A Better Bargain: Worker Solidarity and Mutual Support

A proposal for channeling labor resources away from politics and toward worker benefits and training

Download PDF

Executive Summary

Workers in the modern labor market have a common interest in the security and flexibility of basic supports. They depend on benefits and services from unemployment benefits to health insurance to training to navigate inevitable market frictions and increase their value and thus their wages.

Americans take for granted that some combination of government programs and employers must provide these supports. But this system works poorly, and deprives workers of the responsibility and opportunity to build institutions of their own.

An approach that gave worker-led organizations a leading role could be more flexible to local conditions and more responsive to changing needs, while also advancing solidarity and self-determination among workers. Existing models in Europe and parts of the United States demonstrate the promise of quasi-public labor organizations delivering public programs like unemployment insurance, portable benefits like health insurance, and sector-wide worker training.

Straightforward federal reforms would enable state and local governments to partner with new labor organizations in administering such programs. The organizations would be funded not only by member dues but also by public funding and employer contributions. Their activities would be explicitly apolitical and focused solely on meeting workers’ direct needs. Reform along these lines is overwhelmingly popular across political parties and economic classes.

This paper explores the promise of worker-led organizations in the provision of benefits and services; outlines the basic parameters that policymakers must establish for labor organizations; and suggests additional considerations for ensuring that organizations best serve workers’ interests and the common good.


In The Quest for Community, the sociologist Robert Nisbet described the importance of associations in civil society. They secured a domain for individual freedom, supplied a means for collective self-sufficiency, and served as a countervailing force against impersonal, often bureaucratic institutions. In their absence, our common life risked crumbling into an “atomized mass” in which individuals possessed “neither the will, nor the incentive, nor the ability to combat […] collectivism.”

The same was true in the labor market. “Associational realities of the labor union,” according to Nisbet, supplied the “real defenses […] of economic freedom.” They were grounded in self-organization, self-determination, and self-governance that fostered local allegiance and independence rather than passive receipt of professional services or submission to centralized power. Understood in this way, the decline of organized labor represents not merely a loss of power to secure a fair deal from employers, but also the loss of solidarity that channels workers’ common interests into institutional supports from within and beyond the labor market.

The decline of organized labor represents not merely a loss of power to secure a fair deal from employers, but also the loss of solidarity that channels workers’ common interests into institutional supports from within and beyond the labor market.

While we think of collective bargaining as “what unions do,” that is not all they are meant to do. Nor is it all that workers wish they would do. The American Compass Better Bargain Survey asked workers how they would allocate a labor organization’s resources to different activities. They give just one-fourth to collective bargaining. They ranked training and benefits a close second. Nearly 60% of workers who say they would vote to unionize their workplace cite training and benefits as a reason for doing so.

Labor unions should be uniquely positioned and equipped to support workers in this way—after all, their origins lie in the guilds and mutual-aid societies of old. Indeed, the National Labor Relations Act (NLRA) underscores this function and defines the rights of employees not only “to bargain collectively through representatives of their own choosing,” but also “to engage in other concerted activities for the purpose of collective bargaining or other mutual aid or protection.”

Unfortunately, the NLRA also narrowly prescribes how workers may engage in such activities. In the United States, workers organize workplace by workplace, so membership depends principally on where they work. They must win a majority vote for the union in their workplace before they can enjoy its benefits and services. An individual worker may opt out of union membership in right-to-work states, but nowhere can he opt in on his own. While this structure may be necessary for collective bargaining in the absence of a broad-based bargaining regime (see: A Better Bargain: Worker Power in the Labor Market, by Oren Cass), it can actively obstruct solidarity and mutual support.

Workers suffer as a result. Instead of turning to their fellow workers for support, they depend on their employer (if they are traditional employees) and the government. The hodgepodge of private benefits and public programs helps workers manage market frictions and increase their value in the labor market, but it offers them little input and assigns them little responsibility. This only makes unions less relevant, and encourages them to turn toward political action in pursuit of policy outcomes that increase the generosity of public programs and mandate more from employers.

And so long as national political agendas inevitably divide workers, a worker organization focused on advancing any given agenda will be a source of division rather than solidarity.

In practice, this political agenda has ballooned into a laundry list of generic progressive priorities, which has only further alienated workers from the labor movement. Unions’ political involvement is the most commonly cited reason workers give for opposing unionization in their workplaces, and it is the activity to which they would like to see unions allocate the least attention. The emphasis on politics also changes the very character of association. Even stipulating that politics is effective at securing economic gains, the approach undermines the role of the union worker himself—from an active member working in solidarity to an essentially passive beneficiary of political advocacy and public benefits. As AFL founder Samuel Gompers warned in 1912, “[The labor movement] must not, we cannot, depend upon legislative enforcements. When once we encourage such a system, it is equivalent to admitting our incompetency for self-government and our inability to seek better conditions.” And so long as national political agendas inevitably divide workers, a worker organization focused on advancing any given agenda will be a source of division rather than solidarity.

An organization of, by, and for workers that provides benefits and services complementary to employment would be in many respects the antithesis of the standard American union: It would operate independent of particular employers, not workplace by workplace. It would be open and voluntary, not dependent on an up-or-down vote. It would be excluded from politics. By virtue of all these features, such organizations would focus relentlessly on meeting workers’ needs more effectively than employer- and government-run programs. The benefits they provide would remain available to a worker through periods of employment, unemployment, retraining, and contract work, affording a combination of flexibility and security unattainable today.

This model is within reach. Straightforward reforms to federal and state policy could spur the emergence of worker-led organizations that forego politics and instead channel public, employer, and worker resources toward providing workers the support they need.

Part I: The Provider of First Resort

Workers navigating the modern labor market depend on a range benefits and services to manage inevitable market frictions, increase their productivity, and provide for their families. Most Americans take for granted that these benefits and services are the responsibility of government or employers. But in parts of Europe as well as in some U.S. industries, labor organizations have adopted a leading role. Not only are their approaches more responsive to workers’ needs and more effective at meeting them, they offer a value proposition to workers beyond the traditional benefits of representation and bargaining coverage. Policymakers should encourage the expansion of this model.

Unemployment Insurance

In European countries with the highest and most stable levels of unionization, unemployment insurance is voluntary and administered through unions in an arrangement known as the “Ghent system.” The popular system draws on state funding and employer contributions to provide generous benefits to workers for modest fees. Even though workers have the choice to receive insurance either through a government agency or a union, most choose a union because its services are more readily accessible and easier to navigate. This creates effective incentives for workers to join and remain members of unions. Administering unemployment insurance is estimated to unionize an additional 20% of the workforce and rates of unionization where the Ghent model exists are among the highest in the world.

In stark contrast, the American unemployment system finds itself on life support. The state-administered program has been chronically under-resourced and its systems are outdated. Some offices process claims using Excel spreadsheets; others rely on an archaic programming language, COBOL, that dates back to the 1950s. The federal government offers little guidance or support (about 40 Department of Labor employees oversaw a system that disbursed $27 billion in benefits and managed more than 11 million claims in 2019). Administrators face threats of funding cuts and pressure to target fraud even though the share of unemployed workers receiving the benefit has been steadily declining for decades. The COVID-19 pandemic further exposed the system’s weaknesses as offices struggled to implement new eligibility requirements and unemployed workers struggled with onerous processes.

Labor unions already demonstrate competence in helping members navigate bureaucracy and receive public benefits. Union members are more likely to receive workers’ compensation and 23% more likely to receive unemployment insurance than comparable nonunion workers. A Ghent system would build upon these existing competencies while making them more widely available. Whereas the existing unemployment system struggles to reach more than 30% of unemployed workers, Ghent unions achieve far higher coverage rates, as high as 85%, while reducing public administrative burdens and operational costs.

Portable Benefits

The livelihoods of workers and their families often depend on certain benefits—health insurance, retirement accounts and pensions, paid leave, and more. But the American benefits system relies upon traditional employer arrangements at the expense of workers’ needs and economic interests. The best example is also the most consequential: health insurance.

As with much of the nation’s program of social insurance, our health benefits system was built around a legal category of “employees,” where contributions to employer-sponsored plans are tax-free. Still, only about half of private-sector workers receive health insurance from their employers, and in the Better Bargain Survey, only 37% said that their employer would be their preferred source of health insurance coverage.

The most obvious flaw in an employer-based system of health insurance is that it excludes those without traditional employers: independent contractors, freelancers, temps, and gig workers. Worse, the cost of providing health benefits gives employers a strong incentive to avoid taking on employees, instead holding labor in these arms-length, independent-contractor relationships. This has accelerated the transition to “fissured” workplaces, in which businesses offload the employment relationship to subcontractors and other third-parties or else misclassify their employees as independent contractors. Alternative arrangements have been steadily normalized: Between 2005 and 2015, the United States added nearly 9 million such jobs, while there was virtually no job growth for traditional employment. Workers are left without health insurance, and also without the many other benefits of stability and security, legal protection, and employer investment that accompany traditional employment.

Less obvious, but also costly, is the effect of the employer-based-insurance regime on traditional employees. They do have health insurance. But they have no say over which insurance plans they can access, and they find themselves handcuffed to their particular employer lest their coverage lapse. If they lose their job, they also lose their insurance right when they can least afford to. Changing jobs poses a risk too: What insurance will the new job have, on what network? Employer-sponsored health plans have been found to reduce job-switching by 15–25%, discouraging employees from pursuing opportunities that may improve their lives and economic fortunes—finding a new employer, moving to a new location, seeking training and education, and more. In few contexts would Americans accept such a bizarre constraint on their access to such an essential service, yet in health care it is now taken for granted.

Policymakers attempt to address the challenge of disruptions in health coverage with a patchwork of different programs, each tailored to different populations: Medicaid, Medicare, COBRA, the Affordable Care Act’s exchanges, and so on. But this compounds the disruption—from one employer on to a public program, then from that program back to a new employer—and introduces another layer of bureaucracy unaccountable to workers as customers or members. Market-oriented reform proposals attempt to grant each household freedom to seek its own coverage a la carte. But the importance of risk-pooling and the problems of asymmetric information and moral hazard make health insurance a difficult market for this approach.

In few contexts would Americans accept such a bizarre constraint on their access to such an essential service, yet in health care it is now taken for granted.

Labor organizations could hold the key to a better model. Operated by and for workers, they could offer fully portable benefits, conditioning them on membership rather than employment status. Organizations could receive contributions from employers, public funding, and member dues. For work in non-traditional arrangements, organizations could collect pro-rated contributions from employers proportional to time worked or income earned.

Some workers already receive benefits in this manner. For instance, the Culinary Health Fund, run jointly by the Bartenders and Culinary Workers Unions in Las Vegas, provides portable benefits to its members, funded both by employer contributions and member dues. Similarly, New York’s Black Car Fund raises money from small surcharges on black car and limo fares to fund a range of different benefits—including workers compensation, vision, telemedicine, and death benefits—to drivers, who tend to be independent contractors.

Workforce Training

Both workers and employers stand to benefit when workers develop skills that increase their productivity. But this potential win-win also poses a difficult question, one of the oldest in labor economics, of who will make the necessary investment.

Workers may wish to invest in their own training but, especially early in their careers, they may not have the resources to do so. They also lack knowledge of what training to pursue and opportunities for on-the-job experience. Employers will invest in training to the extent that they can capture the returns. But this means getting the higher productivity without paying a higher wage, in which case the desired outcome for the worker is not realized. Employers then face the risk that workers will demand higher pay to match their skills or else find employment elsewhere, and respond by trying to limit workers’ options.

The best way to square this circle is for workers to invest collectively in their training.

Several working models of this approach already exist, with labor organizations taking the lead. In the U.S. construction industry, for instance, programs jointly sponsored by unions and employers have higher completion rates than those sponsored by employers alone. In the United Kingdom, union members receive more and better training than nonunion workers, while in Germany unions tend to increase training in existing apprenticeship programs.

For labor organizations eager to boost their own numbers and their value as partners for employers, training offers a unique opportunity—not only to help onboard workers as they enter relevant occupations, but also to retrain workers further along in their careers.

The German model, in particular, deserves attention for its integration of labor organizations into workforce development. Germany runs a robust vocational education and training system, including an apprenticeship program in which students seeking certification earn money and learn skills on the job. The national standards for such programs are established jointly by tripartite committees of experts from government, industry, and trade unions. In the course of collective bargaining (see A Better Bargain: Worker Power in the Labor Market, by Oren Cass), trade unions negotiate the salaries of apprentices. The administration of worker training is also handled by worker organizations. For instance, the examination bodies that oversee the creation and administration of tests for certification include worker representatives nominated by workers. Works councils (see A Better Bargain: Worker Voice and Representation, by Chris Griswold), which include worker representatives, oversee the local administration of vocational training programs at individual worksites.

A fully integrated, German-style model may not be achievable in the near- to mid-term. But empirically, involving labor organizations in the design and delivery of training for an entire sector is just a better way to train workers. Unions already improve and complement worker training programs in measurable ways. Union involvement in worker training has been shown to produce better outcomes in the workplace and sector-wide training programs have proved to be more effective than employer-specific training models.

For labor organizations eager to boost their own numbers and their value as partners for employers, training offers a unique opportunity—not only to help onboard workers as they enter relevant occupations, but also to retrain workers further along in their careers. In Las Vegas, for instance, the Bartenders and Culinary Workers Unions partner with major properties to provide not only training, but also a “hiring hall” to ensure that graduates are hired. They do both pre-employment training for workers entering the hospitality industry as well as advanced skills training for union members. Training is free for union members, creating a material incentive to join.

Policy Design Considerations

By definition, worker-led organizations must be led by workers, and so will emerge gradually over time. Indeed, some already exist. The role for government, then, is not to create them but rather to make their creation more attractive, both by removing obstacles to their formation and operation, and by creating interfaces with existing public programs that allow for delegation of services and distribution of funds.

Organization Structure. Policymakers must define the parameters of organizations eligible to fill the quasi-public role of administering subsidized training and benefits programs, by creating space within federal labor law for nonunion labor organizations to form, partner with employers, receive funding, and provide services. The NLRA, which has near-exclusive jurisdiction over the relationship between employers and organizations of workers, should be amended to govern only unions created through the NLRA’s process. Its prohibition against nonunion employer-worker collaboration, as set out in Section 8(a)(2), should be removed.

Next, policymakers should formally recognize a new form of labor organization that is controlled by dues-paying members and operated for the purpose of providing benefits and services to them. Policymakers will need to set the parameters within which an organization could limit its membership—who can join, when, on what terms. They should also establish governance standards appropriate to the organizations’ role in partnering with employers and government to deliver benefits and services while retaining worker control.

Ideally, these labor organizations should take a form that could also engage in broad-based bargaining (see A Better Bargain: Worker Power in the Labor Market, by Oren Cass), and policymakers should establish a recognition process for designating an organization as a bargaining representative.

Benefits and Services. Federal policymakers should identify public programs that labor organizations could deliver—including unemployment insurance, workers compensation, health benefits, workforce training, trade adjustment assistance, and more—and permit the state and local governments that are typically responsible for administering the programs and disbursing the funds to do so through the organizations. Unemployment benefits, for example, are administered by state governments, which already have broad discretion over their programs. Nevertheless, the Department of Labor would need to approve labor organizations as administrators before state governments could involve them.

Labor organizations also have the potential to experiment with programs adapted to local conditions and workers’ needs. Policymakers should encourage this by offering waivers from the standard requirements imposed on funds, where a suitable alternative is proposed. For instance, a labor organization might modify unemployment insurance to operate more as wage insurance for workers forced to move to a lower-paying job.

Funding. While labor organizations would receive contributions from members, they would also receive taxpayer money for the government-funded or government-subsidized programs that they administer. Additionally, policymakers should recognize the value of partnerships between labor organizations and employers and should permit employer contributions to the organizations to fulfill legal requirements associated with providing benefits like health insurance or retirement plans.

With these funding sources would also come constraints. As labor organizations would be entrusted with public money and serve as proxy administrators of public and quasi-public programs, they should be precluded from direct political activities, just as other such entities like nonprofit organizations are today.

Part II: Unbundling Organized Labor

Organized labor has long engaged in politics. In fact, political action committees (PACs) are themselves a union creation. The CIO formed the first one in 1944 to support the (third) re-election campaign of FDR. But contrary to claims of union activists, politicking is not essential to a labor organization. As Harvard Law School’s Benjamin Sachs has noted, “There is nothing in the nature of unionization that requires the bundling of economic and political functions.” Rather, it’s “an artifact of history and, more to the point, of law.”

For labor organizations that could accept employer and public funds to provide benefits and services, removing that artifact of law and “unbundling” the worker-focused, economic function from the political function would be desirable for three reasons.

First, labor organizations would function as quasi-public entities administering public programs subject to constraints set forth by policymakers. To avoid an obvious conflict of interest, labor organizations should not be permitted to influence how policymakers operate or oversee such programs.

Second, labor organizations would receive funding from different sources—including member dues, employer contributions, and public funds—to administer these public programs and other services. It would be inappropriate for organizations to divert any of these fungible funding streams to political spending. Of course, a labor organization and its members would still be permitted to set up a separate PAC and raise dedicated money for political operations.

Third, part of the public purpose for recognizing and supporting these labor organizations is to create the opportunity for workers to advance their common interests and build solidarity. Political activism plainly conflicts with this purpose and with what workers say they want. When the Better Bargain Survey presented workers with a list of nearly 20 different political issues that unions are working on today, and asked which they would want to see a worker organization speaking out on, not a single one received majority support. By a three-to-one margin, potential union members said they would prefer a union “that devotes its resources only to issues facing you and your coworkers in the workplace” to one that devotes resources to “national political issue” as well.

Policy Design Considerations

Prohibiting political spending and activity is straightforward. Two options are available to federal policymakers. They could structure the prohibition as an explicit exclusion—as applies to 501(c)(3) nonprofit organizations, for instance—on political spending of any kind. Alternatively, policymakers could enumerate a specific set of activities that these labor organizations could pursue, limited to the provision of services and benefits that create direct value for workers. Of course, not every organization that provided such services and benefits or served workers would be subject to these rules. But to enjoy the special legal status that would facilitate both public funding and employer partnership, while guaranteeing worker-controlled governance, labor organizations would need to play by the rules.


Workers possess the resources, capabilities, and interests to address their own needs better than anyone else, but they lack the institutional forms to do so. Public policy can help, both by creating the legal infrastructure and by supplying funding. Public support should be conditioned on confining activities to serving members directly.

Americans support this simple bargain—get unions into providing benefits and out of practicing politics—by two-to-one margins across political parties and classes.

The European experience suggests that this approach also holds promise for revitalizing the American labor movement broadly. At the time of its development during the Great Depression, the Ghent system was unpopular with labor unions in most countries. Thus, it was adopted only where organized labor was weakest; countries with stronger labor movements kept their centralized, government-run models. Today, it is countries with the Ghent system that tend to enjoy higher rates of unionization and a stronger labor movement.

Embracing a new model could also help to redefine organized labor in the American imagination as less an instigator of adversarial workplace fights and more an institution that relates to workers as full people with families, aspirations, values, and common interests. It would mark a return to the origins of the labor movement itself: solidarity and mutual aid.

See more from this series

Wells King is the research director at American Compass.