The Yale Law Journal


The Once and Future Countervailing Power of Labor

03 Feb 2021

A Response To

abstract. This Essay looks to the history of the labor movement to engage with Kate Andrias’s and Benjamin I. Sachs’s argument that labor law provides a model of how law can support organizing. While embracing the solidarity of labor scholarship with the vibrant scholarly conversation about law and progressive social movements, I sound four cautionary notes drawn from past repression of labor and civil-rights activism. Labor’s experience is that law tends not only to thwart and suppress but also to channel movement activity in ways that weaken threats to the hegemony of the wealthy under capitalism. Sketching three specific ideas about how law can enable the exercise of countervailing power without also channeling activism away from radical challenges to elite power, I note how law has made it both essential and difficult for labor and racial justice activists to organize together against inequality. Although today’s inclusive activism gives hope for completing unfinished work of the New Deal, Civil Rights, and Great Society eras, we should be clear-eyed about the role law played in leaving that work undone and the difficulties of using law to build sustainable class-based social movements.

Introduction: Using Law To Transform Today’s Mobilization Into Tomorrow’s Countervailing Power

Don’t waste any time mourning—organize!”

- Joe Hill1

By now it borders on cliché to note the similarities between 2020 and 1918, 1930, and 1968. Yet the comparisons are useful if they serve as a guide to legal reforms that will complete the unfinished business of the progressive movements of the past. This pandemic, like that of 1918, reveals that universal access to health care is necessary to protect the common good. This Gilded Age, like the one that crashed to an end in 1930, reveals the problems with unregulated capitalism and the gaping holes in the social safety net as people—especially people of color—face unprecedented levels of unemployment and a predicted tidal wave of evictions without access to adequate unemployment benefits, debt relief, or a right to housing.2 The massive Black Lives Matter (BLM) protests—perhaps the largest protests in American history—like the uprisings of 1968, reveal that racist police killings are the tip of an iceberg of systemic racism.3

What will it take for these cataclysmic events to produce a new and more racially inclusive New Deal that will complete the unfinished work of the labor and civil-rights movement activism of 1930-1970? One need not squint too much to see the possibility of dramatic change in protests against the extreme inequalities exacerbated by the pandemic and police violence.4 The dawning awareness among elites that structural racism and inequality threaten our democracy might finally give political traction to working class organizing. Hope is manifest in the BLM protests, and in the upsurge of labor protest, including teachers’ strikes in 2019, and the 2020 strikes, protests, and whistleblowing by Amazon warehouse workers,5 Instacart shoppers,6 and even Facebook staffers angry at the company’s refusal to address false and hatemongering posts by President Trump.7

Labor was one of the major successful social movements of the twentieth century. It was one of the few that aspired to build countervailing power against the hegemonic rule of capital. Building countervailing power of the working class, Kate Andrias and Benjamin I. Sachs remind us, was the way in which Americans built a new political economy in the New Deal, and it could build a new political economy now.8 Their timely article, Constructing Countervailing Power: Law and Organizing in an Era of Political Inequality, argues that law can be a tool to enable the kind of sustainable organizations that exercise countervailing power. They urge reformers to consider the example of labor law, especially as it existed before major legislative change in 1947. Labor law was unusual in American law in promoting organizing and collective action of working people with the goal of enabling them to exert countervailing power against the concentrated economic, social, and political power of corporations and employers.

In this Essay, I embrace and build on Andrias and Sachs’s law-and-organizing frame and their ideas about how laws can enable the formation of working people’s movements. I also suggest some cautionary notes drawn from law’s response to uprisings of the past, where labor and civil-rights activism were repressed by antiradical, anti-immigrant, and antilabor legislation.9 As labor scholars and former labor organizers, Andrias and Sachs know as well as anyone the many ways law can both support and thwart the formation of sustainable and transformative class-based social movements. Law tends not only to thwart and suppress but also to channel movement activity in ways that weaken threats to the hegemony of the wealthy under capitalism.10 I seek, therefore, to elaborate on their ideas while bearing in mind lessons from history about how law enabled, channeled, and thwarted labor’s quest to be a transformative social movement. In exploring specific ideas about how law can enable worker organizing without also channeling activism away from radical change and thwarting real challenges to elite power, I sketch ways in which law has made it both essential and difficult for labor and racial justice activists to organize together against inequality around the shared identity as have-nots.

Part I of this Essay explores four lessons for today’s class-based social movements from the history of law’s successful channeling of labor. Part II sketches three ways that labor organizations can navigate the law to become the countervailing power that Andrias and Sachs promote. Part III concludes with some reminders that, while law can enable today’s activism to complete some of the work that was left undone in the New Deal, Civil Rights, and Great Society eras, we should be clear-eyed about the role law played in leaving that work undone and the difficulties of using law to build sustainable class-based social movements.

I. constructing countervailing power: lessons from the mid-twentieth century

Andrias and Sachs argue that mass-membership organizations were instrumental in winning changes in law and policy that benefitted ordinary people rather than elites.11 In a political economy dominated by huge and concentrated power on the one side, organization on the other side is necessary for law to be responsive to the non-elite. The dearth of countervailing power, as John Kenneth Galbraith called it,12 or what Andrias and Sachs call “the paucity of collective organization among non-elites,” is not “a natural occurrence” stemming from inexorable economic principles (e.g., collective action problems). Rather, it is a product of legal rules that make it hard to organize.13

The elements of a legal regime that will nurture class-based organizing essential to economic and political democracy, Andrias and Sachs argue, are found in the National Labor Relations Act (NLRA).14 The NLRA was the single most significant government effort to remove legal obstacles to collective action. Andrias and Sachs make the case that effective organizing requires legal rights to access spaces in which to organize, to reliably collect funds from members, to be free from retaliation, and to bargain with relevant actors over conditions—to do all the things that the National Labor Relation Board (NLRB) under the Trump Administration and the Supreme Court are busy using law to block.15 Both the conservatives who are rolling back protections for labor and liberals like Andrias and Sachs share a belief that law matters; law is a cause and power is an effect. It is well known that after Congress enacted legal rights to unionize in 1935, union membership exploded.16 It is less well known that after Hawaii enacted legal rights to organize agricultural workers, union membership in that state skyrocketed, forever transforming a red-state plantation economy into a blue state.17 Thus, the case for a law to help the development of countervailing power is compelling.

But if the history of the labor movement offers inspiration for those who believe law can promote organizing, it also offers cautionary tales. Here I sketch four.

A. Promotion of Organizing Entails Regulation of Organizations

Government protection of labor organizing came with government regulation of labor organizations. The labor organizations to which the NLRA gave a special role in the self-regulation of industry were no longer just private-membership organizations but were now imbued with the public interest.18 Law therefore created a tangle of incentives towards less radical, less political, more self-interested behavior.19 The Norris-LaGuardia Act20 and the NLRA reduced outright repression of labor as a social movement, but they channeled union activism towards a state-preferred goal—collective bargaining—and away from more radical movement objectives.21 The NLRA’s promotion of union organizing on a certain model changed the structure under which unions organized.22 As Andrias and others have argued, it did this by requiring unions to adhere to the NLRB’s determination of what constituted an appropriate bargaining unit (typically based in a single location of a single enterprise), rather than what the union considered best for its organizational objectives and worker interests (which could have been industry-wide or sectoral).23 The protection of worker power through mandated union security based on exclusive representation enabled racist unions to exclude people of color24 while empowering conservative and quiescent unions to push out more radical unions.25 And, as we now see with police, legal protections for majority unions have enabled union leaders to thwart challenges from reform-oriented or minority groups.26 Antilabor interpretations of the NLRA increased the costs of movement activism, including sit-down strikes (which had been crucial to organizing Detroit in 1937), slow-downs (useful for workers who fear being fired for striking), and secondary picketing or boycotts (which are necessary to exert effective power in supply chains and in the service economy).27 Courts and, to a lesser extent, the NLRB, imported pre-NLRA notions from master-servant law into the new labor law in ways that constrained the rights of workers to act collectively.28 Regulation of the internal affairs of unions increased financial transparency but did not invariably make unions more democratic and in fact increased the power of government and union opponents to channel union activity.29

The vast literature on twentieth-century labor and civil rights shows how business skillfully deployed the menaces of radicalism, socialism, and communism to tar ambitious labor and civil-rights organizing. The Taft-Hartley Act’s anticommunist oath requirement removed all NLRA protections from unions whose leadership refused to sign oaths repudiating communism, and refused access to the NLRB for such unions.30 The statute thus leveraged the protections of law and administrative processes to force unions to push radicals out of leadership positions.31 The National Association for the Advancement of Colored People and some other civil-rights groups sought to preserve their legitimacy and influence by taking a strong stance against communists in their ranks.32 Although there is plenty of blame to go around—indeed, some American communists did the cause of labor and civil rights no favors33—the purge of progressives and radicals reduced the vigor of organizing and derailed certain efforts to end race and sex discrimination on the part of employers and

History invites us to think about what happens after unions or other social-movement organizations gain power. At the end of World War II, just as unions were poised to win (through a massive strike wave) the wage increases that had been long delayed by wartime wage and price controls, business used the labor upsurge as a wedge issue to win Congress for the first time since 1932. Congress scaled back labor protections over President Truman’s half-hearted veto, insisting that labor had become too powerful. The Taft-Hartley Act of 1947 adopted a multipronged approach to reducing union power. It granted legal protections to workers who resisted unionization (thus enabling employers to undermine solidarity).35 It denied unionization rights to independent contractors and low-level supervisors, even though they historically had belonged to unions and needed the protections of collective representation.36 It prohibited certain forms of labor protest.37 It forced leftists out of leadership
positions.38 It made unions subject to suit as entities.39 It made collective-bargaining agreements enforceable by judges40 who used that power to enjoin strikes in violation of collectively bargained no-strike clauses and to award damages against unions that were insufficiently vigorous or competent in enforcing contracts against employer breach.41 The Landrum-Griffin Act continued the work of the Taft-Hartley Act by eliminating legal protection for secondary protests and by regulating internal union affairs.42

The ostensible purpose of this legislation was to curb union abuses of power and corruption, and make unions more responsible to their members and to their contracts with employers.43 But an important purpose and effect was also to weaken unions and reduce labor protest. As I have documented elsewhere, the legislation used unions’ institutional power as leverage to reduce their activism by punishing union picketing with crushing damages liability and injunctions.44 This, in turn, required union lawyers to protect the union by counseling against certain forms of protest, reviewing the content, timing, and location of union picket signs and other protest tactics, and even censoring union newspapers.45

B. Political Compromises that Exclude Some from Protection

If the original sin of America was slavery, the original sin of the New Deal was building on the legacy of slavery in the political compromise that got the New Deal passed. To get the votes of legislators from agricultural states, Congress excluded farmworkers and domestic workers from the NLRA right to unionize, which meant that Black and Latinx workers and women disproportionately were excluded.46 The Social Security Act of 1935 likewise disproportionately omitted the jobs dominated by people of color and women from the unemployment and old age insurance systems it created.47 These political choices were possible only because the political parties and the interest group organizations that drafted and lobbied for the legislation did not represent women and people of color. A legal regime like the NLRA that grants power to organizations risks entrenching inequality if the organizations that gain power under the law are not representative, as many unions during the New Deal were not.

A challenge in implementing the Andrias-Sachs proposal is to ensure that the organizations that law supports remain responsive to the diversity of working-class interests. In today’s polarized and highly segmented political culture, no one should underestimate the difficulty of forming broadly inclusive unions. For instance, Whites without a college degree (a common definition of working class) are much more likely to describe themselves as Republicans or conservative than are working class Blacks or Latinx.48 Building a political agenda across the divides of race and ideology is tremendously difficult because race has been used for so long and still is used so effectively to divide the working class.49 Claims of racial justice or racial resentment are powerful appeals in organizing. As long as racial and class oppression exist, racial and class resentment can be stoked, and opponents of change will stoke them.

Unions with diverse membership and leadership are well-suited to organize inclusively along lines of occupational, class, racial, and other identities. As membership organizations committed to principles of solidarity, they are communities in which injustice can be confronted and reconciliation can be cultivated. To ensure that inevitable political compromises in legislation do not unfairly sacrifice the interests of some group, it is essential that the organizations involved in crafting the legislation be diverse and representative of the full swath of America. But the times in which this has been successful are regrettably few.

C. Regulating Countervailing Power

A third lesson from history is that when organizing succeeds and unions build political power for their members, opponents will argue that concentrated political power disserves the public interest. As noted above, the Taft-Hartley Act is one of many examples when working people’s organizations were weakened or destroyed when business organizations and elites regrouped and deployed law to suppress the threat to what they argued was the public interest in “labor peace” or “free movement of commerce” or “property rights.” But one need not go back to the 1940s to find examples of law empowering skeptics of unions, and the fear of union power has crossed political lines.

In Wisconsin in 2010, the Republican legislature eliminated collective-bargaining rights for workers whose unions opposed the candidacy of Scott Walker and preserved bargaining rights for workers—police, prison guards, firefighters—whose unions support Republicans.50 This was political payback, but it was also defended as serving the public interest in reducing labor costs and taxes and making the government more responsive to the people.51 Sometimes the critique of union power is bipartisan. The campaign to reduce collective-bargaining rights of teachers had bipartisan support, for the ostensible reason that teachers’ unions make it too hard to reform education in ways that would serve Black and Brown children.52 Even the Left attacks the concentrated power that a union provides, as seen in current proposals to eliminate or curtail police and prison-guard union rights as obstacles to police reform,53 or because police unionization is said to be associated with an increase in police violence.54

Organizations that have power to influence the regulatory processes designed to ensure they serve the public interest are rarely ones that represent the dispossessed. And when they do, elites label them a threat to capitalism. If unionization and collective bargaining are a privilege only for good workers, rather than a fundamental right, every group of workers who gain power is at risk of being found unworthy because its collective action will be portrayed as inimical to the public welfare.55

D. Transforming a Social Movement into a Sustainable Institution

A fourth cautionary note from history draws from the experiences of the United Farm Workers (UFW) in the 1970s. Enacting even a good law to enable organizing is not enough to build power if a union fails to do the work of balancing its social movement side with its role as an institution that exerts power in the workplace and in the economy.The UFW had secured the enactment of one of the most progressive labor laws in the country; California’s Agricultural Labor Relations Act contains most (but certainly not all) of the five elements of an Andrias-Sachs law.56 Only a few years after the law was enacted, however, the UFW began a downward slide, from which it has yet to recover (at least as measured by membership and collective bargaining agreements).57 In part, that was caused by employer and political opposition—employers managed to get the state to defund the agency and to bog it down with delays.58 And partly it is because of sectarian struggles among labor. The Teamsters, who had never had the slightest interest in organizing farmworkers, swooped in to replace the UFW at many farms. Growers were only too happy to replace the militant UFW with the more accommodating Teamsters.59 But partly it was because the UFW leadership did not aggressively pursue the traditional union sources of power, such as organizing dues-paying members and negotiating and enforcing collective-bargaining agreements. The institutional power the union could exercise through the negotiation and administration of collective agreements required a commitment of resources that some activists and leaders preferred to spend on more organizing, on political action, and on building and maintaining a regional and national movement of Latinx people.

Winning civil-rights or labor legislation does not transform workplaces or neighborhoods or politics if there is no mechanism to make the legal change stick on the ground. Unions as institutions have done this. But when a social movement becomes an institution, as unions did, the challenge is to continue the work of progressive transformation that is an unending project in capitalism.

II. law channels: thoughts on building a new political economy

In this Part, I will expand on the Andrias-Sachs proposal by sketching three examples of how law can promote worker organizing in a way that responds to the needs of the current moment. Bearing in mind the lessons of the past, close attention should be paid to institutional design to minimize the risks of relying on law to support organizing, especially when the organizing is oriented around challenging the existing distribution of wealth and power.60

First, law could promote worker power and voice immediately by ordering every workplace to recognize a worker-elected workplace safety and health committee.61 Workers often know better than the public or supervisors where the risks of illness and injury are, and they could form common cause with one another and with consumers, workers at other firms, and the public over the need to prevent the spread of infectious disease or other hazards. The public interest in allowing workers to organize around coronavirus infections is obvious because of the abject failure of the Occupational Safety and Health Administration (OSHA). With nearly twenty-five million confirmed COVID-19 cases and more than 400,000 deaths in the United States by mid-January 2021,62 OSHA had proposed only $3.5 million in total penalties by mid-December, mainly against relatively small businesses, such as nursing homes.63 Workers walked out of an Amazon warehouse—and later filed a public nuisance suit—when they contracted the coronavirus from coworkers who were encouraged to come to work sick and who lacked adequate paid sick leave. Amazon defended its conduct by saying it complied with OSHA guidelines, which was a plausible position because OSHA has failed to enforce existing law or to issue new regulations.64 State attorneys general pressured Walmart to bring its policies into compliance with law protecting workers from coronavirus because OSHA had done nothing.65 In the face of the failure of OSHA to protect workers, state and local governments could use their power to protect public safety by ordering every workplace to recognize an employee workplace health committee.66 This would build worker power on a broadly inclusive basis and demonstrate the public interest in empowering worker collective action.

A second proposal draws on what teachers unions pioneered beginning in 2012: bargaining for the common good.67 Bargaining for the common good (BCG) is a practice and a philosophy under which unions work with the community to identify collective needs and then negotiate with the employer to achieve those goals.68 BCG has at least three obvious advantages for building worker power. It makes collective action appealing to a wide swath of workers (beyond those concerned with the usual paycheck issues). It makes the union responsive to the needs and wishes of a diverse workforce and a diverse community. And it builds power for workers and for the community together by preventing antilabor groups from portraying the workers’ interests as being in conflict with the public interest. BCG also has some promise for making unions more accountable to the public by institutionalizing a role for the community in helping workers and the union identify problems and form priorities for negotiation. These are crucial if unions are to avoid some of the attacks sketched above in Part I.

At a minimum, state and local public-sector labor law should redefine the subjects of bargaining to allow BCG. More ambitiously, law could consider ways to incentivize unions to choose BCG by allowing workers whose unions practice it greater protections for resolving negotiating disputes. Public involvement in police-union negotiations, for example, might force departments and unions to rethink use-of-force policies or disciplinary processes that result in police violence.69 It might incentivize and enable radical restructuring of policing so that traffic, social service, and other functions are assigned to unarmed officers or civilian agencies; perhaps officers could be trained for these jobs rather than instinctively oppose the reorganization out of fear of job loss.70 Rather than prohibiting unions from negotiating over disciplinary procedures, law might mandate some kind of regular collective scrutiny of police contracts when they are up for renegotiation. Law might facilitate peer review and public scrutiny of police uses of force so that the union, rank-and-file officers, management, the public, and others could consider reforms, including early warning systems to identify problem officers and better ways to train rookies.71

A third possible way to use law to enable organizing to build power is members-only unionism. In the private sector and in most states, workers have a right to unionize only in a union chosen by the majority, and only the union chosen by the majority has any bargaining rights.72 This is majority unionism and exclusive representation. A union chosen by a minority of workers (what is known as a members-only or minority union) has no right to bargain.73 Much has been written about how members-only unionism could enable workers in nonunion workplaces to experience the power of collective action without waiting for a majority of workers to join the union and vote for exclusive representation.74 Andrias and Sachs’s concern that law makes it too difficult for workers to act collectively would be addressed by a law requiring management to meet and confer with a representative of a smaller group than a majority.

Members-only unionism could also be adopted to force unions to be more responsive to minority interests. This would address the concentrated power of police unions. Some assert, based on systematic and anecdotal evidence, that involving the rank-and-file is not only possible but essential to reducing or eliminating police violence75 or other abuses of the collective power of workers.

Consider what empowering minority unionism and greater public involvement in police-union negotiations might do in a situation like that of the Minneapolis Police Department in the months preceding the killing of George Floyd on May 25, 2020. Derek Chauvin, the officer who killed Floyd, had drawn twenty-two complaints or internal investigations in his nineteen years on the force,76 and two of the three officers who stood by and watched Derek Chauvin kill George Floyd were rookies in their first weeks on the job.77 One rookie was African American, and had joined the police because he wanted to help the community.78 Their prompt prosecution—and the fact that two chose to cooperate rather than adhere to the customary police code of silence—show the power of the global protests sparked by the killing. These facts also demonstrate the possibility that creating institutional supports for some officers to challenge the hegemony of a police culture might help identify problematic officers and training situations before they result in death or injury.

Law could help promote community organizations to partner with progressive or reform-minded public employees to transform public safety unions and criminal-law enforcement. With civil-rights laws granting broad immunity to municipalities and to individual officers for liability for misconduct, police departments and unions have no reason not to trade expansive protections against discipline for smaller increases in pay and benefits.79 After all, limits on discipline appear to cost the city little, because municipal and qualified immunity for the city and the officers mean the city faces reduced risk of liability for employing a bad cop.80 But empowering the public and rank-and-file officers to oversee contract negotiations, or to participate in them as a minority union, would ensure representation of the interests of the public in changing police-citizen encounters. Even if the city negotiators and the majority union agreed to exchange generous protections against discipline for smaller wage increases, public representatives or minority union representatives could disagree. Adding such voices might change the discipline system more effectively than statutory prohibitions on police unions negotiating over discipline.81

* * *

These three proposals are based on the common theme that those seeking legal protection for organizing must be attentive to the historical fact that law has been used more often to squelch unions that get too powerful than to empower those that are weak. The contemporary debate over whether law should curb the power of police or prison guards’ unions is a reminder that the public often gets antsy about the collective power of workers. As Sachs has recently argued, “[w]hen unions use the power of collective bargaining for ends that we, as a democratic society, deem unacceptable it becomes our responsibility—including the responsibility of the labor movement itself—to deny unions the ability to use collective bargaining for these purposes.”82 The problem is that working class people are rarely the ones to determine when the collective bargaining or labor protest is used for unacceptable purposes.83 When the business community decided that secondary boycotts gave labor too much power, they excoriated the Teamsters Union’s so-called “top down” organizing,84 but Congress outlawed the secondary boycott for everybody.85 The result is that janitors now get punished when they picket outside office buildings pleading for tenants to support their effort to get an abusive supervisor fired and get a pay raise.86 The International Longshore and Warehouse Union is facing an enormous damages judgment because a number of workers engaged in secondary activity to protest two job assignments at the Port of Portland.87 The farmworker and longshore worker unions both paid substantial damages when the dock workers supported the famous 1966 grape boycott by refusing to load struck grapes onto a ship.88

All three of these proposals develop the Andrias-Sachs suggestion that progressives use law to facilitate organizing, while seeking to ensure that the power the organizing creates remains accountable to the diverse workers in whose name it is exercised. Constraints on that power should come from transparency, diversity, and accountability rather than from judges who rarely in American history have had the interests of the have-nots foremost in their mind. Under capitalism, labor is rarely in a position to control how capital uses law.

Conclusion: People Make Their Own History, But Not Just As They Please

The 2020 upsurge prompts those who have waited for decades for a labor civil-rights resurgence to wonder if the moment has finally come for law to help ordinary people challenge systemic racism and the most unequal distribution of wealth the United States has ever seen. It is the moment for the Andrias-Sachs prescription for law to help build countervailing power by reviving and improving upon one of the great regulatory visions of the New Deal. They make wise use of history to argue that class-based organizing can address the concentrated power of capital and promote self-government in the economy and in politics. To that end, they advance the idea that law should facilitate the formation of self-governing civil-society organizations and empower them to meet as equals with the institutions of capitalism.

But, as Marx famously observed, people “make their own history, but they do not make it just as they please.”89 The law that promoted organizing in the past also channeled labor organizations away from radical challenges to the concentrated economic and political power of business. Andrias and Sachs are cognizant of the reasons why the New Deal legal structure and civil-society institutions perpetuated racial and gender subordination and eventually allowed a return of Gilded Age levels of economic inequality. Their proposal challenges all interested in using law to construct countervailing power to do so in a way that is more inclusive than the institutions that gained power in the New Deal.

To that end, I have explored three ways law could do more to promote some of the forms of organizing that are happening in this moment. Law could empower workers through workplace safety committees, it could support efforts to build worker-community coalitions through bargaining for the common good, and it could help enable collective action for the unrepresented or reform oligarchic unions through members-only unions. Relatively modest statutory change at the state or federal level could enable the proliferation of these three forms of worker collective action beyond the places they have already been proposed or tried. They could build worker power while encouraging diversity, racial, ethnic, and gender equity, accountability, and transparency.

Barbara Nachtrieb Armstrong Professor of Law, University of California, Berkeley. I am grateful to my colleagues Erwin Chemerinsky and Diana Reddy for their comments and contributions to the ideas essayed here and to the editors of the Yale Law Journal for the invitation to write this Essay and their thoughtful edits.