Labor & Employment Law Journal 37, no. 2: 331–78.
Sprague, Robert. 2008. “Rethinking Information Privacy in an Age of Online Transparency.” Hofstra Labor & Employment Law Journal 25, no. 2: 395–417.
Stone, Will. 2020. “An ER Doctor Lost His Job After Criticizing His Hospital on COVID-19. Now He’s Suing.” NPR, May 29.
Tippett, Elizabeth C. 2007. “The Promise of Compelled Whistleblowing: What the Corporate Governance Provisions of Sarbanes Oxley Mean for Employment Law.” Employee Rights and Employment Policy Journal 11, no. 1: 1–52.
Tippett, Elizabeth C. 2018. “Harassment Trainings: A Content Analysis.” Berkeley Journal of Employment & Labor Law 39, no. 2: 481–526.
United Auto Workers (UAW) and Ford Motor Company. 2019. Agreements Between UAW and the Ford Motor Company, Volume 1, Agreements Dated October 30, 2019 (Effective November 18, 2019).
VanderVelde, Lea. 2012. “The Proposed Restatement of Employment Law at Midpoint.” Employee Rights and Employment Policy Journal 16, no. 2: 359–93.
Volokh, Eugene. 2012. “Private Employees’ Speech and Political Activity: Statutory Protection Against Employer Retaliation.” Texas Review of Law & Policy 16, no. 2: 295–336.
Weil, David. 2011. “Enforcing Labour Standards in Fissured Workplaces: The U.S. Experience.” Economic and Labor Relations Review 22, no. 2: 3–54.
Zatz, Noah. 2020. “Get to Work or Go to Jail: State Violence and the Racialized Production of Precarious Work.” Law and Social Inquiry 45, no. 2: 304–38.