A proposal for empowering workers to help set the terms and conditions of their employment
A well-functioning capitalist system relies upon workers and employers both possessing sufficient power in the labor market to defend and advance their own interests. This is not the natural state of affairs, as economists since Adam Smith have warned. Absent worker power, policymakers must turn to redistribution to ensure widely shared prosperity and regulation to govern the workplace.
Unfortunately, America’s enterprise-level system of workplace-by-workplace collective bargaining has proved unworkable and is descending toward irrelevance. Only 6% of private-sector workers are union members and the labor movement now operates more as a political force than an economic one, which serves only to alienate workers further. Workers have seen wages stagnate despite dramatic economic growth, while both redistribution and regulation have metastasized.
Fortunately, an alternative, broad-based framework for labor organizing exists and offers real promise for restoring economically sustainable worker power. Federal reforms should create space for state and local policymakers to experiment with broad-based models in which an organization representing all workers across a region, industry, and/or occupation bargains with one representing all of the counterpart employers. The gig economy, where enterprise-level bargaining is futile, and where policymakers are actively seeking alternative approaches to regulation that preserve flexibility, offers an especially attractive starting point.
To give workers and employers greater control over their own workplaces and expand the scope of bargaining in ways attractive to both sides, federal policymakers should also permit broad-based collective bargaining agreements to depart from federal regulatory standards. Today, federal regulation serves as the floor atop which any bargaining must occur. It should instead provide the default, in effect wherever workers lack the power to bargain collectively, but accepting of departure where workers and employers on equal footing might agree on some other approach.
This paper explains the advantages of broad-based bargaining, the key parameters that policymakers must establish, and the gradual process of experimentation by which it could gain prevalence in the American economy.
The American labor movement has been in decline for decades. In 2020, collective bargaining agreements covered just 7% of private-sector workers. A lower share, 6%, were union members themselves.
Some economists and commentators theorize that an efficient labor market renders collective representation unnecessary, because the market’s self-regulating forces will ensure that individuals receive compensation commensurate with the value they create. Adam Smith did not share this view. “Upon all ordinary occasions,” he warned in The Wealth of Nations, employers “have the advantage in the dispute, and force [workmen] into a compliance with their terms.” Likewise, John Stuart Mill, a favorite of libertarians, lamented that without sufficient union strength, “the laborer in an isolated condition, unable to hold out even against a single employer … will, as a rule, find his wages kept down.”
Smith and Mill were correct. The American Compass Better Bargain Survey shows that, absent collective representation, employers largely dictate labor-market outcomes. Two-thirds of workers have not requested and achieved a significant change to their compensation, benefits, or some other term or condition of their employment in the last five years; for most, the last time this occurred was “never.” Employers surely respond over time to competitive and regulatory pressures, but markets do not move by magic, they move through the push and pull of offer and counteroffer, leverage and concession. If workers are unable to engage effectively in that process, they will not fare well.
And indeed, they have not. While GDP per capita rose 92% from 1979 to 2019 and corporate profits per capita rose 77%, wages for nonsupervisory workers rose only 9%, according to the Bureau of Labor Statistics. Over the same period, average compensation for CEOs at the 350 largest, publicly traded U.S. firms went from being roughly 30 times larger than the average worker’s to being 300 times larger. Nor has the rise in national wealth and productivity translated for the typical worker into the sort of secure employment that policymakers and their staffs take for granted. The American Compass Better Bargains Survey finds that less than one-third of nonsupervisory workers in July 2021 held secure jobs, defined as annual income of $40,000 or more, predictable earnings, steady hours, and health benefits. Among workers without college degrees, that figure falls to one-in-five.
The protections that workers do have, and the income gains they have achieved, are in part a function of government programs of regulation and redistribution. Employment law provides for the 40-hour work week, the minimum wage, overtime pay, paid time off, safety standards, and myriad other protections. Minimum wage increases have produced above-average wage growth for low-wage workers, which otherwise would have lagged. Analysts include transfer payments in their calculations of household income when making the case that the economy is spreading prosperity widely.
This should satisfy no one. “We must not, we cannot, depend upon legislative enforcements,” said Samuel Gompers, the first and longest-serving president of the American Federation of Labor. “When once we encourage such a system, it is equivalent to admitting our incompetency for self-government and our inability to seek better conditions.” Workers deprived of power in the labor market lose their agency and become beholden to bargains struck by political actors for whom many other interests may take precedence. And while labor activism is often associated with the left, the right too should prefer self-constituted institutions advancing their own interests in the market over government programs of redistribution and regulation. Unions, observed conservative sociologist Robert Nisbet in The Quest for Community, are “the true supports of economic freedom.” Conversely, as unions focus on political and legislative outcomes rather than economic and negotiated ones, they further alienate their erstwhile constituents—who say by enormous margins that they don’t want worker organizations involved in politics. The end point of this vicious cycle is a labor movement asymptotically approaching irrelevance, workers alienated from organizations that might provide them solidarity and voice, and businesses faced with constantly multiplying diktats from Washington.
The failure of organized labor as an economic institution and a source of worker power is an inevitable consequence of America’s enterprise-level system, in which unionization is a fight waged workplace by workplace, bargaining only occurs after a successful vote, and employers who reach agreements with unions find themselves at a competitive disadvantage. American labor law should shift toward a broad-based model, common in many parts of the world, where unions represent workers across entire regions, industries, and occupations and negotiate with comparably broad trade associations representing employers. This model has been proven to yield economic arrangements both attractive for workers and sustainable for employers, while providing a superior alternative to top-down regulation.
Reforms in this direction can and should occur gradually, with experimentation by states and localities, in particular industries and occupations, focused on specific terms and conditions of employment. Alongside changes in labor law that permit and encourage broad-based bargaining, the federal government should reform employment law to allow bargains struck this way to supersede bureaucratically established defaults. If successful, the model would scale naturally toward a new structure for organized labor compatible also with the proposals described elsewhere in this collection.
Part I: Broad-Based Bargaining
Americans take for granted the idea of a union as something that workers organize within their place of business, to deal with their particular employer; Sally Field as Norma Rae, standing atop her table and turning silently with her hand-written sign that reads “UNION.” Globally, that model is an aberration, with good reason. One problem with this business- or enterprise-level system is that union membership depends entirely on place of employment. Want to be a union member? No luck, unless you get half of your coworkers to vote yes.
This yields what is commonly called a “representation gap,” where a substantial share of workers desires union representation, but not enough that the share in a given workplace will typically reach the 51% needed to unionize. The share that wants a union will be much lower than the share that has one. For instance, in the Better Bargain Survey, 35% of nonunion workers say they would vote for a union. That is likely to yield far less than 35% of workers ultimately employed at workplaces with successful union drives—and of course, those who do end up unionized will only partly overlap with those who say they want to be unionized.
Another problem with enterprise-level bargaining is that it yields fragmented arrangements within markets. Management at each company rationally fears that dealing with a union will put it at a competitive disadvantage against other, non-union companies. This leads to combative unionization campaigns and adversarial relations where unions do gain a foothold.
The alternative model, much more common in Europe, is called sectoral or broad-based bargaining. In this model, unions represent workers of common interest regardless of their place of employment—generally across an entire occupation, industry, or region. These unions bargain with trade groups representing all of the relevant employers, reaching agreements that apply broadly. In many respects, these agreements are closer to regulations governing industries than to contracts governing workplaces; but unlike regulations, they are chosen by the participants themselves rather than by bureaucrats. This model has several important advantages:
- Worker Power. Workplaces have to be governed and terms and conditions of employment set. The question is who should play what role in that process. One option would be to leave employers in control and allow individual workers to attempt negotiations—in practice, this leaves workers with little voice in decision-making and little capacity to demand change. Another option is to rely on government to establish rules. In this case, workers can attempt to exert influence through the political process, but that influence will be greatly diluted by the myriad other issues unrelated to the workplace that policymakers must address. Workers are likely to disagree on many of those issues, precluding them from operating as an effective bloc when it comes to the economic issues on which their interests most likely align.
Guaranteeing an environment in which workers are empowered to bargain collectively—and thus on equal footing—with employers is a much better approach. The parties most affected are the ones who will make the decisions, and they can do so with consideration of both their own immediate interests and the long-term interests of both sides in what is ultimately a symbiotic relationship. Bargains struck in this way are likely to prioritize those things which are of highest value to each side, making tradeoffs that are mutually beneficial and thus create economic value in the process. Workers will achieve better outcomes than they could individually, while employers will not perceive every concession as potentially fatal if they are all making it together.
- Basis of Competition. By encompassing all competitors in negotiations, broad-based agreements take some areas of potential differentiation off the table and thus channel competitive energies in other directions. If labor relations are standardized, no one can seek to outperform everyone else by squeezing workers harder on wages or skimping on safety, potentially triggering a “race to the bottom.” Conversely, investing in productivity gains, innovation, customer retention, and so forth becomes that much more important.
- Flexibility. Independent contractors and “gig workers” are easily covered by broad-based agreements, whereas the 1930s-style “vote at the worksite” mode of unionization is plainly inapposite. Likewise, structural obstacles posed by franchises, joint employers, fissured workplaces, and so on become resolvable—workers can be covered regardless of who signs their paycheck. Employers are covered based on the activities they engage in, not how they’ve structured their ownership and whether they’ve beaten back an organizing campaign recently.
- Institution Building. Unions can do many things for workers beyond collective bargaining, but in an enterprise-level system they cannot be relied upon because they will often be inaccessible. Whether a worker can benefit from a union depends on a workplace vote, rather than the worker’s own preference. In a broad-based system, unions operate as institutions of civil society outside the confines of any particular workplace. A union is something a worker chooses to join because he wants to have his voice heard and because the union provides something that he values. Such unions can operate as a quasi-public provider of benefits, for instance offering group health insurance and administering unemployment benefits, or facilitating engagement with schools and running training programs. Likewise, in a broad-based system, unions are reliably present in communities, contributing to the social fabric, promoting solidarity, and helping workers support each other.
America will not suddenly and smoothly transition to such a system. The sheer scale of legal infrastructure governing the modern labor market makes overhaul in one fell swoop, with some 5,000-page piece of legislation, wildly implausible. Even if technically feasible, the everyday arrangements of one-hundred-million workers and their employers would not adjust so quickly. As importantly, a well-functioning system is not merely or even primarily a matter of law. It requires institutional capacity, establishment of norms and obligations, and trust between the parties. It must evolve and grow over time.
A Plausible Transition
Fortunately, broad-based bargaining is well suited to experimentation and gradual implementation. While “broad” as compared to enterprise-level bargaining, it confines itself along dimensions like occupation, region, and contractual term.
For instance, a particular city or state could choose to experiment with a broad-based regime for a particular industry—say, janitors in New York City or home health aides in Ohio. The jurisdiction could likewise specify the scope of bargaining. For instance, the minimum wage might be an ideal starting point. Minimum-wage laws frequently operate at the state and local levels, and some states already have wage boards. States could build industry-specific processes, or the federal government could establish a framework for states to operate within.
The gig economy, for which governments are actively seeking viable regulatory frameworks, provides another starting point. Traditional unions make little sense in this context and efforts to reclassify workers as standard employees have fared poorly and often conflict with what the workers themselves want. Cities, states, or the federal government could designate representatives for a class of gig workers—say, drivers—and then mandate that firms operating relevant platforms join together to negotiate with those representatives.
Federal policymakers can play a direct role in structuring and spurring initial pilots but, generally speaking, it is state and local leaders that must lead on these efforts if they are to succeed, with federal government’s primary role being to ensure that its own laws do not interfere. It is local leaders in government, the business community, and among workers who must build the necessary institutions and relationships and then arrive at and implement agreements proving that they can improve upon what top-down regulation achieves. For governments, the task is to set the rules and designate the parties who will participate; the parties themselves must prove themselves honest brokers, earn the trust of their constituents and each other, and then do the work of bargaining.
Policy Design Considerations
Broad-based bargaining renders irrelevant many of the major issues often associated with labor law today but introduces an entirely new set. Gone are concerns about ground rules for organizing campaigns, litigation of unfair labor practices, and administration of elections. Instead, questions arise of how to determine which organization represents workers and how to ensure workers have the option to participate. Gone too are issues of “closed” versus “open” shops and “right-to-work” laws—no one has to join a union. Instead, questions arise about the bargained terms and union services that are or are not available to non-members.
A framework for broad-based bargaining should include:
- Legitimate representatives. Policymakers must establish a basis for selecting the organizations that will bargain on behalf of workers and employers. Any number of organizations, including trade associations, existing unions, alternative worker organizations, and newly constituted groups, can play the necessary role. Most important are the rules governing who can participate and who makes the decisions. All participants—whether workers or employers—who would be governed by an eventual agreement must be eligible to join on equivalent terms and all who choose to join must be given proportional voice in the selection of leaders. Workers will generally all have equal standing, though, in a context like the gig economy, workers might also have more or less influence based on their hours worked or income earned.
- Scope of bargaining. Policymakers must establish those terms and conditions of employment that an agreement might cover. These should be some subset of those traditionally addressed by collective bargaining agreements, which include wages and hours, benefits, holidays and leave, grievance and discipline procedures, and health and safety standards. Some traditional subjects are inappropriate to bargain industry-wide, such as processes employed in particular facilities; for this reason, local mechanisms of workplace governance like works councils are a valuable complement to broad-based regimes (see A Better Bargain: Worker Voice and Representation, by Chris Griswold). Conversely, broad-based organizations are better positioned to provide benefits to workers. For instance, in a forthcoming Better Bargain paper on Ghent-style unions, Wells King describes how a broad-based worker organization might accept public and employer funds and take responsibility for providing health or unemployment benefits or administering training programs.
- Concerted action. Workers must have the ability to act collectively, so that they have some manner of leverage in the bargaining process. The rights of workers protected under the National Labor Relations Act should be extended to those whose representatives are engaged in broad-based bargaining. In some respects, those rights would give them substantial power—an industry-wide strike sounds even more disruptive than one at a particular company—but in other respects the power would be less than that held by enterprise-level unions. Whereas an organized workplace will generally see all workers strike together, only those workers choosing membership in a broad-based organization might participate in an industry-wide strike. And whereas one business might face catastrophe if its workers strike while competitors continue with business as usual, businesses at least find themselves on comparable footing if all face the same labor action simultaneously. A sensible provision employed in some broad-based bargaining frameworks eliminates the right to strike while an agreement is in effect, creating strong pressure to reach a new agreement before the old one expires.
- Right of communication. To facilitate a worker organization’s selection of representatives, feedback on policy issues, and collective action, workers across employers must have an ability to communicate with each other through the representatives. Communication amongst workers who choose to join the organization will be straightforward, but representatives must also have occasional opportunity to contact all eligible workers, update them on relevant developments, and invite their participation.
- Limitations on political action. Organizations representing workers and employers must forego direct involvement in political processes, a constraint similar to that faced by 501(c)(3) nonprofit organizations. One reason for this requirement is that these worker organizations will depend upon a public designation as bargaining agents and hold quasi-public power to reach agreements with broad effect. They should not use that power to influence elections of the officials who grant them that power. A second reason is that these organizations may receive public funding to carry out bargaining duties and administer public programs. Such funds should not be intermingled with ones used for political expenditures. Finally, broad-based bargaining requires the worker organization to retain the trust and represent the shared economic interests of diverse constituents with divergent political views. It cannot do this if engaged simultaneously in partisan political maneuvering.
- Federal flexibility. Without dictating the specifics of broad-based bargaining regimes, federal policymakers will need to adopt several reforms to create the space for its emergence. For instance, provisions of the National Labor Relations Act that designate its type of union as the only valid form of organizing must be scrapped or at least waived. Likewise, employers, and perhaps workers as well, will require antitrust exemptions permitting them to coordinate and use their collective power in ways competition law generally seeks to prevent. And then, to increase the potential for bargains that benefit both parties (the prerequisite for any durable agreement), employment law must itself become more flexible. That flexibility is the subject of Part II.
Part II: Departing from Regulatory Defaults
The theoretical justification for employment law, and the practical reason for its rapid growth in recent decades, is that individual workers lack sufficient leverage in the labor market to effectively defend their own interests. These concerns do not apply where workers are organized and can bargain through their representatives on equal footing with employers. America’s hybrid system of labor and employment law fails to account for this distinction, instead taking employment laws designed to protect unorganized workers and applying them just as stringently where organizing has occurred. In other words, employment law has become a floor atop which organized workers can seek additional protections and benefits. But more sensibly, it should be a default that operates for unorganized workers, but which organized workers and their employers can depart from as they see fit.
This option for departure should not apply everywhere. For instance, anti-discrimination law is premised in part on the need to protect minority groups from harms that the majority may not face or may even be causing. A majority of workers should not be able to bargain that away. Likewise, some areas of law that affect the workplace are intended to protect interests besides those of workers. An environmental regulation may affect how workers do their jobs, but it is often there to protect people outside the workplace from byproducts of the work. Both businesses and workers might gladly waive the requirement. For that matter, it might very well be a foolish requirement that should be removed. But broad-based bargaining would not include the relevant parties in the decision; the issue must be left to the political process.
These exceptions still leave plenty of rules: the wages and overtime provisions of the Fair Labor Standards Act, the paid- and unpaid-time-off provisions of the Family and Medical Leave Act, many of the mandates in the Occupational Safety and Health Act (OSHA), various mechanisms for delivering publicly funded training programs, unemployment insurance and workers’ compensation systems. Retailers and retail workers could agree that workers would always be provided at least two weeks’ notice of their schedules, but also that overtime would pay less than time-and-a-half. Machine technicians could agree to waive reams of safety rules wherever works councils exist and give workers on the shopfloor sufficient input into process and equipment changes.
In a law review article published long before his tenure as Secretary of Labor, Eugene Scalia argued for this harmonization of labor and employment law, using the example of OSHA’s inspection regime. “Workplaces where employees are empowered to address safety and health conditions are less in need of government scrutiny,” he noted, and unionized employees “could still summon OSHA inspectors if the company failed to address safety problems the union identified.” While a worker organization may lack the expertise to set “permissible exposure limits for toxic substances,” it could always incorporate OSHA standards by reference while still asserting a greater role in enforcement.
Some worker advocates worry that permitting departure from default regulations could leave workers worse off than they started. But this misunderstands the nature of bargaining, which can succeed only when each side feels it is obtaining something it wants and could not otherwise get. A fatal flaw in the current American system is that it accords no such opportunity to employers, who consider themselves better off in all respects if they do not have to negotiate at all. Placing on the table onerous and poorly tailored regulations, which have been designed to protect workers who cannot take an active role in representing themselves, has the prospect of making collective bargaining something that employers want to do. Every rule that has less value to workers than it has costs for employers, meanwhile, represents an opportunity for a mutually beneficial exchange. And if workers feel they would do better under the regulatory default, their representatives will always have that option.
A more salient objection is that the opportunity to depart from regulatory defaults will set off a “race to the bottom” in which employers threaten to decamp to whichever state’s workers will offer the best deal. Some risk of this does exist, but it is mitigated in two respects. First, most service-sector jobs are quite immobile. As in the examples described above, one cannot move janitors, home health aides, or ride-share drivers from one state to the next. They have to work where the buildings, recipients of care, and riders are. Second, employers would need somewhere to go. During an initial period of experimentation, their only alternative would be jurisdictions still operating under the default. If broad-based bargaining were to become widespread, employers would not be appealing to legislators to cut them a deal but to workers, themselves organized and able to exert power and defend their interests.
Empirically, rather than triggering a race to the bottom, broad-based bargaining triggers a virtuous cycle by creating actual value for the parties. Employers benefit from the bargains struck and relationships deepened, and workers are able to extract some share of that value for themselves. Workers, for their part, find it in their own interest to ensure the competitiveness and success of employers, believing that they will share in the upside. In “Workers of the World,” Wells King explains:
Broad-based bargaining improves the performance of unionized industries and firms along a number of different dimensions. It has been shown to reduce employee turnover and to establish better, and also more flexible, safety standards for particular industries. By including all employers within a given industry, it creates new incentives and collaborative forums for worker training; industries covered by sector-level agreements are more likely to invest in workforce development and devote greater resources to firm-sponsored training.
Broad-based bargaining’s benefits also spill over into the broader economy, improving both labor-market and social outcomes. It increases national employment by both reducing unemployment and increasing labor force participation, and also boosts productivity rates for covered industries. Meanwhile, it compresses wage distributions across entire industries, much as enterprise-based bargaining does within unionized firms, reducing economic inequality.
In countries with broad-based bargaining—particularly those where agreements are national in scope—unions are responsive to macroeconomic issues like wage-driven inflation and international competitiveness. They tend to strike a balance that accepts relatively lower wages but promotes healthier firms and rising productivity, which supports higher wage growth in the long run. In Germany, for instance, trade unions have agreed to set wages below marginal productivity in order to increase the competitiveness of export sectors.
Someone must regulate the workplace, and as between the regulator doing so from the capital, the employer doing so unilaterally, or workers and employers doing so together, this third option strikes the best balance of ensuring that both sides have their interests protected in ways that promote the success of their shared enterprise. “Those in government who profess confidence in labor unions need to consider how federal employment law could better reflect that faith,” Scalia concluded. “And those in government who advocate less federal regulation, and more local control, need to consider how unions could help bring that about.”
Policy Design Considerations
Federal policymakers will need to decide which regulations a collective bargaining agreement can depart from, and under what circumstances this can occur.
- Potential departures. In conjunction with permitting broad-based bargaining as described above, Congress should provide for departure from large areas of federal employment law by collective bargaining agreements reached through that arrangement. Existing law would in all cases remain in force by default, but an agreement that explicitly chose to depart in some way would be permitted to do so. As a general rule, if current labor law considers a topic to be a bargainable term or condition of employment, employment law should recognize that parties can bargain around its default. By contrast, regulations that unions do not bargain over today—discrimination, environmental protection, taxation, etc.—are likewise inappropriate for departure. Congress could begin by amending a concrete set of statutes and regulations to specify that collective bargaining agreements are not bound by them. Alternatively, Congress could define a process by which the Department of Labor expands the scope of potential departures on a case-by-case basis as parties to bargaining identify provisions that they would like to address. Once an exemption has been established, it would be universally available for bargaining.
- Approval process. The goal should be for bargaining to occur by right, not by application for approval to a federal agency. A categorical exception established by Congress is therefore the ideal, but may be too aggressive as a starting point. As an intermediate step, the Department of Labor (DOL) could be authorized to review bargaining arrangements and approve exceptions. Where this occurs, it should happen ex ante, before bargaining begins. Developing an agreement and then asking DOL to approve it ex post would distort the bargaining process and leave anyone unhappy with the agreement lobbying for its alteration. Collective bargaining agreements should not require federal approval to take effect. Likewise, once some provision of federal law has become “bargainable,” it should be available in all cases. By a gradual process of accumulation, governance of workplaces would shift from regulators to bargained agreements to the extent those agreements prove successful.
- Localization. An important feature of broad-based bargaining, when coupled with workplace governance mechanisms like works councils, is the ability to exert worker power at a sectoral level but then delegate decision-making to the local level. Thus, for instance, a broad-based collective bargaining agreement might establish the requirement that workers review and approve the safeguards implemented with the introduction of any new equipment, without attempting to codify safeguards directly. This process should apply to departures from federal regulation as well, for instance a collective bargaining agreement might establish the programs of paid time off and family leave while allowing local agreements between workers and management to set eligibility criteria.
A well-functioning capitalist system is not one in which capitalists generate for themselves the greatest possible profit; it is one in which capitalists, in their pursuit of the greatest possible profit, take actions that generate widespread prosperity for the whole nation. Equalizing the power of workers and employers in the labor market is one of the most powerful mechanisms that policymakers have at their disposal for achieving this latter condition. Otherwise, they are left to impose by regulation and redistribution what the market will fail to deliver. Tight labor markets accord workers significant power, and Americans have seen firsthand in recent years the dramatic, positive effects this can have on outcomes in the labor market. But business cycles boom and bust, and workers have many interests that they cannot meet through individual negotiations that they are unlikely to undertake anyway.
Workers need access to collective action, but America’s dysfunctional, enterprise-level system no longer affords it. The journey to a broad-based model is long, but innovative policymakers should recognize its promise and take the first steps.