American workers face the threat of AI-based job displacement and the reality of eroded working conditions without a real voice on the job and with an increasingly threadbare safety net. A future presidential administration committed to shared economic prosperity has an opportunity to reimagine the safety net’s role in our society, moving toward what I call a “solidarity net.”

A solidarity net would support building worker power and voice while fostering Americans’ economic security. Here’s what a solidarity net could look like in practice. Right now, a non-union worker whose working conditions have substantially degraded—for example, because their employer rolled out invasive AI-powered surveillance tracking every minute of work—has few options for pushing back. Without a union, a worker who speaks up about these changes risks being ignored at best and disciplined or fired at worst.

Imagine instead if that worker were eligible to receive unemployment benefits for quitting their job in the face of those changes or because their employer fired or disciplined them for organizing with their coworkers. And imagine if, instead of contending with an overburdened state unemployment agency, that worker could apply for benefits directly from a trusted union or worker organization. In the process, the worker would have the chance to connect with labor organizers, learn about their labor rights, discover labor-sponsored training programs, or even find out how to form a union at their new job. A solidarity net would involve an expanded role for unions and other worker organizations in directly administering safety net programs intended to explicitly support worker voice.

With a solidarity net in place, workers could feel more confident speaking out about working conditions, including taking steps to organize with their coworkers to form unions and bargain with employers over technology in the workplace. Strengthening workers’ exit options from bad jobs—including jobs degraded by AI-based automated surveillance and management—could also pressure employers to meet workers’ demands. And by creating more touchpoints between workers and labor organizers through social programs, a solidarity net could seed new organizing opportunities for workers interested in unions.

A solidarity net would represent a shift in how presidential administrations approach social programs. For many decades, federal policymakers have debated social policy in terms of impacts on poverty and hardship, cost, and long-run effects on health, wellbeing, and mobility. These are all worthy considerations, but this framing neglects worker power and organizing. A solidarity net perspective would add to this list the question of how changes in social programs could better support workers’ opportunities for labor organizing and exercising their voice on the job.

So, how should this solidarity net be built? While the details will naturally vary across different social programs, policymakers should think about two key design features: buffers and bridges.

First, social programs should better buffer the risks workers face when they undertake collective action—organizing with coworkers, starting a union, going on strike or protesting, or bargaining with employers. Under our weak system of labor law protections, workers face substantial risks when exercising what should be their legally protected rights to collective action. The fact that employers can act with such impunity in illegally firing or disciplining workers for collective action is a major reason why worker interest in unions far outpaces actual union representation.

More ambitious legislative reform of labor rights could better protect workers by increasing the costs to employers of violating the law. But absent such reform, a future presidential administration can expand the safety net to cushion workers who undertake collective action by reducing the costs of job loss. For example, a future administration could pass regulations requiring states to bolster the generosity and reach of unemployment benefits in ways that could support worker collective action.

Second, the safety net should build stronger bridges between social program applicants and worker organizations. In survey after survey, workers express high levels of interest in collective action but do not know where to start. My own surveys indicate that only about one in ten non-union workers say they would know how to start a union at their job if they wanted to. By building worker organizations into the administration and delivery of social programs, policymakers can accomplish multiple goals simultaneously.

These partnerships can help more people access benefits and services for which they are eligible. A major problem with our safety net is that many individuals and families fail to receive benefits and services to which they’re entitled because of a lack of awareness, language accessibility issues, and daunting paperwork requirements—often dubbed administrative burdens.

Trusted worker organizations rooted in local communities could serve as social program navigators or even administrators, enabling more outreach and application assistance to workers who might be eligible but aren’t currently receiving support. As I previously recounted in Democracy, pilot programs funded by the Department of Labor under the Biden Administration supported unions and other worker organizations in Maine in helping unemployed workers access unemployment insurance. My research found that these partnerships enabled more unemployed workers to access benefits—and to do so faster and with less stress. More generally, research suggests that workforce training programs work best for workers when in partnership with strong worker organizations, especially unions. Strong worker representation can ensure that programs are not captured by business interests at the expense of worker outcomes.

Beyond helping social programs reach more workers, partnerships with worker organizations can also help those groups have more conversations with workers about their labor rights and connect them with organizing opportunities. That is exactly what happened in the Maine pilot. The research I helped carry out suggested that unemployed workers assisted by worker organizations may have developed a stronger interest in workplace collective action and worker voice as a result of their engagement with worker navigators.

Formalizing these partnerships can also provide worker organizations with new revenue streams through government contracts and grants. This financial support can unlock more opportunities for worker organizations to scale their reach while helping government agencies achieve their enrollment goals.

While a solidarity net perspective is new to contemporary social policy, it has deep historical roots in the American welfare state and labor movement. The Congress of Industrial Organizations (CIO), which led ambitious and creative organizing in the New Deal era, saw FDR’s new social programs not just as economic support but as a platform for building worker power.

In 1937, the CIO’s national leadership instructed its affiliate unions to establish “unemployment committees” that would help unemployed workers file for benefits while simultaneously pressuring state capitals and Congress for more generous relief. In some states, especially inventive CIO unions worked with state welfare agencies to gain authority to directly certify and process benefit claims for union members. CIO unions thus served as navigators and even direct administrators of safety net programs.

As one account in the CIO’s national newsletter explained of innovative organizing in Schenectady, New York: “[u]pon being laid off our members were instructed to report immediately to [union] headquarters. Our office for the unemployed was furnished with Unemployment Insurance Application blanks which were made out for our members, and followed through to their proper destination…All applications…are handled through [the union’s] Unemployment Office. A committee from this office meets regularly with the City Welfare Bureau and the Unemployment Insurance Office on relief grievances and delayed insurance checks.” The union served as a trusted, familiar one-stop shop for unemployed workers to process their safety net applications—and as an advocate when bureaucratic burdens or delays arose.

The CIO saw this work not just as helping individual workers claim benefits, but also, as its national leadership put it, as a means to “further the interests of our unions and to advance the CIO cause of organizing the unorganized.” CIO unions used their safety net interactions to keep laid-off workers engaged and to reach out to non-union jobless workers, helping them understand their labor rights and seeding organizing opportunities. The CIO also pushed states to expand UI benefits for striking workers, further subsidizing the costs of collective action and strengthening the CIO’s hand in organizing drives and contract negotiations.

Beyond the CIO’s history, there’s an important international precedent for using the safety net to build worker power. While union membership has declined across rich democracies, countries where unions have a formal role in administering unemployment benefits—so-called Ghent system countries—have seen much smaller declines and still retain membership rates above or close to 50 percent of workers. Research has consistently documented how having unions (or union-affiliated funds) administer unemployment benefits and reemployment support helps keep unemployed workers engaged and encourages more workers to join.

What concrete steps might a future administration take to achieve this vision? As an example, take unemployment insurance. A future administration could encourage states to partner with worker organizations as navigators for, or even administrators of, state UI programs by sharing guidance that such partnerships are permissible under current law (as other legal scholars have argued)—bringing the Ghent model to American states.

A future administration could also encourage states—through guidance and model legislation—to expand eligibility to workers whose working conditions have been substantially eroded, including through AI-based tools, and to ensure that UI benefits are adequate for working families.

And a future administration should assert its statutory authority to define standards for state UI administration through new rulemaking that ensure that benefits reach a majority of unemployed workers–not just the paltry 14 percent of jobless workers covered by the program in recent years.

These steps could be replicated across other social programs, with federal agencies identifying more opportunities for partnering with—and funding—worker organizations to facilitate access to programs, ensuring that benefits are adequate for workers who lose their jobs while taking risky collective action, and creating more on-ramps for social program applicants and beneficiaries into unions and other movement groups.

The stakes for rebuilding worker power are high. A majority of American workers say they should have more influence in their workplace over their working conditions on the job. To take one example, 55 percent of American employees say they lack the voice they want over the adoption of new technologies, like AI. As our economy faces the risks of dislocations and degradation from new technology, worker voice is more important than ever. The safety net is one underappreciated platform for workers to build that voice–and one that a future administration committed to building more inclusive economic growth could pursue.