Cynthia L. Estlund<1>;
Introduction
Imagine a society whose citizens had free speech rights but no due process rights. The government could imprison citizens or banish them for any reason, or no reason at all, without notice or a hearing or proof of the charges; but it could not punish citizens based on their criticism of the government or other protected speech. The citizen who believed she had in fact been punished for speaking against the government could go to court and, if she could prove it, secure relief. How free would speech be in such a system? Would the citizens feel free to challenge the regime without fear of retaliation?
The questions are not idle ones, for the system that I have described is essentially that which prevails in the American workplace for the at-will employee, both public and private. Employees in both the public and the private sectors have free speech rights: the Constitution, state and federal statutes, and the common law purport to protect employees' rights to speak on various matters free from employer retaliation. But this "system of freedom of expression<2>; in the workplace lacks a crucial element of the system of freedom of expression in the society under the federal Constitution: many employees in both the public sector and the private sector enjoy no general legal protection against arbitrary discipline or discharge. They can be fired at will, for any reason or for no reason at all, subject to their ability to mount a legal challenge and to prove that the employer's true reason was the desire to punish the employee for protected speech (or was some other forbidden motive). Under this system, as in my hypothetical regime, the formal principle of freedom of expression may be largely illusory, for retaliation is readily cloaked in the guise of arbitrariness, and uncloaking it is likely to be too difficult, too time-consuming, and too costly to offer much reassurance to the typical employee.
The difficulties of proving an employer's motivation and the costs and delays of litigation undermine the effectiveness of many wrongful discharge remedies, including those against discrimination based on race, religion, ethnicity, sex, age, and handicap. But there is an additional element that should concern us when we seek to protect voluntary conduct by an individual, such as reporting of employer wrongdoing, rather than classes of individuals. The hurdles faced by the employee who is fired because of protected speech are not invisible to the employee who is deciding whether to speak up. An employee who considers whether to blow the whistle on her employer for unlawful conduct or to advocate unionization among her coworkers, but who fears she may lose her job with no readily available means of redress, is likely to be deterred from speaking. Notwithstanding the formal prohibition of retaliation against certain kinds of speech, we should expect reasonable employees to be "chilled" from speaking freely when it may put their jobs at risk. This is an important but largely unrecognized cost of the "at-will" presumption that underlies employment law for most American employees.
In this description of the workplace, it may seem that I am ignoring a crucial distinction. It is commonly thought that public employees enjoy freedom of expression; private sector employees do not. But I want to challenge that dichotomy and suggest another more important one. I will argue that real freedom of expression in the workplace depends less on the applicability of First Amendment protections than on the existence of independent protections against discharge without just cause and associated procedural rights. These substantive and procedural due process rights, whether grounded in the Constitution or elsewhere, are the key to realizing the free speech rights that we already afford in principle to both public and private employees. I do not aim in this Article to question the doctrine of state action that keeps the First Amendment out of virtually all employment disputes in the private sector.<3>; Rather, putting the state action question to one side is part of my strategy for uncovering and comparing the functional elements of the two systems of freedom of expression.<4>;
I will thus contend that at-will public employees—though they enjoy free speech rights against their employer under the First Amendment—are extremely vulnerable without the underpinnings of basic due process protections against unjustified discharge. The Supreme Court's recent decision in Waters v. Churchill<5>; takes a small, but still too small, step toward recognizing this link between free speech and due process. Turning to the private sector, I will argue that there is a "system of freedom of expression" consisting of a patchwork of discrete protections against employer retaliation for socially valued speech. But I will argue that the virtual absence, except in the dwindling union sector, of any functioning system of fair treatment and due process renders those speech protections nearly meaningless for the vast majority of employees. By contrast, in workplaces where the employer must show an impartial decisionmaker that there is good cause for discharge, as in the typical unionized workplace, employees enjoy much greater freedom of expression—perhaps more than many public employees protected by the First Amendment.
In short, I am making claims about the freedom of speech that employees experience at work in light of their legal environment. These contentions should be stated in the form of hypotheses. No analysis of the doctrine or case law can demonstrate that the legal environment which I describe "chills" employee freedom of speech. Evidence for my contentions would be found not in the cases adjudicating claims of retaliation, but in the silence that surrounds them—the silence of the typical employee who is neither uncommonly brave nor litigious, whose job is crucially important to her and her family, and who is guided in her actions by expectations about the consequences of those actions. There is some empirical evidence to support these claims, but the data are incomplete.<6>; In the meantime, it seems fair to assume that employees respond sensibly to the expected costs and benefits of speaking out. As things stand, much speech that is ostensibly protected remains very costly and produces little or no private benefit to the employee.
However we define the speech that we value and seek to protect in the workplace, that speech will not be free if the speaker must prove her right to speak and must bear the burdens of delay, uncertainty, or error in decisionmaking. Freedom of expression in the workplace, as in the society at large, requires "breathing room." I submit that the best way to give breathing room to valued workplace speech is to support existing protections of expression with a regime of substantive and procedural due process. Due 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. Due process would thus help to provide the necessary breathing room for the freedom of expression that we already purport to recognize.
I anticipate three fairly sweeping objections to a universal just cause requirement: It won't work; it isn't necessary; and it costs too much. The objection that a legally mandated just cause regime will not protect employees against retaliation and will not support freedom of expression strikes at the heart of my thesis. My belief in the value of due process is based partly on standard economic reasoning: employees should normally be expected to act on the basis of reasonable expectations about the consequences of speaking out and particularly about the possibility of discharge. The greater ease, availability, and timeliness of a remedy against retaliatory discharge offered under a due process system rather than under a wrongful discharge system reduces the cost of speaking out. Empirical evidence bolsters this analysis, including experience in the union setting. But that experience may indicate the need for some form of institutional support for employees seeking to challenge their discharge, such as some type of employee representation within or outside the workplace, in addition to individual due process.
The second objection is that a legal mandate of due process is not necessary to support legally protected speech because the threat of wrongful discharge liability will deter employers from wrongful conduct or will lead them to take precautions in the form of procedural safeguards. But the speech that is protected by law can be quite costly to employers for the same reasons that it is valuable to the public; that is why legal protection is necessary. Given the real benefits that employers might gain from the discharge of a disruptive speaker, the deterrence of other potential speakers, and the hurdles that employees face in remedying a retaliatory discharge, the prospect of liability may often be inadequate to deter punishment of valuable and legally protected speech.
The third and related objection, that universal due process is too costly, is for some the most telling objection. Given the very small percentage of discharges that are probably based on valuable or protected speech, universal just cause protection appears to be a very big and intrusive solution to a relatively small problem. Of course, the current wrongful discharge regime, only a part of which concerns the protection of speech, is itself extremely costly to employers, employees, and the economy as a whole. If it is as ineffective as I claim it is, then the question of costs begins to look quite different. But the question of whether due process is too costly goes beyond the present project, for there are both costs and benefits to due process that lie far beyond the issue of workplace speech. My argument is therefore not that the benefits of due process for employee freedom of expression are sufficient to justify instituting a universal just cause regime, but that securing employee freedom of expression is an important and largely unrecognized benefit of moving toward such a regime.
There is one final response to my claim of drastic underenforcement of speech protections: "You get what you pay for." More protection would entail greater interference with managerial prerogatives and greater administrative or litigation costs as well as more of the ostensibly desirable speech. Perhaps the existing weak enforcement mechanisms afford all the protection of employee speech for which we as a society are willing to pay. For those who take this view, the greater part of my argument, which exposes the enormous gap between what the law apparently promises in terms of freedom of speech and what the law delivers, will fall on deaf ears: "So what? You simply failed to read the fine print." Stated differently: Why should we protect the employee speech that we claim to protect?
It is partly in anticipation of this question that I begin in Part I by discussing what is at stake in the protection of employee speech. Against this background I proceed in Part II to describe what employee speech the law currently protects. I conclude that existing law represents a promise of employee freedom of expression that is limited in scope—in some cases too limited—but that it is a promise well worth fulfilling, for the speech that is protected is of great value in the workplace and the society. In Part III, I examine the limited empirical evidence of the effectiveness of existing legal protections. I look at both the prevalence of discharges based on protected speech, and more importantly, the extent of freedom of speech that employees experience. The limited empirical evidence available suggests a large gap between what the law promises and what it delivers: many employees do not experience even the limited freedom of expression that the law purports to grant them.
In Parts IV and V, I return to existing law, first in the public sector and then in the private sector, to seek an explanation for that gap. But I seek that explanation not in the content of what is protected, but in the procedures and structures through which these protections are administered. In particular, I examine how those structures and procedures operate against the backgrounds both of at-will employment—which underlies the system of freedom of expression in many public sector and most private sector employment relations—and of a system of due process and just cause. In both the public and private sectors, I conclude that the current regime of wrongful discharge remedies, against the heavy backdrop of the at-will presumption, provides plainly inadequate remedies for employees fired on the basis of protected speech, but that a functioning due process system acts as an indirect but effective bulwark for the protection of employee speech. Finally, in Part VI, I will discuss very briefly some aspects of how due process safeguards might be implemented as part of the system of freedom of expression in the public and private sector workplace.
I. The Value of Employee Freedom of Speech: Three Views of the Workplace
Why should we care about freedom of expression in the workplace?<7>; To begin with, most adults spend a great part of their waking hours at work. The intrinsic value of free speech for individual autonomy, and its value in fostering individual self-realization and fulfillment, transcends the boundaries established by the state action doctrine. Freedom of expression at work enhances human development and well-being no less than freedom of expression in civic life.<8>;
But speech is not a solitary exercise; it presupposes a listener, a conversation, and a community. Speech in the workplace, like speech in the society, can benefit the audience and the society as a whole as well as the speaker. The values served depend on how we view the workplace as an institution. First, we can see the workplace from the inside as a microcosmic society governed by management. We can also see the workplace from the outside as an object of societal control. Finally, we can see the workplace as an intermediate institution, an institution that mediates between individuals and the society as a whole, in which individuals cultivate some of the values, habits, and traits that carry over to their roles as citizens. Each of these perspectives on the workplace sheds light on the value of employee speech.<9>;
A. Speech and the Workplace as a Self-Governed Community
The workplace is an internally governed community.<10>; The employee-citizens of the workplace community are governed by management.<11>; The form of government may vary. Some workplace governments are highly autocratic, relying heavily on hierarchy and close supervision. Few are governed "democratically" by the citizen-employees themselves, though many workplace governments include mechanisms for participation and representation of employees.
There is something like "law" within the workplace community as well. Most of the "law" that governs the conduct of the citizen-employees is the law enacted by management more or less unilaterally: rules of conduct, standards of performance, and the hierarchies and systems established by management for enforcing these rules and standards. In the unionized workplace, the employees, through their union, play a role in establishing some of the "law of the shop." The law established by management, written or unwritten, with or without the participation of a union, is what regulates the day-to-day conduct of the government and its citizens.
The law of the shop includes regulation of speech. Employers might seek to prohibit or suppress speech that helps to enforce external regulatory norms and that threatens the firm with outside scrutiny or sanctions. Employers might also regulate or punish speech that threatens the internal structures, relationships, and hierarchies through which it governs, such as harassment, rudeness, or insubordination. And just as governments may punish sedition, employers may seek to punish speech that, from their vantage point, foments rebellion against the existing workplace regime.<12>; In many nonunion workplaces it appears to be widely understood that employees shall not discuss the possibility of unionization and that violation of this "rule" is punishable by discharge.<13>; From the employee's perspective, this is the law of the shop.<14>;
The workplace government is constrained, however, by something analogous to a constitution: the whole body of public law that governs the workplace from the outside, including the laws protecting some kinds of employee speech against employer sanctions. These laws, the content of which is discussed below, establish the framework and the minimum standards within which workplace law operates. They are no more subject to change by the internal governing structure of the workplace than is the federal constitution by a local government. The amalgam of public laws protecting speech in the private sector workplace thus functions as a kind of quasi-First Amendment for the workplace just as the First Amendment itself, in modified form, functions like a constitutional limit on the internal law of the public sector workplace. I will discuss the content and structure of this quasi-First Amendment below. For the moment, I want to focus on the protections of workplace speech.
Thinking about the workplace as a management-governed community puts a new spin on the relation between free speech and intelligent self-government. It is an article of faith that the free speech of the citizenry is essential to democratic self-governance in the society at large. Might employee free speech play an analogous role in fostering informed self-governance within the workplace?
To pose that question may seem to finesse an enormous difference between the political community and the workplace. In our society the ideal of democratic self-government is a given. It constitutes the society and the manner in which we pursue other societal objectives, and that is what makes free speech on public issues so self-evidently necessary. By contrast, the ideal of workplace democracy is at best contested. The private firm, in particular, is organized to compete effectively in the relevant product market and to generate profits for its owners, not to foster the domestic well-being of its citizen- employees. Even in the public sector workplace, the instrumental goals of carrying out governmental objectives and delivering public services take precedence over the democratic nature of the workplace as an institution.<15>; So whatever role free speech may have in improving democratic self-government, the question may seem misplaced in the workplace and particularly in the private profitmaking firm.
But recent developments in industrial relations scholarship and practice, as well as in corporate law and scholarship, have laid a foundation for asking this question. While employee self-governance is hardly a given, as it is in the polity as a whole, employee participation in workplace governance is increasingly viewed as both an intrinsic and an instrumental good. The intrinsic value of employee participation can be seen in the simple fact that employees are more satisfied with their work when they have an opportunity to express themselves and potentially influence workplace decisions.<16>; But more than job satisfaction is at stake, for 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.
Moral claims such as these have long been overshadowed by issues of economic efficiency in discussions of private firm governance,<17>; but they have a new generation of champions in the field of corporate law. For many observers, the wave of mergers and acquisitions of the 1980's exposed the extent to which shareholder primacy subordinated the pressing concerns of other "stakeholders" in the private corporation, including workers and their communities.<18>; In the wake of these developments has come a broad current of dissatisfaction with the long-prevailing paradigms of corporate law, which give shareholders the dominant claim on legal protection while leaving other parties in the "nexus of contracts" to bargain with whatever chips their circumstances endow them.<19>; For some of these new communitarians of corporate law and theory, employee voice—the opportunity to speak and to have an influence on decisions in the firm—is valued because it fulfills human needs and helps to foster a humane working environment.
Employee voice has also won champions among labor economists. In much modern thinking about workplace relations and the economy, the ideal workplace of the future is the "high-performance," high-wage, high-skill workplace. It is widely held, and was recently affirmed by the Presidential Commission on the Future of Worker-Management Relations (or the Dunlop Commission), that thoroughgoing cooperation and partnership between workers and management is a crucial prerequisite to a high-performance economy.<20>; Although economists remain divided over these claims, a good deal of research shows that employee-management cooperation, particularly through the medium of collective bargaining, tends to improve productivity at the firm level by supporting employees' investment in firm-specific skills, enhancing the exchange of knowledge, and improving employee morale, loyalty, and longevity.<21>;
Cooperation, in turn, requires mechanisms for fostering employee "voice" in the workplace. The Dunlop Commission, among others, has focused on the organizational structures for employee voice.<22>; There has been a profusion of scholarship chronicling the decline of traditional union representation and the gradual rise, at the margins of existing law, of nonunion, management-initiated forms of employee participation.<23>; Legal scholars have asked how the law can foster effective forms of employee representation including, but not limited to, traditional union representation and collective bargaining.
There is good reason to focus on participatory institutions. Employee voice, to be effective in workplace governance and in monitoring regulatory compliance, must be channeled into workable and representative structures with power within the workplace; it is important to explore alternative forms of employee representation as well as means of reinforcing the traditional right to choose union representation. But I believe that in touting the virtues and evaluating the organizational structures of employee voice without directly examining the system of employee freedom of expression, we have neglected the obvious. To be free to participate in workplace governance, to initiate new or better alternative forms of employee representation, and to choose traditional union representation, employees must be free to speak out, among themselves and with supervisors and management, about workplace issues.<24>; Employees cannot be expected to act as full partners and to take the initiative in forming and reforming effective representative structures in an environment where they do not feel free to criticize and question management without risking their jobs.<25>;
Many firms have recognized the benefits of employee voice and participation in firm governance, at least to some degree, and many educated observers have urged a much wider dissemination of participatory mechanisms.<26>; It is evident that a wholesale shift toward codetermination and away from unfettered management prerogatives would require a dramatic cultural transformation in American business. Cultural transformation is not easily legislated. But freedom of expression, and the legal structures necessary to protect it, is something with which the American legal system has a wealth of experience. We know that self-governance in the political sphere depends on free public debate, and that free public debate requires dissident viewpoints and strong legal guarantees against government censorship and repression of information. Now that the economists have rediscovered the instrumental virtues of participation in workplace governance, it is time to take another look at the system of freedom of expression that, in the workplace as in the society, lies at the foundation of a participatory system of governance.
The workplace is not only a self-governing community but is also an object of regulation by the government and the citizens at large. Protection of some employee speech facilitates the informed and intelligent exercise of societal control of the workplace. This is true in different ways in both the public sector and the private sector. Citizens govern the public sector workplace—government agencies and institutions—both through legislation and through political processes, including elections. Public employees are an important source of information about government operations and conduct. Protecting public employee speech from employer reprisals potentially enhances the quality of citizens' political judgments and decisions about how public agencies should run and who should run them.<27>;
Much of what the government does, however, is to regulate the private sector. Just as public employees may provide information to the public about how government works, private employees may provide information to the public about how private firms operate with regard to working conditions, product safety, environmental practices, and other matters in which the society has a well-established regulatory interest. Assuming (as I will here) that there is a legitimate public interest behind the underlying regulations—workplace safety and antidiscrimination laws, environmental and product safety regulations, etc.—then there is a public interest as well in information of the sort that employees often have about compliance with the regulations.<28>; Employee speech about corporate conduct may also inform public debate about whether more or less or different forms of regulation are wise.<29>;
Whistleblowing directed to public or government officials is the most dramatic example of employee speech that enhances the regulatory process. But less visible, less dramatic forms of employee speech—the day-to-day speech of ordinary employees within the workplace—may be the necessary first step toward improving safety and other working conditions as well as compliance with environmental, safety, financial, and other regulatory requirements.
Employee speech on these matters can be understood not simply as good for the public but as a "public good" in the economic sense.<30>; Most dramatically, reporting of employer misconduct and threats to public safety produce goods, in the form of information, for the public at large. Moreover, the society relies heavily on employee monitoring and reporting as key elements of the enforcement scheme for most workplace regulations, many of which necessarily protect workers as a group as well as interests of the public at large in adequate working conditions.<31>; Both types of employee reporting produce public goods. The employee cannot keep to herself the benefits of the information, nor can she generally share with others the significant costs she incurs in producing the information—the cost of learning about legal requirements and about the employer's conduct, as well as the risk of retaliation from the employer.<32>; Protecting employee speech that informs the government or the public or even the employer about illegal or otherwise harmful conduct or conditions within the workplace may mitigate the public goods problem, and encourage employees to produce this valued information.<33>;
C. Speech and the Workplace as Intermediate Institution
The workplace functions not only as a self-governing institution and as a regulated institution; it also functions as a crucial intermediate institution that stands between the individual and the state.<34>; The workplace is one institution in which most adults can and must interact with others—initially strangers, often from diverse cultural, ethnic, political, and religious backgrounds—in a constructive way toward common aims. Indeed, as other intermediate communities and institutions such as neighborhoods, families, and religious congregations are battered and destabilized by the pressures of economic change and geographic mobility, the workplace is increasingly one of the few organic communities of a human scale in which many members of the society participate on a regular basis.<35>; To that degree, the workplace has become an increasingly important site for the forging of those crosscutting ties that help bind together a diverse society and for the formation of "civic virtues": the habits and traits and beliefs that make good citizens.<36>;
This picture of the workplace community is surely idealized, but to the extent that it is an ideal worth aspiring to, it requires that people feel reasonably free to speak their minds on what matters to them at work.<37>; The ideal of the workplace as a "school for democracy" requires respect for some of the basic values and freedoms of political democracy. I do not think such respect in the workplace requires the full range of freedom of expression that the First Amendment demands in the society at large, however.
Not all employee speech advances these values. The ideal of the workplace as an effective mediating institution requires a degree of civility, mutual respect, and tolerance that some speech may undermine.<38>; These concerns should not overshadow the value of freedom of expression in the workplace in cultivating habits of active and responsible citizenship in the society at large. But they do complicate the picture by raising the prickly problem of harmful workplace speech.
By "harmful workplace speech" I do not mean speech that the employer deems disruptive and harmful for the same reasons that it benefits workers or the public, such as speech that exposes the employer to deserved criticism or legal liability<39>; or discussion among coworkers of unionization.<40>; That is not what I mean by "harmful speech.<41>
But other speech, such as personal insults, divisive gossip, racial slurs, sexual taunts, propositions, innuendo, and the like may harm coworkers and undermine employee morale, cooperative relations, and ultimately productivity within the workplace. Some speech undermines the norms of equality that are, like the speech protections discussed here, part of the "constitution" of the workplace community. Virtually all employers rely to some degree (and I have argued that they should rely increasingly) on cooperation and harmony among employees at different levels of management. How do we treat employee speech that erodes the relationships on which a more democratic and less hierarchical workplace depends?
These same dilemmas have generated passionate debate within constitutional law and theory over the extent to which the government may or even must restrict some kinds of harmful speech in order to effectuate equality, to enable the often silent to speak, or to preserve the civility and tolerance that holds together a community within which free speech is possible and meaningful.<42>; These dilemmas are largely absent from the positive law of freedom of expression within the workplace, for in the private sector workplace there is no First Amendment at all, nor indeed any general principle of freedom of expression; the legal protections of socially valuable speech are like islands of protection in a sea of employer discretion.<43>; The landscape of the public sector, in which the First Amendment does apply, is not so very different, for speech is protected only if it is on matters of public concern, and even then only if it is not too "disruptive.<44>; So speech that is harmful, and that is not protected by positive law, is freely punishable by the at-will employer, public or private.
Of course, to the extent that the government itself punishes workplace speech in the private sector, the First Amendment is implicated. The emerging debate over the constitutionality of Title VII's prohibition of sexual and racial harassment parallels in many respects the broader controversy over the regulation of hate speech and raises important questions about workplace speech within the constitutional scheme.<45>; Those questions are not unrelated to the questions explored here, but they would take me well into the realm of First Amendment jurisprudence and far afield from the focus of this Article. I leave them for another day.<46>; For the moment I am concerned with why we might choose to extend legal protection to some employee speech. While some employee speech may erode the network of relations that make the workplace a crucial mediating institution, and may to some extent be appropriately proscribed by law, there can be no doubt that some level of employee freedom of expression is also essential to that mediating function, and calls for the protection of some speech from employer reprisals.
These three perspectives on the workplace as an institution—the workplace as a self- governed community, the workplace as an object of regulation, and the workplace as intermediate institution—provide a normative background against which to examine the positive law. To what extent does the law reflect a societal appreciation for the value of employee speech and the importance of freedom from employer reprisals? What employee speech does the law protect? I take up these questions in Part II.
II. The Promise of Free Speech in the Workplace: What Employee Speech Does the Law Protect?
Most basically, legal protection of free speech takes the form of a prohibition against punishment or suppression of speech. The First Amendment thus bars most government punishment or censorship of its citizens' speech. When we consider freedom of speech within the workplace, the employer takes the place of the government, and the prohibition and punishment of speech is accomplished primarily through discipline and discharge. Freedom of speech in the workplace, in this Article, is the freedom to speak out at or about the workplace free from the threat of discharge or serious discipline.<47>; Thus, the First Amendment bars the government employer in most cases from discharging a public employee because of protected speech. Similarly, the many speech protections governing the private workplace prohibit or make actionable the discharge of an employee based on a particular sort of message or speech.
Beginning with the enactment of the National Labor Relations Act in 1935, and increasingly since 1960, legislatures and courts have extended the principle of freedom of expression into public and private workplaces by prohibiting employer reprisals based on various forms of communication.<48>; I will begin with the more familiar developments in the public sector, where the First Amendment applies directly. Then I will turn to the more interesting and motley picture in the private sector, where a patchwork of legal protections form a kind of quasi-First Amendment for private sector employees.
A. The Scope of Protected Speech in Public Sector Employment
In Pickering v. Board of Education,<49>; the Supreme Court announced that the First Amendment constrains government not only in its capacity as sovereign, but also, albeit to a lesser degree, in its capacity as employer. In Connick v. Myers,<50>; the Court gave sharper form to the doctrine, establishing a threshold requirement that the employee's speech be on a matter of public concern and, for speech that meets that threshold, a balancing test in which the disruption (or fear of disruption) of the agency's operations is generally decisive.
The public concern test concisely, if not neatly, circumscribes the content of the speech of public employees that is protected by the First Amendment against employer punishment.<51>; Quintessentially, employees speak on matters of public concern when they report dereliction of public duties, corruption, or threats to public health or safety.<52>; On the other hand, the public concern test is often used to deny protection to employees' expression of grievances or criticism concerning workplace conditions.<53>; Indeed, in the not uncommon cases in which speech is both a workplace grievance and of potential concern to the public, the employee's personal interest in the matter often disqualifies her speech from protection.<54>; First Amendment doctrine thus embraces the view of the public sector workplace as an object of governance and gives no support to the value of speech in promoting self-governance and participation within the workplace,<55>; nor in fostering the mediating function of the workplace.
The First Amendment is not the only source of protection for public employee speech.<56>; Federal civil service employees are legally shielded by the Whistleblower Protection Act<57>; against reprisals for reporting fraud, waste, and abuse.<58>; Many states have also enacted "whistleblower protection" laws that generally prohibit reprisals against public employees who report serious fraud, waste, or illegality to some proper authority or law enforcement official.<59>; In addition, most federal employees and some state and municipal employees enjoy statutory protections of speech relating to unionization and collective bargaining.<60>;
But only the First Amendment protects all public employees, and for present purposes it will serve as a useful paradigm for protected public employee speech. I will return below to the question of how and how effectively these protections work. For now, however, let us turn to the private sector, in which the content of the employee speech that is legally protected is much less susceptible to concise description than it is in the public sector.
B. The Scope of Protected Speech in Private Sector Employment
Employees in the private sector have, of course, no constitutional free speech rights to raise against their employer's decision to fire them. Indeed, normally employers have the right to fire their employees at will, for good reason, bad reason, or no reason at all. But the right to fire for "bad reasons" has come under sustained legal assault. In particular, a variety of federal and state statutes and common law doctrines protect particular kinds of employee speech from employer retaliation. This amalgam of legal constraints forms, in effect, a quasi-First Amendment doctrine that constrains the governance of the workplace by management. It stands outside of and above the government of the workplace much as the Constitution stands outside of and above the governments of the society.
One of the most dramatic developments in employment law has been the rise of "whistleblower protection" laws and doctrines. The common law protections, which take the form of "public policy exceptions" to the employer's power to fire employees at will, are perhaps best-known.<61>; Very recently, for example, a California court recognized the claim of an employee who was fired for complaining to his employer that many employees were working overtime without being paid overtime wages.<62>;
Less well-known are a large and growing number of statutory whistleblower protections that prohibit reprisals against employees who make complaints, bring charges, or participate in proceedings under various regulatory statutes. Some of these statutes protect employees seeking to vindicate their own interests, individually or as part of the workforce. Antiretaliation provisions are thus routinely included in federal statutes regulating terms and conditions of employment, such as antidiscrimination laws, health and safety laws, minimum wage and maximum hours laws, and pension laws.<63>; Other whistleblower provisions protect information of concern outside the workplace to the public at large. Thus, many federal laws that regulate the conduct of private firms outside the employment context, such as pollution control laws, prohibit retaliation against employees who report violations or participate in proceedings against their employer.<64>;
Several states have enacted broader "whistleblower protection" statutes providing limited remedies for employees who suffer employer retaliation for reporting specified kinds of wrongdoing to public authorities.<65>; The protection that these statutes afford varies widely and is severely limited in many instances; most significantly, employees may not be protected when they report wrongdoing to officials within their own organization.<66>; Still, these laws and doctrines together protect a good deal of employee speech that employers might be inclined to suppress. These provisions suggest an emerging principle of protection for employees who report employer wrongdoing.
If whistleblowing is our paradigm of protected employee speech, as it is in the public sector under the "public concern" requirement, then we would expect very limited protection of more mundane speech relating to internal workplace governance: employee grievances, criticism of supervisors, suggestions for improved work practices, and even discussions of unionization. But ironically, as a formal matter, most private sector employees enjoy greater freedom of expression than do most public employees on the issues that matter to them most at work.<67>; Indeed, the single most sweeping legal protection of employee speech in the private sector is the National Labor Relations Act, which gives employees the right to engage in "concerted activity for . . . mutual aid or protection." Section 7 applies not only to employees who are, or who seek to be, represented by a union but to almost all private sector employees, provided that they act "in concert" with one or more coworkers.<68>; The NLRA is rarely used by and is largely unfamiliar to nonunion employees outside the organizing context.<69>; But section 7 is a potentially significant source of free speech rights in the workplace on issues of concern to workers; it protects speech about unionization or other forms of employee representation, discussion of work-related grievances, and petitioning for their redress.<70>; The NLRA protects a great deal of employee communication that promotes participation and self-governance within the workplace.
What unites these various legal protections of employee speech in the private sector is more axiomatic than insightful: Each protects speech that some employers have shown a proclivity to punish. In each case some legal authority has concluded that the interest of the employee in speaking freely, and in many cases the interest of the public or of coworkers in hearing the message, outweighs the employer's interest in suppressing that message. But how effective are these protections? How free is speech that the law purports to protect?
III. Wrongs Unremedied and Speech Unspoken: Some Empirical Evidence on the State of Freedom of Expression in the Workplace
What is most important to know about the actual state of freedom of expression in the workplace cannot be gleaned from ordinary legal materials. First, how prevalent is employer retaliation against protected speech? And second, to what extent do employees feel free—free from the threat of employer retaliation—to engage in protected speech? Even the first question seems virtually unanswerable, for proof of the actual prevalence of retaliation would require a mini-trial of unknown cases that never occurred.<71>; But the available evidence strongly suggests that employees continue to experience retaliation and the threat of retaliation for speech that the law purports to protect and that employee silence resounds where speech should be free.
The most detailed information on these questions comes from the federal government, whose employees are protected by the First Amendment and by the 1989 Whistleblower Protection Act.<72>; The Whistleblower Protection Act prohibits reprisals against employees for reporting illegal or wasteful activity within the agency; the study, for the Merit Systems Protection Board ("MSPB"), defined whistleblowing as reporting such activity to someone other than a friend, coworker, or family member.<73>; The 1992 study of over 13,000 federal employees found the following: 18% had personally observed or seen evidence of illegal or wasteful activities in their agencies.<74>; Approximately 50% of those observers had reported the wrongdoing to someone who might be in a position to address the problem.<75>; Of those who did not report the wrongdoing, 59% explained that they did not think anything would be done; 33% said they feared retaliation.<76>;
Of those who did report wrongdoing, 37% believed they had been subject to some form of retaliation.<77>; But almost all of the reported retaliation was more subtle than discharge, demotion, or suspension. For example, employees cited poor performance appraisals (47%), shunning by coworkers or managers (49%), verbal harassment or intimidation (47%), or assignment to less desirable duties (37%).<78>; Of those employees who took action against perceived reprisals, only 3% reported positive outcomes; most said that reporting the reprisals either caused them more trouble or made no difference.<79>; As compared to earlier studies, the 1992 study found significantly greater willingness to speak out. The study also found more reported reprisals, albeit fewer of the more official and more serious kind.<80>;
There is one glaring problem with this study and with much of the existing data on whistleblowers and retaliation: because the study surveyed current employees, it systematically excluded any employees who had been fired or who had quit in response to perceived reprisals. Still, the study is instructive. It indicates that many employees who observe wrongdoing do not report it, that many of those who do so perceive employer retaliation, albeit of a comparatively mild variety, and that many of those who do not report wrongdoing attribute their unwillingness to speak out to the fear of retaliation. That these findings were made in the federal civil service, whose employees enjoy greater job security than just about any other group of employees, raises serious concerns about the state of freedom of expression in the private sector workplace.
Other studies of whistleblowing reached similar findings. These studies tend to focus on managerial or professional employees; some fail to distinguish between public and private sector employees, and most fail to distinguish between disclosures that are protected by law and other disclosures of wrongful conduct within the organization. But these studies, together with the MSPB's study of federal employees, make up the largest body of empirical data on employee free speech. The studies generally have found that, as in the federal government, many employees who observe misconduct do not report it to anyone.<81>; A significant number of those who do speak out report that they suffered retaliation, though most reported retaliation is of a comparatively mild type, and the reported incidence of retaliation varies dramatically from one study to another.<82>; Still, it has proven harder to demonstrate decisively that the silence of many is due to a fear of retaliation.<83>;
The data are less detailed and less rigorous as to other employee groups and other kinds of speech. The single most prolific source of employee free speech claims in the private sector is the NLRA's prohibition of employer retaliation based on union activity and other "concerted activity" on workplace issues. In recent times, approximately 10,000 employees per year have been ordered reinstated as a result of findings that they were discharged on the basis of protected activity. Many of these were fired on the basis of union organizing activity such as solicitation of union support, discussions with coworkers about unionization, and other forms of expressive activity.<84>; These reinstatements represent only a small fraction of the discrimination charges filed. Even if many of the charges are frivolous, it seems probable that many meritorious charges do not result in a remedy or even an adjudication.<85>; It thus appears that nonunion employers are demonstrably and increasingly willing to suppress union talk in violation of a sixty- year-old federal statute.
In light of prevalence of discriminatory discharges, it is not surprising that most employees—more than in the past—believe that pro-union talk meets with serious employer reprisals. The Dunlop Commission recently reported the results of a 1991 poll in which 79% of the employees surveyed agreed that it was either "very" or "somewhat" likely "that nonunion workers will get fired if they try to organize a union." Of employed nonunion respondents, 59% said it was likely they would lose favor with their employer if they supported an organizing drive; 41% agreed with the statement "it is likely that I will lose my job if I tried to form a union.<86>; It is worth noting as well that union organizing activity is at nearly unprecedented low levels.<87>; Scholars have disagreed, however, over whether these phenomena are related.<88>;
Union organizing seems to be the most harshly suppressed and frequently adjudicated of all forms of protected workplace speech. The effectiveness of union talk depends on the speaker's ability to gain a following; thus retaliation may appear to be a particularly attractive option for the employer who deeply opposes the formation of a union. And because, by definition, there is a well-informed employee advocate on the scene, claims of anti-union retaliation are particularly likely to become formal charges. But these data strongly suggest that many employers do violate the law and fire employees who speak out contrary to the perceived interests of the firm.
The picture is a bit different, and the data still more incomplete, with respect to other kinds of speech. For example, in spite of the dramatic expansion of wrongful discharge doctrines, civil litigation under all those doctrines remains "rare" in relation to both total employment and the number of discharges,<89>; and it is concentrated among more highly paid employees who can afford attorneys.<90>; In states with private sector whistleblower protection statutes, there is very little litigation under those statutes.<91>; With the glaring exception of union speech under the NLRA, there are surprisingly few proceedings under the federal statutes protecting private sector employee speech; for example, most of the whistleblower protection provisions of federal health, safety, and environmental statutes are virtually dormant.<92>; Perhaps the most widely used of these provisions is the Occupational Safety and Health Act, 29 U.S.C. §§ 651-678. Several thousand complaints are filed each year, but only a tiny fraction of them result in any relief.<93>;
The paucity of remedies and even adjudications under these antiretaliation statutes may be due to the infrequency of serious retaliation. It may also be due to the ignorance of individual employees as to what their rights are and how to protect them. Most disturbing, and most difficult to ascertain, is the possibility that many employees remain silent because they are simply unwilling to take the chance that they will suffer reprisals. There is evidence, for example, that employees routinely ignore workplace safety violations, although safety violations are more regularly reported and redressed in union shops than in nonunion shops.<94>;
The same pattern is repeated in other areas in which workers have the "right" to complain about unlawful working conditions free from employer reprisals. Employees have the right under Title VII, for example, to report incidents of discrimination and harassment, and employers are prohibited from retaliation against them.<95>; Yet many of those who do complain of sexual harassment end up being fired or quitting in response to what they regard as employer retaliation.<96>; Many incidents of perceived sexual harassment go unreported, and many would-be complainants cite a fear of reprisals as a reason for remaining silent.<97>;
The data, though they are incomplete and uneven, portray a sizable gap between the freedom of expression that the law purports to secure to employees and the freedom of expression that many employees experience. Clearly there are forces other than fear of employer reprisals that silence employees: concern about peer disapproval, pessimism about the utility of speech, and sheer inertia are powerful censors.<98>; But fear of discharge or other serious economic reprisals is an especially strong deterrent, and it is one that the law purports to redress. Yet many employees appear unwilling to test the protection that the law affords them, and many of those who do so experience retaliation notwithstanding the law's protections. This incongruence between what the law promises and what employees experience raises a further question: Why does the law fail to effectively secure freedom of expression even in the limited areas singled out for protection?
Thus far I have described the content of the employee speech that is protected against employer retaliation. But the "system of freedom of expression<99>; in the workplace, like the system of freedom of expression under the First Amendment generally, is not only the content of the law's protections but also the mechanisms by which the law operates: the procedures, burdens of proof, and presumptions through which the law's protections are administered. A closer look at these aspects of the system of freedom of expression in the workplace offers some strong clues to the very partial success of that system. Part IV discusses these issues in the public sector workplace; Part V concerns the more complex—but, as it turns out, largely parallel—system of freedom of expression in the private sector.
IV. The System of Freedom of Expression in the Public Sector Workplace and the Role of Due Process
All public employees are protected by the First Amendment. If a public employee is fired on the basis of protected speech, she may bring a constitutional "wrongful discharge" action against her employers. But not all public employees enjoy freedom of speech. I will argue that those employees whose sole recourse against a retaliatory discharge is the slow, costly, and unreliable wrongful discharge remedy—that is, the many public employees who are terminable at will, without a reason and without a hearing—enjoy correspondingly limited freedom of expression. But those employees who enjoy independent due process rights—who cannot be fired without a good reason or without notice and a hearing—should be expected to enjoy much greater freedom of expression as well. I will begin by examining the workings of a public employee's First Amendment lawsuit challenging her discharge.
A. The Wrongful Discharge Model of Free Speech Protections in the Public Sector Workplace
The First Amendment operates in public employment quite differently than it does in society at large. The First Amendment generally establishes a broad principle of protection with a number of discrete exceptions for "unprotected speech." It is for the government to bear the burden of proving that a particular message or communication falls outside the realm of protected speech. Under Connick v. Myers, by contrast, the First Amendment in the governmental workplace creates an exception—speech on matters of public concern—to a general background principle of no protection, or employer discretion; the burden of proving protection rests on the employee.<100>; The principle of freedom of expression in public employment is, in short, a "defining-in" sort of system, in Professor Schauer's terminology, as opposed to the "defining-out" structure that prevails under the First Amendment generally.<101>;
Once the public employee-speaker has overcome the threshold "public concern" requirement and has shown that what she said constitutes protected speech, she must undertake the formidable burden of proving that her discharge was motivated by her speech. This is no mean feat in the case of a reasonably knowledgeable employer and an imperfect employee. Even if the employee can prove the existence of an unconstitutional motive, the employer has the chance to prove that it would have fired the employee anyway for reasons unrelated to her speech.<102>;
Even if the employee proves that her discharge was based on constitutionally protected speech, she must still contend with far greater deference to the government's claimed justification for the discharge than would the citizen-at-large under the First Amendment: the government is required to show neither a compelling government interest nor the necessity of its means, but rather only "disruption" or a threat of disruption that outweighs the employee's interest in speaking out.<103>;
The Supreme Court's decision last term in Waters v. Churchill<104>; offers a window on how these elements interact with the "public concern" test, and adds an important twist. In Waters, a public hospital fired nurse Cheryl Churchill—apparently an at-will employee—based on what her superiors believed were disparaging comments about her supervisors and the department.<105>; Churchill sued the hospital under § 1983<106> for firing her on the basis of constitutionally protected speech. According to Churchill, she did not disparage the department, but rather criticized a hospital policy of dealing with staff shortages by using "cross-trainees" from other departments, a policy she believed threatened patient care.<107>; Churchill argued that her comments were clearly on a matter of public concern and were not sufficiently disruptive to justify discharge.<108>;
The question before the Supreme Court was which of three possible versions of the employee's speech should be subjected to the Connick test: Is it what the employer honestly believed the employee said? Is it what the employer reasonably and honestly believed the employee said, based on its investigation of the facts? Or is it what the employee actually said, according to the ultimate judicial factfinder? Given these three possible standards of liability—intentional wrongdoing, negligence, or strict liability—the Court split three ways.
Justice Scalia, joined by Justices Kennedy and Thomas, took the narrowest view of the First Amendment: The government employer violates the First Amendment rights of public employees only if it acts on the basis of an unconstitutional retaliatory motive. The employee should thus have no claim if her statements, as the employer honestly (but perhaps unreasonably) believed them to be, were unprotected. Churchill should therefore lose unless she proves on remand that the unprotected statements at issue were a mere pretext for firing her, and that she was actually fired because of past protected statements about hospital policy.<109>;
Justices Stevens and Blackmun, on the other hand, took the broadest view of First Amendment liability: all that should matter is what the employee actually said, regardless of how reasonably the employer believes that she said something different and unprotected. The government may be wise to investigate what was said before firing an employee based on her speech, but under the dissenters' view the government would be liable nonetheless if the ultimate judicial factfinder later found that the employee actually said something else that was protected.<110>;
But the plurality—Justices O'Connor, Rehnquist, Souter, and Ginsburg—took a middle view: The government does not incur First Amendment liability if the employee's statements, as the employer honestly and reasonably believed them to be, were unprotected. On this view the hospital in Waters, having conducted a reasonable investigation of what Churchill said,<111>; could thus prevail even if Churchill might ultimately persuade the judicial factfinder that she said something different that was protected. The case was remanded, however, to give Churchill a chance to prove that her unprotected statements—or the version of those statements that the employer reasonably believed—were a pretext for firing her, and that she was in fact fired for other protected speech.<112>;
The Waters plurality built on a line of decisions imposing procedural safeguards as necessary adjuncts to the First Amendment.<113>; But "First Amendment due process," as it has evolved in these cases, is only a minor manifestation of a much more vital connection between free speech rights and due process of law. Basic procedural constraints on adverse government action provide an indispensable foundation for freedom of speech: Without independent due process rights against government deprivations of life, liberty, and property, the First Amendment would offer dissenters a very fragile shield against state retaliation. How free would speech be if the government could punish the speaker, ostensibly for speeding or theft or murder or even some unwritten crime, without notice of the prohibition and of the charges, and without probable cause or a hearing or the burden of proving its case to an impartial tribunal? How free would speech be if the speaker's only remedy in such a case were to sue the government and prove that it acted on the basis of her dissent? Yet this is roughly where existing doctrine, and every member of the Court in Waters, would leave the at-will public employee—an employee who, like Churchill, has no civil service protection or tenure, and who can be fired for any reason at any time.
The analogy may seem hyperbolic: discharge from a job is not equivalent to imprisonment or execution. But in fact termination of employment may be more serious than many smaller criminal penalties that trigger the full panoply of constitutional due process rights. Termination of employment is likely to have a harsher impact on one's life and well-being, and carry a greater social stigma, than would a modest criminal fine. It is no accident that, in the law and language of labor arbitration, discharge is regarded as "industrial capital punishment.<114>;
Notwithstanding the existence of free speech rights under the First Amendment, the dissenting employee who can generally be fired for no good reason and without a hearing is extremely vulnerable to retaliatory discharge. Justice Douglas made this connection in his dissent in Board of Regents of State College v. Roth: "Without a statement of the reasons for the discharge and an opportunity to rebut those reasons . . . there is no means short of a lawsuit to safeguard the right not to be discharged for the exercise of First Amendment guarantees.<115>; He thus argued for the extension of due process rights to all public employees faced with dismissal, in part to realize and support these employees' substantive constitutional rights. But the majority of the Court rejected this view in Roth,<116>; and none of the opinions in Waters resurrects it.<117>;
Consider the hurdles that face an at-will public employee who has publicly criticized her superiors' performance of their public duties, and who is subsequently fired. She may be given no reason, or she may be told that her performance has been unsatisfactory. She has no right to notice or documentation of the alleged inadequacies, or an opportunity to refute the charges. If the employee believes that she was fired because of protected speech, and if she can find a lawyer to represent her, she can file a lawsuit in which she may attempt to prove—largely through the employer's documents and witnesses still employed by the employer—that her speech was on a matter of public concern, that her discharge was actually motivated by her speech, and that the speech was not unduly disruptive.<118>; In the meantime the employee may be unemployed under circumstances that may handicap her in getting another job.
In making the link between procedural safeguards and substantive protection of speech for public employees, the Waters plurality took an important step. But in squeezing this insight into a realm in which due process is generally absent ("at-will" public employees who have no "property interest" in continued employment) the plurality created a minefield of uncertainty and absurdity that Justice Scalia gleefully exposed. Most strikingly, Justice Scalia pointed out that "[i]n the present case, for example, if the requisite `First Amendment investigation' disclosed that Nurse Churchill had not been demeaning her superiors, but had been complaining about the perennial end-of-season slump of the Chicago Cubs, her dismissal, erroneous as it was, would have been perfectly OK.<119>;
The problem is exaggerated in a bureaucratic context in which the ultimate decision to discharge is typically made not by the immediate supervisor or by anyone with direct knowledge of the employee's conduct and performance, but by a higher management official based on information from others. Thus, if a dissenting employee is reported, perhaps by a hostile supervisor, to have made some disruptive, unprotected comments, the decisionmaker presumably has some duty under the plurality opinion to investigate the report's accuracy before discharging the employee on that basis. But if the same employee is reported by the same hostile supervisor to have been late to work six times, or to have done her job badly, or to have worn a disagreeably brightly colored dress, the decisionmaker may fire her summarily with no need to show that its grounds were reasonable or based in fact (subject to the employee's possible effort in litigation to prove an unconstitutional motive). The Waters plurality effectively imposed a duty to investigate only where "an employment action is based on what an employee supposedly said, and a reasonable supervisor would recognize that there is a substantial likelihood that what was actually said was protected.<120>;
Justice Scalia ascribed these absurdities to the plurality's requirement that the government act reasonably in cases involving potentially protected speech. But I would put it differently: Absurdities arise when the government is required to act reasonably only in cases involving speech—that is, when the government is not generally required to act reasonably in its treatment of those employees, like Churchill, whom it has chosen to employ on an at-will basis. The general power of the government to act arbitrarily and without reasons in its treatment of some of its employees—to fire them if it wishes for "complaining about the perennial end-of-season slump of the Chicago Cubs,<121>; or for wearing a brightly colored dress—renders somewhat incongruous and ill-fitting the plurality's minimal requirements of reasonableness in a particular kind of case.
The Waters decision adds nothing—and even Justice Stevens' dissent adds nothing—to the protection of the at-will public employee against retaliatory discharges that are unexplained or explained by something other than speech. The members of the Court divide over what happens in the case in which the employer relies explicitly on speech as the basis for a discharge. But where the employer gives no reason, or gives a reason that has nothing to do with speech, all of the opinions in Waters would leave the employee where she was: Her only remedy comes by way of a federal lawsuit in which she must prove that the discharge was based on protected speech. For the employee contemplating public criticism of her employer, the prospect of prevailing, perhaps after many years, in such a lawsuit, must look like a long shot. It is a gamble that many employees may reasonably decline to take.
Waters v. Churchill adds little or nothing to the First Amendment protection of at-will public employees. It also adds nothing to the protection of those public employees who, unlike Churchill, have some form of civil service or "just cause" protection. As to those employees, the government already has the burden to reasonably investigate and evaluate whatever facts it is relying upon before imposing serious discipline, and if challenged to prove its case in a post-termination hearing. Even before termination, "[t]he tenured public employee is entitled to oral or written notice of the charges against him, an explanation of the employer's evidence, and an opportunity to present his side of the story.<122>; This alone is more process than the Waters plurality found adequate for the at- will employee who is fired explicitly on the basis of speech.
I will call this right—the right to notice of the charges and to a hearing before an impartial decisionmaker at which the employer must show just cause for discipline or discharge and the employee may respond—"due process." Under prevailing constitutional doctrine, the substantive requirement of just cause or the like arises from statute or contract, not from the due process clause itself; the entitlement to retain one's job absent specified conditions is the "property interest" that triggers the constitutional requirement of due process.<123>; But due process rights assume the existence of substantive protections, and when I refer to "due process," I include both the procedures and their substantive predicate—the right not to be fired without a good reason.
The tenured civil servant who is fired after having publicly criticized her superiors' performance of their public duties is in an altogether different position than the at-will employee discussed above. She cannot be fired but for reasons prescribed by regulation, or for just cause. She must be given notice of the charges and of the evidence, and an opportunity to present her side of the story before termination; if she is fired she may challenge the employer's claims in an administrative hearing before an impartial decisionmaker. The administrative process for adjudicating just cause discharges is much cheaper and more prompt, and consequently more accessible to most employees, than civil litigation. The employee who believes that she was actually fired because of her public criticism need not prove that the employer was so motivated, although she may seek to do so to rebut the employer's claim of just cause.
Due process rights undergird the free speech rights of public employees much as due process rights in criminal matters undergird the free speech rights of citizens generally. Indeed, I suspect that whatever greater margin of freedom of expression public employees have over private employees is enjoyed only by those protected by some form of civil service regulations or tenure, and rests as much on due process protections as it does on the ability to bring a First Amendment claim.
Waters thus suggests, but does not give much substance to, the idea of a "First Amendment due process" for public employees. That is unfortunate. Given the great difficulty and burden of litigating these cases, due process—requiring the government to give notice and a hearing and to prove a rational basis for discharge—would do more to protect the freedom of speech of many public employees than would a more generous standard of liability in a First Amendment lawsuit.
V. The System of Freedom of Expression in the Private Sector Workplace
Employees in the private sector do not enjoy even the limited constitutional right of freedom of expression at the workplace that public employees have, for their employers' actions are not state action. Yet there is a system of freedom of expression in the workplace. Its content, described above, differs from that in the public sector, but I will argue that its structure and the probable consequences are surprisingly similar: as in the public sector, those employees who can be fired at will cannot be expected to enjoy the freedom of speech the law purports to provide them, while those employees who enjoy "due process rights" can be expected to enjoy much greater free speech rights as well.
A. The Wrongful Discharge Model of Free Speech Protections in the Private Sector At-Will Workplace
I have suggested that we think of the workplace as a community in which management governs the citizens-employees, subject to an amalgam of speech-protective laws that serves as the functional equivalent of a First Amendment doctrine (just as the various employment discrimination laws can be seen as the equivalent of the Equal Protection Clause). This quasi-First Amendment creates no general principle of freedom of speech, subject to certain defined exceptions; this is not a "defining-out" system.<124>; Unlike the First Amendment generally, but like the First Amendment in public employment, the system of freedom of expression in private employment is a "defining-in" system,<125>; in which the employee bears the burden of showing that her speech is within limited protected categories. But speech that is "defined-in" in the private sector consists of a number of particular, discrete exceptions stemming from different sources, rather than a single broad exception, to a general principle of no protection.
But no protection in the private sector is harsher than it is in most of the public sector, for in the great majority of private sector workplaces there is no independent principle of due process. I do not refer simply to the absence of "state action" and the inapplicability of constitutional due process; I refer rather to the absence of any general principle under which the "government"—here, firm management—must prove, by fair procedures, a rational basis for adverse action against an individual. In the private sector, of course, the background rule is quite the opposite: the law generally gives employers plenary power to discharge employees at will, for good reason, no reason, or even bad reason, including expressive activity. The at-will presumption has been dramatically eroded by a growing array of discrete "wrongful discharge" doctrines against discrimination and retaliation of various kinds, including the speech protections at issue here. But each of these protections is superimposed upon and, I will argue, undermined by the background rule of employer discretion.
The typical private sector employee is thus in much the same situation as the sizable minority of public employees who are at will and cannot rely on any general procedural or substantive guarantee of fairness to support their supposed protection against discharge based on protected speech. Without a just cause requirement for discharge, employees have no right to notice of the grounds for discharge, to a hearing before an impartial tribunal, or to proof by the employer of a fair and rational basis for the decision. That void makes employees very vulnerable to employer retaliation or discrimination. A discharge—even one that the employee claims was based on "protected" speech—stands unless and until the employee can prove a hidden unlawful motive by way of drawn-out and often costly procedures.
The at-will rule, which lies at the foundation of employment law, casts the heavy burdens of proof and of delay on the discharged employee. The at-will rule thus continues to exercise a tremendous gravitational pull on the status and security of the employee even in the context of the many exceptions to the at-will rule. In particular, the at-will background to the system of freedom of expression undermines legal protections of speech in the private sector, whether they are enforceable through litigation or administratively.<126>;
Let me take as an example of the latter the NLRA's protection of union organizing. An employee who believes he was fired for union activity may file a charge with the National Labor Relations Board ("Board"), but only the Board may prosecute a complaint.<127>; Given the burden of proof and limited prosecutorial resources, it is often difficult to persuade the agency to pursue these cases without a "smoking gun," and "smoking gun" testimony on behalf of a recently discharged union activist is very hard to come by.<128>; Even when the Board does prosecute, a recalcitrant employer's decision reigns until and unless a court finally orders reinstatement or some other remedy.<129>; A fully litigated discrimination case lasts an average of three years under the NLRA.<130>; Nor is there a pot of gold at the end of this long process. The Board's remedial powers are limited to ordering reinstatement and backpay; the NLRA provides for neither full compensation nor penalties.<131>;
These features of the NLRA—the lack of a private cause of action, the heavy burden of proving employer motive, the economic burden of delay, and the lack of fully compensatory (or deterrent) remedies—are typical of the administrative schemes by which most statutory antiretaliation provisions are enforced. An agency may reject many meritorious but hard-to-prove cases, particularly when enforcement resources are tight.<132>; Because these statutory schemes provide no private cause of action, an agency's refusal to act means the end of the employee's case. Remedies in administrative cases are typically limited to reinstatement and backpay, and are not fully compensatory. Fines, when they are allowed, are generally minimal.<133>;
But there is good reason to believe that many potentially meritorious retaliation claims are never filed. Perhaps because of the lack of a private cause of action, most of these antiretaliation provisions are virtually unknown to employees and even to lawyers. And for most protected employee speech that is unrelated to union activity, there is typically no informed agent, such as a union, on the scene. That difference alone may account for much of the otherwise surprising contrast between the voluminous activity under the NLRA and the virtual dormancy of most other statutory antiretaliation provisions.<134>; Widespread ignorance of the law combines with what is typically a very short statute of limitations to cut off many potential claims before they are recognized.<135>; All this suggests that it is not the rarity of employer reprisals but rather weaknesses in the system of enforcement—some combination of employee ignorance about these provisions, inadequate agency resources, lack of political will, and vagaries of proof—that have rendered many statutory antiretaliation provisions virtually dormant.<136>;
The situation is different but not markedly better with respect to those retaliation claims that are in the hands of private litigants. The employee still bears the difficult burden of proving a retaliatory motive where the employer has created the relevant documents and employs most of the potential witnesses. Further uncertainty results from the limited and indeterminate scope of the common law protections. Moreover, litigation is notoriously slow. In the face of these difficulties, lawyers play a screening role similar to the role of administrative agencies.
The decisions of private attorneys, however, cannot help but be greatly influenced by the prospective plaintiff's wealth: few low- and middle-income employees who are fired from their jobs have the savings or income to pay a lawyer's fee, and, given that backpay is a chief determinant of damages, few have the realistic prospect of a recovery sufficient to attract a lawyer to any case that would require a significant investment of time.<137>; Notwithstanding the popular perception created by some highly publicized large verdicts, recovery in most wrongful discharge lawsuits amounts to the equivalent of half a year's salary.<138>; Punitive damages are sometimes available, and may be necessary in some contexts.<139>; But the few resulting mega-verdicts create a misleading public perception of wrongful discharge remedies, lending what the Dunlop Commission calls a "lottery-like" character to wrongful discharge litigation:<140>; a few plaintiffs hit the jackpot, but the great majority of wrongfully discharged employees get little or no relief.
B. The Rationality of Employer Retaliation and Employee Silence in an At-Will Environment
Let us assume that employers and employees respond rationally to the costs and benefits they face, and let us see where those assumptions take us. Discharges of productive employees, especially discharges that other employees regard as unfair, have costs.<141>; Unfair discharges may bruise employee morale and loyalty; they may lead to lower productivity, greater turnover, or perhaps an interest in unionization; a reputation for unfairly discharging employees may hurt the firm's ability to compete for some of the best prospective employees.<142>; Unfortunately, the economic deterrents to unfair discharges may be much weaker than they are often portrayed.<143>; This is particularly true in the case of discharges based on protected speech.
Consider first that the speech the law protects often harms the employer for the very reason for which it is protected: It brings information to the public or spreads ideas among the workforce that may threaten the employer's chosen way of doing business. So the employer often stands to gain a great deal from the discharge of an employee speaker. For example, the discharge of a union activist will remove one disruptive voice (for that is, at best, how most nonunion employers view union organizing<144>;) and will very likely deter others. Similarly, the discharge of an employee who has complained internally of unsafe working conditions may neutralize her complaints, avert a potentially expensive OSHA investigation, and deter other employees from making similar complaints. Unfortunately, the incremental benefit of a single speech-based discharge—suppressing one speaker and deterring others from speaking in ways that harm the organization—will often outweigh the costs of that discharge.<145>;
Unfortunately, the unwieldy procedures and limited remedies for enforcing those protections mean that a retaliatory discharge may be cost-effective even after taking into account the prospect of a legal challenge. Speech protections that are enforced administratively present an almost negligible risk, given the unlikelihood of a prosecution and the extremely limited remedies. Even the prospect of civil litigation may not be an adequate deterrent: The hurdles facing potential plaintiffs—the narrow and uncertain scope of wrongful discharge protections, the difficulty of proving motive, the limited recovery in most cases, the delay, and the array of economic and noneconomic costs of both the discharge and the litigation to the employee—should lead employers to steeply discount the expected cost of a wrongful discharge.<146>;
Moreover, to the extent that employers can predict that some employees are more likely than others to litigate—the most highly paid, the most assertive, or the most litigious—they may discount the risk selectively. Those employees who are most vulnerable and most in need of the protections of unionization and health and safety laws, for example—unskilled and semiskilled workers in production, maintenance, and service jobs—are least likely to litigate. The cost of discharging these workers is further reduced to the extent that employee turnover is high and the employer counts on little loyalty, morale, or investment in and sharing of skills on the part of employees. In those cases especially, it may be rational for the firm to fire an employee who speaks against the interests of firm management.<147>;
Not all employers will act "rationally" in disregard of the law. To various degrees and in different ways, many employers feel constrained to obey the law.<148>; But others will be sorely tempted to violate the law by firing a disruptive speaker—particularly if the speaker is not a model employee—where the risks and consequences of detection are very low. Many employers will occupy a comparatively safe middle ground: they will simply communicate, explicitly or otherwise, that certain topics are off limits, letting employees translate this if they choose into a threat of retaliation. Indeed, the veiled threat of retaliatory discharge may be the most cost-effective approach for employers, for it may yield most of the benefits of a discharge—suppression and deterrence of speech—with few of the costs.
The problem is not simply that wrongful discharges go unremedied. I am ultimately concerned less with the potential wrongful discharge claims that are not vindicated than with the potential speech that is not spoken. If we assume that employees are generally aware of what it takes to redress a wrongful discharge, and that they tailor their conduct accordingly,<149>; then we can comfortably predict that all but the most intrepid employees will be deterred, or "chilled," from speaking out in ways that might provoke the employer's displeasure.<150>; Keeping in mind that much of this speech is a "public good" and carries no tangible benefit for the speaker, many employees are likely to be "chilled" from speaking out by even a slight and unspoken threat of discharge. The resulting silence of employees about workplace issues and corporate conduct impoverishes public discourse, hinders the enforcement of workplace and economic regulations, and undermines the formation and reform of institutions for employee involvement in workplace decisionmaking. If I am right about the "unfreedom" of employee speech, the workplace may be less a school for democracy than a training ground for apathy and disengagement.
Of course, some people do speak out. There are whistleblowers who defy fears of retaliation and union activists who brave the often open threat of reprisal. But what does it mean for workplace discourse if the only people who speak out against the employer are those who fail or refuse to succumb to rational fears and economic pressures? The few who do speak out may appear to be (and may sometimes be) idiosyncratic, odd, or a little nutty; sometimes the very qualities that enable them to overcome the fears that silence most employees may isolate them and make them both less appealing and less effective in their articulation of potentially shared concerns. As many union activists have learned, their speech itself may heighten the anxiety and even provoke the hostility of the more typical, less brave employee. The current system of freedom of expression in the at-will workplace—the existing patchwork of speech protections against a background of employer power and discretion—gives a false promise of protection that few employees may be willing to test. It fosters an atmosphere in which workplace debate, criticism of current practices, and employee monitoring of workplace conditions are not normal and ongoing, but exceptional, dramatic, and seemingly irrational.
C. Two Models of Due Process in the Private Sector Workplace
For the vast majority of private sector employees who are employed at will, the wrongful discharge system offers the only hope of a remedy. But due process, as I have defined it, is not wholly absent from the private sector workplace. I will turn next to two very different systems of "due process" in the private sector workplace and discuss how they affect the protection of valuable employee speech.
1. "Industrial Due Process" and the Protection of Employee Speech
Virtually all employees represented by a union are protected by a just cause requirement for discipline and discharge, together with a grievance-arbitration system for enforcing it. The just cause clause has generated a body of arbitral law known as "industrial due process.<151>; As the unionized sector shrinks to near ten percent of the private sector workforce, this system may seem to be of increasingly marginal importance.<152>; But it remains important at least as a model of how "due process" could work in the private sector to protect employee free speech.
Under a just cause provision, the employer must have a valid basis for discipline or discharge. If an employee "grieves" the decision, the union may take the grievance through several steps of discussion with successively higher management officials, and eventually to arbitration. The arbitrator is an impartial third party with experience in labor relations, and often in the particular collective bargaining relationship, who is chosen jointly by the employer and the union, and whose future employment as an arbitrator depends on a reputation for impartiality and evenhandedness. The hearing, at which the employer bears the burden of proof,<153>; resembles a trial, but is less formal. An attorney or other representative for each side presents arguments, documents, and witnesses, but briefing is limited or nonexistent. The arbitrator is typically required to render a decision and some short explanation within a few weeks. The standard remedy for a discharge without "just cause" is reinstatement and backpay. The entire process, from the employer's discharge decision to the arbitrator's ruling, may take as little as two to four months.<154>;
How does this process protect employee free speech? Industrial due process typically affords no direct, explicit protection of speech at all. Arbitrators are bound to apply the contract; this generally includes principles of industrial due process that have developed out of the just cause requirement, but there is considerable dispute as to whether it includes external law such as protections of whistleblowers and the like.<155>; But it is rare, in any event, that an employer will expressly base a discharge on speech that is protected by external law.<156>; More typically, the employer denies that the discharge was based on speech and claims some other cause. In that case, the issue is the purely factual one of whether the employer's reason is just cause and is proven by the evidence. Simply putting the employer to its proof will protect employee speech indirectly by making it more difficult to discharge a workplace critic on pretextual grounds.
In addition, on the purely procedural front the employee may challenge the adequacy of the employer's pre-termination process, including the adequacy of notice, compliance with "progressive discipline" in the form of warnings and suspensions prior to discharge, and equitable treatment vis-à-vis other offenders. These procedural requirements are supplemented by the grievance process itself, the first steps of which give the union employee an opportunity to probe the employer's case and often to settle the dispute prior to arbitration. All of this process enforces norms of fair and equal treatment that make it hard for the employer to seize upon minor infractions or lags in performance to discharge employee dissenters.
The employee may also challenge the substantive justification for discipline: she may contest the reasonableness of the employer's rules, the job-relatedness of alleged misconduct, and the severity of the penalty.<157>; Through these substantive limitations on discharge, the just cause requirement inevitably generates some protection for "low level" freedom of speech: speech that is not protected by external law, but that is also not harmful enough to justify discharge. Most complaints about workplace conditions, criticism of supervisors or management or aspects of the workplace regime, and workplace arguments will be "protected" simply because they do not violate reasonable workplace rules of which the employee had notice, or are not serious enough to justify the penalty.<158>; Appreciation for the value of free expression can readily creep into the determination of just cause, for example, in deciding whether criticizing or challenging a supervisor is "insubordination," or whether heated discussions, and even some rude or offensive language, are part of the "shop talk" that is acceptable under past practice in the workplace.<159>;
The employee protected by a just cause clause, like the tenured civil servant, has a kind of property right in continued employment, the termination of which must be justified by good reasons and preceded by fair procedures. If she has the support of her union, the protection of a just cause clause, and a union that believes her discharge was based on legally protected speech, she is in much the same position as the tenured civil servant with a potential First Amendment claim: the employee may challenge her discharge and demand a prompt hearing before an impartial arbitrator, at which the employer must show just cause. The delay and the cost of the proceeding is minimal, and the burdens of proof and uncertainty fall on the employer.
Industrial due process should deter some retaliatory discharges, and it should offer a more accessible remedy for claims of retaliation. But it should also enhance employee freedom of speech at the most basic level by giving employees a measure of breathing room and confidence that they cannot easily be fired or penalized on manufactured grounds. In large part because of these protections, union employees with a well- functioning grievance apparatus and an effective union should experience much more freedom of expression—including in areas in which all employees are supposedly protected—than do nonunion private sector employees.
There is one major functional difference between constitutional due process and industrial due process that has important implications for the protection of speech: unlike constitutional due process, industrial due process is a collective right. Under a collective bargaining agreement, the union decides whether to carry a grievance through arbitration.<160>; This makes the employee's position especially vulnerable if her speech is critical of or simply contrary to union policies or procedures. <161>; The union is subject to a duty to fairly represent employees, but that does not guarantee every employee a hearing.<162>; Moreover, the duty of fair representation is enforceable only through litigation; an employee who is forced to file suit in order to get a hearing on her discharge carries all the burdens that, I have argued, undermine the practical effectiveness of wrongful discharge remedies.
On the other hand, the presence of an active, effective collective bargaining representative bolsters the employee's position in a number of ways. The union is a repeat player standing in a position of rough equality with the employer in selecting an arbitrator; the arbitrator's future livelihood depends on dealing fairly with both sides. The union also provides representation, either by a union official experienced in handling arbitrations or by an attorney. Moreover, if the arbitrator reinstates the individual, the union remains on the scene to help protect the employee's position.
The union's presence in the workplace also supports employees' freedom of expression in ways that go beyond just cause protection. The presence of a strong institutional voice not only backs up employees' freedom to discuss their shared concerns, but gives them tools for redress and reform in the workplace. Unionization, when it works well, brings several elements of democratization to the workplace beyond the elements of due process and free expression. But given the steady decline of union membership over the last few decades, I am loath to conclude that unionization is the necessary precondition for either effective due process or genuine freedom of expression. I therefore turn briefly to another form of due process in the private sector.
2. The Growth and Limitations of "Corporate Due Process"
The expansion of wrongful discharge liability has prompted a proliferation of internal grievance procedures established unilaterally by firm management since the mid- 1960's.<163>; Such systems are most prevalent among large firms and firms with extensive dealings with government; they can be traced in large part to the enactment of substantive constraints on the power to discipline and discharge, particularly the antidiscrimination mandates beginning with the 1964 Civil Rights Act, but also the antiretaliation mandates that I have discussed here.<164>; It appears that many private employers, faced with the threat of wrongful discharge liability, are willing to supply some form of due process. Might this be an answer to the problem of free speech and due process in the workplace?
If due process is conceived of as a kind of precaution against wrongful discharge, it may also be a sensible precaution for employers to take on their own. As noted above, unjustified discharges have costs, including legal consequences. We would expect those costs to lead employers to institute procedures for monitoring and reviewing discharge and disciplinary decisions. The basic economic theory of tort law would lead us to expect employers to add precautions up to the point that additional precautions would cost more than would the discharges those precautions might avoid.
Internal due process mechanisms are well suited to avoiding discharges that, on balance, do not serve the firm's interests. Under these systems, decisionmaking is more centralized and more elaborate, providing some mechanism for employees to challenge, within the firm, what they regard as unfair decisions and for employers to avoid potentially costly mistakes. But "mistakes" in this context are discharges, the expected cost of which exceeds their expected benefits. Unfortunately, many wrongful discharges based on protected speech probably cost less, even considering the prospect of liability, than they gain for the firm.<165>; The shape of internal due process systems reflects this limitation.
Most companies' grievance procedures consist of either an informal or a formal "open door" policy: a policy allowing an aggrieved employee to approach anyone in management regarding her complaint.<166>; Relatively few private companies have more elaborate procedures, including the right to a hearing of some kind.<167>; Even those firms with more elaborate forms of due process almost invariably retain ultimate decisionmaking power within the company; the final decision in almost all corporate due process systems, unlike both constitutional and industrial due process systems, lies with upper management.<168>; The threat of wrongful discharge liability has not led companies to submit nonunion personnel disputes to an independent arbitrator as an added precaution against mistaken discharges.<169>;
The continued resistance to outside arbitration may be changing in the wake of the Supreme Court's decision in Gilmer v. Interstate/Johnson Lane Corp.,<170>; which upheld an employee's pre-hire agreement to submit employment disputes to final and binding arbitration, and foreclosed litigation of an employee's claim under the Age Discrimination in Employment Act.<171>; A broad interpretation of Gilmer would make arbitration not simply an additional precaution against mistaken wrongful discharges, but a shield against liability. It seems likely that many employers will be willing to give up final control over discharge decisions in exchange for freedom from most wrongful discharge litigation.<172>; Whether that is a fair exchange, and one that will vindicate the policies underlying the various wrongful discharge doctrines, is another matter with which I will wrestle below. But Gilmer notwithstanding, few corporate due process systems currently provide for review of discharge decisions by an impartial outside decisionmaker.
Nor has the threat of liability (nor any other incentive) led management in most cases to make substantive promises of fair treatment. On the contrary, employers increase their exposure to one kind of liability—liability for breach of an implied contract—by explicitly limiting their right to fire at will in employee handbooks and the like.<173>; Along with the growth of internal corporate grievance procedures for challenging discipline and discharge, we have seen a flight from substance. The trend is toward adding to employee handbooks express language affirming the right to discharge at will, and toward removing language that appears to limit the permissible grounds for discharge.<174>; A typical corporate due process scheme thus offers nothing like the "just cause" requirement that, in my view, is a necessary element of due process.
Employers seek to minimize liability in a variety of ways, many of which do nothing to protect employees against unjustified discharge. Employers may not simply take more precautions against injury; they may also reduce their "activity level"; that is, they may employ fewer individuals. A RAND Corporation study found that the expansion of wrongful discharge liability—particularly tort liability—was associated with a reduction in employment levels that is quite staggering in relation to the comparatively minimal direct costs of litigation.<175>; Even if these findings are exaggerated, it is clear that employees do not gain security against wrongful discharge in direct proportion to the increase in wrongful discharge liability.
This evidence suggests that employers are acutely sensitive—even oversensitive—to the threat of wrongful discharge liability.<176>; But neither the desire to avoid liability nor the desire to remain nonunion while attracting and retaining skilled and productive employees is pushing employers toward due process as I have defined it. Corporate due process mechanisms are proliferating, but their prevailing forms are shorn of two crucial features: impartial resolution of grievances and substantive guarantees that discipline will be just and fair.
We can assume that these systems deliver something that employees value, for if they did not, they would seem to accomplish few of management's own objectives. But internal grievance systems are unlikely to provide a meaningful check against retaliatory measures. Rather, these systems are more likely to curb overreaching or hasty actions by lower level supervisors, provide for more even-handedness in discipline, and implement more routine systems of employee evaluation, and even forestall some justified discharges. These results may give some limited "breathing room" for employee speech, such as criticism of supervisors and the like. But purely internal corporate due process seems unlikely effectively to support employee freedom of speech in areas in which management feels threatened.
I have sought to explain a gap between law and reality, a gap between the freedom of expression that the law purports to afford employees and the freedom that employees seem to experience, based on the incidence and apprehension of retaliation and the willingness to speak out. I have argued that much of the gap results from an underlying presumption of at-will employment that surrounds and constrains each of its exceptions, including the speech protections at issue here. I have also contended that actual freedom of speech in both the public and private sector workplace depends to a great extent on the presence of independent due process protections in the form of procedural and substantive guarantees that discharges and serious discipline are for just cause.
Unfortunately, I know of no studies examining the relationship between employee freedom of expression and job security or due process rights. We are left to draw inferences from the spotty evidence discussed above. The massive study of federal employees discussed above<177>; involved the civil service only; given the lack of a comparison group of at-will employees, it can tell little about the significance of due process rights. Still, the study injects a strong cautionary note, for it suggests a fairly high level of managerial hostility and employee "silence" (failure to report perceived wrongdoing) even in the most legally protected environment. Even with the civil service laws and due process rights backing up the explicit protection of federal whistleblowers, many employees experience subtle forms of retaliation that are not easy to detect or challenge, particularly through a due process regime. At least partly as a result, many who observe wrongdoing choose to remain silent.
These data discourage any expectation that due process will insure complete freedom of expression. But my argument is a comparative one: I contend that employees protected by due process will enjoy much greater freedom of expression than purely at-will employees. Unfortunately, that claim is fully consistent with the rather discouraging portrayal of free speech in the federal sector. It may be that more severe reprisals are threatened, feared, or carried out, and that the silence resounds all the more loudly in the absence of due process protections.
Some very indirect evidence to this effect may be found in the data on union talk in the pervasively at-will environment of the nonunion private sector. Recall that nearly eighty percent believed that employers will fire employees who try to organize a union, and over forty percent of nonunion employees believed that they personally would be fired for engaging in union activity.<178>; At least as to union talk, many more employees expect much harsher forms of reprisal than was reported in the federal civil service. The comparison is rough but suggestive: perhaps the difference is partly due to strong due process protections in the latter sector, and their complete absence in the former. On the other hand, there is another obvious difference between the two sectors that probably accounts for at least as much of the apparent difference in levels of retaliation: Market forces in the private sector may drastically increase the incentives to suppress speech that is believed to be economically threatening.
Another suggestive piece of evidence is the higher reporting of OSHA violations and the safer conditions present in union as compared to nonunion workplaces.<179>; This evidence offers some confirmation of my strong impression that union employees, who are protected by a just cause clause and who have access to a grievance process, feel much more free than nonunion employees to discuss workplace issues with their coworkers and bring grievances to management. Unfortunately, other differences between nonunion and union shops again confound the picture. It would appear nearly impossible to isolate the influence of industrial due process—as opposed to the other advantages of having an informed agent for the collective body of workers and a collective bargaining relationship with the employer—in fostering greater freedom of speech.
There is currently little empirical evidence for my claim about the fundamental significance of due process in the system of freedom of expression in the workplace. On the other hand, I do not believe that the propositions are counterintuitive, or that they require big inferential leaps, for they follow quite reasonably from an assumption that employee behavior is basically rational. Speaking out on issues of public and collective concern promises little direct benefit to the employee and poses a significant risk, or at least a perceived risk, of direct harm in the form of employer retaliation. Under the existing "wrongful discharge model" of speech protections, such retaliation, even when it is unlawful, may be very hard to redress. Under a due process model, speaking out is less costly because redressing employer retaliation is easier, faster, and cheaper. Recognizing that the proof is incomplete, I hope to have laid the groundwork for more research and analysis of the contours of the existing and the best system of freedom of expression in the workplace.
VI. Securing Free Speech Through Due Process in the Workplace: Some Preliminary Observations on Implementation
I have argued for the recognition of a basic principle of fair treatment and due process to provide a foundation for the system of freedom of expression in the workplace. The proposal is not narrowly tailored to meet the free speech problem I have identified; on the contrary, I have argued that the existing narrowly tailored approaches to protecting valuable speech—which take the form of discrete statutory and common law remedies against particular kinds of retaliatory discharge—have failed to overcome the shadow cast by the basic at-will presumption.
The implications of my analysis obviously reach far beyond the protection of freedom of expression, into parallel issues of "equal protection" in the workplace, and into the ideology and economics of managerial prerogatives that form the background for all workplace regulations and employee rights. So I do not pretend to have offered a complete case for the abandonment of the at-will rule and its replacement with a regime of just cause and due process. Rather, I have sought to inject into the broader debate over at-will and just cause regimes a recognition of some of the less visible costs of the existing at-will rule. Among the significant costs of at-will employment is the undermining of existing legal protections of employee speech and the resultant unfreedom of workplace discourse. But in this final Part, I will offer some preliminary thoughts on how to implement universal due process in the public sector (in Part VI. A) and in the private sector (in Part VI. B), with an eye toward the present objective of supporting freedom of expression.<180>;
Universal due process for discharge is not hard to envision in the public sector; it was indeed envisioned and rejected over twenty years ago in the early public employee due process cases. But the logic is sound: The government is subject to the substantive constitutional constraints of the First Amendment and the Equal Protection Clause, as well as the minimal obligation to act reasonably—not "arbitrarily and capriciously"—in its dealings with citizens. When the government is dealing with its own employees, weighty interests are affected, for one's job is the foundation of economic, psychic, social, and moral well-being, and discharge from employment has particularly devastating consequences. So, in the context of termination of employment, the inherent constitutional constraints on government action should translate into an entitlement protected by the due process clause: All public employees should be entitled, before they are finally discharged, to a hearing before an impartial decisionmaker at which the government must demonstrate a constitutionally valid reason for discharge: a reason that is not arbitrary and is not otherwise unconstitutional.<181>;
Under this broad view of due process—admittedly a blast from the Warren Court past—due process would serve as a bulwark for other substantive constitutional protections, including the First Amendment. Speech protected by law, such as speech on matters of public concern that is not unduly disruptive, could not constitute a valid reason for discharge.<182>; Simply requiring the government to prove a nonarbitrary and valid reason for discharge would give added protection to speech of public concern and would effectively protect some speech that, while not deemed important enough to merit First Amendment protection, is not so disruptive as to constitute a legitimate basis for discharge. This would create a buffer zone around the critical core of speech on matters of public concern, and allow some breathing room for employee speech.<183>;
What effect would universal due process have on the level of employee speech, the number of wrongful discharges, or the volume of litigation? These are all empirical questions on which I can currently offer only informed speculation. If I am right about the value of due process, it should encourage more employee speech critical of or disagreeable to public employers. As the federal whistleblower data demonstrate, universal due process would not produce complete freedom of expression, but it would surely give added security to otherwise at-will employees against the most powerful forms of retaliation for protected speech.<184>; By the same token, due process safeguards should reduce public employers' resort to retaliation. The need to justify discharge would make such retaliation easier to uncover and harder to carry out than it currently is for employers.
What about the volume of litigation? Less retaliation should result in less litigation. But we might see less litigation even without optimistic assumptions about less retaliation. Due process allows employees who are fired, and who challenge their discharges, to peek into the black box of the personnel process and learn something about the employer's justifications for these decisions. This gives them some free discovery if they decide to litigate and in any event educates them about the employer's defense. Some employees who would otherwise not litigate might be encouraged to do so by information they discover in the administrative proceedings. But some who win at the hearing will be satisfied with the outcome and forgo the additional cost, stress, and risks of litigation. Some who lose at the hearing will conclude that their case is weaker than they had thought. It is possible, though hardly certain, that the net effect of a due process regime would be less First Amendment litigation.<185>; The litigation that does occur even after a due process hearing should, in any event, be based on a more educated assessment of the facts.
A more modest alternative is also possible: employers might be required to give all public employees at least the minimal pre-termination process called for by Loudermill:"[o]ral or written notice of the charges against him, an explanation of the employer's evidence, and an opportunity to present his side of the story" before termination.<186>; Employees who are then in fact discharged, and who claim that their discharge is actually based on protected speech (or some other illegal ground), would be entitled to a full post-termination process of the sort available to tenured civil servants. In the absence of a substantive just cause requirement, the issue at the hearing would be simply whether the discharge resulted from protected speech; the burden of proof would logically rest with the employee.
Such a hybrid procedure would be less costly and would offer some support to free speech rights by providing an inexpensive and relatively prompt hearing on the retaliation claim.<187>; But it would leave in place the hurdle of proving motive, and it would not confer the indirect protection of speech—the breathing room—that just cause or its equivalent would. It would also yield some of the same anomalous results as the Waters plurality: If the requisite predetermination inquiry or the later hearing disclosed that the employee "had been complaining about the perennial end-of-season slump of the Chicago Cubs, her dismissal, erroneous as it was, would have been perfectly OK.<188>;
Just as in the case of the plurality's approach in Waters, these complications and anomalies can be traced to the attempt to preserve a sphere of at-will employment—of virtually unfettered employer discretion—in the public sector. I am inclined to view that sphere as an anachronistic holdover from the days of the "rights/privileges distinction" that long kept the Constitution out of the sphere of government employment and benefits. It is hard to reconcile with, and in fact undermines, the full range of substantive constitutional and statutory protections that have arisen in the last thirty years.
The proposed universalization of due process would do more to protect the speech rights of now-at-will public employees than would any of the opinions in Waters. But it would do so by reaching far beyond the relatively few cases in which public employees claim they are fired for protected speech. Securing the free speech rights of public employees may not be a sufficient justification for extending due process rights to all. But when we assess the costs and the benefits of such a regime—both of which go beyond the impact on employee speech—we should attempt to take into account the significant cost of silence that I believe attends the present regime and the largely unquantifiable benefits of supporting the freedom of public employees to speak out on matters of concern to them and the public.
B. Universal Just Cause and Free Speech in the Private Sector: Lessons from Industrial and Corporate Due Process
The analogue to these proposals in the private sector—statutory just cause protection—has been debated among labor law scholars for many years.<189>; Admittedly, employee speech is involved in only a small fraction of the discharges that would be affected by a shift to just cause. Still, I maintain that among the virtues of just cause protection is the important and largely unrecognized benefit of expanding real employee freedom of speech, both that speech which currently enjoys formal protection and some speech that currently does not.<190>;
Innumerable questions of forum and procedure would arise under such a proposal: Should it be a state or federal forum?<191>; What kind of tribunal: arbitral or administrative? What time limits for a hearing and a decision? How much discovery? While these and other questions must be answered before a proposal can be implemented, I choose to put these questions largely beyond the scope of this Article.<192>; But a few observations are in order.
The union experience demonstrates that it is both important and possible to design a process that is expeditious and inexpensive (compared to litigation), with adequate but not extravagant remedies, and with fair and impartial decisionmakers. In two respects, however, the system proposed here would differ from industrial due process. First, any system of universal due process should entertain not only the just cause issue but any argument that the discharge is in fact for reasons that are unlawful; external law should thus be fully in force within the due process proceedings. Moreover, the individual should have full control over her case; no intermediary could decide not to seek a hearing.
But the union experience raises an important concern on the other side. Unions, as potentially powerful collective institutions in the workplace, bring far more than the industrial equivalent of due process to the employees they represent. Would a universal mandate of just cause support employee free speech in the absence of institutions that perform at least some of the functions that unions perform? The short answer is that just cause and an individual right to a hearing should accomplish no less in the private sector than it does in the public sector. But this answer may be too short. It would not be surprising to discover greater depths of resistance in the private sector to the demise of the right to fire employees at will and to an order to reinstate an unjustly fired employee.<193>;
A "works council" within the workplace, or even an employee membership organization operating outside the workplace, could perform some of the functions of a union, such as monitoring the performance of decisionmakers, providing experienced representation, and perhaps helping to secure the notoriously vulnerable status of the rei