The Yale Law Journal


“There Is No Such Thing as an Illegal Strike”: Reconceptualizing the Strike in Law and Political Economy

06 Jan 2021

Diana S. Reddy

abstract. In theorizing the lack of strikes in recent decades, legal scholars have focused on the rules that make striking difficult, even counterproductive, for workers. But while American law on striking is ambivalent at best, the impact of legal restrictions on labor protest has always been mediated by broader political economy. One hundred years ago, as the critique to Gilded Age inequality raged, strikes by emergent labor unions made the stakes of the “labor question” apparent. These strikes were subject to legal repression, and in the shorter term, were often seen as failures. In the longer term, though, they laid the foundation for the New Deal order and its modest compression of economic inequality.

That order—built in part on a depoliticization of unions’ role in the workplace—proved unable to withstand the neoliberal turn. Yet, in today’s new Gilded Age, workers are tentatively rediscovering the power of the strike, bad law notwithstanding. Labor activists have long proclaimed, “There is no such thing as an illegal strike, only an unsuccessful one,” inviting the question of how strikes can be successful today, in spite of an inhospitable legal regime. Strikes are not just “economic weapons”; they are political protest. And the success of strikes today should be measured not only by their immediate economic wins, but also by their longer-term sociopolitical ones—to the extent they help upend the dominant regulatory and discursive impulses of recent decades, and like the strikes of the Progressive Era, legitimate a new vision of law and political economy.


In the wake of the 2020 killing of George Floyd by a Minneapolis police officer—in the midst of a pandemic that had already resulted in 100,000 deaths1 and prompted forty million Americans to apply for unemployment2— protests erupted in Minneapolis and then nationwide. As protestors exercised their constitutional rights in the streets, union bus drivers sought to protest on the job. In an online petition, bus drivers proclaimed, “We are willing to do what we can to ensure our labor is not used to help the Minneapolis Police Department shut down calls for justice.”3 Drivers refused to transport arrested protestors to jail.4

The Amalgamated Transit Union, the international affiliate of the workers’ local union, quickly weighed in with a statement of support. The union’s president spoke out against racism in policing, but he also framed the employees’ refusal to work more narrowly. He emphasized worker safety rather than solidarity with a broader struggle: “[O]ur members—bus drivers—have the right to refuse work they consider dangerous or unsafe during the pandemic.”5 While arguably lacking the moral force of the drivers’ statement, the president’s frame was legally important. In seeking to protest not just as polity members, but as workers, the bus drivers had entered the complex world of labor law and its distinct logic regarding when, why, and how protest is sanctioned.6 When it comes to protest in the public sphere, the First Amendment protects political expression above all.7 But strikes occur in the context of ongoing economic relationships; they implicate questions of property and contract, not just liberty.8 As such, when it comes to strikes, legal protections are far narrower, prioritizing the “economic” over the “political.”9

Strikes were once relatively common in the United States, as a tactic through which working people sought to increase their power in the employment relationship, and periodically, as a fulcrum of massive social change.10 Strikes were a causal factor in the enactment of American labor law, and in helping it withstand constitutional
11 Over the past several decades, however, strike activity has dwindled down to the lowest level in 150 years.12 In theorizing this decline, legal scholars have focused on the rules that make striking difficult, even counterproductive, for workers. But while American law on striking is ambivalent at best, the impact of these rules on labor protest has always been mediated by political economy—the policies, social practices, and beliefs that constitute the arena of social life we refer to as the “economy.”13

In this Essay, I present a brief history of the strike within American law and political economy, through four eras of transition and contingency—the Gilded Age-into-Progressive Era, the New Deal, the neoliberal turn, and today. Given the shorter form of the piece, I paint with broad strokes, sometimes sacrificing granularity for the sake of clarity. At the end of the nineteenth century, strikes—borne of the new circumstances of industrial wage labor—politicized the employment relationship; they rendered private employment an arena of public concern, appropriate for state regulation. Notwithstanding legal impediments and many short-term failures, these strikes enabled New Deal labor law and its limited but meaningful redistributive effects. The contours of that regime rendered it particularly vulnerable to changing political economy, however. When the neoliberal turn undermined the economic case for unions, labor found itself restricted not just by an increasingly archaic legal regime, but by an underspecified normative claim. Even as inequality increased and job quality decreased, worker protest dried up.

Today, as a new Gilded Age gives rise to a new Progressive Era, the ossified remains of New Deal labor law discourage striking in a host of ways. Yet, in 2018 and 2019, strikes were suddenly on the rise. Labor activists have long proclaimed, “There is no such thing as an illegal strike, only an unsuccessful one,” inviting the question: “Can strikes be successful, in spite of an inhospitable legal regime?” Strikes are not just “economic weapons”; they are political protest. And the success of strikes today is about more than their immediate economic wins; their success must be in their longer-term socio-political ones—to the extent they help upend the dominant regulatory and discursive impulses of recent decades, and like the strikes of the Progressive Era, legitimate a new vision of law and political economy.

In Part I, I focus on the construction of the Progressive Era “labor question,” and the divergent responses of Progressive reformers and labor unions. In Part II, I reflect on the jurisprudence of the strike, focusing on current law’s deep ambivalence about labor power, manifest in the socio-legal construction of strikes as hard bargaining, rather than contentious politics. In Part III, I discuss whether and how worker protest today may have the capacity to build a new, and more enduring, social compact. My argument is this: the legal construction of strikes as “economic weapons” has obfuscated their role as normative claims-making, as inherently political. And while somber assessments of the ways in which current law deters striking are not inaccurate, legal rules are not the sole determinant of labor’s power, or legitimacy. In this turbulent moment, strikes can be a tipping point for political economic change, bad law notwithstanding. I conclude with some brief thoughts about the role of law in social change.

I. labor’s progressive era challenge to laissez faire: strikes as protest

Legal scholars who study organized labor have little choice but to be historians. Labor’s legal regime is old, outdated, and so incompatible with both the modern organization of work and most people’s common-sense beliefs about it, that explaining it requires reconstruction of a world which no longer exists. We have often limited our historical horizons to the New Deal and its unraveling.14 Yet, the New Deal’s specific regulatory apparatus was built upon the period of socioeconomic change which proceeded it—the Gilded Age-turned-Progressive Era.15

This moment is an opportune one to investigate the longer-term origins of present predicaments. Over the past decade, a host of scholars and commentators have claimed that we live in a new Gilded Age.16 While historical metaphors are necessarily imperfect, there are parallels: rapid technological change; economic growth; inequality and instability; a populist challenge to the political establishment; even a pandemic.17 As the protests of 2020 now augur the rise of a new Progressive Era, there is much to be learned about how the popular activism of one hundred years ago came to shape the institutional order that followed—for all of its successes and failures.

In this Part, I focus on the role of labor protest in the Gilded Age and Progressive Era as a response to the evolving institution of waged labor. Protest made the stakes of what came to be called the “labor question” apparent, rendering the employment relationship political. Yet, fearful of a state that had proven hostile to unions, leading Progressive Era labor leaders rejected a political vision of unions and the strike, in turn.

A. The “Labor Question”

Is labor getting its due?” neoclassical economist John Bates Clark asked in 1912.18 “[H]ow are the men and women who do the daily labour of the world to obtain progressive improvement in the conditions of their labour reflected President Wilson in 1919.19 Amidst the rapid social change of the late 1800s and early 1900s, these questions—collectively understood as the “labor question”—occupied a central space in theorizing how industrialization and incorporation were reshaping American life.20

To be clear, this was not just a question of economic policy; it was “the constitutive moral, political, and social dilemma of the new industrial order.”21 Historian Rosanne Currarino describes the labor question of the late nineteenth century as whether “democracy [could] survive in industrial America.”22 The fact that it is hard to imagine a “labor question” as implicating the fate of democracy shows how much is not the same today. While there is currently greater concern about the structure of work and its rewards than in decades past, these concerns are not generally thought of as threats to democracy.23 A century ago, however, the linkage was more palpable.

Prior to rapid industrialization in the 1860s, the northern economy was based primarily on small-scale farming and artisanal production.24 In the course of a generation or so, this mode of production was displaced by the growth of large corporations, and with them, waged labor.25 This transition to waged labor in the North, as formerly-enslaved African American workers experienced new forms of economic subjugation in the post-Civil War South, challenged confidence in the promises of “free labor.” It did more than create new economic inequalities. It also embedded inequalities of control, independence, and self-determination into economic structures.26 Small-r republican-minded thinkers worried that these structural inequities harkened back to the European aristocracies many had left behind.27 The structure of work was accordingly very much a question of democracy.

B. The “Progressive” Answer

The Progressive Era involved a diverse set of responses to the problems of the Gilded Age. Its hallmark, however, was a new philosophy of governance. For Progressives, the solution to the labor question was a proactive state; reformers during this period were the first to harness state regulatory authority to benefit workers and consumers. While this instinct may seem common-sensical now, it involved a fundamental rethinking of the logic of state power.

Prior to the Gilded Age, government power—as the most visible form of concentrated resources and coercive authority—was largely conceptualized as something the American people needed protection from. But in the late 1800s, the proliferation of the corporate form, along with its use by industrialists to accumulate the massive surpluses of industrialization, created new private centers of power.28 In response, Progressive reformers began to imagine a new role for government. Rather than being a threat to individual liberty, government could be a protector of it, by undertaking a positive role in regulating corporate power and promoting the interests of workers and consumers.29

Recasting state regulation as in the general welfare was difficult work.30 A host of legitimating narratives made the inequalities of the Gilded Age appear just: Spencerian survival of the fittest and Horatio Alger stories in public discourse;31 and laissez faire and liberty of contract in the legal realm.32 Legal historian Barbara Fried’s influential recounting of what she calls the “first law and economics movement” details the ways in which legal scholars at the time gradually undermined the laissez-faire view that “liberty in economic affairs” should be beyond government authority.33 Progressive Era legal scholars emphasized the potential for coercion and unfreedom in the private sphere. “We live . . . under two governments,” Robert Hale argued, “‘economic’ and ‘political,’ the second public and hence visible, the first private and hence invisible.”34 This reframing changed the terms of the debate about government regulation of the workplace. It meant that “when the government intervened in private market relations to curb the use of certain private bargaining power, it did not inject coercion for the first time into those relations. It merely changed the relative distribution of coercive power.”35

While some Progressive Era attempts to use the state to resolve the labor problem fell to Lochner Era judicial review, reformers did not give up.36 Child labor laws, workers’ compensation systems, antitrust laws, and consumer regulations were all Progressive Era legal innovations.37 Perhaps more importantly, Progressive Era law and political economy set the stage for imagining, and ultimately sustaining, the New Deal.

C. Illegal Strikes

These ideological and legal changes did not take place in a vacuum. If America was born of protest, so too was the modern American state. As historian David Huyssen recently explained, “Although middle-class philanthropists and technocratic politicians gave voice to policies that began to curtail inequality, they did not generate the conditions that made such policies either politically possible or effective.”38 That, he argues, “took decades of . . . organizing from working people—in labor unions, youth groups, radical political parties, and coalitions of mass protest—from the 1870s through the 1940s.”39

Labor unrest was massive during the Gilded Age and Progressive Era. Prior to 1877, most labor conflicts were small and localized.40 The large-scale conflicts of the late 1800s were unprecedented, and “the new American state remained unprepared for [what was to come,] the great upheaval of 1885-86, the 1892 Homestead strike, and the 1894 Pullman strike.”41 Between 1901 and 1910, there were 162 strikes per million nonagricultural workers.42 During the 1919-1920 strike wave, twenty-two percent of the nonagricultural workforce went on strike.43 In comparison, in 2015, there were thirteen “major work stoppages” total, involving less than .04% of the nonagricultural workforce.44

These strikes were part of a new strategic repertoire45 for the incipient labor movement, a form of protest made possible by the unique circumstances of industrial waged labor.46 Striking was risky, and not all labor unions were initially sanguine about the tool. The Knights of Labor, for instance, originally insisted that striking was counter-productive, too prone to backlash.47 Strikes were largely deemed illegal at the time, as criminal conspiracies and then as antitrust violations, and subject to court injunction.48 But workers kept striking, anyway. In the 1880s, workers struck throughout the country for the eight-hour day, the ability to share in the improved quality of life rapid growth had enabled. They proclaimed, “Eight hours for work, eight hours for sleep, eight hours for what we will.”49 In 1902, mine workers in eastern Pennsylvania struck, seeking shorter hours, higher pay, and recognition of their union.50 In 1912, the well-known “Bread and Roses” strike took place, in reaction to a pay cut. Female textile workers in Lawrence, Massachusetts walked out en masse, proclaiming “Hearts starve as well as bodies; give us bread, but give us roses!”51

The immediate outcomes of these strikes were mixed. With the help of government intervention, the 1902 coal strike was a relative victory; workers secured a nine-hour day and a pay raise, albeit no union recognition.52 But government intervention was usually not neutral. Strikes were deemed unlawful conspiracies, or anti-competitive cartel action. They were subject to massive legal repression by state police, federal military power, and federal courts.53 In contrast to the Progressive hope for state power, it was employers, not workers, who tended to benefit from state intervention during on-the-ground disputes between capital and labor. In the face of employer resistance, facilitated by law, workers often lost.

The “Great Steel Strike” of 1919-20—the last large strike of the Progressive Era—illustrates all that seemed possible, yet turned out not to be, in this Era. In the fall of 1919, more than 350,000 steel workers across the Northeast and Midwest walked off the job, bringing half of American steel production to a halt.54 But the Russian Revolution of 1917 had turned public opinion against labor, and the federal government opted not to intervene on behalf of the striking workers.55 State militias and local police imprisoned strikers, and employers brought in strike-breakers, weakening worker solidarity. In some areas, local police rounded up striking workers from their homes and forced them back to work.56After this loss, virtually no union organizing occurred in the steel industry for fifteen years.57

But these immediate losses were not the end of the story. These strikes grew the labor movement, creating the material (organized workers) and ideological (something must be done about the “labor problem”) infrastructure for the legal reforms to come.58Importantly, they changed public consciousness. By ensuring that workers’ experience of the new economy was a part of public discourse, strikes contributed to the Progressive challenge to laissez faire. As Louis Brandeis proclaimed following the 1902 coal strike, “The growth in membership has been large, but the change in the attitude toward unions both on the part of the employer and of the community marks even greater progress. . . . That struggle compelled public attention to the trades union problem in a degree unprecedented in this country.”59

The path was not linear. During the 1920s, a host of factors—including pandemic fatigue60—prompted the country to revert to its Gilded Age habits. But when the Great Depression hit, both the ideas and the on-the-ground power built in the decades prior allowed for rapid deployment of pro-labor legislation at just the moment when it was politically possible to implement it. The Norris-LaGuardia Act became law in 1932. The National Labor Relations Act (NLRA) followed in 1935.

The Great Steel Strike’s legacy extended beyond its immediate aftermath. When the Supreme Court upheld the NLRA against constitutional challenge in 1937, it cited the strike—that great failure—as evidence of the constitutional propriety of the Act.61 This “illegal strike” became part of the legitimating narrative for why government intervention in support of unionization was appropriate. “The Government aptly refers to the steel strike of 1919-1920, with its far-reaching consequence,” read the opinion; the “[r]efusal to confer and negotiate has been one of the most prolific causes of strife.”62 Industrial unrest disrupted the stream of commerce; government regulation to prevent such disruption was constitutional.

D. Labor’s Approach to the Political

Still, the bumpy road to the New Deal changed the labor movement, altering the scope of its demands. The labor movement which emerged at the end of the Progressive Era was different from what it had been, what it seemed it might become. While historians disagree about what exactly had changed and why, their stories are uniformly of paths not taken. In the late 1800s, consistent with the broad framing of the labor question, elements of organized labor fought for broad social and political reform, not just industrial citizenship for those skilled white men able to form unions. “Labor republicans” of the mid-1800s believed that the disparities in wealth and control linked to corporate wealth and waged labor were fundamentally incompatible with democratic governance. Terence Powderly, head of the Knights of Labor, sought “to forever banish that curse of modern civilization—wage slavery.”63 These reformers demanded “industrial democracy”—a term that once connoted collective ownership of industry, or at least broad political co-determination of its terms. By the end of the Progressive Era, however, industrial democracy was well on its way to becoming synonymous with collective bargaining.64

Whether due to employer opposition,65 judicial repression,66 American
67 or state violence,68 the “pure and simple unionism” of the American Federation of Labor (AFL) emerged as the leading vision for the role of labor unions in American society.69 Pure and simple unionism was less about democracy and more about wages, less about using the state to counterbalance corporate power and more about being left alone to engage in “self-help”70—economism and voluntarism, respectively.71 For this reason, when the Great Depression hit and political opportunity availed itself, labor’s primary request from public law was solitude. The Norris LaGuardia Act, labor’s first big legislative victory of the 1930s, deprived federal courts of jurisdiction over labor conflicts. Three years later, the NLRA gave workers the right to unionize, strike, and engage in other concerted activity for the purpose of collective bargaining. “More” was the AFL’s answer to the labor question—but it was not the “more” of a broad, redistributive state; it was “more” at the bargaining table, by those workers with enough power to command it.72

William Forbath sums up the AFL’s ethos in the Progressive Era as, ironically, its own version of laissez faire.73 In an era in which “Progressive” reformers were unified by the “bedrock conviction that positive statism was necessary to remedy the ills of modern life, and that continued negative statism invited disaster,”74 the AFL insisted that the state could not be trusted to help workers. In other words, the form of organized labor which was becoming dominant was not “Progressive” at all.

Built on these foundations, the NLRA ended up codifying a relatively economistic and voluntarist understanding of organized labor’s role in a democratic polity.75 Historian Leon Fink attributes this directly to the legacy of the Progressive Era:

[F]or all the “triumph” of long-sought principles in the 1930s act, a mortal weakness was also left over from its Long Gilded Age inheritance. Appealing at best to the interstate commerce clause allowing Congress to set appropriate economic policy for the national interest, the Wagner Act was not rooted in deep constitutional principles (such as the Bill of Rights) that would demand respect for unions as a necessary pillar of American democracy.76

As I discuss below, the vision of the strike which developed under the NLRA was similarly shaped by this “mortal weakness”—an apoliticism, in process and substance. With the enactment of the NLRA, the strike became lawful as a bargaining tactic of last resort, not as a political right, necessary for co-determination.

II. legal ambivalence towards the strike: constructing the “economic weapon”

The strike has never fit easily within extant legal categories. According to Craig Becker, “the law has variously categorized strikes as criminal activity, as an invasion of property rights, and as a fundamental component of labor’s right to engage in collective bargaining.”77 Jurisprudentially, striking has been theorized as either an associational freedom upon which law cannot intrude, or in the alternative, conduct so coercive and disorderly as to be antithetical to the rule of law—industrial vigilante justice.78 Following enactment of the NLRA, strikes ostensibly became legal for the private sector workers covered by it. But especially after the 1947 Taft-Hartley Amendments to the NLRA, striking’s legality was tied to an increasingly narrow understanding of its purpose. In this Part, I provide a brief overview of how current law—shaped by its Progressive Era mortal weakness—codifies long-lasting legal ambivalence about striking, by constructing the strike as an “economic weapon,” and in so doing, as apolitical.

A. The “Right” to Strike

Under the NLRA, workers are generally understood to have a “right” to strike. Section 7 of the Act states that employees have the right to engage in “concerted activities for . . . mutual aid or protection,”79 which includes striking. To drive this point home, section 13 of the NLRA specifies, “Nothing in this [Act] . . . shall be construed so as either to interfere with or impede or diminish in any way the right to strike . . .”80 Note that it is a testament to deeply-held disagreements about the strike (is it a fundamental right which needs no statutory claim to protection, or a privilege to be granted by the legislature?) that the statute’s language is framed in this way: the law which first codified a right to strike does so by insisting that it does not “interfere with or impede or diminish” a right, which had never previously been held to exist.81

To say that a strike is ostensibly legal, though, is not to say whether it is sufficiently protected as to make it practicable for working people. Within the world of labor law, this distinction is often framed as the difference between whether an activity is legal and whether it is protected. So long as the state-as-regulator will not punish you for engaging in a strike, that strike is legal. But given that striking is protest against an employer, rather than against the state-as-regulator, being legal is insufficient protection from the repercussion most likely to deter it—job loss.

Employees technically cannot be fired for protected concerted activity under the NLRA, including protected strikes. But in a distinction that Getman and Kohler note “only a lawyer could love—or even have imagined,”82, judicial construction of the NLRA permits employers to permanently replace them in many cases. Consequently, under the perverse incentives of this regime, strikes can facilitate deunionization. Strikes provide employers an opportunity, unavailable at any other point in the employment relationship, to replace those employees who most support the union—those who go out on strike—in one fell swoop. As employers have increasingly turned to permanent replacement of strikers in recent decades, strikes have decreased.83 A law with a stated policy of giving workers “full freedom of association [and] actual liberty of contract” offers a “right” which too many workers cannot afford to invoke.84

It is not just that the right is too “expensive,” however; it is that its scope is too narrow, particularly following the Taft-Hartley Amendments. Law cabins legitimate strike activity, based on employees’ motivation, their conduct, and their targets. The legitimate purposes are largely bifurcated, either “economic,” that is to provide workers with leverage in a bargain with their employer, or to punish an employer’s “unfair labor practice,” its violation of labor law (but not other laws). A host of reasons that workers might want to protest are unprotected—Minneapolis bus drivers not wanting their labor to be used to “shut down calls for justice,” for instance. Striking employees also lose their limited protection if they act in ways that are deemed “disloyal” to their employer,85 or if they engage in the broad swath of non-violent activity construed to involve “violence,” such as mass picketing.86 Tactically, intermittent strikes, slow-downs, secondary strikes, and sit-down strikes are unprotected.87 Strikes are also unprotected if unionized workers engage in them without their union’s approval,88 if they concern nonmandatory subjects of bargaining,89 or if they are inconsistent with a no-strike clause.90 Independent contractors who engage in strikes face antitrust actions.91 Labor unions who sanction unprotected strikes face potentially bankrupting liability.92

The National Labor Relations Board—the institution charged with enforcing the policies of the Act—summarizes these “qualifications and limitations” on the right to strike on its website in the following way:

The lawfulness of a strike may depend on the object, or purpose, of the strike, on its timing, or on the conduct of the strikers. The object, or objects, of a strike and whether the objects are lawful are matters that are not always easy to determine. Such issues often have to be decided by the National Labor Relations Board. The consequences can be severe to striking employees and struck employers, involving as they do questions of reinstatement and backpay.93

The “right” to strike, it seems, is filled with uncertainty and peril.

Collectively, these rules prohibit many of the strikes which helped build the labor movement in its current form. Ahmed White accordingly argues that law prohibits effective strikes, strikes which could actually change employer behavior: “Their inherent affronts to property and public order place them well beyond the purview of what could ever constitute a viable legal right in liberal society; and they have been treated accordingly by courts, Congress, and other elite authorities.”94

B. The Limits of Legal Categories

Legal scholars have long tried to make sense of the law’s ambivalent treatment of worker power and collective self-determination. Here, I emphasize one point. What makes striking, and the legal apparatus for it, so complicated is that the strike transcends easy legal categorization. In one respect, it is a form of protest, fundamentally normative and political. Yet, that protest takes place within an economic relationship where property and contract law reign supreme.95 To the extent a strike looks like protest, it approximates the kind of activity that should be a fundamental right. But because it takes place not in the public sphere, but at work—within these “authoritarian, private government[s],” as philosopher Elizabeth Anderson recently described them—such rights do not carry over.96

Our eighty-five-year-old labor-law regime circumnavigates these complex jurisprudential issues by conceptualizing the strike as economic activity. In the almost anachronistic language used by the Supreme Court, strikes are “economic weapons in reserve.”97 And “their actualexercise on occasion by the parties, is part and parcel of the system that the Wagner and Taft-Hartley Acts have recognized.”98

A full genealogy of the term “economic weapon,” and how it came to such prominence in legal discourse, is beyond the scope of this short Essay.99 But it is worth noting that in pre-New Deal labor discourse, labor activists used the term to emphasize voluntarism over electoral politics, material power over normative. In 1924, the AFL’s magazine, the American Federationist, reflected on the voluntarism of American unions:

[Electoral politics] is the fatal lure in other countries, the deadly trap, the rock on which labor goes smash, soon or late.

It has never fooled the labor movement in America. The American worker goes into politics and uses his ballot according to his convictions, but he does not tie his economic weapon into a bundle with his political power and then find he has made a slapstick at which in the end everybody laughs.100

The AFL’s view of the strike, as an “economic weapon” to use on the market battlefield, eventually became the legal understanding. Reflecting on the evolving law of the strike in 1920, Supreme Court Justice Taft proclaimed the strike an instrument “in a lawful economic struggle . . . between employer and employees as to the share . . . between them of the joint product of labor and capital.”101

Political scientist Josiah Lambert accordingly argues that law came to protect the strike as a “commercial right,” not a “citizenship right.”102 Legal scholars similarly highlight that a more expansive, quasi-constitutional view of striking as a “civil right” warred with the purely statutory conception of striking as an “economic[]interest.”103 The latter won the day, eclipsing early conceptions of a worker as more than “a factor of production” but “a self-governing citizen with rights and duties beyond those enumerated in the labor contract.”104 While visions of how else a strike might have been conceptualized within American law are varied and inchoate, the common thread is, again, that something else was possible.105 In 1914, the Clayton Act proclaimed that “[t]he labor of a human being is not a commodity or article of commerce.”106 By 1937, labor’s commerce-ness had become the source of its rights.

Consistent with the construction of the strike as economic (and not political), public-sector employees excluded from the NLRA largely lack a legally protected right to strike. Most federal and state labor relations laws prohibit strikes by public
107 Where strikes are prohibited by law, workers can not only be fired for them but can also face loss of government benefits, fines, and even jail time.108 Their unions face similar consequences.109 Within a world in which strikes are understood as “a legitimate aspect of the market or enterprise economy,” public-sector strikes lack comparable legitimacy.110 They are “deemed inappropriate because the government is not merely an employer participating in the economy, but is the lawgiver for the economy.”111 In other words, they are political. In Janus v. AFSCME, the Supreme Court finally ruled public-sector agency-fee laws unconstitutional, on the grounds that line drawing between the economic and the political in the public sector is untenable.112 And it is, but Janus does little to bring coherence to the law.113

To be clear, there is a way in which categorizing the strike as an economic act is radically pro-labor. According to labor law, the collective withholding of labor, facilitated by law, is neither an illiberal conspiracy nor an anticompetitive restraint of trade. It is a legitimate bargaining tactic.

C. Alternate Visions

And yet, the pre-New Deal history of the American labor movement reminds us that alternate visions for the strike once existed, and with them, for the role of organized labor in a democratic polity. Legal scholars of the first law-and-economics movement argued that corporations were akin to governments in their coercive power.114 They did so in order to politicize the private, to render it a legitimate site of democratic governance.115 Jurisprudentially, much was up for grabs at this point. Corporations might have been analogized to governments, rather than persons, owing constitutional rights to their constituents, rather than holders of those rights against government regulation. Consistent with this view of employment as a political relationship, the right to strike could have been a right of a different sort (political? civil? property?), not a commercial one.116

The labor movement which emerged from the Progressive Era, however, did not advance this vision of political economy. Fearing that any broadly conceived “public interest” would prioritize capital over labor, the AFL preferred to build worker-led institutions, organizations that would be corporations’ counterparts. It advanced a corporatist solution rather than a statist one. And within the world of industrial democracy qua collective bargaining, the strike became understood as an economic right, hard bargaining, a last resort in negotiations gone awry.117

The limitations of this understanding of the strike did not immediately become apparent. Among other factors, Keynesian political economy gave labor unions legitimacy as agents of a common good.118 Employers—constrained in part by law, and even more by institutional understandings of what was acceptable—pulled their punches. When the NLRA was first enacted, strike activity, union membership, and worker wages grew concurrently.119 And even after enactment of the Taft-Hartley Amendments—in the era of the so-called “labor-capital” accord—strikes continued to occur with some frequency, and to be associated with wage growth for workers.120

As I discuss further below, this changed in the 1970s, as the Keynesian political economic commitments that had bolstered labor law gave way. Between 1983 and 2015, total union membership declined by approximately 44%, and private sector union membership declined by 60%.121 Over a similar period, major work stoppages declined by 90%.122 And the few strikes that still did occur were qualitatively different. They were “desperate measures,” which failed to advance union objectives.123 They did not create wage growth, the “more” that Progressive Era labor leaders had envisioned.124 As an economic weapon, the strike malfunctioned.125

III. labor’s twenty-first century challenge to neoliberalism: striking in a new gilded age

A. Work Law in Political Economic Context

In analyzing labor’s current predicament, most legal scholarship emphasizes the limitations of law. “So it is,” Ahmed White concludes, “that workers have found themselves with a right to strike that equals little more than a right to quit work—and maybe lose their jobs or their houses and savings in the balance.”126 The unstated assumption in many of these pieces is that workers rarely strike because the law is bad. And there is little question that current legal impediments powerfully influence unions’ and workers’ calculus about whether striking is worth it. But if the massive labor unrest of the Gilded Age and Progressive Era tells us anything, it is that the relationship between bad law and worker action is not as direct as legalistic accounts suggest. In this Part, I draw from the experiences of labor in the Gilded Age and Progressive Era to discuss the relationship between law, political economy, and social change today, and how strikes may serve the labor movement and the polity—bad law, notwithstanding.

Regulation of the workplace is inexorably connected to political economy.127 Work law, like all law, is mediated by institutionalized practices and legitimating
narratives.128 And so it is for the law of the strike. Nowhere is this insight better illustrated than in the regulation of “permanent replacements.” In explaining the lack of strikes in recent years, many scholars immediately point to the rule that permits employers to permanently replace economic strikers.129 Yet, the rule precedes the drastic decline in strikes by decades. It was not a reactionary Burger, Rehnquist, or Roberts Court which construed the NLRA to permit permanent replacement of strikers. Rather, the Supreme Court unanimously adopted this construction of the NLRA in 1938, one year after it affirmed the constitutionality of the Act.

In a classic example of what law-and-society scholars describe as the gap between law on the books and law in action,130 employers largely did not enforce the “right” to permanently replace striking employees until four decades later.131 As labor scholar John Logan tells it, “The legal status of permanent replacements did not seem terribly important in the 1940s-1960s, as relatively few employers used or threatened to use permanent replacements.”132 This is partly because unions were powerful during these decades, and it was strategically unwise for employers to escalate labor conflict (not to mention that it was difficult to recruit workers to serve as “scabs”).133 But it was also because the New Deal changed more than law; it changed the institutional practices and cultural commitments that followed from it.134 For a time, unions became a relatively unquestioned part of the status quo.135 This is not to say that the scope of their purview was uncontested; it always has been. Their fundamental legitimacy, however, was not in question, as it has been of late.

The labor-capital accord ended with the neoliberal turn of the late 1970s.136 And while “neoliberalism” can be a frustratingly capacious term, its analytical purchase here is in situating labor’s current struggles in relation to regulatory and discursive changes that have prioritized the “free” market over other forms of ordering, like unions. Under neoliberalism, existing arguments for why unions served a common good were subject to epistemic erasure.

Labor’s historical antipathy to the “political” rendered it particularly vulnerable to the neoliberal turn. Labor’s New Deal accommodation, born of its Progressive Era “mortal weakness,” made an economic case for labor unions, and left the normative case for them underspecified. Keynesian demand-side economic policy and industrial peace were the NLRA’s leading rationales.137 Neoliberal political economy undermined both. It flipped the Keynesian script, depicting capital as the driver of economic well-being, rather than worker income-qua-consumer spending. And, a few crushed strikes provided a watered-down version of industrial peace. Ronald Reagan’s decision to use the power of the federal government to end the 1981 federal air controllers’ strike (an illegal and unsuccessful strike) by firing striking workers, replacing them, and decertifying their union,138 dramatized this shift. Private employers began replacing workers on strike too. Workers in turn stopped striking.139 And for some, this was taken as proof that labor law had done its job of achieving industrial peace, and was no longer needed.140

And so, while there is a tendency to see hostile law as the reason for worker quiescence, it is not formal law alone that is the problem. The problem is equally an economic and cultural milieu which renders these rules cognizable, legitimate, and enforceable. In recent decades, a vision of capital as the driver of economic growth and unions as rent-seeking interest groups has enabled the permanent replacement of strikers, the deployment of union avoidance consultants, and a host of anti-union practices that used to be illegitimate, but not necessarily illegal. Labor lost “the contest of ideas.”141

At one level, the importance of political economy is acknowledged in the literature. Craig Becker concedes in a footnote that “[t]o emphasize the law’s role is not to imply that the efficacy of strikes rests solely on formal legal rights, for strikes were waged with success prior to the advent of legal protection.”142 And White notes that law can be “malleable and . . . within the province of workers to reshape around their own interests and visions.”143 Still, through emphasis on formal legal rules, legal scholarship has at times failed to recognize the magnitude of neoliberalism’s impact, not just as a material change in conditions, but as an ideological change in what is possible to imagine.144

Reflecting a more nuanced approach to the relationship between formal legal rules and what happens on the ground, labor organizers are themselves ambivalent about how much bad law matters. A well-known maxim within the movement declares that “there is no such thing as an illegal strike, only an unsuccessful one.”145 Labor-movement activists speak matter-of-factly about a union’s strategic choice to disregard legal rules:

A union that decides to break anti-worker laws should do so united, and with a plan for the consequences.

Is your leverage great enough to make the law moot? (They can’t fire us all.) Do you have lawyers on hand for the fallout?

Can you make withdrawal of legal charges part of the strike settlement? Will the public put the fear of God into politicians or police chiefs that try to harm the union? Balance the potential risks against the possible gains.146

Law is one source of leverage, activists proclaim, but it is not the only one.

This more nuanced account of the relationship between law, power, and culture is particularly important in the current historical moment. While the law of the strike has not changed in recent years, public consciousness about economic inequality and the potential role of labor unions in combatting it has changed—dramatically.147 Sociological research traces the origins of this shift, in significant part, to the Occupy Wall Street protests of 2011. These broad, public-facing protests did not change law, but they did propel economic inequality back into public discourse.148 Given that awareness of inequality increases support for labor unions,149 it is perhaps not surprising that public support for organized labor is now at a fifteen-year-high.150 That this sea change in public consciousness was led not by the labor movement, but by a motley crew of unorganized middle-class precariat151 says much about labor’s long turn away from the public sphere.

Consistent with an understanding of labor law and political economy as mutually constitutive (but not with a legal formalist account), in the past few years, there has been an uptick in strikes.152 The trend thus far is meaningful, albeit short. The number of workers who went on strike in 2018 and 2019—485,200 and 425,500 respectively—reflects the largest two-year-average in thirty-five years.153 Some of these strikes adhere to the narrow confines of labor law; many do not. Yet, like one hundred years ago, the efficacy of these strikes is not determined by their adherence to legal rules.

B. Striking as Political

For those who believe that a stronger labor movement is needed to counterbalance the concentrations of economic and political power in this new Gilded Age, the question is not just whether the law is bad (it is), but whether strikes can be effective nonetheless. If labor activists are correct that there is “no such thing” as an illegal strike, just an unsuccessful strike, the question follows: what makes a strike successful enough, under current conditions, to transcend legal constraints?154 To some extent this is an empirical question, and one on which there are many opportunities for generative research. Beginning with the theoretical, however, I suggest that the success of strikes must be measured in more than economic wins in the private sphere. Like their Progressive Era progenitors, their success must be in raising political consciousness in the public sphere—in making the stakes of the twenty-first century labor question apparent.155

As noted above, under current labor law, strikes are conceptualized as “economic weapons,” as hard bargaining.156 And while legal terminology is distinct from on-the-ground understandings, unions have often emphasized the economic nature of the strike as well. Strikes are “[t]he power to stop production, distribution and exchange, whether of goods or services.”157 A strike works because “we withhold something that the employer needs.”158 At the same time, there has been a corresponding tendency to dismiss the more symbolic aspects of the strike. To quote White again, “while publicity and morale are not irrelevant, in the end, they are not effective weapons in their own right.”159

These arguments are important. A strike is not simply protest; it is direct action, material pressure. But with union density lower than ever, ongoing automation of work tasks that renders employees increasingly replaceable, and decades of neoliberal cultural tropes celebrating capital as the driver of all economic growth and innovation, it is a mistake to think of publicity and morale as nice-to-haves, rather than necessities. Instead, striking must be part of building what sociologists have described as the “moral economy,” cultural beliefs about fair distribution untethered to technocratic arguments about what is most efficient.160 And in that way, striking is and must be understood as political.

The term political, of course, has many meanings—engendered by law, culture, and the relationship between the two. Building on the work of other scholars, I have argued that neo-Lochnerian readings of the First Amendment which have categorized labor protest as solely economic, and therefore apolitical, are one mechanism by which unions have lost legitimacy (and legal protection) as a social movement.161 Under current law, what precisely constitutes the political is less than clear, though. In distinguishing “political” speech from other kinds of speech for the purpose of First Amendment analysis, the Supreme Court has at times equated the political with: electioneering;162 speech directed to or about the government;163 or most broadly, “speech and debate on public policy issues.”164 Within labor parlance, by contrast, the term “political strike” is specifically used to refer to strikes that are “designed to win a specific political outcome, such as the passage of legislation or a change in
regulation.”165 Consistent with the NLRA’s construction of unions as economic entities, strikes which are solely “political” and without sufficient nexus to the employment relationship, are deemed unlawful secondary boycotts.166

But my argument here for reconceptualizing the strike as political is not about more “political strikes,” or about electoral politics, or even necessarily about state action. Based on a vision of the “political” as normative engagement directed towards collective decision-making—it is about destabilizing jurisprudential line drawing between the economic and the political in the first place.167 It is recognizing that all strikes are political or have the potential to be—in that all strikes are protest meant to transform collective conditions, not merely bargaining towards immediate, transactional ends. To use political science terminology, strikes are contentious politics: “[E]pisodic, public, collective interaction among makers of claims and their objects.”168 They are a way through which workers engage in claims-making when business and politics as usual have proven nonresponsive.169 They do not only address the employer; they engage the polity.

The need to reconceptualize the strike as outward-facing towards the public, not just inward-facing towards the employer, is partly a function of material changes, both in economic production and union density. As labor scholar Jane McAlevey points out, “Today’s service worker has a radically different relationship to the consuming public than last century’s manufacturing worker had . . . In large swaths of the service economy, the point of production is the community.”170 For this reason, she argues that effective strikes today must engage the public to be successful.171 Union density is also many times higher now in the public sector than in the private one, an upending of the realities of unionization mid-century.172 As illustrated by the Supreme Court’s decision in Janus v. AFSCME, it is easier to see the economic work of unions as political (qua affecting government policy, spending, and debt) in the public sector.173

Yet, the shift is also about recognizing that it was a legal and an ideological accommodation that made the work of unions in their representative capacity appear as “economic,” and thus outside politics. The work of unions has been artificially “bifurcated” vis-à-vis the political realm.174 For years, as Reuel Schiller has argued, unions have engaged in “two sets of activities that appear barely related to one another”: private, transaction bargaining in the workplace; combined with broad, public mobilization around electoral politics. But there were always alternate visions of the relationship between the economic and the political within union advocacy and workplace governance.175 If “establishing terms and conditions of employment [is] a political act involving not just a worker and an employer, but also a union, an industry as a whole, and the state,” then union advocacy is a political act too.176 Strikes are part of the “contest of ideas.”

Reconstructing a purposefully political philosophy, jurisprudence, and tactical repertoire of collective-labor advocacy is a project that is new again; and it will inevitably require deliberation, debate, and compromise.177 For the time being, though, one thing seems apparent. Strikes must be a part of engaging a broad swath of the public in reconceptualizing political economy.

C. New Strategies, New Risks

In recent years, consistent with this vision, there has been a shift in the kinds of strikes workers and their organizations engage in—increasingly public-facing, engaged with the community, and capacious in their concerns.178 They have transcended the ostensible apoliticism of their forebearers in two ways, less voluntaristic and less economistic. They are less voluntaristic in that they seek to engage and mobilize the broader community in support of labor’s goals, and those goals often include community, if not state, action. They are less economistic in that they draw through lines between workplace-based economic issues and other forms of exploitation and subjugation that have been constructed as “political.” These strikes do not necessarily look like what strikes looked like fifty years ago, and they often skirt—or at times, flatly defy—legal rules. Yet, they have often been successful.

Since 2012, tens of thousands of workers in the Fight for $15 movement have engaged in discourse-changing, public law-building strikes. They do not shut down production, and their primary targets are not direct employers. For these reasons, they push the boundaries of exiting labor law.179 Still, the risks appear to have been worth it. A 2018 report by the National Employment Law Center found that these strikes had helped twenty-two million low-wage workers win $68 billion in raises, a redistribution of wealth fourteen times greater than the value of the last federal minimum wage increase in 2007.180 They have demonstrated the power of strikes to do more than challenge employer behavior. As Kate Andrias has argued:

[T]he Fight for $15 . . . reject[s] the notion that unions’ primary role is to negotiate traditional private collective bargaining agreements, with the state playing a neutral mediating and enforcing role. Instead, the movements are seeking to bargain in the public arena: they are engaging in social bargaining with the state on behalf of all workers.”181

In the so-called “red state” teacher strikes of 2018, more than a hundred thousand educators in West Virginia, Oklahoma, Arizona, and other states struck to challenge post-Great Recession austerity measures, which they argued hurt teachers and students, alike.182 These strikes were illegal; yet, no penalties were imposed.183 Rather, the strikes grew workers’ unions, won meaningful concessions from state governments, and built public support. As noted above, public-sector work stoppages are easier to conceive of as political, even under existing jurisprudential categories.184 But these strikes were political in the broader sense as well. Educators worked with parents and students to cultivate support, and they explained how their struggles were connected to the needs of those communities.185 Their power was not only in depriving schools of their labor power, but in making normative claims about the value of that labor to the community.

Most recently, 2020 saw a flurry of work stoppages in support of the Black Lives Matter movement.186 These ranged from Minneapolis bus drivers’ refusal to transport protesters to jail, to Service Employees International Union’s Strike for Black Lives, to the NBA players’ wildcat strike.187 Some of these protests violated legal restrictions. The NBA players’ strike for instance, was inconsistent with a “no-strike” clause in their collective-bargaining agreement with the NBA.188 And it remains an open question in each case whether workers sought goals that were sufficiently job-related as to constitute protected activity.189 Whatever the conclusion under current law, however, striking workers demonstrated in fact the relationship between their workplaces and broader political concerns. The NBA players’ strike was resolved in part through an agreement that NBA arenas would be used as polling places and sites of civic engagement.190 Workers withheld their labor in order to insist that private capital be used for public, democratic purposes. And in refusing to transport arrested protestors to jail, Minneapolis bus drivers made claims about their vision for public transport.

Collectively, all of these strikes have prompted debates within the labor movement about what a strike is, and what its role should be. These strikes are so outside the bounds of institutionalized categories that public data sources do not always reflect them.191 And there is, reportedly, a concern by some union leaders that these strikes do not look like the strikes of the mid-twentieth century. There has been a tendency to dismiss them.192 In response, Bill Fletcher Jr., the AFL-CIO’s first Black Education Director, has argued, “People, who wouldn’t call them strikes, aren’t looking at history.”193 Fletcher, Jr. analogizes these strikes to the tactics of the civil-rights movement.

As Catherine Fisk and I recently argued, law has played an undertheorized role in constructing the labor movement and civil-rights movement as separate and apart from each other, by affording First Amendment protections to civil rights groups, who engage in “political” activity, that are denied to labor unions, engaging in “economic” activity.194 Labor unions who have strayed from the lawful parameters of protest have paid for it dearly.195 As such, it is no surprise that some unions are reluctant to embrace a broader vision of what the strike can be. Under current law, worker protest that defies acceptable legal parameters can destroy a union.

Recasting the strike—and the work of unions more broadly—as political is risky. Samuel Gompers defended the AFL’s voluntarism and economism not as a matter of ideology but of pragmatism; he insisted that American workers were too divided to unite around any vision other than “more.”196 He did not want labor’s fortunes tied to the vicissitudes of party politics or to a state that he had experienced as protective of existing power structures. Now, perhaps more than ever, it is easy to understand the dangers of the “political” in a divided United States. Through seeking to be apolitical, labor took its work out of the realm of the debatable for decades; for this time, the idea that (some) workers should have (some form of) collective representation in the workplace verged on hegemonic.

And yet, labor’s reluctance to engage in the “contest of ideas” has inhibited more than its cultivation of broader allies; it has inhibited its own organizing. If working people have no exposure to alternative visions of political economy or what workplace democracy entails, it is that much harder to convince them to join unions. Similarly, labor’s desire to organize around a decontextualized “economics” has always diminished its power (and moral authority), given that the economy is structured by race, gender, and other status inequalities—and always has been. During the Steel Strike of 1919, the steel companies relied on more than state repression to break the strike. They also exploited unions’ refusal to organize across the color line. Steel companies replaced striking white workers with Black workers.197 Black workers also sought “more.” But given their violent exclusion from many labor unions at the time, many believed they would not achieve it through white-led unions.198

From this perspective, labor’s “mortal weakness,” derived from its Progressive Era legacy has not just been failing to articulate its moral vision; it has been lacking one sufficiently capacious to build the power necessary for the task at hand. As David Huyssen argues:

Progressives and New Dealers . . . achieved their reforms by reaffirming the Gilded Age’s ideological and legal commitments to white supremacy, imperialism, and xenophobia. . . . Signature New Deal legislation—the Social Security Act and the National Labor Relations Act—discriminated against women and African Americans by excluding domestic and agricultural workers, valorizing the white male family wage earner.199

“The ‘solutions’ that ended Gilded Age inequality,” he concludes, “became a crucial seedbed for our own era’s historically distinct expressions of inequality.”200 It is unsurprising, then, that worker protest today so often goes against the grain of the existing legal infrastructure.

D. Law and Social Change

In arguing that strikes can be successful today, in spite of law, my argument is not that law is unimportant; it is that law does not exist independently of other social processes. Political contestation will always be on multiple fronts, and success in one arena shifts the balance of power in others.

Tumultuous times come to an end. Social agitation is difficult to sustain, and the gravitational pull of existing power structures tugs us towards a less tumultuous, if forever altered, new normal. Institutional change is what lasts beyond the tumult. For a new Progressive Era to correct the multiple, overlapping forms of inequality which have led to the current moment, a broad new social compact is necessary. That requires political co-determination and law.201 Law matters (it is just not the only thing that matters).

Perhaps unsurprisingly, the limitations of labor’s current legal regime have led to a resurgence in labor’s laissez-faire instincts. Legal scholars caution skepticism about the state; they urge labor to seek “freedom” but not “rights.”202 And they are correct that labor must strategically approach state power from a place of political realism. But whatever those realities, the important work of small-p politics and small-c constitutionalism remains.203 For better or worse, the Progressive view of the state has won the day, and legal contestation is the crucible in which many public values are built. To hive off the economic from the political, the private from the public, cedes too much.

From this perspective, the phrase “there is no such thing as an illegal strike” is itself a remnant of labor’s sticky voluntarism, its apoliticism. Having been cast as outlaws in the Gilded Age, union leaders “internalize[d] this negative identity” and have wielded it since.204 Labor scholars today question why unions’ long history of illegal strikes is not told, alongside the tactics of other movements, as a civil disobedience story.205 But law-breaking by itself is not civil disobedience. It is through articulating an alternate normative vision—through law-making—that illegal strikes become civil disobedience.


The Progressive Era involved massive protest, and a host of different explanations and proposed solutions for how to reconcile the inequalities of capitalism with the egalitarian commitments of a democracy. The ideas and institutions that stuck helped create the scaffolding for the legal reforms of the New Deal Era. That resolution of the labor question settled on one particular vision of organized labor as a social good: organized labor would be an economic entity that would engage in economic conflict with employers, and, in turn, the strike would be an economic weapon rather than a political one. That vision of labor proved insufficient to counter the challenges of the neoliberal period.

How we understand today’s “labor question” will inform the next legal regime for labor and whatever new normal it ushers in. It took economic devastation and protest to bring about the partial victory of the first New Deal. Workers across the country, including Minneapolis bus drivers, are now risking their livelihoods to bring about something better.206

University of California, Berkeley. I am grateful to Kate Andrias, Catherine Fisk, David Singh Grewal, Deborah Malamud, Yan Fang, Yael Plitmann, and Kavitha Iyengar for their comments on previous drafts of this Essay; to John Logan, Joseph McCartin, and Dylan Penningroth for their counsel; to the editors of the Yale Law Journal for their vision and skillful editing; and to the workers and teachers whose loving care for my two young children in the midst of a pandemic enabled me to take on this project.